Recession has put many Eurozone labour markets under stress, particularly those in Mediterranean countries that have inflexible markets. As part of ongoing reforms, many have tackled the labour market – under the explicit encouragement of the EU and the ECB.
Of course, labour-market reforms and macroeconomic policy matter in their own right, but an emerging consensus on the resolution of the Eurozone crisis has firmly linked the two in the minds of policymakers and economists, as well as market forces (Wyplosz 2013). While the importance of labour reforms has been well documented, the interaction with macroeconomics has not. The empirical and theoretical literature shows that firing costs reduce hiring and firing (see, e.g., OECD 2004). The interaction of firing costs and macroeconomic policies, by contrast, has received little attention so far from empirical economists. There are good reasons for thinking the two are linked.
In our recent work, we argue that larger firing costs lead to a more severe policy trade-off for central banks – i.e. more price stability is associated with more inefficient employment fluctuations (Faia, Lechthaler and Merkl 2013). This is particularly worrisome in a monetary union with heterogeneous firing costs, such as the Eurozone, and calls for a harmonisation of firing costs across countries to make sure that the ECB can perform an efficient policy for all member states.
Since Bentolila and Bertola (1990), we have known that firing costs reduce both hiring and firing. Since these effects work in opposite directions on aggregate unemployment, the overall effect on unemployment is ambiguous. This ambiguous relationship is confirmed by theory (e.g. Pissarides 2000) and cross-country studies.
By contrast, the interaction of firing costs and the business cycle has been neglected until recently. Two recent studies (Abbritti and Weber 2010, and Merkl and Schmitz 2011) have empirically shown that larger firing costs reduce output fluctuations (see Figure 1 for an illustration for the Eurozone). Given that firing costs matter for the business cycle, firing costs may be of relevance for monetary policy, too.
Figure 1. Output volatility and employment protection index (1/1999-2/2008)
Since the existing literature has neglected the optimal design of monetary policy in this context, we make a first step in this direction. We use a selection model based on Lechthaler et al (2010), who provide a simple and intuitive framework to analyse the consequences of labour turnover costs for business cycle fluctuations. The model is based on the idea that employers do not randomly hire workers that they are matched with, but rather assess the suitability of job applicants and select only those workers who are most productive. A similar logic applies to the decision of whether to keep a worker or to fire her. When the losses generated by the worker are higher than the cost of firing, the worker will lose her job.
Naturally, the magnitude of firing costs plays an important role for both decisions. Higher firing costs do not only increase the costs of separations and thus reduce the firing rate. They also put restrictions on the behaviour of firms and reduce their profitability. This tends to decrease the hiring rate: firms become more selective when choosing their workers because they know that it is expensive to separate from them. Thus, the model can very well explain the empirical relationship between firing costs and hiring/firing rates discussed above. It can also replicate the stylised fact that the economies of countries with higher firing costs tend to be less volatile over the business cycle (See Figure 2).
Figure 2. Output volatility and firing costs in the model simulation
What is important for macroeconomic policy is the fact that this relationship implies important externalities, i.e., the market left alone will not work efficiently. By hiring a worker today, firms reduce the future pool of applicants. This implies that firms h