March 31, 2016

How strong and weak ties help you find a job

By: Moira Burke, Jason J. Jones, Laura Gee

Networking is important to finding a job. But who’s more helpful: a very close friend or an acquaintance? Previous academic work finds mixed results [1-5]: weaker ties may bring novel information, but strong ties might put more effort into helping you. With so many people searching for work, it’s important to understand how different kinds of connections are helpful.

In a new paper appearing in the Journal of Labor Economics, using de-identified, aggregated U.S. data based on employment start dates people choose to list in their Facebook profiles, we find a seemingly contradictory set of results:

  1. Most people find a job through one of their numerous weaker ties
  2. An individual stronger tie is more likely to help than a individual weaker tie

How is this possible? Well, weak ties are important collectively because of their quantity, but strong ties are important individually because of their quality.

Measuring job “help”

How do we know if a friend helped another friend to get a job? We use a rough proxy: whether a person eventually ends up working at the same place as a pre-existing Facebook friend. It’s an overestimate with a lot of noise (people living in the same area often become friends and also happen to work at the same employer), but we used some strict temporal thresholds to reduce noise. The figure below shows the process. Let’s consider two people (called “ego” and “alter”). By our proxy “job help” measure, if ego and alter were Facebook friends, and then alter joins Firm A, and then more than a year later ego joins Firm A, there’s a reasonable chance that alter helped ego find her job. Alter may have known about a job opening, referred ego, helped ego practice her interview, or simply made ego aware that Firm A was a good place to work.

post00087_image0002
Figure 1. We say that a friend (“alter”) may have helped someone (“ego”) find a job if ego and alter were friends, then alter begins a job at Firm A, and then more than a year later, ego begins a job at Firm A. We measure tie strength for the year before ego starts at Firm A.

Measuring tie strength

There are lots of ways to measure how close two people feel, known as “tie strength” to social scientists. We chose a few different measures spanning the year before ego started her job at Firm A: how often ego tagged alter in photos, how many posts ego wrote on alter’s wall, and how many mutual friends they had. Results were similar for all three measures of tie strength.

Who provides more job help?

Then, we tested if most jobs collectively come from weaker ties by counting the percentage of jobs that came from weaker or stronger ties. The left side of Figure 2 below shows that over 90% of job-helping friends are weak ties. However, as the right side of Figure 2 shows, most friends are weak by our metric (photo tagging).

So although weak ties are collectively very helpful in job finding, that is because we have so many weak ties. Think about your own social network, how many close friends do you have? How many acquaintances?

post00087_image0003
Figure 2. The majority of job-help friends were weak ties, but this is because weak ties comprise a majority of a person’s network.

Weak ties are collectively useful, but are they individually more useful too? To test that, we estimated the probability that each type of tie would be helpful. Imagine assigning a probability to each one of your friends about how likely they are to help you find a job. We found those individual probabilities using a regression. And we found that the stronger ties had higher probabilities of helping a person find a job as illustrated in Figure 3. That means that an individual stronger tie was more likely to be helpful than an individual weaker tie by our tie-strength metric (photo tagging).

We already know that networking is important to finding a job. Our study shows that weaker ties are useful because they are numerous, but that a single stronger tie is more useful than a single weaker tie.

post00087_image0004
Figure 3. Individual strong ties are more likely to help than weak ties.

There are some major challenges to studying how networks relate to job-finding:

  1. It is easy to say that increased tie strength is associated with increased job help, but hard to say that increasing tie strength causes it. We have data about social ties that already exist. The stronger ties you already have are also likely to be the people who you are already more likely to work with. Your strong ties are probably strong because you enjoy the same things, have a similar work ethic, and have similar backgrounds. Researchers can’t measure all those hidden similarities, but they could be driving our results. We were able to control for many similarities like being close in age, going to the same schools, and both being the same gender. We also ran a placebo test where we pretended that alter got her job from ego (instead of the other way around). We still found that strong ties are important individually. But we can never fully account for all those hidden similarities between strong ties.
  2. We don’t have data about all people in the US and all their friends. So these results may not apply to the average US worker. However, in the US over 54% of adults have a Facebook account and 40% of social network users have “friended” their closest friends [6].
  3. Not all interactions take place on Facebook, so our measures of tie strength may not reflect how close two people are in reality. However this isn’t such an issue, because previous academic work finds that Facebook interactions are a good predictor of real world friendships [7-8].

We did our best to overcome these challenges. And after a lot of careful testing we still found that a person is more likely to work with a weaker tie because weaker ties collectively make up most of a person’s social network. But, strengthening an existing tie should increase the probability that you will work with that specific friend.

How to act on these results

If you’re currently unemployed and looking for a job, or your current job is expiring, you might consider broadcasting your job-seeking status to your weak ties. Consider writing a status update letting people in your network know that you’re looking for a new job. Narrowly target specific requests for job help to your close friends, such as sending a message asking about their workplaces. Our findings suggest time-consuming, costly communication should be directed at strong ties, but informing weak ties about your job search is also a good investment as long as you can reach many weak ties quickly.

Weak ties are important collectively to job finding, but individually strong ties are more helpful.

References

[1] Granovetter, Mark S. (1973). The strength of weak ties. American Journal of Sociology 78(6), pp. 1360-1380.
[2] Granovetter, Mark S. (1983). The strength of weak ties: A network theory revisited.’ Sociological theory 1(1), 201-233.
[3] Kramarz, Francis, and Oskar Nordstrom Skans (2014). When strong ties are strong: Networks and youth labour market entry. The Review of Economic Studies
[4] Burke, M., & Kraut, R. (2013). Using Facebook after losing a job: Differential benefits of strong and weak ties. Proc.2013 conference on computer-supported cooperative work (CSCW). p. 1419-1430).
[5] Aral, S. and Van Alstyne, M. (2011) The Diversity-Bandwidth Trade-off. American Journal of Sociology 117(1), 90–171.
[6] Hampton, Keith, Lauren Sessions Goulet, Lee Rainie, and Kristen Purcell (2011). Social networking sites and our lives.’ Pew Internet and American Life Project
[7] Jones, Jason J., Jaime E. Settle, Robert M. Bond, Christopher J. Fariss, Cameron Marlow, and James H. Fowler (2012). Inferring tie strength from online directed behavior. PLOS ONE
[8] Gilbert, Eric, and Karrie Karahalios (2009) `Predicting tie strength with social media.’ CHI ’09 Proceedings of the SIGCHI Conference on Human Factors in Computing Systems