All the Variables in IS-LM Model - United States of America (1930-95)

Taposh Kumar Roy 4th year, Roll no: 119 Dept. of Economics University of Dhaka

Dwaipayan Chakma 4th year, Roll no: 162 Dept. of Economics University of Dhaka April, 2010

1|P a g e

Table of Contents
1. Introduction------------------------------------------------------------ 01 2. Methodology----------------------------------------------------------- 01 3. Concept about IS-LM Model----------------------------------------- 01
a. IS Curve-------------------------------------------------------------------------- 02 b. LM Curve------------------------------------------------------------------------ 02

4. Background ------------------------------------------------------------ 04
a. Causes of Great Depression---------------------------------------------------- 06 b. Could Depression Happen Again?-------------------------------------------- 07 c. The Depression: Issues and Ideas --------------------------------------------- 07

5. Sample Description and Estimation Strategy------------------------09

a. IV Estimation----------------------------------------------------------------------12 b. Regressions on IS-LM Equation---------------------------------------------- 15

6. Findings------------------------------------------------------------------18 7. Conclusion-------------------------------------------------------------- 18 8. References-------------------------------------------------------------- 20

2|P a g e

Of all the economic fluctuations in the world history, the one that stands out as particularly large, painful, and intellectually significant is the Great Depression of the 1930s. During this time, the United States and many other countries experienced massive unemployment and greatly reduced incomes. In the worst year 1933, one-fourth of the labor force was unemployed, and real GDP was 30 percent below its 1929 level. This devastating episode caused many economists to questions the validity of classical economic theory. Classical theory seemed incapable of explaining the Depression. After the onset of the Depression, many economists believed that a new model was needed to explain such a large and sudden economic downturn and to suggest government policies that might reduce the economic hardship so many people faced. In 1936 the British economists John Maynard Keynes revolutionized economics with his book The General Theory of Employment, Interest, and Money. Keynes proposed that low aggregate demand is responsible for the low income and high unemployment that characterize economic downturns. He criticized classical theory for assuming that aggregate supply alone-capital, labor, and technology- determines national income. The Keynesian theory of aggregate demand concerns IS-LM model, which shows what determines national income for any given price level.

Introduction
The economy of the United States is the biggest economy in the world and almost all the economy of the world is closely related to the US economy. So, it is hardly surprising that economists pay close attention to IS-LM variables in the United States, since it has great importance to know or predict different economic aspects. Most of this attention has been devoted to traditional sorts of economic concerns, including GDP, consumption, investment, net expenditure, disposable income etc. for various economic research and developments. For some reasons, this IS-LM variable is important. Firstly, it shows the relationship between national income and other variables. Secondly, it shows the short run and long run equilibrium condition between money markets and goods markets, and so on. The goal of our paper is to show the changing patterns of GDP and the related concepts or variables that affect the GDP to change over times. Besides, the IS-LM model or variables show what determines national income for any given price level and how national income changes when its determinants changes. There are two ways to view this exercise. We can view the ISLM model as showing what causes income to change in the short run when the price level is fixed. Or we can view the model as showing what causes the aggregate demand curve to shift.

Methodology
Our empirical analysis involves several methods. We will apply these methods successively for estimating the results more accurately. Firstly, we will give brief concepts about the theoretical IS-LM model, because it will help us to know what IS-LM variables are. Secondly, we will review the data on the basis of Great Depression (1929-1939) and the rationality for this is that it will help us to compare the trends of GDP between the Great Depression and after the depression period. Thirdly, we will use the statistical software package STATA for empirical estimation and sample description of the data through OLS and IV (2SLS) regressions. Finally, we will demonstrate the findings from the above empirical estimates.

3|P a g e

Concept about IS-LM Model

The IS-LM model was introduced in a classic article by the Nobel-Prize winning economists John R. Hicks. The two parts of the IS-LM model are the IS curve and LM curve. IS stands for investment and saving, and the IS curve represents whats going on in the market for goods and services. LM stands for liquidity and money, and LM represents whats happening to the supply and demand for money. The IS Curve One way to think about the IS curve is that it describes the combinations of income Y and the exchange rate e that satisfy an equation: Y = C(Y - T) + I(r) + G + NX (e) This equation combines the national income accounts identity, the consumption function, net export function and the investment functions. It states that the quantity of goods produced, Y, must equal the quantity of goods demanded C+ I + G +NX We can learn more about the IS curve by considering the special case in which the consumption function and investment function are linear. We begin with the national income accounts identity Y = C + I + G + NX Now suppose that the consumption function is C = a + b(Y - T), where a and b are numbers greater than zero and the parameter b is the marginal propensity to consume, so we expect b to be between zero and one. The investment function is I = c - dr, where c and d also are numbers greater than zero. The parameter d determines how much investment responds to the interest rate; because investment rises when the interest rate falls, there is a minus sign in front of d. The net export equation is NX = f(e) Where e is the exchange rate and it is negatively related with the net export (NX) schedule. From these three equations, we can derive an algebraic expression for the IS curve and see what influences the IS curves position and slope. If we substitute the consumption and investment functions into the national income accounts identity, we obtain Y = [a + b(Y - T)] + (c - dr) + G +NX (e) This equation expresses the IS curve algebraically. It tells us the level of income Y for any given interest rate r and fiscal policy G and T. Holding fiscal policy and interest rate fixed, the equation gives us a relationship between the exchange rate and the level of income: the higher the exchange rate, the lower the level of income. The IS curve graphs this equation for different values of Y and e given fixed values of G and T and r. The LM Curve The LM curve describes the combinations of income Y and the interest rate r that satisfy the money market equilibrium condition M/P = L(r, Y).

4|P a g e

This equation simply equates money supply and money demand. We can learn more about the LM curve by considering the case in which the money demand function is linearthat is, L(r, Y) = eY - fr, where e and f are numbers greater than zero. The value of e determines how much the demand for money rises when income rises.The value of f determines how much the demand for money falls when the interest rate rises. There is a minus sign in front of the interest rate term because money demand is inversely related to the interest rate. The equilibrium in the money market is now described by M/P = eY - fr. To see what this equation implies, rearrange the terms so that r is on the left hand side. We obtain r = (e/f) Y - (1/f) M/P. This equation gives us the interest rate that equilibrates the money market for any values of income and real money balances. The LM curve graphs this equation for different values of Y and r given a fixed value of M/P. From this last equation, we can verify some of our conclusions about the LM curve. First, because the coefficient of income is positive, the LM curve slopes upward: higher income requires a higher interest rate to equilibrate the money market. Second, because the coefficient of real money balances is negative, decreases in real balances shift the LM curve upward, and increases in real balances shift the LM curve downward. From the coefficient of income, e/f, we can see what determines whether the LM curve is steep or flat. If money demand is not very sensitive to the level of income, then e is small. In this case, only a small change in the interest rate is necessary to offset the small increase in money demand caused by a change in income: the LM curve is relatively flat. Similarly, if the quantity of money demanded is not very sensitive to the interest rate, then f is small. In this case, a shift in money demand caused by a change in income leads to a large change in the equilibrium interest rate: the LM curve is relatively steep. Thus we get the IS and LM equations in case of open economy is as follows: Y = C(Y - T) + I(r*) + G + NX (e) M/P = L(r*, Y). IS* LM*

5|P a g e

Exchange rate, e

IS*

LM*

e*

Y*

Income, output, Y

Figure-1: The IS-LM model

Background
The data for our analysis mainly comes mainly from the period of Great Depression in 1929 in USA. The Great Depression had its negative impacts on the economy, especially on the US economy, for roughly two or three decades. So, a basic concept about Great Depression should be gathered to analyze this data, because Great Depression is mainly related with IS-LM variables. The best-known fact about the great depression is the stock market crash. Between September 1929 and June 1932, the market fell by 85 percent, which means that stocks worth $1,000 at the stock market peak were worth only $150 at the bottom of the market in 1932. The depression and the stock market crash are popularly thought of as almost the same thing. In fact, the economy started turning down in August 1929, before the stock market crash, and continued falling until 1933.
Figure-2: GDP Growth During Great Depression

600 1930

700

GDP (Real) 800

900

1000

1932

1934 Year

1936

1938

1940

6|P a g e

Between 1929 and 1933, GNP fell by nearly 30 percent and the unemployment rate rose from 3 to 25 percent. Until early 1931, the economy was suffering from a very severe depression, but not one that was out of the range of the experience of the previous century.' It was in the period from early 1931 until Franklin Roosevelt became president in March 1933 that the depression became "great." More than anything else, the great depression is remembered for the mass unemployment that it brought. For the 10 years 1931 to 1940, the unemployment rate averaged 18.8 percent, ranging between a low of 14.3 percent in 1937 and a high of 24.9 percent in 1933.2 By contrast, the post-World War II high, reached in 1982, was under 11 percent. Investment collapsed in the great depression; indeed, net investment was negative from 1931 to 1935. The consumer price index fell nearly 25 percent from 1929 to 1933; the stock market fell 80 percent between September 1929 and March 1933. In the recovery, from 1933 to 1937, real GNP grew at a rapid annual rate of nearly 9 percent, but even that did not get the unemployment rate down to normal levels. Then, in 1937-1938 there was a major recession within the depression, pushing
Table-1: ECONOMIC STATISTIC OF THE GREAT DEPRESSION
year

1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939

GNP 1972 $billion (1) 314.7 285.2

263.3

1/GNP (%) (2) 17.8

13.5 9.0

G 1972 $billion (3) 40.9

44.6 46.2

Unemployment Rate % (4) 3.2

8.7 15.9

CPI 1929 =100 (5) 100.0

97.4 88.7

Commercial paper rate % (6) 5.9

3.6

AAA Rate, % (7) 4.7

4.6

Stock market index (8) 83.1

67.2

M1, 1929 100 (9) 100.0

96.2

Full employment Surplus/Y*% (10) -0.8 -1.4 -3.1 -0.9 1.6 0.2 -0.1 -1.1 1.8 0.6 -0.1

2.6 2.7 1.7 1.0 0.8 0.8 0.9 0.8 0.6

4.6 5.0 4.5 4.0 3.6 3.2 3.3 3.2 3.0

43.6 22.1 28.6 31.4 33.9 49.4 49.2 36.7 38.5

89.4 78.0 73.5 81.4 96.6 110.6 114.8 115.9 127.3

226.8 222.1 239.4 260.8 296.1 309.8 297.1 319.7

3.5 3.8 5.5 9.2 10.9 12.8 8.1 10.5

44.0 42.8 48.7 49.8 58.5 56.3 61.3 63.8

23.6 24.9 21.7 20.1 16.9 14.3 19.0 17.2

79.7 75.4 78.0 80.1 80.9 83.3 82.3 81.0

NOTE: Stock market index is Standard & Poor's composite index, which includes 500 stocks; September 1929 = 100. y* denotes full-employment output. SOURCES: Cols. 1, 2, 3: U.S. Department of Commerce, The National Income and Product Accounts of the United States, 19291974. Col. 4: Revised Bureau of Labor Statistics data taken from Michael Darby, "Three-and-a-Half Million U.S. Employees Have Been

7|P a g e

Mislaid: Or, an Explanation of Unemployment. 1934-1941," Journal of Political Economy, February 1976. Cols. 5,6,7: Economic Report of the President, 1957. Col. 8: Standard & Poor's Statistical Service, Security Price Index Record, 1978. Col. 9: Milton Friedman and Anna J. Schwartz, A Monetary History of the United States, 1867-1960 (Princeton, N.J.: Princeton University Press, 1963), table Al, col. 7. Col. 10: E. Cary Brown, "Fiscal Policy in the Thirties: A Re appraisal" American Economic Review, December 1956, table 1, cols. 3, 5, and 19.

the unemployment rate back up to nearly 20 percent. In the second half of the decade, short-term interest rates, such as the commercial paper rate, were near zero. Causes of Great Depression What caused the Great Depression, the worst economic depression in US history? It was not just one factor, but instead a combination of domestic and worldwide conditions that led to the Great Depression. As such, there is no agreed upon list of all its causes. Here instead is a list of the top reasons that historians and economists have cited as causing the Great Depression. The effects of the Great Depression were huge across the world. Not only did it lead to the New Deal in America but more significantly, it was a direct cause of the rise of extremism in Germany leading to World War II. a. Stock Market Crash of 1929 Many believe erroneously that the stock market crash that occurred on Black Tuesday, October 29, 1929 is one and the same with the Great Depression. In fact, it was one of the major causes that led to the Great Depression. Two months after the original crash in October, stockholders had lost more than $40 billion dollars. Even though the stock market began to regain some of its losses, by the end of 1930, it just was not enough and America truly entered what is called the Great Depression.
b.Bank Failures

Throughout the 1930s over 9,000 banks failed. Bank deposits were uninsured and thus as banks failed people simply lost their savings. Surviving banks, unsure of the economic situation and concerned for their own survival, stopped being as willing to create new loans. This exacerbated the situation leading to less and less expenditures. c.Reduction in Purchasing Across the Board With the stock market crash and the fears of further economic woes, individuals from all classes stopped purchasing items. This then led to a reduction in the number of items produced and thus a reduction in the workforce. As people lost their jobs, they were unable to keep up with paying for items they had bought through installment plans and their items were repossessed. More and more inventory began to accumulate. The unemployment rate rose above 25% which meant, of course, even less spending to help alleviate the economic situation. d.American Economic Policy with Europe As businesses began failing, the government created the Smoot-Hawley Tariff in 1930 to help protect American companies. This charged a high tax for imports thereby leading to less trade between America and foreign countries along with some economic retaliation.

8|P a g e

e. Drought Conditions

While not a direct cause of the Great Depression, the drought that occurred in the Mississippi Valley in 1930 was of such proportions that many could not even pay their taxes or other debts and had to sell their farms for no profit to themselves. This was the topic of John Steinbeck's The Grapes of Wrath. Could the Depression Happen Again? Economists study the Depression both because of its intrinsic interest as a major economic event and to provide guidance to policymakers so that it will not happen again. To state with confidence whether this event could recur, we would need to know why it happened. Because there is not yet agreement on the causes of the Great Depression, it is impossible to rule out with certainty another depression of this magnitude. Yet most economists believe that the mistakes that led to the Great Depression are unlikely to be repeated. The Fed seems unlikely to allow the money supply to fall by one-fourth. Many economists believe that the deflation of the early 1930s was responsible for the depth and length of the Depression. And it seems likely that such a prolonged deflation was possible only in the presence of a falling money supply. The fiscal-policy mistakes of the Depression are also unlikely to be repeated. Fiscal policy in the 1930s not only failed to help but actually further depressed aggregate demand. Few economists today would advocate such a rigid adherence to a balanced budget in the face of massive unemployment. In addition, there are many institutions today that would help prevent the events of the 1930s from recurring. The system of Federal Deposit Insurance makes widespread bank failures less likely. The income tax causes an automatic reduction in taxes when income falls, which stabilizes the economy. Finally, economists know more today than they did in the 1930s. Our knowledge of how the economy works, limited as it still is, should help policymakers formulate better policies to combat such widespread unemployment. The Great Depression: Issues and Ideas The depression was the greatest economic crisis the western world had experienced. In the 1930s, by contrast with 1990, it was the economy of the Soviet Union that was booming while western economies seemed to be collapsing. The questions of what caused the great depression, whether it could have been avoided, and whether it could happen again have therefore to be taken seriously. The classical economics of the time had neither well-developed theory that would explain persistent unemployment nor any policy prescriptions to solve the problem. Many economists then did, in fact, recommend government spending as a way of reducing unemployment, but they had no macroeconomic theory by which to justify their recommendations. Keynes wrote his great work, The General Theory of Employment, Interest and Money, in the 1930s, after Britain had suffered during the 1920s from a decade of double-digit unemployment and while the United States was in the depths of its depression. He was fully aware of the seriousness of the issues. As Don Patinkin of the Hebrew University puts it:
9|P a g e

... the period was one of fear and darkness as the Western world struggled with the greatest depression that it had known.... There was a definite feeling that by attempting to achieve a scientific understanding of the phenomenon of mass unemployment, one was not only making an intellectual contribution, but was also dealing with a critical problem that endangered the very existence of Western civilization. Keynesian theory explained what had happened, what could have been done to prevent the depression, and what could be done to prevent future depressions. The explanation soon became accepted by most macroeconomists, in the process described as the Keynesian revolution, even though the Keynesian revolution did not have much impact on economic policy making in the United States until the 1960s.
The Keynesian Explanation

The essence of the Keynesian explanation of the great depression is contained in the simple aggregate demand model. Growth in the 1920s, in this view, was based on the mass production of the automobile and radio and was fueled by a housing boom. The collapse of growth in the 1930s resulted from the drying up of investment opportunities and a downward shift in investment demand. The collapse of investment, shown in Table -1, fits in with this picture. Some researchers also believe there was a downward shift in the consumption function in 1930s. Poor fiscal policy, as reflected in the perverse behavior of the full-employment surplus from 1931 to 1933, shares the blame, particularly for making the depression worse. It was also widely believed that the experience of the depression showed that the private economy was inherently unstable in that it could self-depress with no difficulty if left alone. The experience of the 1930s was, implicitly or explicitly. The basis for the belief that an active stabilization policy was needed to maintain good economic performance. The Keynesian model not only offered an explanation of what had happened. but also suggested policy measures that could have been taken to prevent the depression, and that could be used to prevent future depressions. Vigorous use of countercyclical fiscal policy was the preferred method for reducing cyclical fluctuations. If a recession ever showed signs of deteriorating into a depression, the cure would be to cut taxes and increase government spending. And those policies would. too, have prevented the depression from being as deep as it was. What about monetary factors in the depression? The Fed argued in the 1930s that there was little it could have done to prevent the depression, because interest rates were already as low as they could possibly go. A variety of sayings of the type, "You can lead a horse to water but you can't make it drink," were used to explain that further reductions in interest rates would have had no effect if there was no demand for investment. Investment demand was thought to be very unresponsive to the rate of interest-implying a very steep IS curve. At the same time, the LM curve was believed to be quite flat, though not necessarily reaching the extreme of a liquidity trap. In this situation, monetary expansion would be relatively ineffective in stimulating demand and output.
The Monetarist Challenge

The Keynesian emphasis on fiscal policy, and its downplaying of the role of money, was challenged by Milton Friedman and his coworkers during the 1950s.9 They emphasized the role
10|P a g e

of monetary policy in determining the behavior of both output and prices. If monetary policy was to be given an important role, though, it was necessary to dispose of the view that monetary policy had been tried in the great depression and had failed. In other words, the view that "You can lead a horse to the water, etc.," had to be challenged. The view that monetary policy in the thirties had been impotent was attacked in 1963 by Friedman and Schwartz in their Monetary History. They argued that the depression, far from showing that money does not matter, "is in fact a tragic testimonial to the importance of monetary factors They argued, with skill and style, that the failure of the Fed to prevent bank failures and the decline of the money stock from the end of 1930 to 1933 was largely responsible for the recession's being a serious as it was. This monetary view, in turn, came close to being accepted as the orthodox explanation of the depression.

Sample Description and Estimation Strategy

It is seemed that it is necessary to know the changing trend of variables. The variable gdp shows that it had been declined gradually throughout the first half of 1930s due to the Great Depression. Then it had been increasing throughout the rest periods up to 1995. The variable cons shows the same pattern as the gdp does. Then in case of invest, it shows that it had been decreasing gradually for the 1930s and first half of 1940s due to Great Depression; and then it continued its growing behavior. The variable g_exp shows that it had great fluctuations from the first half of 1930s to the first half of 1950s; and throughout the rest periods, it continued its growing pattern, in spite of, sudden fall in government expenditure in some periods. The net_exp shows that it had almost all negative values for the whole period that means the US economy were in a acute adverse situation in case of international trade, and it seems that Great Depression more adversely affected this sector. The variable xrate shows the great fluctuations all over the period. The variable cpi shows that inflation rate was negative that means deflation throughout the 1930s. After 1930s, besides some negative values inflation rate was positive where there were reasonable fluctuations throughout the whole periods. The variable m1 shows that real money supply had been decreasing throughout the first half of 1930s and then it continues its increasing pattern up to 1995. In case of real_lrate, the long term interest rate had been gradually decreasing for the 1930s and first half of 1940s to encourage investment, because the amount of idle money increased gradually throughout this period. Then interest had been increasing for the rest periods. The variable u shows that unemployment rate was high during first half of 1930s. Then besides some fluctuations, it gradually decreased for the rest periods. After all, the analysis of changing trend of variables shows that the bad effect of Great Depression retained for the two decades, specifically for 1930s and 1940s, and the US economy continued to recover after 1940s. The estimation strategy of this data involves the regression equation. The specification, we estimate is the following: gdp = 0 + 1 cons + 2 invest + 3. g_exp + 4 net_exp + 5 ypd + 6 xrate + 7 cpi + 8 m1 + 9 real_lrate + 10 u + 11 umale + 12 ufemale + 13 uwhite + 14 ublack e 1
11|P a g e

Then a simple regression has been estimated shown in the following table 3. Table -2 Sample descriptive statistics (N=66) Mean Standard Dev. Minimum Maximum Independent variables cons invest g_exp net_exp ypd xrate cpi m1 real_lrate u Dependent variables gdp Instrumental variable umale ufemale uwhite ublack 1984.756 1243.771 396.4182 284.7993 690.4136 354.881 -30.57424 48.86357 2193.274 1356.816 109.9636 13.15394 3.553846 4.311579 630.1727 199.6766 2.280303 4.524913 7.621212 5.351363 526.6 34.7 103.8 -163.9 511.1 86.6 -10.3 223.5 -10.3 1.2 4577.8 1010.2 1263 98.2 4945.8 143.2 14.4 1091.2 14.3 25.2

3041.005

1822.529

661.5

6742.9

4.727083 5.391667 5.307143 13.38333

1.61795 1.236817 1.302417 2.522536

2.1 2.9 3.1 9.4

8.9 8.3 8.6 19.5

The above table-2 shows the sample descriptive statistics. The table contains mean, standard deviation, minimum, and maximum values for the observation of different variables.

12|P a g e

Table- 3 . reg gdp cons invest g_exp net_exp ypd xrate cpi m1 real_lrate u umale ufemale uwhite ublack Source | SS df MS -------------+---------------------------------------Model | 21363660.1 14 1525975.72 Residual | .041241542 9 .004582394 -------------+----------------------------------------Total | 21363660.1 23 928854.788 Number of obs F( 14, 9) Prob > F R-squared Adj R-squared Root MSE = 24 = . = 0.0000 = 1.0000 = 1.0000 = .06769

----------------------------------------------------------------------------------------------------------------gdp | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+--------------------------------------------------------------------------------------------------cons | 1.000556 .0006486 1542.70 0.000 .9990889 1.002023 invest | 1.000152 .0011576 863.99 0.000 .9975335 1.002771 g_exp | .9997504 .0011907 839.64 0.000 .9970569 1.002444 net_exp | .9989238 .0007899 1264.60 0.000 .9971369 1.000711 ypd | -.0005035 .0007048 -0.71 0.493 -.0020979 .0010909 xrate | .0018204 .0035032 0.52 0.616 -.0061044 .0097452 cpi | .0281297 .0125936 2.23 0.052 -.000359 .0566183 m1 | -.0001978 .0008505 -0.23 0.821 -.0021218 .0017263 real_lrate | -.012287 .0191985 -0.64 0.538 -.0557171 .031143 u | .2844555 .6620371 0.43 0.678 -1.213177 1.782088 umale | .1952372 .2567979 0.76 0.467 -.3856801 .7761545 ufemale | .0635175 .2013114 0.32 0.760 -.3918805 .5189156 uwhite | -.3952621 .4763658 -0.83 0.428 -1.472876 .6823523 ublack | -.0938682 .0711181 -1.32 0.219 -.2547484 .0670121 _cons | .1097303 1.087655 0.10 0.922 -2.350717 2.570178 The above statistical table-3 shows the numerical value of the regression equation. The coefficient of cons shows that, holding all other things constant, $1increase in the consumption expenditure had increased the gdp by the amount $1.000556 and the coefficient of cons is also statistically significant. Then the coefficient of invest shows that $1 increase in the investment expenditure had increased the gdp by the amount $1.000152 and it is also statistically significant. The coefficient of g_exp shows that $1 increase in the real government expenditure had increased the gdp by the amount $0.9997504 and it is also statistically significant. The coefficient of net_exp shows that $1 increase in the net export of goods and services had increased the gdp by the amount $0.9989238 and it is also statistically significant. The coefficient of ypd shows that $1 increase in the national savings had decreased the gdp by the
13|P a g e

amount $0.0005035and it is also statistically insignificant. The coefficient of xrate shows that 1% increased had increased the gdp by 0.18% and it is also statistically insignificant.The coefficient of cpi shows that 1% increase in inflation had increased the gdp by 2.81% and it is also statistically significant. The coefficient of m1 shows that $1 increase in real money supply had increased the gdp by $0.0001978 and it is also statistically insignificant. The coefficient of real_lrate shows that 1% increase in the interest rate had decreased the gdp by 1.22% and it is also statistically insignificant. The coefficient of u shows that 1% increase in the civilian unemployment had increased the gdp by 28.45% and it is also statistically insignificant. The value of the intercept term shows that holding all other things constant gdp had increased by 10.97% per year and it is also statistically insignificant. IV Estimation We know that gdp has a close relation with civilian unemployment rate (u). Now we can calculate a simple regression equation to estimate the relationship between gdp and u. gdp = 0 + 1u + e (2) For comparison, we first obtain the OLS estimates: gdp= 3957.997 120.321 u (3) (369.948) (39.826) n = 66 R2 = 0.125 We have already shown that gdp is closely related with the u (unemployment). The above regression equation (gdp on u) shows a negative relationship between u and gdp, shown by the coefficient of u (-120.321). Notice that unemployment may be correlated with the statistical error term. If this fact happens, there will be endogeneity problem in OLS estimation. Hence an instrumental variable may be used to solve the endogeneity problem. The probable instrumental variable for the u may be umale, ufemale, uwhite, and ublack. First we can use Unemployment Rate, Civilian, Age 20+, male (umale) as an instrumental variable for u. We have to maintain that umale is uncorrelated with e. The second requirement is that u and umale are correlated. We can check this very easily using a simple regression of u on umale: u = 1.32 + .9398 umale (.198) (.0398) (4)

n= 48 R2 =0.924 The t statistic on umale is 23.61, which indicates that u and umale have a statistically significantpositive correlation. In fact, umale explains about 92.4% of the variation in u in the sample.Using umale as an IV for u gives gdp = 770.006 + 520.576 u (794.366) (133.379) (5)

14|P a g e

n= 48

R2 =0.261

Though umale explains significant variations in u, we cannot use umale as an IV for u in this case. Because the IV (2SLS) regression shows the positive relationship between gdp and u. Actually, there is a negative relationship between gdp and u. Now we can check for ufemale as an IV for u. As usual, ufemale have to satisfy the above mentioned two requirements. Then the simple regression of u on ufemale is: u = -0.965 + 1.248 ufemale (0.2295) (0.042) n= 48 R2 = 0.952 (6)

The t statistic on ufemale is 29.714, which indicates that u and ufemale have a statistically significant positive correlation. Besides ufemale explains about 95.2% of the variations in u in the sample. Using ufemale as an IV for u gives gdp = 1280.851 + 431.926 u (787.068) (132.027) (7)

n = 48 R2 = 0.254 Though ufemale explains significant variations in u, we cannot use ufemale as an IV for u in this case. Because the IV (2SLS) regression shows the positive relationship between gdp and u, and this creates contradiction. Similarly, we can check for uwhite as an IV for u. As usual, uwhite have to satisfy the above mentioned two requirements. Then the simple regression of u on uwhite is: u = 0.031 + 1.126 uwhite (0.091) (0.017) (8)

n=42 R2 = 0.991 The t statistic on uwhite is 66.235, which indicates that u and uwhite have a statistically significant positive correlation. Besides uwhite explains about 99.1% of the variations in u in the sample. Using uwhite as an IV for u gives gdp = 1824.123 + 372.826 u (903.677) (146.239) n=42 R2 =0.172
15|P a g e

(9)

Though uwhite explains significant variations in u, we cannot use uwhite as an IV for u in this case. Because the IV (2SLS) regression shows the positive relationship between gdp and u and this creates contradiction. Similarly, we can check for ublack as an IV for u. As usual, ublack have to satisfy the above mentioned two requirements. Then the simple regression of u on ublack is: u = .370 + 0.479 ublack (0.473) (0.035) (10)

n = 24 R2 = 0.896 The t statistic on ublack is 13.686, which indicates that u and ublack have a statistically significant positive correlation. Besides ublack explains about 89.6% of the variations in u in the sample. Using ublack as an IV for u gives gdp = 5885.965 110.194 u (1149.707) (166.968) n = 24 R2 = 0.035 This IV (2SLS) regression (11) shows that there exists a negative relationship between gdp and u when we use ublack as an IV for u. But the above three IV (2SLS) regression equations (5) (7) and (9) show positive relationship between gdp and u that is not acceptable, because we know that gdp falls as u rises and gdp rises as u falls. So we can use only ublack as an IV for u among the above mentioned IVs. Now we can compare OLS and IV regression of gdp on u, that means we have to compare equation (3) with (11). The IV (2SLS) estimate of the return to unemployment is -11019.4%, that means 1% increase in unemployment will decrease the GDP by about -11019.4%, and this is less than the OLS estimate. This suggests that the OLS estimate is too high and is consistent with omitted ability bias. But we should remember that these are estimates from just one sample: we can never know whether -110.194 is above the true return to unemployment, or whether -120.321 is closer to the true return to unemployment. Further, the standard error of the IV (2SLS) estimate is higher than the OLS estimate. Besides, the sample size in (3) is 2.75 times higher than the sample size in (11) and for this reason, the R2 in (3) is about 3.57 times higher than the R2 in (11). Therefore, although the differences between (3) and (11) are practically large, we cannot say whether the difference is statistically significant. (11)

16|P a g e

But it can be concluded, in the case of using ublack as well as umale, ufemale and uwhite as an IVs of u, that we should not use IV for u. Because the OLS regression in (3) shows significant relationship between gdp and u. Besides, the R2 in (3) also significant, because in the OLS regression, other explanatory variables are controlled. The IV (2SLS) regressions are run only for checking whether IVs is applicable for u. Regressions on IS-LM Equation The familiar ISLM equations are as follows: Y = C(Y-T) + I(r*) + G + NX (e). M/P = L(r*, Y). IS* LM*

Now, consider the IS equation. First we can run a regression of consumption on disposable income. cons = -23.532 + 0.916 ypd (13.949) n= 66 (0 .005) (12)

R2 = 0.998

The above OLS regression (12) shows that there is very statistically significant relationship between consumption and disposable income. The R2 is also very high. This shows that consumption increases as disposable income increases and vice-versa. Now, consider the IS equation. First we can run a regression of investment (invest) on real longrun interest rate (real_lrate). Invest = 338.325 + 25.476 real_lrate (36.232) n= 66 (7.194) R2 = 0.164 (13)

The above OLS regression (12) shows that the relationship between invest and real_lrate is statistically significant but economically insignificant, because there exists positive relationship between these. Besides, the R2 is also low. Now we can run another regression of net export (net_exp) on exchange rate (xrate). net_exp= -49.247 + 0 .170 xrate (51.366) n =66 (0.464) R2 = 0.002 (14)

17|P a g e

The above OLS regression (13) shows that the relationship between net_exp and xrate is both statistically and economically insignificant, because there exists positive relationship between these and the t statistics is also low. Besides, the R2 is very low.
Figure-3: IS Variables Contribution to GDP

3000

4000

5000

1000

2000

Contribution to GDP

2000

4000 6000 Gross Domestic Product (GDP) CONS G_EXP INVEST NET_EXP

8000

Now we can run a regression on the basis of IS equation. gdp = .0093 +1.000 cons + 1.000 invest + 1.000 g_exp + 1.000 net_exp (.019) (.000) (0.000) (0.000) (0.000) (15)

n = 66 R2 = 1.000 In the regression equation (14), all the explanatory variables are statistically and economically significant. The R2 is also high and it shows very good fitness of data. Thus the IS equation is statistically very significant. Then, consider the LM equation. First we have to run an regression of real money supply (m1) on gdp and real long run interest rate (real_lrate). m1 =337.906 + 0.106gdp 13.524 real_lrate (21.646) (0.007) (2.663) (16)

n= 66 R2 = 0.805 The LM regression equation shows that gdp and real money supply are positively related shown by the coefficient of gdp (0.106). Besides the real money supply and the real long run interest rate are negatively related as shown by the coefficient of real_lrate. The both gdp and real_lrate
18|P a g e

are statistically significant. The value of the R2 shows the very good fit of the data.
8000

Figure-4: LM Variables

0 200

GDP and Interest Rate 2000 4000 6000

400

600 800 Real Money Supply GDP

1000

1200

REAL_LRATE

So, from (14) and (15), we see that both IS and LM regression equations have demonstrated an strong relationships. We know that the combination of the IS and LM equations shows an important relationship between GDP and exchange rate. Now a simple regression of gdp on xrate is as follows: gdp = 7673.632 - 42.12872 xrate (1827.054) (16.50) n= 66 R2 = 0.093 This regression equation shows that there exists a negative relationship between xrate and gdp. Because as exchange rate increases, import increases and export falls; and thus gdp falls. For this, reason the IS curve is downward slopping and LM curve is vertical, because exchange rate does not enter into LM equation. Besides the relationship between gdp and xrate is also statistically significant, but the R2 is too low as other explanatory variables are controlled for. (17)

19|P a g e

Figure-5: GDP and Exchange Rate Over Time

140 80 0 Exchange Rate 100 120

2000

4000 6000 Gross Domestic Product (GDP)

8000

Findings
We can summarize our empirical estimations from the above procedures by the followings ways. We summarize as follows: 1. As our main goal is to show what factors determine the GDP and how through IS-LM analysis, we have followed the initial periods economic downturn and this evidence shows that the Great Depression is due to supply shock and investment crisis. Due to the inactiveness of the monetary policy, the expansionary fiscal policy was used to prevent the economy from the Great Depression and to stimulate the economic growth, as suggested by the great economist John M. Keynes. 2. The data shows that the bad consequences of Great Depression prevailed in the economy for the next two decades i.e. 1930s and 1940s. 3. The OLS regression model based on the IS equation shows that the GDP is well determined by consumption, investment, government expenditure and net export. The OLS regression model based on the LM equation shows that the real money balances is also well determined by GDP and real interest rate.

Conclusion
Actually, the IS-LM variable data is a vast topic and almost all the important variables used in macroeconomics are focused on the IS-LM variable analysis. For the convenience of the study, the number of variables shown in the original data sheet has been reduced into a short form. Because, it will help us to estimate or forecast more accurately. As our data ranges from the period 1930 to 1995 in USA, it is necessary to analyze the initial economic situation that means Great Depression in 1929 in USA. The analysis on Great Depression in the paper has focused the fact that the depression mainly occurred due to
20|P a g e

investment shock, unexpected rise in national saving, inefficient macroeconomic policy and as consequently, stock price index fall to a great extent. Then the bad effect of depression had been sustaining for about two decades, specifically 1930s and 1940s, and then US economy marched for a better improvement. In this paper, we have already examined about the effects of monetary and fiscal policy on GDP that means, the effects of real consumption expenditure, real gross private investment, real government expenditure, exchange rate etc. on GDP both in short run and long run. Our data shows several important dimensions about the determination of GDP. Actually, real gross domestic consumption is the most important factor in determining GDP. Consumption explains about 65 percent variations in GDP over time. The second important determinant is the government expenditure and it explains about 22 percent variations in GDP. Next the third important determinant is the real gross private domestic investment and it explains about 13 percent variations in GDP. Then the fourth important determinant is the real net export of goods and services. These above mentioned four variables are the fiscal policy factors and any change in fiscal policy changes these variables. Then the monetary policy factors are shown by the LM equation. These monetary policy factors are the real interest rate, real money supply, and money demand. Any change in monetary policy changes these factors. Due to lack of proper information, time, and experience, probably, we have not presented this paper accurately. But we have spent my best effort in making this paper. If we will get such an opportunity like this in future, then we will apply our best effort to correct the existing problems in this paper and to make next paper more accurate and attractive. We also strongly believe that this paper will be very helpful for our future research.

21|P a g e

References
Mankiw, N. Gregory (1997). Macroeconomics, 5th edition. Worth Publishers New York: pages (295-311). Rudiger Dornbush, Stanly Fisher, Macroeconomics,6th edition. Pages (437-445). Hall and Taylor "Macroeconomics" page. http://www.wwnorton.com/college/econ/macro/ Economics Web Institute. http://www.economicswebinstitute.org

22|P a g e