| The state of the economy undoubtedly affects wage claims by trade unions. In particular, there are feedback effects from recent investment-capital decisions to wage setting. In the absence of binding contracts, one expects that a higher capital stock will lead the union to make a higher wage claim. Intuitively, since an increase in the capital stock increases the demand for labor, wage claims are expected to increase with the capital stock. In turn, this leads to under-investment.
However, the optimal response of wages to investment-capital decisions is also affected by changes in the socio-economic environment. For instance, changes in social security, employment protection legislation or macroeconomic policy are expected to systematically affect the behavior of wages.
The objective of this paper is twofold. First, it formulates the dynamic interaction between firms and trade unions. The model yields simple testable predictions about the optimal reaction of wages to the predetermined capital stock. It shows that this optimal reaction is nonmonotonic. It depends on the markup of market wages over reservation wages and the elasticity of wage income to wages.
The second objective of the paper is to test the theoretical predictions of the model for the greek manufacturing during 1954-1990. In Greece, there are (at least) two distinct labor market regimes during this period. An early period until 1974, during which unions were weak (1954-66) or state-controlled (1967-74). And a second period 1975-90, during which unions have been strong operating within an environment of relaxed fiscal policy and regulated labor market. These characteristics make Greece an ideal candidate for testing the applicability of our model.
The theoretical model is simple. We use a dynamic monopoly union model with capital accumulation in the spirit of Ploeg (1987). The interaction between firms and unions is modeled as a leader-follower game with the union acting as a leader. Then, applying standard dynamic programming techniques, we solve for Markov perfect equilibria. The solution specifies how wages react to the state of the economy.
The main theoretical results is as follows. If (i) reservation wages are close to market wages and (ii) wage income is insensitive to wages, then wages increase with the predetermined capital stock. Intuitively, if these conditions hold, the union can afford to take advantage of favorable states in the capital stock and go for higher wage claims. On the contrary, if these conditions do not hold, it is risky for the union to go for higher wages and in that case wages can decrease with the capital stock. This places the socio-economic environment as a key factor for wage behavior.
In the empirical study for the greek manufacturing 1954-1990, we estimate real wage equations as functions of the predetermined capital stock. Our methodology is as follows. First, we detect possible regime changes. According to our theory, regime changes can be associated with changes in the markup of market wages over reservation wages and/or changes in the sensitivity of wage income to wages. Second, we test whether the response of wages to the predetermined capital stock systematically changes as regimes change.
Our main empirical results are as follows. It is reasonably accurate to say that the postwar history of the greek labor market is divided into three regime periods. An early period 1954-66, the period of the dictatorship 1967-74, and a late period 1975-90. The distinguishing feature of these regimes, as we move from one regime to the next, is the diminishing markup of market wages over reservation wages. This diminishing markup is due to a constantly rising government consumption, which has allowed Greek government since 1975 to follow a very lax employment policy in the public sector (including public enterprises) where a de facto job is guaranteed.
Then, by estimating real wage equations, we test whether regime changes affect the feedback coefficients on the predetermined capital stock. Using alternative estimating techniques (multiplicative dummies, window regressions and Kalman filters), we show that wage claims increase (decrease) with the capital stock when the markup of market wages over reservation wages is low (high). Thus, the response of wages to the predetermined capital stock is significantly negative during 1954-66 (when the markup was high as a result of tight fiscal policy), but is significantly positive during the late period 1975-90 (when the markup was low as a result of loose fiscal policy). During the dictatorship 1967-74, when market forces played a minor role, the response is negative but insignificant. Finally, we report that there is no evidence of regime changes associated with conservative and socialist administrations (1975-81 and 1982-90 respectively). |
|
