VoxEU Column Labour Markets

Where are the jobs? Out there, somewhere. Perhaps.

If you lose your job, can you find a new one with a little help from your friends? This column presents evidence that displaced Italian workers with more employable friends and social contacts are unemployed for a shorter period of time.

The global crisis hit jobs hard. According to the OECD, between 2007 and 2010 the number of employed people fell by almost 5 million throughout OECD countries and the number of job seekers rose by over 16 million. It is now about two years since the trough of the recession, and unemployment rates remain at historical highs in many advanced economies despite signs of recovery in economic activity and labour demand.

Vacancies and hires in the current recession

Most notably in the US, a debate has arisen over whether this signals an increase in the structural rate of unemployment or a lack of aggregate demand (Kocherlakota 2010; Hall 2010; Valletta and Kuang 2010). The answer probably lies in the middle. A small portion of the unemployment does indeed seem to be structural in nature. The increased number of vacancies, taken by many commentators as evidence of a major shift in the natural rate of unemployment, appears to be due to a lower vacancy yield – i.e. the number of hires per vacancy (Barnichon et al. 2010). While in principle this might reflect a more pervasive mismatch between demand and supply, the available evidence does not seem to support this (Dickens 2010; Sahin et al. 2010). Among the reasons why hiring rates may have decreased, a prominent explanation relates to the way information on employment opportunities is generated and circulated in modern labour markets.

Information asymmetries are pervasive. Workers have to learn about job opportunities, firms have to learn about workers, and both have to learn about the suitability of an employment relationship. Survey evidence suggests that job seekers rely on a wide array of tools to collect information about suitable employment opportunities. Some of these channels are formal (e.g. employment agencies or newspapers) but the most widespread turn out to be informal (e.g. friends, family, social and professional acquaintances). Informal channels are quantitatively relevant even in advanced economies. For example, in the US about 20% of jobseekers reports checking with acquaintances and friends and about 50% of existing jobs are reported to have been obtained through social connections.

What happens when a crisis hits? Several theoretical contributions have shown that an aggregate negative employment shock may have long lasting effects if the amount of job-relevant information circulated within social networks depends on the employment status of a person’s social contacts, presumably because of their privileged access to soft information. That is, a recession disrupts usual information channels and the ensuing damages to the efficiency of the matching process further delay the recovery.

Social networks and finding a new job: Evidence from Italy

How relevant are information networks for job search? In a recent paper (
Cingano and Rosolia, forthcoming), we exploit a large matched employer-employee dataset covering all employment relationships in two northeastern Italian provinces over 20 years that allows us to assign to each individual at each point in time a specific social network, the one formed by his co-workers up to that point of his career. With this definition, a reasonable proxy for the amount of job-related information in the network is the proportion of employed social contacts at a given moment. We thus focus on workers displaced by firm closures and relate the length of their subsequent unemployment spell to his or her contacts’ employment rate when displacement occurs. Information spillovers are identified by comparing the performances of workers contemporaneously displaced by the same firm, while also controlling for a wide range of individual and group characteristics.

We find that the proportion of employed contacts plays a significant role in shaping re-employment patterns. Increasing the employment rate of contacts by one standard deviation reduces the job-seeker’s unemployment duration by about 10%; this effect is only slightly less than the reduction associated to a one standard deviation increase in own wage at displacement, a proxy of individual ability. Contacts’ employment status plays a stronger role if they recently searched for a job, and thus collected useful and up-to-date information, and if their current employer is closer (spatially and technologically) to the unemployed. We rule out that this evidence reflects a referral mechanism, whereby employed friends and contacts are also more skilled and thus reliable and more successful when recommending a worker to a prospective employer, by showing that proxies of their ability are not related to the displaced duration into unemployment.

Our findings show that job-relevant information generated and disseminated by employed workers through informal channels is a prominent element of a frictional labour market. The employment fall in recessionary periods can thus be quite disruptive, determining a lasting loss of efficiency of the matching between job seekers and firms. Consistently, recent evidence shows that temporary hikes in the number of vacancies relative to unemployment are a persistent feature of post-recession periods when the relationship between vacancies and unemployment is looked at over a longer horizon (Tasci and Lindner 2011). Our findings suggest that this may result from imperfect information about job opportunities rather than from a mismatch between labour demand and supply characteristics. As a consequence, employment-oriented stimulus policies may have had spillover effects in helping restore the information facilitation process that have so far gone unrecognised.


Barnichon R, M Elsby, B Hobijn, and A Șahin (2010), “Which industries are shifting the Beveridge curve?”, JOLTS Symposium.

Cingano F and A Rosolia (forthcoming), “People I know: job search and social networks”, Journal of Labor Economics. 

Dickens, W. (2010). “Has the recession increased the NAIRU?”. 

Hall, RE (2010), “The Labor Market in the Current Slump”, Working Paper, Stanford University. 

Kocherlakota, N (2010), “Inside the FOMC”, Federal Reserve Bank of Minneapolis. 

Șahin A, J Song, G Topa, and G Violante (2010), “Measuring Mismatch in the U.S. Labor Market”, New York Federal Reserve. 

Tasci, M. and J. Lindner (2011), “Has the Beveridge curve shifted?, Federal Reserve Bank of Cleveland, Economic Trends. 10 August. 

Valletta, R and K Kuang (2010), “Is Structural Unemployment on the Rise?”, Federal Reserve Bank of San Francisco, Economic Letter 2010-34.

420 Reads