GL OB AL INTEGRITY

AN INVESTIGATIVE REPORT TRACKING CORRUPTION, OPENNESS AND ACCOUNTABILITY IN 25 COUNTRIES

SOUTH AFRICA
Integrity Scorecard Integrity Assessment Corruption Notebook Corruption Timeline Indicator Scores Country Facts

THE CENTER

FOR

PUBLIC INTEGRITY

INTEGRITY SCORECARD
Overall, South Africa ranks 6th out of 25 countries on the Public Integrity Index, falling into the strong tier. South Africa scores in the very strong tier (90-100) for Category 5, in the strong tier (80-90) for Categories 1, 3 and 6, and in the weak tier (60-70) for Categories 2 and 4. 1. Civil Society, Public Information and Media, Strong, ranking 5th 2. Electoral and Political Processes, Weak, ranking 18th 3. Branches of Government, Strong, ranking 3rd 4. Administration and Civil Service, Weak, ranking 5th 5. Oversight and Regulatory Mechanisms, Very strong, ranking 2nd 6. Anti-Corruption Mechanisms and Rule of Law, Strong, ranking 6th

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

INTEGRITY ASSESSMENT
By Marianne Camerer Civil Society, Public Information and Media
Almost a decade after transition to democratic rule, South Africa has a vibrant and active civil society and progressive and diverse media, unfettered in keeping vigilant watch over those in power. With the right of access to information assured by the Constitution and the Promotion of Access to Information Act (PAIA), the safeguards to secure an open society in which all citizens can enjoy their rights of freedom of expression and association are guaranteed in law, and exercised in practice. There is an active trade union movement. The Congress of South African Trade Unions, with its 2 million affiliate members, is an influential part of the ruling tripartite alliance (the other members are the African National Congress and the South African Communist Party). Civil society organizations, of which there are close to 100,000, may voluntarily register with the government to gain tax-exempt status. In practice, a limited number of CSOs engage actively in public advocacy campaigns. Those that do, such as the CSOs associated with the Treatment Action Campaign, have through intense mass mobilization and court actions been remarkably successful in getting the government to adhere to their various demands, such as the free supply of anti-retroviral drugs for those who are HIV-positive. Only a few CSOs actively engage in monitoring corruption and governance issues. These are not broad-based organizations, but rather focus on applied policy research. Civil society activists working on corruption issues are safe, with none that I am aware of having been threatened, physically harmed, imprisoned or killed. One strength of the governments recent anti-corruption strategy, with the launch of the Moral Summit in 1998 and various other initiatives, has been the call for a partnership approach between government, civil society, business and labor to address corruption as a new struggle that affects everyone. However, the formation of the National Anti-Corruption Forum, in June 2001, has so far failed to deliver concrete inter-sectoral cooperation in this area. The right of access to information, which covers both the public and private sectors, is a powerful tool in the fight against corruption and is increasingly being tested. A recent survey by the Open Democracy Advice Center, a group specializing in monitoring the implementation of the PAIA Act, found, however, poor responsiveness to requests for information by most government departments: 61 percent of public bodies polled did not respond to requests filed under the act, and necessary information policies and processes required by the act have not been implemented. In the event of non-response, application must be made to the High Court or Magistrates Court, which is both time-consuming and beyond the financial means of most citizens. There is a healthy and vibrant media, both print and broadcast. The Independent Communications Authority of South Africa is responsible for regulating broadcasting and telecommunications and issuing broadcast licenses. While there are more than 100 community radio stations, concerns have been raised around the time taken to issue these licenses and how best to promote media diversity. The media report vigorously on corruption (mainly in the public sector). In particular, two weekly publications, the Mail & Guardian and the Sunday Times, have brought key issues to light around a US$6 billion arms deal that involves allegations of fraud and corruption in the procurement process. There is some sensitivity on the side of government, particularly within the African National Congress (ANC), around media reporting on corruption. Whats more, accusations of racism abound, with the media accused of being white-owned, although this is no longer the case. No journalists investigating corruption have been killed in South Africa, although independent journalists report a degree of self-censorship in writing about certain cases involving highranking politicians.

Electoral and Political Processes

Under the Constitution all citizens over 18, if registered on the common voters roll, may vote. The Electoral Act was recently amended to broaden the scope of voters abroad, following an uproar by various opposition parties who argued that voting is a fundamental right of all South African citizens wherever they may live. South African government officials, students and holiday makers living abroad at the time of the 2004 general election will now be allowed to vote, but not those who are working abroad. Elections are held according to a regular timetable, and there is a constitutionally enshrined process whereby an election must be held within 90 days of Parliament being dissolved. The electoral system is characterized by proportional representation at the national level, though at local government this is undertaken through a mixed member arrangement. All citizens are free to form political parties and run for elected office; 27 political parties contested the first democratic election in 1994, and 26 during the 1999 election. If measured by turnout, voter confidence in the electoral system was nearly 90 percent in the 1999 national poll. During election time, equitable free broadcast time is assigned to political parties. In 1999 this was determined by a formula based both on the current seats held by parties and on the number of candidates fielded, according to specific guidelines of the Independent Communications Authority of South Africa. South Africa scores well on gender representation; in the national legislature, 30 percent of MPs are female. The ruling African National Congress and its alliance partners hold more than a two-thirds majority in Parliament.

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

The Independent Electoral Commission (IEC) is a constitutional structure with a key role in managing elections. While it appears to be sufficiently funded and has a number of powers to sanction offenders, it rarely does so, according to civil society observers such as the Electoral Institute of Southern Africa. The nomination process of commissioners is structured by the Constitution to be nonpartisan. A committee (composed of the President of the Constitutional Court, the Public Protector, and individuals from the Human Rights Commission and the Gender Commission) proposes a list of at least eight nominees. A proportionally representative all-party committee then chooses five commissioners who are endorsed by a majority vote in the National Assembly. The most glaring omission in electoral processes and accountability provisions is the lack of rules governing either the receipt or disclosure of private or foreign contributions to political parties. A public and open debate on this issue has been initiated by the Institute for Democracy in South Africa (IDASA), Institute for Security Studies (ISS) and other CSOs. Using PAIA, they have requested information from several political parties about monetary donations and contributions and requested disclosure information from top companies on the Johannesburg Securities Exchange (JSE, formerly the Johannesburg Stock Exchange). The parties failure to respond has resulted in a court case brought by IDASA. Political parties represented in Parliament do receive taxpayer money under the Public Funding of Represented Political Parties Act (Act Number 103 of 1997). This law provides for the establishment of a fund through which political parties receive funds equitably and proportionately in terms of the Constitution. During 2003, 66 million rand is being distributed to political parties by the fund, which is made up of money both appropriated by Parliament and contributed by other sources. Political parties must account for monies received from the fund. A report on the management and administration of the fund is presented by the IEC to Parliament on an annual basis. Since there is no requirement for disclosure of private funds received, there are no audited political

party accounts publicly available. At the moment, it is unclear which private or foreign interests support South Africas political parties.

Branches of Government
There is a constant balance and tension between the executive, legislature and judicial branches in South Africa. The ruling partys majority position in Parliament makes it even more crucial that the legislature play its constitutionally enshrined duty of executive oversight. There is judicial review of the executive, in that any action has to stand up to constitutional scrutiny. There have been a number of challenges in this regard. Neither members of the executive nor the legislature are immune from prosecution while in office, and are treated as ordinary citizens in this regard. Recently, the National Directorate of Public Prosecutions (NDPP) investigated Deputy President Jacob Zuma for allegedly soliciting a half-million Rand bribe a year from a defense company that benefited from the controversial arms deal. The NDPP dropped the charges, although its director said there was prima facie evidence of corruption. The decision not to prosecute has not been well-received. A smear campaign, allegations of being an apartheidera spy, and charges of abuse of power emerged against the head of the NDPP. A commission of inquiry, set up to interrogate these allegations has now reported and as expected the NDPP has been largely exonerated. The cloud over the deputy president, who also heads the Moral Regeneration Movement, is bringing the government unwelcome notoriety. The importance of conflict-of-interest and financial disclosures are increasingly emerging as important anti-corruption safeguards for both the executive and the legislature. The implementation and effectiveness of South Africas impressive ethics regime, such as the Executive Members Ethics Act and the parliamentary code of conduct, are continually being tested with recurrent media allegations of non-disclosure by members of Parliament. There appears to be limited internal capacity to monitor or enforce these regulations. While the Public Protectoras an inde-

pendent constitutional bodycan investigate breaches of the executive code if it receives a complaint, no independent agency monitors the legislature in this regard. The registrar of members interests in Parliament has limited capacity to conduct investigations and can only do so at the behest of the ethics committee. The registrar relies largely on the good faith of members to make disclosures of financial interest. A number of recent media reports of non-disclosure by MPs indicate that more active monitoring may be required by Parliament. The recent discovery that records from the confidential part of the financial disclosure forms had been destroyed, allegedly due to lack of space in a decision apparently never approved by the speaker or the national assembly, points to an apparent failure to grasp the importance of retaining these records as an accountability mechanism. The budget process is fairly transparent and the Standing Committee on Public Accounts (SCOPA) plays an important role in monitoring how public funds have been spent. Its independence as a committee was seriously weakened during the arms deal episode, with the previously nonpartisan committee split along party lines and former Chair Gavin Woods resigning in protest over the failure of SCOPA to conduct a proper investigation into the largest-ever allocation of taxpayer money (60 billion rand). The current committee has also shown no willingness to challenge executive actions, although it does generate some pressure on a number of government institutions to conduct their financial affairs in a proper manner, facilitated by the Public Finance Management Act. South Africas judiciary is fiercely independent and is transforming rapidly. According to the Constitution, appointments for the judiciary

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

are made on merit; however, it is noteworthy that in South Africa merit is not the only criteria, and appointments to key positions need to reflect the demographic reality and address past inequalities. The composition of the Judicial Services Commission, which appoints judges, is another safeguard to ensure that appropriate candidates are appointed to the bench. While access to justice is a guaranteed constitutional right, the reality is that accessing the courts is expensive and timeconsuming. The court rolls are full, while the criminal justice system is perceived by many as ineffective, with certain aspects of the system corrupt. Legal aid does exist, offering legal defense to the poorest in criminal cases. There are no cases reported in the past year where High Court judges have either been harmed or killed as a consequence of adjudicating corruption cases.

Administration and Civil Service

The past 10 years have seen extensive transformation in the civil service and administration. Both the Constitution and Public Service Act provide the framework for transforming the public service to one characterized by professionalism and ethical standards. The difficulties in transforming a society as complex, diverse and unequal as South Africa, however, mean that, in practice, implementation of these new standards of public management has not been uniform, particularly in provinces such as the Eastern Cape, where the legacy of apartheid is more pronounced. There are national conditions of employment in place and regulations requiring an impartial and independent civil service. In practice, however, the model of public administration in South Africa is such that politicians are very involved in setting policy priorities, for instance, and actively participating in the appointment of heads of departments (director-generals). There are regulations preventing nepotism and cronyism in the civil service, where any deviations from highly structured appointment procedures need to be explained. Civil servants found guilty of corruption are not prohibited from future employment in the civil service. Currently, there are also no restrictions on senior

public servants entering the private sector, although proposals to prevent this have been accepted by the cabinet. There are efforts underway to improve information management systems within the administration, which would give managers a further tool to combat corruption. There are conflict-of-interest prohibitions and financial disclosure requirements for senior civil servants, although implementation of these systems has not been uniform and the monitoring of such information is restricted by the limited capacity of the independent Public Service Commission. The Department of Public Service and Administration is initiating a more proactive approach to enforce financial disclosure regulations, with letters from the minister threatening disciplinary action for non-compliance. Citizens have no ready access to this information, and access to these records has yet to be tested under PAIA. Comprehensive whistle-blower legislation exists in South Africa in terms of the Protected Disclosures Act, No. 26 of 2000, whereby employees in both the public and private sectors are legally protected from recrimination or other negative consequences by their employer should they make a disclosure covered by the terms of the act. In practice, it is rarely safe to blow the whistle without suffering negative consequences. There are concerted attempts by CSOs working with the Public Service Commission to establish awareness around the act and assist in designing whistle-blower policies that will provide confidence to civil servants to raise their concerns. Mechanisms for public servants to report corruption internally are currently being formalized; by the terms of a Department of Public Service and Administration directive following an intense audit, all national departments are required to have basic anti-corruption capacity in place by early 2004. A national hotline with a proper case management system will also become operational in 2004. Procurement processes are open, transparent and mostly require competitive bidding and limited sole-sourcing. In terms of the Constitution, contracts for goods or services must be managed in accordance with a system that is fair, equi-

table, transparent, competitive and costeffective. This does not prevent organs of state or institutions from implementing a procurement policy that provides for preferential allocation of contracts to certain categories. Government tenders are available on the Department of Finance Web site. Unless there is a high public interest, the outcomes of contracts are not automatically available. But interested parties have a right to be given reasons for tender decisions, and losing bidders have a right of appeal. In terms of debarment, companies found guilty of violations such as bribery or incompetence are included on a list compiled and circulated by the treasury to government departments. In theory, companies on the list are precluded from receiving government contracts; in practice, however, this list is rarely consulted. The Prevention of Corrupt Activities Bill provides detailed proposals for a register to prohibit corrupt enterprises and their agents from doing business, and its believed that this legal prohibition will make enforcement around this issue easier in the future. The rules of the game on privatizing state assets are publicly available on the Department of Public Enterprises Web site. All groups are allowed to compete for privatized state assets, although broadening the participation of historically disadvantaged groups is a priority.

SOUTH AFRICA

Oversight and Regulatory Mechanisms

When it comes to effective oversight and regulatory mechanisms to ensure sound administration and protection of the public purse, South Africa has a range of institutions. There are fully functioning offices of the Public Protector and Auditor-General, with transparent appointment and removal procedures and guar-

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

anteed independence enshrined in Chapter 9 of the Constitution. In 2001, the report of the Joint Investigation Team into the strategic defense procurement package (the arms deal) tested the publics confidence in both these offices with respect to their operational independence from the executive. (Both offices denied there had been any pressure.) The appointment of the current Public Protector, former ANC MP Leonard Mushwana, raised concerns at the time about whether the office, tasked with investigation of maladministration across government, could withstand executive pressure and political interference with regard to certain investigations involving high-level officials. The South African Revenue Services, which regulates revenue and customs matters, actively pursues tax evaders and has an excellent reputation as a professional body, adopting a zero tolerance attitude towards corruption, including within its own ranks. The independence of the South African Reserve Bank is protected in law by the South African Reserve Bank Act. The current governor of the bank, former minister of labor Tito Mboweni, has done a remarkable job in protecting the independence and autonomy required for this position, which has earned him the confidence of the business and international communities at large. All policy decisions of the bank are available to the public. The JSE Listings Division regulates the reporting and financial records of public companies, as mandated by the JSE Control Act. GAAP practices are followed. General business licensing occurs at the local government level, and licenses are available to all. Licenses are not restricted to domestically owned enterprises, although some concessions may be limited to citizens. Citizen access to information around business licenses has improved vastly over the past five years, and licenses can be obtained within a reasonable time period and at reasonable cost. Small businesses can, however, rarely access financing from commercial banks, and instead rely on development banks created by the state to fill this gap.

Anti-Corruption Mechanisms and Rule of Law

A comprehensive Prevention of Corrupt Activities Bill is currently before Parliament. When it comes into force in early 2004, it will replace the Corruption Act of 1992, which has been largely ineffective in terms of bringing those charged with corruption to book. The new legislation, introduced in April 2002, is comprehensive and farreaching, creating a range of new and tightly defined offenses and penalties covering corrupt activities in both the public and private sectors. These include various offenses: the general offense of corruption; corrupt activities relating to specific persons (including public officers, agents, members of the legislative authority, members of judicial authority, members of the prosecuting authority, witnesses and foreign public officials); corrupt activities relating to specific matters (including procuring and withdrawal of tenders, auctions, contracts, sporting events, gambling and disclosure of information for gratification) and a range of other offenses. Chapter 5 of the bill deals specifically with investigations regarding possession of property, where public officials under suspicion for corruption are required to disclose the source of their unexplained wealth. Under the bill, a general corruption register is envisaged to debar corrupt individuals and corporations from doing business. The bill also creates a duty to report corrupt transactions in both the public and private sectors. Once passed, this law will complement the money-laundering and prevention of organized crime statutes already in place, and assist the range of anti-corruption agencies to do their job more effectively. As mentioned, South Africa has adopted a multi-agency approach to fighting corruption. The Directorate of Special Operations, known as the Scorpions, is the key agency in the fight against corruption and has successfully prosecuted two leading members of the ANC on fraud charges (Tony Yengeni and Winnie Madikizela-Mandela). Falling under the National Directorate of Public Prosecutions, the Scorpions enjoy wide respect among the public. The agency has suffi-

cient funding and powers to effectively address serious cases of corruption, although there is constant tension between its role and that of the police. Abuse of power charges by the deputy president against the head of the NDPP, Bulelani Ngcuka, as well as accusations that he was an apartheid-era spy resulted in the appointment of a commission of inquiry whose recent public findings exonerated Ngcuka. President Mbeki has repeatedly stressed that the law must take its course when it comes to investigations of corruption against high-level political figures. The rule of law is generally upheld in South Africa, although access to the criminal justice system is uneven. There is a right of appeal and citizens are protected from detention without trial. Prison overcrowding is an enormous problem. Those charged with petty offenses, who are unable to afford bail, are often detained for long periods, although there are moves to address this. Individual property rights are upheld by the Constitution, and in practice there is no expropriation of property without appropriate compensation. Legal contracts are generally honored by the state. Law enforcement has been tested by crime in South Africa. The police service is generally thought to be under funded and perceived as the most corrupt government agency in numerous public opinion surveys. There are, however, sound accountability mechanisms in place to deal with police abuse of power, such as the Independent Complaints Directorate, which ensures that all cases of police misconduct, as well as death in police custody (there are still over 500 a year), are thoroughly investigated.

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

CORRUPTION NOTEBOOK CORRUPTION NOTEBOOK

By Justin Arenstein
The battle for South Africas democratic soul took center stage this year, publicly pitting the young nations most powerful public institutions and dynamic leaders against each other. The unprecedented battles, some telecast live to an enthralled public, saw initial victories for those championing clean government, starting with the conviction of colorful parliamentarian Tony Yengeni and struggle icon Winnie MadikizelaMandela on fraud charges, and the African National Congresss (ANC) reprimand of Defense Minister Mosiuoa Lekota for financial impropriety. But these landmark victories were soon clouded by growing ambiguities around just who the good guys really are in a war of attrition that has repeatedly stretched the publics respect for the rule of law. The inquisition started at the top: There were sharp questions about South African President Thabo Mbekis role in a dodgy government-to-government oil deal with Nigerian entrepreneurs, Lekota was grilled for failing to declare business interests, and former Transport Minister Mac Maharaj was hounded off the board of a private bank for having accepted gifts from contractors. Even the integrity of respected watchdog institutions like the Auditor General and Public Protector were questioned following revelations that potentially politically damaging information was edited out of public reports. The most dramatic clash, however, saw anti-corruption czar and public prosecutions director Bulelani Ngcuka name popular Deputy President Jacob Zuma and his financial advisor, Schabir Shaik, as two of the key suspects in South Africas contentious US$4.8 billion Strategic Defense Procurement Package, or arms deal. After months of speculation, Ngcuka, himself a ruling African National Congress stalwart, finally announced that he had prima facie evidence that Zuma solicited a 500,000-rand-a-year protection fee from French armaments company Thales during the audit process. Ngcuka declined, however, to prosecute Zuma, saying he was not confident of a conviction. The announcement unleashed a political firestorm that quickly dwarfed the arms controversy itself, and even more pressing challenges such as South Africas fumbling attempts to tackle HIV/AIDS. Certain elements of the ANC rallied to Zumas defense and counterattacked, questioning the patriotism and legitimacy of both Ngcuka and his elite investigation unit, the Scorpions. What had begun as a cavalry charge against perceived corruption rapidly devolved into a painful dissection of South Africas psyche. South Africa [is] on trial: a trial of conscience, integrity and commitment to the values enshrined in our republics founding document. In the dock [are] the countrys leading institutions, warned leading independent editor Mondli Makhanya, in the Mail & Guardian. The arms deal, South Africas largest ever, was signed in 1999 as a 20-year plan to re-equip the countrys atrophying military machine with state-of-the-art warplanes, warships, and submarines. In return, suppliers were expected to create 65,000 jobs and transfer valuable technology as part of a series of offset or countertrade deals. Within months, the price tag ballooned and watchdogs began questioning the involvement of senior ANC figures in the consortia jostling for contracts. Concerns initially focused on the failure of arms acquisition chief Chippy Shaik to recuse himself during adjudication of a bid by his brother and former ANC banker Schabir Shaik for key tenders to supply combat systems for the navys four new warships. A string of fiercely contested, quasijudicial probes by Parliament, the Auditor General, Public Protector, and Ngcuka publicly confirmed transgressions during the procurement process, but declared the arms deal untainted. As a result of the investigation, just two state officialsincluding Chippy Shaik lost their jobs. Public unease remained, however, essentially because the probes examined contracts covering just 5 percent of the to-

tal arms package. Charging into the gap, feisty local arms contractor Richard Young created legal precedent in 2003 by using the countrys new access to information laws to extract detailed dossiers and early draft reports from the auditor general, proving that investigators findings had been significantly edited in an apparent attempt to shield senior figures. The uproar sparked renewed media questions about Zumas role in the deal. South Africas philosopher-king, Mbeki, led a counterattack, arguing in May 2003 that Some in our country have appointed themselves as fishers of corrupt men. Our governance system is the sea in which they have chosen to exercise their craft so convinced are they of the outcome of their fishing expedition that they regularly describe the defense procurement as the arms deal scandal or debacle. They say a deepening shadow of allegations is threatening to engulf the highest reaches of government. Accusing the media of stoking public perceptions that government is corrupt, Mbeki warned he would not abandon the offensive to defeat the insulting campaigns that seek to further entrench a stereotype that has, for centuries, sought to portray Africans as a people that is corrupt, given to telling lies, prone to theft and self-enrichment by immoral means, a people that is otherwise contemptible in the eyes of the civilized. What our country needs is substance and not shadows, Mbeki concluded, facts instead of allegations, and the eradication of racism. And, almost in answer, facts and substance finally began to emerge. The Londonbased Guardian newspaper reported that Britains largest arms manufacturer, BAE Systems, paid as much as 160 million in secret commissions to secure a 1.5 billion contract to supply Hawk trainer jets and Gripens fighter jets to

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

South Africa. The Guardian also reported charges that an additional 500,000 had been paid to the ANC through outgoing defense minister Joe Modise, whose undeclared interests in a certain defense subcontractor continued making waves even after his death in November 2001. Yengeni also finally admitted that he had accepted a 47 percent discount on a luxury Mercedes Benz from another foreign beneficiary of the arms deal, the European Aeronautical Defense & Space Company. Yengeni, who was ANC chief whip and head of Parliaments Joint Standing Committee for Defense at the time, failed to disclose the benefit. He tried to quash press investigations into the issue, and vigorously proclaimed his innocence, until finally and dramatically caving in to criminal fraud charges. A hurried plea agreement with prosecutors resulted in a four-year prison sentence, with presiding magistrate Bill Moyses berating Yengeni. Parliamentarians are leaders of the nation, he said, and should set an example to their constituents. I regret to say the example you have set as chief whip of the ANC is shocking. A defiant Yengeni responded by appealing the sentence. His gumption may in part be explained by the ANCs own ambivalence about the scandal. Although Yengeni lost his parliamentary seat, he retained his ANC membership and was fined merely US$700 after an internal disciplinary hearing. His was not the only slap on the wrist by a party that appeared either unable or unwilling to purge itself of compromised leaders. Lekota, the ANCs national chairman and Modises successor as defense minister, also received nothing more than a rap on the knuckles for failing to fully disclose his private business interests including directorship of a Free State winery that regularly supplied wine for local provincial government functions. ANC Secretary General Kgalema Motlanthe responded to the growing scandals by insisting an individuals conscience was more important than party doctrine. The ANC only moves in when your own conscience fails to guide you properly, he said. He was, however, more emphatic about Ngcuka, whom he accused of or-

chestrating corruption investigations as if they were Hollywood movies. The criticism appeared to embolden those under investigation, with presidential financial advisor Schabir Shaiks brother, former ANC intelligence operative Mo Shaik, joining Maharaj to brand Ngcuka a turncoat spy. Citing a struggleera ANC intelligence report to President Zuma, the two claimed Ngcuka had become a double-agent for the apartheid security police. The charge stunned the nation. The very future of the Scorpions, widely considered South Africas most effective law enforcement agency, appeared under threat. This is clearly a time for us, as a nation, to examine our collective conscience and ask ourselves what exactly our attitude to malfeasance is, Mondli Makhanya wrote in the Mail & Guardian. Are we on the side of strengthening the hand of those who promote good governance, or do we sanction the undermining of the institutions that have made us the evolving model of democracy we are admired for being? The question soon played itself out on live television, with Mbeki confounding everyone by insisting the allegations be dealt with publicly through a specially appointed commission headed by retired judge Joos Hefer. The resulting soap opera, complete with American-style grandstanding by both lawyers and witnesses, was soon dubbed the Mac & Mo Show. Their carefully constructed case against Ngcuka began crumbling, however, with the first brutal cross-examination. Maharaj was eventually forced to concede he could not prove the charges, and Shaik was forced to apologize to Ngcuka. The media emerged from the battle rather less gloriously. Journalist Ranjeni Munusamy, who leaked the spy allegations to a competing newspaper when her own editors refused to publish, emerged as a duped stooge. The editor who agreed to publish, Vusi Mona of the City Press, was forced to admit he was reckless and had broken his professional code. Their public disgrace came at the peak of a crisis in the media that saw the countrys leading satirist fired for repeated plagiarism, Mona forced to resign for conflict of interests, and the head of South Africas

most influential paper, Sunday Times editor Mathatha Tsedu, fired over irreconcilable differences with management. Perhaps the medias greatest failing, however, was its continued disdain for the nations model Promotion of Access to Information Action (PAIA). Journalists failed to bring a single significant application using PAIA, even as officials made access more difficult. A culture of secrecy is starting to take root, warned political commentator William Gumede. Sadly, access to information from government departments in a democratic South Africa has become a treacherous affair. But, while the media struggled to come to grips with its mandate, civil society embraced the complex new laws. The tiny South African History Archives set the tone with 120 applications delving into everything from South Africas dismantled nuclear weapons program to pioneering requests for Apartheid-era security records. South Africas largest policy think tank, IDASA, also harnessed PAIA to lodge what is perhaps 2003s most significant legal challenge: a High Court bid to force political parties to disclose their funding sources and so expose influence peddling. Transparency would be essential for winning rapidly waning public trust. A global Transparency International survey found this year that most South Africans expect corruption to increase. Of the 47 countries polled, South Africans were the most pessimistic about government graft, with unusually high numbers also mistrusting the police, the medical system, and the education system. But it isnt just government that is rotten. A global corporate crime survey by PricewaterhouseCoopers indicated that South African companies fall victim to more crime than companies almost everywhere else in the world.

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

A full 71 percent, or double the world average, have uncovered serious irregularities in the past two years. The scourge prompted legislators to draft a sweeping new bill, The Prevention and Combating of Corrupt Activities Act, that seeks to hold accountable both the bribe givers and takers. Other initiatives in 2003 included the creation of a national database of corrupt businesses banned from doing business with government, plans for the blacklisting of corrupt civil servants, and moves to strengthen South Africas asset-forfeiture laws for confiscating the proceeds of crime. A new beefedup Executive Members Ethics Act and updated 2003 Ministerial Handbook will also begin forcing politicians and senior officials to disclose all private shares and directorships, while Parliament began drafting similar regulations for judges, and government laid the foundations for a specialized anti-corruption unit for municipalities, which operates with little public oversight. Fighting corruption may even help in taming South Africas other enemypoverty. Poverty appears to be steadily worsening despite widespread praise for South Africas macroeconomic policies. The Institute of Race Relations benchmark annual survey claims the poverty rate rose from 41 percent in 1996 to 49 percent in 2001, with corruption bedeviling attempts to bring development and jobs to the countrys poorest regions. But, while corruption hogged the limelight in 2003, anti-HIV/AIDS activists spearheaded by the Treatment Action Campaign won the years most momentous victory: finally pressuring South Africas dissident government to agree to make antiretroviral drugs available to all citizens. With an estimated 5 million South Africans (out of a population of 44.8 million) infected with HIV/AIDS, the announcement in November literally saw people dancing in the streets. Though no dates have been set for the rollout, the turnabout is a watershed for a government that once branded antiretroviral drugs toxic. We have built something unique on the southern tip of the African continent, says journalist Mondli Makhanya. Out of the ruins of a despotic, corrupt state we

have, in less than a decade, constructed a nation with institutions and values that took other peoples decades and centuries to mold. The trick is to make these institutions and values work, and not let them buckle under any pressure.

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

CORRUPTION CORR UPTION TIMELINE

February 1990 President F. W. de Klerk lifts a decades-long ban on several dozen anti-apartheid organizations, including the African National Congress (ANC), the Pan-African Congress (PAC), and the South African Communist Party, and releases many anti-apartheid activists from prison, including Nelson Mandela. June 1991 The government repeals the last remaining apartheid laws, including the Groups Areas Act, which segregated residential neighborhoods; the Land Act, which restricted blacks rights to purchase land; and the Population Registration Act, which classified the population by race. October 1991 The Goldstone Commission is established as part of the National Peace Accord to investigate instances of political violence, which have sharply increased since 1990. In some instances, police and security force members are deliberately provoking the violence. April 1993 South African Communist Party leader Chris Hani is assassinated outside his home. In October, Polish immigrant Januzs Walus and former Conservative Party MP Clive Derby-Lewis are found guilty of the murder and are sentenced to death, later commuted to life imprisonment. April 1994 The first democratic, nonracial elections in South Africa are held. The once-banned ANC becomes South Africas first democratically elected government, and party leader Nelson Mandela becomes president. April 1994 The 10 ethnically determined homelands of the apartheid era are dissolved and incorporated into nine new provincial administrative regions. February 1995 Allan Boesak is forced to withdraw from an ambassadorial posting while police investigate charges that he misappropriated aid money. Boesak is found guilty of theft and fraud in 1999 and is sentenced to prison. February 1995 The Constitutional Court holds its first sessions. In June, the court declares the death penalty unconstitutional. July 1995 President Mandela signs a law creating the Truth and Reconciliation Commission to investigate human rights violations committed from March 1960 to May 1994. The commission, chaired by Archbishop Desmond Tutu, begins holding public hearings in April 1996, completes most of its fact-finding work by late July 1998, and releases a final report three months later. February 1996 Minister of Welfare Abe Williams, accused of accepting bribes, resigns from the cabinet. In 2000, he is convicted of multiple counts of corruption and theft and serves one year of a threeyear prison sentence. Spring 1996 Health Minister Nkosazana Zuma is accused of misleading Parliament about the awarding of a contract to produce an AIDS-awareness musical, Sarafina 2. An inspector general finds that bidding procedures were violated. March 1996 The South African Human Rights Commission, a watchdog organization that derives its powers from the Constitution, is officially launched. May 1996 Parliament adopts a new Constitution, which goes into effect in February 1997. The Constitution guarantees a broad range of political, civil, economic, and cultural rights to all South Africans and establishes a number of independent bodiesso-called Chapter 9 institutions, including the Public Protector, the Auditor-General and the Human Rights Commissionthat are charged with safeguarding these rights. July 1996 President Mandela fires Bantu Holomisa, deputy minister for environment and tourism, after he accuses Mandela, Deputy President Thabo Mbeki, and other ANC officials of accepting money from Sol Kerzner, a businessman

under investigation for bribery. Mandela admits a month later that the ANC accepted a campaign donation from Kerzner but denies any wrongdoing. September 1996 Eugene de Kock, commander of the secret Vlakplaas unit of the security police during the previous government, is convicted of six murders and 83 other crimes. Before he is sentenced, de Kock testifies that other senior members of the former government, including former presidents P.W. Botha and de Klerk and several army generals and police officials, were involved in dirty tricks against the anti-apartheid movement. March 1997 President Mandela appoints Judge Willem Heath to head the Special Investigating Unit (SIU), one of South Africas most powerful anticorruption investigative bodies. April 1997 Investigators uncover widespread corruption in Mpumalanga regional driving license testing centers, where licenses are being issued in exchange for bribes. Parliamentary deputy speaker Baleka Mbete-Kgositsile received an improperly issued driving license, but police and prosecutors find no proof that she acted with criminal intent. May 1997 A 198 million rand contract to build more than 10,000 houses is awarded to an unknown company run by a close friend of Housing Minister Sankie MthembiMahanyele. Billy Cobbett, the director general of the housing department, is fired after he objects to the scheme. The commission of inquiry looking into the matter finds no proof of outright corruption, but criticizes the Mpumalanga

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

10

governments irresponsible handling of housing funds. May 1997 Five of the seven publicly appointed members of the Independent Broadcasting Authority, which regulates broadcasting in South Africa, resign after the auditor general issues a report detailing widespread financial mismanagement at the authority. June 1997 The Code of Conduct for the Public Service becomes part of the regulations for all public servants. October 1997 An inter-departmental cabinet committee is appointed to consider proposals for a national anti-corruption campaign. The committees proposals are approved by the cabinet in September 1998. December 1997 President Mandela resigns as president of the ANC. Thabo Mbeki is nominated as his successor. October 1998 President Mandela and other participants at a Moral Summit endorse a code of conduct for people in leadership positions. October 1998 The Truth and Reconciliation Commission releases its final report. The 3,500-page report brands the apartheid-era government the primary perpetrator of gross human rights violations, but also holds the ANC and other anti-apartheid activists, including Winnie Madikizela-Mandela and her Mandela United Football Club, accountable for violations. November 1998 Parliament hosts the Public Sector Anti-Corruption Conference, attended by over 200 delegates representing a broad spectrum of national and regional governments and the public service. At the conference, deputy president Thabo Mbeki promises to adopt a zero tolerance approach to corruption. March 1999 The Sunday Times reveals that Home Affairs Director General Albert Mokoena is using his office to run a professional basketball team without approval. Mokoena, who is found guilty of

misconduct after an investigation, resigns in September. April 1999 Parliament hosts the National Anti-Corruption Summit, bringing together more than 300 representatives of government, business, organized religion, NGOs, the media, organized labor, and academiapolitical parties were not invited. Summit resolutions are implemented the following August by the Public Service Commissions Cross Sectoral Task Team on Corruption June 1999 The ANC wins a landslide victory in general elections. Thabo Mbeki, who had replaced Nelson Mandela as the partys leader, becomes president. President Mbeki announces the establishment of a special criminal investigation unit that will tackle high-profile crime, including public corruption. This unit becomes known as the Directorate of Special Investigations, or, more popularly, the Scorpions. August 1999 To help combat whitecollar crime, a specialized commercial crime court is established in Pretoria. A second court is set up in Johannesburg in January 2003. September 1999 The South African government announces the strategic arms procurement packagethe acquisition of aircraft, helicopters, submarines, and ships at a cost of 29 billion rand. PAC MP Patricia de Lille expresses concerns about corruption in the deal, which will develop into one of South Africas biggest corruption scandals, the so-called arms deal. November 1999 Correctional Services Commissioner Khulekani Sitole resigns after Parliaments public accounts committee finds prima facie evidence that he abused his position as chief of South Africas prisons, citing several instances of misappropriation of government funds and conflicts of interest. Subsequent investigations uncover other instances of mismanagement and corruption during Sitoles three-year tenure, including a secret organization that lavished family members and friends with plum jobs and lucrative contracts.

March 2000 A freedom of information law called the Promotion of Access to Information Act (PAIA) goes into effect. The law gives effect to the publics broad right to access information as set out in the 1996 Constitution, which requires private as well as governmental bodies to disclose information. March 2000 Former Mpumalanga Parks Board (MPB) Finance Director Nico Krugel is sentenced to 10 years in prison for his role in a scheme involving the funneling of funds from MPB coffers to the ANC and various politicians. Also implicated in the scandal are MPB chief executive Alan Gray, former MPB accountant Maxi Green, ANC Youth League Secretary James Nkambule, and ANC Youth League organizer Alfred Thumbathi. April 2000 South African cricket team captain Hansie Cronje admits that he and other players engaged in match-fixing, sparking the countrys worst-ever sports scandal. The government appoints a commission headed by retired judge Edwin King. Cronje was banned for life from cricket; he died in a 2002 plane crash. King concluded his investigation unsure if Cronje had been entirely forthcoming. November 2000 After the auditor general finds substantial irregularities in the arms deal, Parliaments Standing Committee on Public Accounts (SCOPA) calls for an investigation to be conducted by a Joint Investigating Team (JIT) consisting of the SIU, the auditor general, the national director of public prosecutions, and the public protector. After weeks of political squabbling, the SIU is excluded from the team. November 2000 Mineral and Energy Affairs Minister

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

11

Phumzile Mlambo-Ngcuka uncovers a secret deal between the Strategic Fuel Fund (SFF), the state-owned oil company, and two private oil trading companies that effectively sold off the countrys oil trading operations without the governments knowledge. Several officials involved admit they accepted bribes. The following month, Mlambo-Ngcuka fires the entire SFF board and repudiates the deal. November 2000 The newly created Moral Regeneration Committee, established by President Mbeki, launches a nationwide crusade to combat crime and corruption by strengthening the moral fiber of the country. February 2001 The Protected Disclosures Act comes into effect, providing legal protection to public- and private-sector whistle blowers who report corruption or illegal activities at their places of employment. March 2001 The South African government signs onto the U.N. Global Programme Against Corruption, which prescribes a country assessment of corruption and anti-corruption activities. March 2001 The Sunday Times reports that Tony Yengeni, ANC chief whip and former chair of the defense committee, received a substantial discount on a Mercedes Benz from an arms deal bidder. In October, he is arrested and resigns from Parliament. In March 2003, he receives a four-year sentence for defrauding Parliament, which he appeals. May 2001 The JIT begins holding public hearings into the arms deal. June 2001 The National Anti-Corruption Forum is held in Langa township outside of Cape Town, bringing together representatives from government, business, labor, and civil society to coordinate national anti-corruption strategies targeting all sectors of society. July 2001 The Star publishes a list of 33 cars supplied at a discount by the European Aeronautic Defence and Space Company to prominent officials including

Tony Yengeni and Siphiwe Nyanda, the head of the defense force. July 2001 The Mail & Guardian reveals that former minister of defense Joe Modises luxury home was being built by Denel, a state-owned arms manufacturer. After he leaves Parliament in June 1999, Modise becomes chairman of Conlog holdings, which had an indirect stake in the arms deal. August 2001 ANC MP and SCOPA member Andrew Feinstein resigns, expressing disappointment with the way the ANC has handled the arms deal investigation. November 2001 The JIT investigative panel publishes its report on the arms deal, clearing members of President Mbekis cabinet of wrongdoing but also criticizing the governments tendering procedures. November 2001 The ruling ANC and the New National Party enter a cooperation agreement to form a loose coalition. December 2001 Judge Thabani Jali leads a commission to investigate wrongdoing in the countrys prison system. The commission is soon inundated with allegations of corruption and maladministration in prisons throughout the country. January 2002 The cabinet adopts the Public Service Anti-Corruption Strategy, which includes a review and consolidation of the legislative framework to fight corruption. February 2002 Dr. Gavin Woods resigns as chairperson of SCOPA in protest over the ANCs obstruction of SCOPAs investigation into the arms deal. March 2002 President Mbeki announces the creation of a U.S.-style presidential press corps, which will have privileged access to the president, including exclusive briefings and the chance to accompany the president on official trips. April 2002 The Scorpions begin investigating former transport minister Mac Maharaj for misconduct in the award-

ing of licensing and toll road contracts to a businessman under criminal investigation. An independent inquiry completed by the FirstRand bank in August 2003 finds no evidence to support media allegations of corruption, but Maharaj resigns from the banks board due to the simmering controversy. July 2002 Andile Nkuhlu, a chief director of the Department of Public Enterprises, is suspended, later resigns, following allegations he received payments from a company that won a state forestry contract. November 2002 Richard Young, director of a computer systems firm that lost out on a defense contract, wins a landmark legal battle when the Pretoria High Court grants his request under PAIA to compel the release of documents relating to the arms deal. November 2002 The Mail and Guardian reveals that the mysterious Mr. X who is being investigated by the Scorpions for accepting a 500,000 rand bribe from a defense firm bidding on a contract to supply ships to the navy is none other than Deputy President Jacob Zuma, who is re-elected vice president of the ANC at the partys national conference the following month. January 2003 President Mbeki sends a specially appointed Interim Management Team (IMT) to the Eastern Cape Province to tackle governmental inefficiency and corruption in the scandalplagued region. The IMT will expose hundreds of cases of corruption, many of which result in criminal convictions. February 2003 The Scorpions arrest former Western Cape Premier Peter Marais and former Deputy Minister David Malatsi on corruption

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

12

charges. They are alleged to have accepted kickbacks from the developer of a golf resort. Their trial begins in November, but adjourns after two weeks to resume in June 2004. April 2003 A Pretoria court sentences Winnie Madikizela-Mandela to five years in jail for a string of theft and fraud offenses. Madikizela-Mandela, who had pleaded not guilty, was convicted of abusing her position as head of the ANCs womens league to defraud a bank and dozens of people. June 2003 The Guardian confirms that British arms manufacturer BAE Systems paid millions in secret commissions to win a contract to supply Hawk jets to South Africa. July 2003 After a preliminary investigation uncovers widespread fraud in magistrate courts, President Mbeki signs a proclamation authorizing a high-level investigation of the financial affairs of senior magistrates, prosecutors, and court clerks. National and regional offices of the department of justice will also be investigated. July 2003 Representatives of 11 South African civil society organizations meet in Cape Town and establish the Civil Society Network Against Corruption, a loose coalition of civil society groups working in the area of anti-corruption. August 2003 Although there is a prima facie case, according to chief prosecutor Bulelani Ngcuka, corruption charges against Deputy President Jacob Zuma are dropped because we are not sure if we have a winnable case. Zuma had been accused of soliciting a US$68,000 bribe from a foreign company seeking a contract to supply the South African military with weapons. November 2003 The Prevention of Corrupt Activities Bill is accepted by Parliament. December 2003 South Africas four biggest political partiesthe ANC, the Democratic Alliance, the Inkatha Freedom Party, and the New National Partypre-

pare to challenge a PAIA claim by watchdog group Institute for Democracy in South Africa (IDASA) compelling the parties to disclose their funding sources, including private-sector donors.

SOUTH AFRICA

GLOBAL INTEGRITY / THE CENTER

FOR

PUBLIC INTEGRITY

South Africa: Civil Society, Public Information and Media

Sub-Category: I-1/Civil Society Organizations Indicators
2 Can citizens organize into trade unions? 3 In practice, do CSOs actively engage in public advocacy campaigns? 4 Are citizens able to form CSOs? 5 Are civil society activists safe when working on corruption issues?

Scores
1.00 1.00 1.00 1.00

1 In law, do citizens have a right to form civil society organizations (CSOs)? 1.00

Indicator and sub-Indicator Details

1 In law, do citizens have a right to form civil society organizations (CSOs)? Score: yes/1.00 Comments: Yes. The right to assemble is guaranteed by the Constitution. South Africa has a very active civil society sector. The Idasa/CORE survey on "The state of civil society in South Africa" found that civil society organizations appear to be well spread throughout South Africa. Most of the CSOs that participated in the study run their programs and projects country-wide with 40 percent in metropolitan areas, 44 percent in small urban areas, 34 percent in informal settlements in urban areas, 27 percent in informal settlements in rural areas and 52 percent in rural areas. Another study found that civil society, in general, was strong with close to 100,000 nonprofit organizations in the country, more than half of which had cooperative relationships with government. References: Non Profit Organisations Act No 71 of 1997, and Non Profit Organisations Amendment Act No 17of 2000. www.sangoco.org.za, www.idasa.org.za; Human Rights Watch World Report 2003: Africa: South Africa 2 Can citizens organize into trade unions? - 2a: In law, citizens have a right to organize into trade unions. Score: yes/1.00 Comments: Yes, the right to assemble and organize is guaranteed by the Constitution. Section 17 of Chapter Two of the Bill of Rights notes everyone has the right, peacefully and unarmed, to assemble, to demonstrate, to picket and to present petitions. References: www.polity.org.za; Constitution of South Africa - 2b: In practice, citizens are able to organize into trade unions. Score: almost always/1.00 Comments: In practice, citizens are able to organize into trade unions. South Africa has a long history of unionization and many firmly entrenched unions. The Congress of South African Trade Unions (COSATU), a federation of trade unions, including the National Union of Mineworkers (NUM) and National Union of Metalworkers of South Africa (NUMSA), claiming membership of two million affiliate members, forms part of the tri-partite governing alliance with the ANC and South African Communist Party. References: www.cosatu.org.za Peer Review Comments: Trade unions may have been active since the beginning of the 20th Century, but black unions were banned by the apartheid state and were only allowed to legally organize from the 1970s. [Comment 2]: Trade union membership has been declining over the last two years as a consequence of unemployment. The third largest union grouping, NACTU (with approximately 350,000 affiliate members) is on the verge of financial collapse, a consequence in part of mismanagement. 3 In practice, do CSOs actively engage in public advocacy campaigns? Score: almost always/1.00 Comments: Yes. CSOs do actively engage in public advocacy campaigns. The Treatment Action Campaign (TAC) is a vivid example of an NGO which engages the government actively on HIV/AIDS issues. Other examples include the Black Sash Trust, which has actively advocated for a Basic Income Grant for all South Africans. However, the number of CSOs that actively engage in public advocacy is rather limited and does

not represent all groups in society. References: Interview with Idasa researcher: www.idasa.org.za Peer Review Comments: Very few CSOs are prepared to engage in advocacy for the purposes of holding government accountable for its performance in the management and use of public resources. Many progressive CSOs are locked into a relationship with government premised on the notion of an equal "partnership" as opposed to a relationship based on the concept of "accountability." Moreover, there is very little distinction on the part of CSOs between which arm of government, the executive or the legislative, they should best engage with and build strategic relationships with. CSOs' understanding of the division of powers and the oversight role of the legislature remain weak, both provincially and nationally. [Comment 2]: It must be noted that it is mostly well-resourced CSOs that are able to participate and advocate for policy. Many of the community based organizations still find it difficult to access institutions such as Parliament and advocate for particular policies. 4 Are citizens able to form CSOs? - 4a: In practice, the government does not create barriers to the organization of new CSOs. Score: almost always/1.00 Comments: In South Africa, citizens are easily able to form CSOs. Should they want to register to obtain tax exemption status, they are assisted by the department of Social Development as well as NGOs versed in the Non Profit Organisations Act No 71 of 1997, and Non Profit Organisations Amendment Act No 17of 2000. The objects of the Act are "to encourage and support nonprofit organizations in their contribution to meeting the diverse needs of the population of the Republic by: a) creating an environment in which nonprofit organizations can flourish; b) establishing an administrative and regulatory framework within which nonprofit organizations can conduct their affairs; c) encouraging nonprofit organizations to maintain adequate standards of governance, transparency and accountability and to improve these standards; d) creating an environment in which the public may have access to information concerning registered nonprofit organizations; and e) promoting a spirit of cooperation and shared responsibility within government, donors and amongst interested persons in their dealings with nonprofit organizations." Regarding registration, the Idasa/Core survey on the state of civil society in South Africa found 8 percent of the organizations reported that they had not registered at all, 38 percent were registered as Section 21 not-for-profit companies and 36 percent registered as non-profit organizations (with the Department of Welfare under the NPO Act of 1997). Thirty-six percent said the registration process was difficult. References: Interview with Idasa researcher: www.idasa.org.za; www.polity.org.za; Non Profit Organisations Act No 71 of 1997; and Non Profit Organisations Amendment Act No 17of 2000. Peer Review Comments: While to my knowledge there are no formal constraints, government has been able to exert some influence through its control of purse strings. One important funding mechanism for NGOs in the past was lotteries. When government licensed a national lottery a few years ago, that became the only legal lottery. To make up for the lost revenue to NGOs, part of the national lottery proceeds goes to a charity fund which a government-appointed body divides between NGOs. There have been complaints about tardiness, and the specter of bias, in the way payments are allocated to applicants. - 4b: In practice, citizens can obtain any necessary license to form a CSO within a reasonable time period. Score: almost always/1.00 Comments: It is relatively painless to obtain the necessary certificate of registration to register a CSO, although most CSOs still employ outside consultants to help them, despite the new NPO Act of 2000 which attempted to make it easier to obtain NPO status. It takes CSOs approximately three to six months to complete the process. References: Interview with Idasa researcher: www.idasa.org.za - 4c: In practice, citizens can obtain any necessary license to form a CSO at a reasonable cost. Score: yes/1.00 Comments: Yes. There is no cost involved other than sending the necessary documentation to the department in Pretoria. References: www.polity.org.za Non Profit Organisations Act No 71 of 1997, and Non Profit Organisations Amendment Act No 17of 2000. - 4d: In practice, in the past year, no existing CSO has been banned by the government for nonviolent advocacy. Score: yes/1.00 Comments: Yes, no CSO has been banned by the government in the past year. The Constitution guarantees freedom of association. Strategic consideration 9 of the National Anti-Corruption Strategy stresses the importance of social analysis, research and policy advocacy, recognizing the role of society and in particular organizations within civil society that have an interest in corruption matters, and that they should

be encouraged to undertake the following: a) Ongoing social analysis and research on the trends and causes of corruption as well as the impact of anti-corruption measures; b) Advocacy of preventative measures and the promotion of a culture of whistle blowing within their constituencies; and c) Ongoing monitoring and evaluation of corruption trends, causes and measures in all sectors. CSOs actively engage in public advocacy campaigns. The Treatment Action Campaign (TAC) is a vivid example of an NGO which engages the government actively on HIV/AIDS issues. Other examples include the Black Sash Trust who has actively advocated for a Basic Income Grant for all South Africans. However, the number of CSOs that actively engage in public advocacy is rather limited and do not represent all groups in society. References: www.polity.org.za; Constitution of South Africa; http://www.dpsa.gov.za/AntiCorruption/Assessmentreport2003.pdf; Interview with Idasa researcher (www.idasa.org.za) Peer Review Comments: Although no organization has been banned, it is an open secret that some organizations have been monitored by the National Intelligence Agency such as the Landless Peoples Movement and possibly the Treatment Action Campaign (TAC). This was particularly the case prior to the World Summit on Sustainable Development (in late 2002). 5 Are civil society activists safe when working on corruption issues? - 5a: In practice, in the past year, no civil society activists working on corruption issues have been imprisoned. Score: yes/1.00 Comments: Yes, in the past year no civil society activists working on corruption issues have been imprisoned. References: The main CSOs in South Africa looking at corruption issues are: The Center for the Study of Violence and Reconciliation (www.csvr.org.za) The Institute for Security Studies (www.iss.org.za), The Institute for Democracy in South Africa (www.idasa.org.za), The Open Democracy Advice Center (www.opendemocracy.org.za), the Public Service Accountability Monitor (PSAM - www.psam.org.za); Transparency South Africa (www.tisa.org.za). See UN/DPSA report on corruption http://www.dpsa.gov.za/AntiCorruption/Assessmentreport2003.pdf. Peer Review Comments: One Bheki Jacobs who runs a freelance information agency, which interacts with CSOs and supports them was imprisoned on what seems to be trumped up charges in an extraordinary arrest, involving many policemen and an unprecedented flight from Cape Town to a Pretoria jail. The case is ongoing. - 5b: In practice, in the past year, no civil society activists working on corruption issues have been physically harmed. Score: yes/1.00 Comments: Yes, no civil society activists working on corruption issues have been physically harmed. References: The Center for the Study of Violence and Reconciliation (www.csvr.org.za) The Institute for Security Studies (www.iss.org.za), The Institute for Democracy in South Africa (www.idasa.org.za), The Open Democracy Advice Center (www.opendemocracy.org.za), the Public Service Accountability Monitor (PSAM - www.psam.org.za); Transparency South Africa (www.tisa.org.za). Peer Review Comments: While those working on corruption issues often experience pressure, such as being called "racists" or facing other attempts at discrediting them, I cannot immediately think of any cases where physical harm has resulted. CSO activists working in fields other than corruption, and especially in fields where there are strong traditional value systems being challenged, such as HIV/Aids, have however been exposed to community/gender violence. [Comment 2]: Very few anti-corruption advocacy CSOs actually follow-up on the details of individual corruption cases. The PSAM, which is an independent monitoring institute based at Rhodes University in South Africa's Eastern Cape province, has followed-up and databased its investigations into a limited number of corruption cases. PSAM staff members have received telephone threats implying that they would no longer be safe during the course of certain of these follow-up investigations. - 5c: In practice, in the past year, no civil society activists working on corruption issues have been killed. Score: yes/1.00 Comments: Yes, no civil society activists working on corruption issues have been killed. A few years ago a civil society activist who disclosed her HIV status, was killed. References: The Center for the Study of Violence and Reconciliation (www.csvr.org.za) The Institute for Security Studies (www.iss.org.za), The Institute for Democracy in South Africa (www.idasa.org.za), The Open Democracy Advice Center (www.opendemocracy.org.za), the Public Service Accountability Monitor (PSAM - www.psam.org.za); Transparency South Africa (www.tisa.org.za). Peer Review Comments: The civil society activist who disclosed her HIV status and was killed was not murdered by a government official; her death was a symptom of continuing and widespread social prejudices in the community in which she lived and worked.

South Africa: Civil Society, Public Information and Media

Sub-Category: I-2/Access to Information Law Indicators
6 In law do citizens have a right of access to information? 7 In practice, is the right of access to information effective?

Scores
1.00

Indicator and sub-Indicator Details

6 In law do citizens have a right of access to information? - 6a: In law, do citizens have a right of access information and basic government records? Score: yes/1.00 Comments: Yes, there is the Promotion of Access to Information Act No 2 of 2000. The need for transparency in government is laid down as a basic value in Chapter 10 of the Constitution that states that transparency must be fostered by providing the public with timely, accessible and accurate information. In terms of S 32 of the Constitution everyone has the right to access any information held by the state, and any information held by another person that is required for the exercise or protection of any rights. There are, however, some exemptions to this law, such as documents that may be in the interest of national security. References: www.polity.org.za -The South African Constitution, Act of 1996; The Promotion of Access to Information Act No 2 of 2000; www.freedominfo.org; Interview with the Open Democracy Advice Center www.opendemocracy.org.za Peer Review Comments: Accessibility of government records in practice remains a problem. [Comment 2]: It is possible for citizens to obtain access to basic government records. There is a schedule of costs which applies to public and private bodies and sets out standard requester fees, as well as costs for various types of copies ranging from A4 paper copies to CD and visual images. Under the Promotion of Access to Information Act, the cost for accessing basic government records (other than a birth certificate, passport, etc., for which there is generally no cost) varies. [Comment 3]: The larger problems than costs for access are a lack of implementation by government bodies, poor record keeping and extensive delays. See recent reports by ODAC and SAHA. - 6b: In law, do citizens have a right of appeal if access to a basic government record is denied? Score: yes/1.00 Comments: Yes, chapter four of the Promotion of Access to Information Act deals with the right of appeal both internal and external (through the courts). In the case of external appeal this would be dealt with by the high court and could cost in the area of R30,000 and depending on the court rolls, could take several years. There is an internal appeal in terms of the Act S75(3) and a fixed appeal fee of R50 and a time period of 30 days within which such an appeal needs to take place. References: The Promotion of Access to Information Act No 2 of 2000; www.freedominfo.org; Interview with the Open Democracy Advice Center - www.opendemocracy.org.za 7 In practice, is the right of access to information effective? Score: Composite value, See commentary Comments: This indicator is a composite of access to information scores in all sectors of government. Based on those measures, this country scored a 53 out of a best possible score of 100. References:

South Africa: Civil Society, Public Information and Media

Sub-Category: I-3/Freedom of the Media Indicators Scores

8 9

In law, is freedom of the media guaranteed? In law, is freedom of speech guaranteed?

1.00 1.00 0.75 0.83

10 Are citizens able to form media entities? 11 Is the media able to report on corruption?

12 Are journalists safe when investigating corruption? 1.00

Indicator and sub-Indicator Details

8 In law, is freedom of the media guaranteed? Score: yes/1.00 Comments: Yes, freedom of expression is guaranteed by Section 16 of the Constitution where 16. (1) Everyone has the right to freedom of expression, which includes freedom of the press and other media; freedom to receive or impart information or ideas; freedom of artistic creativity; and academic freedom and freedom of scientific research. (2) The right in subsection (1) does not extend to propaganda for war; incitement of imminent violence; or advocacy of hatred that is based on race, ethnicity, gender or religion, and that constitutes incitement to cause harm. References: Bill of Rights in the Constitution; Interview with independent journalist; Sparks, A. 2003. "Beyond the miracle: Inside the new South Africa." Jonathan Ball Publishers, Cape Town. 9 In law, is freedom of speech guaranteed? Score: yes/1.00 Comments: Yes, freedom of speech is guaranteed by Section 16 of the Constitution where: (1) Everyone has the right to freedom of expression, which includes freedom of the press and other media; freedom to receive or impart information or ideas; freedom of artistic creativity; and academic freedom and freedom of scientific research. (2) The right in subsection (1) does not extend to: propaganda for war; incitement of imminent violence; or advocacy of hatred that is based on race, ethnicity, gender or religion, and that constitutes incitement to cause harm. References: Bill of Rights in the Constitution; Interview with independent journalist; Sparks, A. 2003. "Beyond the miracle: Inside the new South Africa." Jonathan Ball Publishers, Cape Town. 10 Are citizens able to form media entities? - 10a: In practice, the government does not create barriers to forming a media entity. Score: almost always/1.00 Comments: The type of barriers depends on the type of media entity being formed; community radio stations still experience barriers. References: Interview with ICASA councilor for Global Access SA pilot report 2002. Peer Review Comments: Online publications and newspapers are not regulated, nor do they require licensing, so anyone with funds can start them. However, ICASA does award and regulate broadcast stations, and any citizens wishing to start a radio or TV station is dependent on that award. There is only one free to air broadcaster, and its award was controversial. Similarly, applications for community radio station licenses have been turned down in the past, meaning that some groups have been denied the right to start community radio stations, and after an initial burst of licensing, ICASA has been sluggish at licensing more stations. So not everybody and anybody can start a radio or TV station. - 10b: In law, where a license is necessary, there is an appeal mechanism if a license is denied or revoked. Score: yes/1.00 Comments: Yes, there is an appeal mechanism, as well as using the courts. References: www.polity.org.za - Independent Broadcasting Authority Act 153 of 1993. - 10c: In practice, where necessary, citizens can obtain a media license within a reasonable time period. Score: < 6 months/0.50 Comments: It depends on the type of license. Community broadcasters have launched marches and protest action, voicing their frustration at the slow pace of licensing. Some community radio stations took up to three years to get a license because initially there were 237 applications, which in terms of the act required public hearings. The act has subsequently been amended so that public hearings are not required for community radio stations. For private/commercial radio licenses, this takes six to seven months. ETV's

license process took from July 1997 to March 1998 and longer to the commencement date. References: Communications Portfolio Committee minutes; www.pmg.org.za; Interview with ICASA councilor for Global Access SA pilot report 2002. Peer Review Comments: In practice, setting up a print media entity is not restricted (except by own funding, of course). - 10d: In practice, where necessary, citizens can obtain a media license at a reasonable cost. Score: sometimes/0.50 Comments: The cost depends on the type of license. For example, a community radio application fee costs a R3,000 once off minimal issuing license fee. A private/commercial radio license fee application costs R30,000 non-refundable and the issuing of the license costs R2,500. Any application for amendment costs R30,000 and the annual licensing fee is 1 percent turnover minus agency fees and other deductions (if a nonprofit, there are no costs). ETV (independent television station) license application fee of R300,000, an issuing license fee of R5,000, any application for amendment is R200,000 non-refundable and an annual licensing fee of 2 percent turnover less agency fees and other deductions. References: Interview with ICASA councilor for Global Access SA pilot report 2002. Peer Review Comments: In practice, setting up a print media entity is not restricted (except by own funding, of course). 11 Is the media able to report on corruption? - 11a: In law, it is legal to report accurate news even if it damages the reputation of a public figure? Score: yes/1.00 Comments: Freedom of the media is guaranteed under section 16 of the Constitution, although it does not take primacy over other rights. The Bogoshi test of "reasonable and not negligent" as noted in the Hefer judgment is important: "The publication in the press of false defamatory allegations of fact will not be regarded as unlawful if, upon consideration of all the circumstances of the case, it is found to be reasonable to publish the particular facts in the particular way and at the particular time." References: Interview with independent journalist; Sparks, A. 2003. "Beyond the miracle: Inside the new South Africa." Jonathan Ball Publishers, Cape Town. - 11b: In practice, the government does not encourage self-censorship of corruption-related stories. Score: sometimes/0.50 Comments: There is a certain self-censorship that goes on in this regard as pressure is put on journalists reporting on corruption to see things in a certain way. Elements within the ANC are very critical of the media's role in reporting on corruption, accusing it of being racist. Two publishing houses are black controlled, 13 of the 22 editors of mainstream editors are black, 34.7 percent of journalists in those papers are black and 34.5 percent are women. The Human Rights Commission inquiry into racism in the media in 2000 was prompted by complaints that contended that because the Mail and Guardian newspaper had published 14 reports on black corruption and only four on white corruption over a given period of time it had shown racial bias and given the impression that black people are essentially corrupt and incompetent. There have been a number of attacks on the media throughout its reporting of corruption in the Strategic Defense Procurement Package, a.k.a. "the arms deal." References: Interview with independent journalist; Sparks, A. 2003. "Beyond the miracle: Inside the new South Africa." Jonathan Ball Publishers, Cape Town. Peer Review Comments: Perhaps the most high-profile recent attack has been President Mbeki's allegation last year condemning "fishers of corrupt men" motivated by racist attitudes that black people are naturally corrupt. On the media racism inquiry: It was not only the "Mail & Guardian" against which the original charge was made -- there was a similar and simultaneous charge against the "Sunday Times." Of course the inquiry itself was widened to look at the media in general. [Comment 2]: The self-censorship issue is important but is probably more relevant to corruption in the provincial and national executives. Equally concerning is the fact that few newspapers invest in investigative capacity among journalists. Also the fact that a few large companies own most print titles has meant that many newspapers duplicated the same news, only adding a local spin. Complicated corruption cases are, therefore, not reported. [Comment 2]: The politics of media freedom is a contentious one, especially for the public who are mostly removed from the power play that goes on between the media bosses and the politicians. It begs the question, whether media through its journalists are always neutral in the positions they take when covering alleged corruption stories. [Comment 3]: Some journalists who investigate corruption have been marginalized by the media industry, and find it difficult to get jobs after being black-listed by the ruling party. There are some media organizations which appear to have no real investigative journalists left on their staff, and certainly offer very little support to those who wish to pursue difficult stories that embarrass those in power. The "Mail and Guardian" and the "Sunday Times" are exceptions. Few journalists follow up on their stories. There is appreciation for corruption stories, but many news managers are too nervous, inexperienced and intimidated to pursue them. I would say that in many instances well-timed telephone calls from powerful people can affect the news agenda, both in terms of softening stories and in terms of getting stories on

diary to reflect political agendas. [Comment 4]: The South African Public Broadcaster (SABC) rarely covers corruption stories involving senior government officials with any vigor. For instance, SABC TV news took some three to four days to cover allegations of corruption and conflicts of interest involving deputy president Jacob Zuma. The story was only covered after it had been widely exposed by other media. [Comment 6]: In the past year no libel laws have been invoked for truthfully reporting on corruption. Threats may have been made in this regard. In the past year, no libel laws have been used to close a media entity. Threats may have been made in this regard. In the past year, no libel laws have been used to imprison members of the media. - 11c: In practice, there is no prior restraint on publishing corruption-related stories. Score: almost always/1.00 Comments: This is quite rare. Rather there is some self-censorship as described above. References: Interview with independent journalist. 12 Are journalists safe when investigating corruption? - 12a: In practice, in the past year, no journalists investigating corruption have been imprisoned. Score: yes/1.00 Comments: Yes, in the past year no journalists have been imprisoned for investigating corruption. References: Interview with independent journalist. Peer Review Comments: Mechanisms to dissuade investigative journalists tend to be rather more nuanced in South Africa than physical harm/threats of imprisonment. These revolve, as mentioned, around discrediting actions including the "racism" card. - 12b: In practice, in the past year, no journalists investigating corruption have been physically harmed. Score: yes/1.00 Comments: Yes, in the past year no journalists have been physically harmed for investigating corruption. References: Interview with independent journalist. Peer Review Comments: Although I agree, this does not cover intimidation that many journalists and whistle-blowers face. Such pressure can often be equally effective as physical abuse. - 12c: In practice, in the past year, no journalists investigating corruption have been killed. Score: yes/1.00 Comments: Yes, in the past year no journalists have been killed for investigating corruption. References: Interview with independent journalist.

South Africa: Electoral and Political Processes

Sub-Category: II-1/National Elections Indicators
14 Can all citizens exercise their right to vote freely and fairly? 15 Do citizens participate in the political process?

Scores
1.00 0.80

13 In law, is universal and equal adult suffrage guaranteed to all citizens? 1.00

Indicator and sub-Indicator Details

13 In law, is universal and equal adult suffrage guaranteed to all citizens? Score: yes/1.00 Comments: Yes, section 19 of the Bill of Rights in the Constitution notes: (1) Every citizen is free to make political choices, which includes the right to form a political party; to participate in the activities of, or recruit members for, a political party; and to campaign for a political party or cause. (2) Every citizen has the right to free, fair and regular elections for any legislative body established in terms of the Constitution. (3) Every adult citizen has the right to vote in elections for any legislative body established in terms of the Constitution, and to do so in secret; and to stand for public office and, if elected, to hold office. Therefore, everyone over the age of 18 years old who is a South African citizen (permanent residents are excluded) and if registered on the common voters roll is entitled to vote in an election, even if you are in prison. References: www.polity.org.za - Constitution of South Africa Electoral Institute of Southern Africa: www.eisa.org.za; Global Access pilot report (South Africa) 14 Can all citizens exercise their right to vote freely and fairly? - 14a: In practice, all adult citizens can vote. Score: almost always/1.00 Comments: The Electoral Act was recently amended to broaden the scope of voters abroad, following an uproar by various opposition parties who argued that voting is a fundamental right of all South African citizens wherever they may live. South African government officials, students and holiday makers living abroad at the time of the 2004 general election will now be allowed to vote, but not those who are working abroad. In the 1999 elections, 18,356,851 voters registered out of a total voting population of 22,799,626 (80.5 percent registered). There has been a mixed voter turnout: a high voter turnout of 89.3 percent in 1999, but only 50 percent for the municipal elections. References: Electoral Institute of Southern Africa: www.eisa.org.za Peer Review Comments: The new IEC regulations of January 2004 made provision for South Africans temporarily abroad, including students, to vote at embassies; the procedure is quite cumbersome, though, and will not result in a significant number of overseas voters. - 14b: In practice, ballots are secret or equivalently protected. Score: almost always/1.00 Comments: This is guaranteed in the Constitution Section 19, (3): Every adult citizen has the right to vote in elections for any legislative body established in terms of the Constitution, and to do so in secret; and to stand for public office and, if elected, to hold office, and in practice ballots are secret. References: www.polity.org.za - Constitution of South Africa Electoral Institute of Southern Africa: www.eisa.org.za Peer Review Comments: It must be noted that there are still areas in South Africa where citizens do not vote as a result of fear. In particular Kwazulu Natal remains one province where political tolerance is still an issue. - 14c: In practice, elections are held according to a regular schedule. Score: almost always/1.00 Comments: Elections are held according to a regular schedule - every five years according to the Constitution. The founding democratic election was in 1994. In 1999 President Thabo Mbeki was elected and the next election is in April 2004. When the national assembly is dissolved, or when its term expires, an election must be held within 90 days. References: www.polity.org.za - Constitution of South Africa Electoral Institute of Southern Africa:

www.eisa.org.za Peer Review Comments: Elections for members of national and provincial legislatures take place simultaneously. Local government elections are held separately. 15 Do citizens participate in the political process? - 15a: In practice, all citizens have a right to form political parties. Score: almost always/1.00 Comments: In terms of the Constitution, section 19. (1) Every citizen is free to make political choices, which includes the right to form a political party; to participate in the activities of, or recruit members for, a political party; and to campaign for a political party. In practice, a total of 27 political parties contested the first democratic election in 1994 and 26 during 1999. References: www.polity.org.za -Constitution of South Africa; Global Access pilot report (South Africa) Peer Review Comments: South African electoral law contains quite detailed procedures for the formation of political parties. There are limits, such as registered parties are not allowed to promote racial animosity. - 15b: In practice, all citizens have a right to run for public office. Score: almost always/1.00 Comments: In terms of the Constitution, section 19 (3) every adult citizen has the right to vote in elections for any legislative body established in terms of the Constitution, and to do so in secret; and to stand for public office and, if elected, to hold office. References: www.polity.org.za - Constitution of South Africa Electoral Institute of Southern Africa: www.eisa.org.za Peer Review Comments: Prisoners can not run for political office. This is an important point given the extremely high inmate ratio in South Africa. - 15c: In practice, at least 30 percent of national legislators are women. Score: yes/1.00 Comments: Following the 1999 elections, 119 women were elected to Parliament out of a total of 400 seats (30 percent women representation in Parliament). The ANC has an internal party list quota of 30 percent women. References: Electoral Institute of Southern Africa: www.eisa.org.za Peer Review Comments: There is also a 50/50 campaign underway calling for equal gender representation in the legislature. This has not received any backing by political parties as yet. - 15d: In practice, the ruling political party controls less than 2/3 of seats in the legislature. Score: no/0.00 Comments: With the crossing of the floor legislation in 2003, the ANC, as the ruling party, now controls more than 2/3 of seats in the legislature. References: Sparks, A. 2003. Beyond the miracle: Inside the new South Africa. Jonathan Ball Publishers, Cape Town. - 15e: In practice, during the most recent election, political parties received media coverage roughly proportional to their popular support. Score: almost always/1.00 Comments: Registered political parties received proportional coverage according to regulations set out by the independent broadcasting authority where the IBA Act section(s) ensures equitable treatment of political parties by all broadcasting licensees during any election period. References: www.polity.org.za - Independent Broadcasting Authority Act 153 of 1993. Peer Review Comments: Note that the IBA (now ICASA) regulations apply to broadcast media only. If "media coverage" is taken to include print media, obviously there is no regulation. On average, however, I think print media coverage tends to be balanced during elections. [Comment 2]: Recent reports show that the ANC is getting greater coverage from the national broadcaster than other political parties, for the upcoming elections. It is a point which is being vigorously debated in the media, with many newspapers critical of the SABC.

South Africa: Electoral and Political Processes

Sub-Category: II-2/Election Monitoring Agency

Indicators
17 Is the election monitoring agency effective?

Scores
0.90

16 In law, is there an election monitoring agency? 1.00

Indicator and sub-Indicator Details

16 In law, is there an election monitoring agency? Score: yes/1.00 Comments: Yes, there is an Independent Electoral Commission (IEC) established by the constitution. The IEC reports directly to Parliament. A five-person commission is appointed following nominations from an all-party committee in Parliament. A committee (composed of the President of the Constitutional Court, the Public Protector, and individuals from the Human Rights Commission and the Gender Commission) proposes a list of at least eight nominees. A proportionally representative all party committee then chooses the five commissioners who are endorsed by a majority vote in the National Assembly. References: Section 181 of the Constitution, the Electoral Commission Act no 51 of 1996 and Electoral Act no 73 of 1998 and Municipal Electoral Act no 27 of 2000; Electoral Institute of Southern Africa: www.eisa.org.za Peer Review Comments: Monitoring or evaluating the conduct of elections is only one of the IEC's functions. The commissioners preside over a large bureaucracy, which exists mainly to administer and manage elections. 17 Is the election monitoring agency effective? - 17a: In law, the agency is protected from political interference. Score: yes/1.00 Comments: According to the Constitution, the IEC is one of the state institutions that strengthens constitutional democracy and is protected from political interference. References: See Section 181 of the constitution. Electoral Commission Act - 17b: In practice, agency appointments are made that support the independence of the agency. Score: almost always/1.00 Comments: The nomination process of commissioners is structured by the Constitution to be nonpartisan. A committee (composed of the president of the Constitutional Court, the public protector, and individuals from the Human Rights Commission and the Gender Commission) proposes a list of at least eight nominees. A proportionally representative all party committee then chooses five commissioners, who are endorsed by a majority vote in the National Assembly. References: Interview undertaken for Global Access 2002 pilot report. Peer Review Comments: The IEC appears to work comparatively well at this stage. However, there are unverified reports of dubious appointments within the organization from time to time. [Comment 2]: IEC commissioners are not supposed to be politically prominent at the time of their appointment and so far this principle has been honored. - 17c: In practice, the agency has a professional, full-time staff. Score: almost always/1.00 Comments: Yes, the IEC has a professional, full-time staff. References: IEC Web site: www.elections.org.za - 17d: In practice, the agency makes reports to the legislature following an election cycle. Score: almost always/1.00 Comments: Yes, the IEC reports regularly to Parliament. References: IEC Web site: www.elections.org.za - 17e: In practice, when necessary, the agency imposes penalties on offenders. Score: sometimes/0.50 Comments: The Electoral Court settles election-related disputes. Parties must submit complaints within 48 hours of the announcement of results. However, if objections are raised regarding the results of an election, the IEC can initiate an investigation. The IEC almost never initiates investigations. For example, if objections are raised regarding the results of an election, the IEC can initiate an investigation but does not

have the power to deal with criminal investigations. The CEO may institute civil proceedings before a court, including the Electoral Court, to enforce a provision of the Electoral Act or its Code of Conduct (EA S95). The Electoral Court has final jurisdiction in respect of all electoral disputes and complaints about the infringement of the Code. Its decisions and orders are not subject to appeal or review (EA S96). References: Interview with Electoral Institute of Southern Africa: www.eisa.org.za Peer Review Comments: In the last general election the IEC did occasionally impose penalties on transgressors of its Code of Conduct including the ANC for minor offences. Its preferred approach is to deal with complaints and conflict through its party liaison committees and through mediation. The Electoral Court held no hearings in the 1999 election.

South Africa: Electoral and Political Processes

Sub-Category: II-3/Political Party Finances Indicators
18 Are there regulations governing political party finances?

Scores
0.00

19 Are the regulations governing political party finances effective? 0.17 20 Can citizens access the financial records of political parties? 0.00

Indicator and sub-Indicator Details

18 Are there regulations governing political party finances? - 18a: In law, there are regulations governing private contributions to political parties. Score: no/0.00 Comments: There is currently no regulation of private domestic or foreign contributions to political parties. There is no duty on political parties to disclose the identities of donors of funds they receive directly, nor how they raise funds. References: Electoral Institute of South Africa: www.eisa.org.za Media report: Business Day 13 August 2003: www.bday.co.za; IDASA Position paper: Regulation of private funding to political parties; September 2003. www.idasa.org.za Peer Review Comments: The lack of regulation on private and foreign funding is one of the greater dangers to our democracy, both in terms of the integrity of political outcomes and in terms of levels of corruption. While it is generally agreed and accepted that when individual politicians take money in return for political favors the crime of corruption is committed, there seems to be a large perceptional gray area in situations where parties, rather than individuals, accept financial benefit. [Comment 2]: The Public Funding of Represented Political Parties Act (103 of 1997) provides for the establishment of a fund through which political parties receive funds equitably and proportionately in terms of S 236 of the Constitution. In 2004, R66 million will be distributed this way. The fund is credited by money appropriated by Parliament, contributions and donations from any source and interest. Political parties must account for money allocated from the fund and report on the management and administration of the fund, which is presented to Parliament on an annual basis. - 18b: In law, there are limits on individual donations to candidates and political parties. Score: no/0.00 Comments: There is no law regulating limits on individual donations to candidates or political parties. The Institute for Democracy in South Africa (Idasa) is advocating that parties open their books on donations particularly if they exceed a threshold of R20,000 a year or cumulatively equal more than that amount in a year from a single donor. References: IDASA Position paper: Regulation of private funding to political parties; September 2003. www.idasa.org.za - 18c: In law, there are limits on corporate donations to candidates and political parties. Score: no/0.00 Comments: There is no law regulating limits on corporate donations to candidates or political parties. References: IDASA Position paper: Regulation of private funding to political parties; September 2003. www.idasa.org.za - 18d: In law, there are limits on total party expenditure.

Score: no/0.00 Comments: There is no law regulating limits on total party expenditure. References: IDASA Position paper: Regulation of private funding to political parties; September 2003. www.idasa.org.za 19 Are the regulations governing political party finances effective? - 19a: In practice, when necessary, an agency monitoring political party finances independently initiates investigations. Score: almost never/0.00 Comments: The IEC almost never initiates investigations. For example, if objections are raised regarding the results of an election, the IEC can initiate an investigation but does not have the power to deal with criminal investigations. The CEO may institute civil proceedings before a court, including the Electoral Court, to enforce a provision of the Electoral Act or its Code of Conduct (EA S95). The Electoral Court has final jurisdiction in respect to all electoral disputes and complaints about the infringement of the Code. Its decisions and orders are not subject to appeal or review (EA S96). References: Interview for Global Access Pilot Report - 19b: In practice, when necessary, an agency monitoring political party finances imposes penalties on offenders. Score: rarely/0.25 Comments: The IEC almost never initiates investigations. For example, if objections are raised regarding the results of an election, the IEC can initiate an investigation but does not have the power to deal with criminal investigations. The CEO may institute civil proceedings before a court, including the Electoral Court, to enforce a provision of the Electoral Act or its Code of Conduct (EA S95). The Electoral Court has final jurisdiction in respect to all electoral disputes and complaints about the infringement of the Code. Its decisions and orders are not subject to appeal or review (EA S96). References: Interview for Global Access Pilot Report Peer Review Comments: Two opposition politicians are currently on trial for corruption after a controversial Italian count made donations to the New National Party simultaneously with official objections to a golf estate development of his having been withdrawn. But this prosecution is an exception to the rule. - 19c: In practice, contributions to political parties are audited. Score: rarely/0.25 Comments: Only public contributions are audited. In terms of section 4(1) of the act, the chief electoral officer is responsible for the management and administration of the fund, and is its accounting officer and CEO. In effect, this means that the fund is administered through the IEC, which keeps parties informed of the relevant rules and regulations. There are a number of responsibilities of each political party receiving an allocation from the fund. The party must keep a separate account with a bank in the republic, into which money allocated from the fund must be deposited. It must appoint an official within the party as the accounting officer to take responsibility for the money received in this bank account and ensure that the party complies with the requirements of the act. The accounting officer must keep separate books and records for this money in the manner prescribed. An income and expenditure statement, showing for what purposes the money has been applied, must be audited annually. The auditor is to express an opinion as to whether the allocation has been spent for purposes not authorized by the act. The accounting officer must submit the financial statement and the auditors report to the commission annually. References: Public Funding of Represented Political Parties Act (103 of 1997). IEC Web site: www.elections.org.za Peer Review Comments: Political parties do not currently disclose contributions from private sources. 20 Can citizens access the financial records of political parties? - 20a: In law, citizens can access the financial records of political parties. Score: no/0.00 Comments: In law, citizens can access the IEC reports on public granted funding given to political parties, but in practice this is very difficult. Information on private funding is not available. References: Interview with Idasa researcher. Peer Review Comments: Possibly using the PAIA legislation there is a test case underway in the courts but it rests in part on the premise that political parties are public and not private bodies. Most parties are reluctant to disclose funding sources. [Comment 2]: The IEC reports are published and distributed quite widely to interested parties and many end up in libraries. Citizens cannot, of course, get any kind of access to the financial records of political parties themselves, only to the IECs accounts of how they spend public money.

- 20b: In practice, citizens can access the financial records of political parties within a reasonable time period. Score: > 1 year/0.00 Comments: Idasa has been trying to get records of donations to political parties, and this is proving difficult. References: Interview with Idasa researcher. Peer Review Comments: The ANC has on various occasions been accused of having a foreign policy "for sale" after news leaks of donations from "skunk" states (Taiwan, Suharto's Indonesia, etc.). There have also been strong allegations that the awarding of tenders/licenses, etc. may be influenced by whether companies make donations to the party or even give a secret shareholding to the party. The fact that private funding for parties is usually impenetrable means that it is impossible to correlate whether tenders, policies, or other governmental outcomes are influenced by donations. - 20c: In practice, citizens can access the financial records of political parties at a reasonable cost. Score: almost never/0.00 Comments: Citizens can access the IEC reports on public funding given to political parties for free, but there is no access to private funding. References: Peer Review Comments: At present the cost to see the records of political parties is, in the Idasa case, litigation. This is an expensive process that the majority of citizens can not afford.

South Africa: Branches of Government

Sub-Category: III-1/Executive Indicators
21 In law, can citizens sue the government for infringement of their civil rights? 22 Can members of the executive be held accountable for their actions? 23 Is the executive leadership subject to prosecution?

Scores
1.00 1.00 1.00

24 Are there regulations governing conflicts of interest by the executive branch? 1.00 25 Can citizens access the asset disclosure records of the head of state? 26 In practice, is the ruling party distinct from the state? 0.33 0.50

Indicator and sub-Indicator Details

21 In law, can citizens sue the government for infringement of their civil rights? Score: yes/1.00 Comments: Yes, citizens can sue the government in terms of the law of delict if the government were negligent and basic human rights have been abused. References: Interview with member of the parliamentary justice committee. 22 Can members of the executive be held accountable for their actions? - 22a: In practice, members of the executive give reasons for their policy decisions. Score: almost always/1.00 Comments: Yes, members of the executive need to give reasons for their policy decisions to Parliament which exercises an oversight function on behalf of the people over the executive. See 42 (3): the National Assembly is elected to represent the people and to ensure government by the people under the Constitution. It does this by choosing the president, by providing a national forum for public consideration of issues, by passing legislation and by scrutinizing and overseeing executive action. Where rights have been adversely affected by administrative action, it is a citizen's right to be given written reasons for the decision. These principles apply to administration in every sphere of government and organs of state. See Bill of Rights S33.2. References: Constitution of South Africa; Interview with member of the parliamentary justice committee. Interview with Idasa researcher. Peer Review Comments: While citizens, through Parliament and legal mechanisms, have a right to explanations on policy decisions, there have been some areas (such as HIV/Aids policy) where explanations have been superficial, contradictory and confusing in the extreme, and where arguably Parliament has not fulfilled its oversight role to demand proper explanations. - 22b: In law, the judiciary can review the actions of the executive. Score: yes/1.00 Comments: Yes, in terms of the Constitution, every executive act has to stand up to constitutional scrutiny if challenged. References: Constitution of South Africa; Interview with IDASA (www.idasa.org.za) - 22c: In practice, when necessary, the judiciary reviews the actions of the executive. Score: almost always/1.00 Comments: Yes, this happens in practice. For example, in the Treatment Action Campaign (TAC) case which challenged the minister of health's policy in relation to the distribution of anti-retroviral drugs. This case relied on the Grootboom case which dealt with socio-economic rights in the Constitution promoting access to housing and land. In the TAC case the constitutional court spelled out various time-frames by which its rulings must be complied. References: Constitution of South Africa; Interview with member of the parliamentary justice committee. Interview with Idasa researcher. 23 Is the executive leadership subject to prosecution?

- 23a: In law, the head of state is not immune from prosecution. Score: yes/1.00 Comments: Yes, South African law does not provide any individuals or organizations special grants of immunity from indictment, prosecution, or preventive custody. They are treated as ordinary citizens in front of the law. References: April 2003 DPSA/UN Country Corruption Assessment Report. (www.dpsa.gov.za) Peer Review Comments: South African law does not provide any individuals or organizations special grants of immunity from indictment, prosecution, or preventive custody. However, no head of state has been prosecuted for corruption. The way in which allegations of corruption surrounding the Deputy President Jacob Zuma have been handled, suggests that for a variety of reasons it is unlikely that this would happen. - 23b: In law, ministerial-level officials are not immune from prosecution. Score: yes/1.00 Comments: Yes, South African law does not provide any individuals or organizations special grants of immunity from indictment, prosecution, or preventive custody. They are treated as ordinary citizens in front of the law. References: April 2003 DPSA/UN Country Corruption Assessment Report. (www.dpsa.gov.za) Peer Review Comments: When cases of corruption come to light and there is sufficient evidence, it is likely that in practice a ministerial level official will sometimes be prosecuted. There have been a number of cases in which senior politicians have been prosecuted on corruption related charges, such as Abe Williams, a previous minister of welfare, Winnie Madikizela-Mandela, a previous deputy minister for arts, science and culture, and Tony Yengeni, chief whip of the ANC. There is however, some reluctance to prosecute current Deputy President Jacob Zuma, on corruption charges although the National Director of Public Prosecutions says he has a prima facie case of corruption for allegedly accepting a R500,000 a year bribe relating to a company which benefited from the arms deal. 24 Are there regulations governing conflicts of interest by the executive branch? Score: yes/1.00 Comments: Yes, the Executive Members Ethics Act no. 82 of 1998 was passed to "provide for a code of ethics governing the conduct of members of the cabinet, deputy ministers and members of provincial executive councils; and to provide for matters connected therewith." The act requires that the president must, after consultation with Parliament, by proclamation in the Gazette, publish a code of ethics prescribing standards and rules aimed at promoting open, democratic and accountable government and with which cabinet members, deputy ministers and MECs must comply in performing their official responsibilities. In 2000 after consultation with Parliament, the acting president published the Executive Ethics Code in terms of section 2 (1) of the Executive Members Ethics Act, 1998. References: Executive Members Ethics Act no 82 of 1998. IDASA report on Ethics 2003. 25 Can citizens access the asset disclosure records of the head of state? - 25a: In law, citizens can access the asset disclosure records of the head of state. Score: yes/1.00 Comments: The Executive Ethics Code of 2000 states that the register must have a confidential and a public part. Section 7.5 of the code states that any person has access to the public part of a register during the office hours of the Secretary concerned. References: Executive Ethics Code of 2000; Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za) Peer Review Comments: Disclosure by the executive is often forced by a media scandal. Time and time again, executive members have been caught out with interests that were not declared, which introduces conflicts of interest. Many get away with their conflicts of interest despite the media attention, particularly in provincial administrations. - 25b: In practice, citizens can access the asset disclosure records of the head of state within a reasonable time period. Score: > 1 year/0.00 Comments: Idasa researchers were not able to access the records of the head of state as his, as well as a number of other executive members, had not been filed. The act makes specific timeframe provisions for filing and the Code of Ethics of 2000 stipulates that the first disclosure must be made within 60 days after the promulgation of the code or of a members assumption of office, or a member becoming aware of such interests. Although the code was promulgated on July 28, 2000, the records provided to Idasa researchers show that the first disclosures were only filed in 2002.

References: Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za) - 25c: In practice, citizens can access the asset disclosure records of the head of state at a reasonable cost. Score: almost never/0.00 Comments: Idasa researchers had to travel from Cape Town to Pretoria to the offices of the presidency and were allowed to see, but not make copies of the records. References: Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za) 26 In practice, is the ruling party distinct from the state? Score: sometimes/0.50 Comments: There are frequent accusations by the opposition of state funds being used to finance political party activity. At the national level efforts are being made to spell out this separation, but at provincial and local levels, this is less clear. References: Media reports Peer Review Comments: The party often uses its access to power to benefit its electoral chances, but there have been very few accusations of "crude" or direct party abuse of public funds. [Comment 2]: The ruling party is often not distinct from the state. [Comment 3]: The most important channel through which state funds may be promoting political party activity illicitly is through party controlled business enterprises receiving official contracts after tendering rules have been broken. This is not proven, but there has been extensive reporting of such cases in the last couple of years.

South Africa: Branches of Government

Sub-Category: III-2/Legislature Indicators
27 Can members of the legislature be held accountable for their actions? 28 In law, are members of the legislature subject to prosecution? 29 Are there regulations governing conflict of interest by members of the legislature? 30 Can citizens access the asset disclosure records of members of the legislature? 31 Can citizens access legislative processes and documents? 32 Does the legislature have control of the budget? 33 Can citizens access the national budgetary process?

Scores
1.00 1.00 1.00 1.00 0.92 0.38 0.92

34 In law, is there a separate legislative committee which provides oversight of public funds? 1.00 35 Is the legislative committee overseeing the expenditure of public funds effective? 0.44

Indicator and sub-Indicator Details

27 Can members of the legislature be held accountable for their actions? - 27a: In law, the judiciary can review the actions of the legislature. Score: yes/1.00 Comments: Yes, acts of the legislature are subject to judicial review. References: Interview with MP - 27b: In practice, when necessary, the judiciary reviews the actions of the legislature. Score: almost always/1.00 Comments: All legislation has to pass the constitutionality requirement. There are cases where laws have had to be amended to meet this requirement. References: Interview with MP 28 In law, are members of the legislature subject to prosecution?

Score: yes/1.00 Comments: Yes, South African law does not provide any individuals or organizations special grants of immunity from indictment, prosecution, or preventive custody. They are treated as ordinary citizens before the law References: April 2003 DPSA/UN Country Corruption Assessment Report. (www.dpsa.gov.za) Peer Review Comments: There have been rare prosecutions at the national level. Corruption at the national legislature level is not so much of a problem as it is at provincial level where a number of members of the executive committee (MECs) have engaged in corrupt activities and have not been effectively sanctioned. 29 Are there regulations governing conflict of interest by members of the legislature? - 29a: In law, members of the legislature are required to file an asset disclosure form. Score: yes/1.00 Comments: Whilst not law, there are regulations in terms of a parliamentary code of conduct. The public section of the register of members' interests requires the disclosure of the following items: The number, nature and nominal value of share-holdings; The identity of any extra employment (extra employment has to be approved by the MPs political party); The identity of any directorship or partnership; The identity of remuneration of any consultancy (lobbying is prohibited); The source and description of sponsorship; Any interest in property; Details of foreign travel; Pensions Other benefits; Gifts and hospitalities above the value of R350. The confidential section requires disclosure of the following items: Remuneration of extra employment; Remuneration of directorship or partnerships; The amount of value of any other benefit; The private residence of a member of parliament; The value of any pension. References: Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za); Parliamentary code of conduct 1996 - 29b: In law, there are regulations concerning gifts and hospitality for members of the legislature. Score: yes/1.00 Comments: Yes, in terms of the code, gifts and hospitalities above the value of R350 must be disclosed. References: IDASA report on ethics 2003; Parliamentary code of conduct 1997 30 Can citizens access the asset disclosure records of members of the legislature? - 30a: In law, citizens can access the asset disclosure records of members of the legislature. Score: yes/1.00 Comments: Yes, in terms of the code, the public section of the register is accessible to citizens. The first register of members interest was published in February 1997. It was recently discovered that in 1997 the ethics committee had decided to shred documents from the confidential record one year after a member of Parliament had left office, although this decision was never referred to the national assembly for ratification as it should have. The destruction of the confidential records in June 2000, after Mac Maharaj, the previous transport minister left Parliament in June 1999, made it impossible to later check whether he had in fact declared in the confidential section of his declaration of interests the benefits he allegedly received. References: Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za); Parliamentary code of conduct 1996; Media reports - www.bday.co.za - 30b: In practice, citizens can access these records within a reasonable time period. Score: < 1 month/1.00 Comments: Yes. The public section is now available on an electronic register. References: Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za) Peer Review Comments: In my experience, the electronic register (at least the online version) is not necessarily up to date, and it also does not give a year-by-year breakdown. But the Registrar's office has, in my own experience, always promptly faxed any record instantly, and for free, on demand. Also, while Parliament scores 100 percent on access, the same cannot be said of provincial legislatures. - 30c: In practice, citizens can access these records at a reasonable cost. Score: almost always/1.00 Comments: Initially in order to access the public section of the register members of the public had to pay R5 to see the entry for each Member of Parliament. At this stage there was no consolidated register implying that seeing the whole register would mean photocopying the entire register for R5 per MP. Under pressure from civil society, in particular PIMS of Idasa, the committee reduced the payment to R1 per MP entry. However there has been a significant development since 1997. In recent years, the register has been consolidated and can be accessed electronically. No fee is now required to access the register.

References: Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za) 31 Can citizens access legislative processes and documents? - 31a: In law, citizens can access records of legislative processes and documents. Score: yes/1.00 Comments: The Constitution and the rules of the National Assembly make it very difficult to prevent the public from attending any legislative process and from receiving any document which is made available in the process. There are very few instances where the public could be prevented from such access. Legislation (Access to Information Act) has further strengthened the public's rights in this regard. References: Interview with former chair of the Standing Committee on Public Accounts. - 31b: In practice, citizens can access these records within a reasonable time period. Score: < 3 months/0.75 Comments: Committee records and meetings are available within a week from the Parliamentary Monitoring Group (PMG), which is an independent network of monitors. Official Hansard proceedings usually take slightly longer. There could be cases where the confidential status of a document might be challenged, which in turn could make such an application more protracted, especially if legal actions are instituted. References: Parliamentary Monitoring Group; www.pmg.org.za ; Interview with former chair of the Standing Committee on Public Accounts. Peer Review Comments: PMG minutes are free to nonprofit organizations or citizens subscribing to its email service. [Comment 2]: One or two committees have ruled to go on camera when there seemed to be little reason for them to do so other than political sensitivity. [Comment 3]: No transcripts of the meetings of national Parliament's portfolio committee meetings are available. Idasa's PMG simply provides summaries of these meetings (which may vary in quality and accuracy). By contrast, in some of the provincial legislatures, such as the Eastern Cape, verbatim transcripts of standing committee meetings are available for public inspection within a six-month period. - 31c: In practice, citizens can access these records at a reasonable cost. Score: almost always/1.00 Comments: If such documents are requested through the legislature's library service, there is a nominal cost. If members of the public request such documents through a member of the legislature it is likely to be received without any cost. References: Interview with former chair of the Standing Committee on Public Accounts; www.pmg.org.za 32 Does the legislature have control of the budget? - 32a: In law, the legislature can amend the budget. Score: no/0.00 Comments: No. The budget, which is written into law via the annual "Appropriation Act" is what is known as a "'money bill," which has to either be accepted or rejected in total by Parliament. However, the 1996 Constitution does instruct that the legislature be given powers to amend money bills. These powers have yet to be introduced. In terms of the Annual Appropriation Act, Parliament is supposed to pass legislation which will give it the power to amend the budget. This has not happened, although the budget committee has been looking into this. The point is that retrospective oversight over the budget is not desirable; this issue was one of the reasons that ANC MP Barbara Hogan has threatened to resign. References: Interview with former chair of the Standing Committee on Public Accounts; Interview with Idasa. - 32b: In practice, significant public expenditures require legislative approval. Score: often/0.75 Comments: Significant public expenditures are specified in the budget, which in turn is approved by the legislature. There is scope for the executive to make decisions regarding unbudgeted expenditures (which can be significant amounts) if an unexpected and important enough reason calls for this in the passage of a financial year. In such cases, the executive will ask the legislature to give retrospective approval through an "adjustments or supplementary estimate bill." References: Interview with former chair of the Standing Committee on Public Accounts. Peer Review Comments: Yes, Parliament does approve the budget, but where the executive enters contracts with significant budgetary implications (such as in the arms deal) there is no necessary budget pre-approval by the legislature, and the legislature has no way to retrospectively cancel the contracts, meaning it has no choice but to approve the budget items. It has been claimed by proponents of the arms

deal that Parliament had pre-approved that expenditure by approving the so-called "force design," i.e. the ideal number of ships, aircraft, tanks, etc. But this was only an in-principle approval; no costed version was presented to Parliament for approval. [Comment 2]: It should be pointed out that the bulk of South Africa's public services, and its public expenditure, is administered through its nine provincial administrations. National government accounts for 30 percent of expenditure and local government accounts for 10 percent, while the provinces account for 60 percent of expenditure. Roughly 70 percent of South Africa's 100,000 public servants are employed in the provinces. 33 Can citizens access the national budgetary process? - 33a: In practice, the national budgetary process is conducted in a transparent manner in the debating stage. Score: almost always/1.00 Comments: The budgetary process is conducted in a very transparent manner in the debating stage. In the course of these individual department's budgets being studied in the legislature's committees and being debated in the legislature itself, there is total transparency as members of the public may attend such deliberations and the debates are televised live. References: Interview with former chair of the Standing Committee on Public Accounts. - 33b: In practice, citizens provide input at budget hearings. Score: often/0.75 Comments: As the legislatures committees have become more attentive to the budget's content, they have invited relevant individuals and representative civil society groups to express their opinions on the budget at the committee hearings. References: Interview with former chair of the Standing Committee on Public Accounts. - 33c: In practice, citizens can access itemized budget allocations. Score: almost always/1.00 Comments: The budget has been increasingly well documented over recent years and all such documents which are given to the legislature are "public documents" and are thus available to members of the public. References: Interview with former chair of the Standing Committee on Public Accounts; www.finance.gov.za 34 In law, is there a separate legislative committee which provides oversight of public funds? Score: yes/1.00 Comments: The Standing Committee on Public Accounts (SCOPA). The Constitution instructs the legislature to oversee all actions of the executive government and permits the legislature to set up "mechanisms" through which to do this, but it does not actually establish any of the committees. The rules of Parliament do however establish a committee system through which to process legislation and to exercise oversight of government. The rules specifically establish a Standing Committee on Public Accounts and mandate it to exercise oversight over public funds. References: Interview with former chair of the Standing Committee on Public Accounts. 35 Is the legislative committee overseeing the expenditure of public funds effective? - 35a: In practice, department heads regularly submit reports to this committee. Score: sometimes/0.50 Comments: Reports are only submitted upon request of the committee. References: Interview with former chair of the Standing Committee on Public Accounts. - 35b: In practice, a member of an opposition party presides over this committee. Score: sometimes/0.50 Comments: While not a requirement of law, this has become the established practice. References: Interview with former chair of the Standing Committee on Public Accounts. Peer Review Comments: Since the arms deal debacle, SCOPA has been chaired by a member of the New National Party (NNP), which is a coalition partner of the ANC. Arguably the chair has not been quite as robust in his rulings as his predecessor (whose party, the IFP, was admittedly also a coalition partner of the ANC, but not as close a coalition). [Comment 2]: The undermining of SCOPA over the arms deal was a watershed for Parliament, as it seriously undermined a tradition of accountability that had developed. It is unlikely to regain its strength and independence in the near future. Currently, an MP from the ANC's coalition partner, the NNP, presides over SCOPA, so this person is not an effective member of the opposition. [Comment 3]: Both the past and present chairperson is drawn from a party that is in a coalition

government with the ruling party at national or provincial government. This has not compromised the oversight role of SCOPA, but could prove costly if SCOPA is required to vigorously pursue a large scale corruption case such as the arms deal in future. - 35c: In practice, this committee is protected from political interference. Score: rarely/0.25 Comments: Whereas before 2001 SCOPA had a proud tradition of working consensually as a committee and exercising its oversight role in a nonpartisan fashion, the arms deal history significantly weakened the committee which then voted along party lines with direct political interference from the ruling party, directing particular courses of action for members serving on the committee. Thus, while there is seldom political interference into the work of the committee, on the few occasions it happened in recent times, it did not receive any protection, not even from the overall legislature from whence such protection should be forthcoming. References: Interview with former chair of the Standing Committee on Public Accounts. - 35d: In practice, when necessary, this committee initiates independent investigations into financial irregularities. Score: sometimes/0.50 Comments: Sometimes SCOPA does not have the capacity or resources to investigate matters. Firstly, it is under resourced and is, therefore, unable to cover many areas of public spending. Secondly, it tends to rely only on the reports of the auditor general (the supreme audit institution) for its information. And thirdly, the current committee has shown an unwillingness to challenge actions which the executive does not want challenged. Notwithstanding these limitations, the committee does generate some pressure on a number of government institutions to conduct their financial affairs in a proper manner. References: Interview with Idasa researcher; Interview with former chair of the Standing Committee on Public Accounts.

South Africa: Branches of Government

Sub-Category: III-3/Judiciary Indicators
36 In law, is the independence of the judiciary guaranteed? 37 Is the appointment process for high court judges effective?

Scores
1.00 1.00

38 Can members of the judiciary be held accountable for their actions? 0.33 39 Can citizens access the judicial system? 40 In law, is there a program to protect witnesses in corruption cases? 41 Are judges safe when adjudicating corruption cases? 0.81 1.00 1.00

Indicator and sub-Indicator Details

36 In law, is the independence of the judiciary guaranteed? Score: yes/1.00 Comments: Yes, under section 165 of the Constitution, the courts are independent and subject only to the Constitution and the law, which they must apply impartially and without fear, favor or prejudice. References: Constitution of South Africa 37 Is the appointment process for high court judges effective? - 37a: In practice, there is a transparent procedure for selecting high court judges. Score: almost always/1.00 Comments: Section 174 of the Constitution deals with the appointment of judicial officers and spells out a transparent procedure. References: Constitution of South Africa - 37b: In practice, there are certain professional criteria required for the selection of high court

judges. Score: almost always/1.00 Comments: Section 174 of the Constitution on the appointment of judicial officers states: (1) Any appropriately qualified woman or man who is a fit and proper person may be appointed as a judicial officer. Any person to be appointed to the Constitutional Court must also be a South African citizen; (2) The need for the judiciary to reflect broadly the racial and gender composition of South Africa must be considered when judicial officers are appointed. With respect to the progress made in changing the composition of the judiciary, the following figures were recently cited: as of April 27, 1994, there were no black female judges, three black males, two white females, and 160 white males. The intervention of the Judicial Service Commission had seen the numbers change to 12 black women, 61 black males and 13 white female judges in June 2003. References: Constitution of South Africa; "Mail and Guardian" newspaper (www.mg.co.za) - 37c: In law, there is a confirmation process for high court judges (i.e. conducted by the legislature or an independent body). Score: yes/1.00 Comments: Yes in terms of section 174, (3), the president, as head of the national executive, after consulting the Judicial Service Commission and the leaders of parties represented in the National Assembly, appoints the president and deputy president of the Constitutional Court and, after consulting the Judicial Service Commission, appoints the chief justice and deputy chief justice. In terms of the Constitution, judges are appointed by the president on the advice of the Judicial Service Commission. This commission is established by section 178 of the Constitution. Its functioning is regulated largely by the Judicial Service Commission Act (9 of 1994). The Judicial Service Commission consists of the Chief Justice of South Africa, president of the Supreme Court of Appeal, judge president of the High Courts, designated by the judges' president of the High Courts, the cabinet member responsible for the administration of justice, representatives of the legal profession and members of Parliament, among others. References: Constitution of South Africa - 37d: In law, high court judges are protected from removal without relevant justification. Score: yes/1.00 Comments: Section 177 of the Constitution deals with the removal of judges. (1) A judge may be removed from office only if : a) the Judicial Service Commission finds that the judge suffers from an incapacity, is grossly incompetent or is guilty of gross misconduct; and b) the National Assembly calls for that judge to be removed, by a resolution adopted with a supporting vote of at least two-thirds of its members. (2) The president must remove a judge from office upon adoption of a resolution calling for that judge to be removed. (3) The president, on the advice of the Judicial Service Commission, may suspend a judge who is the subject of a procedure in terms of subsection (1). References: Constitution of South Africa - 37e: In practice, high court judges are protected from political interference. Score: almost always/1.00 Comments: Section 165 of the Constitution deals with judicial authority 165 and states (3) No person or organ of state may interfere with the functioning of the courts, and (4) Organs of state, through legislative and other measures, must assist and protect the courts to ensure the independence, impartiality, dignity, accessibility and effectiveness of the courts. In practice, there is minimal political interference in the judiciary, which is regarded as independent. References: Constitution of South Africa 38 Can members of the judiciary be held accountable for their actions? - 38a: In law, members of the judiciary are obliged to give reasons for their decisions. Score: yes/1.00 Comments: Yes, the Constitution specifies that reasons need to be given in the case of administrative action. References: Interview with Idasa researcher. - 38b: In practice, members of the judiciary give reasons for their decisions. Score: almost always/1.00 Comments: Judges always give reasons for judgments. When they dont, parties to litigation can request them to hand down a written judgment.

References: Interview with Idasa researcher. - 38c: In law, there is an ombudsman (or equivalent agency) for the judicial system. Score: no/0.00 Comments: Not yet, although the Judicial Officers Amendment Bill seeks to create mechanisms to maintain oversight of judicial conduct and ethics. References: www.pmg.org.za; Interview with member of the parliamentary justice committee; Interview with Idasa researcher. Peer Review Comments: The greatest weakness within South Africa judicial system is that there is no mechanism for ensuring critical oversight of judicial conduct and ethics. There is currently a wave within among some academia in South Africa that the Dutch-Roman law is inadequate for South Africa judicial system because the society is rights based. Reference to Professor S. Gutto: Centre for African Renaissance, University of South Africa (UNISA). [Comment 2]: Up until now the Judicial Services Commission has dealt with complaints about judicial misconduct. However, because of its powers ( regarding impeachment) it has no mechanisms for dealing with less serious cases. - 38d: In law, the judicial ombudsman (or equivalent agency) is protected from political interference. Score: no/0.00 Comments: Not yet in place. References: Interview with member of the parliamentary justice committee; Interview with Idasa researcher. - 38e: In practice, when necessary, the judicial ombudsman (or equivalent agency) initiates investigations. Score: almost never/0.00 Comments: Not yet in place. References: Interview with member of the parliamentary justice committee; Interview with Idasa researcher. - 38f: In practice, when necessary, the judicial ombudsman (or equivalent agency) imposes penalties on offenders. Score: almost never/0.00 Comments: Not yet in place. References: Interview with member of the parliamentary justice committee; Interview with Idasa researcher. 39 Can citizens access the judicial system? - 39a: In practice, citizens earning the median yearly income can afford to bring a legal suit. Score: often/0.75 Comments: Right of access to justice is guaranteed in Section 34 of the Bill of Rights: Access to Courts 34. Everyone has the right to have any dispute that can be resolved by the application of law decided in a fair public hearing before a court or, where appropriate, another independent and impartial tribunal or forum. Bringing a legal case depends on a number of factors including the nature of the case, quality of legal practitioners, etc. The small claims court exists for cases under R5,000. It costs about R1,000 to sue for debt and R5,000 for a motor vehicle case, as such the costs are reasonable and the courts are full. References: Interview with legal practitioner. - 39b: In practice, a typical small retail business can afford to bring a legal suit. Score: sometimes/0.50 Comments: It depends on the nature of the case. References: - 39c: In practice, the state provides legal counsel for defendants in criminal cases who cannot afford it. Score: almost always/1.00 Comments: Yes, there is legal aid for the poor (mainly for criminal cases). References: Legal Aid Board; Law clinics - 39d: In practice, all citizens have access to a court of law, regardless of geographic location.

Score: almost always/1.00 Comments: Yes, efforts have been made in this regard. References: Justice Committee Hearings; www.pmg.org.za 40 In law, is there a program to protect witnesses in corruption cases? Score: yes/1.00 Comments: Yes, the Witness Protection Act (112 of 1998) provides for the protection of witnesses through witness protection programs, which are administered by a Central Office for Witness Protection. Offenses in which a witness or related person may be placed under witness protection and which have a bearing on corruption directly or indirectly, include: Any offence relating to dealing in dependence-producing substances if the value thereof is more than R10,000, or if the value thereof is less than R5,000 and the offense is committed by a group of persons or by a syndicate or if the offense was committed by a law enforcement officer; Any offense referred to in section 1 or 1A of the Intimidation Act, 1982; Any offense relating to exchange control, corruption, extortion, fraud, forgery, uttering or theft, involving amounts of more than R50,000, or involving amounts of more than R10,000 if the offense is committed by a group of persons or by a syndicate or if the offense was committed by a law enforcement officer; Any offense referred to in the Prevention of Organized Crime Act, 1998; Any conspiracy, incitement or attempt to commit any offense referred to in the Schedule to the Act; Any other offense which the Minister has determined by regulation; Any other offense in respect of which it is alleged that the offence was committed by a person, group of persons or syndicate in the execution or furtherance of a common purpose or a conspiracy or by a law enforcement officer, and in respect of which the Director of the Witness Protection Program is of the opinion that the safety of a witness who is or may be required to give evidence or who has given evidence in respect of such offence, warrants protection. References: Witness Protection Act 112 of 1998; 41 Are judges safe when adjudicating corruption cases? - 41a: In practice, in the last year, no high court judges have been physically harmed because of adjudicating corruption cases. Score: yes/1.00 Comments: Yes, no high court judges have been physically harmed. References: Interview with Constitutional Court for Global Access 2002. - 41b: In practice, in the last year, no high court judges have been killed because of adjudicating corruption cases. Score: yes/1.00 Comments: Yes, no high court judges have been killed, although a magistrate was killed in a gang/drugrelated cape in the Western Cape. References: Interview with Constitutional Court for Global Access 2002. Interview with Idasa researcher.

South Africa: Administration and Civil Service

Sub-Category: IV-1/Civil Service Regulations Indicators
42 Are there national regulations for the civil service? 43 Is the law governing the administration and civil service effective? 44 In law, are there conflict of interest regulations for senior civil servants?

Scores
0.67 0.47 1.00

45 Can citizens access the asset disclosure records of senior civil servants? 0.00

Indicator and sub-Indicator Details

42 Are there national regulations for the civil service? - 42a: In law, there are regulations requiring an impartial and independent civil service. Score: yes/1.00 Comments: Yes, the Public Service Act confers powers on the Minister for Public Service and Administration to set a framework of norms and standards relating to employment. Section 11: (1) In the making of appointments and the filling of posts in the public service due regard shall be had to equality and other democratic values and principles enshrined in the constitution. (2) In the making of any appointment or the filling of any post in the public service - the evaluation of persons shall be based on training, knowledge and the need to redress the imbalances of the past to achieve a public service broadly representative of the South African people, including representation according to race, gender and disability. Constitution S195 (4) The appointment in public administration of a number of persons on policy considerations is not precluded, but national legislation must regulate these appointments in the public service. S 197 (3) No employee of the public service may be favored or prejudiced only because that person supports a particular political party or cause. References: Public Service Act no. 103 of 1994 and Public Service Regulations of 2003. Constitution Section 195, 197. Interview with senior official at the Department of Public Service and Administration. - 42b: In law, there are regulations to prevent nepotism within the civil service. Score: yes/1.00 Comments: Yes, Part 2D of the Public Service Regulations prevents nepotism. Appointment procedures are structured to reduce undue influence. Any deviation from the established appointment procedures needs to be explained. References: Public Service Regulations. Interview with senior official at the Department of Public Service and Administration. - 42c: In law, there are regulations to prevent cronyism/patronage within the civil service. Score: yes/1.00 Comments: Yes, Part 2D of the Public Service Regulations prevents nepotism. Appointment procedures are structured to reduce undue influence. Any deviation from the established appointment procedures needs to be explained. References: Public Service Regulations. Interview with senior official at the Department of Public Service and Administration. - 42d: In law, civil servants convicted of corruption are prohibited from future government employment. Score: no/0.00 Comments: This is being proposed as part of the Public Sector Anti-Corruption Strategy, but has not been made into law. References: Public Sector Anti-corruption strategy. Interview with senior official at the Department of Public Service and Administration. - 42e: In law, there are restrictions for senior public servants entering positions in the private sector. Score: no/0.00

Comments: This is being proposed as part of the Public Sector Anti-Corruption Strategy, accepted by cabinet in January 2002, but has not been made into law. References: Public Sector Anti-corruption strategy. Interview with senior official at the Department of Public Service and Administration. - 42f: In law, there are regulations and registers concerning gifts and hospitality for senior civil servants? Score: yes/1.00 Comments: Yes, there are Treasury as well as departmental regulations to have gift registers. The Public Service Commission is working on a gift policy. References: Interview with senior official at the Department of Public Service and Administration. 43 Is the law governing the administration and civil service effective? - 43a: In practice, civil servants are protected from political interference. Score: sometimes/0.50 Comments: The model of the South African civil service is that politicians are very involved in the administration of public services with politicians sitting on the appointment panels for senior civil servants and involved in the structuring of departmental policies and objectives. References: Interview with senior official at the Department of Public Service and Administration. Peer Review Comments: The ANC's strategy of "redeployment" within the public service may make some officials vulnerable to political interference. - 43b: In practice, civil servants are appointed according to professional criteria. Score: often/0.75 Comments: Professionalism is one of several criteria. Public administration must be broadly representative of the South African people, with employment and personnel management practices based on ability, objectivity, fairness, and the need to redress the imbalances of the past to achieve broad representation. References: Constitution of South Africa; Public Service Act No 103 of 1994. Interview with senior official at the Department of Public Service and Administration. Peer Review Comments: It is generally accepted that, at the higher levels of the civil service in particular, political acceptance plays as much a role as professional criteria. The South African model is somewhere between the British model, where the civil service is supposed to be apolitical, and the U.S. model, where top positions are overtly politicized. [Comment 2]: This is not necessarily the case in South Africa's provincial administrations. For example in the Eastern Cape, there is the reported case of two officials being appointed to an advisory post in the provincial department of education. One was the ex-provincial speaker's wife, the other was related to the premier's wife. See psam.org.za. - 43c: In practice, civil servants are not employed based on nepotism. Score: sometimes/0.50 Comments: This is difficult to quantify. There are repeated allegations of nepotism in certain departments, such as correctional services, however these are mostly at the provincial rather than national level. A report on nepotism by the public protector found there was no substance to these allegations. References: Media reports. Public Protector Report on nepotism. Peer Review Comments: The issue of nepotism is very difficult to ascertain, but the practice seem to be prevalent within public service. There are allegations that job advertisements are placed in national newspapers specifically designed to fit a particular individual. Anecdotal evidence abounds to attest to this assertion. - 43d: In practice, civil servants are not employed based on cronyism/patronage. Score: often/0.75 Comments: This is difficult to quantify. There are repeated allegations of cronyism/patronage in certain departments, such as Correctional Services, however mostly at the provincial rather than national level. References: Media reports. - 43e: In practice, civil servants have clear job descriptions. Score: sometimes/0.50 Comments: In law and in practice, clear job descriptions are required by the Public Service regulations. A study by the Public Service Accountability Monitor found that at the provincial level, 60 percent of employees interviewed in several departments in the Eastern Cape administration had no clear job

description. For senior managers, there is a much clearer delineation of roles and responsibilities linked to performance. References: Interview with senior official at the Department of Public Service and Administration.; Public Service Accountability Monitor (www.psam.ru.ac.za) - 43f: In practice, in the past year, the government has paid civil servants on time. Score: often/0.75 Comments: There is no problem with paying civil servants on time. There was one instance in the Eastern Cape where teachers in a deep rural area were paid late, but this had more to do with the processing of the salaries rather than insufficient funds. References: Interview with senior official at the Department of Public Service and Administration. Peer Review Comments: There have been ongoing instances of educators not being paid their salaries in a timely manner in the Eastern Cape. The South African Democratic Teachers Union (SADTU) staged a protest march to express their frustration with this state of affairs in Bisho (the provincial capital) in October 2003. In November the Education Department reported that out of a figure of 15,000 unpaid teachers, 5,228 still had to be paid at that point in time. - 43g: In practice, civil servants convicted of corruption are prohibited from future government employment. Score: almost never/0.00 Comments: There is no prohibition from future government employment, although this has been proposed under the Public Service Anti-Corruption Strategy, along with creating a more comprehensive information management system (Personnel Salary System PERSEL) to capture this sort of information from public sector employees. References: Interview with senior official at the Department of Public Service and Administration; - 43h: In practice, restrictions on private sector employment for senior public officials are enforced. Score: almost never/0.00 Comments: There are currently no restrictions, except for the intelligence community. The Joint Investigations Team's report into the arms deal recommended placing such restrictions and these were accepted by cabinet in 2002, but have yet to be regulated in law. References: Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za) 44 In law, are there conflict of interest regulations for senior civil servants? Score: yes/1.00 Comments: Yes, there are Treasury as well as departmental regulations around conflict of interest and financial disclosure. References: Interview with senior official at the Department of Public Service and Administration. 45 Can citizens access the asset disclosure records of senior civil servants? - 45a: In law, citizens can access the asset disclosure records of senior civil servants. Score: no/0.00 Comments: No, these are confidential. References: Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za) Peer Review Comments: The greatest problem in this area is that there is no public access to civil servants' declarations. To the extent that the PSC and individual departments don't do verifications internally (and, as mentioned, that is not consistently done) there is no one else to catch errant civil servants. - 45b: In practice, citizens can access the asset disclosure records of senior civil servants within a reasonable time period. Score: > 1 year/0.00 Comments: No, these are confidential. References: Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za) - 45c: In practice, citizens can access the asset disclosure records of senior civil servants at a reasonable cost. Score: almost never/0.00 Comments: No, these are confidential. References: Government ethics in post-apartheid South Africa, IDASA 2003. (www.idasa.org.za)

South Africa: Administration and Civil Service

Sub-Category: IV-2/Whistle-blowing Measures Indicators
46 Are employees protected from recrimination or other negative consequences when reporting corruption (i.e. whistle-blowing)? 47 Is there an effective internal mechanism (i.e. phone hotline, e-mail address, local office) where civil servants can report corruption?

Scores
0.69 0.31

Indicator and sub-Indicator Details

46 Are employees protected from recrimination or other negative consequences when reporting corruption (i.e. whistle-blowing)? - 46a: In law, civil servants who report corruption are protected from recrimination or other negative consequences. Score: yes/1.00 Comments: Yes, they are protected from occupational detriment in terms of the Protected Disclosures Act, which includes protection from various forms of recrimination and other negative consequences. References: Interview with Open Democracy Advice Center (ODAC) www.opendemocracy.org.za - 46b: In practice, civil servants who report corruption are protected from recrimination or other negative consequences. Score: rarely/0.25 Comments: Bona fide whistle-blowers are protected from Occupational Detriment in terms of the Protected Disclosures Act, which includes various forms of recrimination and other negative consequences, but in practice, despite CCMA and Labour Court rulings, this still occurs. It also depends very much on the organization, what the whistle has been blown on, and on whom the whistle has been blown. The Department of Public Service and Administration is developing policies and mechanisms to assist in the implementation of the whistle-blowing law. References: Interview with Open Democracy Advice Center (ODAC) www.opendemocracy.org.za; Interview with senior official at the Department of Public Service and Administration Peer Review Comments: Whistle-blowing measures need to be strengthened by practice. There is not yet a culture of the protection of whistle-blowers even though the laws are in place. - 46c: In law, private sector employees who report corruption are protected from recrimination or other negative consequences. Score: yes/1.00 Comments: They are offered protection from Occupational Detriment in terms of the Protected Disclosures Act (PDA). Amendments to the Labour Relations Act last year, included under the definition "Automatically Unfair Dismissal," an Occupational Detriment committed in terms of the PDA. This now entitles employees who are dismissed in terms of the PDA, or who resign and make a case for "Constructive Dismissal" in terms of the PDA, entitled to compensation of up to 24 months salary. This might be helpful for some whistle-blowers, but not all, especially those mid-career employees with long service records and others, who might battle to find alternative employment. References: Interview with Open Democracy Advice Center (ODAC) www.opendemocracy.org.za - 46d: In practice, private sector employees who report corruption are protected from recrimination or other negative consequences. Score: sometimes/0.50 Comments: They are offered protection from Occupational Detriment in terms of the PDA. Amendments to the Labour Relations Act last year, included under the definition "Automatically Unfair Dismissal" an Occupational Detriment committed in terms of the PDA. This now entitles employees who are dismissed in terms of the PDA, or who resign and make a case for "Constructive Dismissal" in terms of the PDA, entitled to compensation of up to 24 months salary. This might be helpful for some whistle-blowers, but not all, especially those mid-career employees with long service records and others, who might battle to find alternative employment.

References: Interview with Open Democracy Advice Center (ODAC) www.opendemocracy.org.za 47 Is there an effective internal mechanism (i.e. phone hotline, e-mail address, local office) where civil servants can report corruption? - 47a: In practice, the internal reporting mechanism for public sector corruption has a professional, full-time staff. Score: sometimes/0.50 Comments: The hotlines which are in existence in various departments are relatively effective, although most hotline services are outsourced. The Department of Public Service and Administration is currently looking at the best way to implement a national hotline. Under current consideration is whether to centralize or decentralize the system. A national hotline to report public sector corruption should be in place by April 2004 and will include a sophisticated case management system located in the Public Service Commission, which will direct, monitor and track the outcomes of cases. References: Interview with senior official at the Department of Public Service and Administration - 47b: In practice, the internal reporting mechanism for public sector corruption receives regular funding. Score: sometimes/0.50 Comments: Outsourced hotline services do receive regular funding. References: - 47c: In practice, the internal reporting mechanism for public sector corruption acts on complaints within a reasonable time period. Score: > 1 year/0.00 Comments: There are no actual figures on this, but in the recommendations to the Public Service Commission on the installation of a whistle-bowing infrastructure in the public service, the whistle-blowing policy contained certain guidelines in terms of acting on complaints and responding to them in a timely manner. No fixed time limits were suggested, but it was recommended that matters be dealt with in a reasonable time period. Reasonableness when it comes to such time frames has been well tested in terms of labor legislation and would apply here too. Usually disciplinary cases take a minimum of six months to a year to resolve. References: Interview with Open Democracy Advice Center (ODAC) www.opendemocracy.org.za; Interview with senior official at the Department of Public Service and Administration Peer Review Comments: It is unclear what this "internal mechanism for reporting public sector corruption" currently consists of in South Africa. Particularly within provincial administrations such as the Eastern Cape. A PSAM survey into the experiences and perceptions of corruption by Eastern Cape government officials, undertaken in 2001, found that 64 percent of respondents were unaware of the existence of mechanisms for reporting corruption in their departments. Some figures are available, however. The PSAM's case monitoring project, which tracks corrective actions taken in response to cases of misconduct, maladministration and corruption in the Eastern Cape has found that of 436 cases reported in the period between 1996 and 2004, only 32 have reportedly been met with a satisfactory resolution (defined as the initiation of internal disciplinary action, institutions of civil proceedings for the recovery of public funds, and referral for criminal prosecution, where relevant). This amounts to a reported resolution rate of 7.3 percent for all cases. Specifically with regard to corruption cases, only 13 out of 166 cases have reportedly been resolved. (Figures correct as of Jan. 27, 2003). See www.psam.ru.ac.za and click case monitoring. - 47d: In practice, when necessary, the internal reporting mechanism for public sector corruption initiates investigations. Score: rarely/0.25 Comments: The capacity that currently exists is limited and not uniform, although the minimum requirements for department's internal corruption capacity, which include the ability to initiate investigations on corruption matters, need to be in place by July 1, 2004. References: Interview with senior official at the Department of Public Service and Administration Peer Review Comments: A PSAM survey into the experiences and perceptions of corruption by Eastern Cape government officials, undertaken in 2001, found a lack of confidence in using the government's own internal reporting mechanisms at that point in time. 41 percent of respondents expressed the fear that "syndicates" would attempt to intimidate them if they reported corruption, 30 percent felt that "senior officials" in their department might attempt to intimidate them if they were to report corruption, 23 percent said that intimidation from "syndicates" or "senior officials" would be sufficient to prevent them from making a report of corruption. Reference: Allan, Colm, Mattes, Robert and Millie, Unathi (2002) Government corruption seen from the inside: A survey of public officials perceptions of corruption in the Eastern Cape. PSAM Research Series No 1, Grahamstown: PSAM.

South Africa: Administration and Civil Service

Sub-Category: IV-3/Procurement Indicators
48 Is the public procurement process effective?

Scores
0.72

49 Can citizens access the public procurement process? 0.90

Indicator and sub-Indicator Details

48 Is the public procurement process effective? - 48a: In law, there are conflict of interest regulations for public procurement officials. Score: yes/1.00 Comments: Yes, the existing tender regulations (ST 36/37) cover conflict of interest regulations. The public service procurement system is in the process of revision and the decentralized system following the Australian model, is expected to be in place by April 5, 2004 and to require declaration of financial interests of employees involved in procurement as well as employees responsible for negotiating with service providers/contractors. References: April 2003 DPSA/UN Country Corruption Assessment Report. (www.dpsa.gov.za); Interview with senior official at the Department of Public Service and Administration Peer Review Comments: The Supply Chain Management Framework came into effect on Dec. 5, 2003. This now governs the tender process. See: www.treasury.gov.za - 48b: In practice, the conflict of interest regulations for public procurement officials are enforced. Score: sometimes/0.50 Comments: This varies across government departments. At the provincial level, conflicts of interest are rarely declared. References: Interview with senior official at the Department of Public Service and Administration Peer Review Comments: The claim that the departments of Public Works and Trade and Industry are meticulous in their implementation of conflict of interest regulations cannot be sustained. There are numerous allegations of serious conflicts of interest in both departments. - 48c: In law, there is a mechanism that monitors the assets, incomes and spending habits of public procurement officials. Score: no/0.00 Comments: No, not yet. However, part of the revision of the procurement system calls for an enforced declaration of conflict of interests and adjudication on declared conflicts by a competent authority. References: April 2003 DPSA/UN Country Corruption Assessment Report. (www.dpsa.gov.za) Peer Review Comments: In an interesting recent case the head of the government statistics service was placed under Public Service Commission investigation when it was noticed that his properties seemed to be more valuable than he could afford. But the investigation came as a result of a specific complaint and not a general check to see whether civil servants live above their means. - 48d: In law, all major procurements require competitive bidding. Score: yes/1.00 Comments: Yes, procurement regulations provide for competitive bidding. Section 217 (2) of the Constitution allows organs of state or institutions to implement a procurement policy which provides for categories of preference in the allocation of contracts and the protection or advancement of persons, or categories of persons, disadvantaged by unfair discrimination. References: Interview with senior official at the Department of Public Service and Administration Peer Review Comments: There are some criteria, such as urgency, mechanisms, and roster systems for professional services, which in specific cases allow competitive tendering to be bypassed. These exceptions have on occasion been abused. - 48e: In law, strict formal requirements limit the extent of sole sourcing. Score: yes/1.00 Comments: Yes, as far as possible with respect to the Constitution, which requires a diversity of suppliers.

References: Interview with senior official at the Department of Public Service and Administration - 48f: In law, unsuccessful bidders can instigate an official review of procurement decisions. Score: yes/1.00 Comments: Yes, there are tender regulations and there is a review process through the State Tender Board. References: Interview with senior official at the Department of Public Service and Administration Peer Review Comments: There are many instances when losing bidders challenge tender outcomes. One of the more extraordinary involves a tender for construction on a road in the Eastern Cape. The losing bidder took the final decision to court and the judges reallocated the tender, indicating that the tender board had distorted the scoring. - 48g: In law, unsuccessful bidders can challenge procurement decisions in a court of law. Score: yes/1.00 Comments: Yes, there are tender regulations and there is a review process through the State Tender Board. Losing bidders have a right to review all administrative actions. References: Interview with senior official at the Department of Public Service and Administration - 48h: In law, companies guilty of major violations of procurement regulations (i.e. bribery) are prohibited from participating in procurement bids. Score: yes/1.00 Comments: Debarring of companies is not provided for in law yet, although this is being proposed by the Prevention and Combating of Corrupt Activities Act, which should come into effect in April 2004 where companies and businessmen found guilty of corruption in government tender processes will almost certainly be barred from bidding for government contracts. However, there are treasury regulations that list companies precluded from receiving government contracts, either for nonperformance or corruption. Departments are required to consult this list before contracting companies or individuals. References: Prevention of Corrupt Activities Bill 2003; www.pmg.gov.za; Interview with senior official at the Department of Public Service and Administration - 48i: In practice, companies guilty of major violations of procurement regulations (i.e. bribery) are prohibited from participating in future procurement bids. Score: almost never/0.00 Comments: In practice, no departments ever consult the list from treasury and formalizing this into law as proposed in the Prevention of Corrupt Activities Bill will help. References: Prevention of Corrupt Activities Bill 2003; www.pmg.gov.za; Interview with senior official at the Department of Public Service and Administration Peer Review Comments: We will also have to wait and see if any company is debarred and not only shamed. 49 Can citizens access the public procurement process? - 49a: In law, citizens can access public procurement regulations. Score: yes/1.00 Comments: See the government tender bulletin (www.finance.gov.za). This is guaranteed in the Constitution, Section 217 Procurement: (1) When an organ of state in the national, provincial or local sphere of government, or any other institution identified in national legislation, contracts for goods or services, it must do so in accordance with a system which is fair, equitable, transparent, competitive and cost-effective. (2) Subsection (1) does not prevent the organs of state or institutions referred to in that subsection from implementing a procurement policy providing for: a) categories of preference in the allocation of contracts; and b) the protection or advancement of persons, or categories of persons, disadvantaged by unfair discrimination. References: www.finance.gov.za; Public Finance Management Act regulations regarding Supply Chain Management Framework, gazetted on 5 Dec 2003. State Tender Board Act. - 49b: In practice, citizens can access public procurement regulations within a reasonable time period. Score: < 1 month/1.00 Comments: Yes, citizens can access procurement regulations that are published by the relevant department. References: Public Finance Management Act (www.treasury.gov.za )

- 49c: In practice, citizens can access public procurement regulations at a reasonable cost. Score: almost always/1.00 Comments: There is no cost. References: - 49d: In practice, major public procurements are widely advertised. Score: almost always/1.00 Comments: Yes, as far as possible through the government tender bulletin and media. References: State Tender Board - 49e: In practice, citizens can access the results of major public procurement bids. Score: sometimes/0.50 Comments: They are accessible, but not publicized, although there are moves afoot to publish outcomes on the Web. It is a constitutional right to be given reasons for tender decisions if one has an interest, therefore winning and losing bidders are notified, but only if it is a matter of national interest, such as Telkom (the main telephone service provider in South Africa), is this made public. References: State Tender Board

South Africa: Administration and Civil Service

Sub-Category: IV-4/Privatization Indicators
50 Is the privatization process effective?

Scores
0.92

51 Can citizens access the terms and conditions of privatization bids? 0.92

Indicator and sub-Indicator Details

50 Is the privatization process effective? - 50a: In law, all businesses are eligible to compete for privatized state assets. Score: yes/1.00 Comments: Yes, in law, all businesses are eligible to compete for privatized state assets. References: Interview with senior official at the Department of Public Service and Administration - 50b: In law, there are conflict of interest regulations for government officials involved in privatization. Score: yes/1.00 Comments: The general conflict of interest requirements for public servants apply. References: Interview with senior official at the Department of Public Service and Administration - 50c: In practice, conflict of interest regulations for government officials involved in privatization are enforced. Score: often/0.75 Comments: Public enterprises have internal policies on conflict of interest. Officials involved in a privatization process are specifically required to declare assets and conflicts. For each process a file is kept on each official involved with the relevant declarations. Should a probity question arise the department will just draw the file and check. References: Interview with senior official at the Department of Public Service and Administration Peer Review Comments: There have been few test cases on this, but the one case where a senior official was caught having accepted benefits (a contribution for his wedding during a major forestry privatization tender from a bidding company) led to the termination of his employment with the Department of Public Enterprises. [Comment 2]: The Zama Forestry Deal indicated how bribery had crept into the privatization process. It was only when a key official from the Department of Public Enterprise was exposed by the media as having taken a bribe that the deal collapsed, weeks before finalization. Certainly, government

officials were not monitoring the situation and only appeared to react under the glare of the media. 51 Can citizens access the terms and conditions of privatization bids? - 51a: In law, citizens can access the terms and conditions of privatization bids. Score: yes/1.00 Comments: Some restrictions apply, but no one is excluded in law. The restrictions are, for example, that a pre-qualification briefing process applies and a deposit to obtain the terms and conditions apply. Of course, the Access to Information Act applies, but it has not been tested. References: Interview with senior official at the Department of Public Service and Administration; www.dpe.gov.za - 51b: In practice, citizens can access the terms and conditions of privatization bids within a reasonable time period. Score: < 3 months/0.75 Comments: Some restrictions apply, but no one is excluded in law. The restrictions are, for example, that a pre-qualification briefing process applies and a deposit to obtain the terms and conditions apply. Of course, the Access to Information Act applies, but it has not been tested. References: Interview with senior official at the Department of Public Service and Administration - 51c: In practice, citizens can access the terms and conditions of privatization bids at a reasonable cost. Score: almost always/1.00 Comments: Some restrictions apply, but no one is excluded in law. The restrictions are, for example, that a pre-qualification briefing process applies and a deposit to obtain the terms and conditions apply. Of course, the Access to Information Act applies, but it has not been tested. References: Interview with senior official at the Department of Public Service and Administration

South Africa: Oversight and Regulatory Mechanisms

Sub-Category: V-1/National Ombudsman Indicators
52 In law, is there a national ombudsman, public protector or equivalent agency covering the entire public sector? 53 Is the national ombudsman effective? 54 Can citizens access the reports of the ombudsman?

Scores
1.00 0.69 1.00

Indicator and sub-Indicator Details

52 In law, is there a national ombudsman, public protector or equivalent agency covering the entire public sector? Score: yes/1.00 Comments: Yes, there is a national ombudsman known as the Public Protector. References: Chapter 9 Institution, Section 182 of the Constitution of South Africa, Act 108 of 1996; Public Protector Act. 53 Is the national ombudsman effective? - 53a: In law, the ombudsman is protected from political interference. Score: yes/1.00 Comments: Yes. All Chapter 9 institutions are protected by the Constitution and can operate "without fear or favor" in fulfilling their mandate. References: Section 181 of the Constitution - 53b: In practice, the ombudsman is protected from political interference. Score: sometimes/0.50 Comments: The current public protector, Leonard Mushwana, was previously a member of Parliament for the ANC. There have been concerns that he could have been pressured with respect to certain investigations involving senior members of the ANC. References: Media reports. Peer Review Comments: The public protector has been fairly disappointing in its performance. Appointing a serving MP (i.e. a person who was an MP directly before his appointment) was a move that hardly enhanced the credibility of the office. In this field there is no shortage of expertise among black professionals. [Comment 2]: The problem, I suspect, is not so much that the public protector is "leaned on" during sensitive investigations; rather that he knows what is expected of him or that his allegiance, as an ANC "deployee", lies with the ruling party. The reason for the latter, of course, is that the ANC majority in Parliament can use its muscle to ensure candidates acceptable to it are appointed to head all Chapter 9 institutions. [Comment 3]: The current National Director of Public Prosecutions, Bulelani Ngcuka was ANC Deputy Chair of the NCOP until his appointment. He has proved, on most accounts, exemplary in his independence. I don't think one should single Mushwana out. The argument should rather be that one could try and set a principle in future that the head of certain institutions may not be a card carrying member of any political party. - 53c: In practice, the ombudsman is protected from removal without relevant justification. Score: almost always/1.00 Comments: Yes, under the Constitution the public protector may only be removed from office on the grounds of misconduct, incapacity or incompetence. References: Section 194 of the Constitution - 53d: In practice, the agency has a professional, full-time staff. Score: often/0.75 Comments: Yes, the agency has a professional, full-time staff which has expanded in recent years to enable the agency to fulfill its mandate at provincial level. References: Public Service Commission National Review of Anti-Corruption Agencies, August

2001.www.psc.gov.za Peer Review Comments: In practice the agency is hopelessly understaffed and is incapable of meeting its mandate. - 53e: In practice, agency appointments support the independence of the agency. Score: sometimes/0.50 Comments: Criteria for the appointment of the public protector have recently been amended so that experienced members of parliament (10 years of experience) are now eligible for the appointment. The Democratic Alliance rejected this amendment, arguing "such a person had to be seen to act independently in the eyes of the public." The memo on the objects of the bill notes that "The Public Protector is an institution incorporated in Chapter 9 of the constitution, dealing with state institutions supporting constitutional democracy. As such the public protector is independent and accountable to Parliament." The amendments also propose one deputy public protector be appointed by the president with the involvement of parliament. References: Public Protector Amendment Act 2003; www.pmg.org.za Peer Review Comments: Both incumbents to date (Baqwa and Mushwana) have given the impression of being relatively independent-minded generally, but both seem(ed) prone to using contorted logic to arrive at executive-minded decisions in high-profile, sensitive cases. [Comment 2]: Neither the current nor the past incumbent have successfully supported the independence of the agency in the face of executive pressure. - 53f: In practice, the agency receives regular funding. Score: almost always/1.00 Comments: Yes, the agency receives regular funding, allocated by the justice department and approved by a vote in the legislature. This funding is not always seen as sufficient. References: Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za - 53g: In practice, the agency makes regular reports to the legislature. Score: often/0.75 Comments: Yes, the public protector reports on an annual basis to parliament and issues reports if requested to investigate. The most recent reports include investigations into the conflict of interests and financial disclosure or the minister of defense, and the deputy president. References: Media reports; Public Protector Web site; http://www.polity.org.za/html/govt/pubprot/ Peer Review Comments: It depends on the case. If the case involves a member of the executive the agency may prevaricate, delaying its investigation as well as its reports to parliament. The case of its investigation into Jacob Zuma, the deputy president , is a case in point. - 53h: In practice, the government acts on the findings of the agency. Score: rarely/0.25 Comments: The public protector can investigate and then make recommendations to Parliament. In many cases these are ignored by the ruling party, which holds a majority in Parliament. References: Interview with Public Protector (Global Access pilot) - 53i: In practice, the agency acts on citizen complaints within a reasonable time period. Score: < 6 months/0.50 Comments: It depends on the nature of the report. It varies from a few months to a few years. References: Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za 54 Can citizens access the reports of the ombudsman? - 54a: In law, citizens can access reports of the ombudsman. Score: yes/1.00 Comments: Yes, these reports are publicly available. References: http://www.polity.org.za/html/govt/pubprot/ - 54b: In practice, citizens can access the reports of the ombudsman within a reasonable time period. Score: < 1 month/1.00

Comments: Yes, they are available as soon as they are published in a hard copy. Internet accessibility takes longer. References: Peer Review Comments: They're not necessarily available as soon as published, since some reports go to the president first, who then has to comment on them and submit them to Parliament, from where they are made public. The president has a set time within which to submit to Parliament. The reports that go via the president are ones dealing with complaints under the Executive Members' Ethics Act. - 54c: In practice, citizens can access the reports of the ombudsman at a reasonable cost. Score: almost always/1.00 Comments: There is no cost involved. References:

South Africa: Oversight and Regulatory Mechanisms

Sub-Category: V-2/Supreme Audit Institution Indicators
55 In law, is there a national supreme audit institution, auditor general or equivalent agency covering the entire public sector? 56 Is the supreme audit institution effective? 57 Can citizens access reports of the supreme audit institution?

Scores
1.00 0.89 1.00

Indicator and sub-Indicator Details

55 In law, is there a national supreme audit institution, auditor general or equivalent agency covering the entire public sector? Score: yes/1.00 Comments: Yes, there is an Auditor General, an agency with a constitutional mandate to audit and report on the accounts, financial statements and financial management of all public sector agencies. References: Section 188 of the Constitution Act no 108 of 1996; The Auditor-General Act (1995) and the Public Finance Management Act (1999).www.polity.gov.za; www.agsa.co.za Peer Review Comments: For the most part the Office of the Auditor-General (A-G) is effective. However, there have been some exceptions to the effectiveness of this agency at the local government level. 56 Is the supreme audit institution effective? - 56a: In law, the supreme audit institution is protected from political interference. Score: yes/1.00 Comments: Yes, the independence of the auditor general is guaranteed by the Constitution. References: Constitution Act no 108 of 1996. Peer Review Comments: The auditor general can only audit the financial statements of government departments if these departments allocate budget funds to pay for these services. In the Eastern Cape, in the past, a number of local government councils have not been audited due to non-payment of audit fees. The auditor general has a national office as well as nine provincial offices. There have been some questions raised about the national auditor generals stance on the exclusion of the Heath Special Investigating Unit from the Arms procurement deal investigation. He backed the executives decision to exclude Heath Special Investigating Unit. The role he played in directly managing the auditor general's investigation into the primary contracts as part of the arms deal investigation has also been called into question by the ex-chairperson of SCOPA, Gavin Woods. On the other hand, within certain provinces, the work of the Auditor General's Office is of an exceptionally high standard in the face of extreme political pressure from the provincial executive. - 56b: In practice, the head of the agency is protected from removal without relevant justification. Score: almost always/1.00 Comments: The auditor general, as a chapter nine institution of the Constitution, may only be removed from office on the grounds of misconduct, incapacity or incompetence. This has never happened.

References: Section 194, Constitution Act no 108 of 1996 - 56c: In practice, the agency has a professional, full-time staff. Score: almost always/1.00 Comments: Yes, the agency has a professional, full-time staff. References: Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za - 56d: In practice, agency appointments support the independence of the agency. Score: often/0.75 Comments: Yes, although the independence of the current auditor general was questioned by opposition parties and the independent media in terms of how his office dealt with the investigation into the arms deal. References: Peer Review Comments: While all the comments on the auditor general are true, it must be said that his office was seriously compromised by its actions on the arms deal. The two most qualified auditors were mysteriously taken off the investigation and it degenerated into a murky and compromised squabble between political parties, the media and the office of the auditor general, with some going so far as to suggest that the auditor general had misled Parliament. It is unlikely that the office will be able to recover from that under the present incumbent. I would say, as a commentator, that he bowed to political pressure by showing the executive drafts of his report on the arms deal investigation prior to showing them to SCOPA. - 56e: In practice, the agency receives regular funding. Score: almost always/1.00 Comments: The agency receives regular funding from the audits it conducts. References: Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za - 56f: In practice, the agency makes regular reports to the legislature. Score: almost always/1.00 Comments: Yes, the auditor general reports regularly to the legislature through the Standing Committee on Public Accounts. References: Peer Review Comments: Many departments are very late in submitting reports to the auditor general and, therefore, the auditor general has a backlog of departments that need to be audited. This is impinging on the office's ability to ensure that Parliament is provided with an up-to-date fully audited report. - 56g: In practice, the government acts on the findings of the agency. Score: sometimes/0.50 Comments: Under the Public Finance Management there are more incentives for government departments to act on the recommendation of the auditor general. References: Peer Review Comments: It is not easy to establish compliance with auditor general's recommendations as departments are not currently obliged to report on actions taken in response to the auditor general's and SCOPA resolutions. The level of implementation of auditor general's resolutions in the Eastern Cape is weak. The PSAM has established that not a single resolution passed by SCOPA in the province has been responded to by the provincial executive between 1996 and 2002. Most of these resolutions would have been informed by the auditor general's findings. 57 Can citizens access reports of the supreme audit institution? - 57a: In law, citizens can access reports of the agency. Score: yes/1.00 Comments: Yes, reports of the Auditor General's Office are available to citizens. References: www.agsa.co.za - 57b: In practice, citizens can access the agency reports within a reasonable time period. Score: < 1 month/1.00 Comments: Citizens can access these reports within a reasonable time period. References: www.agsa.co.za

Peer Review Comments: Copies of the auditor general's financial audit reports by law have to be published as part of all government departments annual reports. These reports should be tabled in the relevant Legislature by the 31st of August each year. After this point they should be available for public inspection. - 57c: In practice, citizens can access the agency reports at a reasonable cost. Score: almost always/1.00 Comments: Citizens can access these reports for free. References: www.agsa.co.za

South Africa: Oversight and Regulatory Mechanisms

Sub-Category: V-3/Taxes and Customs Indicators
58 In law, is there a national tax collection agency? 59 Is the tax collection agency effective? 60 In practice, are tax laws enforced uniformly and without discrimination? 61 In law, is there a national customs and excise agency? 62 Is the customs and excise agency effective?

Scores
1.00 1.00 1.00 1.00 1.00

63 In practice, are customs and excise laws enforced uniformly and without discrimination? 1.00

Indicator and sub-Indicator Details

58 In law, is there a national tax collection agency? Score: yes/1.00 Comments: Yes, there is the South African Revenue Services (SARS). References: www.sars.gov.za Peer Review Comments: SARS is one of the New South Africa's success stories. It is unlikely that taxes have ever been as efficiently collected from the formal sector until now. However, the informal sector, particularly industries such as the taxi industry, which generate great wealth, often escape the taxman's net. [Comment 2]: The SARS has also introduced an option of resolving tax disputes within an alternative dispute resolution (ADR) procedure. ADR is not intended to be a "court case," but an opportunity to settle differences of fact and interpretation of law. This option is available in addition to a taxpayer's right to appeal to the Tax Court or Board, and any delays caused through the ADR procedure will not affect this right. 59 Is the tax collection agency effective? - 59a: In practice, the tax collection agency has a professional, full-time staff. Score: almost always/1.00 Comments: Yes, the agency has a professional, full-time staff and in terms of its act, it has the power to determine its own staff establishment, appoint employees and determine their terms and conditions of employment. In the past year, the agency has gone out of its way to professionalize its image and services and is very strong in cracking down corruption, issuing press releases welcoming the conviction and jail sentences of previous employees to spread the message that corruption will never be tolerated. SARS Commissioner Praving Gordhan noted, "We are committed to rooting our fraud and corruption whether it comes from within our ranks or outside and we make no apologies for that. Law abiding South Africans should know that we will never hesitate to use any means to defend our integrity and the confidence that the public has entrusted on us as a government institution." References: South African Revenue Services Act no 34 of 1997; www.sars.gov.za; - 59b: In practice, the agency receives regular funding. Score: almost always/1.00 Comments: Yes, the chief source of income for the SARS is money appropriated annually by Parliament

for its services. SARS may also charge an organ of state or institution concerned at an agreed rate. References: South African Revenue Services Act number 34 of 1997, Section 25. South African Revenue Services (www.sars.gov.za) - 59c: In practice, the agency makes regular reports to the legislature. Score: almost always/1.00 Comments: Yes. In terms of the Act, the commissioner of SARS must annually submit a report to the minister of finance on the activities of SARS during the financial year. The minister must without delay table a copy of the report in the National Assembly and submit a copy to the National Council of Provinces. References: South African Revenue Services Act no 34 of 1997, section 29; www.sars.gov.za; 60 In practice, are tax laws enforced uniformly and without discrimination? Score: almost always/1.00 Comments: The agency has professionalized its images and services tremendously in the past few years, and aims to enforce laws uniformly and without discrimination. References: 61 In law, is there a national customs and excise agency? Score: yes/1.00 Comments: Yes, the South African Revenue Services serves this function. References: www.sars.gov.za 62 Is the customs and excise agency effective? - 62a: In practice, the customs and excise agency has a professional, full-time staff. Score: almost always/1.00 Comments: Yes, the agency has a professional, full-time staff and in terms of its act has the power to determine its own staff establishment, appoint employees, and determine their terms and conditions of employment. References: www.sars.gov.za - 62b: In practice, the agency receives regular funding. Score: almost always/1.00 Comments: Yes, the chief source of income for the SARS is money appropriated annually by Parliament for its services. SARS may also charge an organ of state or institution concerned at an agreed rate. References: www.sars.gov.za Peer Review Comments: It is understood that the Customs and Excise are understaffed and currently cannot cover every port and airfield in the country, meaning that goods can be smuggled in to South Africa quite easily. This contrasts with the zealousness with which officials line up to fine returning tourists for bringing back too many goods. - 62c: In practice, the agency makes regular reports to the legislature. Score: almost always/1.00 Comments: Yes. In terms of the Act the Commissioner of SARS must annually submit a report to the minister of Finance on the activities of SARS during the financial year. The minister must, without delay, table a copy of the report in the National Assembly and submit a copy to the National Council of Provinces. References: www.sars.gov.za 63 In practice, are customs and excise laws enforced uniformly and without discrimination? Score: almost always/1.00 Comments: The agency has professionalized its images and services tremendously in the past few years, and aims to enforce laws uniformly and without discrimination. References: www.sars.gov.za

South Africa: Oversight and Regulatory Mechanisms

Sub-Category: V-4/Financial Sector Regulation

Indicators
64 In law, is there a central bank? 65 In practice, is the central bank independent of the executive?

Scores
1.00 1.00

66 In law, is there a financial regulatory agency overseeing publicly listed companies? 1.00 67 Is the financial regulatory agency effective? 68 Can citizens access the financial records of publicly listed companies? 69 Are business licenses available to all citizens? 1.00 0.92 1.00

Indicator and sub-Indicator Details

64 In law, is there a central bank? Score: yes/1.00 Comments: Yes, there is the South African Reserve Bank. The bank was established in 1921 in terms of the Currency and Bank Act of Aug. 10, 1920. References: South African Reserve Bank Act 90 of 1989; www.reservebank.co.za 65 In practice, is the central bank independent of the executive? Score: almost always/1.00 Comments: Yes, the bank prizes its independence. References: Peer Review Comments: The bank seems highly independent operationally. But the system does allow for government to make overall policy inputs and perhaps even give directives. A couple of years ago the bank changed its mandate to base monetary decisions solely on inflation-targeting criteria, which was an instruction/recommendation from the Treasury. 66 In law, is there a financial regulatory agency overseeing publicly listed companies? Score: yes/1.00 Comments: Yes, the Johannesburg Stock Exchanges Listing division regulates the reporting and financial records of all public companies. The Financial Services Board was set up by the Financial Services Board Act, 97 of 1990, and is financed by the financial services industry to supervise the control over the activities of non-banking financial services. References: JSE Control Act Peer Review Comments: There is a growing trend for companies to delist for various reasons, including the governments surprisingly (for a developing country) lax control over foreign listings. 67 Is the financial regulatory agency effective? - 67a: In law, the financial regulatory agency is protected from political interference. Score: yes/1.00 Comments: Yes, the financial regulatory agency is protected from political interference. References: Interviews conducted for Global Access 2002 pilot report - 67b: In practice, the agency has a professional, full-time staff. Score: almost always/1.00 Comments: Yes, the agency has a professional, full-time staff. References: Interviews conducted for Global Access 2002 pilot report - 67c: In practice, the agency receives regular funding. Score: almost always/1.00 Comments: Yes, it is self-funded and self-regulated. References: Interviews conducted for Global Access 2002 pilot report - 67d: In practice, when necessary, the financial regulatory agency independently initiates investigations. Score: almost always/1.00

Comments: Yes, the JSE Listings and compliance division can discipline and sanction offenders according to a process with timeframes for compliance, which consist of three stages: 1) private censure, 2) public censure, and 3) a fine. Therefore, depending on the nature of the noncompliance, the results of such actions would be public. References: Interviews conducted for Global Access 2002 pilot report - 67e: In practice, when necessary, the financial regulatory agency imposes penalties on offenders. Score: almost always/1.00 Comments: Yes, the JSE Listings and compliance division can discipline and sanction offenders according to a process with timeframes for compliance, which consist of three stages: 1) private censure, 2) public censure, and 3) a fine. Therefore, depending on the nature of the noncompliance, the results of such actions would be public. References: Interviews conducted for Global Access 2002 pilot report Peer Review Comments: The FSB has, on occasion, been criticized for its toothlessness on insider trading matters. Successful prosecutions have been extremely rare. 68 Can citizens access the financial records of publicly listed companies? - 68a: In law, citizens can access the financial records of publicly listed companies. Score: yes/1.00 Comments: Yes, all publicly listed companies need to make their financial records available. References: Interviews conducted for Global Access 2002 pilot report - 68b: In practice, the financial records of publicly listed companies are regularly updated. Score: almost always/1.00 Comments: Yes, the financial records of publicly listed companies are regularly updated. References: Interviews conducted for Global Access 2002 pilot report - 68c: In practice, the financial records of publicly listed companies are audited according to international accounting standards. Score: almost always/1.00 Comments: Yes, GAAP practices are followed and the JSE Listings and compliance division monitors such reporting. References: Interviews conducted for Global Access 2002 pilot report - 68d: In practice, citizens can access the records of disciplinary decisions involving publicly-listed companies. Score: sometimes/0.50 Comments: Yes, the JSE Listings and compliance division can discipline and sanction offenders according to a process with timeframes for compliance, which consist of three stages: 1) private censure, 2) public censure, and 3) a fine. Therefore, depending on the nature of the noncompliance, the results of such actions would be public. References: Interviews conducted for Global Access 2002 pilot report - 68e: In practice, citizens can access the financial records of publicly listed companies within a reasonable time period. Score: < 1 month/1.00 Comments: Yes, almost immediately. References: Interviews conducted for Global Access 2002 pilot report - 68f: In practice, citizens can access the financial records of publicly listed companies at a reasonable cost. Score: almost always/1.00 Comments: Yes, there is no cost. References: Interviews conducted for Global Access 2002 pilot report 69 Are business licenses available to all citizens? - 69a: In law, business licenses are not restricted to domestically-owned enterprises. Score: yes/1.00

Comments: General business licensing occurs at the local government level. Business licenses are not restricted to domestically-owned enterprises, although some concessions may be limited to citizens only. References: Department of Trade and Industry (www.dti.gov.za) - 69b: In law, a complaint mechanism exists if a business license request is denied. Score: yes/1.00 Comments: There is a complaint mechanism to query a business license denied. References: Department of Trade and Industry (www.dti.gov.za) - 69c: In practice, citizens can obtain any necessary business license (i.e. for a small import business) within a reasonable time period. Score: < 1 month/1.00 Comments: It usually takes a few days to get a business license once the paper work is filled out. References: Department of Trade and Industry (www.dti.gov.za) - 69d: In practice, citizens can obtain any necessary business license (i.e. for a small import business) at a reasonable cost. Score: almost always/1.00 Comments: Citizens can obtain a license at a reasonable cost. References: Department of Trade and Industry (www.dti.gov.za)

South Africa: Anti-Corruption Mechanisms and Rule of Law

Sub-Category: VI-1/Anti-Corruption Law Indicators
71 In practice, are anti-corruption laws enforced?

Scores
0.25

70 In law, is there legislation criminalizing corruption? 1.00

Indicator and sub-Indicator Details

70 In law, is there legislation criminalizing corruption? - 70a: In law, attempted corruption is illegal. Score: yes/1.00 Comments: Yes, attempted corruption is illegal both in terms of the old law as well as the new bill. References: The Corruption Act, No 94 of 1992; The Prevention of Corrupt Activities Bill 2003; Parliamentary Monitoring Group (www.pmg.org.za); Member of the Justice Committee - 70b: In law, extortion is illegal. Score: yes/1.00 Comments: Yes, extortion is illegal. References: The Corruption Act, No 94 of 1992; - 70c: In law, offering a bribe (i.e. active corruption) is illegal. Score: yes/1.00 Comments: Yes, offering a bribe is illegal, both in terms of the old law and the new bill, where the common law crime of bribery, which was repealed in 1992, is going to be reintroduced in the new bill. References: The Corruption Act, No 94 of 1992; The Prevention of Corrupt Activities Bill 2003; Parliamentary Monitoring Group (www.pmg.org.za); Member of the Justice Committee - 70d: In law, receiving a bribe (i.e. passive corruption) is illegal. Score: yes/1.00 Comments: Yes, receiving a bribe is illegal, both in terms of the old law and the new bill where the common law crime of bribery, which was repealed in 1992, is going to be reintroduced in the new bill. References: The Corruption Act, No 94 of 1992; The Prevention of Corrupt Activities Bill 2003; Parliamentary Monitoring Group (www.pmg.org.za); Member of the Justice Committee - 70e: In law, bribing a foreign official while in domestic territory is illegal. Score: yes/1.00 Comments: No, not yet, although the new bill creates a separate set of offences related to "corrupt activities relating to foreign public officials." References: The Prevention of Corrupt Activities Bill 2003; Parliamentary Monitoring Group (www.pmg.org.za); Member of the Justice Committee - 70f: In law, using public resources for private gain is illegal. Score: yes/1.00 Comments: Yes, the new Prevention of Corrupt Activities Act makes it a crime to use public resources for private gain. References: - 70g: In law, using confidential state information for private gain is illegal. Score: yes/1.00 Comments: Yes. Under the new bill there is a specific section dealing with offenses in respect to corrupt

activities relating to disclosure of information for gratification. References: The Prevention of Corrupt Activities Bill 2003; Parliamentary Monitoring Group (www.pmg.org.za); Member of the Justice Committee - 70h: In law, money laundering is illegal. Score: yes/1.00 Comments: Yes, the Prevention of Organized Crime Act, no. 121 of 1998 makes money laundering an offense. References: The Prevention of Organized Crime Act no 121 of 1998 (www.polity.gov.za) - 70i: In law, conspiracy to commit a crime (i.e. organized crime) is illegal. Score: yes/1.00 Comments: Yes, the Prevention of Organized Crime Act, no 121 of 1998, makes conspiracy to commit a crime an offense. References: The Prevention of Organized Crime Act no 121 of 1998 (www.polity.gov.za) 71 In practice, are anti-corruption laws enforced? Score: rarely/0.25 Comments: The Corruption Act has rarely been enforced and there have been serious difficulties in obtaining convictions under the act, where police have charged suspects with fraud and theft as these charges have been easier to prove. The new bill that creates a range of prosecutable offenses should make a difference in the enforcement against acts of corruption. References: April 2003 DPSA/UNODC Country Corruption Assessment Report (www.dpsa.gov.za); Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za Peer Review Comments: Yes, the most serious deficiency of the old law was that it basically imposed on prosecutors the obligation to prove not only that gratification had been demanded or offered for the illicit exercise of power, but that in fact the official had indeed had the power demanded or offered. In the Tony Yengeni case, for example, it was not clear that Yengeni, in any formal sense, had the power to affect tender outcomes or that he had done so -- which meant prosecutors had to concentrate on the side issues of how Yengeni had sought to mislead Parliament, etc. The new bill seems to have gone out of its way to remedy that, and many other, deficiencies. Interestingly, it seems that, under the new bill (were it to apply retrospectively, which it won't) Jacob Zuma would have stood a much greater chance of being convicted. He did receive, on his own version, "interest-bearing loans" from businesspeople for whom, hard evidence tends to show, he had done favors. The new bill specifically includes loans in its definition of "gratification". [Comment 2]: The extraordinary case of Jurgen Harksen indicates how poorly the state was able to implement anti-corruption laws. Fugitive from justice, Harksen, was able to battle the state in the courts, with stolen funds, for about seven years, while living in style in Cape Town, before he was finally deported from South Africa to face trial in Germany. While he was in South Africa, he managed to steal more money and corrupt politicians. Similarly the continued presence of alleged mafia kingpin, Vito Palazzolo, in South Africa indicates how slow the state is to deport and catch a big-time crook, although at the time they had evidence to charge him. State officials pursuing him have either been co-opted by him, or beaten out of the system. These are two examples to demonstrate the point.

South Africa: Anti-Corruption Mechanisms and Rule of Law

Sub-Category: V1-2/Anti-Corruption Agency Indicators
73 Is the main anti-corruption agency effective? 74 Can citizens access the main anti-corruption agency?

Scores
0.92 0.88

72 In law, is there an agency (or group of agencies) with a legal mandate to address corruption? 1.00

Indicator and sub-Indicator Details

72 In law, is there an agency (or group of agencies) with a legal mandate to address corruption? Score: yes/1.00

Comments: Yes, there are a range of anti-corruption agencies with a legal mandate to address corruption. References: April 2003 DPSA/UNODC Country Corruption Assessment Report (www.dpsa.gov.za); Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za 73 Is the main anti-corruption agency effective? - 73a: In law, the agency is protected from political interference. Score: yes/1.00 Comments: This section will deal with the Scorpions, also known as the Directorate of Special Operations (DSO) within the National Prosecuting Authority. Three investigating directorates were established in terms of the National Prosecuting Authority Act 32 of 1998, with a mandate to cover serious economic offences, organized crime and public violence, and corruption. The National Prosecuting Authority Amendment Act 61 of 2000 consolidated the directorates into one Directorate of Special Operations (DSO) headed by a deputy national director of prosecutions. Therefore, the operational independence of the DSO within the National Prosecuting Authority is protected from political interference. References: National Prosecuting Authority Amendment Act 61 of 2000 (www.polity.gov.za) Peer Review Comments: The DSO is not protected in law from interference as the NDPP reports to the executive (through the Minister of Justice). Although not a political puppet, the DSO is not a completely independent body and can be subject to pressure from the executive. - 73b: In practice, the agency is protected from political interference. Score: often/0.75 Comments: The NPA has shown its independence by successfully prosecuting two high level members of the ANC (Winnie Madikizela Mandela and Tony Yengeni) with fraud. In late 2003 the National Prosecuting Authority's (NPA) freedom from political interference was tested. The head of the agency, Bulelani Ngucka, was subjected to a smear campaign and allegations that he had been an apartheid-era spy. A subsequent commission of inquiry appointed by President Mbeki, the Hefer Commission, cleared Ngucka of these charges. References: media reports: www.bday.co.za; www.mg.co.za; www.iol.co.za Peer Review Comments: While Ngcuka's image as an independent prosecuting/investigations head has been publicly revived by his tough pursuit of Winnie, Yengeni and Zuma (or at least of Zuma's associates), his image was not that good during the main arms deal investigation. It may be argued that Ngcuka has, at the very least, applied some "self-censorship". Some would perceive a pattern where individuals who are out of favor with the presidency (such as Zuma and Winnie) are much more likely to be "fair game" than individuals who are close to the president. [Comment 2]: The president's comments regarding the future of the Scorpions could be interpreted as a failure to defend the DPP from political interference. Arguably, this statement contributed to the atmosphere instability and anxiety within the unit after it had announced that there was a prima facie case of corruption to be answered by the Deputy President, Jacob Zuma. Correction. Yengeni was criminally convicted of fraud for misleading Parliament. He was not convicted of corruption. - 73c: In practice, the head of the agency is protected from removal without relevant justification. Score: almost always/1.00 Comments: Yes, the head of the DSO is appointed by the president but may only be removed if there are sound reasons for doing so. In April 2003, Leonard McCarthy was appointed deputy national director of public prosecutions and head of the Scorpions. References: National Prosecuting Authority Amendment Act 61 of 2000 (www.polity.gov.za); www.iol.co.za - 73d: In practice, appointments to the agency are based on professional criteria. Score: almost always/1.00 Comments: Yes, there are public service regulations which govern this agency and the merit principle prevails in terms of this law as well as the constitutional principles governing public service. References: Public Service Regulations. Constitution of South Africa. - 73e: In practice, the agency has a professional, full-time staff. Score: often/0.75 Comments: Yes, the DSO has a professional, full-time staff and forms part of the 5,000 people making up the National Prosecuting Authority. It has not been without problems. In June 2003, the entire human resources staff of the NPA was suspended as a precautionary measure for allegedly creating ghost workers and pocketing their pay. References: April 2003 DPSA/UNODC Country Corruption Assessment Report (www.dpsa.gov.za); Public

Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za; www.sundaytimes.co.za - 73f: In practice, the agency receives regular funding. Score: almost always/1.00 Comments: Yes, the agency receives regular funding and is rather better funded than other criminal justice agencies. References: April 2003 DPSA/UNODC Country Corruption Assessment Report (www.dpsa.gov.za); Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za - 73g: In practice, the agency makes regular reports to the legislature. Score: often/0.75 Comments: Yes, the DSO reports through the head of the National Prosecuting Authority directly to Parliament. References: www.pmg.org.za Peer Review Comments: The director of public prosecutions reports directly to the minister of justice. The minister of justice reports to Parliament. The director of public prosecutions may table reports in Parliament as he or she deems necessary. Reference: National Prosecuting Authority Act, 1998. - 73h: In practice, the agency has sufficient powers to carry out its mandate. Score: almost always/1.00 Comments: The agency has extensive powers to carry out its mandate. References: Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za - 73i: In practice, when necessary, the agency independently initiates investigations. Score: almost always/1.00 Comments: Yes, the DSO has operational independence to initiate its own investigations. References: Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za 74 Can citizens access the main anti-corruption agency? - 74a: In practice, the main anti-corruption agency acts on complaints within a reasonable time period. Score: < 3 months/0.75 Comments: The DSO is seen as more efficient than the normal police service, but the nature of its cases are more specialized and complicated, and may take some time to resolve. References: Interview: Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za - 74b: In practice, citizens complain to the agency without fear of recrimination. Score: almost always/1.00 Comments: The DSO is widely respected and seen as accessible to citizens to complain to without fear of recrimination. References: Interview: Public Service Commission National Review of Anti-Corruption Agencies, August 2001.www.psc.gov.za

South Africa: Anti-Corruption Mechanisms and Rule of Law

Sub-Category: VI-3/Rule of Law and Access to Justice Indicators
76 In law, is there a general right of appeal?

Scores
1.00

75 In practice, does the criminal justice process function according to the rule of law? 1.00

77 Are citizens protected from detention without trial? 78 Are individual economic rights guaranteed?

0.75 1.00

Indicator and sub-Indicator Details

75 In practice, does the criminal justice process function according to the rule of law? Score: almost always/1.00 Comments: Rule of law is upheld in South Africa. While access to the criminal justice system is uneven, since 1994, efforts have been made to ensure a non-arbitrary code of practice when it comes to the functioning of the criminal justice process. The courts function within the Criminal Procedure Act and there is generally consistency. References: Interview with chair of the Human Rights Commission 76 In law, is there a general right of appeal? Score: yes/1.00 Comments: Yes, a right of appeal is guaranteed under the Constitution under the right to a fair trial, Section 35 3 (o). References: Interview with chair of the Human Rights Commission; Constitution of South Africa, Act no. 108 of 1996. 77 Are citizens protected from detention without trial? - 77a: In practice, the government does not detain anyone without charging them for more than 48 hours. Score: almost always/1.00 Comments: In practice this generally does not occur, however there are circumstances when a person may be taken into custody on a Wednesday evening and it may be too late on a Friday to charge them in a court of law. References: Interview with chair of the Human Rights Commission - 77b: In practice, the government does not detain anyone accused of petty theft for longer than two weeks without a resolution in a court trial. Score: sometimes/0.50 Comments: In general, those accused of petty theft are released on bail after being charged. This generally occurs within two weeks of being taken into custody. At the moment there are efforts underway to release from detention those who cannot pay bail of R1,000 or less in order to lesson overcrowding. Prisons in South Africa remain seriously overcrowded, with the total prison population reaching 175,290 at the end of 2001, outstripping the system's capacity by around 65,000. About a third of the total of the prison's population consisted of person's awaiting trial, with more than half of incarcerated juveniles (under the age of 21) still awaiting trial, and 40 percent of pretrial prisoners unable to afford relatively low bail. References: Interview with chair of the Human Rights Commission; Human Rights Watch World Report 2003: South Africa Peer Review Comments: There are many stories of suspects in minor crimes (often possession of soft drugs) being detained for months before their trials start. I suspect that that means, in practice, that suspects in cases of petty theft would very often be detained for more than two weeks, purely because of their inability to afford bail. [Comment 2]: Many people are held for minor crimes for lengths of time because they simply do not have enough money to pay for bail. 78 Are individual economic rights guaranteed? - 78a: In law, individual property rights are protected Score: yes/1.00 Comments: Yes, Section 25 (1) of the Constitution protects property in that no one may be deprived of property except in terms of law of general application, and no law may permit arbitrary deprivation of property. References: Constitution of South Africa, Act 108 of 1996 (www.polity.gov.za) - 78b: In practice, individual property rights are protected. Score: almost always/1.00 Comments: Yes, individual property rights are upheld and there are a range of sophisticated procedures in

place. References: Interview with chair of the Human Rights Commission. - 78c: In practice, the government does not expropriate property without appropriate compensation. Score: almost always/1.00 Comments: According to Section 25 (2) of the Bill of Rights, property may be expropriated only in terms of law of general application for a public purpose or in the public interest; and subject to compensation, the amount of which and the time and manner of payment of which have either been agreed to by those affected or decided or approved by a court. The amount of the compensation, and the time and manner of payment must be just and equitable, reflecting an equitable balance between the public interest and the interests of those affected, having regard to all relevant circumstances, including the current use of the property; the history of the acquisition and use of the property; the market value of the property; the extent of direct state investment and subsidy in the acquisition and beneficial capital improvement of the property; and the purpose of the expropriation. References: Interview with chair of the Human Rights Commission. Constitution of South Africa, Act 108 of 1996 (www.polity.gov.za) - 78d: In practice, legal contracts are honored. Score: almost always/1.00 Comments: Yes, the South African state has a strong tradition of honoring legal contracts. In terms of breach of contract among private individuals, this may happen and there is a legal process to address this. References: Interview with chair of the Human Rights Commission.

South Africa: Anti-Corruption Mechanisms and Rule of Law

Sub-Category: VI-4/Law Enforcement Indicators
79 Is the law enforcement agency (i.e. the police) effective?

Scores
0.58

80 Can law enforcement officials be held accountable for their actions? 0.94

Indicator and sub-Indicator Details

79 Is the law enforcement agency (i.e. the police) effective? - 79a: In practice, appointments to the law enforcement agency are made according to professional criteria. Score: often/0.75 Comments: There are regulations governing the appointment of law enforcement officials. Efforts have been made to diversify the composition of the force since 1994. References: Interview with chair of the Human Rights Commission. - 79b: In practice, the agency has a budget sufficient to carry out its mandate. Score: sometimes/0.50 Comments: The South African Police Service is generally perceived to be under funded. The fact that an organization like Business Against Crime supplements basics such as vehicles for the police service, suggests that they are under funded. References: Interview with chair of the Human Rights Commission. Peer Review Comments: Police remain very badly paid: for example an inspector, a relatively senior official, after taxes earns roughly R5,000 a month, about the same as a diploma qualified high school teacher. [Comment 2]: The issue of funds is two ways: it is partly an issue of insufficient budgetary allocation and misallocation of resources. The political interference within the police service in South Africa needs to be understood within the context of political transition and need to transform the racial imbalances within the security sector. [Comment 3]: The criminal justice system, the police and the prisons do not have enough resources to adequately police the country, enforce the law and rehabilitate prisoners. In some instances, for example, a police station might have only one van, and would be staffed with policemen

without driving licenses. Salaries are too low, making it easy for police to be corrupted. - 79c: In practice, the agency is protected from political interference. Score: sometimes/0.50 Comments: The police are generally perceived to operate without political interference. This is difficult to quantify, but there are cases, for example a stronger policing presence and crackdown on illegal aliens, which precede an election, suggesting some political interference. The civilian controlled Secretariat for Safety and Security which was established to provide oversight has largely become an extension of the SAPS. References: Interview with chair of the Human Rights Commission. Peer Review Comments: The police are vulnerable to political interference. In many instances, there may be no interference but when there is, it is damaging. For example, the police investigated three top ANCaligned businessmen - potential challengers to the ANC presidency - and the investigation, which seemed without basis, was announced to the media, effectively smearing their reputations and warning them off trying for president. Similarly, the police recently arrested, on trumped up charges, a political and commercial risk analyst, Bheki Jacobs, whisked him from Cape Town to Pretoria in a jet for reasons nobody has yet been properly been able to explain. Charges that were put to him that he was involved in a plot to kill the president have been dropped. These are examples of extraordinary "political" actions by the police. [Comment 2]: The agency is not protected from political interference, the head of police is appointed by the president. 80 Can law enforcement officials be held accountable for their actions? - 80a: In practice, there is an independent mechanism for citizen complaints about police action. Score: almost always/1.00 Comments: The Independent Complaints Directorate (ICD) is a statutory body set up to receive citizen complaints of police misconduct and abuse of power. The ICD reported 214 deaths in police custody and 371 otherwise as a result of police action in the year to April 2002, a 12.8 percent decline from the previous year, although complaints to the ICD had increased by 15 percent. The ICD has called for greater civilian oversight of police to help discover and prevent torture. References: www.icd.gov.za; Human Rights Watch World Report 2003: South Africa - 80b: In law, there is an agency to investigate and prosecute corruption committed by law enforcement officials. Score: yes/1.00 Comments: Yes, although it is not an agency outside of the South African Police Services. Until recently there used to be a dedicated police anti-corruption unit. In October 2002, the SAPS Anti-corruption Unit was shut down with its members and functions integrated into the Organized Crime Unit and the Serious and Violent Crime units of the SAPS, in line with re-organization of the Detective Service "in order to pool resources and make the SAPS more efficient." Established in 1996, at its full strength, the unit consisted of a total of 240 staff (201 police members, 39 civilian staff). References: Newham, G. & Gomomo, L. 2003. "The closure of the SAPS Anti-corruption unit", Center for the Study of Violence and Reconciliation. www.csvr.org.za - 80c: In law, law enforcement officials are not immune from prosecution. Score: yes/1.00 Comments: South African law does not provide any individuals or organizations special grants of immunity from indictment, prosecution, or preventive custody (e.g. Members of Parliament, judges, prosecutors, etc.). In effect, prosecutions against law enforcement officials are taken more seriously and deciding not to prosecute requires a higher level of authority. References: Criminal Procedure Act. Interview with chair of the Human Rights Commission. - 80d: In practice, law enforcement officials are not immune from prosecution. Score: often/0.75 Comments: Law enforcement officials are not immune from prosecution, with even high ranking police officials prosecuted. A recent example is that of Piet Meyer, head of Organized Crime in KwaZuluNatal. Between its establishment in 1996 and the end of 2001, the SAPS Anti-corruption Unit received a total of 20,779 allegations of police corruption, 3,045 police members were arrested and 576 were convicted. References: Newham, G. & Gomomo, L. 2003. "The closure of the SAPS Anti-corruption unit", Center for the Study of Violence and Reconciliation. www.csvr.org.za Peer Review Comments: There is still a strong culture in the police where colleagues protect each other. So while there may be the will from higher levels to act against officers suspected of crime, in practice such investigations are often hard to conduct and prosecute because of the system of informal protection.

[Comment 2]: The Police Anti-Corruption Unit was disbanded in 2003 and absorbed into the SAPS organized crime unit, some of whose members it was allegedly investigating.

South Africa: Country Facts

Land
Land area (sq km)

Facts
1.22 million

Population
Adult illiteracy rate (% of people ages 15 and above) Life expectancy at birth (years) Infant mortality rate (per 1,000 live births) Population growth (annual %) Population Poverty headcount (% of population living below the national poverty line) Ethnic breakdown Religious breakdown Languages Murder rate (per 100,000)

Facts
14.39 47.14 56 0.78 43.24 million NO INFO Black, 77.8%; White, 10.2%; Colored, 8.7%; Asian (Indian), 2.5%; other, 0.85% Predominantly Christian; traditional African, Hindu, Muslim, Jewish Afrikaans, English, Ndebele, Pedi, Sotho, Swazi, Tsonga, Tswana, Venda, Xhosa, Zulu (all official languages) 114.84

Political
Capital city Character of government Current president/prime minister Executive branch description Legislative branch description

Facts
Administrative, Pretoria; Legislative, Cape Town; Judicial, Bloemfontein Republic; current constitution effective since 1997. Thabo Mbeki President named for a five-year term by the National Assembly. A bicameral Parliament consists of the 400-seat National Assembly (members are elected by popular vote under a system of proportional representation) and the 90-seat National Council of Provinces (10 members elected by each of the nine provincial legislatures). Members of both serve five-year terms. Constitutional Court; Supreme Court of Appeals; High Courts; Magistrate Courts 30

Judicial branch description Percentage of women in legislature

Economics
Aid (% of central government expenditures) Aid per capita (US$) Net foreign direct investment (current US$) GDP growth (annual %) GDP per capita (constant 1995 US$) Consumer prices inflation (annual %) Military expenditure (% of central

Facts
1.3 9.91 10.85 billion 2.22 4,068 4.83 5.4

government expenditure) Military expenditure (% of GDP) Tax revenue (% of GDP) 1.6 26.51

Exchange rate to US$1 (as of January, 6.92 2004) Currency Public spending on education (% of GDP) Public health expenditure (% of GDP) Main exports Unemployment (% of total labor force) External debt (current US$) Rand 5.5 3.71 Gold, diamonds, metals and minerals, machinery 23.3 24.05 billion

Information/Technology
Radio: Radios (per 1,000 people) Telephone: Telephone mainlines (per 1,000 people) Mobile phones (per 1,000 people) Television: Television sets (per 1,000 people) Television broadcast stations Other Media: Daily newspapers (published at least four times a week) in circulation per 1,000 people Internet users

Facts
338

112 252

152 Three stations controlled by the South African Broadcasting Corporation and one private station.

32

3.07 million

A Comparative Study of Inequality and Corruption Author(s): Jong-Sung You and Sanjeev Khagram Source: American Sociological Review, Vol. 70, No. 1, (Feb., 2005), pp. 136-157 Published by: American Sociological Association Stable URL: http://www.jstor.org/stable/4145353 Accessed: 19/04/2008 18:22
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=asa. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We enable the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

http://www.jstor.org

A Comparative Study and of Inequality Corruption

You, Jong-sung HarvardUniversity Sanjeev Khagram HarvardUniversity

increasesthe level of corruption Thisarticlearguesthatincomeinequality through and The mechanisms. wealthyhave bothgreatermotivation materialand normative to whereasthepoor are morevulnerable to moreopportunity engage in corruption, as extortionand less able to monitorand hold the rich andpowerfulaccountable and also adverselyaffectssocial normsaboutcorruption inequalityincreases.Inequality it easier therebymaking people's beliefsaboutthe legitimacyof rulesand institutions, Thiscomparative as analysisof 129 for themto toleratecorruption acceptablebehavior. variables witha varietyof instrumental countriesusing two-stageleast squaresmethods measuresof corruption(the World the authors' usingdifferent hypotheses supports International's Indexand the Transparency Bank'sControlof Corruption Corruption as powerof inequalityis at least as important Index).Theexplanatory Perceptions The such as economicdevelopment. causes of corruption conventionally accepted as and authorsalsofound a significantinteraction effectbetweeninequality democracy, aboutcorruption well as evidencethatinequality using affectsnormsandperceptions to also contributes incomeinequality, the World Values Surveysdata. Becausecorruption and societies oftenfall into viciouscirclesof inequality corruption.

orizing and research on corruptionis surprisingly rare.Whereaspoliticalscientistsandeconomists have examined corruptionprimarilyin relation to economic developmentratherthan inequality,sociologists who examine inequality have paid scant attentionto the problemof As corruption. a consequence,the relationships between inequality and corruptionare grossly understudied. This articlemakesthreesets of contributions to Direct all correspondence You, Jong-sung, to the literatureon cross-nationalvariationin F School corruption.First, it gives a theoreticalaccount John Kennedy of Policy, Department Public MA Harvard of Government, of why income inequalityincreases corruption University, Cambridge, The 02138 (youjong@ksg.harvard.edu). authors generally and counterintuitively democratic in Daron thank Alberto Abadie, Acemoglu, Christopher political systems. We arguethatinequalityfosErzo John ters a norm of corruptionas acceptablebehavJohann Lambsdorff, Luttmer, Graf Jencks, Suzanne SusanRose-Ackerman, Shanahan, ior, that corruption is likely to reinforce or Meyer, on andDavidWisefortheirhelpfulcomments earwiden existing inequalities, and that vicious Gili Drori, lier drafts.The authorsacknowledge are circles of inequality-corruption-inequality and RafaelLaPorta, DaniRodrik, DanielTreisman thus likely to manifest. for kindly providing their data as well as the Second, this article shows the explanatory Bank the Institute, International, World Transparency for theirdata ability of income inequalityand the interaction DavidDollar, AartKraay making and testedempiralso acknowledge betweeninequalityanddemocracy The publiclyavailable. authors the Hauser for Center research ically against competing conventional explaby provided support nationsof corruption.Indeed,to the best of our at University. Nonprofit Organizations Harvard income inequality affect corruption? importantquestion has seldom been addressedby social scientists. Although crossthe studiesinvestigating causnationalstatistical haveproliferated es of corruption (Ades andDi Tella 1999; Montinola and Jackman 2002; Paldam2002; Treisman 2000), sociologicalthe-

Does This

2005o,VOL. (February.i36-x57) REVIEW, AMMRICAN 70 SocIOLoGICAL

AND INEQUALITY CORRUPTION 137

knowledge, this article offers the first systematic cross-countrystatistical study focused on the causal effect of income inequality on corruption. Also, evidence for the relationship among inequality,perceptions, and norms of corruptionis provided, as well as the reverse causationfrom corruptionto inequality. The third set of contributions is methodological. Previous cross-national studies on causes of corruptionhave primarilyused ordinary least squares(OLS) methods.As a result, these studies did not directly address critical issues of simultaneous causation or measurement error.We have used a range of instrumental variables and two-stage least squares (2SLS) methods to correct substantially for these problems.We also have used a wide array of controls as well as different measures for corruptionto test the robustnessof our results. We first briefly review major sets of theoretical explanationsand results from previous empiricalstudies on corruption.In the second section, we advocatemuch greaterattentionto inequality,a factorthathas receivedlittle attention by scholars.We describethe dataandmethods used to investigate our hypotheses in the third section. In the fourth section, empirical findings and theoretical interpretationsfrom workarepresented. The final secour statistical tion concludes with some researchand policy implications.

EXISTING EXPLANATIONS AND EMPIRICAL FINDINGS

We use a somewhatnarrowbut widely accepted definition of corruption: abuse of public power (or public office) for private gain. Althoughwe do not see any reasonsto exclude corporateembezzlement,fraudin the nonprofit sector,and the like from the definition, there are no available cross-nationalmeasures that capturethis fuller range of corruption. Until recently, statisticalstudiesof corruption havebeen hinderedby the lack of reliablequantitativedata.As the dataon the (perceived)levels of corruptionbecame available for a large number of countries, cross-country statistical research bourgeoned. Although these studies generateda considerableconsensus regarding the negative effects of corruptionon economic development,underminingthe long-advocated functional view of corruption, studies on the

causes of corruptiondid not produce general agreement (Andvig et al. 2000; Lambsdorff 1999). Numerousvariableshave been suggested as causes of corruption. These variablescan be classified into three broad categories: economic, political, and culturalexplanations. Economic factors often are considered the prime causes of corruption.Many studies have found economic development (per capita income), ostensibly throughthe spreadof education, creationof a middle class, and so forth to be the strongest determinantfor reducing corruption(Paldam2002; Treisman2000). In contrast, Kaufmannand Kraay (2002) argued that causation runs from lower corruption to economic development and not from higher income to less corruption.Ades and Di Tella (1999) and Treisman(2000) found tradeopenness, presumablythroughincreased economic competition and economic growth,to be associatednegativelywith corruption a significant to Torrez (2002) found that its degree, although significance depended on the choice of corruption index. Findings have shown that the significance of the relative wages paid to public servants in controlling corruption depend on the measuresand specificationsused (Evans and Rauch 1999;RijckeghemandWeder2001). Countries with larger endowments of natural resources were found to be significantly more corrupt,probablybecause windfall gains offer greateropportunitiesfor corruption(Ades and Di Tella 1999; Gylfason 2001; Leite and Weidman 1999). Political explanationsof corruptioninclude variables such as democracy,governmentsize, and decentralization. Althoughdemocracy(e.g., electoral competition,political rights) theoretically is supposed to provide checks against corruption, empirical studies have found differing results.' Treisman(2000) concludedthat democracies are significantly less corruptonly after 40 years. Montinola and Jackman(2002) demonstratedthatthe effect of democracymay be nonlinear,that partialdemocratizationmay but increase corruption, thatonce past a threshold, democracy inhibits corruption.The larger

(1999) arguedthat elections 1 Rose-Ackerman of increasethe accountability politicians, that but they also producenew incentivesfor corruption needsincrease. becausepoliticalfinancing

138

REVIEW SOCIOLOGICAL AMERICAN

size of governmentor the greaterextentof governmentinterventionwas proposedto increase corruption(LaPalombara1994), yet empirical are evidenceshows thatlargergovernments less corrupt(Friedmanet al. 2000; La Porta et al. 1999). Moreover,the findings on the effect of decentralization contradictory are (Fismanand Gatti 2002; Treisman2000). Culturaland historical explanationsof corruptionhave highlightedthe effects of religion, cultural values, colonial heritage, legal traditions, and ethnolinguistic fractionalization. Egalitarianor individualisticreligions such as Protestantism may encourage challenges to abuses by officeholders, whereas hierarchical religions such as Catholicism, Eastern Orthodoxy, and Islam may discourage such link with economchallenges. Protestantism's ic developmentand democracyoffer two additional causalpathways.Many empiricalstudies have found Protestantism be associated sigto with less corruption(La Portaet al. nificantly 1999; Paldam 2001; Sandholtz and Koetzle 2000; Treisman2000). Colonial experience and legal system are closely correlated.La Porta et al. (1999) proposed that legal systems reflect the relative power of the state vis-a-vis property owners. Whereas the British common law system was developed as a defense of property owners against attemptsby the sovereign to expropriate property,civil law was developed as a sovereign instrument for state building and economic control. Treisman (2000) further argued that British legal traditions tend to fairness.He foundformer emphasizeprocedural Britishcolonies to be significantlyless corrupt. Althoughcountrieswith Frenchlegal or socialist origins have higher levels of corruption, legal origins are insignificant when control is used for other factors (La Portaet al. 1999). Huntington's cultural areas of Western Europe, Latin America, and old Communist countries (Paldam 2002) and Hofstede's culturalvalues such as power distance, uncertainty avoidance, and masculinity (Husted 1999) also were significant predictorsof corruption. Although some scholars have suggested that would increase ethnolinguisticfractionalization corruption(Mauro 1995), its significance disappearedafter per capita income and latitude controls were added (La Portaet al. 1999).

INEQUALITYAND CORRUPTION The relationship between inequality and corruption was not rigorously theorized or systematically examined in any of the aforementionedstudies. Our overall argument is that greater levels of inequality are social structurallyconducive to higher levels of corruptionthroughmaterialand normativemechanisms. Thus we should find a directempirical betweeninequalityand corruption, relationship factors being equal. In this article, we other focus on income inequality.2 A centraltheoreticalargumentin the literature maintainsthat corruptionis a functionof motivationsand opportunities (Klitgaard1988; Rose-Ackerman 1978, 1999). As income inequalityincreases,the rich have moreto lose and throughfair political, administrative, judithe cial processes.Withthe increasedinequality, rich also have greaterresourcesthatcanbe used to buy influence, both legally and illegally (Glaeser,Scheinkman,and Shleifer2003). The rich,as a class or as interest groups,canuse legal or lobbying and political contributions bribery (grand political corruption)to influence lawmakingprocesses. The rich, as interestgroups, as firms, or as individuals,may use briberyor connections to influence law-implementing processes (bureaucratic corruption)andto buy favorable interpretationsof the law (judicial corruption). As inequalityincreases,most of the population will be relatively poorer,3and likely will demandmore extensive redistribution through higher levels of progressivetaxation (Meltzer and Richard 1981). As the redistributive pressures increase, the rich correspondinglywill have greater motivation to use political corruptionto lower the tax rates and bureaucratic 2Twostudiesfounda significant effectof gender and (Dollar,Fisman, Gatti equalityon corruption foundthat et 2001;Swamy al. 2001).Thesestudies in and women wereless involved bribery less likely bribes. to condone practice taking the of Theyalso women that was where found corruption less severe share thelaborforceandheld a of comprised larger seats. a larger shareof parliamentary with typicallyis associated 3 Higherinequality bothgreater skewness therightanda greater to gap the between richandthepoor.

AND INEQUALITY CORRUPTION 139

corruptionto furthercircumventthe collection of taxes. Whereasthe rich have more motivationand capability to behave corruptlyat higher levels of inequality,the non-rich have more to gain from combating corruption.The middle class and the poor generally have cause to monitor, expose, and halt the corruptionby the rich and the powerful. However,the poor lack material resources to organize, and a thin middle class is likely to exert less influence in high-inequality societies. High levels of inequality (and associated poverty), with other factors held as equal, arethus likely to inhibitthe capacitiesof middle-class poorgroupsto monitorthe corand rupt activities of the rich and powerful (McCarthy and Zald 1977; Tarrow 1994). Greaterequalityis likely to entail a largermiddle class that can act to protect its interests (Husted 1999). Moreover, in high-inequality societies, the large numbers of poor are more likely to be deprivedof basic public services such as education and health care than in low-inequality countries.Hence, they aremore likely to rely on petty corruptionor to be the targetsof bureaucratic extortion in their efforts to secure basic services. Although the amount of their actual kickback payments may be small because of theirlimitedabilityto pay,the poorwill perceive corruptionlevels to be very high and will come to see corruption as an appropriateform of behavior. incomeinequalityis assoHypothesis1: Greater ciated with higher levels of corruption. However, the impact of income inequality on corruptionmay differbetween more demoIn craticandless democratic countries. countries with authoritarianregimes, the rich and the powerful can use or promote repression to advancetheirinterests.In democraticcountries, however,oppressionas a substitutefor corruption cannotbe used, so the richmustrely on corruptionmore and more as inequalityincreases and redistributive pressures grow. Whereas countrieswith more authoritarian regimes are on likely to havehigherlevels of corruption, the average,the effect of inequalityon corruption may be higher in more democraticcountries. in Furthermore, a highlyunequalsociety with of elections,a largenumber poorpeople arelikely to sell their votes in exchange for money,

gifts, or other favors, whereas the rich and the powerful will buy votes to maintainthe status quo of inequality.The poor are likely to be satisfied with small benefits by participatingin petty corruptexchanges and patronageinstead of resistinggrandcorruption the rich andthe by powerful, thus allowing them much larger benefits. Hypothesis 2: The adverse effect of inequality on corruptionis largerin more democratic countries. Human behavior is powerfully determined by values, norms, and perceptions(Dowse and Hughes 1986; Marchand Olson 1989). Values of integrity may differ across individuals, groups, and societies. Religion may have an impact on values and norms about corruption. countries However, people acrosshighly corrupt with widely different religious traditions,and even those who engage in corruption themselves often are found to dislike corruption, combining excuse with condemnation(Miller, Grodeland,and Koshechkina2002). Toleranceof corruptionas acceptablebehavior may be explainednot only by religious values, but also by perceptions concerning the extentof corruptionand associatedwidespread social networks for corruption. If people are or surrounded corruption perceiveit to be the by case, they may have to accept and even participate in corruptiondespite their values. In surveys, people justify their corruptbehavior by citing its prevalence (Rose-Ackerman 2001). Corrupttransactionsoften requirethe involvement of multiple actors, and the consequent networks of corruptionwill offer more social structuralsupport for participationin corrupt activities (Warburton 2001). Correspondingly, we argue that income inequality affects perceptions concerning the extentof corruption habituates and normsabout At higher levels of inequality,the corruption. rich are likely to believe increasinglythat corruptionis an acceptableway of preservingand advancingtheir societal position as this behavior goes unpunishedand social networksof corruptionexpand.Also, people aremore likely to consider political institutions and rules in unequalsocieties as favoringthe rich, as unjust, andas lackinglegitimacy.Morepeople arelikely to circumvent laws and regulations when they are considered illegitimate. Thus, people

140

AMERICAN SOCIOLOGICAL REVIEW

will more easily justify their corruptactivities as inequality increases. most Moreover,at higherlevels of inequality, non-richpeople arelikelyto believe thatthe rich and powerful must be corrupt, and that it is impossible to do well honestly.Hence, they are likely to justify their own corrupt behavior, while finding it difficult to hold the rich and powerful accountable.As the rich and the nonrich engage in corruption, corrupt practices spread, and corrupt networks further expand becomes a norm. and deepen.Thus, corruption As corruptpractices spreadand become habituated as "how things are done" in highly unequal societies, the norm of corruption is socialized by subsequentgenerations. Hypothesis 3: Perceptions of the extent and norms for acceptabilityof corruptionare higher in more unequalsocieties. Corruption tends to reinforce or widen already existing inequalities (Johnston 1989). Corruptioncontributesto inequalityby faciliof tating the unequalappropriation wealth and andby inhibitinginstitutional changes privilege, that could threatenexisting advantages.Thus, we expect to see a persistence of corruption with the persistence of inequality,and hence a mutually reinforcing relationship between inequalityand corruption. Hypothesis 4: Higher levels of corruptionare associatedwith higherlevels of inequality. We did not find any statistical examination specifically focused on the effect of income inequality on corruption,althoughtwo empirical studies(Husted1999;Paldam2002) included income inequality as one of many explanatory variables and tested its effect throughOLS regressions.Neither study found a statistically significant effect, but this negative finding probablywas the result of ineffibias from measurement ciency and attenuation errorsin the inequalityand corruptionindicators used. The empirical tests conducted by Husted (1999) and Paldam(2002) were far fromrigorous, and both authorsfocused primarilyon cultural variables. Husted (1999) used Transparency International's Corruption Perceptions Index (CPI) for 1996 and the income share of the top 10 percent in 1996 from the World Bank's data for a sample of only 36 countries. Paldam (2002) used International's for 1999 and CPI Transparency

gini coefficients for different years from the World Bank's data for samples of 85 to 100 countries.The OLS regressionsof bothauthors showedthat only per capitaincome andcultural variables(culturalvalues for Hustedandculturalareas for Paldam)were significant. Althoughboth studiesfound inequalityto be errorin the income insignificant,measurement inequalitymeasuresmay have caused substantial attenuationbias in their estimates. Both authors used a single measure of corruption and inequality for a single year, making their to resultsparticularly vulnerable chargesof spuriousness. Husted's (1999) small sample size furtherraises the possibility of selection bias. Both authorsused OLS regressions, which cannotaddresspotentialbiases associatedwith measurement error, omitted variables, and The relative size of reverse reverse causation.4 be greaterfor per capitaincome, causalitymay causing it to be relatively overestimated.The erroris likely to be far greater for measurement inequality,causing it to be relativelyunderestimated. Thus, we believe that there is substantial room for more rigorousstatisticalanalysis of the effect that inequalityhas on corruption. The effect of corruption on inequality, in was contrast, morerigorouslyexamined.Gupta, Davoodi,andAlonso-Terme(2002) andLi, Xu, and Zou (2000), using cross-countryanalysis, found a significant effect of corruption on inequality.Gupta et al. (2002) suggested that corruptionincreases inequality by perpetuating an unequaldistributionof asset ownership the andunequalaccess to education, minimizing of the tax system, loweringthe progressiveness level and effectiveness of social spending,and loweringeconomic growth.

DATAAND METHODS
Differentstudiesfrequently haveproduced varyresults depending on the model specificaing tions, statistical methods, and measuresused. We addressedmany of these problems in our analysis.

4Paldam varithat (2002)stated theinstrumental subables2SLSmethod usedforoneeconomic was werenotpresented. model,buttheresults

AND INEQUALITY CORRUPTION 4i

INCOME MEASURBs INEQUALITY

We used gini coefficients based on the highquality income distributiondata compiled by Dollar andKraay(2002). These authorsassembled datafromfoursourcesincludingDeininger and Squire(1996) and the UN-WIDERIncome InequalityDatabase. The gini coefficient rangesfrom 0 to 1, with a gini of 0 representing and perfectequality a gini of 1 meaningthatonly one personor household has the total income in the country. Because income-basedginis aresubstantially higherthan expenditure-based ginis, andbecausedifferences also exist betweengross income-based ginis and net income-based ginis as well as betweenhousehold-based ginis and person-based ginis, we appropriately adjustedthe rawdataon the basis of differentdefinitionsto make them comparable with householdnet expenditure-based ginis. Our adjustmentsfor the data were as follows: adjustedgini (GINI) = gini - .0398 income .0123 gross + .0112 person. The coefficients were based on the regressionof gini in relation to these threevariables,countrydummies, and decadedummies(fordetails,see ourASR appendix online supplement http://www.asanet.org/ at journals/asr/2004/toc043.html). Becausethe effect of inequalityon corruption is likely to be long term, it is betterto use averages over a longer period than datafor a single year. The average value of the adjusted gini (GINI) for the period of 1971-1996, available for 129 countries,was used as an independent variablewhen perceivedcorruption the perifor od 1996-2002 was the dependentvariable.The averagevalue for the periodof the 1990s, available for 114 countries, was used as a dependent variable when we examined the reverse causation from corruptionto inequality. Also, by extendingthe period,we also could minimize measurementerror.Analysis of variance (ANOVA) showed that variation within countriesover time explainedonly 2.1% of the whereasvariation total variation, betweencounA tries explained91.3% of the total variation.5

substantialpart of the variation in inequality within countriesacrosstime was likely to result from measurementerror,and averaginghelped to reduce it.6 CORRUPTION MEASURES The main indicators for corruption were the World Bank Institute'sControl of Corruption Index (CCI) and the Transparency International'sCorruption Perceptions Index (CPI). The two data sets are regarded as the most reliable for cross-national comparisons and cover a large numberof countries.We also used the Political Risk Service's International CountryRisk Guide (ICRG) index of corruption for a robustnesscheck.7These indices representthe perceivedlevel of integrityor freedom from corruptionbecause a highernumberindicates a lower level of corruption.The CCI is a standardizedscore, with a mean of zero and a standarddeviation of one (Kaufinann,Kraay, and Mastruzzi2003). The CPI rangesfrom 0 to 10 (Lambsdorffforthcoming), and the ICRG index ranges from 0 to 6. We used the averagevalues of these indices for corruptionduringthe period 1996-2002 as a dependentvariable.The CCI for 1996-2002 (average for 1996, 1998, 2000, and 2002) was available for 195 countries, including all the 129 countries for which we had the GINI 1971-1996 data.The CPI for 1996-2002 (average for 1996-2002) was availablefor 109 countries. As an independentvariable, the CCI for 1996-1998 (averagefor 1996 and 1998) andthe CPI for 1996-1999 (average for 1996-1999) was used with the dependent variable inequalof ity for the period of the 1990s. The World Bank Institute's CCI and the International's CPI are based on Transparency various sources of survey data, whereas the Political Risk Service's ICRG index of corruption is assessed by its countryexperts.The surveys that provide the basis of the CCI and the

6Assuming measurement hasa normal that error distribution a meanof zeroanda variance c2, with of and that 5 Li, Squire, Zou(1998)alsoshowed 92% will of N observations reducethe variaveraging in of thevariance Deininger Squire and on anceto c2/N. (1996)data for for indexof corruption widethe is 7 Although ICRG ginicoefficients 112countries theyears1947 to 1994is cross-country whereas 1% used by scholars, Lambsdorff(forthcoming) variance, only is ly variance. is evidence thepersistence warned This over-time for its that noting it measures against reliability, of income across countries time. over than risks" rather thedegrees corruption. of inequality "political

142

REVIEW AMERICAN SOCIOLOGICAL

CPI mostly reflect the opinions of international business people and country experts. The various sources differently measure the perceived level of overallcorruption,fromthe frequency of additional payments to get things on done to the effects of corruption the business to The environment grandcorruption. CCIgives more weight to data that are more highly correlated with the resulting aggregate index (Kaufmannet al. 2003). Our corruption data had some limitations. are Measuresfor variouskindsof corruption not the CCI, CPI, or ICRG.Hence it is providedby impossible to test whether inequality affected crossMoreover, types of corruption. particular based on the perceptionsof surcountryratings international business vey respondents(mainly or the subjectivejudgments of experts people) are not only imprecise, but also can be biased. Thus, measurementerrorand systemic bias is concern.Some criticsraisethe spea particular cific possibility that rich countriesare favored by equatingrichnesswith cleanness(Kaufmann et al. 2003). Yet, countrycorruptionindices based on the judgments of experts and internationalbusiness people are highly correlatedwith domestic public perceptions.8 Moreover, it is practically impossible to measure the actual levels of corruptionacross countries.We minimized the estimation inefficiency resulting errorby using averageddata frommeasurement for severalyears(from 1996 to 2002) insteadof data for a single year. We furtherreduced the possibility of reachingspuriousconclusionsby using three differentmeasures of corruption.
AND PERCEPTIONS NORMS

interestslooking out for themselves, or (0) that it is run for the benefit of all the people? Question 2. How widespread do you think are bribetakingand corruption in this country? (1) Almost no public officials, (2) a few public officials, (3) most publicofficials, or (4) almost all public officials are engaged in it. Questions 3 and 4. Tell whether you think each of the followingstatements (10) canalways be justified, (1) neverjustified, or (2-9) something in between. Question3. Cheatingon taxes if you have a chance? Question 4. Accepting a bribe in the course of one's duties? VARIABLES CONTROL We included several economic, political, and cultural variables identified as significant by previous studies: Economic development.The naturallog of gross domesticproduct(GDP) per capita,averaged for 1971-1996, were calculatedfromthe WorldBank's WorldDevelopment Indicators. usingthe La Missing valueswere supplemented Portaet al. (1999) andTreisman (2000) dataset. Tradeopenness. The naturallog of percentfor age importsplus exportsoverGDP,averaged 1971-1996, were calculated from the World Development Indicators.Missing values were supplementedusing the La Portaet al. (1999), Treisman (2000), and Rodrik, Subramanian, and Trebbi(2002) data set. Natural resource abundance. The share of fuel, ore, and metal exportsfrom the totalmerfor chandiseexports,averaged 1971-1996, were calculated from the World Development Indicators. Democracy.The political rights score, averaged for 1972-1996, were calculatedfromthe Freedom House.9 The political rights score reflectsthat(1) thereare free and fairelections, (2) those elected rule, (3) there are competitive parties or other competitive political groups, (4) the opposition has an importantrole and

and As the measureforperceptions normsabout we used the WorldValues Surveys, corruption, conductedbetween 1995 and 1997 in 50 countries (Inglehartet al. 2003). The following four questions were relevant: Question 1. Generallyspeaking, would you say (1) that this country is run by a few big

The perceptions the domesticpublicconof the extent of corruption(WorldValues cerning of.85 have correlation a coefficient 1995-97) Survey, withtheCCIand.86 withtheCPI.3
8

House'scivil libertiesscore or 9 The Freedom is because scoreof freedom combined rating notused extreme has thechecklist civilliberties included for in indifference its 14eleand corruption government mentssince 1984.

AND INEQUALITY CORRUPTION 143

power, and (5) the entities have self-determination or a high degree of autonomy.The original scores were converted such that a higher score representsmore freedom. For countries thatbecameindependent afterthe collapseof the Soviet Union and other former communist regimes,the politicalrightsscore for the former regimeswas appliedfor the periodbefore independence. Federalism.The sum of the following five indicators of federalism, averaged for 1975-1996, were calculated from the World Bank'sDatabaseof PoliticalInstitution(Keefer 2002): (1) the existenceof autonomous regions; (2) whethermunicipalgovernmentswere locally elected; (3) whether state/provincegovernments were locally elected; (4) whether the had overtaxing,spendstate/provinces authority or legislating; and (5) whether the coning, stituencies of the senators were the states/provinces. Religion. The percentages of Protestants, Catholics, and Muslims in 1980 were determined from La Portaet al. (1999). Legal origins. English CommonLaw (reference category), French Commercial Code, Socialist/Communist Laws, German Commercial Code, and Scandinavian Commercial Code were determined from La Portaet al. (1999). The Ethnolinguistic fractionalization. average value of ethnic fractionalization and linguistic fractionalization was determinedfromAlesina et al. (2003).
INSTRUMENTAL VARIABLES

To addressthe potential issue of simultaneous causation and the problem of measurement we variables. the one On error, used instrumental hand, because corruption also is likely to increase inequality,OLS may overestimatethe coefficient for inequality. On the other hand, measurement error in inequality may cause attenuation bias. Omitted-variables bias can be either positive or negative. Following Leigh (2003), we used "mature cohort size" relative to adult population as an instrumentalvariable for inequality. Higgins and Williamson (1999) theorized the effect of cohortsize on inequality. Because "fatcohorts" tend to get low rewards,when these fat cohorts lie at the top of the age-earningscurve,earnings

inequalityis reduced.When the fat cohorts are old or young adults, earningsinequalityis augmented. They show that the relative size of the cohort 40 to 59 years of age is a powerfulpredictor of inequality,both across countries and within the United States. Indeed,maturecohort size (ratioof the population 40 to 59 years old to the population 15 to 69 years old, averaged for 1971-1996, calculated from the United Nations (2000) population data)is a powerfulpredictorof inequality (see TableA3 on ourASR appendixonline supplement). It is reasonable to believe that this indicatorwill not directlyinfluence or be influenced by the level of corruption other than throughits effect on inequality,when controlis used for othervariables.10 high correlation The = -.72) between the instrument (mature (r cohort size) and the endogenous variable (inequality) as well as the presumably very weak correlation, if any, between the instrument and the errorterm of the regression likely minimizes the bias for the 2SLS estimator. We also consideredthe endogeneityof other variables.Economic development,tradeopenness, and democracyalso may be influencedby corruption.Democracy (political rights score) also is likely to sufferfroma largemeasurement whereaseconomicdevelopment error, (percapita income) and trade openness (exports plus imports over GDP) are likely to be more precisely measured. To obtain an unbiased estimate for the inequalitycoefficient,we had only to controlfor other variables that may have been correlated with maturecohortsize. We did not need instruments for otherendogenousvariables.However, we attemptedto instrumentother endogenous variables as well to get better unbiased estimates for them andto comparethe resultsbased on different sets of instrumentsas robustness checks. We used reasonablygood instruments

10Doubtaboutthe authors' instrument couldbe raisedby arguing the mature that cohort(individuals 40 to 59 yearsof age) mayhavemoreopportuand to nityforcorruption be moreprone corruption. ValuesSurveys datashowthat the However, World the maturecohortis slightlyless but not signifiand their cantly likelytojustifybribe taking, that perare to about ceptions corruption similar thoseof the (Table 4). remaining population

x44

AMERICAN REVIEW SOCIOLOGICAL

for economic developmentandtradeopenness, butwereunableto find a suitableone fordemocracy. (2000), we used distance FollowingTreisman from the equator(absolutevalue of latitude)as an instrumentalvariable for economic development(per capitaincome). Latitudeis known to be associatedsignificantlywith levels of economic development, probably through the of unevendistribution climates(tropicalvs temperate climates) and diseases (Gallup, Sachs and Mellinger 1999; McArthur and Sachs 2001). Because there is no reasonwhy latitude is directly correlatedwith corruption,latitude is a potentiallyuseful instrumentfor economic development."1 that Gallup and Sachs (2000) demonstrated malariaprevalence is a strong determinantof economic development,but thatmalariais very geographically specific and little affected by economic development. They presented evidence showingthatthe ecologic conditionssupporting the more efficient malaria mosquito vectorsprimarily determinethe distribution and we Therefore, used the intensityof the disease.12 index of malaria prevalence (the fraction of land area subject to malariatimes the fraction of falciparummalariacases, averagedfor 1966 and 1994; from Gallup and Sachs 2000) as anotherinstrumentfor economic development. We confirmed that both latitude (absolute value) and malariaprevalencehave very strong predictivepower (TableA3 on ourASRappendix online supplement).The simple correlation between latitudeand per capita income is .54, and that between the malaria index and per capita income is -.63. Together,they are highly significant for economic development,with other covariates held as equal. Although the exact causal relationship between geography

and economic developmentstill is being debated, these two variableswere reasonablygood instrumentsfor our purposes.13 As an instrument tradeopenness,we used for "constructedopenness" (naturallogarithmof predicted trade shares from a bilateral trade comequationwith "pure" geographyvariables, Rodriket al. 2004), following Frankel putedby and Romer (1999). Because geographyshould not inherently be correlatedwith corruption, this is a valid instrument. statistiOurinstruments satisfiedthe required cal propertiesvery well. The F statisticsfor the are null hypothesisthatthe instruments partially uncorrelated(when control is used for exogewithinequality, capita nousvariables) income, per and opennesswere sufficientlyhigh (all greater than 10), and the values ofR2 were sufficiently large (between 0.42 and 0.73; TableA3 on our We ASRappendixonline supplement). also conductedoveridentification whenever tests possible. The reported values for the overidentification p tests generally were sufficiently large for not that rejectingthe nullhypothesis ourinstruments withthe errortermof the corwereuncorrelated ruptionequation(also see TablesA7-A9 on our ASRappendixonline supplement). RESULTSAND INTERPRETATIONS First,we demonstratethe utility of using multiple measures for (freedom from) corruption and data averagedover many years for corrupWe tion and inequality. thenreportandinterpret our results.
DATAANDDIFFERENT THE USE OFAVERAGED

OF MEASURES CORRUPTION Table 1 presentsthe OLS regressionresults,in which single-yeardataand averageddatawere

" Acemoglu, and Johnson, Robinson (2001)and and that Subramanian, Trebbi Rodrik, (2002)argued not the of determines quality institutions, geography, McArthur levels of economicdevelopment. and that Sachs(2001)argued bothgeography instiand tutions matter. 12Acemogluet al. (2001) arguedthat malaria is and prevalence endogenous, thatit is the poorer that have been countrieswith worse institutions the unable eradicate malaria. is beyond scopeof It to this articleto examinethe conflictingarguments of about prevalence malaria. the

or thatlatitude malaria 13Evenif the possibility with is prevalence correlated the errortermof the ? it (i.e., regression Corr[z,u] 0) is considered, is not is Corr(z,x) substantially likelyto be large.Because is that largeandthe authors suspect Corr(x,u) quite unbiased is theIV estimator likelyto bebetter large, variable The thanthe OLS estimator. endogenous income)is denoted X, theinstrumenby (percapita the and talvariables index) z, and by (latitude malaria error term of the corruption regression by u 2002). (Wooldridge

AND INEQUALITY CORRUPTION 145 Table 1. OLSRegressionResultsfor Single-year Data versusAveraged Dependent

Variable

CPI98 Data

CPI96-02Data
1995 -1.377 (1.039) -0.071 Average -2.210* (1.120) -0.108

CCI98 Data
1995 -1.023* (0.467) -0.113 Average

CCI 96-02 Data

1995 Average

1995 Average Gini b -0.558 -1.593 (SE) (1.363) (1.540) B -0.028 -0.077 In GDPpc b 0.953*** 0.736*** (SE) (0.150) (0.197) B 0.641 0.461 PoliticalRights 0.041 b 0.301" (SE) (0.089) (0.137) B 0.033 0.236 InOpenness b 0.310 0.696* (SE) (0.228) (0.284) B 0.068 0.159 %Protestant 1980 b 3.356*** 2.893*** (SE) (0.562) (0.583) B 0.349 0.283 Constant b -4.735** -5.064** (SE) (1.570) (1.898) N 72 83 0.753 0.752 R2

Independent Variables

-1.642*** -0.916 * -1.500 ** (0.447) (0.439) (0.463) -0.172 -0.107 -0.167

0.867*** 0.659*** 0.403*** 0.313*** 0.376*** 0.287*** (0.047) (0.060) (0.120) (0.154) (0.049) (0.062) 0.629 0.442 0.632 0.451 0.623 0.438 0.097 (0.069) 0.081 0.210 (0.203) 0.047 0.336** (0.104) 0.278 0.603* (0.255) 0.143 0.067* (0.028) 0.127 -0.081 (0.096) -0.039 0.176*** 0.078** 0.174*** (0.027) (0.040) (0.043) 0.315 0.157 0.332 0.029 (0.114) 0.015 -0.025 (0.083) -0.013 0.088 (0.100) 0.047

3.067*** 2.795*** 1.129*** 0.945*** 0.946*** 0.822*** (0.227) (0.233) (0.496) (0.512) (0.233) (0.238) 0.185 0.211 0.308 0.265 0.238 0.171 -3.546* -4.019** -2.651*** -2.530*** -2.791*** -2.638*** (0.536) (0.562) (1.362) (1.509) (0.606) (0.626) 128 91 102 109 109 129 0.775 0.769 0.779 0.768 0.748 0.749

Note: ForGini,In GDPpc, andInOpennes,averaged dataare for the periodof 1971-1996, andfor Political the thereis no averaged data.Throughout tables, Protestant, Rightsit is for 1972-1996. Forpercentage standard errorsarepresented parentheses, standardized in and coefficients(B) aregiven heterokedasticity-robust coefficients.The levels of statistical togetherwith regression significancearedenotedas follows, unlessindicated otherwise.CCI= Controlof Corruption Index;GDP= gross domestic Index;CPI= Corruption Perceptions OLS= ordinary least squares; = per capita. product; pc *p <.05, **p < .01, ***p <.001

used for the dependent and independentvariables. TransparencyInternational's CPI and WorldBankInstitute's CCIfor 1998 or the average for 1996-2002 was used in each regresthe sion as the measureof corruption, dependent variablein this analysis. Inequality(GINI), per capita income (naturallog of GDP per capita), the political rights score, and trade openness (naturallog of percentageimportsplus exports over GDP) for 1995 or their averagevalues for 1971-1996 (1972-1996 for political rights)14

14 Note that Freedom House began to publish countryratingsin 1972.

were used as explanatoryvariables,with control used for the percentageProtestants 1980. in For the same dependent variable (measure of corruption), the estimated coefficients for inequality and political rights always become larger in magnitude and more significant, whereas those for per capita income always decrease when averageddata of the independent variables are used instead of single-year data. For CPI 1998, the magnitudeof the estimated standardizedcoefficients for inequality increased from .03 to .08, and that for political rights increased from .03 to .24, whereas the estimate for per capita income decreasedfrom .64 to .46 as we switched the independentvariable measures from single-year to averaged

x46

SOCIOLOGICAL REVIEW AMERICAN

data.This resultsuggeststhatOLS estimatesfor single-yeardataarebiased upwardfor per capita income, and towardzero for inequalityand political rights, because the lattertwo contain largermeasurementerror. Table 1 also shows thatthe estimatedcoeffivaricients for inequalityandotherexplanatory ables vary dependingon whetherCPI or CCIis used as the dependent variable. Inequality is insignificant when the single-yearCPI is used, whereasit is significantwhen eitherthe singleyear or averaged CCI is used. This suggests thatmeasurementerrorsin CPI or CCI or both are correlatedwith inequality and other independentvariables.Thus the measures of (freedom from) corruptioncontainsubstantialerror and also may be systemically biased. Whenthe averageddataareused for boththe variables,however, dependentand independent is significant regardlessof the corinequality ruptionmeasure,with controlused for per capita income, political rights,tradeopenness, and Protestantism.The standardized coefficients for inequality (GINI 1971-1996) are -.11 for CPI 1996-2002 and -.17 for CCI 1996-2002. Thus, averaginghelps to reduce measurement errors,althoughit may not solve the problemof systemic bias in corruptionmeasures.

The countries. measuresof the fourkey explanatory variables (inequality,per capita income, political rights, and openness) all are averaged for the period 1971(1972) to 1996. A striking contrast is evident between the bivariateregression (OLS 1) and the simplest multiple regression (OLS 2). The simple correlationbetween inequality(GINI) andcontrol of corruption(CCI) is -.39 and highly significant. However,afterthe inclusion of per capita income, inequality is insignificant, with the standardizedcoefficient of -.06, whereas per capita income is highly significant, with the coefficient of .79. Is it economic standardized developmentandnot inequalitythatmattersfor corruption? However, the situation is reversed as more controls are introduced. Including either democracy (political rights) or socialist legal origin contributes to a magnification of the coefficient for inequality(from -.06 to -. 13 or -.23) and to a decrease of that for per capita income (OLS 3 and 4). We see that democracy is highly correlated positively with both per capita income and (freedom from) corruption, and somewhat negatively correlated with inequality (see Table A2, on our ASR appendix online supplement). Hence, omission of this variablecauses substantialupward bias for per capita income and some attenuaOF ON THEINFLUENCE INEQUALITY tion bias for inequality. CORRUPnON Countrieswith socialist legal origins have a We tested our hypotheseswith a sample of 129 of significantlymore equaldistribution income countriesfor which both GINI 1971-1996 and than others, but are significantly more corrupt CCI 1996-2002 were available.Forrobustness on the average (TableA4 on our ASR appenchecks, we comparedthe regressionresultsfor dix online supplement).The omitted-variables CCI with those based on differentmeasuresof bias from a lack of control for legal origins corruption (CPI and ICRG average for considerably reduces the magnitude of the coefficient for inequality. The results from 1996-2002).15 The tables of the results for the CPI and ICRG index of corruption are prerunning the same regression as OLS 2 sepasented in the Appendix on our ASR appendix rately for countries with a socialist and those online supplement. with a nonsocialist legal origin demonstrate Table2 presentsthe OLS regressionresults that greater inequality is significantly associof various specifications for a sample of 129 ated with higher corruptionwithin both sets of countries (Table A5 on our ASR appendix online supplement). Thus, failure to consider different conditions between socialist and 15 number countries of covered varies The depend- nonsocialist legal origins obscuresthe effect of measureused. The current ing on the corruption inequality on corruption. 102 for included countries CPIand110counsample The OLS 5 controls for openness, with the for triesforICRD. Running regressions CCI Protestantism, legal origins, federalism, ethor thesample 102countries 110countries of produced estimatessimilarto those for the sampleof 129 counnoliguistic fractionalization, and natural tries. resource abundance, as well as for per capita

AND INEQUALITY CORRUPTION x47 Table2. OLS RegressionResultsfor Controlof Corruption Models (CCI96-02) forVarious OLS Models (1) Gini 71-96 b (SE) B InGDPpc71-96 b (SE) B PoliticalRights72-96 b (SE)
B

(2)

(3)

(4)

(5)

(6)

(7)

(8)

3.528***-0.569 -1.171* (0.782) (0.527) (0.460) -0.063 -0.130 -0.392

-2.072*** -2.964***-2.114*** 2.146*** 1.869*** (0.624) (0.684) (0.514) (0.582) -0.585 -0.230 -0.240 -0.210 -0.230 -0.304

0.513*** 0.329*** 0.454*** 0.337*** 0.320*** 0.284*** 0.301*** (0.039) (0.057) (0.038) (0.074) (0.068) (0.068) (0.068) 0.490 0.785 0.518 0.430 0.460 0.503 0.695
-

0.188*** (0.041)
0.358

-0.004 (0.064)
-0.007

0.035 (0.052)
0.070

0.347**-0.012 (0.122) (0.235)

0.660 -0.020 0.034*

PoliticalRightsSquared
b
---

(SE)
B

(0.017)
0.570 --0.774** -0.606*

Gini*(PoliticalRights 4)
b

(SE)
B

(0.301) (0.316)
-0.550 -0.721*** -0.937***-0.786*** -0.430 -0.610"* -0.686***

SocialistLegal Origin
b

(SE)
B -

(0.134)
-0.290

(0.265)

(0.203)

(0.206) (0.206) Yes Yes 129 129 0.814 0.808

OtherControls N
R2

No 129 0.153

No 129 0.662

No 129 0.722

No 129 0.723

Yes Yes 114 129 0.791 0.793

Note: OLS 5 through8 haveadditional Protestant controlssuch as tradeopenness,percentage French, population, and fractionalizaGerman, Scandinavian legal origins(Britishlegal originas referencecategory),ethnolinguistic but resourceabundance, federalism. coefficientsfor these controlvariablesarenot reported, and The tion, natural in those for OLS 6 arereported Table3. OLS 5 is basedon listwisedeletion,while OLS 6 through8 arebasedon of least squares. OLS = ordinary multipleimputation missingdata.GDP= gross domesticproduct; *p <.05, **p < .01, ***p < .001

income andpoliticalrights.16 Because 15 countries had missing values for either natural resourceabundanceor federalism,OLS 5 covered 114 countries only. To use the available informationas fully as possible andto maintain the sampleof 129 countries, usedthe method we of multiple imputationfor the missing data in OLS 6 (Allison 2002; King et al. 2001).17

16 Neitherthe percentageof Catholicsnor the percentage of Muslims was significant.These variables were droppedin the reportedregressions. 17 King et al.'s software Amelia (available at: was used for multiple http://GKing.Harvard.Edu) imputation.The same regressions were run for the five imputeddatasets, andthe resultswere combined

As OLS regressions 5 and 6 show,the coefficients for inequality all are statistically significant at the 1 percent level, and their magnitudeis substantial.A one standarddeviation reduction in income inequality (0.11 decreasein GINI)is associatedwith 0.30 (OLS deviationimprove5) or 0.23 (OLS 6) standard as measured by CCI. ment in corruption, Because we controlledfor the most significant variablesin previousstudies, omitted-variables bias is not a great concern.These estimates are arguablythe best that can be obtainedby using OLS regression methods.
to produce a single set of estimates for each regression.

SOCIOLOGICAL REVIEW x48 AMERICAN Runningthe same regressionsusing CPI and ICRGmeasuresof corruptionas the dependent variableproduces similar results. Inequalityis significant for CPI and ICRGat the 1 or 5 percent level using the same controls as for OLS 5 and 6. A one standarddeviationreductionin income inequality is associated with about a in deviationimprovement 0.21 or 0.26 standard CPI (or ICRG) (TableA6, OLS 2 and 6; on our ASR appendix online supplement). This finding is radicallydifferentfrom that of Husted(1999) and Paldam(2002). Theiruse of a single-yearCPI and a single-yearmeasure of inequalityprobablyproducedbiased results to from attenuationbias attributable measurement errorin inequality,inefficiency attributable to measurementerrorin CPI, and perhaps bias bias attribomitted-variables andadditional utable to systemic bias in CPI.18 coefficients We also see thatthe standardized for per capita income are the largest and highly significant, as previous studies have consistently found. Democracy (political rights) is significant without control for legal origins (Table 1 and Table 2, OLS 3), but generally insignificant with control for legal origins and othervariables (Table2, OLS 5 and6). However, when a quadratic term of political rights is included, it is significant (OLS 8), consistent with the findingof democracy's nonlineareffect (2002). Tradeopenby MontinolaandJackman ness generally is insignificant for CCI, but A6 stronglysignificantfor CPIandICRG(Table on our ASR appendixonline supplement). We see that Protestantism is significantly associated with freedomfrom corruption,irrespective of choice of corruptionmeasure, as previous studies have found. Ethnolinguistic is fractionalization significantlyassociatedwith higher levels of corruptionfor CCI, but not for CPI and ICRG(TableA6 on ourASRappendix online supplement). Federalism and natural resource abundanceare not significant. THE EFFECT INTERACTION OF AND DEMOCRACY INEQUALrTY Figure 1 shows thatcontrol of corruption(CCI 1996-2002) and income inequality (GINI 1971-1996) are highly correlated negatively with each other for more democraticcountries with a mean political rights score (1972-1996) of 4 or more,but the correlationis weakfor less democraticcountrieswith a score less than 4. This is consistent with our hypothesisthatthe effect of inequalityon corruptionis greaterin more democraticcountries. To test this hypothesis more rigorously,we included an interaction term of inequality (GINI) and democracy (difference from the mean political rights score of 4). Table2 shows thatthe estimatedcoefficient for the interaction term is highly significant and large in OLS 7. A one standard deviation improvement in inequality(0.11 reductionin GINI) is associated with -2.146*(-0.1 1) = 0.24 points, or a 0.24 standarddeviation improvementin control of corruption(CCI) at the mean political rights score of 4. But it increases in magnitude to (-2.146-0.774*3)*(-0.11) = 0.49 points, or 0.49 standarddeviation of CCI at the maximum political rights score of 7, whereas it decreasesas the country'spolitical rightsscore declines. The interactionterm still is significantwith the inclusion of a quadraticterm of democracy (OLS 8). Thus the adverse effect of inequality on corruption is higher in more democratic countries, even when the nonlinear effect of democracy is taken into account. The interaction term also is highly significant for CPI and ICRG (TableA6 on our ASR appendixonline supplement). Separateregressionsalso yield a highly significant and large coefficient estimate of inequality for the sample of more democratic countries,whereasthe estimatedcoefficientfor inequality is insignificant and small for the sample of less democraticcountries(TableA8 These on ourASRappendixonline supplement). results supportourhypothesisthatthe effect of is inequalityon corruption greaterin moredemocratic countries. THE EFFECT OF CAUSAL ON INEQUALITYCORRUPTION We establishedthat a significant partialcorrelationbetweeninequalityand corruption exists, with controlused for many otherplausiblevarivariables(IV) ables. We next used instrumental to providebetterevidence ofa causalinfluence

possiblethatCCIdoes andCPIdoes not bias. havesystemic

18 It is

AND INEQUALITY CORRUPTION x49

Less Democratic 3
2.5 2 0.5

More Democratic 3
2.5 2

o0 -0.5 10.5
-1--1.5

-0.5 vv

0.
-11

-1.5 GINI 1971-1996 GINI 1971-1996

in and of Countries. Figure 1. Inequality Control Corruption Less andMoreDemocratic = = Note:Less Democratic Political Rights(1972-1996) < 4. MoreDemocratic Political Rights(1972-1996) 2 4.

frominequalityto corruption to obtaineven and better unbiased estimates for the effect of inequalityon corruption. We began from the simplest IV regressions, in which only inequality(GINI) is instrumented: = + + + CCIi Po P1 GINIi (P2 Xi) Ei
+ GINIi = *yo 1 Maturei + (y2 Xi) + rli,

whereX denotescovariates, e is the random and errorterm. The OLS estimates for P13 be biased can becauseof omittedvariables, error measurement in inequality, and reverse causality from corA variruptionto inequality. good instrumental able can cure all these potential biases. We instrumented inequality with mature cohort size. If mature cohort size was not correlated with any otherindependent we variables, did not need to control for other variables. However,maturecohort size was correlated with bothper capitaincome andpoliticalrights. Hence, we presentIV estimationboth with and without controls to compare the results with those of OLS regressions.To make the results comparable with those of OLS for the same sample of 129 countries, we present the IV regression results based on the multiple imputation of missing datatogether. The instrumental-variables two-stage leastsquares (IV 2SLS) estimated coefficients for inequalityin Table3 are much largerthantheir correspondingOLS estimates in Table2, both with and without controls. Without controls, the magnitude the standardized of coefficientfor increases from 0.39 to 0.82 (OLS 1 inequality

vs IV 1). With the same set of controls as for OLS 6, the standardized coefficient for inequalincreasesfrom0.23 to 0.63 or 0.73 (IV 2 with ity list-wise deletion, or IV 3 with multiple imputation), and it is significant at the 1 percent level. This probablyis our best estimate of the causal effect that inequalityhas on corruption. The standardizedcoefficient for per capita income is smallerthanthat for inequalityin IV 2 and 3. Tradeopenness is significant only in IV 3. To obtain better estimates for per capita income, trade openness, and inequality, we instrumented these three endogenous variables with the fourinstruments maturecohortsize, of latitude, malariaindex, and constructedopenness in IV 4 and 5. We also introduced same the set of controls as for OLS 6, and used multiple for imputation the missing datain IV 5. The estimated standardizedcoefficient for inequality, with control used for other factors, is -.81 (IV 4 and5), whichis muchlargerin magnitude than its OLS counterpart of-.23. However,the standardized coefficient for per capita income is .31 or .29, which is much smallerthan its OLS counterpartof .49 (OLS 6). Inequality is significant at the 1 percent level, but per capita income and tradeopenness are not significant. In summary, the various IV 2SLS regressions give substantialevidence that inequality has a significant and large causal effect on corruption.A one standarddeviation reductionin inequality causes about a two-thirds standard deviation improvementin freedom from corruption(CCI), other factors being equal (IV 2, IV 3). The use of instrumentalvariables consistently increasesthe magnitudeof the coefficient for inequality and decreases that for per

150

REVIEW SOCIOLOGICAL AMERICAN

with OLS(6) Resultsfor CCI(1996-2002), Compared Table3. IV 2SLS Regression IV (1) IV (2) IV (3) IV (4) IV (5) OLS(6)

Gini 71-96 -7.419*** -6.163* -6.613* b (2.721) (2.415) (1.316) (SE) -0.633 -0.820 -0.730 B In GDPpc71-96 0.293*** 0.210* b (0.081) (0.098) (SE) 0.450 0.320 B InOpen71-96 0.267 0.181 b (0.129) (0.139) (SE) 0.093 0.140 B PoliticalRights72-96 -0.026 0.003 b (0.071) (0.064) (SE) -0.048 0.010 B 80 %Protestant 0.888* 1.415* b (0.640) (0.433) (SE) 0.188 0.290 B FrenchLegalOrigin -0.135 -0.135 b (0.163) (0.149) (SE) -0.067 -0.070 B SocialistLegalOrigin -1.399** -1.386** b (0.438) (0.461) (SE) -0.560 -0.534 B German LegalOrigin -0.357 -0.436 b (0.273) (0.269) (SE) -0.073 -0.090 B Scandinavian LegalOrigin -0.244 -0.679 b (0.594) (0.420) (SE) -0.045 -0.120 B BritishLegalOrigin(reference category) frac. Ethnoling. -0.132 -0.337 b (0.347) (0.293) (SE) -0.080 -0.033 B resourceabundance Natural -0.152 -0.206 b (0.203) (0.204) (SE) -0.071 -0.050 B Federal 0.025 0.022 b (0.044) (0.043) (SE) 0.040 0.043 B Constant 0.305 2.974*** -0.097 b (1.553) (1.268) (0.538) (SE) 114 129 127 N
p (overidentification test)
-

-8.186** -7.334*** -2.114*** (2.508) (0.514) (2.954) -0.230 -0.810 -0.806 0.197 (0.185) 0.307 0.269 (0.280) 0.123 0.007 (0.097) 0.013 1.109 (0.574) 0.227 -0.034 (0.229) -0.016 0.187 (0.188) 0.290 0.133 (0.247) 0.070 0.020 (0.089) 0.040 1.622* (0.669) 0.340 -0.122 (0.194) -0.060 0.320*** (0.068) 0.490 0.182 (0.108) 0.100 0.035 (0.052) 0.070 0.609 (0.356) 0.130 -0.219 (0.115) -0.110

-1.464** -1.407*** -0.786*** (0.388) (0.203) (0.526) -0.320 -0.570 -0.427 -0.336 (0.300) -0.071 -0.421 (0.489) -0.080 0.098 (0.426) 0.025 -0.211 (0.211) -0.073 0.014 (0.062) 0.024 0.824 (1.792) 98
0.284

-0.473 (0.292) -0.100 -0.898 (0.550) -0.160 -0.340 (0.322) -0.080 -0.127 (0.193) -0.040 -0.005 (0.055) -0.010 1.259 (1.725) 129
0.763

-0.184 (0.207) -0.040 0.149 (0.314) 0.030 -0.385 (0.212) -0.090 -0.248 (0.198) -0.080 0.034 (0.042) 0.060 -2.037*** (0.525) 129
-

cohortsize (1971-96). IV regresGini 3 Note: IV regressions1 through instrumented (1971-96) with Mature cohortsize, Gini sions 4 and 5 instrumented (71-96), InGDPpc(71-96), andInOpen(71-96) with Mature 3 Malaria index,andConstructed Latitude, openness.IV regressions and5 andOLS 6 arebasedon multiple GDP frac.= ethnolinguistic of fractionalization; = gross domesticproduct; imputation missingdata.Ethnoling.
OLS = ordinary least squares; pc = per capita.

* p < .05, ** p <.01, ***p <.001

AND INEQUALITY CORRUPTION I51

capita income. Although each of our instruments may not be perfect, all of them are not Thus, likely to be wrongin the same direction.19 the weightof the evidencesupports hypotheour ses that inequality increases corruption, and that large measurement error for inequality causes substantialattenuationbias in OLS.20 Instrumentalvariables 2SLS results using othermeasuresof corruption also areconsistent with our hypotheses. These estimates for inequalityarealwayslargerthancorresponding OLS estimates, regardless of the corruption measurebeing used,whereasthose forper capita income are smallerthan OLS estimates. We also find thatthe IV 2SLS estimatesfor inequality and per capita income differ depending on the choice of corruptionmeasures. The estimatedstandardized coefficient for inequalityis smallest for CPI:-.26 when only inequality is instrumented -.33 when threeendogenous and variablesare instrumented (TableA7, IV 2 and 4; on our ASR appendix online supplement). However,it still is statistically significant and OLS largerin magnitudethanits corresponding estimateof-.21 (TableA6, OLS 2; on ourASR variInstrumental appendixonlinesupplement). able regressionsof ICRGalso produceda larger standardized coefficient for inequality of -.43 (with inequalityonly instrumented; Table A7, IV 6, on ourASR appendixonline supplement) or -.57 (with threevariablesinstrumented;TableA7, IV 8, on ourASRappendixonline thantheirOLS counterpart of-.26 supplement) (TableA6, OLS 6, on ourASRappendixonline supplement). A substantialpart of the differencebetween these estimateswas attributable the difference to in the sample because CPI and ICRG were availablefor a smallernumberof countries,but we cannot rule out the possibility of systemic bias in corruptionmeasures. For example, if we runthe IV 2 in Table3 for the same sample of91 countriesfor which both CPI and CCI are available, the standardized coefficients for inequality are -.38 for CCI and -.26 for CPI.

Thus, the difference in the estimates is not as large as it seemed. measure,the By using CCI as our corruption and samplesize was increased selectionbias was minimized. Previous studies often were based on relatively fewer countries, which raises the question of externalvalidity.Although we cannot produce a single, reliable estimate of the causal effect that inequalityhas on corruption, we can confirm the existence of a statistically significant and substantivelyimportantcausal effect runningfrom inequalityto corruption. ROBUSTNESS CHECKS FURTHER We conducted multiple robustness checks in measadditionto the use of differentcorruption ures. First,we ran OLS and IV regressionsseparately for high- and low-income countries (TableA9 on ourASR appendixonline supplement). Both OLS and IV regressionsproduced significant coefficients for inequality for the sample of high-income countries,but insignificant coefficients for the low-income countries. Given the high correlationbetween per capita income and political rights, this is understandable. However, it raises the possibility that an interaction effect between inequality and per capita income exists, and that our results were biased from omission of this variable. When we included both interactionterms, the interactionof inequalitywith democracyoutweighed that with economic development,and only the formerwas significant (TableA10 on ourASR appendix online supplement). Second,we ranOLS andIV regressions,with control for region dummies (TableAl on our ASR appendix online supplement). Because Africa high levels of inequalityin sub-Saharan and LatinAmerica may account for higher levin els of corruption these countries,regiondummies may weaken the effect of inequality. However, inequality generally was significant even within regions.All these tests demonstrate that our findings are robust.
AND PERCEPTIONS CORRUPTION OF NORMS

test. way,theycanpasstheoveridentification 20Also, the magnitude reverse of from causality to economic is perhaps corruption development largthat erthan from to whichwill corruption inequality, causefurther bias income. upward forpercapita

19 If the instruments all wrongin the same are

Now that we have found substantialempirical supportfor a causal relationshipfrom inequality to corruption,it is importantto test empirically our hypothesis concerning the effect of inequality on norms and perceptions of corruption, using the WorldValues Surveys data.

I52

REVIEW AMERICAN SOCIOLOGICAL

of Table4. Predictors Normsand Perceptions aboutCorruption (OLSEstimates) Variable Dependent Variables Country-level Gini 71-96 PoliticalRights72-96 In GDPpc71-96 In Open71-96 Frenchlegal origin Socialistlegal origin German legal origin Scandinavian legal origin Individual-level Variables Mature(age 40-59 yr) Education Income class Subjective Unemployed Female Catholic Protestant Muslim (n) Respondents Countries (n)
R2

BribeJustified CheatTaxJustified Runby Big Interests Perceived Corruption B B B B 0.152*** -0.080*** 0.060*** 0.035*** 0.024** 0.052*** 0.017** 0.031*** -0.035*** 0.036*** 0.001 0.008 0.043*** -0.041*** -0.019** -0.019** -0.007* 41476 31 0.032 0.067*** -0.038** 0.150*** 0.083*** 0.026** 0. 184*** -0.006 0.034*** -0.041*** 0.052*** 0.009 0.016** 0.036*** -0.058*** -0.041*** -0.042*** -0.009* 41049 31 0.076 0.116*** 0.070*** -0.065*** 0.042*** -0.019 0.202*** 0.038*** -0.076*** -0.013* 0.043*** 0.048*** -0.077*** 0.007 0.028*** -0.072*** -0.034*** 0.020*** 36530 31 0.077 0.068*** -0.051*** -0.209*** -0.071*** 0.091*** 0.227*** 0.046*** -0.046*** -0.001 -0.016** 0.037*** -0.079*** 0.029*** 0.018*** -0.039*** -0.004 -0.021*** 40005 31 0.146

and ValuesSurveysandEuropean ValuesSurveys(1995-97). "Bribe Note: MicrodataarefromWorld justified" extent taxesjustified"takethe valuesof one to ten, "Runby big interests" zero or one, and"Perceived "Cheating of corruption" leastsquares; = percapita. fromone to four.GDP= gross domesticproduct; OLS = ordinary pc * p < .05, ** p < .01, *** p < .001

The sample covers more than 36,000 individuals in 31 countries. As the OLS regression results in Table 4 show, people in countrieswith higher inequality are more likely to perceive that the society is runby a few big interests,andthatmost public officials are corrupt.The people in these countries also tend to justify bribe taking and cheatingon taxes as acceptablebehavior,when such as income, eduindividualcharacteristics cation, and other macro factors are held constant. This evidence supports our hypothesis thatincome inequalityaffects people's perceptions concerning the extent of corruptionand habituatesnorms about corruption. Interestingly,individuals in countries with higherper capitaincome aremore likely tojustify bribetakingandcheatingon taxes, but their perceived extent of corruptionis lower on the average.Individualincome has no effect on the norms.Individuals moredemocratic in countries are less likely to justify bribe taking, whereas people in countries with a socialist origin are more likely to justify bribe taking and tend to

perceive that more public officials are corrupt. However,religionhas little impacton the norms and perceptionsabout corruption.21 Although women are slightly less likely to justify bribe taking and cheating on taxes than men, consistent with the finding of Swamy et al. (1999), the genderdifferencewas negligibly small. People ages 40 to 59 years (mature cohort) were slightly (almost negligibly) less likely to justify bribe taking or cheating on taxes, and showedno differencewith otherpeople in the perceivedextent of corruption. THE EFFCTOFCORRUPTIONINEQUALITY ON OurOLS regressionsconfirmthatcorruption is significantly associatedwith income inequality, consistentwith the previousfindings(Gupta et al. 2002; Li et al. 2000).Weregressed inequal-

21 Standardized coefficientsless than0.05 are

as regarded negligible.

AND INEQUALITY CORRUPTION 153

ity (average adjusted Gini for the 1990s) on two measures of perceived freedom from corruption (CCI average for 1996 and 1998, and CPI averagefor 1996-1999). Table 5 indicates that the OLS-estimated coefficients for CCI 1996-1998 are large and significantat the 1 percentlevel.A one standard deviation increase in CCI is associated with a .44 standard in deviation reduction inequality for a sampleof 114 countries,with controlused for per capita income, political rights, tradeopenness, Protestantism, legal origins, ethnolinguistic fractionalization, natural resource abundance,and federalism(OLS 3). Similarly, a one standard deviationincreasein CPIis associated with a .31 standarddeviationreduction in inequality,with the same controlsused for a sample of 77 countries(TableA12 on our ASR appendixonline supplement). However, these estimates may be biased because of reverse causation as well as measurement errorin corruption. Althoughwe experimented with various sets of instruments for corruption, we were not able to find a good candidate (see our ASR appendix online supplement). Thus, our results are inconclusive aboutthe causaleffect of corruption inequalon

ity. The coefficientfor corruption may have been overestimated becauseof reversecausation, but it also may have been underestimated because of measurementerror in corruption. Because these two sources of bias are likely to cancel out ratherthanmagnify,we suspect that the effect of corruptionon inequalityis in fact

significant.
Thus, there is evidence of reciprocalcausation betweeninequalityand corruption. Greater inequality causes higher levels of corruption, and higher levels of corruption intensify inequality.As a result, many societies are likely to be trappedin vicious circles of inequality and corruption.This mutuallyreinforcingrelationshippossibly explainswhy income inequality persists within countriesover time.

CONCLUSIONS AND IMPLICATIONS In summary, incomeinequality likelyto be a is and determinant of significant no lessimportant thaneconomicdevelopment (and corruption thusmanyother for The variables, thatmatter). effect of inequality likely to be greaterin is moredemocratic countries. Therealso is evidence suggestingthatinequalityfostersper-

Table 5. OLSRegressions Inequality of (GINI 1990s) Models OLS 1 B CCI96-98 InGDPpc71-96 PoliticalRights72-96 InOpen71-96 %Protestant 80 FrenchLegalOrigin SocialistLegalOrigin German LegalOrigin Scandinavian Origin BritishLegalOrigin(reference category) fraction Ethnolinguistic Natural resourceabundance Federal N
R2

OLS2 B -0.458*** 0.037 -0.142 0.013 0.292** 0.065 -0.563*** -0.119* -0.250** 0.077 -0.032 0.011 102 0.581

OLS 3 B -0.443*** -0.049 -0.147 0.042 0.407** 0.006 -0.550*** -0.148*** -0.316** -0.016 -0.034 0.004 114 0.523

-0.431*** -0.057 -0.137 0.044 0.402** 0.009 -0.540*** -0.141*** -0.311** 114 0.521

Note: Gini index(1990s), the dependent variable, rangesfrom0 and 100. OLS2 takesthe methodof listwise deletion.OLS 3 uses the whole sampleof countries, whichbothGINI(1990s) andCCI(96-98) areavailable, for for Index;GDP= gross domestic employingthe multipleimputation missingdata.CCI= Controlof Corruption OLS= ordinary least squares; = per capita. product; pc * p <.05, **p <.01, ***p < .001

154

SOCIOLOGICAL REVIEW AMERICAN

ceptions of widespread corruptionand correspondingly habituatesnorms of corruptionas "the way things are done."' Corruptionalso is likely to reproduceand accentuateexisting inequalities.Countriesmay thus be trappedin vicious circles of inequality and corruption,or liberatedin virtuouscircles of equality and integrity (freedom from corruption). This study identified the likely significant betweenincome inequalityandcorrelationship ruption. Furtherinvestigation of the relationships between other kinds of inequality (in wealth, education, political participation,and social opportunitiesas well as gender and ethnic inequality)andcorruption may be revealing as well. Our analysis also suggests that curmeasuresof (perrentlyavailablecross-country have systemic and ceived) corruption may ideological bias. More work is needed to minimize both bias and measurementerror. Our findings may contribute to an undersubjects. standingof threeadditionalimportant First, corruption is likely to be an important channel through which inequality adversely affects economic growth. Inequalityincreases which in turndetersinvestmentand corruption, growth. Although Alesina and Rodrik (1994) andPerssonandTabellini(1994) arguedthatthe adverseeffect of inequalityon economicgrowth to is attributable high ratesof taxationandredistribution, our results suggest an alternative explanation,with corruptionas a causal pathway. Second, our findings may help to explain why higher levels of marketincome inequality are not associated with higher levels of redisto tribution, contrary the predictionof the median voter theorem (Iversen and Soskice 2002; Meltzerand Richard1981). Inequalityincreases corruption,especially in democracies, and corruptionproducespolicy outcomes closer to those preferredby the rich than those favored by the medianvoter.Hence, taxationandredistribution in high-inequality societies will be lower thanpredictedby the median voter theorem. Thus, inequalitytends to persist without convergenceacross countriesover time. Third,the significant effect of inequalityon corruptionalso may help to explain why larger government size is not associated with a higherlevel of corruption.One recentpuzzling empirical finding was that smaller,not larger,

government size was associated with higher levels of corruption(Friedmanet al. 2000; La Portaet al. 1999), contradicting previousstudcan ies. Extensive redistribution both increase if size government andlowercorruption it effectively reduces inequality. The corruptionliteraturein recent decades has tended to focus on the corrupt and rentseeking behaviorof publicofficials. Whencorruptionis exclusivelyassociatedwith thepublic sector,the remedy is simple: if you wantto cut (Becker 1995). But corruption,cut government if corruptionis the resultof the rich attempting to preserve and advancetheir position, and if largergovernmentsize can be associatedwith less corruption, minimizingthe stateis not necthe appropriate policy response. essarily Our study thus stresses the need for considfor ering the motivationsand opportunities the rich and the privatesectorto engage in corruption (Glaeser et al. 2003; Hellman, Jones, and Kaufmann2000). The experience of massive privatizationaccompanied by enormous corruption in Eastern European countries offers further evidence (Black, Kraakman, and Tarassova 2000; Hellmanet al. 2000). We also note that the skyrocketingCEO compensation in the United States, which was supposed to align the interestsof CEOswith those of shareholders, not only increasedincome inequality, as but also stimulatedcorporatecorruption, the recent scandals demonstrate. The relationshipsof governmentsize, quality, and interventionwith corruptionneed to be studied further.Studies investigatingwhat kinds of governmentinterventionare more or less prone to differenttypes of corruptionand whatkinds of governmentaction arenecessary to controlcorruptionmay be fruitful.Although has much of literature stressedthe need to minand imize government regulations the discretion of public officials, it may be the kinds rather thanthe quantitiesof regulationand discretion for thataremorerelevant controlling corruption. Previousstudiesemphasizedthe role of economic developmentand religious and colonial traditionsin determininglevels of corruption. Given the persistence of cultures,one way out has of corruption seemedto be economicdevelis opment,but corruption knownto hindereconomic development. Thus, corruption has seemed to be destiny.

AND INEQUALITY CORRUPTION 155 However, redistribution may turn vicious circles into virtuous circles. Democracy (or political equality) is not sufficient to curb corruption without economic equality, and democratization in highly unequal societies may even generate increased corruption in the short run. One task of politics and public action is to shape institutions and social conditions so that people behave honestly because they believe that the basic structure of their society is just (Elster 1987). Corruption may not be destiny after all. You,Jong-Sung is a PhD Candidatein Public Policy and Doctoral Fellow oflnequality and Social Policy Program at Harvard University. His dissertation research is devoted to developing a theory of corruption as injustice and to conducting empirical studies on the causes of corruptionand the impact on ofinequalityand corruption social trustand redistribution. SanjeevKhagramis Visiting Professorat theInstitute for InternationalStudies at StanfordUniversityon sabbaticalfrom the KennedySchool of Government at Harvard University. His research focuses on transnationaldynamicsand architecturesof governance as well as on the political economy of (demand ocratic)governance,(sustainable)development, (human)security.His bookDams andDevelopment: TransnationalStruggles for Water and Power was published by Cornell UniversityPress in 2004, and he is the lead coeditor of the volume Restructuring World Politics: TransnationalSocial Movements, Networks andNormspublished by the Universityof Minnesota Press in 2002. REFERENCES Acemoglu, Daron, Simon Johnson, and James A. Robinson. 2001. "The Colonial Origins of Comparative Development: An Empirical Investigation." American Economic Review 91:1369-401. Ades, Alberto and Rafael Di Tella. 1999. "Rents, and American Economic Competition Corruption." Review 89:982-94. William Alesina,AlbertoE, ArnaudDevleeschauwer, Easterly, Sergio Kurlat, and Romain Wacziarg. 2003. "Fractionalization." Journal of Economic Growth8:155-94. Alesina, Alberto F. and Dani Rodrik. 1994. "Distributive Politics and Economic Growth." QuarterlyJournal ofEconomics 109:465-90. Allison, PaulD. 2002. MissingData. Thousand Oaks, CA: Sage Publications. Andvig, Jens. C., Odd-Helge Fjeldstad, Inge Amundsen,ToneSissener,andTinaSereide.2000. "Research on Corruption: A Policy Oriented Survey."Bergen/Oslo, Norway: Chr. Michelsen Institute (CMI) & Norwegian Institute of International Affairs (NUPI). Bergen/Oslo, Retrieved 2003 10, Norway. July (http://wwwuser.gwdg.de/-uwvw/downloads/ contribution07_andvig.pdf). Becker, Gary S. 1995. "If You Want to Cut Corruption, Cut Government." Business Week 3454:26. Black, Bernard, Reinier Kraakman, and Anna Tarassova. 2000. "Russian Privatization and Corporate Governance: What Went Wrong?" StanfordLaw Review 52:1731-808. Deininger, Klaus and Lyn Squire. 1996. "A New Data Set Measuring Income Inequality." World Bank Economic Review 10:565-91. Dollar, David, R. Fisman, and R. Gatti. 2001. "Are Women Really the 'FairerSex'? Corruptionand Women in Government."Journal of Economic Behavior and Organization46:423-29. is 2002. "Growth Good Dollar,DavidandAartKraay. for the Poor."Journal of Economic Growth 7: 195-225. Retrieved August 1, 2003 (http://econ.worldbank.org/view.php?type=-18&id= 22018). Dowse, RobertE. andJohnA. Hughes. 1986.Political Sociology. Chichester:Wiley. Elster, John. 1987. "The Possibility of Rational Politics." Archives Europeennes de Sociologie 28:67-103. Evans, Peterand James Rauch. 1999. "Bureaucracy and Growth:A Cross-National Analysis of the Effectsof 'Weberian' StateStructures Economic on Growth." American Sociological Review 64:748-65. Fisman, R. and R. Gatti. 2002. "Decentralization and Corruption: Evidence across Countries." Journal ofPublic Economics 83:325-45. Frankel, JeffreyandDavidRomer.1999. "DoesTrade Cause Growth?" American Economic Review 89:379-99. Friedman,Eric, Simon Johnson, Daniel Kaufmann, and Pablo Zoido-Lobaton. 2000. "Dodging the GrabbingHand:The Determinantsof Unofficial Activity in 69 Countries." Journal of Public Economics 76:459-93. Gallup, John L. and Jeffrey D. Sachs. 2000. "The Economic Burdenof Malaria." AmericanJournal of TropicalMedicine and Hygiene 64:85-96. Gallup, John L. and Jeffrey D. Sachs, with Andrew D. Mellinger. 1999. "Geographyand Economic Development." International Regional Science Review 22:179-232. Glaeser, Edward, Jose Scheinkman, and Andrei Shleifer. 2003. "The Injustice of Inequality." Journal ofMonetary Economics 50:199-222. HamidR. Davoodi,andRosaAlonsoGupta,Sanjeev, Terme. 2002. "Does CorruptionAffect Income

156

AMERICAN SOCIOLOGICAL REVIEW International.Retrieved Germany:Transparency July 15, 2003 (http://www.transparency.org/ working_papers/lambsdorff/lambsdorff eresearch.html). Lambsdorff, J. Graf. Forthcoming. "Measuring Corruption: The Validity and Precision of Subjective Indicators."Measuring Corruption, Brisbane, Australia: International Institute for Public Ethics. LaPalombara, Joseph. 1994. "Structural and Institutional Aspects of Corruption." Social ResearchLXI:325-50. Andrei La Porta,Rafael,FlorencioLopez-de-Silanes, Schleifer, and Robert W. Vishny. 1999. "The Quality of Government." Journal of Law, Economics and Organization15:222-79. Leigh, Andrew. 2003. Does Equality Lead to Dissertation Harvard University, Fraternity? Paper. Cambridge,MA. 1999.Does Mother Leite,CarlosandJensWeidmann. Nature Corrupt?NaturalResources, Corruption, and Economic Growth. WorkingPaperNo. 99/85. Washington,DC: InternationalMonetaryFund. Retrieved July 15, 2003 (http://www.imf.org/ external/pubs/ft/wp/1999/wp9985.pdf). Li, Hongyi, Lyn Squire, and Heng-fu Zou. 1998. "Explaining International and Intertemporal Variations in Income Inequality." Economic Journal 108:26-43. Li, Hongyi, Lixin C. Xu, and Heng-fu Zou. 2000. Income Distribution,and Growth." "Corruption, Economics and Politics 12:155-82. March, James G. and Johan P. Olson. 1989. Rediscovering Institutions: The Organizational Basis ofPolitics. New York:The Free Press. Mauro, Paolo. 1995. "Corruption and Growth." Journal of Economics 110:681-712. Quarterly McArthur, John W. and Jeffrey D. Sachs. 2001. Institutions and Geography: Comment on Acemoglu,Johnson and Robinson (2000). NBER WorkingPaper 8114. Cambridge,MA: National Bureau of Economic Research. RetrievedApril 15, 2004 (http://www.nber.org/papers/W8114). McCarthy, J. and M. Zald. 1977. "Resource Mobilization and Social Movements: A Partial Theory." American Sociological Review 82:1212-41. Meltzer, Allan H. and Scott F. Richard. 1981. "A Rational Theory of the Size of Government." Journal ofPolitical Economy 89:914-27. Miller, William L., Ase B. Grodeland,and Tatyana Y. Koshechkina.2002. "ValuesandNormsversus ExtortionandTemptation." 165-93 in Corrupt Pp. Exchanges:EmpiricalThemesin the Politics and the Political Economy of Corruption,edited by DonatellaDella Portaand SusanRose-Ackerman. Baden-Baden, Germany:Nomos Verlagsgesellschaft. Montinola, G. and R. Jackman.2002. "Sourcesof

Inequality and Poverty?" Economics of Governance3:23-45. Gylfason, Thorvaldur.2001. "Nature, Power, and Growth."Scottish Journal of Political Economy 48:558-88. Hellman, Joel S., Geraint Jones, and Daniel Kaufmann.2000. Seize the State, Seize the Day: State Capture, Corruption, and Influence in Transition. WorldBank Policy ResearchWorking PaperNo. 2444. Washington,DC: WorldBank. Higgins, Matthew,andJeffreyG. Williamson. 1999. Explaining Inequality the WorldRound: Cohort Size, Kuznets Curves, and Openness. NBER WorkingPaper 7224. Cambridge,MA: National Bureau of Economic Research. Retrieved April 1, 2004 (www.nber.org/papers/w7224). Husted, Bryan W. 1999. "Wealth, Culture, and Corruption."Journal of International Business Studies 30:339-60. ValuesSurveys Inglehart,Ronald,et al. 2003. World and European Values Surveys, 1981-1984, 1990-1993, and 1995-1997 [Computer file]. ICPSR version. Ann Arbor, MI: Inter-university Consortiumfor Politicaland Social Research[distributor]. Iversen,Torbenand David Soskice. 2002. Electoral Systemsand the Politics of Coalitions: WhySome DemocraciesRedistribute MorethanOthers.Paper presentedat the 2002 annualmeetingsof American Political Science Association. Retrieved August 1, 2003 (http://www.people.fas.harvard.edu/ iversen/Iversen-Soskice2003.pdf). Johnston, Michael. 1989. "Corruption,Inequality, and Change." Pp. 13-37 in Corruption, Development and Inequality, edited by P. Ward. London and New York:Routledge. Kaufmann,Daniel and Aart Kraay.2002. "Growth WithoutGovernance." World BankPolicyResearch WorkingPaper No. 2928. Retrieved August 1, 2003 (http://econ.worldbank.org/docs/l 199.pdf). Kaufmann,D., A. Kraay,and M. Mastruzzi. 2003. GovernanceMatters III: GovernanceIndicators for 1996-2002. World Bank Policy Research Working Paper 3106. Washington, DC: World Bank.Retrieved August 1,2004 (http://econ.worldbank.org/files/28791_wps3106.pdf). Keefer,Philip.2002. DP12000, DatabaseofPolitical Institutions: Changes and VariableDefinitions. RetrievedApril 15, 2004 (www.econ.usyd.edu.au/ download.php?id=2475). King, Gary, James Honaker, Anne Joseph, and Kenneth Scheve. 2001. "Analyzing Incomplete Political Science Data:An AlternativeAlgorithm for Multiple Imputation." American Political Science Review 95:49-69. Klitgaard, Robert. 1988. Controlling Corruption. Berkeley, CA: Universityof CaliforniaPress. Lambsdorff,J. Graf. 1999. Corruptionin Empirical Research: A Review. Working paper. Berlin,

AND INEQUALITY CORRUPTION i57 Corruption: A Cross-Country Study." British Journal ofPolitical Science 32:147-70. and Paldam,M. 2001. "Corruption Religion:Adding to the Economic Model."Kyklos54:383-413. . 2002. "The Cross-Country Pattern of Corruption:Economics, Culture,and the Seesaw Dynamics." European Journal of Political Economy 18:215-40. Persson, Torsten and Guido Tabellini. 1994. "Is Inequality Harmful for Growth?" American Economic Review 84:600-21. Political Risk Services. InternationalCountryRisk Guide. East Syracuse, NY: The PRS Group. Retrieved April 2, 2004 (http://www. icrgonline.com). VanandBeatriceWeder. 2001. Rijckeghem,Caroline "Bureaucratic Corruption and the Rate of Temptation:Do Low Wages in the Civil Service Affect Corruption,and by How Much?"Journal ofDevelopment Economics 65:307-31. and Rodrik, Dani, Arvind Subramanian, Francesco Trebbi.2004. "Institutions Rule: The Primacy of Institutions over Geography and Integration in Economic Development."Journal of Economic Growth9:131-65. A Rose-Ackerman,Susan. 1978. Corruption: Study in Political Economy. London, New York: Academic Press. 1. 999. Corruption and Government: Causes, Consequences and Reform.Cambridge: CambridgeUniversity Press. . 2001. "Trust Honestyin Post-Socialist and Societies."Kyklos54:415-44. for Sandholtz,W andW.Koetzle.2000. "Accounting Economic Structure, Democracy,and Corruption: 44:31-50. Trade." International StudiesQuarterly Swamy, A., S. Knack, Y. Lee, and 0. Azfar. 2001. "Gender Corruption." and JournalofDevelopment Economics 64:25-55. Tarrow,Sidney. 1994. Power in Movement:Social Movements, Collective Action, and Politics. Cambridge,UK: CambridgeUniversity Press. Torrez, Jimmy. 2002. "The Effect of Openness on Journal of International Tradeand Corruption." Economic Development 11:387-403. International. Perceptions Transparency Corruption Index (CPI). Berlin, Germany: Transparency International. Retrieved August 10, 2003 index.html#cpi and (www.transparency.org/cpi/ wwwuser.gwdg.de/-uwvw/corruption.cpi_2003. html). Treisman,Daniel. 2000. "TheCauses of Corruption: A Cross National Study." Journal of Public Economics 76:399-457. United Nations. 2000. Demographic Yearbook: Historicalsupplement 1948-1997 (computerfile). New York:United Nations Publications. Warburton,John. 2001. "Corruption as a Social Process:FromDyads to Networks." 221-37 in Pp. CorruptionandAnti-Corruption,edited by Peter Australia: LarmourandNick Wolanin.Canberra, Asia Pacific. Wooldridge,JeffreyM. 2002. EconometricAnalysis MA: of CrossSectionandPanel Data. Cambridge, The MIT Press. World Bank Institute. Governance Indicators: 1996-2002. Retrieved August 10, 2003 (www. worldbank.org/wbi/ governance/govdata2002).