اثرات تعادل عمومی مشوق های مالیاتی سرمایه گذاری

A new-Keynesian model with a nominal tax system is developed and used to study the macroeconomic effects of temporary tax-based investment incentives. Two claims regarding the effects of these incentives are examined: first that they are overstated in partial-equilibrium frameworks; and second that repeated use of such incentives by policymakers can ultimately be destabilizing. The results contradict the first claim and imply that the second claim is not general. The model is also used to compute the predicted effects of an investment tax incentive that has figured prominently in recent fiscal stimulus packages. Highlights ► Tax-based investment incentives are examined in a new-Keynesian model. ► Nominal rigidities can amplify the real effects of tax-based investment incentives. ► Partial-equilibrium analyses do not always overstate the real effect of incentives. ► Repeated countercyclical use of investment incentives can be destabilizing.

The past two U.S. recessions have been characterized by a shift towards more active countercyclical fiscal policy (see Auerbach, 2009). In each case, temporary partial expensing allowances on business equipment investment (also called bonus depreciation allowances) have been important components of the policy response. Despite this increased reliance on temporary expensing allowances as an instrument of countercyclical fiscal policy, virtually no attempt has been made to assess the impact of these provisions in a fully specified structural forward-looking general-equilibrium model, such as the new-Keynesian framework that now serves as the workhorse specification for analyzing macroeconomic stabilization policies. Elmendorf and Reifschneider (2002) use a forward-looking macromodel (the FRB/US model) to examine the effect of a permanent investment tax credit, but are unable to use their model to analyze the effect of a temporary credit. House and Shapiro (2006) consider the effect of a temporary bonus depreciation allowance; however, they do not analyze rational-expectations solutions, but instead use an approximation whereby a variable's future expectations are set equal to its steady-state value. In addition, their model is fully real, with no nominal rigidities. This paper adds a tax system with nominal depreciation allowances to an otherwise-standard new-Keynesian model, and uses the resulting framework to analyze the effect of temporary partial expensing allowances on investment and real activity. The paper investigates two generally accepted views concerning the effects of temporary tax-based investment incentives that have emerged from earlier analyses. First, Auerbach and Summers (1979) and Judd (1985) argue that partial-equilibrium analyses overstate the computed impact of temporary tax-based investment incentives. This conclusion turns out to be a property of purely real models—in which the real interest rate rises to offset some of the stimulus from partial expensing—and does not necessarily extend to models with nominal rigidities. In particular, with sticky prices and wages, the effect of a temporary expensing allowance on investment is larger in general equilibrium than it is in partial equilibrium. The second view—advanced by Auerbach and Summers (1979) and later analyzed by Christiano (1984)—is that such investment incentives can ultimately be destabilizing. Intuitively, if agents come to expect that such incentives will be put into place whenever the economy enters a recession, they will postpone their capital expenditures when a negative shock hits the economy—thereby weakening the economy further—until the incentives are enacted. Such instability can in fact emerge, although its existence depends on how adjustment costs are specified in the model. The model's quantitative predictions are also compared with the estimates obtained by House and Shapiro, 2006 and House and Shapiro, 2008 in their study of temporary bonus depreciation over the 2001–2005 period. The larger effects found by our model highlight the potential importance of nominal rigidities in driving responses to investment tax incentives.

This paper has analyzed the effects of a temporary tax-based investment incentive in the context of a fully specified general-equilibrium model with nominal rigidities. The analysis uncovered two noteworthy results. First, the presence of nominal rigidities has an important quantitative effect on the response of investment and output to changes in temporary partial expensing allowances, and needs to be accounted for when assessing the effects of these policies. Second, the conduct of certain types of fiscal policy can in some cases prove to be destabilizing over time, which suggests that the endogenous reactions of private agents to systematic government policy—a topic that takes center stage in the study of monetary policy—might be important to the analysis of fiscal policy as well. A natural next step would be to use the model in a more formal empirical exercise. The recent U.S. experience with partial expensing allowances suggests that numerous complicating factors—for instance, whether state and local governments follow the federal government in enacting parallel provisions in their own tax codes—can influence the real-world impact of these policies.19 An estimated version of the structural model developed here would permit a careful accounting of the forces that affected investment over this period—including the size of any deviation from the model's predictions, which could reflect the presence of these factors or some other departure from neoclassical investment theory.