Competitive jobseekers: When sharing less leaves firms at a loss

For efficient matching of jobs and workers in the labour market, it is essential that information on job postings reaches suitable jobseekers. While social networks play a key role in this regard, competition for jobs may disincentivise information-sharing. Based on an experiment with college students in Mumbai, this article finds that these disincentives reduce the overall quality of applicants and hires.

Jobseekers frequently rely on their social networks to learn about job openings (Trimble and Kmec 2011, Beaman and Magruder 2012, Dustmann et al. 2016), especially in India, where good quality jobs are scarce and the competition for them is intense. For firms aiming to attract skilled candidates, it is essential that job information effectively reaches these jobseekers through their social circles, since formal channels are often ineffective in doing so (Kelley et al. 2022, Fernando et al. 2023). However, in highly competitive environments, this flow of information often becomes constrained. Jobseekers may be hesitant to share leads with peers if doing so reduces their own chances of securing the position, creating “strategic disincentives” to sharing this information within networks.

In recent research (Chiplunkar, Kelley, and Lane 2024), we design a randomised controlled trial (RCT) to first, document whether jobseekers exhibit strategic disincentives in information sharing (that is, share less information about jobs with specific types of peers when the competition for these jobs is made salient to them); second, explore how it impacts the talent pool of applicants received by a firm, the hires that they make, and their performance on the job; and finally, assess how these strategic disincentives interact with a common strategy firms use to attract better talent – offering higher wages.

Study design

We partnered with six colleges in Mumbai that offer multiple programmes of study, including commerce, finance, information technology, and human resources (‘a batch’). We worked with around 500 students who were about to complete their final year of college in 2021 and intended to look for jobs once they graduated. Every week for a month, we created short-term job opportunities that lasted approximately 45 minutes and required students to work independently to research and summarise media articles on a particular topic assigned to them.

Each week, information about these jobs was shared with a randomly selected set of students within a batch – the “entrypoints”. The job itself was randomly categorised as either ‘rival’ or ‘non-rival’ across batches. For rival jobs, all applicants, including the entry points, had to apply and compete against their peers for the position. In contrast, non-rival jobs guaranteed that entry points were hired, but allowed them to share the information with their peers, who could then apply for consideration. Whether a batch received a rival or non-rival job, and who within the batch received information about it (entry points) were randomly chosen every week. Lastly, we ranked all applicants in rival batches by GPA (grade point average) and hired the most qualified set. In non-rival batches, we hired all entrypoint students (if they applied), along with other students with the highest GPA.¹

Impact of competition on information sharing

Our key finding is that jobseekers strategically withhold information from specific peers when competition is salient. When a job was rival, high-ability students in the class had a significantly lower chance of hearing information about it from their friends. On the other hand, when positions were non-rival, jobseekers were more likely to share with their high-ability peers. In both cases (rival and non-rival), jobseekers shared information with lower ability peers, perhaps not viewing them as competition for getting the job. Notably though, having close connections mitigated competitive concerns. Close friends shared job information with each other regardless of its competitive nature, highlighting that social bonds can sometimes counteract the pressures of competition. Lastly, these effects were all driven by men, who were less likely to share rival information, either with their higher-ability friends, close friends, or other male friends. We find no such effects for women.

Competition lowered quality of hires for the job

Figure 1 reports the GPA distribution of hires made for rival and non-rival jobs. We see a clear rightward shift (that is, towards a higher GPA) in this distribution when the information was non-rival as opposed to when it is rival. On average, students hired when job information was non-rival had around 8.4% higher GPA scores, as compared to those hired when the information was rival. This suggests that the lack of sharing rival information, especially with talented peers, had a cascading effect in lowering the overall quality of applicants and hires as well. As a result, new hires in these rival jobs underperformed on key performance metrics as well. They were less likely to finish the job on time, more likely to be absent, finish it slower, and submit poorer quality of work.

Figure 1. Ability distribution of hires

Does increasing wages attract better talent?

A common strategy employed by firms to attract better talent is to increase wages. However, strategic disincentives in information sharing – like the one that we study – could further exacerbate competition for these desirable jobs, thus mitigating firms’ ability to attract better talent. To study this more rigorously, we randomly doubled the wage offered in some jobs when they were advertised (in addition to whether they were rival or not).

We find that simply doubling wages (in status-quo rival jobs) does attract higher quality talent, who do perform better on the job. However, a key advantage of our setting is that it allows us to investigate how competition can dampen this effect. In other words, we find that simply mitigating competition by making job information non-rival, without changing any wages, can have an even larger impact on the quality of hires and their job performance. A back-of-the-envelope calculation suggests that firms would have to increase wages by 3-6 times (rather than doubling it, like we did in our experiment) to get the same increase in ability induced by eliminating the strategic disincentive to share information.

Implications for firms and the labour market

The implications of our study for the labour market are significant. When social networks play a crucial role in communicating job postings, the incentive to withhold information in competitive situations can reduce the overall quality of applicants and hires. While firms can attempt to offset this by offering higher wages or more attractive jobs, our results suggest that the wage increase would need to be substantial.

These findings highlight the value of supporting technologies that facilitate the flow of job information beyond social networks – such as job portals, information campaigns, and public advertising. Firms could also ensure that those relied upon to relay information through social networks are perhaps properly incentivised to do so (through incentivised referrals, for example), reducing the likelihood of competitive resistance. More broadly, by understanding these dynamics, firms and policymakers can better design recruitment strategies that minimise the downside of competition and enhance the overall quality of matches in the job market.

Notes:

In both cases, students were unaware of the selection criteria or the total number of slots available in any given week.