Key message 1 – Reducing information gaps can increase employment quality and earnings for jobseekers.
Several recent studies found that providing information to jobseekers can significantly improve employment outcomes. We describe these interventions below and present a table of treatment effects estimates in the online appendix.3
ENABLING JOBSEEKERS TO COMMUNICATE THEIR SKILLSET BETTER
Recent experimental evidence shows that programmes that certify existing skills can help jobseekers find better jobs. Certification is particularly relevant for young people, who may have limited formal work experience and credentials. IGC research by Abebe et al. (2018) studies job application workshops which provided skill certificates, and instructions on how to present skills in resumes, cover letters, and at job interviews. Four years later, treated jobseekers had significantly higher earnings (+20%), job satisfaction, and employment duration. Encouragingly, these gains were concentrated among those with the least education and experience. Similarly, in South Africa, helping jobseekers signal test results on cognitive and non-cognitive skills to firms increased their employment rate (+17%) and earnings (+32%) three to four months after treatment (Carranza et al. 2018). In Uganda, providing certificates of soft skills led workers who found employment to earn 11% more in the two years after the intervention (Bassi and Nansamba 2018).
Certificates work best when they focus on general skills. In another IGC-funded intervention in Uganda, vocational training (which was focused on general skills) proved more effective than apprenticeships. The certified skills acquired during the training proved useful up to four years after the intervention. Apprentices, on the other hand, learned firm-specific skills that were harder to certify and were valued less by other firms in the market (Alfonsi et al. 2017).
Reference letters also seem to be a useful and undervalued tool to convey information about skills. In an audit study in South Africa, Abel et al. (2017) encouraged some applicants to seek a reference letter and provided them with a template to do so. They found that including a reference letter with a job application increased employer call-backs by 60%, and doubled them for women. However, in a second experiment, Abel et al. (2017) show that jobseekers underestimate the value of providing reference letters. The same study also develops a standardised template that can be used to help jobseekers obtain reference letters from previous employers.
Jobseekers may also struggle to communicate their preferences. One experiment looked at participants in a large vocational training programme in India. In this context, placement officers had poor knowledge of the jobs trainees wanted to be placed in. Providing the placement officers with this information resulted in more job offers and higher retention rates for trainees (Banerjee and Chiplunkar 2016). Treated individuals were also 1.5-2 pp (8-11%) more likely to be employed three to six months after the intervention.
GIVING JOBSEEKERS INFORMATION ABOUT THE LABOUR MARKET
Jobseekers may find it difficult to gather information about existing vacancies, and to accurately assess their prospects in the labour market. For example, in a developed country context, Spinnewijn (2015) finds that unemployed people overestimate how quickly they will find work, and consequently search too little and deplete their savings too quickly. Similarly, for developing countries, Abebe et al. (2017b) find that jobseekers overestimate the probability of being offered a job when they make an application. Providing information can help: Ahn et al. (2018) find that information about the competitiveness of specific vacancies helps jobseekers target their job applications more effectively.
One reason information may be difficult to obtain is that jobseekers may live far away from firms. This is particularly relevant in the labour markets of the poorest countries, where information technology has limited diffusion, and job search and applications require frequent use of public transport. In rural India, Jensen (2012) informed young women about business process outsourcing (BPO) jobs in the city and offered assistance with the application process. Treated women were 4.6 pp more likely to work in BPO jobs, and 2.4 pp (11%) more likely to work outside the home for pay. In more urban contexts, jobseekers who lack the cash to pay for bus fares to search for vacancies or attend job interviews will experience worse labour market outcomes (Abebe et al. 2017b, Abebe et al. 2018). Using a structural model, Abebe et al. (2017b) estimate that about 30% of the jobseekers who call to inquire about a job in Addis Ababa do not apply for the position because of credit constraints.4
Recent interventions in developed countries show that informing jobseekers can also be done cost-effectively through mass media, with flyer campaigns (Altmann et al. 2018), or leveraging the existing services of employment agencies (Belot et al. 2015).
Job fairs offer an alternative way to bridge the information gaps of jobseekers. In IGC research in Ethiopia, a job fair failed to create much employment, but led low-skill candidates to lower their reservation wages to more realistic levels (Abebe et al. 2017a). In the Philippines, a job fair also allowed attendees to learn about their labour market prospects, leading to a 10.6 pp increase (from 7.7%) in the probability of working in a formal job later (Beam 2016).
Information can also help to change norms that distort labour market outcomes. A particularly important set of norms are those related to the participation of women in the labour market. For example, in rural India, Bernhardt et al. (2018) found that women’s work outcomes were strongly associated with whether they thought their husbands approved of them working for pay outside the home. Norms become entrenched when people expect to be sanctioned by the community if they do not conform: in the same paper, men’s approval was correlated with whether they expected to be socially sanctioned. However, people may overestimate others’ disapproval: the men thought community disapproval was almost twice as high as it actually was.
Recent research suggests that correcting these false beliefs about the general support for social norms can improve Female Labour Force Participation (FLFP). In Saudi Arabia, where husbands can typically decide whether their wives can work, the vast majority of young married men in an experiment privately supported FLFP, but also underestimated support from similar men. Informing them of the true level of support increased sign-up for a job-matching service for their wives, and four months later, the wives were more likely to have applied and interviewed for a job outside of home (Bursztyn et al. 2018). Women’s beliefs also matter. In rural Uttar Pradesh, India, McKelway (2018) finds that improving women’s belief in their own ability to attain desired outcomes raises subsequent work for income by 36% (8 pp) after four months, possibly because women were inspired to exert more effort to find employment.
Norms become entrenched when people expect to be sanctioned by the community if they do not conform.
Other norms that are relevant in the labour market are related to the prestige of various occupations. Groh et al. (2015) in Jordan offered a job-matching service to firms and recent graduates. However, youth rejected 28% of interviews they were offered. Those youth who took a job left within a month 83% of the time. Survey evidence suggests that this was because the jobseekers perceived the jobs on offer to have low prestige. So far, there has been no work that tries to identify whether norms related to prestige are malleable.
NOT ALL INTERVENTIONS WORK, OR WORK FOR LONG
Timeframes matter, and not all interventions have long-lasting impacts. For example, Dammert et al. (2015) used SMS messages to provide information on job vacancies. Just giving access to a public dataset had no impact, and while adding other sources of information (e.g. newspaper ads) did increase employment, effects did not persist for more than three months. Further, Abebe et al. (2018) find that the formal employment effects of transport subsidies documented eight months after treatment do not persist in the four year follow up. Piloting and experimentation is thus crucial to determine whether a specific policy will work in a given context.