Social job-search networks and the transition from school to stable employment

Francis Kramarz, Oskar Nordström Skans, 17 October 2013



The challenges faced by young workers transitioning from school into stable employment are a major concern throughout the OECD. The search for stable employment is a time-consuming process, particularly in countries without highly developed apprenticeship systems. Many young workers – especially the least educated – are caught struggling for years.

Considering the importance of this process we know surprisingly little about the strategies used by young job searchers looking for entry jobs. Two key insights arise from the large body of research on job-finding networks:

  • Social ties serve as a key component in the process of matching workers and firms in general, and
  • The role played by these ties are particularly important among the young and low educated (Pellizzari, 2010).

For these reasons, job-finding networks arise as a key research topic for anyone interested in youth unemployment and school-to-work transitions.

Social ties and other informal channels appear to play a role in the majority of labour market matches, perhaps for a multitude of reasons (Ioannides and Loury 2004). We are therefore unlikely to find a ‘one-size fits all’ understanding of the role played by social networks on the labour market. In a recent paper (Kramarz and Skans 2013) we focus on young workers, using a unique population-wide Swedish register dataset. We show that family ties play a central role in many young workers’ transition into the labour market. On average 14% have a close relative (parent, sibling, uncle or aunt) within the establishment of their first stable; this rises to 19% among the least educated.

We start by comparing the entry patterns among graduates from all levels of schooling within the Swedish school system, documenting that the propensity to follow the own parent is at least an order of a magnitude larger than the propensity to follow a classmate’s parent.1

A key finding is that young workers’ tendencies to follow their parents are restricted to the cases when the parents are present within the actual establishment. Youths do not enter into other establishments within the same firm as the parent, and they do not enter the parents’ previous establishments after the parents have moved on. In fact, children disproportionally follow their parents even if we only compare across graduates from the same school, within the same occupational field, whose parents work in different establishments within the same firm. These results indicate a causal impact of social ties.

One way of illustrating the impact of social networks is to analyse the recruitment of graduates of a certain type – defined by the precise (physical) school in a given field of education – by firm over time. This pattern is shown in Figure 1, which displays the probability of hiring a graduate of the same type as the graduating child over time, before and after the graduation year. This probability is low and stable before the graduation of someone with a parent in the plant, but then increases dramatically at graduation, after which it subsequently declines. This gradual decline is consistent with the fact that not all graduates find a stable job immediately after graduation. The fact that there is no decline over time before graduation also implies that employers do not postpone recruitments until the linked child graduates.

Figure 1. Fraction of graduates of a certain type hired by a plant where a parent works, before and after graduation of a child with that exact type of education

Source: Kramarz and Skans, 2013.

Social ties provided by parents are particularly important for young workers who are low educated and have poor grades. In contrast, a complementary analysis of weak ties (e.g. neighbours) suggests that their impact is largely independent of the level of education. Network strength then seems more important the lower the education of the graduate. For a given level of education, strong ties also appear to be more important the less specific the training is (i.e. the more targeted to a specific type of occupation or firm).

Youths who find employment in plants in which they have a strong social tie benefit in terms of:

  • Much shorter school-to-work transitions, and
  • Higher employment rates and wages

at a few years' horizon. These benefits are particularly helpful in the high-unemployment years, when social ties are used more frequently. However, the youths do pay a price in the form of lower entry wages.

A central but difficult and rarely studied (see however Kramarz and Thesmar, JEEA) question in the analysis of social networks is how the use of job-finding networks relates to firm-level efficiency. As with most previous studies we are not able to fully analyse this question, but we do find several indications that networks appear to be used in more efficient conditions, rather than less. For example, hiring of employees’ children is more frequent when the parents are high-wage workers and have long tenure, suggesting that firms use parental quality as a signal of youth reliability. In addition, the use of strong social ties appear to be more frequent among well-performing (productive and high profiting) firms, and during years when performance measures are about to improve. Overall, the firm-side benefits of social networks appear to be large enough to induce small firms to create jobs that otherwise would not have been created, or at least not otherwise filled by youths.

The combined set of results speaks against the notion that networks matter simply by conveying information about available vacancies. Rather, the results indicate that networks transmit information regarding the quality of the young worker, particularly along dimensions that are difficult to observe (as in Montgomery 1991).2 It is conceivable that firms use the observed ability of the parent as a signal of low educated youths’ abilities. Along these lines, although in a different setting, Hensvik and Skans (2013) show that recruitments through coworker networks have a negative relationship to easily observed skills (such as educational outcomes) but a positive relationship to skills that are more difficult to observe (such as test scores).

The results we present in Kramarz and Skans (2013) are important for our understanding of the school-to-work transition process since they show how important strong social connections are to low educated youths entering the labour market. The picture that emerges from this analysis is quite different from the standard modelling of an anonymous job-matching process where job searchers apply in response to publicly posted vacancies. It also differs from the often-emphasised role of weak social ties (Granovetter 1973). Weak ties are often perceived to be particularly important because they bridge new information sets in a process where job searchers look for information about available jobs. This may be an adequate characterisation of job matching among well-established workers with transparent CVs, but the results we present suggest that networks play a different role for young untested workers with low education. Here strong ties – which may convey significantly more credible information about the quality of the workers – may provide crucial additional information for employers who may be reluctant to allow unknown and untested graduates to enter their firms. An important result provided by the analysis is also that the practice of relying on strong social ties appears most prevalent within the private sector (despite decentralised recruitments in the Swedish public sector) and in well-performing, high-wage firms. Although we are unable to fully analyse whether the use of strong social ties in recruitments is efficient or not, we show that the phenomenon is not confined to a non-competitive or secondary labour market.

Policy implications

To derive policy advice from this analysis it is necessary to generalise beyond the actual setting. With this caveat in mind, it does not seem particularly far-fetched to interpret the analysis as suggesting that the provision of contacts between youths and employers may be a crucial element when trying to assist youths in their job-search process.

Simply put, when designing policies it is useful to acknowledge that firms tend to provide jobs to worker whom they already know rather than to unknown applicants. In addition, our results suggest that it is insufficient to provide causal contacts between the employers and (in particular) low educated job seekers. Although policies cannot provide close substitutes to parental contacts, it is possible that low educated job-seekers benefit more from long-lasting exposure to single employers than from shorter exposure to many employers. Our results suggest that the impact of ties are more pronounced if there is an overlap in skills, a result also found in a replication focusing on contacts established during summer jobs (Hensvik and Skans, forthcoming).

As a final note, we would like to reiterate that we believe that networks and social ties may play very different roles in different segments of the labour market. With this in mind, we believe that the firm-side regularities which we document for entering inexperienced workers should serve as a starting point for future research, hopefully pushing towards a comprehensive documentation of how network recruitments relate to firm performance and overall recruitments at different segments of the labour market. A central part of our own current research agenda is to contribute to this documentation.


Granovetter M (1973), “The Strength of Weak Ties”, American Journal of Sociology, 78 (May), 1360-1380.

Hensvik L and O N Skans (2013), ”Social Networks, Employee Selection and Labor Market Outcomes”, IFAU Working Paper 2013:15

Hensvik L and O N Skans, ”Networks and Youth Labor Market Entry”, Nordic Economic Policy Review, forthcoming.

Ioannides Y M, and L D Loury (2004), “Job Information Networks, Neighborhood Effects and Inequality”, Journal of Economic Literature, 42 (4), 1056.1093.

Kramarz F and O N Skans (2013), “When strong ties are strong: Family networks and youth labour market entry”, CEPR Discussion Paper 9620, forthcoming in Review of Economic Studies.

Kramarz F and D Thesmar (2013), “Social Networks in the Boardroom”, Journal of the European Economic Association, 11, 4, 780-807.

Montgomery J (1991), “Social Networks and Labor-Market Outcomes – Toward an Economic Analysis”, The American Economic Review, vol. 81, 5, 1408-1418.

Pellizzari M (2010), “Do Friends and Relatives Really Help in Getting a Good Job?”, The Industrial and Labor Relations Review, 63, 494-510.

1 Our analysis excludes self-employed parents.

2 An alternative is that they are driven by the presumption that parents may help monitoring the entering youth.

Topics: Labour markets
Tags: graduates, jobs, unemployment, youth unemployment

Director, Centre de Recherché en Economie et Statistique (CREST); Professor, ENSAE; Associate Professor, Ecole Polytechnique
Researcher, Institute for Evaluation of Labour Market and Education Policy (IFAU)