The cost to society created by distortions in the market as a result of a tax is also known as

10.5.2 Elimination of tax wedges

The economic impact of tax distortions can be measured by the improvement in economic welfare when the tax wedges are eliminated. We first analyze the impact of distortions resulting from the taxation of income from capital. We consider the elimination of tax wedges among assets and among sectors. We also consider the elimination of wedges between rates of return before and after taxes. Specifically, we measure the gains from the following changes in the 2010 tax system:

Eliminate tax wedges between short-lived and long-lived assets within each sector.

Eliminate tax wedges between short-lived and long-lived assets in the business sector – corporate and non-corporate.

Eliminate tax wedges for short-lived and long-lived assets among all sectors – corporate, non-corporate and household.

Eliminate all tax wedges in the business sector.

Eliminate all tax wedges in the private sector.

Corporate tax integration.

Eliminate taxation of income from capital

Eliminate capital income taxes and sales tax on investment goods.

Eliminate capital income taxes and property taxes.

Eliminate capital income taxes, sales tax on investment goods and property taxes.

The social rate of return is the rate of return before all taxes, adjusted for inflation. This is calculated by subtracting the rate of depreciation from the price of capital services. The social rate of return includes the inflation-adjusted rate of return after all taxes, together with the tax burdens due to corporate income taxes, individual income taxes and property taxes. The tax burdens are partly offset by capital consumption allowances. In order to eliminate tax wedges among asset categories, we equalize the social rates to return by assigning an appropriate investment tax credit to each category of assets.

Table 10.11 shows the present values of capital consumption allowances z and the rates of economic depreciation δ. Capital consumption allowances are deductions from income for tax purposes and must be distinguished from tax credits, which are deductions from tax liabilities. Table 10.11 also shows the steady-state allocation of capital stock ω and prices of capital services PKS for the base case corresponding to the 2010 tax system. The tax credits required for the first six sets of changes in the 2010 tax system are presented in panel 2 of Table 10.12, along with the corresponding social rates of return and effective tax rates. For comparison base case figures are presented in panel 1.

Table 10.11. Steady state of the base case (rate of inflation: 4%)

Table 10.12. Elimination of interasset and intertemporal tax wedges (rate of inflation: 4%)

The welfare effects of the 10 tax reform proposals are summarized in Table 10.13. We begin with simulations based on a lump-sum tax adjustment to achieve revenue neutrality. This provides a standard of comparison for more realistic policies that achieve revenue neutrality by adjusting distorting taxes, such as labor income taxes, sales taxes and individual income taxes. We find that the welfare gain from the elimination of the tax wedges within the three sectors is $479.0 billion (2011 US$). Under lump-sum tax adjustment, elimination of tax wedges between the corporate and non-corporate sectors yields a welfare gain of only $40.7 billion.

Table 10.13. Welfare effects of tax distortion: 2010 tax law (billions of 2011 US$)a

The economic impact of our third tax reform proposal illustrates the substantial welfare gains from eliminating tax wedges between the business and household sectors. This is intuitively plausible, given the size of the tax wedges between these sectors. The estimated gain is $5347.8 billion. By contrast, the welfare gain from eliminating all tax wedges among business assets alone is only $303.9 billion. The fifth simulation eliminates all the tax wedges among sectors and assets, leading to efficient allocation of capital within each time period. The welfare gain is estimated to be $5567.0 billion. Most of this can be attributed to the elimination of tax wedges between business and household sectors, as in the third simulation.

The sixth simulation, corporate tax integration, is the key to President Bush’s Advisory Panel’s Simplified Income Tax Plan. In this simulation we eliminate tax wedges between the assets in the corporate and non-corporate assets by setting the social rates of return of corporate assets equal to the corresponding rates on non-corporate assets. The tax burdens on the corporate assets are unambiguously reduced without an offsetting increase in other marginal tax rates. The estimated welfare gains are $2320.2 billion, less than half the gains from eliminating all tax wedges among sectors and assets.

In the first six simulations we have focused on the distorting impact of tax wedges among sectors and assets. In the following four simulations, we estimate the welfare cost of tax distortions resulting from wedges between before- and after-tax rates of return. We eliminate the distortions caused by the taxes on capital income, including property taxes and sales taxes on investment goods. In the seventh simulation we set the effective tax rates on all forms of capital equal to be zero.

We find that elimination of capital income taxes at both individual and corporate levels generates a welfare gain of $5176.7 billion. Eliminating sales taxes on investment goods as well increases this gain to $5628.2 billion. Eliminating capital income taxes and property taxes produces a gain of $6054.0 billion, while eliminating taxes on investment goods as well generates a gain of $6421.8 billion.

Table 10.13 also shows that the magnitudes of welfare gains under alternative tax adjustments. Since the elimination of tax wedges is not revenue-neutral, changes in tax rates to generate the missing revenue can produce significant distortions. For this reason the welfare effects are very sensitive to the method for revenue adjustment. These effects are most sensitive to the choice between lump-sum tax adjustment and the distorting tax adjustments. The results are also somewhat sensitive to choices among the distorting tax adjustments, especially when the required revenue is large.

3.1.2 Equilibrium, Tax Distortions, and International Externalities

A competitive equilibrium for the model is a sequence of prices { rt,rt*,qt,qtg,qtg*, wt,wt*} and allocations {kt+1,kt+1*,mt+1,mt+1*,bt+1,bt+1*,xt,xt*,lt,lt*,ct,ct* ,dt+1,dt+1*} for t=0,…,∞ such that: (a) households in each region maximize utility subject to their corresponding budget constraints and no-Ponzi game constraints, taking as given all fiscal policy variables, pretax prices, and factor rental rates; (b) firms maximize profits subject to the Cobb–Douglas technology taking as given pretax factor rental rates; (c) the government budget constraints hold for given tax rates and exogenous sequences of government purchases and entitlements; and (d) the following market-clearing conditions hold in the global markets of goods and bonds:

ωyt−ct−xt−gt+(1−ω)yt*−ct*−xt*−gt*=0,

ωbt+(1−ω)bt*=0,

where ω denotes the initial relative size of the two regions.

The model's optimality conditions are useful for characterizing the model's tax distortions and their international externalities. Consider first the Euler equations for capital (excluding adjustment costs for simplicity), international bonds and domestic government bonds. These equations yield the following arbitrage conditions:

(8) (1+γ)u1(ct,1−lt)β~u1(ct+1,1−lt+1)=(1−τK)F1(mt+1kt+1,lt+1)mt+1+1−δ( mt+1)+τKθδ¯=1qt=1qtg,(1+γ)u1(ct* ,1−lt*)β~u1 (ct+1*,1−lt+1*)=(1−τK*)F1(mt+1*kt+1*,l t+1*)mt+1*+1−δ( mt+1*)+τK*θδ ¯=1qt=1qtg*.

Fully integrated financial markets imply that intertemporal marginal rates of substitution in consumption are equalized across regions, and are also equal to the rate of return on international bonds. Since physical capital is immobile across countries, and capital income taxes are residence-based, households in each region face their own region's tax on capital income. Arbitrage equalizes the after-tax returns on capital across regions, but pre-tax returns differ, and hence differences in tax rates are reflected in differences in capital stocks and output across regions. Arbitrage in asset markets also implies that bond prices are equalized. Hence, at equilibrium: qt=qtg=qtg*.

As shown in Mendoza and Tesar (1998), unilateral changes in the capital income tax result in a permanent reallocation of physical capital, and ultimately a permanent shift in wealth, from a high-tax to a low-tax region. Thus, even though physical capital is immobile across countries, perfect mobility of financial capital and arbitrage of asset returns induces movements akin to international mobility of physical capital. In the stationary state with balanced growth, however, the global interest rate R (the inverse of the bond price, R ≡ 1/q) is a function of β, γ and σ:

R=(1+γ)σβ,

and thus is independent of tax rates. The interest rate does change along the transition path and alters the paths of consumption, output and international asset holdings. In particular, as is standard in the international tax competition literature, each country would have an incentive to behave strategically by tilting the path of the world interest rate in its favor to attract more capital. When both countries attempt this, the outcome is lower capital taxes but also lower welfare for both (which is the well-known race-to-the-bottom result of the tax competition literature).

Consider next the optimality condition for labor:

u2(ct,1−lt)u1(ct,1−l t)=1−τL1+τC F2(kt,lt).

Labor and consumption taxes drive the standard wedge (1 − τW) ≡ (1 − τL)/(1 + τC) between the leisure-consumption marginal rate of substitution and the pre-tax real wage (which is equal to the marginal product of labor). Since government outlays are kept constant and the consumption tax is constant, consumption taxation does not distort saving plans, and hence any (τC, τL) pair consistent with the same τW yields identical allocations, prices, and welfare.

Many Neoclassical and Neokeynesian dynamic equilibrium models feature tax distortions like the ones discussed above, but they also tend to underestimate the elasticity of the capital tax base to changes in capital taxes, because k is predetermined at the beginning of each period, and changes gradually as it converges to steady state. In the model we described, the elasticity of the capital tax base can be adjusted to match the data because capital income taxes have an additional distortion absent from the other models: They distort capacity utilization decisions. In particular, the optimality condition for the choice of mt is:

(9)F1(mtkt,lt)=1+Φt1−τKδ′(mt),

where Φt=η(1+γ)kt+1−(1−δ(mt))ktkt−z is the marginal adjustment cost of investment. The capital tax creates a wedge between the marginal benefit of utilization on the left-hand-side of this condition and the marginal cost of utilization on the right-hand-side. An increase in τK, everything else constant, reduces the utilization rate.w Intuitively, a higher capital tax reduces the after-tax marginal benefit of utilization, and thus reduces the rate of utilization. Note also that the magnitude of this distortion depends on where the capital stock is relative to its steady state, because the sign of Φt depends on Tobin's Q, which is given by Qt = 1 + Φt. If Qt > 1 (Φt > 0), the desired investment rate is higher than the steady-state investment rate. In this case, Qt > 1 increases the marginal cost of utilization (because higher utilization means faster depreciation, which makes it harder to attain the higher target capital stock). The opposite happens if Qt < 1 (Φt < 0). In this case, the faster depreciation at higher utilization rates makes it easier to run down the capital stock to reach its lower target level. Thus, an increase in τK induces a larger decline in the utilization rate when the desired investment rate is higher than its long-run target (ie, Φt > 0).

The interaction of endogenous utilization and the limited depreciation allowance plays an important role in this setup. Endogenous utilization means that the government cannot treat the existing (predetermined) k as an inelastic source of taxation, because effective capital services decline with the capital tax rate even when the capital stock is already installed. This weakens the revenue-generating capacity of capital taxation, and it also makes capital taxes more distorting, since it gives agents an additional margin of adjustment in response to capital tax hikes (ie, capital taxes increase the post-tax marginal cost of utilization, as shown in eq. 9). The limited depreciation allowance widens the base of the capital tax, but it also strengthens the distortionary effect of τK by reducing the post-tax marginal return on capital (see eq. 8). As we show in the quantitative results, the two mechanisms result in a dynamic Laffer curve with a standard bell shape and consistent with empirical estimates of the capital tax base elasticity, while removing them results in a Laffer curve that is nearly-linearly increasing for a wide range of capital taxes.

The cross-country externalities from tax changes work through three distinct transmission channels that result from the tax distortions discussed in the previous paragraphs. First, relative prices, because national tax changes alter the prices of financial assets (including internationally traded assets and public debt instruments) as well as the rental prices of effective capital units and labor. Second, the distribution of wealth across the regions, because efficiency effects of tax changes by one region affect the allocations of capital and net foreign assets across regions (even when physical capital is not directly mobile). Third, the erosion of tax revenues, because via the first two channels the tax policies of one region affect the ability of the other region to raise tax revenue. When one region responds to a debt shock by altering its tax rates, it generates external effects on the other region via these three channels. Given the high degree of financial and trade integration in the world economy today, abstracting from these considerations in quantitative estimates of the effects of fiscal policy is a significant shortcoming.

Zero-Tax Economy Versus Existing-Tax Economy

An immediate implication from the expression for marginal loss, Eq. (13.6), is that if there are no tax distortions in the economy, the imposition of a marginal tax on one of the goods or factors does not generate a deadweight loss even to a second order of approximation. The level of all tax rates is either exactly or approximately equal to zero near the initial no-tax equilibrium such that the marginal loss is also (approximately) zero. In other words, the first marginal distortion is free. The intuition behind this result is that all resource transfers in response to the new marginal tax occur at values (approximately) equal to their marginal costs. If so, returning the tax revenue is sufficient compensation for the distortion.

Of course, the zero-tax, zero-loss result is just a theoretical curiosum. All developed countries have complex tax structures that raise substantial revenue. Thus, the policy-relevant conclusion to be drawn from Eq. (13.6) is that even a marginal tax change can generate substantial welfare losses, precisely because resources are shifting from an initial position in which marginal values may be far from their marginal costs. The government cannot choose to ignore the efficiency implications of minor changes in the tax structure simply because the changes are “small.”

5.3 Distributional considerations and the mcpf

With a heterogeneous population, the provision of public goods and the correction of externalities take on added complications. Even in the absence of distortionary taxation, the optimal rules then reflect the social valuations of utilities of different individuals. In addition, the costs and benefits of public goods, externalities, and the taxes that address them all have distributional consequences. For example, the government might wish to expand provision of public goods that have favorable distributional consequences; Sandmo (1998) offers a detailed analysis of the general problem. Also see Slemrod and Yitzhaki (2001), who illustrate how one can decompose both the costs and benefits of public expenditure projects in terms of efficiency and distributional consequences. However, it is also useful to consider circumstances in which the problem becomes much simpler, which is the case when the government has sufficient flexibility in its choice of tax instruments.

There is a close analogy here to the standard optimal income tax problem, under which it may not be necessary to tax luxury goods more heavily for purposes of distribution if the government can use a nonlinear income tax [as in Atkinson and Stiglitz (1976)]. Indeed, the analysis yields a parallel result, namely that distributional considerations should not enter into the provision of public goods or the correction of externalities when there is a nonlinear income tax and preferences are weakly separable into goods and leisure. This result is described by Kaplow (1996), building on previous work of Hylland and Zeckhauser (1979).

Kaplow’s observation is that the Samuelson rule for public-goods provision is unaffected by the presence of distortionary taxation when preferences are separable and the government uses a nonlinear income tax. The argument has two pieces. First, following the intuition given above for the proportional-tax case, there will be no change in labor supply, so that all of the expenditures on the public good come through reductions in the untaxed numeraire commodity. Hence, there is no tax wedge at the margin between public and private goods. Second, because of the availability of the nonlinear income tax, the distributional consequences of an increase in public goods spending can be offset, so that distributional weights will also be absent from the decision.

To expand on the reasoning Kaplow provides for his result, we present a detailed proof here. Suppose that households vary with respect to wage rates, w, but that each household’s preferences take the form U(υ(c, g), 1 − L), where c is private good consumption, g is the level of the public good, and L is labor supplied. Public goods are financed using a nonlinear tax T(wL;g) on labor income, where T1 is the household’s marginal tax rate. Consider an experiment in which g is increased, with taxes raised on each individual so that net utility is unchanged. (Continuing to spend and tax in this way will eventually lead to an optimal level of public goods provision, if the government persists to the point that marginal revenue from additional spending is zero.) The claim is that this policy results in no change in labor supply.

The household’s initial optimum labor supply decision implies that

(5.11)∂U∂L=U1υ1w−T1w−U2=0

and that Equation (5.11) holds as g changes:

(5.12)υ1dU1dG+ U1υ11dcdg+υ12w1−T1−U1υ1 wT11wdLdg+ T12−U21dυdg−U 22dLdg=0.

The claim is that Equation (5.12) holds with both U and L constant. Note that if U and L remain constant, so must υ, and hence U1. Thus, the claim implies that

(5.13)υ11dcdg+ υ12=υ11−T1T12

or, using dυ/dg = υ1 dc/dg + υ2 = 0 ⇒ dc/dg = −(υ2/υ1),

(5.14)∂υ2/υ1∂c=11−T1T12 .

By the assumption that L is fixed, dc/dg = dT/dg and dT/dg = T2. Thus, υ2/υ1 = T2. Moreover, this equality does not hold simply at a particular point, but rather at all points in the income distribution. That is, the functions υ2/υ1(c, g) and T2(wL;g) are equal for any value of c = wL − T(wL;g). Thus,

(5.15)∂υ2/υ1∂c=T21dwL dcg=T2111−T1.

Because T12 = T21, (5.14) holds, consistent with the initial claim.■

Just as in the case previously considered in Section 5.2, a parallel analysis applies to externalities, with the implication that, under the maintained assumptions regarding preferences and the use of the nonlinear income tax, no adjustment to the standard Pigouvian tax formula is warranted. While these results do depend on two key assumptions, those concerning the separability of individual preferences and the flexibility of the income tax, they are still quite important because they identify the source of deviations from the basic rules of Samuleson and Pigou. As discussed in this Handbook’s chapter 25 by Kaplow and Shavell, they also have additional implications regarding the extent to which government policies should be influenced by distributional issues.

3.3.3 Involuntary unemployment

The previous analysis assumed a well-functioning labor market with a flexible wage rate that assures full employment. In Europe, where involuntary unemployment is widespread, there has been considerable interest in green tax reform as a vehicle for reducing unemployment as well as improving the environment. This has prompted several studies investigating the effects of revenue-neutral environmental tax reforms in situations involving involuntary unemployment.

Bovenberg and van der Ploeg (1998a) analyze the consequences of an environmental tax reform in a model where involuntary unemployment stems from a rigid consumer wage. Hence this model incorporates labor-market distortions as well as the tax distortions already considered. In addition to labor, a clean non-labor production factor, which is fixed in supply, enters production31. Non-labor income is subject to a fixed tax rate of less than 100 percent.

The analysis shows that an environmental tax reform may reduce involuntary unemployment by expanding labor demand. Specifically, in the presence of a non-labor production factor, an environmental tax reform can shift part of the tax burden away from labor to the inelastically supplied non-labor factor. This tax-shifting effect exerts a positive impact on employment because it allows for a fall in wage costs, thereby boosting labor demand. In this model, there is a wedge between the marginal social value and marginal social cost of employment both because of the distortionary labor tax and because of the gap between the actual consumer wage and the reservation wage (i.e. the wage at which households would be willing to work). Hence the increase in employment from the environmental tax reform yields a first-order welfare gain.

These results are another example of how the prospects for the double dividend improve when the initial tax system is inefficient from a non-environmental point of view (see also Subsection 3.3.2). Since the non-labor production factor is fixed in supply, a 100 percent tax on non-labor income would be most efficient. However, if such a tax is infeasible for political or other reasons, a second-best alternative is to introduce the pollution tax, an implicit tax on the fixed factor (and labor). Substituting the pollution tax for the labor tax improves efficiency by shifting more of the burden of taxation onto the fixed factor.

Several papers [see, e.g., Koskela and Schöb (1999), Nielsen, Pedersen and Sørensen (1995), Schneider (1997) and Bovenberg and van der Ploeg (1998b)] explore how a green tax reform affects equilibrium unemployment in models with endogenous wage-setting. In these models, the impact on employment depends mainly on how an environmental tax reform affects unemployment benefits, which determine the threat point of employees in the bargaining process between employers and employees. In particular, if unemployment benefits are a fixed proportion of income in employment (implying a fixed replacement ratio), then all taxes are completely borne by employees [see Layard, Nickell and Jackman (1991)]. Hence, an environmental tax reform affects neither wage costs nor unemployment in equilibrium. The importance of the benefit regime applies to most equilibrium models of unemployment, including models of union–firm bargaining, monopoly unions, efficiency wages, and job search.

Koskela and Schöb (1999) illustrate these principles in the context of a model of wage bargaining between unions and employers. They show that the employment effects of an environmental tax reform involving pollution taxes on dirty consumption depend crucially on the taxation of unemployment benefits. In particular, employment may expand if unemployment benefits are neither subject to the labor-income tax nor indexed to the consumer price index. In that case, the unemployed pay the higher pollution taxes on consumption but are compensated neither by lower taxes on labor income nor by higher gross benefits. Indeed, whereas the pollution tax hits workers and unemployed alike, only workers benefit from the recycled revenues in the form of lower taxes on labor. In this way, the environmental tax reform shifts the tax burden away from workers towards the unemployed. This tax-shifting effect makes the outside option of unemployment less attractive for workers, thereby moderating wage costs and thus boosting labor demand.

Broader Tax Distortions

The broader fiscal system causes important distortions in labor and capital markets and, in some cases, in the pattern of spending across goods.

Taxes on labor income (e.g., personal income and payroll taxes) reduce the returns to work effort by lowering net take-home pay. For example, lower net wages may encourage the partner of a working spouse to stay home rather than join the labor force, an older worker to retire sooner, or an existing worker to put in less effort on the job or accumulate fewer skills. As a result, labor income taxes reduce the overall level of work effort below levels that would otherwise maximize economic efficiency. Personal taxes on dividend or capital gains income from savings, and corporate taxes on investment income, similarly reduce capital accumulation below economically efficient levels. And large tax exemptions and deductions in the United States create distortions by inducing too much spending on, for example, employer medical insurance and home ownership, and too little spending on ordinary (non-tax favored) spending.

Externality taxes interact with these sources of distortion in two important (and opposing) ways.

First, using revenues from externality taxes to reduce broader tax distortions produces gains in economic efficiency. And these gains can be substantial, relative to the net economic benefits from reducing the externality itself. Second, however, as externality taxes are passed forward into higher product prices, this increases the general price level. This reduces real household wages and the real return on capital which, in turn, reduces labor supply and capital accumulation in the same way that a direct tax on wages and savings/investment income does. Consequently, there is an economic cost from compounding tax distortions in labor and capital markets.

If there are no tax preferences, the general finding in the literature is that (with some qualifications) the net impact from shifting taxes off income and onto externalities is to increase the costs of preexisting taxes – that is, the gains from recycling revenues are more than offset by efficiency losses in factor markets from higher prices. As a result, the economically efficient tax is somewhat below the marginal external damage, but only moderately so. However, when account is taken of tax preferences, these findings flip around. That is, the gains from revenue recycling now dominate, as cutting income taxes helps to alleviate excessive incentives for tax-favored spending, as well as alleviating distortions in factor markets. Nonetheless, from a practical perspective, it is probably reasonable to impose taxes equal to marginal external damages in most cases, rather than trying to make adjustments to this rule for net interactions with the broader fiscal system, given the imprecision with which these broader linkages are measured.

The far more important point here, however, is the importance of revenue recycling. If externality tax revenues are not used to increase economic efficiency through cutting distortionary taxes (or funding socially desirable spending), the net benefits from these policies is greatly reduced and, perhaps, even eliminated. This is because the net benefits from correcting the externality itself can be largely offset and, in some cases, more than offset by the economic costs of compounding broader tax distortions through their impact on driving up prices and reducing the real returns to labor and capital.

An important concern about externality taxes, however, is that they are often regressive, meaning that they impose a larger burden as a portion of income for poorer households than for wealthier households (even when income is measured on a lifecycle basis). This problem might be addressed, at least in part, through recycling revenues in tax cuts that disproportionately benefit lower-income households (though this tends to imply smaller gains in economic efficiency from tax reductions). On the other hand, it might be argued that pricing externalities and poverty alleviation are distinct goals, requiring completely different instruments whose design should be kept separate rather than confounded. According to this view, for a given pattern of prices that emerges under corrective taxation, distributional goals should be achieved through targeted educational, health, and other policies to promote social mobility and lift people out of poverty.

To sum up this subsection (leaving aside complications discussed in previous sections), given an overall target level of public spending, specific taxes should be used to internalize all the main externalities, and the rest of this spending target should be met through broader tax policies. Earmarking externality tax revenues is generally not advisable on economic grounds (usually there is no relation between efficient externality tax rates and the efficient level of any earmarking), unless the earmarked spending generates comparable net benefits to those from using funds to cut distortionary taxes.

3.6 Concluding remarks

In this chapter, we presented the DGEM that Dale Jorgenson and I developed to evaluate the welfare effects of tax policy and government spending. Important features of the DGEM include the cost of capital, translog price functions, NL3SLS estimation of the consumer and producer models, and two-stage allocation of lifetime wealth. The intertemporal expenditure function based on the DGEM plays a central role in the welfare analysis of alternative tax policies.

The cost of capital approach provides an excellent framework for modeling taxation of income from capital. All of the current instruments of capital income taxation in the United States are represented in the cost of capital formulas. Indeed, the cost of capital can accommodate practically any reasonable proposals for taxation of income from capital. The cost of capital approach allows us to measure tax wedges and effective tax rates of income from capital. Although tax wedges and effective tax rates alone do not determine the efficiency costs of tax distortion, they provide useful information about the major sources of tax distortion in the allocation of capital.

Our analyses of the efficiency costs of tax revenues from various parts of the US tax system allow us to gain a general view of the structure of efficiency cost of taxation. We found that taxation of income from capital is most expensive in terms of the efficiency cost per dollar of tax revenue. It seems reasonable to conclude that the first priority of tax reform in the United States is to reduce the distortions caused by taxation of income from capital. In view of our analysis, the corporate tax cut of 2018 in the United States seems to be a step in the right direction.

To evaluate the welfare effects of tax reform proposals in a more realistic setting, we simulated the economy with four alternative tax adjustments. We obtain the largest welfare gain when all the tax wedges for capital allocation are removed and capital is efficiently allocated across the entire private sector. The welfare gain further increases as labor income tax is flattened. One important message from these simulations is that we do not have to reduce the effective tax rate of capital to zero to attain a high level of welfare.

One may be reasonably confident that we can design a number of tax reform proposals that attain high levels of welfare for the economy. However, efficiency is not the only virtue that makes a good tax policy. Equity, however it is defined, is another important virtue a good tax policy must support. In a democratic society, equitable taxation is not only an important virtue in itself, it is also essential for securing political support for tax policy.

In reality, we need to be prepared to sacrifice some efficiency for the equity of tax burden. Indeed it is possible to evaluate distributional effects of efficiency enhancing tax reform proposals. OLG models with multiple consumers distinguished by age and level of income can be used to analyze distributional effects of tax policy across age cohorts and income groups. Jorgenson et al. (1997) and Jorgenson et al. (2013) provide even more flexible framework in which distributional effects of tax policy can be analyzed with consumers distinguished by wage rate and many other attributes. The critical issue is to put together an efficient and equitable tax reform proposal that is attractive to majority of politicians and voters.

4.1 Connections to the Optimal Policy Literature

We begin by considering Ramsey policies where the policymaker has an ability to make credible promises about how they will behave in the future, before turning to time-consistency issues below. We start by building on Sims’ (2013) analysis. He considers a simple linearized model of tax smoothing under commitment in the face of transfer shocks and long-term debt. The policymaker can use costly inflation surprises as an alternative to distortionary taxation to ensure fiscal solvency. We extend that work in several ways. Specifically, we allow for a geometric maturity structure which nests single-period debt and consols as special cases, employ nonlinear model solution techniques, and allow for anticipated and unanticipated government spending shocks, in addition to transfer shocks. Nonlinear solutions allow us to consider the way in which the size of the debt stock, together with its maturity structure, influences the optimal combination of monetary and fiscal policy in debt stabilization. Innovations to the expected path for inflation can affect bond prices in a way which helps to satisfy the bond valuation equation even without any fiscal adjustment. These bond price movements are effective only if applied to a nonzero stock of outstanding liabilities so the optimal balance between inflation and tax financing of fiscal shocks depends on both the level of government debt and its maturity structure.

Without an ability to issue state-contingent debt or use inflation surprises to stabilize debt, Barro (1979) showed that debt and taxes should follow martingale processes to minimize the discounted value of tax distortions. While Barro did consider the impact of surprise inflation on the government's finances, these were treated as exogenous shocks rather than something that can be optimally employed to further reduce tax distortions. Lucas and Stokey (1983) is an equally influential paper that reaches quite different conclusions on the optimal response of tax rates to shocks. Lucas and Stokey consider an economy where the government can issue real state-contingent debt and show that it is optimal for a government to issue a portfolio of debt where the state-contingent returns to that debt isolate the government's finances from shocks so that there is no need for taxes to jump in the manner of Barro's tax-smoothing result. Instead, taxes are largely flat and inherit the dynamic properties of the exogenous shocks hitting the economy.

A large part of the post-Lucas and Stokey literature considers the implications of debt that is not state contingent, as well as ways of converting the payoffs from portfolios of nonstate-contingent debt into state-contingent payoffs. A key result is that when debt payoffs are not (or cannot be made) state contingent, then the optimal policy looks more like Barro's tax-smoothing result. Aiyagari et al. (2002) show this by assuming that debt is single period and noncontingent in a model otherwise identical to that of Lucas and Stokey. How might noncontingent debt instruments be made to mimic the payoffs that would be generated by state-contingent debt? Two approaches have been suggested in the literature. First, surprise inflation can render the real payoffs from risk-free nominal bonds state contingent. For example, Chari et al. (1994) use a model where surprise inflation is costless to show that the real contingencies in debt exploited by Lucas and Stokey could be created through monetary policy via surprise inflation when government debt is nominal. This underpins Sims’ (2001) results in a model with costless inflation in which tax rates should be held constant to finance any fiscal shocks solely with surprise movements in inflation.

When we start to introduce a cost to surprise inflation, the optimal policy can be strikingly different. For a jointly determined optimal monetary and fiscal policy operating under commitment, Schmitt-Grohé and Uribe (2004) show that in a sticky-price stochastic production economy, even a miniscule degree of price stickiness will result, under the optimal policy, in a steady-state rate of inflation marginally less than zero, with negligible inflation volatility. In other words, although the optimal policy under flexible prices would be to follow the Friedman rule and use surprise inflation to create the desired state contingencies in the real payoffs from nominal debt, even a small amount of nominal inertia heavily tilts optimal policy toward zero inflation with little reliance on inflation surprises to insulate the government's finances from shocks. As in Benigno and Woodford (2004) and Schmitt-Grohé and Uribe (2004) we return to the tax-smoothing results of Barro (1979) thanks to the effective loss of state-contingent returns to debt when prices are sticky. Sims (2013) argues that this may be due to the fact that Schmitt-Grohé and Uribe only consider single-period debt; with longer term debt the efficacy of using innovations to the expected path of inflation to affect bond prices would be enhanced. This is the first issue to which we turn: to what extent will the optimizing policymaker rely on fiscal theory-type revaluations of debt through innovations to the expected path of prices?

While the state contingencies in real bond payoffs can be generated through the impact of surprise inflation on nominal bonds, an alternative approach when bonds are real is to exploit variations in the yield curve to achieve the same contingencies for the government's whole bond portfolio. With single-period risk-free real bonds, Ramsey policy in the Lucas and Stokey model possesses a unit root as in Barro. Angeletos (2002) and Buera and Nicolini (2004) use the maturity structure of nonstate-contingent real bonds to render the overall portfolio state contingent. With two states for government spending, for example, a portfolio of positive short-term assets funded by issuing long-term debt can insulate the government's finances from government spending shocks. More generally, with a sufficiently rich maturity structure the policymaker can match the range of the stochastic shocks hitting the economy and achieve this hedging. The second broad optimal policy question we consider is: what is the role of debt management in insulating the government's finances from shocks?

Having looked at the ability of the Ramsey policymaker to both hedge against shocks and utilize monetary policy as a debt stabilization tool when complete hedging is not possible, we turn to consider the time-inconsistency problem inherent in such policies. We find that constraining policy to be time consistent radically affects the policymaker's ability to hedge against fiscal shocks and generates serious “debt stabilization bias” problems, as in Leith and Wren-Lewis (2013), that are akin to the inflationary bias problems analyzed in the context of monetary economies.

We begin by considering the role inflation surprises play in optimal policy in our simple endowment economy with a geometrically declining maturity structure. We then generalize these results to a more general maturity structure and consider the role of debt management in hedging for fiscal shocks. We then turn to a simple example where complete hedging is feasible.

8 Proposals for Tax Reform

A number of proposals have been advanced for basic reform of the income tax system, frequently motivated by the desire to improve the terms of the equality–efficiency tradeoff. For example, is it possible to achieve the same degree of redistribution at a lower efficiency cost? Tax reforms in a number of Western countries since the 1980s have tried to move in this direction by decreasing the formal degree of progressivity in the tax system, while at the same time eliminating many of the deductions and exemptions that had contributed to a lower degree of effective progressivity. Studies of the American (Auerbach and Slemrod 1997) and Swedish (Agell et al. 1996) reforms of 1986 and 1991 do in fact conclude that their overall effects have been to leave the redistributive effects of the tax system virtually unaffected, while removing many of the tax distortions both in labor and capital markets.

A more radical proposal for reform is the adoption of a so-called flat tax. There are several versions of this, but one that was proposed by Hall and Rabushka (1995) and has been much discussed in the USA involves imposing a constant marginal tax rate on all income above a certain exemption level. Although the name of the tax indicates that the major point in its favor is the simplification of the tax structure, probably its more important features are the elimination of a large number of deductions and a more uniform taxation of income from capital. In this it is in line with the general trends in the reforms of the 1980s and 1990s.

Another reform, the intellectual history of which goes back a long time and which was much discussed during the late 1970s and early 1980s, was the possible adoption of an expenditure tax (Meade Committee 1978, Kay and King 1990). Although the expenditure tax is often seen as an alternative to income tax, it may also be regarded as simply a particular form of income taxation. Briefly, an expenditure tax is an income tax with full deduction for all saving, but with negative saving being added to income. What is taxed is accordingly expenditure or consumption, not income. There are two main claims for the expenditure tax. First, since differences in the standard of living between persons are primarily related to their consumption, not income, consumption is the more logical base on which to level a progressive tax. Second, the treatment of saving and dissaving leads to a more uniform and therefore more efficient taxation of capital income.

Although neither the flat tax nor the expenditure tax have so far achieved any direct political success, the political and economic debates over these proposals may indirectly have had a considerable influence on the reforms which were actually adopted in the following years.

3.4 Notes on the literature

The theory of voting over social security has followed different approaches. Browning (1975) and Boadway and Wildasin (1989a,b) study the determinants of social security in voting models with commitment, where all voters have the same income and differ only in age. Cukierman and Meltzer (1989) consider public debt (equivalent to social security in their model) in an overlapping-generations economy with income heterogeneity, weak altruism within the family, and policy commitments. Tabellini (2000) formulates a median-voter model with income heterogeneity and weak altruism within the family, but no commitment (i.e., in each period, voters choose a tax rate with lump-sum transfers to the currently old). The model of the present section, where voters differ in age and income, but where there is commitment and no altruism, combines features of all these approaches. These results are perhaps closest to those of Cukierman and Meltzer (1989), though that paper focuses on general-equilibrium effects on the real interest rate and neglects tax distortions. General-equilibrium effects and their effect on voters’ preferences have also been studied by Cooley and Soares (1999). Conesa and Krueger (1999) include in their analysis not only general-equilibrium effects, but also social-insurance benefits of the pension system. A general survey of the positive political theories of social security is provided by Verbon (1988), while Feldstein (1998) and Siebert (1998) discuss the recent reform experiences of various developing and industrial countries.

In the absence of policy commitment, social-security systems can be sustained by reputational equilibria. This idea was pursued by Kotlikoff, Persson and Svensson (1998), and more recently by Boldrin and Rustichini (2000), Cooley and Soares (1999) and Azariadis and Galasso (1997). The idea that altruism within the family also induces voters to support intergenerational redistribution is investigated by Tabellini (1991, 2000).

Some papers have studied the political determinants of social security in settings different from voting. Grossman and Helpman (1998) consider a model where members of different generations lobby the government, as in Part II below. Earlier papers relying on the idea that the ability of different generations to influence the political process affects the size and viability of social security include Patton (1978), Stuart and Hansson (1989) and Loewy (1988). More recently, Lambertini and Azariadis (1998) have focused on legislative bargaining among (representatives of) different interest groups. Mulligan and Sala-i-Martin (1999b) and Galasso and Conde Ruiz (1999) study multidimensional aspects of pension policy in slightly different political models.

The validity of the empirical prediction that more inequality leads to more spending on social security has been investigated by Lindert (1994, 1996) with negative results, whereas Tabellini (2000) obtained more encouraging results. Looking at data of Swedish municipalities, Strömberg (1996) finds support for the prediction that the composition of social spending is systematically related to the age of the median voter. The opinions of European citizens towards welfare-state programs and pension systems have been studied in Boeri, Börsch-Supan and Tabellini (2001).