Exchange Rate Policy
Greek Myths

In this paper I investigate the relationships between wage adjustment, competitiveness, macroeconomic policy and aggregate fluctuations in a small open economy. Based on a model of an economy producing both traded and non-traded goods, and assuming that the traded goods sector is competitive while the non-traded goods sector is oligopolistic, I show that real wages in the traded goods sector are negatively related with competitiveness. Wage setting is what determines competitiveness, output and inflation, and is pivotal for the effects of macroeconomic policies. The model is estimated and tested for the postwar period, and is used to assess the macroeconomic experience and policy options in Greece.

 The model consists of four aggregate relationships, two of which consist the main innovations of this paper, and are derived in detail. The first is a GDP determination equation, derived from the supply side. The traded goods sector is assumed fully open to international competition and therefore a price taker. The non- traded goods sector, a large part of which may be the government, is assumed to be a price setter, setting prices as a mark-up on variable costs. It is also assumed that the wage differential between the two sectors is constant, a common assumption in the Scandinavian model for example. This structure implies that the ratio of output to capital is a positive function of the product wage in the traded goods sector, and also of government expenditure, which is taken as an index of production of public non-traded goods. In addition, competitiveness (defined as the relative price of traded goods) is a negative function of the product wage in the traded goods sector. From these two equations, one can derive a (positive) relationship between the output/capital ratio and competitiveness, given government expenditure. The two additional relationships are a wage determination equation, modelled traditionally as the familiar negative relationship between real wages and unemployment, and a trade deficit determination equation, modelled as a combination of the short- run absorption and elasticity approaches. The model determines output, the price level, competitiveness, wages and the current account.
The model is estimated and tested using data for the Greek economy over the period 1955-85. The model's predictions are consistent with the data, as the signs and orders of magnitude of the estimated parameters are in accordance with the predictions of the theory, there is no evidence of statistical misspecification, and the model's overidentifying restrictions are not easily rejected.
The Greek economy presents a particularly interesting test case for supply-side models of aggregate fluctuations, as after the early 1970s and especially after the second oil shock, it was out of step with most of the other OECD countries. After the fall of the military regime in 1974, the incoming conservative administration inflicted a wage shock in 1975-76, on top of the international oil price shock. This was repeated in 1982, when an incoming socialist government autonomously increased wages by a large percentage, and formally instituted a system of automatic wage indexation. In addition, there were three instances of steep devaluations, in 1975-6 (one disguised as acceleration of the crawling peg), and two discrete devaluations in 1983 and again in 1985. Greece suffers from one of the highest rates of inflation in the OECD area since 1974, while its rate of the inflation was lower than the OECD average previously. Since the first oil price shock it has also witnessed one of the sharpest falls in the rate of GDP growth and one of the highest average ratios of the current account deficit and the PSBR to GDP.
Whereas the experience of many other OECD economies has been extensively examined recently, there is very little formal econometric evidence on the relation between supply shocks, wage adjustment, competitiveness, and macroeconomic policy for Greece. In addition, because of the structural idiosyncrasies of the Greek economy, a number of influential Greek economists and commentators argue that sound macroeconomic policies such as those used in the more developed industrial economies are inappropriate for their own country.
This paper presents a simple two-sector macroeconomic model of a small open economy which is somewhat more general than the models that have been used so far to analyse macroeconomic adjustment in the presence of external input price shocks. The approach is innovative in that the determination of competitiveness focuses on the supply side and the labour market, and it treats government expenditure as an exogenous supply-side variable reflecting changes in the supply of non-traded goods. The results serve to dispel the myth that because Greece has a larger informal economy, and a smaller industrial sector, it should pursue radically different macroeconomic policies than its EEC and OECD partners.
To summarize the findings relating to short-run macroeconomic policy, assume that the economy initially experiences unemployment and a current account deficit:
A rise in government expenditure can contribute to the target of higher employment and, in the short run, higher living standards. The costs are the reduction in competitiveness, the acceleration of inflation, and the medium- and long-term deterioration of the current account as the traded goods sector is squeezed.
A reduction in labour and indirect tax rates can improve all macroeconomic targets in the short run. In the medium and long run, however, the increase in the public sector borrowing requirement which may result from such a tax cut could reverse the improvement in the current account, which is determined by the difference between investment and total savings, including the PSBR.
An incomes policy could also improve almost all macroeconomic targets, but at the cost of a fall in the workers' living standards. Employment and competitiveness will improve with a reduction in real earnings, and so will the current account. Moreover, the public sector borrowing requirement does not increase. Given the possible reactions by workers to a reduction in their living standards, this policy may not be sustainable in the medium term. This dimension cannot be fully analysed in what is essentially a short-term model but a fall in real wages, sustained in the medium term, may lead to a rise in investment and hence a rise in the marginal product of labour, a process which will subsequently allow an increase in both real wages and employment.
Finally, the effects of exchange rate policy depend crucially on the degree of indexation of nominal wages. With full indexation, exchange rate policy cannot affect competitiveness, and has no real effects even in the very short run. With partial indexation or with indexation on the price of non-tradables, the effects of devaluation resemble those of direct controls on nominal wage increases, except that inflation rises temporarily with a devaluation. Like incomes policy, exchange rate policy works by increasing competitiveness and reducing real wages. The costs are a rise in inflation and a reduction in the living standards of workers. Had there been no downward inflexibility of nominal wages and had there been mechanisms for successful implementation of an incomes policy, there would have been no need for an active exchange rate policy. Given this downward inflexibility, however, a temporary shock which necessitates a reduction in real labour costs can be achieved through exchange rate policy.
This paper is not concerned with the optimal design of stabilization policy. It may prove useful in the search for such a design as it puts forward a fairly comprehensive and yet manageable short-run model, which not only accounts successfully for the experience of the Greek economy since 1955, but contains enough policy instruments to allow formal theoretical and empirical evaluations of past and future macroeconomic policies.