Endogenous choices of contract types in an agent-based model of the labor market

International audienceThe Fixed Duration Contracts (FDC) have taken an importance place in the European labor markets, notably in France and Spain. They represent a dominant share of the hires, although most workers hold an Open Ended Contract (OEC) at any given date. There is then a permanent coexistence of the two types of contract that we explain through a trade-off that firms compute between their costs and benefits when deciding to open a vacancy. For the first time are taken simultanously into account for the OEC the firing costs, the advance notice costs, and the losses when the firm is unable to meet the legal requirements to initiate economic dismissals. For the FDC, the specific costs are the termination costs and the waiting cost when a new FDC cannot be opened immediately after a termination. Training and vacancy costs are common to both contracts and included but they are amortized over very different durations among the two contracts and these costs influence the trade-off. We extend WorkSim, an agent-based model of the French labor market which reproduces the gross flows of workers between the different states, employment (FDC and OEC), unemployment and inactivity. The theoretical framework is the costly search by the heterogenous agents, firms and individuals, who interact on the market, taking rationally bounded decisions but learning from their mistakes. The competition takes place in a labor stock-flow consistent framework, taking into account crowding out effects. The model is scaled and calibrated through a poweful algorithm to reach a steady state which reproduces the main observed variables in the labor market in the year 2011 with a correct fit. We generate the main effects of FDC, churning, screening, stepping stone, but also model in detail the buffer effect which is built on an option into an intertemporal decision framework with idiosyncratic anticipations of firms demand. The results of the sensitivity analysis show that pessimistic anticipations and the volatility of demand shocks raise the recourse to FDC but also unemployment. Increasing firing costs also raises unemployment but not in very significant way. Forbidding FDC does not change the employment significantly since the opposite effects of FDC seem to compensate each other. While the model puts into a unified framework the main theoretical ideas that yield the trade-off between FDC and OEC, and can be applied to different countries, it also offers sufficent detail to allow for labor market policy discussion in a given country