CRIMINAL JUSTICE

Edited by.

J. Roland pennock, Swarthmore

and

College

John W. Chapman, [Jniaersity of pittsburgh

j1-:.:,:.:

.:

';:.:

-:1.:i

:.: _

'. :::::.:>.:,.t ..

.: '

New York and London.New york University press, lggS

11

ON THE ECONOMIC THEORY OF

CRIME ALVIN K. KLEVORICK

l INTRoDUCTION At the end of his interesting and important article, "Crime and Punishment: An Economic Approach," published in 1968, Gary
Becker characterized his efforts at developing an "economic" framework for analyzing illegal behavior "as a resurrection, modernization, and thereby I hope imrovement" on the Pioneering efforts of Cesare Beccaria and Jeremy Bentham during the eighteenth and nineteenth centuries.l Becker hoped, at least implicitly, to give new life and direction to cost-benefit analysis of public policy toward crime. In this effort he surely succeeded, as the frfteen yeais since his seminal article apThis chapter is a revised version of a paper prepared for presentation t the January 1983 meeting of he American Society for Political and Legl Philoso-

phy in Cincinnati. I would like to thank Bruce Ackerman, Guido calabresi, Owen Fiss, Joseph Goldstein, Geoffrey Hazard, Reinier Kraakman, Rick lvin, George Loewenstein, Jerry Mashaw, A. Mirchell Polinsky, Rob Prichard, Peter Schuck, Stan Wheele, Kenneth Wolpin, and the Legal Studies Seminar a! the Unive_ sity of Pennsylvania Law School for helpful discussions and for their comments on an earlier daft of this chapte, I am also grateftrl for the helpful research assistance ofJon Pedersen. In revising the paper, I benefited greatly fiom the discussion at the meeting and especially from the extremely thoughtful comments of [he commentators Jules Coleman and Stephen Schulhofer. These people, of course, bear no responsibility for any faults [ha[ emain.

289

290

ALVIN K. KLEVoRICK

peared have seen the development of a substantial literature on the economics of crime and Punishment Nevertheless, there is a Perception, among people in law and

economics, that except

for the highly controversial empirical on the deterrent effect of the death penalty, the econowork mists' literature on crime has not entered the mainstream of legal scholarship.2 This work seems to have remained a literature ihat is princially of interest to economists rather than to legal scholari, lawyers, and criminologists. In this resPect, work on the economics of crime stands in contrast to, lor example, the work economists have done on tort law, which has genuinely engaged the interest of torts scholars and teachers.3 In ihis chapte., I wish to suggest that the economists' literature on crime manifests an incompleteness and also a lack of connection among its various strands. Both of these features of the body of work identified as "the economic theory of crime" have restricted the impact that economists' contributions have had in this area. Moreover, these problems reflect an inherent limitation of any economic theory of crime: any such theory must depend on, and simultaneously be confrned by, a set of political and legal presuppositions. Furthermore, I believe tht an understanding of this inrinsic constrint on the economists' aPproach to crime will make possible a better appreciation of the kinds of contributions and the scope of the contributions economists can make in this area. Strengths sometimes appear more clearly when inherent limitations are recognized.
2. Tns SrnucrunE
oF THE

Ll.rsne.rur

I shall frame the discussion in a general description of the work that has been done on the economic theory of crime.a My intent is to describe the contours of the landscape or the topography of this vast literature rather than to Provide a thorough survey or critical review of it. This literature is comprised of three distinguishable strands. I shall discuss each of these strands in turn and draw out the relations among them as I proceed.
A. The Ind.iai.d'tnl's Decision About Criminl Actaity The first branch of this literature applies the microeconomic theory of chice under uncertainty to the decision process of

On tl Economic Theory of

Crime

291

the individual criminal. For the most part, this set of papers views the potential criminal as a rational actor who balances the costs and benefits, uncertain as they are, of his possible actions and allocates his time to legal and illegal activities accordingly.s Thcse models analyze the choice of whethei or not to commit a criminal act with its potential gain and its potential cost in the form

f punishment, where the probability of winning or loiing the gamble is determined by law-enforcement agencies' success in catching, convicting, and punishing criminal actors. The literatie also models the potential criminal's "labor allocation" between legitimate and illegitimate ctivities. The kinds of results that emerge take the standard form of the economist's comparative static conclusions about how changes in one or more prameters-for example, the probability and severity of punishment-affect the actor's choice. The perspeclive ernphasized by the models is tha the potential criminal responds, as others do, i a rational way to the incentives provided by the environment in which he operates. In some analytical studies of individual criminal behavior the authrs are attentive to the (perhaps) problematic nature of the assumptions about rationality and information concerning the risks presented by the criminaljustice system. For example, Philip Cook suggests bringng to bear the literature on boundr:d rationality6 and rules of thumb or "standing decisions" in analyses of crirnina behavior. Cook also treats in greater detail how the threats that the criminal justice system poses for the criminal are communicated to the actor through the media, the visible presence of enforcers (for example, police), and personal experience and observation.T Other analyses of the criminal's choice processs emphasize the importance of nonpecuniary aspects of the criminal's decision. They show, for example, the problems with postulating that monetary equivalents of some foms of punishment exist and the assumptions that one must make in deriving such equivalents. They also note that because ilegal activity is time-consuming, the criminal's decision must take into account effects on things other than wealth. These considerations lead these authors to focus on labor allocation models that take account of both pecuniary and nonpecuniary characteristics of the potential outcomes that the criminal faces. Fnally, some students of individual participation in criminal

292

aLvIN K KLEVORICK

market environactivity are attentive to the organizational or individual dicisions are made Particular ex-.i'li- which Rose-Ackerman's work on corruptione and i"i"a" s;san "pi. Reuter's work on organized crime l0 Pet'er

B.

of crime While some of the literature on the economic t^^eory fo.,l.., on cost-benefit calculations of individuals-the Poten;i"-;;i"ti;"" second strand is concerned with cost-benefit Becker's un"luti. .t, the social level. Research on the latter' with the most i rt*r. being the first formal treatmenl and still .r.o*irr.nt .*"rn"pl", views "optima policies to combat.illegal "' Llere Lehavior [as'| pari of an optimal allocation of resources results about it '""lt r .t.mic anaiysis are applied to derr9 law' I he Prlnthe socialy oPtimal ways to enforce the criminal of on the dJ-f.* orrr.tt pol.y analysis has been optimal choiceand the setype Lf pttittt."t and the "5J*-u"uiritv models' that the It it assumed, in these "iiTiit""i.tt*r,t.t t;..""'*".ti t of punishmerit can be set directly while the p,.inirhi.,ent is determined indirect by society's tt:."uilr." "r n police, courts, and other law enforcement re*p"nai,.it*

Criminl J ustice PolicY

,otirce..lt ""i.".t.,

precharcterization of an optimal set of policies his 1968 article' Becker a function to be optimized ln whose ",,ooor", u, t. iiriil;J -itti-un a general social loss functon aPpretbe costs ot arsuments were the damages from offenses'

i.it.tg "t convicting oiftndtrs, and the social.cost of punirh-.rr. For the mosi Part, however, his analysis proceeded ;iil;-;tt u.*.al formilation of society's objective .function' of

;il"ut,"B..ker

took the social goal to be minimizatio convtctlons' te total social loss in real income from ollenses' analysis'-though' he did ""i.rt*.t". At some poinu in the introiuce seneralizations-for example, an addrtlonal term.ln "a' ost 'Prtce th. lors fuction to recognize the socia loss due to are ir..i-i"t,io"' .tweenffenses that are not and those that that t4 That is, he recognized explicitly i"t.J-" o""i"ttment." i"."t an additional loss if two individuals who com"""i.i" -ti crime are, in the enil, treated differently' with one it;t ;. being punished while the other is not'

On ,he Economi.c Thedry of

Cri.me

293

Othe writers have generalized the social loss function in a 15 variety of ways. John Harris introduced a term for the social loss from wrongful punishment Then in their 1977 contribu16 tion, Roy Carr-Hill and Nicholas Stern suggested incorporating retributive considerations-specifically, departures from the "rtributionist target" punishment-in the term measuring the net social cost of punishment in Becker's social loss function. But in his 1978 paper, "On the Economic Theory of Policy Towards Crime," Stern withdrew the suggestion. He continued to argue for the importance of retribution as a determinant of p.*irh-.nt levels but indicated that it would not be "particuiarly helpful" to "amalgamate different criteria on nlodes of conuct into a grand social loss function."r? Mitchell Polinsky and Steven Shavell in specifying the social loss function in their 1979 article took explicit account of the attitudes toward risk of those who engaged in activities for which they might be apprehended and find.18 Rinally, Isaac Ehrlichre examined the implications of introducing three alternative elements into Beckr's social loss function. They were terms representing equality under the law, avoidance of legal error, and retribution' These features correspond to suggestions that others had previously made-including Becker himself, Harris, and Carr-Hill and Stern-but Ehrlich's analysis of the consequences of introducing them is more thorough and complete. ll oth.r. variants of the social loss function share one striking feature. The term ePresenting the nt cost ordamage to socity is, in each case, taken to be the difference between the harm to society and the social value of the gain to offenders. That is, the utility of the offender is counted in the social welfare function. Indeed, in two of the more elegant and sophisticated analyses in this literature, by Polinsky and Shaveli2o the social welfare function to be maximized is the sum of the expected utilities of the individual members of the society' This sems plausible in their 1979 article rvhere the analysis focuses on fines-both civil and criminal-and the heuristic example is double parking. The aggregation seems more^ questionable in their ltest contribution, "The Optimal Use of Fines and Imprisonment," where the harms involved are obviously severe enough for imprisonment to be contemplated as a punishment.

social t:* To be sure, lhis apParent anomaly in the Itl::i?] a paper sttmulateo Dy f,", U."r, noted befoi George Stigler' in

294

ALVIN K' KLEVORICK

il.k"tb"il
Becker

and Punishm--ent: n +ppt':n'-1i1 -Economic that worK' maoc ol closely in time upon the publication i.rito*inn the sam observation He wrote that:

introduces the "social value of the gan to of,h. offense The determination of this social ;;;t; to doubt its-use;"" ;;;;;pi"ined, and one is entitledevidence is there f;i;;,. u, ^n *pl"natoty concept: what from a'positive value uion th' utilitv derived il;;;;;t;" branded societv i'il;: 'ii'.] ' ""tl ln fact the as illicit has 2r such activities lr,.'i,Jrt'a.i"a from
The subsequent Iiterature
taken serious account'of though I ,h ;;,. it is an issue to which I shall return below'and losses whose gains il ii"."*r""it t !'"ttio" of exactly in the social welfare be counied ;;'i;;;;;;;,"J' 'noutd that is oPtimized' function ';;;;;. beiween the frrst two strands of th:-:S9"9Ttc of.individual-be.nutur' i.ri,,,. is a close one The models supplv of offenses bv indiilit;"ti.i;;;ictions about thefunctions are aggregated ro il;i. j ihJ. individual suplv known as deterrence ;;. ; ;;;, ,'''pptv r'."'ttiL-also of offensesato the Probnumber ir..i""-tt "t ..l.ie. he tottl deterrence function is lirii,t ""'.*.tity of punishment This used to oltain conclu;it ;;;;;i-h! poii'v modet that is illegal behavior-' itt"-.orr;al'strategies to combat "'"M";.-;;;;'B".ker's moel and the theorical work that fol-.mphasized the two-way inte-raction between l.;.; ;;"; enforcement ;;J;;i v ani the policies adopted-bv-^law at-the same the number of offenses ittJ htter affeit "ir,l"t. level and effects of criminal activity affect the choice ;*.;; ;. i"""i"i-i.i,tt. Policv instruments-the Probabilitvactivity. and 9.1':t* it' nf nrrnishment. ln technical terms' iriminal ai' sim"it"n-'o"slv d".ti1'"d' foi empirical efforts to esdmate the iltr;;;li.ations to #; il;iln, th"t is, the sponsiveness of offenses cnme Becuse of the interdependence between
h-as never

:T'i;i .it";;;;iJb..

*i'i'v

On the Econotnic Theory of

Crime

295

and punishment, simultaneous-equations estimating technigues ought to be used if one hopes to unravel the deterrence effects.
Existence of the Criminal Category The close ties and mutual enrichment that exist between the work on economic theories of individual criminal behavior and the research on economic theories of the criminaljustice system stand in sharp contrast to the lack of communication between the social cost-benefit analysis and the third strand in the economics literature on crime. This last enterprise, to which little effort has been devoted, draws upon economic analysis to explain the existence of the criminal category. It-maintains that conomic analysis can help to explain why we distinguish a set of acts that we call crimes-why we have the criminal law at all. The two principal (perhaps only) contributions to this Part of the literature on the economic theory of crime are by Guido Calabresi and Douglas Melamed in "Property Rules, Liability Rules, and Inalienability: One View of the Cathedral"22 and by Richard Posner in Econornic Anasis of Lau'23 Calabresi and Melamed use the framework developed in their article to consider the use of criminal sanctions in cases of theft and violations of bodily itegrity. They do not believe that problems of detection and apprehension, which rende_r the robability of apprehension less than one, explain fully why_we ;charge"za people who commit these acts more than the value of what they take. They argue instead that even if every such offender were caught, "the penalty we would wish to impose would be greater than the objective damages."25 The explanation, Calabresi and Melamed suggest, "lies in a consideration of the difference between property entitlements and liability entillements. For us to charge the thief with a penalty equal to -an objectively determined value of the property stolen would be to convert all property rule entitlements into liability rule entitlements."26 But they demonstrated earlier in their PaPer that economic efficiency considerations, distributional goals, and "other jrstice reasons" would lead a collectivity not to emPloy only liability rules to Protect entitlements. Instead, as Calabresi and Melamed persuasively argue, the state will employ a mixture of properiy rules, liability rules, and inalienability rules.2? Consequently, from the perspective of their analysis,

C. The

296

ALvIN K. KLEvoRICK

The thief not only harms the victim, he undermines rules and distinctions of significance beyond the specific case. . Since in the majority of cases we cannot be sure of the economic efficiency of the trnsfer by theft, we must add to each case an undefinable kicker which represents society's need to keep all property rules from being changed at will into liability rules.28
Of course, the ned for the "undefrnable kicker" applies as well to each vioation of bodily integrity. The basic distinction that Calabresi and Melamed draw between the criminal who robs or violates another person's bodily integrity and. the injurer in an automobile accident or a polluter in a nuisance case is that the criminal resorts to nonmarket transactions when the costs of market transactions are low, while the driver and the polluter find themselves engaged in nonmarket rather than market transactions when the latter are too costly. Posner's analysis is similar except for the singularly important fact that, within his scheme, the sole reason for p.referiing one fotm of transacjion to another-in Calabresi'Melamed tar-t, on" mode of entitlement Protection to another-is economic efficiency, as Posner defines it' For Posner, the only desideratum in social policy is value maximization: the distributional goas that form an integral part of the Calabresi-Melamed frmework, and the "otherjustice" reasons they discuss, receive no weight in his analysis. Foi Posner, "A 'crime' is simply an act that subiects the perpetrator to a distinctive form of punishment that is meted out in a distinctive kind of proceeding," where the unlawfulness of the act derives from "the policy of some other body of law-the rcrt law, in the case of 'the common law crimes' . . and various laws regulating business and personal behavior, in the case of. . statutory crimes."2e Under this approach, "the criminal sanction is simply a method of pricing conduct"3o and "The function of the criminal law-is to impose additional costs on unlawful conduct where the conventional damages remedy alone would be insufcient to limit that conduct to the efficient level."3l Posner, like Calabresi and Melamed, shows how a probability of punishment that is less than one should lead to a punishment that exceeds the damages caused by he criminal. But he

On the Economic Theory of

Crime

291

also makes the stronger point that although a sanction in which

he

expected punishment cost is equal to the damage to the victim is ;'adequate where the crime in question is simply the violation of a regulatory statute,"32 a heavier penalty is required in othe situations. Specifically, "if the law's purpose in prohibiting the act in question was to channel activity into the market, i.e.1he arena of voluntary transacting,"3s then an expected punishment in excess of the damage in8icted is needed to inuce the perpetrator to substitute a voluntary transac(ion for a coercive ne. When market alternatives are available, the expected punishment for a violation would be set equal to "the ium of (t the social costs of the violation and (2) the additional coits to the legal system of substituting coercive trnsactions for

market ransactions."3a Although Posner is not explicit on this Point, it is consistent with his igument that the "additional costs" under item 2 would include the Calabresi-Melamed "kicker." Of course, where there is no market alternative-Posner offers the example of "a driver who violates the speed limit because the opportunity costs of his time are very high"35-the penalty should be large enough to reflect the hazards the criminal creates, subsumed in his item 1, but there would be no add-on designed to induce substitution of a market for a legal transaction.
oF EcoNoMlc Expl-eNer:roNs o CnrMrNL CerBconv FoR THE EcoNoMlc THEoRY

3. Ivplrcerlo.ts

rns
or Cnrvr'er-

JusrtcE PoLrcY This third, relatively undeveloped, strand of work on the economic tleory of crime has severa important implications for research on the economic theory of policy toward crime. First, it suggests that economists' theoretical work on cost-benefit analyiis of public policy toward criminal activity is best viewed as a partial equilibrium approach. The insights offered by economic theories of the criminal justice system are conditional, for rhose'theories take as given the set of acts or activities that are designated as crimes. The results in what I have labelled the seco-nd strand of the literature on the economic theory of crime tell us how to allocate resources to apprehension and conviction, on the one hand, and to punishment, on the other, once

298
the crimes have been identifi

ALVIN

K KLEvoRtcK

the criminal justice system-a qt':- "eii'-:i.::"'.-l-i, travc peripective' lt might r-',". you will-would take a more gobal appropriate amalgam ol the osner's wealth measure, or an measure as tne or H,t-il;i;-td desiderata' f'.'ll yt anotherwould then vield .' ;; timized rhe "'sol.'tion" to be designated' as crimes the set of actions that are optrmally of punishment tor optimal Probability a" 'euetity ;i;;;

ed'-{ruil'J''iJiHT;i:::il,:i

t'lttltlp,t,

,ri'ti"i-l.t'

of ::; ;;'tt';.."nomic theorvc"n the criminal the problern i*agi"e solving Then o# il',ul"*. "t .ll). o"'.in::*f,:l"T'il:*': iii'"."r..' p.*a :.

R.ecker's measu.re of the supPose one accepted 'rul i.tcome 2s rhe relevant quantlty or coni"

categorv sPeaks

criminaUnoncriminal dtvtston the rurr or general.:t"t::::j;: acttvttrcs o-;rh,;;;;;i;; be the car-egorization ot rat yietded abilitY of Punishment/severtty^ social loss' function' Beckerian the minimum minimorum ot the htm "'ii.;;;'l'ut ut Becker and those who followed policv toward crime deltm ir;;1,";i;;-ttonomic models of or the. nroblemo ": their contribution' Nor ts tt have' In particul':'P':lt' ;i."i; ;;;;..' than thev ought t confit'es of is model But the himself was quite clear ^uouJittt does suggest difficulties for'one Dardal character of these models ot' cnme' Ti""i;il .;; o develop a complete economic theory of the broader problem' in ffir:::; determrnes tne r"n'iin partition 'b"i''t oPtimal which one "f ":l:Ii:T'i' severtty categoiies as well as the optimal

":'1,:l'1.,

;.;;.:

:"tr'*

ii;;;;:"p'.

,'l'ki

*'i:i

"t'""".tr-inal :: J*" ; i

not confined to economlc val

the explanation of, or

;y***3:;*
to""'otiottt'

M'la-'d would not ::"J-il;;';li, u*c"r^u'*i an of results about' the crimthe derivation

l":f;::''i

i' Then
..'

H'

Melamed and Posner

A second imPortant Potn

j,;i!::li:,:iiliY.;:,:'..iT',fi

that un tr'"-"'o"rce allocaiion problem that :;;;i.iil", i'uut ua""td' concerns the function of;::i.:;;;;" totitty to the number of relates the net cost or d"t";;;-;;

which explain the use of the

On the Economic Theory of Crime

299

are appropriatey reckoned in the analysis. In particular, the socially disapproved nature of the crimi_ nal's act, as Calabresi-Melamed and posner explain its distvored status, aises serious questions about the apropriateness of subtracting the criminal's gain from the harm tirt his act does to the rest. of_society. A theory or an explanation ofwhy certain acts are labelled criminal that provides reasons for objcting to the particular action may also provide reasons for dicorr.rfino the criminal's gain. Alternatively, since Becker was carefui though others have not been. to say that what ought to be sub_ tracted is the "social value of the gain to offende,,, the expla_ nation of the criminal category provides a means teindeed the explanation is required to-assess that,,social value." Once again, I do not wish to suggest rhat Becker was not clear about the scope of his analysis; on the contrary, he was precise in drawing the lines aound what he did. As he wrore,
Reasonable men will often differ on the amount of dam_ ages or benefits caused by different activities. . . These differences are basic to the development and implementa_

".ralyreJi, ct is classified as criminal precisely because the hm it does "r, to society exceeds the conventional measure of damages. To spec_ ify the social loss function in a Beckerian analysis, ne musi as_ sess the costs of these additional harms and te sure rhat they

thrust of the CalabresiMelamed and posner

fenses. This function, which is central to the overall objective function in Becker's and others, analyses of public polly to_ ward crime, is intended to measure the damag or nei cos that society sustains as the result of the offenses tht occur. But it is difficult to see how one can assess the social damages from of_ fenses unless one has a theory of the criminal caegory. The

hat

tion of_public policy but have been excluded from my inquiry. I assume consensus on damages and benefits and simply try to work out rules for an opiimal implementation of this consensus.36

His analysis is fine, then, as far as it goes. The difficulty, however, is that by sening aside issues of how social losses are assessed and by not havng a theory of the criminal category to inform that reckoning, ecker,s analysis is

.,

300
make subiect ro characterizations tht

ALVIN

KLEVORICK

it unattractive nd

seem-

;;il:;;i;;':ql-':l'li""'.:1,li,"li::1,'1i; rs
And the characterizatton

9l.""if uri rc. sp.cifr.atly, ne one that is Provided as a carrc ,,criminal activities are an imortant subset of the class writes, with the level of crimina i'r.rr"ti..'irt,t cause diseconomies'of offenses"'3? ln further ;:rt,;.;;;ted by the number says characterizing his analysis' he
of the economst's Our analysis of crime is a generalization Analvticallv' the harm o"r diseconomies ;;ili;;;;;"al costs of apprehenillJ#r",i-.;nsists in introducingorobabilitv ol apprethe mke irir "ia ."t,ior,, *hi.hlmportant ecsion variable' and and conviction an hension

ff '; ;ffi -"":

?r ods as well as bY monetary P

tiH;5i,3 ;ffi ill,

from anv othe

:: ::',itt'u"utvtitiiv

activir'v' that

*ht".t'irn-tt "t:,l""lrhhubl' oroduces external harm "n vanrstr' the analytical differences vrrtually iy fines,
is

effects'" But it The criminal's actions surelv have"external pursuing an or t'.nin"l dti;;;u""t punishmenr law vigorouslv charthat "..i-,. .ri_e and of the theoryielf-consciously external diseconliri'rlr.t'ri*ri " a b'at'th that theory of introducing costs bv ',tt.r';;; thut t*t'nd' and by taking account o-t-nonlor.n.n.io" and conviction *il find s-uch an approach "i.l."tt itts Legal ttrtiuttfundamental contributiorrof even less inviting if they tttuiiitt" externarrtres ;:i;il;;.;';; clarifving tt,e recirocal nature of the harm in "t pu"v cr ;*tr,il'; of th uit*-ihut an externalitY situation'5e origin of the criminal lfhe absence of explicit concern for the analvses of public Pgli:l:?*"lg:''1: "' 1982 article' impliill"".".,,ore, in Polinskv u" snuultt are drawn from ."i";r";; ilt of frnes ad impri'o"ment The'basic aci."b ;i;i ;odel of external istconomies and the tot;"tty l''1"'ih;:;s" external t'"- atitthe heartbenefrcial analvsts' ts of the :ii;r";;;ction, which is of all the members of the com: Jtit "t" utlititt inttt"t roiittttty u"a shavell derive are

;;;;';thei

ffit.;i;

,.i''tt'

On the Economic Theory of

Crime

301

teresting and important for policy settings where both monetary and nonmonetary penalties can be used to control an activity that yields social benefits but also results in harm to others. One can think of many such examples-analyses of public policy toward automobile accidents or pollution are two instances-but the economic literature on the existence of the criminal category suggests that policy about criminal activiy is not among them.

4. THE INHEREN: Lrvrrro o

Tueonv or Cnrv

Arv Ecoxovrc

The efforts to use economic analysis to explain why some acts are distinguished as crimes have yet anolher and more fundamenl implicaton about the kinds of insights economists, or the use of economic analysis, can provide about criminal law. The implication might be viewed by some as negative because it Points to an inherent limitation of the economist's approach. But it seems, to me at least, that recognizing this boundary is not in any way depreciatory. Rather, it is to recognize that although economics c indeed contribute to our understanding of crim-

inal behavior and optimal enforcement policy, it cannot provide a full or complete economic theory of crime. The work of economists in this area must be informed by work in other disciplines.

What, then, is this implication? It is that although an explanation of the criminal category can be stated' in economic terms, that vocabulary and mode of analysis does not, in fact, provide the substantive understanding we seek. To see this, it is helpful to frame more generally the answers offered by Calabresi and Melamed and by Posner. Society (or the collectivity or the state) establishes a "transaction structure" that stipulates the terms on which particular transactions or exchanges are to take place under different circumstances. For example, society might determine that if you and I are walking down the street, having no contact with one anorher, and you desire my watch, then you are permitted to have it if, but only if, we agree to a voluntary exchange of my watch for something you have (or promise to deliver, etc.). Society might also say that if the circumstances are different-for

302

aLvIN

KLEvoRIcK

rhe sideexamDle, as vou pedal your motorized bicycle down and my watch is dest'oyedlou are ;;k;"; kr,.k e over to "have" my watch but you must pay a collectrvely

o.r-,,.d
i"t
ual.re

for it tn Calabresi-Melamed terminology' the hrst srtwalc! u-"ion-.ott.tponds to protecting my entitlement t9 mf property rule, while in the second setting it is protected bv a i" r li"itiru rule. In Posner's terms, the former exchange'
".'1"

.*"..u,", would be a market transaction while the atter would be a nonmarket transaction' ' abr to-" possible exchanges, society might impose ancan-

you solute bar. In many societies (it is at least suted that) the sense that you cannot pay me a ceratn not buy my vote in a*o,rnt o'money in exchange for the capacity to cast my.ballot i"'r'or"*. It iany societis, an individua cannot sell himself ltaug. This iast form of regulating transactions corre-

i".

snonds to

Calabresi and Melanied refer to as Protectrng .i lnii,.-.* to my vote and my freedom from bondage with an inaienabilitY rule. The reason i introduce yet another layer of economics vo.^bh.v, with the term "transaction structure"'is to help us to rn"". Ui,V.. the Calabresi-Melamed and Posner characterizaimposition of .t J'.ti-.. For Calabresi and Melamed' the particular acts it . .ti-i"ut sanction and the characterization of ;;;;t-"t derives from a need to keep prppenty- rules-and inarule-s; iienability rules from being "changed at will" into.liability for Posner, it derives from a desire to induce individuals to for ."lriii"r" tof"",ury transactions with one another cost' coercive the former are possible at low enugh ones when --ot., ho*.u.r, that a soiety that imposes the criminal sancol votes tion on individuals who engage in the selling and buying a property into makes it a crime to change an inalienability rule in Posner's terms, to substitute a market for a nonmarrule or i..i,*^*.,i.t. one can also imagine a society in which it is.dea termined that all decisions about pollution should be made-rn to be tree trom collective setting, that is, in which all entitlements ;;il";"t "r. iotected by liability rules such a society might i"ell aoplv th criminal sanction to parties who engage ln- a "buys o1[" the pollu-".ke rrrsa.tion in which the polluter be applied to an emtees. Similarly, a criminal sanction might p.y.t *ft. hir", ,"o.k"t, at a (mutually. agreeable) wage below

ihat

On the Economic Theory of

Crime

303

the legislatively mandated minimum or to a seller who sells goods (to willing consumers) for more than or less than a egislatively set price. A society that would make each of these last two actions a crime would do so because a liability-rule protection had been converted into a property-rule protection. In each instance, the act that is characterized as a crime involves the actor(s) forcing society to deal with a transaction in a way in which society did not want to reat it What makes the act a crime is that the individual assaults the transaction structure that has been established by society.al One could say, alternatively, that the individual coerces societg into considering and coping with an exchange or transaction in a way that differs from th mode society had chosen. Finally; one might characterize the criminal as arrogating to himself the Power or the authority to

determine at least a part of the societal transaction structure, that is, appropriating to hlmself a power or right that society had reserved to itself. The criminal sanction is then a sanction to enforce the transaction structure hat society has chosen as well as to compensate for the harms to individuals within the
society.a2

But, then, an explanation of the criminal category-even an explanation stated in economc terms-requires answers to qustions and elaboration of concepts that economic analysis is not particularly well-suited to provide.a3 The point is not that the language describing the grounds for invoking the criminal sanction-for calling an act a crime-is not drawn from economics. One can, after all, describe the reasons in terms that are much less evocative and less clearly located in another discipline than are the terms "assault," "coercion," and "arrogatin of power." For example, one could simply say that the crimnal acts contrary to the transaction structure society has established.aa But the critical observation is that the explication of why some acts are crimes while other are not requires an inquiry into the legitimation of the transaction structure. It forces one to confront questions like: Why does the collectivity have the right to decide the terms on which particular transactions will take place under different circumstances? Why do some rights reside in the individual while others rest with the ltate? To be sure, one can ProPose a unified, and unitary, economic basis for determining what acts are categorized as crimes

304

ALVIN K' KLEvoRIcl(

of punishment-is::s-:::19 and what probability and severity one posits that wealth maxrmrzatron'If ili ;; ti-i*r

"'tt i';.;'.','il;i;;;ry:l::l:"ff l,fi ,i"';1rn uated, the objective lunctlon

"it". ..r"aljustice policv which se$ onlv with the tn p""'u'lti"' will be consistent i#:'il;;;;tn r -,hTL ll:,;i.:ii:r:fjff'ff i# as well' lndeed' determrnarlolr ur r"'i,ll]^l'L", l; ined ^c rr,l' l""*r,y .r punishment will be a subordinated 'p"" :t :l,t to the global Problem'

the probatrlr-

i:r

larser problem. ln the sotuuon

for market exchanges

?ie;"i"ii;;'h;'"":3:11J.."[ffi:i:::i'ff are suoJe

-it-will

'i

li;;';'.u;;il*.,;f _':1..:i:,:,'JJ5,,:iT':: minimize the reduction ln socli

orvnecessarilyinvokescertarnpoliticalp.es.,ppositions-towtt' structure that str'ives to max;tir"..b;;';;t a right to choose ado not have the right to restst i"i".."]itt ,;a ,"u i;vu'ls not answer the question: wby i'ii.rt""i. s".n a theory does legitimate? And'ul ;t"nsaction structure ::';;;; -1lY:' of att explanation of rc that question l, " t'ttt"u'ylo"'poent
v'ttu

a"-a acts "'i'"";fi ,;;;;"",..f e:::;';;itri:Hl*'.;:Xii:::; oorv. as I have couched tt tn a microeconomic analysts iiJi,it-.itett" undertaking certain minimum ot po:i#'t.q";t; asi precondition i
e

are treated as crimes'

litical and legal structure'

NOTES

l. Garv S. Becker, "Crime and Punishment: ' (1968): 169' 209' '.r'-," l'Pol Econ' 76 h' Henrv Hansmann"'The Current t. 5., ii xample' the tv6' lttol"ttip"'I L'E' 33 (1983): 217' sute of Law-and-Etot"'n't h11f 111 "' writings o'. 'Ji1" 3. ; ;^; to sav that economists' think about thei-subinfluence on ttt" *"y t"^'iui'i"* 'lhot"it all' for the o:tRo1t'tg':f iect. tt would have been A#tut' "tttt been ignored.Bur the tn"from-the wok-of '.:i;;;:.;r; ;;;k "n the subiect to have
sishts about crime that utJytt

An Economic Ap-

drawn and fragmented rh economrc ::H;;t-;;" been speciaiized anvthin! like the salience of the ;yil;i' * ""i "t"" " nt ee"n accepted as one of the approach to to't''r'itn
economic

n"""

On the Econotnic Theory of

Crirne

305

competing paradigms that torts scholars use in analyzing their subject, In torts, a scholar may reject the economic approach but must reckon with it. This is not the case with the economic theory

4. In keeping with the focus of this volume on theories of criminal

of crime.

justice, I restrict my discussion to economists' theoretical work on crime. The vast body of empirical work that has been informed by that theory and that has, in turn, influenced theoretical developments lies outside the purview of this chapter. An interesting set of empirical papers on the economics of crime is collected in J.M, Heineke, ed., Economic Modtk of Crim;nat Behatior (Arserdam: North-Holland, 1978), and a useful assessment of empirical work on the subject is provided by Philip J. Cook, "Research in Criminal Deterrence: Laying the Groundwork for the Second Decade," in Norval Morrs and Michael Tonry, eds., Crime and Justice: An Anmtal Revie of Rseh, vol. 2 (Chicago:

of Chicago Press, 1980) p. 2

l.

The University

5.

M. Heineke, "A Labor Theoretic Analysis of the Criminal Choice," Am. Econ. Reo. 65 (1915):314; lsaac Ehrlich, "Participation in Il-

See Becker, "Crime and Punishment," Michael K, Block andJohn

J. Pol. Econ. Sl (1973): 68; John M. Heineke, "Economic Models of Criminal Behavior: An Overview," in Economic Models of Criminal Behaaior, volume I. 6.
See Cook, "Research

legitimate Activities: A Theoretical and Empirical Investigation,"

in Criminal Deterrence," p. 220. For the classic discussion of bounded rationality, see Herbert A. Simon, Models of

7. PhilipJ. Cook, "A Unied Treatment of Deterrence, lncapacitation, and Rehabilitation: A Simulation Study" (Durham, N.C : lnstitute of Policy Sciences and Public Affairs, Duke University, 1979) 8. See Block and Heineke, "Labor Theoretic Analysis of the Criminal Choice;" Michael K. Block and Robert C. Lind, "Crime and Punishment Reconsidered," ,1. L. Stud. 4 (1915): 241; Michael K. Block and Robert C. Lind, "An Economic Analysis of Cimes Pu!ishable by Imprisonment," I. L. Stud. 4 (1975): 479; Heineke, "Economic Models of Criminal Behavior,"

Ma (New York: John Wiley, 1957).

9.

Susan Rose-Ackerman, Conul,ion: A Slui) ;n Polililal Econory (New

Tle Economics

York: Academic Press, 1978). 10. Peter Reuter, Disorganized Crime:

of the

Vi-sible

Hanl'

ll.
12.

(Cambridge: MIT Press, 1983). Becker, "Crime and Punishment," p.209.

A related interesting literature, which can be viewed

as an outgrorvth of the early socia cost-benefrt analysis of public policy toward crirne, concerns the comParative advantages of private ver-

306
sus Dublic enforcement

ALVIN K' KLEVoRICK

un George J. Stigler, "Law Enforcemenl' Malfeasance' -r'nfoicers," 7. l" sud . 3 (1974): l; William and Compensation f M. Landes and Richard A. Posner. "The Private Enlorcement ot "Private faw," . t. stud 4 (1975): l; and A Mitchell Polinsky' Fines,"rl Sr1 9 (1980): l0S Their ulrsuruUli. f.tfo.cement of its progeny' analyses rely heavily on Becker's seminal article and in orticulai, on th; results concerning the oPmal probability and ooiimal severitv of punishment. These analyses of the relative eli".v of ptuu,. an public enforcement apply to legal commands eenerallv and not to criminal law alone' rA. ?n ."utiiy, the type and severity of punishment are also set only indirectly. The may be legislative directives about sentences or element nes, for example, but ther is usually at least a residual in setting punishments' of iudicial discretion t4. Be-cker. "Crime and Punishment," p l84' Econ' iS. .onn n. Harris, "On the Economics of Law and Ordet"' l ' Pol 78 (1970): 165 t6. Roy e. arr-Hill and Nicholas H Stern, "Theory and Estimation in "fodels of Crime and Its Social Control and Their Relations to o"..fit of S..i"f Output," in Martin S Feldstein and Robert P Inman, eds., The Econimics of Public Semices 116 (London: Mac-

S. gl.te.

of the law The principal works are Gary

l?. 18.

millan, 1977).

ichoas H. Stern, "On the Economic Theory of Policy Towards in Heineke, ed., Economic ModeLs of Criminal Behaior 723'

.im",'
p.

Mitchell Polinsky and Steven Shavell, "The Optimal Tradeoff Between the ProbaLility and Magnitude of Fines," Am' Econ' Rett' 69 (1979): 880. lg, tsaac fnilich, "The optimum Enforcement of Law and the Conceft ofutti.e' A Positive Analysis," /t? R. ^U^.1cont Z.{t!lZ);1, and A Mitchell ZO. foii"tty and Shavell, "The Optimal Tradeoff"' "The Optimal Use of Fines and lmPolinsky and Steven Shavell,

i.

148.

Econ. 18 (191O1: 526, 527. LiaZZ, ,rio Catubre.i u.td A. Douglas Melamed, "Property Rules' Rules, and Inalienability: One View of the Cathedral"' Ham' bility ,: L e,t.88 (I9?2): 1089, l124-1127. fZS\ ni.truta A. Posnsr, Econom AnL1sis of Lu,2d ed (Boston: Little'

2t.

orisonment," I. Pub. Econ. (forthcoming)' eorg";. Sti!1e., "The Optimum Enfcement of Laws"'./ Pol'

U Brown, 19??), pp. 163-l?2 24. The ".ht.ge;; which Calabresi -*nu,"u.t

and Melamed refer includes m"onetary and nonmonetary penalties are imposed on the

On the Economic Theory of

Crime

307

criminal. There is no assumption that the payment, if it is a fine, is used to make restitution to the victim. 25, Calabresi and Melamed, "Property Rules, Liability Rules, and Inalienability," p. I 125. 26. Ibid. 27. I will discuss the substance of each of these rules in greater detail

28, Calabresi and Melamed, "Property Rules, Liability Rules, and Inalienability," p. I 126. 29. Posner, Ectnomic Anasis of Lau, p. 163. 30. Ibid., p. 172. 31. Ibid., pp. 163-4. 32. Ibid., p. 165. 33. tbid. 34. Ibid., p. I66. 35. rbid. 36. Becker, "Crime and Punishment," p. 209. 37. Ibid., p. 173. 38. lbid., p.201. 39. In his classic article about externalities, "The Problem of Social Cost,"/. . Ei Econ, 3 (1960): 1, Ronald Coase considers, in particular, "those actions of business frrms which have harmful effects on others . . [as] a factory the smoke from which has hamful
effects on those occupying neighbouring Properties:' begins the body of his argument as follows:

below.

(p l). He

Tne Recrpnocel Nlune or ,rxs PnoLsv

The taditional approach bas tended to obscure the natule of the choice that has to be made. The question is commonly thought of as one in which A inflicts harm on B and what has to be decided is: how should we restrain A? But this is wrong. We are dealing with a problem of a reciprocal nature. To avoid the harm to B would inflict harm on A, The real question that has to be decided is: should A be allowed to harm B or should B be allowed to harm A? The problem is to avoid the more serious harm . . .(p. 2).

40. Polinsky and Shavell, "The Optimal Tradeoff," p. 21. In addition,

"it is assumed that each individual is equally likely to be the victim of someone else's harm" (ibid.). This latter assumption, which is
of law and economics, too, for example, the analysis of aternative

made for reasons of analytical tractability, is common in other areas

308

ALVIN K. KLEVORICK is plausible Iiabilitv rules in torts. Whether ot not the assumPtion lor careful scruriny itr a model ana' it calls ;;;;;;.;;..ttext. policy lt is clear that all the esults in theirrirn f"*

ii^"tli

"tfo*.-ent s""." coitribtion are obust to some relaxation of "* for example' if different individuals can hve ;;;t;;;; ".' given robabilities ot suffering h3l- Ho.* iif.*ni*"g"totsly . ..tr,. -"i u. ltittd ii o"t *"nt further and had the probor not one engagd
ititu oi.urr..ing harm depend on wbether
aitiviry is another.matter' The concluit a hne or rmprrsonsions about 1l) ihe optimal punishment-be typ" f putti'h-ent is used and when thee "-*rt.i try ",,L to the class and (2)-the desirability of using frnes

t"'il: ;;-;E.idering
;. niu on" *"ut,t

Delng do not depend on the exogeneity o[ the probabrllty oI likelihood of being harmed rs endogeirrmed. Bt when the

-"*i*u-

feasible extettt before imposing nonmonetary Penalles'

i"tii.i*^..,

r.

turns the plcement of the entitlement itself violation of what Cala42. i;;" ;;;.. include the damages due to " ur.til."liaamed call moralims see Calabresi and Melamed' ;;;".;; Rules. Liabilitv Rules. and Inalienability"' pp llll-2 (see note riJ'.o.tt.urion is "pid" to society As noted before thete are that monetary Payments' lt 24). ther is no assumption ,nn. uae us"d to make restitution to victims' lighr on lhe cong. itrnists haue, for example, never shed much lhe term is otten lnvoKeo cepl of coercion, despite the fact that

some of Poli'sky and Shavell's statemens.abort ol ller results ror ,rrrdeadetara"naa aa,d overdeter ence and some modification' the two-class model may require way in ,1"".r.-or theft. the crininal not only challenges the overbut also which society has chosen to Plotect an entitlement

;;;"-;;;
itself.

;i,..orcion involved in.Sovernmental"intervention the marKet in the market while others speak of the coercion ol

a+. ^'

at .".ting issue, but one that I do not address here' is whethcr *,iue usefully described as acts that are contrary to the "i'.ti-.ttransaction structure For example' can one' withort chosen

45.

that tne srrairins language, describe the act of perjury by saylng with a transaction in a way that sop.,rroi rot. ciety to deat as cietv did not want to treat lt? If all acts that we characterize more.genercriinal cannot be described in this language---or' that tne uti.tg the vocabulary of economics-then it is clear "ttv, of crime is necessarily limted My economic theory ;;.-;i'.t .I*t" "ppfi.t only to thse crimes that the economic theory of the criminl category purPorts to explain' or'tni! i"t'."tiott' I assume guend'o that weakh

;;i;'ilto*.

On the Economic Theory of

Crirne

309

maximization can be given a clear, substantive, and commonly agreed upon defrnition. That providing such a denition is not a stiaightforward task is attested to by the pages of controversy about it in the larv-and-economics literatue. The point that follows ap-

plies if, for example, one takes as the criterion the maximization f the total value of all goods and services the society produces where quantities are valued at the equiibrium prices, or if one
substitutes utilitarianism for wealth maximization.