In medio stat victus Labor Demand Effects of an Increase in the Reterement Age, with Tito Boeri and Espen Moen, accepted, Journal of Population Economics

After falling for four decades, statutory retirement ages are increasing in most OECD countries. The labor market adjustment to these reforms has not yet been thoroughly investigated by the literature. We draw on a major pension reform that took place in Italy in December 2011, increasing the retirement age by up to six years for some categories of workers. We have access to a unique dataset validated by the Italian social security administration (INPS), which identifies in each private firm, based on an administrative exam of eligibility conditions, how many workers were locked in by the sudden increase in the retirement age, and for how long. We find that firms mostly affected by the lock in are those that were downsizing even before the pol- icy shock. The increase in the retirement age seems to displace more the middle-aged than the young workers. Furthermore there is not a one-to-one increase in the number of older workers in the firms where some workers were locked in by the reform. We provide tentative explanations for these results, based on the interaction between retirement, employment protection legislation and liquidity constraints of firms.

The paper previously circulated with different titles: A Clash of Generations? Increase in Retirement Age and Labor Demand for Youth , INPS Working Paper Number 1 and Closing the Retirement Door and the Lump of Labor

Static and Dynamic Inefficiencies in an Optimizing Model of Epidemics (2020) with Chritopher Pissarides and Espen Moen, CEPR Discussion Paper 15439

In an optimizing model of epidemics several externalities arise when agents shield to avoid infection. Optimizing behaviour delays herd immunity but also reduces overall infections to approximately the minimum consistent with herd immunity. For reasonable parameter values, and with no vaccine, we find that agents delay too much because of a rat race to shield: they shield too much in the hope that others catch the disease and reach herd immunity. This and other externalities drive large wedges between private and social outcomes. The expectation of a vaccine reverses the effects, and agents shield too little.

Modeling Contacts and Transitions in the SIR Epidemic Model (2020), with Christopher Pissarides and Espen Moen, Pre Published in CEPR COVID ECONOMICS, Vetted and Real Time Papers, Issue 5


Since the outbreak of the covid-19 pandemic economists have turned to the SIR model and its subsequent variants for the study of the pandemic’s economic impact. But the SIR model is lacking the optimizing behaviour of economic models, in which agents can influence future transitions with their present ac- tions. We borrow ideas and modelling techniques from the Mortensen-Pissarides (1994) search and matching model and show that there is a well-defined solution in line with the original claims of Kermack and McKendrick (1927) but in which incentives play a role in determining the transitions. There are also externalities that justify government intervention in the form of imposing more restrictions on actions outside the home than a decentralized equilibrium would yield

Public Employment Redux (2020), with Pedro Gomes and Thepthida Sopraseuth, published in Journal of Government and Economics, Vol. 1 


The public sector hires disproportionately more educated workers. Using US mi- crodata, we show that the education bias also holds within industries and in two thirds of 3-digit occupations. To rationalize this finding, we propose a model of private and public employment based on two features. First, alongside a perfectly competitive private sector, a cost-minimizing government acts with a wage schedule that does not equate supply and demand. Second, our economy features heterogeneity across indi- viduals and jobs, and a simple sorting mechanism that generates underemployment – educated workers performing unskilled jobs. The equilibrium model is parsimonious and is calibrated to match key moments of the US public and private sectors. We find that the public-sector wage differential and excess underemployment account for 15 percent of the education bias, with the remaining accounted for by technology. In a counterintuitive fashion, we find that more wage compression in the public sector raises inequality in the private sector. A 1 percent increase in unskilled public wages raises private-sector inequality by 0.13 percent.

Output Costs of Education and Skills Mismatch (2020), with Pedro Gomes and Thepthida Sopraseuth


We propose a simple theory of under- and over-employment. Individuals of high type can perform both skilled and unskilled jobs, but only a fraction of low-type workers can perform skilled jobs. People have different non-pecuniary values over these jobs, akin to a Roy model. We calibrate two versions of the model to match moments of 17 OECD economies, considering separately education and skills mismatch. The cost of mismatch is 3% of output on average but varies between -1% to 9% across countries. The key variable that explains the output cost of mismatch is not the percentage of mismatched workers but their wage relative to well-matched workers.

A Tale of Comprehensive Labor  Market  Reforms: Evidence from the Italian Jobs Act, (2019), with Tito Boeri,   Labour Economics, Volume 59, August 2019, page 33-48


The Italian Jobs Act introduced a subsidy for new hirings as well as a new open ended labor contract based on graded security, with severance payments increasing with tenure, while phasing out the compulsory reinstatement of workers in the case of unfair dismissals applied until March 2015.  Simple models of job creation and destruction predict that hiring subsidies and lower firing costs unambiguously increase hirings.  Moreover, lower firing costs associated with graded security should also increase layoffs. These effects need not to be uniform across the size distribution of firms, especially when firms of different size are treated differently by the policy changes as in the case of the Jobs Act. On the one hand,  the hiring subsidy was proportional to wages, but had a cap, hence was more generous for small firms- typically paying lower wages than large firms - making them particularly responsive along the job creation margin. On the other hand, the reduction in firing costs applied mainly to large firms concentrating on them the adjustment along the job destruction margin.  To investigate empirically the effects of the  Italian Jobs Act, we draw on a unique dataset covering the universe of private firms in Italy having at least once 10 to 20 employees in the two years prior to the reform of January 2015.
We find evidence of a substantial increase in open ended hirings, and in the transformation of fixed-term into open ended contracts, in the aftermath of the Jobs Act.  The effects of the Jobs Act on firings- conversely- are much smaller, and are concentrated on large firms, while small firms react more intensively- creating new open ended contracts- to the hiring subsidy.

Ordering History Through the Timeline (2017), with Eugenio Garibaldi

Cepr Discussion Paper Num. 12508


History is a key subject in most educational system in Western countries, and there is ongoing concern about the degree of historical knowledge and historical sensibility that students obtain after their high school graduation. This paper proposes a simple linetime test for quantitatively measuring a human sense of history. The paper reports the results of the test administered to approximately 250 Italian university students. There are two empirical results. First, students have remarkable difficulties in ordering basic events over the time line, with the largest mistakes observed around the events that took place in the Middle Age. Second, the paper uncovers a statistical regularity in the test performance across gender, with female subjects featuring a statistical significant and quantitatively sizable downward score. The gender difference is surprising, since existing literature on differences in cognitive abilities across gender suggests that female subjects outperform male subjects in memory related tests. The paper shows also that the gender difference survives to a variety of sub periods, and falls by only 20 percent when we distinguish between violent and non violent events

Financial Constraints in Search Equilibrium: Mortensen and Pissarides Meet Holmstrom and Tirole (2018), with TIto Boeri and Espen Moen.  Labour Economics, Volume 50, March, 144-155, with Tito Boeri and Espen R. Moen

Cepr Discussion Paper Num. 10266

A key lesson from the Great Recession is that firms’ leverage and access to finance are important for hiring and firing decisions. It is now empirically established that bank lending is correlated with employment losses when credit conditions deteriorate. We provide further evidence of this and make causal inferences on the effect of leverage on job losses drawing on a new firm-level dataset that we assembled on employment and financial positions of European firms. Yet, in the Diamond Mortensen Pissarides (DMP) model there is no role for finance. All projects that display positive net present values are realized and financial markets are assumed to be perfect. What if financial markets are not perfect? Does a different access to finance influence the firm’s hiring and firing decisions? The paper uses the concept of limited pledgeability proposed by Holmstrom and Tirole to integrate financial imperfections and labor market imperfections. A negative shock wipes out the firm’s physical capital and leads to job destruction unless internal cash was accu- mulated by firms. If firms hold liquid assets they may thus protect their search capital, defined as the cost of attracting and hiring workers. The paper explores the trade-off between size and precautionary cash holdings in both partial and general equilibrium. We find that if labor market frictions disappear, so does the motive for firms to hold liquidity. This suggests a fundamental complementarity between labor market frictions and holding of liquid assets by firms.

Inside Severance Pay, (2016) with Tito Boeri and Espen Moen.  Journal of Public Economics., Volume 145, January, Pages 211-225

All OECD countries have either legally mandated severance pay or compensations imposed by industry-level bargaining in case of employer initiated job separations. According to the extensive literature on Employment Protection Legislation (EPL), such transfers are either ineffective or less efficient than unemployment benefits in providing insurance against labor market risk. In this paper we show that mandatory severance is optimal in presence of wage deferrals motivated by deterrence of opportunistic behavior of workers. Our results hold under risk neutrality and in general equilibrium. We also establish a link between optimal severance and efficiency of the legal system and we characterize the effects of shifting the burden of proof from the employer to the worker. Our model accounts for two neglected features of EPL. The first is the discretion of judges in interpreting the law, which relates not only to the decision as to whether the dismissal is deemed fair or unfair, but also to the nature, economic vs. disciplinary, of the layoff. The second feature is that compensation for dismissal is generally increasing with tenure. The model also rationalizes why severance is generally higher in  countries with less efficient judicial systems and why small firms are typically exempted from the strictest EPL provisions.

Financial Frictions, Financial Shocks, and Unemployment Volatility, (2015) with Tito Boeri and Espen Moen

Cepr Discussion Paper 10648

Financial market shocks and imperfections, alongside productivity shocks, represent both an impulse and a propagation mechanism of aggregate fluctuations. When labor and financial markets are imperfect, firms' funding and leverage respond to productivity changes. Models of business cycle with equilibrium unemployment largely ignore financial imperfections. The paper proposes and solves a tractable equilibrium unemployment model with imperfections in two markets. Labor market frictions are modeled via a traditional Diamond Mortensen Pissarides (DMP) model with wage positing. Financial market imperfections are modeled in terms of limited pledgeability, in line with the work of Holmstrom and Tirole. We show analytically that borrowing constraints increase unemployment volatility in the aftermath of productivity shocks. We calibrate the model to match key labor and financial moments of the US labor markets, and we perform two quantitative exercises. In the first exercise we ask whether the interaction between productivity shocks and borrowing constraints increase the volatility of unemployment with respect to models that focus only on the labor market imperfections. In the general specification of the model, both leverage and non pledgeable income move with the cycle. Our calibration exercise shows that the volatility of unemployment in response to productivity shock increases by as much as 50 percent with respect to a pure DMP model with wage posting. The second quantitative exercise explores the role of pure financial shocks on aggregate equilibrium. We calibrate pledgeability shocks to match the frequency of financial crisis and define financial distress as a situation in which internal liquidity completely dries up. The second exercise shows that full dry up of internal liquidity implies an increase in unemployment as large as 60 percent. These results throw new light on the aggregate impact of financial recessions. 

Dismissal Disputes and Endogenous Sorting, (2015) with Gerard Pfann

Cepr Discussion Paper, IZA Discussion Paper, CES-Ifo Discussion Paper, forthcoming.

A dismissal dispute is a difference between an employer and an employee that prevents agreement on work contract termination. Disputed employer initiated separations often lead to costly and lengthy job termination processes. Such disputes can have various forms, as emphasized by the law practice and by country specific legislation. Many relevant and important questions can be asked. How shall we model and classify disputes? How do different grounds for contract terminations sort among different types of disputes? How long and how costly are dismissal disputes? Do courts and other third party institutions respond differently to different disputes? Yet, the economics of dismissal disputes is surprisingly silent. This paper is an attempt to partly fill this gap. Theoretically, the paper proposes a simple accounting framework that is coherent with general dismissal legislation. Empirically, it has access to more than 2000 dismissal disputes that took place in the Netherlands between 2006 and 2009. The data set records dispute level information on both the employer and the employee engaged in the controversy, including firm and worker characteristics, the reason of the dispute, the process duration, the decision, and the associated costs. In addition, the paper models the trade off between a lengthy bureacratic dismissal procedure via the Public Employment Service and a costly court ruling, as typically faced by Dutch firms. The model rationalizes the sorting of disputes among the two institutions and helps understanding the longevity of the Dutch model and its political resilience. Finally, it highlights how different institutions act differently in the face of dismissal conflicts. Such a phenomenon is clearly observed in the real life data.

Graded Security From Theory Practice: with Tito Boeri and Espen Moen

Cepr Policy Insight 82

Competitive On the Job Search, (2016) with Espen Moen and Dag Einar Sommervoll. Published in the  Review of Economic Dynamics, special issue in honour of Dale Mortensen. 
Cepr Discussion Paper Num 10175. October 2014.
SSRN 250340 Available at SSRN:
The paper proposes a model of on-the-job search and industry dynamics in which search is directed. Firms permanently differ in productivity levels, their production function features constant returns to scale, and search costs are convex in search intensity. Wages are determined in a competitive manner, as firms advertise wage contracts (expected discounted incomes) so as to balance wage costs and search costs (queue length). An important assumption is that a firm is able to sort out its coordination problems with its employees in such a way that the on-the-job search behavior of workers maximizes the match surplus. Our model has several novel features. First, it is close in spirit to the competitive model, with a tractable and unique equilibrium, and is therefore useful for empirical testing. Second, the resulting equilibrium gives rise to an efficient allocation of resources. Third, the equilibrium is characterized by a job ladder, where unemployed workers apply to low-productivity firms offering low wages, and then gradually move on to more productive, higher-paying firms. Finally, the equilibrium offers different implications for the dynamics of job-to-job transitions than existing models of random search.