RDP 8611: The Effectiveness of Fiscal Policy in an Economy with Anticipatory Wage Contracts 6. Conclusions

The effectiveness of fiscal policy, in an explicitly non-Keynesian setting, turns on its interaction with monetary policy and the subsequent gap that may be created between current and core inflation. A surprising result of this paper is that the supply-side effectiveness can only exist in a model where goods markets always clear. The inflation gap can then be exploited to alter real wages and output; the vital mechanism is that wage setters suffer a politically-induced cost to the extent that their indexed wage contracts can not perfectly reflect longer term forces. If perfect indexation to long term forces were feasible, output would never deviate from steady state. Given the inefficiency, policy can be used to aid and complement optimal social wage contracts. The art of demand management in this paper is to manipulate the inflation gap to maximise the speed of adjustment of output back to steady state, without getting too close to the critical ceiling on core inflation.

Supply-side effectiveness is lost if goods and money markets do not clear instantaneously. In that case current inflation is no longer anchored by the need to maintain monetary equilibrium. With prices inflexible, but inflation free to take on any value, no inflation gap will emerge, wage setters will suffer no endogenous costs, and there will be no forces to push the system out of classical unemployment. Fiscal expansion will be merely inflationary.

Accepting the conjecture that the productivity slowdown in recent years was caused by supply shocks and inadequately responsive real wages, the result of output effectiveness of policy for flexible price models and of only inflation effectiveness for sticky price models may not mean that the policy recommendation is conditional on these two polar alternatives. Since goods markets can be broken down into flexible and sticky price sectors, a convex combination of the results of the two models may apply. It is an empirical question to decide upon the appropriate weights.^[11]

An important caveat arises if we permit capital accumulation. The crowding out of investment by fiscal policy is a vital issue which will have an important bearing on the results in the current paper. In particular, even if fiscal expansion does speed up real wage and output adjustment, the possible crowding out of investment will reduce future productivity. The critical policy issue would be to determine an optimal balance between current and future output losses.

In the same way, opening the economy to trade will be an interesting extension. From the demand-determined model of output by Buiter and Miller (1981, 1982), we know that fiscal and monetary policy can have serious effects on output through real exchange rate effects. For the sticky real wage model, one may also expect fiscal expansion to cause over-appreciation of the real exchange rate. The model could be developed to allow the exchange rate depreciation gap into the Phillips curve. A positive gap can be created after fiscal expansion, replicating the paper's results. If an external debt crisis arose necessitating improvements in the current account, fiscal policy and the wage process would become complementary forces for correction.

Finally, this paper has been built on the strong assumption that inflation is not something to be feared; it is an intermediate target variable which, under particular circumstances, may be exploited to minimize the period of time that output is away from steady state. However there is no doubt that inflation enters the objective function of many democratic governments. One cost of inflation, amongst others, is that it is an index of the closeness of the fiscal deficit to its maximum sustainable level. In that case, the inflation-output trade off can be optimally determined by constructing social indifference curves of the speed of adjustment of output and core inflation, and equating the marginal rates of substitution and transformation.

Footnote

Blundell-Wignall and Masson (1985) estimate a simultaneous equation model of the Federal Republic of Germany over the period 1973–1982. Their inflation equation assumes a breakdown between a flexible and a sticky price sector with the share of the latter estimated as 0.8 with a standard error of .05. [11]