Abstract
This Selected Issues paper on the United States examines the effect of the structure of the mortgage market on real housing activity and housing prices. The market-based financial structure has reduced the volatility of mortgage lending. Changes in the structure of the mortgage market have coincided with lower volatility of real housing activity. Regional income growth and unemployment rates have statistically significant and correct signed effects on housing prices. Tests of the relative importance of mortgage market structure and macroeconomic variables suggest an important effect from the financial structure.
VIII. Effects of Social Security and Tax Reform in the United States61
A. Introduction
1. The Administration has proposed two major fiscal initiatives in the past year—introducing Personal Retirement Accounts (PRAs) and tax reform. PRAs would allow individuals to divert some of their Social Security payments into private accounts while reducing their “traditional” benefits from the system. A presidential advisory panel has also been asked to report on ways to make the U.S. tax system simpler and more efficient.
2. This paper investigates the macroeconomic effects of introducing PRAs and reducing the taxation of capital. A two-country version of the Global Fiscal Model (GFM) is used to examine the impact of:62
Introducing PRAs. The paper first considers the effects of introducing PRAs without other fiscal measures; it then considers the introduction of PRAs coupled with measures to prevent PRA-related increases in government debt.
“Lockboxing” social security. Next, the effects of fiscal adjustment assuming that the level of government debt is lowered by the asset accumulation associated with Social Security surpluses is examined.
Eliminating the double taxation of saving. Finally, the paper explores how reducing the personal income taxation of capital income could affect the economy.
B. The Model and Calibration
3. GFM is a theory-based annual simulation model that has been developed specifically to examine fiscal issues.63 Its main features include:
A private sector that is assumed to be more “impatient” than implied by the government budget constraint. Because the private sector uses a discount factor that is higher than the real interest rate, the effects of future policy actions are discounted more rapidly than is implied by the government budget constraint. As a result, individuals are only partly Ricardian, and fiscal policy changes influence national saving.
Markets are not fully competitive. Firms and workers have some monopolistic power, so that prices and wages are above their perfectly competitive levels. Labor income taxes affect the work-leisure trade off. In addition, profits reflect both returns to capital and economic rents extracted by firms. Compared with the case of perfect competition, these rents reduce the distortionary impact of corporate and personal income taxes.
4. The model involves a stylized representation of the U.S. tax system and the structure of public expenditures. The analysis incorporates three taxes: a labor income tax levied on wage compensation; a corporate income tax levied on accounting profits of firms; and a personal income tax levied on labor income, accounting profits, government transfers, and interest income (on government bonds and net foreign assets). It is assumed that for all three taxes, there is a single marginal rate, which coincides with the average tax rate. Revenues raised by taxation are spent on lump-sum transfers to consumers, government consumption of nontraded goods, and servicing government debt.
5. Other main aspects of the model can be briefly summarized as follows:
Consumption and production are characterized by constant elasticity of substitution utility and production functions, respectively.
Labor and capital are used to produce either traded or nontraded goods.
Investment is driven by a Tobin’s Q relationship, with firms responding sluggishly to differences between the future discounted value of profits and the market value of the capital stock.
International capital mobility is perfect, implying the equalization of real interest rates across countries over time.
Wages and prices are fully flexible.
The model’s financial market block is highly stylized. There are two kinds of assets, namely government debt (which can be traded internationally) and equity (which is held domestically).
6. The model has been parameterized to reflect some macroeconomic features of the United States and the rest of the world (RoW). In particular, the size of the U.S. economy is assumed to be around one-third of the world economy. Consequently, U.S. policies have a substantial impact on the global rate of interest. The macroeconomic structure of the economy—the ratios of consumption, investment, government spending, wage income, and income from capital relative to GDP—is calibrated to the U.S. economy. Similarly, fiscal variables—taxes rates on capital, labor, and personal income, and government debt—have been calibrated to correspond to the U.S. fiscal structure. Behavioral parameters are based on micro-economic estimates (Box 1).
Parameterizing the Model
Key behavioral parameters are set equal across the United States and the RoW. These include parameters characterizing real rigidities in investment, markups for firms and workers, the elasticity of labor supply, the elasticity of substitution between labor and capital, the elasticity of intertemporal substitution, and the rate of time preference.1
Simulations examine the impact of changing the values of the following key parameters:
The sensitivity of labor supply to the real wage (Frisch elasticity). The baseline value (-0.04) is mid-range in the values found by micro-economic studies. Alternative simulations assume values around the upper and lower limits of these estimates (-0.08 and -0.01, respectively).
The elasticity of substitution between labor and capital in the production function. The baseline value is -0.8, with alternative simulations using values of -0.6 and -1 (the Cobb-Douglas case).
The elasticity of intertemporal substitution. The baseline value for this parameter that describes the sensitivity of consumption to changes in the real interest rate is -0.33. Parameter values in alternative simulations (-0.2 and -0.5) are consistent with the lower and upper end of microeconomic estimates.
The wedge between the rate of time preference and the yield on government bonds: This parameter—which determines consumers’ degree of impatience—has not been subject to extensive microeconomic analysis. Comparing real interest rates charged to consumers on credit card debt (the main source of unsecured loans where the lender takes the full risk of default) and government debt indicates that this wedge may be substantial. Based on this, we set the baseline value of the wedge to 10 percent, with alternative simulations using 5 percent and 15 percent values.
7. The impact of fiscal policy on real activity combines responses from aggregate supply and demand. Aggregate supply changes are triggered through the “distortion channel.” Fiscal policy influences include the impact of wage taxes on the incentive to work and the effect of corporate income tax rates on the rate of return of capital. Aggregate demand changes largely depend on fiscal policy’s effects on wealth and interest rates. The strength of the wealth channel is influenced by the level of consumer “impatience.” Fiscal policy also affects the global real interest rate, with consequences for investment and private savings, and spillovers on the RoW (see also Chapter 5 of this Selected Issues paper).
C. Introducing Private Retirement Accounts
8. PRAs are assumed to follow the Administration’s proposals closely:
Workers can divert part of their labor income taxes from social security
contributions into PRAs. Initially, the amount workers can divert is $1,000, which increases by $100 each year to a maximum of $4,000. PRAs would be phased in gradually for younger workers.
The PRA system matures after 45 years. It is assumed that workers up to the age of 45 can participate and that they retire at 65, so that PRAs start paying benefits after 20 years. However, contributions to PRAs exceed benefit payments for a further 25 years, when the youngest workers that participated at the start of the program (assumed to be 20 years old) reach retirement.
Withdrawals from PRAs result in equal reductions in government transfers. Reflecting the stylized nature of financial markets in the model, there is no equity premium to be exploited by owners of PRAs.
9. The model suggests no significant impact from PRAs on GDP, national saving, and financial markets, but indicates a significant increase in federal deficits and debt over several decades(Figure 1). As payroll contributions are diverted from the Social Security system to PRAs, government revenue declines markedly, falling by 3½ percentage points of GDP relative to the baseline after about 20 years. As a result, government debt is 40 percent of GDP above baseline after 20 years. However, when benefit payments from PRAs start, “traditional” Social Security payments decline by a corresponding amount, which allows government deficits and debts to fall. In the long-run, government debt exceeds the baseline by 30 percentage points of GDP (see Orszag, 2005, for a similar estimate).
Effects of Introducing Personal Retirement Accounts (PRAs)
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of Introducing Personal Retirement Accounts (PRAs)
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of Introducing Personal Retirement Accounts (PRAs)
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.10. As private saving through PRAs offset government dissaving, there is no impact on national saving. Real interest rates are virtually unchanged and there is little effect on investment. It should be emphasized that these results follow from the stipulation that workers cannot borrow against accumulated savings held in their PRAs. In this case, a shift from government to private saving does not affect perceived wealth, and there is no change in consumer behavior.
11. Introducing PRAs could, however, lead to perceptions of higher future transfer payments. Survey evidence suggests that workers, especially younger ones, are skeptical about the value of their future Social Security benefits, possibly reflecting the underfunded nature of the Social Security system. Placing contributions into individual accounts could be interpreted as an obstacle to reneging on future benefit payments. Workers could perceive this as an effective increase in their permanent income.64
12. To simulate this, we assume a smaller decline in Social Security benefit payments after PRAs start, financed by a delayed and gradual increase in labor income taxes (Figure 2). The results suggest a mild rise in consumption and output compared to the baseline. However, this effect dissipates when taxes increase to pay for higher transfer payments, leaving the economy worse off starting after 25 years, before returning to the baseline in the long-run.65
Effects of Introducing PRAs If Future Benefits are Assumed To Be Higher
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of Introducing PRAs If Future Benefits are Assumed To Be Higher
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of Introducing PRAs If Future Benefits are Assumed To Be Higher
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.13. Significant macroeconomic benefits may accrue when PRAs are accompanied by greater fiscal discipline that prevents PRA-related increase in government debt (Figure 3). By making future liabilities explicit, PRAs could increase pressure to offset the resulting increase in government debt. Assuming such deficit reduction is achieved, output falls modestly below the baseline over the short run. Over the longer run, higher government saving and lower government debt reduces the real interest rate and boosts investment. Simulation results suggest that the short-run effects are broadly invariant to whether taxes are raised on labor income, corporate income, or personal income, although labor income tax-based consolidation appears to yield quicker but smaller long-run benefits.
Effects of Introducing PRAs Accompanied by Debt Consolidation
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of Introducing PRAs Accompanied by Debt Consolidation
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of Introducing PRAs Accompanied by Debt Consolidation
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.14. The long-term gains of fiscal consolidation depend particularly on the planning horizon of consumers (Figure 4). Alternative parameterizations of labor income tax-based consolidation suggest greater long-run gains to GDP accrue if consumers discount future tax reductions more rapidly, which induces higher saving and capital accumulation. If capital and labor are closer substitutes, or consumers are less sensitive to changes in the real interest rate, fiscal consolidation leads to smaller output losses in the short run, but has a larger and earlier expansionary effect over the long term.
Effects of Debt Consolidation on Real GDP Under Alternative Parametrizations1
(Percent deviation from baseline)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: Fund staff calculation1 Debt consolidation via labor income taxes; 5 years refers to average impact during first five years, similarly for 15 years, and long-run is new steady state value.Effects of Debt Consolidation on Real GDP Under Alternative Parametrizations1
(Percent deviation from baseline)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: Fund staff calculation1 Debt consolidation via labor income taxes; 5 years refers to average impact during first five years, similarly for 15 years, and long-run is new steady state value.Effects of Debt Consolidation on Real GDP Under Alternative Parametrizations1
(Percent deviation from baseline)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: Fund staff calculation1 Debt consolidation via labor income taxes; 5 years refers to average impact during first five years, similarly for 15 years, and long-run is new steady state value.15. “Lockboxing” Social Security surpluses could yield further long-term benefits (Figure 5). PRAs could lead to fiscal discipline through recognition of future unfunded liabilities. However, the same objective could be achieved faster and more directly by tightening fiscal policy to preserve current and future Social Security surpluses—which accumulate to 28 percent of GDP through 2026—thereby helping to ensure that these assets are available to fund future benefits.66 In the short run, labor supply would decline, and both consumption and output would be modestly adversely affected. However, lower public debt would lead to lower interest rates, boosting investment and real GDP.
Effects of a Social Security”Lock Box”
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of a Social Security”Lock Box”
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of a Social Security”Lock Box”
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.D. The Effects of Reducing the Taxation of Savings
16. Since the personal income tax base includes profits of firms, the return to capital can be taxed twice. GFM reflects the traditional view that taxation of dividends negatively affects capital accumulation.67 Narrowing the personal tax base to labor and interest income—eliminating the personal income taxation of capital—should reduce economic distortions.
17. Eliminating the personal income taxation of capital in a revenue neutral manner has significant long-term positive effects (Figure 6). In the short-run, narrowing the personal income tax base to labor income while raising rates on the remaining tax base to prevent revenue losses causes a small decline in real GDP as higher labor taxes damp consumption. Over time, however, national saving increases substantially, the interest rate declines, and capital accumulation results in output increasing about 2 percentage points above baseline.68
Effects of Revenue Neutral Tax Reform that Lower Personal Taxation of Capital Income
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of Revenue Neutral Tax Reform that Lower Personal Taxation of Capital Income
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of Revenue Neutral Tax Reform that Lower Personal Taxation of Capital Income
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.18. Narrowing the tax base without increasing tax rates, however, is inferior to the status quo as it increases government debt. The consequences of this “failed reform,” which is assumed to be reversed after 5 years, are illustrated in Figure 7. As personal income tax rates are raised to repay the increase in government debt over the reform period and in the long run, there is effectively no beneficial impact on investment or GDP. Simulations in which the reversal of the reform occurs after 10 years (not reported for the sake of brevity) find even larger transition costs.
Effects of Non-Revenue Neutral Tax Reforms that Lower Personal Taxation of Capital Income
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of Non-Revenue Neutral Tax Reforms that Lower Personal Taxation of Capital Income
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.Effects of Non-Revenue Neutral Tax Reforms that Lower Personal Taxation of Capital Income
(Deviation from baseline in percent of GDP unless otherwise noted)
Citation: IMF Staff Country Reports 2005, 258; 10.5089/9781451839647.002.A008
Source: IMF staff estimates.19. The long-run benefits from revenue-neutral tax reform depend on the planning horizon of consumers as well as the substitutability between capital and labor(Table 1). A longer planning horizon (more patient consumers) suggests that incentives to save are stronger after the tax reform, which induces a larger reduction in the real interest rate, greater capital accumulation, and larger long-run gains in output. Higher substitutability between labor and capital amplifies distortions from the taxation of capital; thus, eliminating it can yield larger gains.
Impact of Different Parameters on Simulation Results
Long-term effect on GDP (in percent relative to baseline)
Effects on long-term GDP of eliminating the double taxation of dividends under alternative parametrizations (revenue neutral; see Figure 7 for baseline)
Impact of Different Parameters on Simulation Results
Long-term effect on GDP (in percent relative to baseline)
| Baseline 1/ | 2.35 |
| Longer planning horizon | 4.30 |
| Inelastic labor supply | 2.67 |
| Lower intertemporal elasticity of substitution | 2.82 |
| Higher elasticity of substitution between capital and labor | 3.65 |
| Introducing rule-of-thumb consumers | 2.57 |
| Lower elastacity of substitution between home and foreign goods | 2.33 |
Effects on long-term GDP of eliminating the double taxation of dividends under alternative parametrizations (revenue neutral; see Figure 7 for baseline)
Impact of Different Parameters on Simulation Results
Long-term effect on GDP (in percent relative to baseline)
| Baseline 1/ | 2.35 |
| Longer planning horizon | 4.30 |
| Inelastic labor supply | 2.67 |
| Lower intertemporal elasticity of substitution | 2.82 |
| Higher elasticity of substitution between capital and labor | 3.65 |
| Introducing rule-of-thumb consumers | 2.57 |
| Lower elastacity of substitution between home and foreign goods | 2.33 |
Effects on long-term GDP of eliminating the double taxation of dividends under alternative parametrizations (revenue neutral; see Figure 7 for baseline)
E. Conclusions
20. Personal Retirement Accounts (PRAs) are unlikely to yield significant macroeconomic benefits unless lower social security returns spur additional fiscal consolidation. If the introduction of PRAs is combined with fiscal consolidation to prevent a large increase in government debt, interest rates decline by about 50 basis points in the long run, and output increases to about 3 percentage points above the baseline. Even larger benefits are obtained if the social security surplus is placed in a lockbox while maintaining the same debt target.
21. Lowering the taxation of investment income is beneficial, but only if the reform is revenue neutral. A revenue-neutral personal income tax reform yields a long-run increase in GDP of about 2 percentage points above the baseline. In contrast, a reform that is not revenue neutral and needs to be reversed implies no long-run benefits and entails transition costs over the short to medium run.
References
Bayoumi, T., and others, 2004, GEM: A New International Macroeconomic Model, IMF Occasional Paper No. 239 (Washington: International Monetary Fund).
Bayoumi, T., D. Botman, D. Laxton, and D. Muir, 2005, “Jam Today or More Jam Tomorrow? On Cutting Taxes Now Versus Later,” Canada—Selected Issues, IMF Country Report No. 05/116 (Washington: International Monetary Fund).
Botman, D., and D. Laxton, 2004, “The Effects of Tax Cuts in a Global Fiscal Model,” Box 2.2 in World Economic Outlook, April (Washington: International Monetary Fund).
Botman, D., D. Laxton, D. Muir, and A. Romanov, 2005, “A New-Open-Economy-Macro Model for Fiscal Policy Evaluation,” (unpublished; Washington: International Monetary Fund).
Laxton, D., and P. Pesenti, 2003, “Monetary Rules for Small, Open, Emerging Economies,” Journal of Monetary Economics, Vol. 50, pp. 1109–46.
Orszag, P.R., 2005, “Borrowing from Future Social Security Benefits: the Administration’s Proposal for Individual Accounts,” Testimony for the U.S. House Committee on the Budget (February 9). Available on the Internet at http://www.brookings.edu/views/testimony/orszag/20050209.htm.
Prepared by Tamim Bayoumi, Dennis Botman (FAD), and Manmohan Kumar (FAD).
See Botman and others (2005) for a description of GFM. Botman and Laxton (2004) have applied the model to study the effects of U.S. tax cuts. Bayoumi and others (2005) examine the long-term benefits from reducing government debt by delaying tax cuts as well as issues of tax spillovers within the context of Canada’s rapid reduction in government debt.
However, if workers currently assume that the government will not fully meet its promises, this also implies that workers correspondingly should expect a smaller increase in future government debt or taxes in the absence of PRAs.
Financial markets may also expect PRAs to lead to an additional increase in government debt. Financial markets may underestimate implicit liabilities in anticipation of future reforms of benefits that reduce payment obligations of Social Security. If PRAs increase the estimated size of future liabilities by making implicit debt explicit, the risk premium on government bonds may increase.
Proposals for such a “lockbox” have been discussed repeatedly since at least the mid-1990s with the twin objectives of safeguarding the viability of the Social Security system and highlighting the positive contribution of Social Security to the budget. After 2026, when social security will record deficits, regulations require increasing social security contributions.
The new view argues that borrowing by debt issuance rather than equity issuance is the main form of financing of investment. Since debt financing is tax deductible, capital income is effectively taxed only once, and hence there is no need to reduce the personal income taxation of capital. This has little impact on the simulations in this paper, which focus on the macroeconomic consequences of reducing the taxation of personal capital income, rather than on the welfare implications of taxation across factors of production.
If tax reform results in a reduction in the taxation of overall savings, instead of capital income only, the benefits are smaller. The reason is that increasing labor income taxes to reduce taxes on interest income increases distortions in the economy.
