Real Wages, Business Cycles and the Speed of Adjustment of Employment in Manufacturing Sectors of Industrialized Countries
检验了Otani关于实际工资与经济周期正相关关系的假设，利用短期就业函数模型，通过比较不同国家就业调整速度来间接验证该假设。
In a recent paper, Otani (1978) presents empirical evidence on the relationship between real wages and output in the manufacturing sectors of fourteen industrialized countries. He estimates his coefficient 13 by regressing the percentage change in real wages on the percentage change in real output, finding a negative relationship that is significant at the 5% level for six countries, a non-significant negative relationship for five countries and a positive relationship for three countries (Germany, the Netherlands and the United States). Otani puts forward, but does not test, a hypothesis concerning the inter-country differences in the estimated values of 18. The present paper provides such a test and rejects Otani's hypothesis. Otani offers a tentative explanation of the positive correlation between real wages and business cycles in Germany, the Netherlands and the United States. He hypothesizes that, in these countries, it is too costly for firms to fire the existing employees during a recession and hire new ones during a boom (Otani, 1978, p. 303). Otani does not test this hypothesis. If we had data on the costs of adjusting employment levels in the different countries, then we could provide a direct test of the hypothesis; but such data are not available. However, an indirect test is possible. Following work by Brechling (1965), Brechling and O'Brien (1967), Solow (1968), Ireland and Smyth (1970) and others, there is a substantial body of literature concerned with the short-run adjustment of labor demand to cyclical fluctuations-the short-run employment function literature. This may be used to test Otani's hypothesis. Otani's hypothesis involves the existence of a possible disequilibrium condition, such that a firm may be producing at points off its long-run labor demand schedule. The short-run employment function model is a model specifically designed to handle disequilibrium. It has, as its basic equation, Et E,-, = X(E*t Et__) where E is the natural logarithm of employment, E* is the natural logarithm of desired or optimal employment (which is a function of output and other variables), and subscripts involving t indicate the time period. The proportion of the logarithmic difference between the optimal and actual employment levels eliminated in any period is X (where 0 < X < 1). Employment is only on the long-run labor demand schedule if the system is in equilibrium, otherwise E*t * E,-, and Et Et-, * 0; thus the model does specifically allow for labor inputs off the demand for labor schedule in disequilibrium. The smaller is X, the slower is the speed of adjustment; that is, the less responsive is employment to output variations. Solow (1968) has shown that the speed of adjustment is slower, the more costly it is to adjust employment. Otani's hypothesis is that countries with positive ,Bs, Germany, the Netherlands and the United States, have high costs of adjustment, relative to countries with negative 8s. It follows directly, therefore, that the Otani hypothesis implies that the countries with positive f3s will have low values of X relative to those countries with negative f3 coefficients. It is thus possible to use the short-run employment function evidence to test the Otani hypothesis. Table 1 gives Otani's estimates of / together with two sets of estimates of X: XB is from Brechling and O'Brien (1967); XS is from a set of results prepared as