The Impact of Income Taxes on Consumption: Automatic Stabilizers and Realism, Schemes and Mind Maps of Business

Download The Impact of Income Taxes on Consumption: Automatic Stabilizers and Realism and more Schemes and Mind Maps Business in PDF only on Docsity! FRBNY Economic Policy Review / April 2000 35 The Automatic Fiscal Stabilizers: Quietly Doing Their Thing I. Introduction he cyclical nature of the U.S. economy has undergone profound changes over the past century. As carefully documented by Diebold and Rudebusch (1992) and Romer (1999), since World War II, recessions have become less frequent and business expansions have become substantially longer. In addition, Romer argues that recessions are now less severe: Output loss during recessions is about 6 percent smaller on average in the post–World War II period than in the thirty- year period prior to World War I and substantially smaller than in the 1920 to 1940 interwar period. Furthermore, the variance of output growth has declined as well. Romer attributes these changes largely to the rise of macroeconomic policy after World War II; in particular, she argues that the automatic fiscal stabilizers—including the income-based tax system and unemployment insurance benefits—have played a prominent role in converting some periods of likely recession into periods of normal growth as well as in boosting growth in the first year following recession troughs. Given the Keynesian-style models used by Romer to support her claims, one would expect that personal consumption also would have been stabilized since World War II. Indeed, Basu and Taylor (1999) present evidence that the volatility of aggregate U.S. consumption has declined in the postwar period. This paper presents theoretical and empirical analysis of automatic fiscal stabilizers. Using the modern theory of consumption behavior, we identify several channels through which optimal reaction of household consumption plans to aggregate income shocks is tempered by these stabilizers. Such automatic stabilization occurs even when households have full understanding of the constraints on their behavior implied by the government’s intertemporal budget constraint and have full awareness of the difference between aggregate and idiosyncratic shocks to their labor income. This does not necessarily imply that the current fiscal stabilizers in the United States are set at optimal levels. The analysis of optimal tax rates, for example, is the subject of a large literature that involves comparing the benefits and costs of different settings and would take us well beyond the scope of this paper. Moreover, our theoretical findings raise the issue of whether the insurance, wealth, and liquidity effects of the income tax system that we identify are realistic channels through which the effects of income shocks are stabilized. Furthermore, there is the issue of whether these channels are more or less empirically important than the wealth channel identified in earlier work, a channel whose effect requires that households have incomplete information about the nature of income shocks. We believe that these remain important open issues, although we would not be surprised if elements from each channel eventually were found to be empirically meaningful. However, in an attempt to bring at least some evidence to bear on these issues, we present results from several empirical exercises using postwar U.S. data. Using standard time-domain Darrel Cohen and Glenn Follette are economists in the Division of Research and Statistics of the Board of Governors of the Federal Reserve System. The authors thank Olivier Blanchard, Kevin Hassett, David Lebow, Wolf Ramm, John Roberts, Louise Sheiner, Tom Simpson, and Karl Whelan for helpful comments, as well as Flint Brayton for carrying out the FRB/US model simulations. The authors also appreciate the excellent research assistance of Eliot Maenner, Grant Parker, and Dana Peterson. The views expressed are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Darrel Cohen and Glenn Follette T 36 The Automatic Fiscal Stabilizers techniques, we estimate elasticities of the various federal taxes with respect to their tax bases and responses of certain components of federal spending to changes in the unemployment rate. Using frequency-domain techniques, we confirm that the relationships found in the time domain are strong at the business-cycle frequencies. Together, these results showing strong ties between cyclical variation in income and federal government spending and taxes suggest the potential for the automatic fiscal stabilizers to play a quantitatively important role in the economic stabilization process. Using the Federal Reserve Board’s FRB/US quarterly econometric model, however, we find that the automatic fiscal stabilizers play a rather limited role in damping the short-run effect of aggregate demand shocks on real GDP, reducing the “multiplier” by about 10 percent, although they have a somewhat larger damping impact (in percentage terms) on personal consumption expenditures. Very little stabilization is provided in the case of an aggregate supply shock. Before turning to the details of our analysis, it is worth mentioning the startling result developed by Lucas (1987). In the context of a standard model of an optimizing representative consumer, Lucas argues that perfect stabilization—that is, complete elimination of the variance of consumption in the United States—would yield virtually no utility gain to households both in absolute terms and relative to the huge utility gain associated with only a modest increase in the growth rate of consumption. Moreover, much of the subsequent literature has supported the robustness of this result. As such, this finding calls into question the act of devoting resources to the study (as well as to the practice) of stabilization policy. While a complete response is well beyond the scope of this paper, we would make the following brief points. First, national election outcomes and, indeed, the very cohesiveness of societies appear to depend on the state of the business cycle; such factors generally are not captured in the standard utility- maximizing framework. Second, cyclical downturns have a negative and, quite possibly, sizable impact on a minority of the work force; thus, stabilization policy may generate a large welfare gain even if the gain averaged across the entire population is small.1 Third, business-cycle variation and long- term growth (or the mean level of consumption) may not be completely independent, as assumed by Lucas; for example, the loss of human capital associated with job loss during a cyclical downturn might have long-lasting impacts. Fourth, the Lucas result depends partly on the actual variance of U.S. consumption over the post–World War II period, a variance that has declined relative to the prewar period to a fairly low level. If this outcome has resulted largely from macroeconomic stabilization policy, as argued by Romer (1999), then elimination of stabilization policy might cause a large enough increase in aggregate consumption variance to alter the Lucas result. The rest of our paper is structured as follows. The next section offers three theoretical arguments for the effectiveness of automatic stabilizers; each is formally developed as a variation on the same underlying consumer optimization problem. While these modeling exercises, as well as a brief analysis of firms’ investment demand, are carried out in a partial equilibrium context, there will be some discussion of general equilibrium issues as well. Section III reports simulations of the Federal Reserve Board’s FRB/US quarterly econometric model. Section IV analyzes the business-cycle relationship between income and certain federal government taxes and spending using frequency-domain techniques. Section V presents a complete reestimation of the high- employment budget model used by staff at the Federal Reserve Board and the U.S. Congressional Budget Office (CBO) for the past twenty years. Section VI concludes. II. The Analytics of Automatic Fiscal Stabilizers A. Review of the Literature This section examines theoretically the role of automatic fiscal stabilizers—in particular, the income tax—in modifying the response of consumption to income shocks. Perhaps surprisingly, there has been very little written on this subject in the academic literature since the mid-1980s, despite numerous legislative changes in individual income tax rates beginning with the Tax Reform Act of 1986.2 We will briefly discuss earlier work on the role of automatic stabilizers, drawing on the excellent summary in Blinder and Solow (1974) in the context of the basic Keynesian model and on the seminal work in Christiano (1984) showing the possibility that the automatic stabilizers could work using an explicit framework of an optimizing consumer facing uncertain income prospects.3 We then present new models of the effects of the income tax on optimizing consumers that we feel are a move toward greater realism. In contrast to the earlier Keynesian tradition, our models are not full general equilibrium exercises. However, we would argue that the consumer’s decision problem must be central to any sensible analysis of the role of automatic stabilizers and, at the end of the section, we conjecture that general equilibrium feedback is unlikely to change qualitatively FRBNY Economic Policy Review / April 2000 39 of taxes on the wealth effect. The second term represents the offsetting negative effect on consumption owing to higher precautionary saving: higher aggregate first-period income tax receipts imply a lower second-period income tax rate and thus less insurance against idiosyncratic income shocks. As shown in Barsky, Mankiw, and Zeldes, the precautionary saving effect requires that be positive (so that the covariance term in equation 8 is positive). We assume that the wealth effect dominates the precautionary saving effect and hence that a positive (negative) increment to labor income boosts (reduces) first-period consumption. Differentiation of equation 8 with respect to establishes that a stronger automatic stabilizer (that is, a larger ) reduces the positive impact of a temporary income shock on first-period consumption—that is, it establishes that ; it does so by strengthening the precautionary saving effect. Before moving on to our next models, we briefly discuss the assumption made here and in the prior literature: that interest income is not taxable. The introduction of interest income taxation into our model would tend to strengthen the above results regarding automatic stabilizers for two reasons. First, higher before-tax income in period one would lead to a reduction in the income tax rate in period two for the same reason as before, and because the second-period tax base is larger (higher labor income boosts first-period saving and hence interest income subject to tax in the second period) and total second-period tax receipts are determined completely by first-period taxes and government spending. The resulting lower income tax rate in period two further strengthens precautionary saving. Second, a lower second-period tax rate boosts the after-tax interest rate, for a given before-tax rate, which further increases the incentive to save (if the substitution effect exceeds the income effect). C. New Results—Approach 2 We now modify the model to allow a change in income taxes induced by a temporary income shock to be matched by a change in government consumption spending; both are assumed to occur in the first period. It is thus useful to rewrite equation 7 as follows: (9) . In addition, it is assumed that private and government consumption expenditures are directly substitutable (although not necessarily perfect substitutes) within periods; that is, is a substitute for but not for , and similarly for . Thus, for the utility function in equation 1—that is, for —we assume that and and RU222 U122– τ1 τ1 ∂ ∂C1 ∂µ1⁄[ ] ∂τ1⁄ 0< G1 τ1µ1A τ2µ2AR 1– G2R 1––+= G1 C1 C2 G2 U C1 C2 G1 G2,, ,( ) U13 0< U24 0< . In our example, however, is fixed and hence only the conditions and are relevant.7 To evaluate the effect of a shock to the first-period endowment labor income of each person, we again differentiate the first- order conditions, giving: (10) , where is defined above. The first term on the right-hand side of equation 10, which is positive, represents the “wealth effect” of higher after-tax labor income; before-tax labor income is higher, but this is partially offset by higher income taxes in the first period. This offset, owing to the automatic stabilizers (that is, the income tax), is reinforced by the second term on the right-hand side of equation 10. The latter term, which is negative, represents the direct substitution effect of higher government consumption spending (owing to higher income taxes) on private consumption. We assume that the wealth effect dominates the direct substitution effect and hence . Differentiation of equation 10 with respect to establishes that a stronger automatic stabilizer (that is, a higher ) weakens the positive impact of a temporary shock to before-tax labor income, that is, it establishes that . D. New Results—Approach 3 In the final variant of our model, we introduce explicit constraints on borrowing by households following the approach in Chan (1983). We assume that borrowing cannot exceed a fixed fraction of current after-tax labor income and, for simplicity, that . If denotes household lending or borrowing , the constraint can be written as: (11) , where is some fixed, positive number. For example, if and if the constraint is binding in the sense that household borrowing equals after-tax income, then first-period consumption is double after-tax income (that is, the sum of disposable income and the borrowed amount, also equal to disposable income). Such a constraint is consistent with home mortgage payment rules-of-thumb in which monthly interest payments cannot exceed a fixed fraction of income. The possibility of borrowing or liquidity constraints is appealing, especially in light of recent empirical work, such as that of Parker (1999) and Souleles (1999), which finds that individual consumption rises when fully anticipated increases in after-tax income are realized. The rest of the model is the same as in Section II B, in which future income taxes are assumed to adjust to maintain the U23 U14 0= = G2 U13 0< U23 0= dC1 dµ1⁄ 1 H⁄( ) R– E RU22 U12–[ ]( ) 1 τ1–( )= Nτ1 H⁄( ) EU13( )+ H dC1 dµ1⁄ 0> τ1 τ1 ∂ ∂C1 ∂µ1⁄[ ] ∂τ1⁄ 0< ε1 0= L L 0>( ) L 0<( ) b 1 τ1–( )µ1 L 0≥+ b b 1= 40 The Automatic Fiscal Stabilizers government’s intertemporal budget constraint (and in which is replaced by in equations 3 and 4). We consider households for whom the borrowing constraint (equation 11) is binding. For such individuals, the model solution for first- period consumption follows immediately, as in the example above, because the borrowing constraint (along with current after-tax labor income) completely determines first-period consumption. It follows that a higher income tax rate—that is, stronger automatic stabilizers—reduces first-period consumption and hence reduces the effect of a labor income shock on first-period consumption. With an adverse shock to labor income, for example, private borrowing is reduced but, because income taxes decline, government borrowing is increased. As noted by Chan (1983) in a related problem, the government—which is not subject to a borrowing constraint— is effectively borrowing on households’ behalf, thereby circumventing the household limit. E. Investment and General Equilibrium Considerations We now address some loose ends in the prior analysis. We begin with a discussion of the relationship between investment demand and the automatic stabilizers in a partial equilibrium, optimizing framework. We then discuss general equilibrium issues, offering several conjectures but not the development of a full model. Conventional models of business-fixed investment—under the key assumptions of convex adjustment costs, complete information, and perfect capital markets—imply that a firm’s investment demand depends on marginal “ ,” that is, on the present discounted expected value of profits from new investment. To the extent that business cycles are viewed as symmetric variations of economic activity (and hence profits) about trend, a recession will be followed by above-trend activity, implying that the recession likely would have little effect on the present value of a representative firm’s expected profit stream and hence on investment demand. In this case, a corporate profits tax would not be expected to damp the effect of cyclical swings in economic activity on investment demand. Other models, based on asymmetric information and the resulting incentive problems in capital markets, imply that information costs and the internal resources of firms influence the cost of external funds. Consequently, investment demand depends on the “financing constraint” of a firm’s net worth, proxied for by current after-tax cash flow, in addition to marginal . Hubbard (1998) provides an excellent discussion of such models, whose empirical importance is the subject of W L q q some controversy. These models imply that the impact of a cyclical downturn on before-tax cash flow and hence on investment demand would be attenuated by the presence of an income tax; thus, the tax would serve as an automatic stabilizer for investment demand. We now briefly discuss general equilibrium issues. The most basic question is whether the economy is better modeled using the equilibrium real business-cycle approach, as in Baxter and King (1993), or using an approach that allows for nominal demand shocks to have real effects in the short run, as in New- Keynesian models. Although the appropriate framework has been the source of ongoing tension among macroeconomists, in qualitative terms the effectiveness of automatic stabilizers appears invariant to the choice of framework. For the remainder of this section, we assume that both frameworks embed the basic consumer optimization model analyzed above. In the equilibrium business-cycle approach, a shock that reduces aggregate equilibrium output—such as a temporary negative labor income endowment shock—generally originates on the supply or production side of the economy, and the components of aggregate demand must adjust to maintain goods market equilibrium. Thus, if personal consumption falls (as the above analysis suggests) and if government purchases of goods and services are reduced to offset the budget impact of lower income tax receipts, then investment likely will decline to maintain goods market equilibrium.8 The decline in real income net of tax, as well as the decline in government purchases, has no immediate effect on output unless labor supply adjusts in response to wealth and interest rate effects. However, over time, as the capital stock falls relative to baseline, output also declines, which in turn reduces consumption possibilities. The magnitude of the consumption decline will vary inversely with the strength of the automatic stabilizers. By contrast, in a model with sticky wages and prices, negative shocks to any component of nominal aggregate demand (for example, export demand) can lead to short-run reductions in output as labor demand and hours worked decline. The resulting fall in after-tax income reduces private consumption demand (and government purchases fall if they are adjusted to maintain budget balance); the decline in consumption is mitigated by the automatic stabilizers for the same reasons as discussed earlier. Of course, investment demand likely will be boosted by lower interest rates, which implies subsequent increases in the capital stock and output; again, the magnitude of such increases will vary inversely with the strength of the automatic stabilizers. Simulation results from a general equilibrium econometric model with New– Keynesian-style features are presented in the next section. FRBNY Economic Policy Review / April 2000 41 III. Results from the FRB/US Model In Sections III, IV, and V, we present our empirical results. This section presents estimates of the impact of automatic stabilizers—particularly income taxes—based on simulations of the Federal Reserve Board’s FRB/US quarterly econometric model of the U.S. economy. Detailed discussions of the new model can be found in Brayton and Tinsley (1996) and in Reifschneider, Tetlow, and Williams (1999). Households and firms are optimizers whose current decisions are based on expectations of future conditions. For estimation purposes, sectoral expectations are derived from forecasts of small vector autoregressions (VARs). Each VAR has a common set of variables, including consumer price inflation, the output gap, and the federal funds rate. Inclusion of the funds rate means that this form of expectations incorporates an average sample view of how monetary policy was conducted historically. Simulation exercises in this paper also use the same VAR systems. In terms of dynamic adjustments in the model, financial market variables such as interest rates and stock prices adjust immediately to changes in expectations because financial decisions are assumed unaffected by frictions, given the small cost of transacting in these markets. However, the response of nonfinancial variables such as consumption, investment, and employment to changes in fundamentals is not immediate because of (nonexplicitly modeled) frictions in the dynamic adjustment process such as contracts and capital adjustment costs. Indeed, prices and quantities do not adjust quickly enough to ensure full resource utilization at all times. In the long run, however, all adjustments are complete and all markets are clear. Of particular relevance for the simulation results reported below—as well as for a comparison with the prior theoretical discussion of Section II and subsequent empirical analysis of tax elasticities in Section V—is the modeling of aggregate income taxes and consumption in FRB/US. Starting with taxes in FRB/US, the average federal personal income tax rate is procyclical, implying an elasticity of personal taxes with respect to the taxable income base somewhat greater than the corresponding elasticity of 1.4 estimated in Section V.9 Social insurance contributions are specified as proportional to its tax base, implying a unitary elasticity; in Section V, we estimate that the elasticity is about 0.9. The average corporate income tax rate is mildly procyclical in FRB/US; this contrasts with the mildly countercyclical tax rate found in Section V. Turning to the modeling of aggregate consumption in FRB/US, we see that a small fraction of consumption decisions is made by liquidity-constrained households; the share of after- tax income associated with this group of households is estimated at about 10 percent. This group’s behavior would be consistent with the model in Section II D. However, for most households, consumption depends on current property wealth plus the present value of expected after-tax labor (and transfer) income in FRB/US. Expected future income flows are discounted at a high—25 percent— annual rate in computing present values, because it is argued that households are quite averse to the uncertainty of future uninsurable income. As a result of the heavy discounting, current consumption is not affected much by changes in income taxes in the distant future that might be necessary to satisfy the government’s intertemporal budget constraint. Put another way, the rate used by individuals to discount future taxes exceeds the government’s borrowing rate. Moreover, the simulations below are based on VAR expectations that do not incorporate expectations of future tax rate changes. Thus, a change in income taxes (owing, say, to an aggregate demand shock) has a wealth effect on consumption.10 While this is similar to the wealth effect in the model of Section II C, there is a difference in that current government purchases are not adjusted in FRB/US (and so there is no substitution of private for government consumption). Finally, FRB/US may be consistent with a precautionary saving motive. This is because prudent households act as if they apply a high discount rate to future uncertain income, which is the case in the model. Furthermore, consumption depends positively in FRB/US on the expected output gap, which is viewed as capturing countercyclical variation in the perceived riskiness of future before-tax income. Even granting these interpretations, the model does not capture the insurance effect of income tax rates developed in Section II B; that is, an anticipated change in the income tax rate has no effect in FRB/US on the variance of after-tax income. Summing up, FRB/US captures liquidity and wealth effects associated with the income tax system, but does not capture the insurance effect.11 However, there is a sense in which the impact of changes in taxes (and transfers) on consumption demand is assumed: for example, there is no formal testing of the hypothesis that the effects of changes in before-tax income and in taxes are of equal and opposite signs (and separately statistically significant). Rather, after-tax income is the variable included in the FRB/US model consumption equations. Results of the simulation exercises are reported in Tables 1 and 2. The model has four federal tax rates (for personal income taxes, corporate income taxes, indirect business taxes, and social insurance contributions). The effects of automatic stabilizers are measured by comparing simulations in which each federal tax rate is at its actual value with simulations in which each tax rate is set to zero and an add factor (essentially 44 The Automatic Fiscal Stabilizers Finally, on the NIPA tax side, squared coherencies between social insurance contributions and wages and salaries and between indirect business taxes and nominal GDP are only of moderate size (up to about 0.5). We now discuss results using unified individual income tax data (on a quarterly basis). Because these data are not seasonally adjusted (NSA), we also need an NSA personal income tax base. Since this is not available, we use NSA nominal GDP, which is publicly available. The use of NSA data gives a pure reading of real-time fluctuations in taxes and income faced by households, but at the cost of introducing a lot of noise, especially into the analysis of individual nonwithheld taxes (declarations, paid four times per year, plus final payments, paid once each year). The squared coherency between NSA withheld income taxes and nominal GDP (again, both in quarterly growth-rate form) is sizable, both at the business-cycle frequencies and at the primary seasonal frequency ( ). The former is strongly suggestive of the working of automatic stabilizers during business-cycle swings while the latter reflects seasonal patterns in labor incomes and withheld taxes (such as increases in each that often occur at the beginning of calendar years). The gain varies between 1 and 3 at the business-cycle frequencies, again suggestive of discretionary tax changes in addition to the automatic stabilizer component. Very similar results at the business-cycle frequencies arise when the raw data are filtered using four- quarter growth rates (although the strong seasonal relationship is eliminated, as would be expected). By contrast, the squared coherency between NSA individual nonwithheld taxes and nominal GDP is not large at business-cycle frequencies; indeed, the relatively small coherencies apply both to declarations and final payments. Such results suggest the relative ineffectiveness of automatic stabilizers via this tax channel. Finally, on the spending side of the budget, the squared coherency between unemployment insurance outlays as a percentage of GDP and the unemployment rate is very high at the business-cycle frequencies. Thus, even though there may be a short waiting period to collect benefits, the unemployment insurance program appears to operate as an effective, virtually automatic, income stabilizer for unemployed individuals.18 To sum up, the frequency-domain analysis establishes a very strong relationship between income taxes and tax bases at the business-cycle frequencies. In all cases, this reflects the automatic nature of tax variation—particularly of individual withheld taxes—when incomes change, and in some cases it likely reflects discretionary tax changes as well. Furthermore, unemployment insurance also appears effective as an automatic stabilizer of income. ω π 2⁄= V. The High-Employment Budget Surplus In this section, using standard time-domain techniques, we present updated empirical estimates of the responsiveness of federal taxes and certain spending programs to cyclical swings in the economy. While such estimates are useful for many purposes, they are used here as a basis for computing the cyclically adjusted, or high-employment budget surplus (HEB), of the federal government. Although the HEB is not without its faults, as discussed in Blinder and Solow (1974), it nonetheless has been used as a summary measure of the stance of fiscal policy by many U.S. government agencies (and many countries) since the 1960s. Twenty years ago, an intergovernmental task force developed the “gross-up” methodology currently used by staff at the U.S. Congressional Budget Office and the Federal Reserve Board (see deLeeuw et al. [1980]). Using taxes to illustrate the method, high-employment tax receipts equal a cyclical adjustment, or a gross-up, plus actual (or projected actual) tax receipts. The gross-up is the difference between an estimate of taxes at a benchmark (that is, high- employment) level of economic activity—computed by setting the GDP gap equal to zero in key econometric equations—and at the actual level of economic activity—computed by using the actual GDP gap. As a result, the gross-up method has the property that actual and high-employment taxes are equal when the economy is operating at potential. More fundamentally, the method has the property that unexplained shocks to taxable income shares and tax receipts are allowed to pass through to high-employment estimates. The remainder of this section presents detailed estimates. A. High-Employment Receipts The calculation of high-employment receipts involves three steps. First, income share equations are estimated to determine the level of the tax bases if actual GDP was equal to potential GDP. Second, the tax elasticities with respect to cyclical changes in income must be estimated. Finally, these two estimates are combined to obtain cyclical components of tax revenues, which are added to actual revenues to obtain high- employment revenues. The basic equations for receipts are: (12) (13) (14) (15) , where is the ratio of the tax base to GDP; is the tax base applicable to the th tax; is tax revenues from tax ; and is the sum of all taxes from all sources. SHAREKj t, SHAREj t, Σβi ∗GDPGAPt i––= BASEKj t, GDPKt ∗SHAREKj t,= TAXKj t, TAXj t, ∗ BASEKj t, BASEj t,⁄( )ε j t,( )= RECEIPTSKt ΣTAXKj t,= SHARE BASEj j TAXj j RECEIPTS FRBNY Economic Policy Review / April 2000 45 The suffix denotes a high-employment estimate; is the sensitivity of the share of the tax base in GDP to changes in the GDP gap ( ); and is the elasticity of the tax with respect to cyclical changes in the tax base. On the income side, GDP is composed of labor compensation (wages and salaries, and supplements to wages and salaries such as employer-provided health insurance), capital income (corporate profits, proprietors’ income, rental income, dividends, and net interest), and GDP less national income (the statistical discrepancy between income- and product-side measures of GDP as well as indirect taxes and net subsidies to businesses). We estimate the cyclical properties of each of these income sources using the U.S. Congressional Budget Office’s estimates of potential GDP, the Non-Accelerating Inflation Rate of Unemployment (NAIRU), and the potential labor force. From these estimates, we construct estimates of the GDP gap, , and the employment gap (Table 3).19 Our regression equations for income shares are in first- difference forms of equation 13 because the shares are not stationary over the sample period.20 The cyclically adjusted share is equal to the actual share less the sum of the products of the estimated gap terms and the coefficients. The cyclically adjusted shares are obviously smoother (Table 4). NIPA personal taxes are roughly 45 percent of federal NIPA-based receipts. They are composed of personal income taxes, estate and gift taxes, and nontaxes (essentially fees and fines). As income taxes are about 97 percent of personal taxes, we use the personal income tax elasticity for all personal taxes. This elasticity, , can be decomposed into two elasticities: the change in income taxes with respect to adjusted gross income (AGI), and the change in AGI with respect to NIPA-adjusted personal income, .21 Furthermore, the elasticity of income taxes with respect to a change in AGI is a weighted sum of the elasticity of taxes to number of returns, , and the elasticity of taxes with respect to average income per return, , where the weights equal the relative contributions of changes in returns and average income to cyclical changes in income. Thus, may be written as: (16) ,22 where: is the percentage gap in number of income tax returns, is the percentage gap in AGI per tax return, is elasticity of personal income taxes with respect to the change in number of returns, is elasticity of personal income taxes with respect to the change in AGI per return, and is the elasticity of AGI with respect to NIPA-adjusted personal income. K β GDPGAP ε GDPK GDP–( ) GDPK⁄ Epersonal Eagi En Ey Epersonal Epersonal En∗ngap Ey∗ygap∗ 1 ngap+( )+{ }= ngap ygap∗+ 1 ngap+( )[ ] }∗Eagi⁄ ngap ygap En Ey Eagi Table 3 Potential GDP, NAIRU, and Labor Force Participation Year Potential GDP (Billions of Dollars) NAIRU (Percent) Potential Labor Force (Millions) GDP Gap (Percent) Employment Gap (Percent) 1951 327.5 5.3 61.9 -3.7 -2.3 1952 348.6 5.4 62.2 -2.9 -2.4 1953 367.2 5.4 62.7 -3.4 -3.1 1954 383.9 5.4 63.8 0.7 0.5 1955 402.2 5.4 65.0 -3.2 -1.1 1956 429.2 5.4 66.1 -2.0 -2.1 1957 458.6 5.4 67.1 -0.5 -0.9 1958 485.7 5.4 67.7 3.8 1.5 1959 508.6 5.4 68.2 0.3 -0.2 1960 534.9 5.5 68.9 1.6 -1.0 1961 562.0 5.5 70.1 3.1 0.7 1962 591.7 5.5 71.2 1.1 0.9 1963 622.6 5.5 72.4 0.8 0.8 1964 657.5 5.6 73.6 -0.8 0.2 1965 698.6 5.7 74.8 -2.9 -0.8 1966 749.9 5.8 76.0 -5.1 -1.8 1967 807.8 5.8 77.3 -3.2 -2.2 1968 879.4 5.8 78.5 -3.6 -2.7 1969 957.8 5.8 79.8 -2.5 -3.6 1970 1,046.1 5.9 82.0 1.0 -1.9 1971 1,138.2 5.9 84.4 1.1 -0.0 1972 1,225.9 6.0 86.8 -0.9 -0.7 1973 1,339.7 6.1 89.3 -3.2 -1.4 1974 1,510.5 6.2 91.8 0.9 -0.8 1975 1,705.9 6.2 94.2 4.4 2.9 1976 1,862.7 6.2 96.8 2.3 2.2 1977 2,045.9 6.2 99.4 0.9 1.2 1978 2,269.3 6.3 102.0 -1.0 -0.4 1979 2,544.5 6.3 104.8 -0.5 -0.6 1980 2,860.6 6.2 107.0 2.7 1.0 1981 3,208.4 6.2 108.8 2.9 1.6 1982 3,488.4 6.1 110.6 7.1 4.1 1983 3,721.1 6.1 112.3 5.6 4.4 1984 3,958.5 6.0 114.1 1.4 2.1 1985 4,206.8 6.0 115.9 0.6 1.6 1986 4,442.1 6.0 117.8 0.4 1.0 1987 4,709.1 6.0 119.7 0.4 0.1 1988 5,015.9 5.9 121.6 -0.7 -0.5 1989 5,366.1 5.9 123.6 -1.4 -0.9 1990 5,736.0 5.9 125.4 -0.1 -0.6 1991 6,092.7 5.9 126.9 2.9 1.4 1992 6,382.8 5.8 128.3 2.2 1.9 1993 6,679.4 5.8 129.7 1.8 1.6 1994 6,981.9 5.8 131.2 0.5 0.4 1995 7,312.3 5.7 132.6 0.6 0.1 1996 7,644.9 5.7 134.1 -0.2 -0.2 1997 8,005.5 5.7 135.9 -1.3 -1.1 1998 8,328.8 5.6 137.4 -2.2 -1.4 Source: U.S. Congressional Budget Office. Note: NAIRU is the Non-Accelerating Inflation Rate of Unemployment. 46 The Automatic Fiscal Stabilizers is set equal to 1 by assuming that changes in the number of tax filers occur in proportion to the existing distribution. By assuming that is 1, we see that should account for the elasticity of the tax code, given the distribution of income, and the change in the distribution of income over the cycle. Our estimate of , though, is based solely on the tax structure and the existing distribution of income; thus, it abstracts from any potential cyclical sensitivity of the income distribution. Equation 16 was modified to account for two types of filers, as the number of returns and the incomes of single filers appear to exhibit different cyclical properties than those of nonsingle filers. We calculate for single and nonsingle filers (overwhelmingly married filing jointly, but also heads of households, married filing separately, and surviving spouses) using SOI cross-sectional data for each year. for a given type of filer is the weighted sum of the elasticities of the AGI groups shown in the SOIs where the weights equal the tax shares of the groups. The elasticity is estimated by dividing the effective marginal tax rate by the average tax rate for the En En Ey Ey Ey Ey group.23 The effective marginal tax rates are lower than the statutory rates because the effective rates incorporate the rise in deductions that occurs as income rises and include the tax preference for capital-gains realizations.24 Table 5 displays the resulting elasticity estimates, . Over the 1951-96 period, the AGI per return elasticity for nonsingle returns averaged 1.6, and was 1.5 for single returns. This largely reflects differences in the 1950s and 1960s owing to lower average tax rates faced by nonsingles in the lower income brackets because of the relatively more generous personal exemptions in place at the time. Focusing on nonsingle filers, we see that their elasticity fell by 0.1 as a result of the Reagan tax cuts in the early 1980s and fell by another 0.1 with the 1986 Tax Reform Act. During the 1990s, the overall elasticity of the tax schedule has hardly changed, as the elasticity-boosting effects of the expansion of the Earned Income Credit (EIC) and increased marginal income tax rates for high- income filers have been offset by the decrease in the tax rate on capital-gains realizations and the shift in income distribution toward high-income filers who have lower elasticities. Ey Table 4 Share Equations Dependent Variable Wages Supplements Profits Proprietors’ Income Rental Income Net Interest Other Personal Interest Dividends Constant -0.018 (-1.15) 0.038 (5.27). -0.005 (-0.21). -0.027 (-1.69). -0.010 (-1.56). 0.022 (1.98). 0.010 (2.09). 0.004 (0.85). Gap 0.221 (12.6)# 0.030 (3.81) -0.319 (12.5). -0.009 (-0.53). 0.021 (2.93) 0.030 (2.45). 0.016 (3.20). 0.003 (0.60). Gap[t-1] -0.106 (-5.89) -0.010 (-1.21) 0.054 (2.05). 0.015 (0.85). -0.010 (-1.38) 0.010 (0.76). -0.019 (-3.78). -0.008 (-1.67).. Gap[t-2] -0.059 (-3.26) 0.002 (0.30) 0.052 (1.97) -0.010 (-0.54). 0.002 (0.25) -0.012 (-0.94) -0.005 (-1.05). -0.013 (-2.67).. Gap[t-3] -0.056 (-3.09) -0.011 (-1.31) 0.006 (0.24) -0.023 (-1.30). 0.001 (0.08) -0.016 (-1.28). -0.001 (-0.02). -0.002 (-0.41).. Gap[t-4] -0.018 (-1.06) 0.001 (0.16). 0.067 (2.67) 0.006 (0.34). -0.004 (-0.58). 0.003 (0.26). -0.008 (-1.71). 0.006 (1.20). Sum of gap coefficients -0.018 0.013 -0.139 -0.021 0.010 0.015 -0.017 -0.014 Adjusted R2 0.55 0.07 0.50 0.02 0.03 0.03 0.12 0.06 Durbin-Watson 1.63 1.78 2.20 2.02 2.05 1.27 1.80 1.37 Notes: The sample period is first-quarter 1955 to fourth-quarter 1997. Dependent variables are measured as first differences of the variable divided by GDP. Gap terms are first differences of (GDPK-GDP)/GDPK; t-statistics are shown in parentheses. FRBNY Economic Policy Review / April 2000 49 the taxable portion of earnings of those with earnings above the cap. A little algebra yields the elasticity of taxes with respect to an increase in income, equation 19. (18) , where: = the statutory tax rate, = the average wage of those below the wage cap, = the fraction of wage earners below the wage cap, = the maximum wages subject to taxation, and = the number of wage earners. (19) . Calculations using data on the distribution of earners and earnings above the wage cap from the annual Social Security Bulletin yield the tax-schedule elasticities, , shown in Table 8. The elasticity of FICA taxes with respect to wages and salaries rises after the early 1970s because the share of workers below the wage cap rises as a result of the 1972 and 1977 amendments to the Social Security Act. Similar calculations were made for the elasticity of SECA taxes; the elasticity of the SECA tax schedule is, on average, 25 percent lower than the elasticity of the FICA schedule because T t w y x n, , , ,( ) t∗ y∗x∗n w∗ 1 x–( )∗n+[ ]= t y x w n Ey y∗x( ) y∗x w∗ 1 x–( )+( )⁄= Ey a smaller share of the income earned by the self-employed is earned by those below the caps.29 The next step is to estimate the relative shares of the cyclical changes to aggregate wage and salary income that result from greater employment and greater income per worker. The percentage gap in wage earners and percentage gap in average wages are estimated by the following regressions (with t- statistics in parentheses): FICA: = .001 + 1.00 + .013∗ , (.23)(10.0) (3.74) adj. R2 =.72 = .000 + 1.031∗ , (.20) (12.5) adj. R2 =.79 SECA: = −.013 + 1.71∗ + .066 , (-.61) (2.43) (2.50) adj. R2 =.21 ∆ covemp( )ln ∆ emp( )ln law ∆ avecovwage( )ln ∆ avewage( )ln ∆ covemp( )ln ∆ emp( )ln law Table 8 FICA and SECA Tax Elasticities Ey Esocial Ey Esocial Year FICA SECA FICA SECA Total Year FICA SECA FICA SECA Total 1951 .49 .26 .81 .72 .80 1974 .61 .30 .85 .74 .84 1952 .45 .26 .79 .72 .79 1975 .60 .31 .85 .74 .84 1953 .41 .25 .78 .72 .77 1976 .60 .32 .85 .74 .84 1954 .40 .25 .77 .72 .77 1977 .60 .34 .85 .75 .84 1955 .46 .34 .80 .75 .79 1978 .58 .32 .84 .74 .83 1956 .43 .31 .79 .74 .78 1979 .68 .40 .88 .77 .87 1957 .41 .29 .78 .73 .77 1980 .71 .45 .89 .79 .88 1958 .40 .29 .77 .73 .77 1981 .73 .49 .90 .81 .89 1959 .45 .31 .79 .74 .79 1982 .74 .51 .90 .81 .90 1960 .43 .31 .78 .74 .78 1983 .76 .52 .91 .82 .90 1961 .41 .30 .78 .74 .77 1984 .75 .49 .91 .81 .90 1962 .39 .27 .77 .73 .76 1985 .75 .48 .90 .80 .90 1963 .37 .25 .76 .72 .76 1986 .75 .48 .91 .80 .90 1964 .35 .23 .75 .71 .75 1987 .74 .47 .90 .80 .90 1965 .33 .18 .75 .69 .74 1988 .72 .43 .89 .79 .89 1966 .48 .25 .80 .72 .80 1989 .73 .45 .90 .79 .89 1967 .45 .22 .79 .71 .78 1990 .75 .47 .90 .80 .90 1968 .52 .26 .82 .72 .81 1991 .77 .52 .91 .82 .91 1969 .47 .25 .80 .72 .79 1992 .76 .53 .91 .82 .90 1970 .45 .23 .79 .71 .79 1993 .77 .54 .91 .83 .91 1971 .41 .22 .78 .71 .77 1994 .80 .60 .92 .85 .92 1972 .45 .25 .79 .72 .79 1995 .78 .60 .92 .85 .91 1973 .52 .26 .82 .72 .81 1996 .78 .60 .92 .85 .91 50 The Automatic Fiscal Stabilizers = .027 + .24∗ , (3.30)(3.39) adj. R2 =.25, where: = covered employment, from the Social Security Administration, = civilian employment, = a dummy for changes in coverage, 1 for 1955, 1957, 1966, 1983, 1984, 1988, 1991, = the average wage for covered employment, from the Social Security Administration, = average wage: total wages and salaries divided by civilian employment, and = proprietor’s income divided by covered workers. As with the personal income tax elasticity estimates, the weights on and implied by the regressions move dramatically over time—especially when the sum of and is close to zero—and thus they are very sensitive to estimates of potential GDP. As before, we decided to use the average weight over time, which placed 62 percent of the weight on the employment term for FICA. The resulting point estimate for the weight on the employment elasticity for SECA was 1.1. This value seemed unreasonable and probably reflected the poor fit of the SECA equations, so we opted to use the weights from the FICA. Plugging this information into equation 17 gives the cyclical income elasticities of FICA and SECA, summarized in the Esocial columns in Table 8. The weighted average of these two elasticities is shown in the total columns. The elasticity of unemployment taxes to cyclical income was approached in a distinct manner. The unemployment insurance (UI) tax system has two key features. In most states, the wage cap is quite low: indeed, in twelve states the cap is $7,000, and the weighted average across states was only $9,000 in 1997.30 The second key feature of the system is that tax rates for firms are experience-rated. Thus, tax rates tend to rise for several years after a recession and fall during an expansion. To capture this endogenous behavior, we modeled the UI tax rate ( ) as a function of lagged unemployment rates and changes in federal tax laws concerning the Federal Unemployment Tax (FUTA) wage cap and statutory tax rate.31 Lagged changes in unemployment rates for four years and the change in the wage cap were significant, but changes in the statutory tax rate— which have been small and infrequent—had no explanatory power (with t-statistics in parentheses): = −.026 + .042 + .074 (-2.85)(4.25) (7.77) + .004 + .025 + .60 , (-.32) (2.65) (5.51) adj. R2 =.84 ∆ avecovwage( )ln ∆ avepro( )ln covemp emp law avecovwage avewage avepro En Ey ngap ygap UIrate ∆UIrate ∆URt 1– ∆URt 2– ∆URt 3– ∆URt 4– ∆WAGECAP Corporate profits taxes, excluding Federal Reserve earnings, are about 10 percent of federal revenues. Corporate profits tax liability is defined as the product of the average tax rate on income subject to tax and income subject to tax before credits , less tax credits : . The average tax rate is derived from the data, given the BEA’s estimates for the other three terms. Income subject to tax equals modified NIPA economic profits (corporate profits less Federal Reserve earnings and rest-of-world profits), , less adjustments, . The adjustments are losses and capital gains, which are added to , as well as tax-exempt interest, state and local corporate taxes, and deductions for loss carryovers, which are subtracted. These data are found in SOI Corporate Income Tax Returns and in the BEA’s reconciliation tables between IRS measures of profits and taxes and the NIPA economic profits and profits taxes. Tax credits are primarily for foreign taxes and the investment tax credit. The elasticity of corporate profits taxes to changes in modified corporate profits is determined as follows: (20) , where , and . The elasticity of income subject to tax with respect to modified corporate profits in equation 20 is found by estimating the cyclical sensitivity of the major adjustments to corporate profits (Table 9). The elasticities are calculated in two steps. In the first step, the adjustments and modified profits are regressed against the GDP gap and potential GDP.32 The elasticity with respect to GDP is estimated by evaluating the marginal change at mean GDP. Second, the elasticities of the adjustments with respect to GDP are divided by the elasticity of modified profits with respect to GDP to produce the estimates of the elasticity with respect to modified profits. When we plug these results back into equation 20, we obtain an average elasticity of income subject to tax with respect to modified profits of 0.8; the annual figures vary from 0.3 in 1982 to 0.96 in 1968 (Table 10).33 These estimates are similar to those of deLeeuw et al. (1980). The low elasticity reflects the importance of corporate losses, which is the only adjustment that causes the elasticity to fall below one. is the elasticity of the corporate profits tax rate. This is only slightly higher than zero because the corporate income tax is not very progressive and few corporate profits are generated by firms in the lower tax bracket.34 We have assumed that the elasticity of credits with respect to modified profits varies with the share of credits that are for foreign taxes (which appears to have a zero elasticity) and the share of credits owing to investment tax credits (with an assumed 1.0 elasticity). Combining the elasticities in equation 20 produces an overall CPT( ) τ( ) IST( ) C( ) CPT τ∗IST C–= CP ADJ CP CP( ) Ecpt cp, τ( ∗IST Eτ cp, Eist cp,+( )= C∗Ec c, p ) τ∗IST C–( )⁄– Eτ cp, Eτ ist, ∗Eist cp,= Eist cp, CP ΣADJi ∗Eadj cp,–( ) CP ΣADJi–( )⁄= Eτ ist, FRBNY Economic Policy Review / April 2000 51 Table 10 Corporate Income Tax Elasticities Year : Income Subject to Tax Relative to Modified Profits : Corporate Tax Accruals Relative to Profits Year : Income Subject to Tax Relative to Modified Profits : Corporate Tax Accruals Relative to Profits 1954 .94 1.00 1975 .76 1.11 1955 .94 1.01 1976 .84 1.23 1956 .94 1.01 1977 .88 1.28 1957 .90 .97 1978 .90 1.26 1958 .88 .96 1979 .85 1.29 1959 .92 .99 1980 .68 .90 1960 .85 .92 1981 .54 .64 1961 .89 .98 1982 .31 .22 1962 .90 .98 1983 .52 .60 1963 .90 .99 1984 .58 .67 1964 .92 1.01 1985 .63 .78 1965 .95 1.05 1986 .70 .88 1966 .95 1.06 1987 .66 .81 1967 .94 1.06 1988 .64 .82 1968 .96 1.08 1989 .58 .72 1969 .88 .99 1990 .49 .62 1970 .74 .86 1991 .48 .60 1971 .81 .96 1992 .59 .72 1972 .88 1.05 1993 .70 .85 1973 .87 1.08 1994 .70 .85 1974 .73 1.07 Eist cp, Ecpt ct, Eist cp, Ecpt ct, Table 9 Elasticities of Adjustments to Modified Corporate Profits Dependent Variable Modified Profits State Profits Taxes Tax-Exempt Interest Capital Gains Losses Loss Carryovers Constant 17.4 -2.23 -1.53 -0.25 -1.87 -0.96 (4.26) (-13.0) (-17.0) (-0.50) (-2.36) (-5.62) Gap -0.262 -0.003 -0.001 -0.048 0.038 -.005 (-3.34) (-0.94) (-0.72) (-4.97) (2.47) (-1.66) Potential GDP 0.063 0.006 0.005 0.007 0.015 0.004 (19.0) (44.5) (65.3) (17.2) (22.9) (30.5) Elasticity with respect to GDP at mean 3.75 0.58 0.37 7.05 -2.90 1.64 Elasticity with respect to modified profits N.A. 0.16 0.11 2.13 -0.88 0.50 Adjusted R2 .94 .99 .99 .93 .96 .98 Note: The sample period is 1956 to 1994. 54 The Automatic Fiscal Stabilizers Table 13 High-Employment Receipts Year HEB Receipts (Billions of Dollars) Actual Receipts (Billions of Dollars) Cyclical Receipts (Billions of Dollars) Cyclical Receipts (Percentage of GDPK) GDP Gap Response of Taxes to a 1 Percent GDP Change (Percentage of GDP) 1951 60.2 64.7 -4.5 -1.4 -3.7 0.37 1952 64.5 67.8 -3.3 -1.0 -2.9 0.33 1953 66.5 70.5 -4.0 -1.1 -3.4 0.32 1954 65.7 64.3 1.4 0.4 0.7 0.54 1955 69.3 73.2 -3.8 -1.0 -3.2 0.30 1956 75.9 78.6 -2.7 -0.6 -2.0 0.31 1957 82.2 82.6 -0.4 -0.1 -0.6 0.15 1958 85.7 79.5 6.2 1.3 3.8 0.34 1959 91.2 90.6 0.6 0.1 0.3 0.45 1960 99.3 97.0 2.3 0.4 1.5 0.29 1961 105.0 99.0 6.0 1.1 3.1 0.35 1962 109.1 107.2 1.9 0.3 1.1 0.29 1963 117.0 115.5 1.5 0.2 0.8 0.29 1964 114.5 116.2 -1.7 -0.3 -0.8 0.30 1965 119.5 125.8 -6.4 -0.9 -2.9 0.31 1966 131.5 143.5 -12.0 -1.6 -5.1 0.32 1967 144.8 152.6 -7.8 -1.0 -3.2 0.30 1968 167.2 176.9 -9.6 -1.1 -3.6 0.31 1969 192.0 199.5 -7.5 -0.8 -2.6 0.30 1970 199.5 195.1 4.4 0.4 1.0 0.43 1971 208.9 203.3 5.6 0.5 1.1 0.44 1972 230.5 232.6 -2.0 -0.2 -0.9 0.18 1973 250.9 264.0 -13.1 -1.0 -3.2 0.30 1974 299.5 295.2 4.4 0.3 0.8 0.34 1975 321.9 297.4 24.5 1.4 4.4 0.32 1976 357.6 343.1 14.5 0.8 2.3 0.33 1977 395.0 389.6 5.4 0.3 0.9 0.28 1978 438.8 446.5 -7.7 -0.3 -0.9 0.36 1979 504.1 511.1 -6.9 -0.3 -0.5 0.52 1980 581.9 561.5 20.3 0.7 2.6 0.27 1981 674.3 649.3 25.0 0.8 2.9 0.27 1982 696.9 646.4 50.5 1.4 7.0 0.21 1983 725.5 671.9 53.6 1.4 5.6 0.26 1984 757.5 746.9 10.6 0.3 1.4 0.19 1985 812.4 811.3 1.2 0.0 0.6 0.04 1986 850.5 850.1 0.5 0.0 0.4 0.02 1987 937.5 937.5 0.1 0.0 0.4 0.00 1988 983.9 997.2 -13.3 -0.3 -0.7 0.40 1989 1,056.9 1,079.4 -22.4 -0.4 -1.4 0.31 1990 1,124.8 1,129.8 -5.0 -0.1 -0.2 0.56 1991 1,192.8 1,149.0 43.8 0.7 2.9 0.25 1992 1,240.7 1,198.5 42.2 0.7 2.2 0.30 1993 1,307.1 1,275.1 32.1 0.5 1.8 0.26 1994 1,380.4 1,374.7 5.7 0.1 0.5 0.16 1995 1,466.1 1,460.4 5.8 0.1 0.6 0.14 1996 1,577.8 1,584.7 -6.9 -0.1 -0.2 0.42 1997 1,687.1 1,720.0 -32.8 -0.4 -1.3 0.31 1998 1,788.1 1,844.2 -56.1 -0.7 -2.2 0.31 FRBNY Economic Policy Review / April 2000 55 B. High-Employment Expenditures Among expenditures, only those transfers and grants that are oriented toward income support respond automatically to changes in economic activity. Among these, unemployment benefits rise rapidly during a downturn in activity. The number of beneficiaries of low-income and disability programs—such as Food Stamps, the Earned Income Credit, welfare (Aid to Families with Dependent Children, or AFDC, and Temporary Assistance for Needy Families, or TANF), and disability insurance—expand as well, but only to a small extent. The large retirement transfers are essentially unaffected by fluctuations in the economy.35 Unemployment benefits are available for involuntarily unemployed workers who were recently employed and meet certain criteria. In general, benefits can last for up to twenty-six weeks, or up to thirty-nine weeks under the extended benefits program for workers in areas with high unemployment. This permanent extended benefits program was instituted in 1970. The HEB excludes expenditures by the permanent program. However, both before and after that time, temporary extended benefits programs were enacted near the end of each recession. HEB estimates typically include these expenditures because they are not automatic; they result from discretionary policies. However, for some uses of the HEB it may be appropriate to exclude these payments as well. Table 15 provides a summary of the temporary programs. Unemployment benefits have become less sensitive to business-cycle fluctuations over the past two decades as the criteria for obtaining benefits have been tightened and the taxation of benefits effectively reduced their value. In 1975, 76 percent of the unemployed qualified for benefits, but this share had fallen to only 52 percent by 1992. Excluding the temporary extended benefits programs (but not benefits Table 14 Decomposition of Cyclical Taxes Billions of Dollars Year Total Cyclical Receipts Personal Taxes Corporate Income Taxes Social Insurance Indirect Business Taxes Year Total Cyclical Receipts Personal Taxes Corporate Income Taxes Social Insurance Indirect Business Taxes 1951 -4.5 -1.4 -2.6 -0.1 -0.4 1975 24.5 8.0 11.4 3.9 1.2 1952 -3.3 -1.4 -1.8 0.2 -0.3 1976 14.5 6.4 5.9 1.6 0.6 1953 -4.0 -1.7 -2.3 0.4 -0.4 1977 5.4 3.7 2.3 -0.8 0.2 1954 1.4 0.0 0.8 0.5 0.1 1978 -7.7 -1.4 -4.5 -1.6 -0.3 1955 -3.8 -1.1 -2.7 0.3 -0.4 1979 -6.9 -3.0 -1.6 -2.2 -0.2 1956 -2.7 -1.3 -1.3 0.1 -0.3 1980 20.3 7.8 9.6 2.0 0.9 1957 -0.4 -0.6 -0.2 0.5 -0.1 1981 25.0 13.3 6.9 3.8 1.0 1958 6.2 1.9 3.2 0.6 0.5 1982 50.5 30.4 8.6 9.0 2.5 1959 0.6 0.5 -0.1 0.2 0.0 1983 53.6 29.7 13.5 8.0 2.3 1960 2.3 0.7 1.4 -0.1 0.2 1984 10.6 10.2 3.0 -3.3 0.7 1961 6.0 2.2 2.9 0.5 0.5 1985 1.2 4.1 2.1 -5.4 0.3 1962 1.9 1.0 0.8 -0.1 0.2 1986 0.5 2.1 2.6 -4.5 0.2 1963 1.5 0.8 0.8 -0.2 0.1 1987 0.1 2.9 1.7 -4.8 0.2 1964 -1.7 -0.4 -1.1 -0.1 -0.2 1988 -13.3 -2.8 -4.5 -5.6 -0.4 1965 -6.4 -1.8 -3.6 -0.4 -0.5 1989 -22.4 -8.3 -8.0 -5.2 -0.9 1966 -12.0 -4.2 -6.3 -0.7 -0.8 1990 -5.0 -3.8 0.3 -1.5 -0.1 1967 -7.8 -3.5 -3.6 -0.1 -0.5 1991 43.8 16.5 15.1 9.8 2.3 1968 -9.6 -3.7 -5.4 0.2 -0.7 1992 42.2 19.8 11.7 8.9 1.8 1969 -7.5 -3.8 -3.6 0.5 -0.5 1993 32.1 13.8 12.8 3.8 1.6 1970 4.4 0.3 2.1 1.9 0.2 1994 5.7 4.4 2.3 -1.5 0.5 1971 5.6 1.5 1.7 2.1 0.3 1995 5.8 2.3 4.5 -1.6 0.5 1972 -2.0 -0.4 -2.3 0.8 -0.2 1996 -6.9 -1.4 -2.7 -2.5 -0.2 1973 -13.1 -4.8 -6.8 -0.8 -0.7 1997 -32.8 -12.2 -12.8 -6.6 -1.1 1974 4.4 -0.2 3.3 1.0 0.2 1998 -56.1 -24.1 -20.6 -9.4 -1.9 56 The Automatic Fiscal Stabilizers paid under the 1970 Extended Unemployment Compensation Act), a 1-percentage-point increase in the unemployment rate would boost unemployment benefits by about $5 billion in 1998 and would boost the permanent extended benefits program by varying amounts depending on the level of unemployment.36 The Aid to Families with Dependent Children program was never very cyclically sensitive. Its successor program, Temporary Assistance for Needy Families, is essentially a block grant to states and thus it is no longer sensitive to the business cycle from the federal government’s perspective. Our estimates of the cyclical response of AFDC are based on Blank (1997). She finds that a 1-percentage-point increase in the unemployment rate raises traditional AFDC caseloads (single-parent households) by 3½ percent over an eighteen- month period, which then declines to about a 2 percent increase after three years. About 10 percent of AFDC expenses are for AFDC-Unemployed Parent (AFDC-UP), a program for couples that appears to be much more cyclically responsive. AFDC-UP caseloads rise by about 20 percent during the first one and a half years, before easing to a 15 percent rise after three years.37 The following equation approximates the dynamic response of total caseloads to an increase in unemployment as estimated by Blank: (.006 + .006 + .006 + .006 + .006 + .006 − .003 − .003 − .003 − .003 − .003 − .003 ). A rise in the unemployment rate of 1 percentage point would boost AFDC payments by 5 percent after one and a half years and by only 2½ percent after three years. In its peak year— 1994—the federal government spent $13 billion for program ∆AFDC AFDC∗= ∆URt 1– ∆URt 2– ∆URt 3– ∆URt 4– ∆URt 5– ∆URt 6– ∆URt 7– ∆URt 8– ∆URt 9– ∆URt 10– ∆URt 11– ∆URt 12– Table 15 Temporary Unemployment Insurance Extended Benefits Year Provisions Expenditures 1958-59 Temporary Unemployment Compensation Act provided a voluntary program under which states could extend benefits for up to thirteen weeks. Financed by interest-free loans to the states. 2 million workers received $0.6 billion from June 1958 to April 1959. 1961-62 Temporary Extended Unemployment Compensation Act extended benefits for thirteen weeks. Financed by a temporary tax. 2.8 million workers received $0.82 billion from March 1961 to June 1962 1970 Extended Unemployment Compensation Act initiated permanent extended benefits program. Outlays under this program have been made every year. 1971-72 Emergency Unemployment Act provided thirteen weeks beyond the extended benefits period, for a total of fifty-two weeks. $0.6 billion in 1971 and 1972 1974-78 Emergency Unemployment Compensation Act of 1974 (plus three subsequent extensions) extended benefits for up to sixty-five weeks. $6.5 billion in 1975-78 1974 Emergency Jobs and Unemployment Assistance Act provided a temporary pro- gram for the uninsured: farm workers, domestic workers, and S&L employees. $2.5 billion 1982-85 Federal Supplemental Compensation Program (and six subsequent extensions) $9.3 billion: provided for up to fourteen weeks of assistance to workers who had exhausted $1.2 billion in 1982 their benefits. $5.4 billion in 1983 $2.3 billion in 1984 $0.7 billion in 1985 1991-94 Emergency Unemployment Compensation Act (and four extensions). $27.8 billion: $0.8 billion in 1991 $13.6 billion in 1992 $11.9 billion in 1993 $1.4 billion in 1994 FRBNY Economic Policy Review / April 2000 59 awards are equivalent. In each case, a rise in the unemployment rate of 1 percentage point raises awards by 2 percent.42 In the case of the DI program, new awards represent about 10 percent of the total caseload. For SSI, only half of the caseload is disabled working-age adults (the rest are disabled children and the elderly), and new awards are about 10 percent of this subset of the overall caseload. In 1998, expenditures on these two programs were $50 billion for DI and $30 billion for SSI. Thus, a 1-percentage-point increase in the unemployment rate would boost outlays by $0.1 billion in the DI program and by $0.03 billion in the SSI program.43 C. The High-Employment Surplus As shown in Table 16, in 1998 the actual unemployment rate was 1.1 percentage points below the CBO estimate of the NAIRU, which depressed expenditures by $7 billion, about 0.4 percent of total expenditures (a 4 percent increase in the affected programs).44 Most of the increase occurred as increased unemployment benefits (Table 17). To put this in context with receipts, a 1 percent fall in GDP is comparable to about a ½ percent increase in unemployment; thus, a 1 percent fall in GDP would boost expenditures by $3 billion, compared with a $30 billion reduction in receipts in the first year. Table 18 shows the effects of the business cycle on the budget surplus. Over the past decade, the cyclical component of the surplus has swung by 1.5 percentage points of GDP, from adding 0.8 percentage point to the deficit in 1992 to boosting the surplus by 0.7 percentage point in 1998. Table 18 Current Surplus (+)/Deficit (-) HEB Actual Cyclical HEB Actual Cyclical HEB Actual Cyclical HEB Actual Cyclical Year (Billions of Dollars) (Percentage of GDP) Year (Billions of Dollars) (Percentage of GDP) 1951 5.2 10.3 -5.1 1.6 3.1 -1.6 1975 -45.3 -73.9 28.6 -2.7 -4.3 1.7 1952 0.2 4.5 -4.3 0.1 1.3 -1.2 1976 -39.4 -57.2 17.8 -2.1 -3.1 1.0 1953 -2.7 2.4 -5.1 -0.7 0.7 -1.4 1977 -39.2 -46.3 7.1 -1.9 -2.3 0.3 1954 0.3 -1.2 1.5 0.1 -0.3 0.4 1978 -39.8 -31.7 -8.1 -1.8 -1.4 -0.4 1955 2.1 6.3 -4.2 0.5 1.6 -1.0 1979 -26.6 -18.5 -8.1 -1.0 -0.7 -0.3 1956 5.3 8.6 -3.3 1.2 2.0 -0.8 1980 -38.5 -60.9 22.4 -1.3 -2.1 0.8 1957 3.3 4.2 -1.0 0.7 0.9 -0.2 1981 -28.8 -57.8 29.0 -0.9 -1.8 0.9 1958 1.7 -5.5 7.1 0.3 -1.1 1.5 1982 -73.9 -134.7 60.8 -2.1 -3.9 1.7 1959 3.3 2.6 0.7 0.6 0.5 0.1 1983 -110.5 -174.4 63.9 -3.0 -4.7 1.7 1960 9.7 7.3 2.4 1.8 1.4 0.4 1984 -140.8 -156.0 15.2 -3.6 -3.9 0.4 1961 9.5 2.8 6.7 1.7 0.5 1.2 1985 -159.3 -163.0 3.7 -3.8 -3.9 0.1 1962 4.8 2.8 2.0 0.8 0.5 0.3 1986 -174.3 -177.5 3.2 -3.9 -4.0 0.1 1963 6.9 5.3 1.5 1.1 0.9 0.2 1987 -128.0 -128.9 0.9 -2.7 -2.7 0.0 1964 -1.1 0.9 -2.0 -0.2 0.1 -0.3 1988 -136.3 -121.3 -15.0 -2.7 -2.4 -0.3 1965 -3.7 3.4 -7.1 -0.5 0.5 -1.0 1989 -138.7 -113.3 -25.4 -2.6 -2.1 -0.5 1966 -10.6 2.6 -13.2 -1.4 0.4 -1.8 1990 -161.5 -154.7 -6.8 -2.8 -2.7 -0.1 1967 -17.5 -8.3 -9.2 -2.2 -1.0 -1.1 1991 -147.5 -196.1 48.5 -2.4 -3.2 0.8 1968 -14.1 -2.8 -11.3 -1.6 -0.3 -1.3 1992 -229.1 -280.9 51.8 -3.6 -4.4 0.8 1969 -0.7 8.7 -9.4 -0.1 0.9 -1.0 1993 -211.3 -250.7 39.4 -3.2 -3.8 0.6 1970 -10.7 -14.1 3.4 -1.0 -1.3 0.3 1994 -178.6 -186.7 8.0 -2.6 -2.7 0.1 1971 -19.6 -25.4 5.7 -1.7 -2.2 0.5 1995 -170.2 -174.3 4.2 -2.3 -2.4 0.1 1972 -22.8 -20.5 -2.3 -1.9 -1.7 -0.2 1996 -119.8 -110.3 -9.5 -1.6 -1.4 -0.1 1973 -25.6 -11.1 -14.5 -1.9 -0.8 -1.1 1997 -58.0 -21.1 -36.9 -0.7 -0.3 -0.5 1974 -13.6 -16.9 3.3 -0.9 -1.1 0.2 1998 9.3 72.8 -63.5 0.1 0.9 -0.8 60 The Automatic Fiscal Stabilizers VI. Conclusion This paper presents theoretical and empirical analysis of automatic fiscal stabilizers, such as the income tax and unemployment insurance benefits. Using the modern theory of consumption behavior, we identify several channels through which the optimal reaction of household consumption plans to aggregate income shocks is tempered by the automatic fiscal stabilizers. The insurance channel—through which higher anticipated income tax rates reduce the variance of uncertain future after- tax income—is effective, provided that the precautionary motive for saving is important and that individuals understand the implications of the government’s intertemporal budget constraint. The wealth channel—in which current income taxes are lower as a result of, say, a recession—is effective if individuals expect government purchases (rather than income tax rates) to adjust to maintain the government’s intertemporal budget constraint. This channel can also be effective if the rate used by individuals to discount future income tax hikes exceeds the government’s borrowing rate (as in the FRB/US model). The liquidity channel—in which lower current income taxes relax borrowing or liquidity constraints—is effective to the extent that such constraints are in fact binding for a nontrivial fraction of the population. To bring some evidence to bear on these issues, we present results from several empirical exercises using postwar U.S. data. Using standard time-domain techniques, we estimate elasticities of the various federal taxes with respect to their tax bases and responses of certain components of federal spending to changes in the unemployment rate. Such estimates are useful for analysts who forecast federal revenues and spending; the estimates also allow high-employment or cyclically adjusted federal tax receipts and expenditures to be estimated. Using frequency-domain techniques, we confirm that the relationships found in the time domain are strong at the business-cycle frequencies. Such results suggest the potential for the automatic fiscal stabilizers to play a quantitatively important role in the economic stabilization process. However, in one large-scale, macroeconometric model of the U.S. economy—FRB/US—the automatic fiscal stabilizers are found to play a modest role in damping the short-run effect of aggregate demand shocks on real GDP, reducing the multiplier by about 10 percent, although they have a somewhat larger damping impact (in percentage terms) on personal consumption expenditures. Very little stabilization is provided in the case of an aggregate supply shock. In light of the findings from the FRB/US simulations, perhaps the title and conclusion of our paper should be “The Automatic Fiscal Stabilizers: Quietly and Modestly Doing Their Thing.” Appendix: Empirical Results from the Frequency Domain FRBNY Economic Policy Review / April 2000 61 0 0.005 0.010 0.015 0.020 0.025 0.030 0.035 3.53.02.52.01.51.00.50 Spectral density Source: National Income and Product Accounts. Note: Growth rate, seasonally adjusted annual rate. Chart A1 Personal Receipts Second-Quarter 1946 to First-Quarter 1999 Frequency 0 0.05 0.10 0.15 0.20 3.53.02.52.01.51.00.50 Spectral density Source: National Income and Product Accounts. Note: Growth rate, seasonally adjusted annual rate. Chart A2 Corporate Receipts Second-Quarter 1946 to First-Quarter 1999 Frequency 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 3.53.02.52.01.51.00.50 Spectral density Source: National Income and Product Accounts. Note: Growth rate, seasonally adjusted annual rate. Chart A3 Social Insurance Contributions Second-Quarter 1946 to First-Quarter 1999 Frequency 0 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 3.53.02.52.01.51.00.50 Spectral density Source: National Income and Product Accounts. Note: Growth rate, seasonally adjusted annual rate. Chart A4 Indirect Business Taxes Second-Quarter 1946 to First-Quarter 1999 Frequency Endnotes 64 The Automatic Fiscal Stabilizers 1. For an opposing view, see Atkeson and Phelan (1994). 2. One notable exception is Auerbach and Feenberg (1999). 3. Blinder and Solow (1974) do not evaluate automatic fiscal stabilizers under the assumption of rational expectations. In a rational expectations macroeconomic model, McCallum and Whitaker (1979) establish that automatic stabilizers can be effective at stabilizing output; however, like those of Blinder and Solow, their results are not based on an explicit set of optimizing models for consumers and firms. 4. The basic Keynesian model result—that the automatic stabilizers, in fact, stabilize—obviously does not hold in versions of the model in which the aggregate demand multiplier is zero and stabilization is unnecessary. For example, the automatic stabilizers are irrelevant when there is a completely inelastic aggregate supply of goods (the full-employment version of the model) or flexible exchange rates. In these cases, flexible wages, prices, or exchange rates do the stabilizing. In addition to possibly being irrelevant, the automatic stabilizers may be a destabilizing force. An example involves forward-looking expectations (and thus deviates from the basic Keynesian framework). If an income tax rate is varied countercyclically (but not completely automatically, if Congress must first recognize that a recession is under way), employed households may optimally reduce labor supply at the start of a recession (in response to an anticipated increase in after-tax wages), further reducing output. Similarly, if an investment tax credit were varied countercyclically, firms might postpone investments at the start of a recession and accelerate them during booms, thereby exacerbating cyclical fluctuations. 5. Indeed, it is not clear how best to incorporate (expected) future taxes into the basic Keynesian framework; perhaps the present discounted value of the tax stream could be included as a component of private nonhuman wealth, itself a determinant of consumption demand. Alternatively, one could modify the simple textbook model to allow for dynamic and forward-looking elements along the lines of Blanchard (1981). 6. Chan (1983) and Christiano (1984) both allow for households to invest in a private, risky asset. We abstract from such investment opportunities. Also, implicitly households are allowed to borrow and lend at the risk-free interest rate, . 7. A way to motivate this setup is found in Aschauer (1985). In his model, utility is a function of effective consumption in period , , r i C∗i defined as , where is positive if and are substitutes. That is, for a given level of effective consumption, an additional unit of government spending will induce the individual to reduce private consumption by units. Defining , since . corresponds to our assumption . 8. For example, if income falls temporarily by $100, personal consumption should fall by about $5 (given econometric estimates of the marginal propensity to consume out of wealth); with a 20 percent income tax rate, taxes and hence government purchases would fall by $20. Thus, investment would fall by $75. The decline in investment likely will be larger if future income taxes are raised, rather than if government purchases are reduced, to maintain a budget balance. 9. The elasticity estimated in FRB/US probably captures discretionary changes in the tax code as well as endogenous changes in receipts. 10. When fully rational (rather than VAR) expectations are incorporated into the simulations, the model assumes that the government’s intertemporal budget constraint is satisfied by altering future income tax rates to stabilize the government’s debt-to-GDP ratio. 11. In addition, after-corporate-tax cash flow has a positive impact on investment in producers’ durable equipment and on personal consumption expenditures (via stock market wealth) in FRB/US. However, these channels of influence play only a minor role in the subsequent simulation results. 12. Note that there can be a slight tension between the expected federal funds rate generated by the VAR system and the “actual” federal funds rate resulting under either of the two monetary policy assumptions; that is, policy misperceptions are possible, at least in the short run. 13. Note that the table shows increases in the percentage deviation from baseline in real PCE; this translates into an increase in the PCE “multiplier” from about 0.2 to 0.3. 14. Although we have not explicitly considered non-oil-price supply shocks, results reported in Reifschneider, Tetlow, and Williams (1999) suggest that the role of the automatic stabilizers in the face of other supply shocks would differ somewhat from those described above. For example, in FRB/US, a productivity shock affects supply and demand (the latter by altering permanent income) and thus the impact of the automatic stabilizers on model multipliers would be intermediate to the separate demand and supply shock cases considered above. Ci θiGi+ θi C G θ U1 ∂U ∂C∗1⁄= ∂U1 ∂G1⁄ U11∂C∗1 ∂G1⁄ U11θ1 0<= = U11 0< ∂U1 ∂G1⁄ 0< U13 0< Endnotes (Continued) FRBNY Economic Policy Review / April 2000 65 15. The idea is that the sensitivity of the output gap (actual minus potential) with respect to an aggregate supply shock is greater the stronger the automatic stabilizers are in a simple textbook model of aggregate demand and supply. For example, with a negative aggregate supply shock that reduces desired output, actual output will also decline as prices rise; however, the price rise will be smaller—and hence the narrowing of the output gap more limited—the stronger the automatic stabilizers are. 16. See the appendix in Cohen (1999) for a review—aimed at the practitioner—of the key results of spectral analysis used in this paper as well as references to the literature. 17. We utilize PROC SPECTRA from SAS to generate the basic spectral densities and squared coherencies. We use kernel estimation of the spectrum with a bandwidth parameter of 4. The respective 95 percent confidence bands were programmed by us. On rare occasions, the squared coherencies will lie outside the lower 95 percent confidence band; this is possible because of the squaring operation. 18. Furthermore, as shown in Cohen (1999), other federal transfer programs—such as Social Security, Medicare, Medicaid, and Food Stamps—have low squared coherencies with the unemployment rate at business-cycle frequencies, implying that these programs are weak automatic stabilizers at best. 19. The CBO data are based on Bureau of Economic Analysis (BEA) estimates of GDP before the comprehensive revision, which was published in October 1999. Our estimates use the same data. 20. Indeed, this was a problem for deLeeuw et al. (1980), which they addressed by using time trends. Our difference approach creates stationary series and does not rely on deterministic time trends. That said, levels specifications, using cubic-spliced time trends, yield similar results for the coefficients on the terms. 21. Using the annual Statistics of Income (SOI) data on tax liabilities implies that we are estimating a liability elasticity. Both the NIPA budget estimates and the unified budget record taxes on a payments basis. Our estimates may not capture the precise timing of the changes in payments being estimated. For example, during a downturn in the economy, tax payments may be accelerated relative to liabilities. 22. Simplifying to the case where AGI equals adjusted personal income , equation 16 is obtained by taking the total differential of the tax function, —which implicitly allows tax revenues GDPGAP Eagi 1=( ) T F n y,( )= to respond differently to changes in the number of returns and changes in income per return—and dividing the resulting expression by the total differential of the aggregate income function, , all multiplied by . 23. The elasticity of each group equals the slope of the line traced out by the natural logarithms of average taxes and average income. The slope for an AGI group is estimated by calculating the derivative of the parabola defined by three points consisting of the group and the groups above and below. 24. Some deductions—mortgage interest, for example—may be more closely related to permanent income than cyclical income while other deductions—such as state and local income taxes—are closely related to cyclical income. Thus, our calculations may understate the true cyclical marginal tax rate. The lower tax rate for capital gains may also unduly reduce the effective cyclical marginal tax rate to the extent that realizations do not reflect cyclical factors. 25. The chief problem is that the weights become unstable when the gaps are very small. By contrast, our 0.5 estimate is consistent with the swings in the gaps—and the weights—from business-cycle peaks to troughs throughout the sample period. For example, we estimate that the gap in the number of returns swung from -1.0 in 1989 to 1.6 in 1991, while the gap in the average income per return swung from -0.5 to 2.0 over the same period. Thus, the changes in the gaps were approximately equal. Similar results were obtained across earlier business cycles. 26. In addition, there is a small amount collected for veteran’s life insurance, workmen’s compensation, CHAMPUS (the military health program for dependents), and private employer pension benefits (PBGC premiums). 27. The elasticity of federal employee retirement contributions is assumed to be zero because there have been no endogenous changes in federal employment or pay owing to the business cycle. The income elasticity of SMI is approximately zero because Medicare status is based largely on age. 28. Our analysis indicates that the distribution of income between those above the taxable wage cap and those below the cap is not sensitive to the business cycle. We developed two parameters that are sufficient to describe the distribution of wages to make OASDI tax calculations—the share of wage earners below the cap, and the ratio of wages of those above the cap to those below the cap. The former is not AGI n∗y= AGI T⁄ 66 The Automatic Fiscal Stabilizers Endnotes (Continued) Note 28 continued correlated with the business cycle; the latter has only a weak correlation. Thus, we ignore cyclical sensitivity of the income distribution. 29. For example, in 1997 6 percent of the self-employed had income exceeding the caps, and they earned 21 percent of total self-employed income. Among wage earners, only 5 percent were above the caps, and they earned 14 percent of total income. 30. Program specifics are legislated at the state level subject to general federal criteria as well as strong incentives to tax at least $7,000. 31. This exercise may also capture legislated changes by state governments in response to UI trust fund reserves. 32. This step is identical to the deLeeuw et al. (1980) procedure, which has potential econometric problems as the adjustments and potential GDP are nominal values in level terms. “Share style” equations showed no explanatory power. 33. The elasticity tends to fall during recessions owing to the rise in losses. 34. deLeeuw et al. (1980) estimated that the elasticity of the tax code declined from 0.08 in 1955 to 0.02 in 1979. We have assumed that it has remained at that level. 35. Medicare enrollments are insensitive to business-cycle fluctuations because enrollment is based largely on age. OASI enrollments and outlays are boosted during recessions because some workers take early retirement when faced with poor employment prospects. This factor would raise benefit payments by about 0.3 percent for each percentage- point change in the unemployment rate. However, OASI payments are held down by the effects of previous recessions because the additional claimants from those recessions receive lower benefits than they would have if they had retired at the normal age. Given that the present value of the benefit stream is approximately the same for those who take early retirement as it is for those who retire at age sixty-five, we have assumed that the net cyclical effect for the government is zero. 36. Until the extended benefits program is triggered by high levels of unemployment, an increase in the unemployment rate will have little effect on these expenditures. For example, in 1982 $2.5 billion was spent on extended benefits, but only $0.3 billion was spent in 1991, largely because the latter recession was milder. 37. This result appears to be dependent on the states included in the sample. The reported result is obtained when the sample is limited to the nineteen states that provided the AFDC-UP program continuously over the 1975-95 period. When the sample is enlarged to include states that were forced to initiate the program in the 1990s, the unemployment rate becomes insignificant. 38. The zeroing out of welfare abstracts from the small contingency program ($2 billion over five years) for states with high and rising unemployment. 39. The rest appears as lower taxes and is captured by our tax elasticity estimates. 40. The regressions for the personal income tax indicated that the number of nonsingle filers is not sensitive to the business cycle and the lion’s share of EIC beneficiaries is nonsingles. 41. The actual elasticity for the expenditure portion may be smaller, as the refundable portion (about $24 billion of the $28 billion in 1996) would be less heavily weighted in the phase-out region. 42. See Rupp and Stapleton (1995). 43. These calculations ignore any hysteresis that is probably especially evident in the DI program, where few leave the rolls. But if the rolls do tend to ratchet up over time, it is not clear that the increases owing to recessions should be included in cyclical measures. 44. The ultimate effect would be somewhat larger owing to the lagged response of these programs.