On 15 January, workers at the historic “Mirafiori” plant in Turin voted to approve a new labour contract that had been signed by Fiat and by various trade unions but rejected by the powerful, leftist union of metal workers. The vote was close and workers did not fully follow their leadership; 54% voted for the reform – a percentage significantly lower than share workers belonging to the unions that signed the new contract (60-65%).
The new contract provides an exception to the current system of industrial relations which is based on a “national contract”. The key deviations are part of an exchange. On one hand, it allows more flexibility in work shifts, tougher controls on absenteeism, stricter rules for illness-leave, and a reduction in work breaks. On the other hand, the Fiat CEO, Sergio Marchionne, promised new investments (up to €1 billion after 2012), and higher wages (possibly €3,500 a year more). Mr. Marchionne welcomed the result by saying that the vote proved that workers had “confidence in themselves and in their future”.
The debate leading to the vote has been divisive and heated. Those in favour welcome this as an epoch-defining change in industrial relations – the end of the 1970s style class struggle. It settles the stand-off that long involved what many saw as a frontal attack on workers’ rights based on outsourcing “blackmail”.
The debate has mostly focused on issues of workers’ representation. In a functioning system of industrial relations, workers should delegate their elected representatives to negotiate with the firm. In the present case, paradoxically, the new contract establishes that only the trade unions that have signed the contract will be entitled in the future to represent the employees. As a result, one worker out of two will now be left without legitimate representation.
The concerns of Fiat can be understood by what economists call the hold-up problem in negotiations. Before making the investments, the bargaining power of the Fiat is very high, because it can credibly threaten to move production abroad if its demands are not met. After making the investment, that threat lacks credibility as it is already locked-in, so the bargaining power of Fiat is weak. This is why Fiat wanted to predetermine its counterparts in future negotiations.
So far, though, the debate has largely ignored the macroeconomic consequences of the Fiat plan.
A key macroeconomic question is: What would happen to employment and real wages if the Fiat plan were to be extended to the whole system of Italian industrial relations?
The answer is that, in the short term – i.e. before the new investments – the real wage is likely to fall, not rise, while employment is likely to increase. In the medium term – i.e. after the investment – the real wage should increase along with employment.
Here's how to think about it. First, we must define what we mean by the new “Fiat contract”. Initially, the company introduces more effective discipline and monitoring that reduces work-breaks and provides tighter controls for leave, sickness, and absences. In a second step, the company carries out productive investments that increase the productivity of labour.
Second, we must be clear about the logical framework used to study the linkages. The most natural conceptual framework here is the Shapiro and Stiglitz (1984) efficiency-wage model (see Appendix for details).
The first effect of the new contract is to make it easier to discover “shirking” workers, for whom the risk of being laid off increases. As a result, in the logic of the model, a lower wage is required to elicit the workers’ effort. But the logic doesn’t stop there.
As the bargaining power of workers falls, so does the real wage. In turn, this makes it more convenient to the firm to recruit new workers. Thus in the short term, employment grows and real wages fall. In the second step, Fiat invests in the plant and this raises the productivity of labour. The demand for labour by the company increases and these results in a higher employment and wages.
This logic suggests that it is reasonable to believe that the new Fiat contract – if it were extended more widely – will lead to a rise in employment in both the short and medium term. However, the deterioration in the bargaining position of workers should lead to a temporary fall in the real wage.
Shapiro, C and J Stiglitz (1984), “Equilibrium Unemployment as a Worker Discipline Device”, American Economic Review, 74(3):433-444.
The model of Shapiro and Stiglitz is well suited to answer our question. The model, shown in the Figure below, consists of two curves.
The downward sloping curve (Ld) is the labour demand by firms, and relates the number of workers hired (NL, on the x axis) and the real wage (w, on the y axis); the higher the wage, the lower the firm’s demand for labour. The second curve (NSC, No shirking condition) is positively sloped, and describes the real wage which makes the worker indifferent between shirking (and risking to be fired) and eliciting effort on the job. If employment is low (and unemployment high) it will be difficult to find a new job if one is fired, and therefore a relatively low salary will be enough to convince the worker to be exert effort. Conversely, if employment is high, being laid-off is not too costly, since it will be easy to find a new job. Hence a higher salary will be required to convince the worker not to shirk. For these reasons, the NSC curve is positively sloped. In the model, wages and employment are jointly determined when both relations are satisfied, at the intersection of the two curves, point Q.
The first effect of the new contract is to make it easier to detect shirkers, for whom the risk of redundancy rises. Thus a lower wage is sufficient to convince workers to be productive and NSC curve shifts down. As the bargaining position of workers falls, the real wage declines and employment rises (point Q shifts down to the right). When the Fiat's investments occur (they should be unexpected for their effects to materialise only at the time when they are made), the productivity of workers rises and the demand for labour increases; the Ld curve shifts upward. This effect produces an increase in employment as well as in the real wage.