Model Properties - Tax and the Supply-Side

One important feature of the Virtual Economy model is the strong supply-side implications of changes in taxes. A cut in tax rates or an increase in allowances permanently reduces unemployment. While most macroeconomic models agree that cuts produce a short-run stimulus to the economy, there is disagreement over whether any permanent benefit is obtained. These differences are shown by experiments carried out on the full versions the Treasury (HMT) and London Business School (LBS) models held at Warwick.

Differences in model behaviour following changes to tax rates reflect the way in which wages are modelled. A stylised form of a typical wage equation may be written, omitting dynamics, as

w-p+t^e = a(tⁱ+t^e +t^d+t^m)+pr-b.u

where w, p and pr are the log of nominal earnings, producer prices and trend productivity respectively, u is the unemployment rate, t^d, tⁱ and t^e are the average direct, indirect and employers' tax rates (expressed in percentage form) respectively, and t^m is the 'tax' imposed by high import prices. These last four terms drive a 'wedge' between employers' real wage costs and employees' real consumption wages, whose long-run impact on the bargained wage outcome, and consequently on the NAIRU, the non-accelerating-inflation or 'natural' rate of unemployment, is measured by the parameter a. If a does not equal 0 then a permanent change in any of the wedge terms has a permanent effect on the NAIRU. In the model of Her Majesty's Treasury a=0.5, whereas in the London Business School model a=0.

Intuitively, if a is non-zero the gain from the tax cut is shared between employers and employees. Employers care about the cost of employing a worker while employees are concerned with take home pay. Following an income tax cut, firms can pay a lower gross wage (and can therefore afford to take on more labour) while workers enjoy a higher net wage.

The implications of these differences for unemployment and inflation following a reduction of one pence in the standard rate of income tax are shown in Figure 1. In this simulation real interest rates and the real exchange rate are held fixed at baseline values ensuring that cross-model differences are not due simply to differences in behaviour from these sources. It is also implicitly assumed that this tax cut is sustainable in terms of the public finances. The models reach different conclusions on both the speed with which unemployment is reduced and whether this fall is sustained. In the HMT model the existence of the long-run wedge effect ensures a fall in the unemployment rate of 0.34 percentage points by the end of the tenth year, with a gradual decline continuing although it is not clear where the inflation rate is going to settle down. In the LBS model the unemployment rate is within 0.05 percentage points of its baseline values by this time and the model appears to be settling down at a stable equilibrium. (This model has a longer baseline solution available.) This diagnostic simulation usefully demonstrates the sensitivity of the models to a different treatment of the labour market, but it is not a realistic representation of the policy environment currently experienced in the UK.

Changing nominal interest rates in each period by the same amount as the change in the inflation rate is a fairly crude monetary policy rule which in this case is seen to fail to control inflation. A more sophisticated approach uses the interest rate as an instrument through which to minimise the difference between the inflation rate and some target level over the control period. Figure 2 shows the results from the HMT model when this approach is adopted. The other important difference from the previous simulation is that the exchange rate is now determined endogenously in the model, so interest rate changes directly affect the exchange rate and hence prices. The main conclusion remains unchanged. In the Treasury model it is possible to reduce the NAIRU permanently in this scenario, while keeping inflation under control. However this result is entirely dependent on the presence of long-run wedge effects. Removal of these effects leads to the unemployment rate returning to its original level in the tenth year, as shown by line HMT* in Figure 2.

In summary, different treatments of the wedge in the wage equation lead to different conclusions about the long-run benefits of a possible income tax cut in the forthcoming budget. If it is believed that the equilibrium real wage is reduced then the NAIRU will fall. However an alternative view as expressed by Bean (1994) suggests that a permanent fall in tax rates and rise in the permanent income of workers gives an increase in the reservation wage which they are prepared to accept, offsetting the downward pressure on wages. This implies that no long-run effect from taxes should appear in the wage equation although this does not preclude the possibility of substantial short-run effects. The Treasury specification is nevertheless supported by econometric evidence.

Bean, C.R. (1994). European unemployment: a survey. Journal of Economic Literature, 32, 573-619.

Intro | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13