Upward nominal wage rigidity: extensions to wage agreements

July 12, 2017

In Portugal, as in many other countries in continental Europe, the collective wage agreements between trade unions and employer associations are systematically extended to non-subscriber workers and employers. Since these agreements establish wage floors for most job titles, their frequent extension is equivalent to setting a wide range of compulsory minimum wages, which are regularly adjusted upward. In some firms these extensions can result in a wage structure that may not be appropriate, causing fewer hirings or added dismissals. With some trepidation, we call this phenomenon “upward nominal wage rigidity” (UNWR), in close symmetry with the Keynesian notion of downward nominal wage rigidity.

According to our estimates, firms that are more heavily affected by the change in the bargained wage floors decrease their hiring rates and, more importantly, significantly increase their separation rates. We estimate that an increase of 10 percent in our measure of implied wage bill growth (the increase in the wage bill that firms have to undertake as result of the UNWR) leads to a small fall in the hiring rate (0.5 percentage points) but to a significant increase in the job separation rate (3.2 percentage points), leading to a considerable job destruction. We also estimate how those externally driven wage increases impact on the probability of firm exit:  a 10 percent increase in the wage bill generated by the increase in the bargained wage floors significantly increases the probability of firm closure (2.2 percentage points).

These empirical results should call into question the functionality of the architecture of the Portuguese wage setting system. In particular, it raises serious concerns with respect to the widespread use of extension mechanisms. Firms confront not only severe downward nominal wage rigidity because nominal wage cuts are forbidden, but also because of what we tentatively call “upward nominal wage rigidity”.

Click here to go to the paper by Paulo Guimarães, Fernando Martins and Pedro Portugal.


Share this content