What Does Ex-post Evidence Tell us About the Output Effects of Future Tax Reforms?

As a result of the global economic recession there is renewed interest in the effects of taxation on economic performance: many countries attempted to use tax cuts as a means to stimulate aggregate demand and now require future tax increases to reduce large public sector deficits partly in response to the crisis. Correctly forecasting the short‐run output effects and the long‐run growth impact of such tax changes is central from a policy perspective. In practice however, this task is complex: tax changes affect output through a number of transmission channels, and they impact both on demand and supply conditions. There is a large empirical literature estimating the effects of tax changes ex‐post that is a potentially useful source of information for policy makers. These studies can be broadly grouped according to whether they present structural estimates, which mostly refer to the medium‐ and long‐run relationships and employ panel methods, or whether they are based on time series for particular countries, generally refer to shorter run effects, make few structural assumptions and are primarily based on vector auto regressions. Both strands of the literature and the short‐ and long‐run effects of tax changes are usually considered separately, even though from a policy perspective and to understand potential trade‐offs, this division is not justified. However, irrespective of this distinction a general conclusion would be that, while most studies agree that changes in at least some types of taxation are associated with changes in growth rates or output levels in the short and long run, studies neither agree on the magnitude nor on the direction of the effects. The objective of this paper is to summarize and to evaluate existing ex‐post evidence on the shortand long‐run output effects (i.e. short‐run tax multipliers and long‐run growth effects) effects of changes in taxation from the point of view of a policy maker in an OECD country who wishes to predict the output effects of a tax reform ex‐ante. For this reason we propose a set of criteria against which we evaluate existing evidence that will ensure distinctness from a more traditional literature review. In particular, a given parameter estimate is useful for predicting the effects of tax reforms according to whether 1) it is relevant, 2) it can be replicated and 3) it is robust and reliable. While it is relatively straight forward to determine whether a particular study is relevant and whether the results are robust, determining whether the tax reforms can be replicated turns out to be much more complex and requires the papers to recognize the government budget constraint in the empirical specification (that every fiscal change needs to be somehow offset by some compensating fiscal item) among other issues. By applying these criteria we narrow down the number of papers that we consider as useful for policy. We include in the paper detailed summary tables that allow a more systematic and enlightened comparison of conflicting estimates between papers. Overall, our review suggests that based on the literature surveyed, at least the direction of the longrun and to a lesser extent of short‐run output responses to tax changes can be predicted with some degree of certainty. While the magnitudes of the estimated long‐run output effects differ, these differences can often be attributed to differences in the measurement of the tax burden or the assumed offsetting change. In contrast to this, differences in terms of the magnitude of time series studies are often more difficult to reconcile. In addition, the long‐run literature generally takes greater care to identify the exact fiscal change underlying the estimates contrary to the time series literature. Our review also suggests that there may be trade‐offs between short‐run output stabilization and long‐run growth depending on the tax reform considered, but more research focusing on this question is required.

Kneller, Richard and Florian Misch (2011), What Does Ex-post Evidence Tell us About the Output Effects of Future Tax Reforms?, ZEW Discussion Paper No. 11-029, Mannheim.