The Incidence and Economic Effects of the Financing of Unemployment Insurance

Abstract

<p>This thesis deals with the incidence and economic effects of payroll taxes earmarked for unemployment insurance. A major objective is to provide an appropriate theoretical framework for a discussion of this issue. In cases where D.I. coverage is reasonably comprehensive, S6 that the D.I. tax can be regarded as a broad based tax, it is argued that the proper engine of analysis is basically the standard macroeconomic general equilibrium model. The macroeconomic effects of taxation, whether they originate from the demand or supply side, are regarded as an integral part of incidence analysis.</p> <p>The standard macro framework requires modification in one direction, which is a ~ore detailed development of the aggregate labour supply function. This reflects the view that the most important macroeconomic effects of unemployment insurance are likely to emerge from the supply side, via work incentives.</p> <p>A number of variants of a small macroeconomic model are developed, each incorporating an explicit modelling of a hypothetical D.I. system. Qualitative incidence results are obtained using the traditional method of comparative statics, while a quantitative dimension is added in static and dynamic simulation exercises with plausible parameter values drawn from the relevant econometric literature. Different versions of the model employ various alternative hypotheses about the way in which the labour market operates and/or different specifications of the aggregate labour supply function.</p> <p>The incidence results depend largely on the effect of payroll tax increases on labour supply. In the so-called neoclassical version of the model, for example, payroll tax increases reduce both participation and average weeks worked by participants, but tax and benefit rates are connected via the D.I. budget constraint, and benefit rate changes also effect labour supply. An increase in the benefit rate will tend to reduce average weeks worked by participants but to increase participation itself. Therefore a balanced budget increase in payroll tax rates has a potentially ambiguous effect on labour supply. If the net impact on lahour supply is negative we obtain the 'standard' incidence results. A balanced budget increase in payroll tax rates reduces output and emplo)~ent, increases the general price level, and reduces both capital and labour income. (Similar results also occur in other versions of the model in which the labour market does not clear due to (e.g.) real or money wage rigidity.) On the other hand, if the net impact on labour supply is positive we obtain 'perverse' results, increases in output and employment, reductions in the price level and so on.</p> <p>The comparative static analysis and simulation exercises enable us to identify the key parameters in the aggregate labour supply function, and their critical values. For plausible parameter values, chosen on the basis of the available empirical evidence, it would appear that perverse results are not likely. However, there is clearly a need for more empirical investigation in this area.</p> <p>These results conflict with the traditional view that labour bears the full burden of payroll taxation, but we conclude that this view depends heavily on the assumed inelasticity of the aggregate labour supply function. The latter assumption is demonstrated to be inconsistent with the bulk of the empirical evidence on labour supply.</p> <p>Finally, some attention is also paid to the aggregate demand effects of unemployment insurance, in particular the case where the savings propensity out of U.I. benefits is less than that out of private factor incomes.</p>