The Eurozone, in contrast to the US, exhibited remarkable labour market resilience in the aftermath of the Lehman shock that lead to the Great Recession. Conversely, as the debt crisis developed, labour markets in the Eurozone weakened and unemployment started growing above what was predicted on the basis of GDP growth (Figure 1).
Figure 1 Unemployment changes, actual and predicted on the basis of the Okun law, Eurozone
Note: The predicted unemployment rate is an out of sample prediction of an Okun's law estimate on the period starting with 1996q1 and ending with the last quarter of before the start of the recession.
What factors can account for the lost resilience of labour markets in the Eurozone? A key and often-mentioned explanation (European Commission 2011; ECB 2012) is the reduced room for adjusting labour inputs in terms of working hours. At the start of the crisis, including as a result of government-sponsored “Short Time Working Schemes”, a number of Eurozone countries managed to prevent major job shedding while reducing average hours worked. As output recovered, hours recovered as well, but remained below pre-crisis levels. As compared with the initial phase of the crisis, the current growth of unemployment is mostly linked to historically low job-finding rates rather than job separation. Why are firms so hesitant in their hiring decisions? Why is employment growth even more disappointing than output growth?
Explaining low employment growth
Following the lead of recent research (Baker, Bloom, and Davis 2012; Kose and Terrones 2012), we investigate the hypothesis that the lost labour market resilience in the Eurozone is linked to the disincentives to hire and to keep workers arising from an uncertain environment. The underlying idea is that, in an uncertain environment, firms would prefer to delay hard-to-reverse decisions such as hiring, so that job creation would be missing despite a possible recovery in output (see for example Leduc and Liu 2012). By uncertain environment, we especially mean the uncertainty over the resolution of Eurozone economic problems, most notably the debt crisis.
To test our hypothesis, we develop a VAR model to analyse the interaction between GDP, unemployment, and the index of policy-related uncertainty developed in Baker, Bloom and Davis (2012). Our aim is to check if uncertainty has a direct impact on unemployment in addition to that possibly arising via GDP (see European Commission 2012; Arpaia and Turrini 2012)1. The sample for Eurozone variables spans the period 1996Q1-2011Q4. Shocks are identified by means of a Cholesky decomposition with the following order: uncertainty, GDP and unemployment rate. This is equivalent to imposing that uncertainty is a forward-looking variable and that GDP and unemployment respond contemporaneously and with lags to uncertainty shocks. The impulse response function (Figure 2) suggests that unemployment responds not only to GDP shocks, as expected, but also to policy-related uncertainty. This means that falling GDP coupled with increased uncertainty implies unemployment raising more than predicted by a standard Okun relation. Additionally, the impact of uncertainty on unemployment appears to be rather persistent. Following a one standard deviation shock to uncertainty, unemployment rate rises for two years before reverting back to the pre-shock level.
Figure 2 Unemployment response to GDP and uncertainty shocks: Eurozone, US
How much of the fluctuations in unemployment can be explained by shocks to GDP and uncertainty? Table 1 reports the percentage of the variance of the error made in forecasting a variable due to a specific shock at a specific time horizon. It suggests that uncertainty explains a rather limited share of the error variance at a one quarter horizon, but a considerable one at longer horizons (between 20-30%).
Table 1 Forecast error variance decomposition of the unemployment rate
(% of total variance explained by shocks to the three variables)
What drives unemployment? Is it job separations or job-finding rates? Does uncertainty influence unemployment fluctuations and, if so, how? To shed light in this direction, we fit vector autoregression on uncertainty, GDP, job-finding and separation rates over the same period. Job-finding and separation rates are constructed as in Arpaia and Curci (2010) following the methodology proposed by Shimer and adapted to OECD countries by Elsby et al (2009). Figure 3 depicts the impulse responses to output and uncertainty shocks. They show that within a one year horizon, the separation rate is relatively unresponsive to uncertainty (the response is much below that to GDP), while the response of the job-finding rate is relatively stronger, of the same order of that to GDP. At sufficiently long horizons, uncertainty shocks account for the highest share of the error variance of job-finding rates (Table 2).
Figure 3 Response to GDP and uncertainty shocks of job finding and separation rates, Eurozone
Table 2 Forecast error variance decomposition of finding and separation rates, Eurozone
(% of total variance explained by shocks to the three variables)
|Finding rate||Separation rate|
Overall, the evidence appears consistent with a possible role of policy uncertainty in explaining the lost resilience of Eurozone labour markets and current lacklustre job creation. The main implication for policy is rather straightforward: credible solutions to the Eurozone debt crisis will contribute to the critical unemployment situation of a number of Eurozone countries not only by helping to kick start investment and production, but also by an additional, direct boost to job creation linked to confidence.
Arpaia, A and Curci, N (2010), “EU labour market behaviours during the Great Recession”, European Economy Economic Papers 405.
European Central Bank (2012), “euro Area Labour Markets and the Crisis”, Structural Issues Report.
European Commission (2012), “Labour Market Developments in Europe”, European Economy 5/2012.
Baker S, N Bloom and S Davis (2012), “Measuring Economic Policy Uncertainty”, mimeo, June.
Elsby, M W L, B Hobijn and A Şahin (2009), "Unemployment Dynamics in the OECD", NBER Working Paper 14617.
Kose, M A and M E Terrones (2012), "Uncertainty weighing on the global recovery", VoxEU.org, 18 October.
Leduc, S, and Z Liu (2012), "Uncertainty Shocks are Aggregate Demand Shocks", Federal Reserve Bank of San Francisco Working Paper Series No. 2012-10.
Shimer, D (2007), “Reassessing the ins and outs of unemployment”, NBER Working Paper 13421.
1 The Baker, Bloom and Davis (2012) index builds on two components. A first component combines information on the frequency of newspapers articles containing terms uncertain or uncertainty, economic or economy and policy relevant words such as taxes, policy, spending, regulation etc. A second component includes a measure of the disagreement among forecasters on future inflation and government purchases. The index exhibits spikes in correspondence of notorious events of policy uncertainty (e.g. the gulf War, 9/11, Lehman bankruptcy, the Greek bailout).