Barriers to Search and Hiring in Urban Labour Markets

In rapidly growing and changing urban labour markets in lower income countries, firms may also face barriers in the recruitment and retention of workers, perhaps contributing to the disproportionately large number of small firms that exist in these countries. In this subsection, we aim to shed light on firm-level factors that influence hiring in lower-income countries. 

Limited information

We start by considering the employer’s hiring problem. Once an employer decides to hire a new worker, they must both locate and screen applicants. To locate applicants, the employer must decide how and where to post a vacancy— i.e. whether they should pursue referrals, post on online/offline job boards, work with employment agencies, etc. Upon gathering applications, the employer must screen applicants to determine their suitability for the job, paying particular attention to characteristics related to productivity. As limited information[1]1. We use the term “limited information” rather than the more common “asymmetric information” as it is now understood that workers may also have imprecise beliefs about their own productivity-relevant traits.  on productivity-relevant traits can lead to costly hiring mistakes, employers’ beliefs and experiences about different hiring strategies are therefore bound to shape their choice of recruitment methods and, potentially, the growth prospects of their business. We first consider evidence for how the screening problem shapes hiring and then whether firms face significant costs in attracting qualified applicants to vacancies. 

Referrals / hiring via networks 

Hiring through networks—i.e. through family, friends, co-workers or their referrals— is very common in labour markets throughout the world. A literature examines why firms find it advantageous to hire through networks. A common hypothesis is that hires through referrals mitigate an information asymmetry (either about the worker’s unobserved type or their effort), help match workers to jobs which they are less likely to leave, or reduce search frictions.[2]2. Evidence that referrals provide information about a worker's type (Pallais and Sands 2016) or improve match quality (Burks et al. 2015, Friebel et al. 2023) comes from papers that focus on high-income countries.

In a low-income country setting, a weaker set of institutions and legal infrastructure may exacerbate moral hazard issues in the workplace and raise the gains to firms for hiring referred workers whose referrers can monitor them and increase their effort. Heath (2018) argues that referrals help solve a moral hazard problem in garment factories in Bangladesh: factories appear to give a linked contract to the referral provider and referral recipient that punishes the referral provider if the recipient does not perform well, which provides the referral recipient incentives for high effort.

However, referrals may not improve efficiency. In a model in which a job provides a rent because wages are above equilibrium, firms may use the ability to grant a referral as a reward to a favoured employee. Using Chinese data, Wang (2013) finds that the death of a father-in-law decreases a man’s wage by 7%. She interprets the fact that this effect is stronger in state-owned enterprises as evidence that the channel for this effect is nepotism, rather than a model in which the father-in-law provides information about the son-in-law that is valuable to firms. In the Egyptian context, Osman et al. (2022) find that when high-productivity establishments hire individuals socially connected to the owner, these employees tend to be less productive, consistent with nepotism. By contrast, social connections of employees who are not owners tend to be more productive.[3]3. Relatedly, Magruder (2010) documents that intergenerational correlations in industry-specific employment in South Africa are consistent with parents helping children find jobs, which might reflect information transmission to firms about their children or directly to their children about applying for jobs. The paper does not assess the extent to which network-based hiring is productive.

A second body of research has examined the distributional consequences of network-based hiring. This is a natural question to study given that social networks often exhibit homophily — a tendency for similar individuals to connect to one another. Beaman and Magruder (2012) demonstrate in India that an employee’s choice of referral is affected by experimentally-manipulated rewards for the employee’s performance. This result suggests that, while referrals do contain valuable information, they also operate within networks of favour-exchange, which might not work in firms’ favour. Moreover, these networks differentially hurt women jobseekers, whom men seem reluctant to refer (Beaman et al. 2018).[4]4. This echoes findings from the US suggesting that referral networks might be concentrated within race groups (Ioannides and Loury 2004).  More broadly, Chandrasekhar et al. (2020) show that network-based job search and hiring can be a locally efficient solution to imperfect information but be globally inefficient by keeping hiring too narrow.

It is also important to consider that social networks respond to the incentives generated by policies in ways that can worsen the distributional consequences of network-based hiring, and also harm firms. Caria et al. (forthcoming), discussed in the section on jobseekers learning through networks, find that a job-search assistance intervention actually decreases information-sharing between treated individuals and the network members with whom they search for jobs. This is consistent with a model in which individuals garner benefits from membership in job-search networks, but if they gain a sufficiently large individual-specific advantage, it is no longer valuable to help other network members.[5]5. Beaman (2012) documents a similar pattern of rivalry among job-seeking refugees in the United States; larger previous cohorts of co-nationals improve a refugee's labour market outcomes (because these cohorts are likely to already have jobs and thus help newly arrived via referrals), while arriving in a larger current cohort hurts a refugee's labour market outcomes (because these are rivals).  Chiplunkar et al. (2023) show that when job information becomes more rival, jobseekers share less information and tend to skew their information sharing towards lower ability peers. This decreases the quality of the applicant pool that can be attracted by relying on information diffusion in social networks.

Finally, it is important to understand the global efficiency implications of the common reliance on network-based hiring. Chandrasekhar et al. (2020) study this question. The key intuition of the paper is that even if individual firms find it beneficial to hire using referrals because of information frictions, overall output may still be lower in a world of hiring through referrals because hiring is limited by the social networks of existing employees. Indeed, they demonstrate theoretically that hiring through networks lowers efficiency and can result in what they call an “information poverty trap”, where entrepreneurs do not make investments in learning the true distribution of worker ability. 

Hiring outside networks 

If networks limit the scale and scope of hiring, in order to grow, firms must be able to successfully recruit from the external labour market. A number of recent experimental studies investigate how limited information can hinder firms’ willingness to hire outside their networks and shed light on interventions that can improve hiring.

Interventions to signal jobseekers’ work experience to firms

Hardy and McCasland (2023) evaluated a government apprenticeship programme in Ghana that placed young unemployed people with small firms. They find that total employment at treated firms increases by approximately half a worker and these firms experience higher revenues and profits. The authors present evidence that the extensive time commitment of the apprenticeship programme served as a non-monetary screening device retaining higher-ability apprentices who were then successfully placed and employed. Loiacano and Silva-Vargas (2023) find similar results of firms learning from directly observing their own workers’ performance in a more specific labour market in Uganda - they find that working with one refugee worker increases firms’ likelihood of hiring more refugees by adjusting owners’ beliefs about refugees’ skills. References from previous employment can also help more directly - Abel et al. (2020) show that female South African jobseekers can increase employment by using reference letters from past employers.

Beam and Quimbo (2023) find that subsidised summer employment for low-income youths in the Philippines led to increased employment a year later, and find that this is most consistent with work experience serving as a signal of unobservable applicant quality. Beam et al. (2020) supplement this with the finding that on average, work experience, and not technical and vocational education and training (TVET) or college experience, affects callback rates in Manila, though they cannot distinguish between the channels of human capital accumulation or signalling of worker quality driving this result. Godlonton (2019) finds in urban Malawi that work experience gained via experimentally varying the probability of a short-term job offer to youth leads to an 11 percentage point increase in average employment, and finds that effects are highest for workers with low numeracy and literacy scores. She does not find evidence for expanded social networks or letters of reference being the relevant mechanisms, suggesting the work experience itself may be the predominant positive signal to firms that employ the treated youths.

Certification

Certification interventions — which we have reviewed in Section IIA —. have been shown to have consistent positive effects for firms and jobseekers.

Firm Beliefs

In a lab-in-the-field experiment in Ghana, Caria and Falco (2022) show that employers’ beliefs about worker trustworthiness are biased and affect willingness to hire. In the game, the authors allow real employers the opportunity to hire an anonymous worker for a simple task. They elicit beliefs from employers about shirking on the task and find that employers underestimate worker trustworthiness, as 20% more workers successfully complete the task than estimated by employers. These beliefs matter for hiring: employers who hire more in real-life have more positive expectations of worker trustworthiness and are more willing to trust in the game. 

Fernando et al. (2023) show that combining applicant identity verification with advertising services on an online job portal increases hiring from the job portal and improves overall vacancy filling. In the experiment, advertising services expand access to skilled applicants and identity verification enables improved screening of these unfamiliar applicants. Verification information likely reassured employers of applicants’ skills, trustworthiness, or interest in the job, leading to greater recruiting effort on these out-of-network applicants.

In contrast to the studies discussed thus far, Hensel et al. (2022) do not find evidence of firm response to an applicant screening intervention combined with subsidies for formal vacancy creation for small Ethiopian firms. However, treated firms had also shifted their hiring efforts towards white-collar vacancies, which can be harder to fill and which the firms had little experience recruiting for. This may explain the lack of impacts on hiring. Groh et al. (2016) also find that an initial wage subsidy voucher given to female graduates in Jordan lead to a short term increase in employment, but this subsides in the long run, and the authors hypothesise that this is likely due to productivity levels not rising above a binding minimum wage in this setting. 

Finally, a paper by De Mel et al. (2019) provides evidence on the long-term impacts of a temporary wage subsidy offered to small firms in Sri Lanka. This tests for hiring frictions: in the absence of frictions, a wage subsidy should boost employment only for the duration of the subsidy, whereas, in the presence of frictions, impacts on employment should be persistent or at least should decline only gradually over time. The paper finds that the wage subsidy leads to a 10 percentage point increase in employment one year after the end of the subsidy, but effects decline quickly after that. However, when offered in conjunction with either capital or training, the wage subsidy has a positive impact on employment that lasts at least four years after the end of the subsidy. The effect of the combined treatments are significantly greater than the sum of the individual effects of wage subsidy alone and capital or training alone, suggesting that there may be a genuine complementarity between these interventions. 

Policy take-aways

Overall, several studies suggest that limited information on worker experience, skills and trustworthiness can indeed limit hiring and are promising targets for intervention, with a now-substantial evidence base. Furthermore, the impacts of information and other interventions may be muted if they are not combined with complementary treatments that address other constraints (as in both Fernando et al. 2023 and De Mel et al. 2019).

Cost of vacancies and attracting applicants

Alongside the screening problem, firms could face high costs for posting vacancies through formal channels or be unfamiliar with methods to attract skilled applicants to vacancies. 

Two studies directly subsidise vacancy posting costs and do not find evidence of greater hiring. Hensel et al. (2022) subsidise vacancy creation for small firms in Ethiopia. In their context, the cost of posting a vacancy through formal channels like online/offline job boards or newspapers are significant, exceeding the average monthly wage of employees. Treated firms respond by creating more vacancies through formal means and searching for higher-skilled workers, but do not alter the total number of vacancies. In fact, as desirable applicants fail to materialise, treated firms report a lower likelihood of hiring overall. Their results suggest that costs of formal vacancy creation are not a barrier to firm hiring in their context. 

In a similar vein, one of the interventions in Fernando et al. (2023) involves preferential advertising for small firms already posting vacancies on an online job portal in India.[6]6. Unlike in Ethiopia, the costs for this preferential advertising was less than 5% of posted monthly wage for the vacancy.  While this advertising doubles the number of applicants and improves access to observably skilled candidates for firms, treated firms do not increase their hiring significantly more than control firms. The authors explain that larger applicant pools likely increased the screening burden for firms and, without technologies to assist with screening, employers may not benefit from the larger recruitment pools accessible through internet-based recruitment. 

On the other hand, two studies that consider alternative interventions to improve the chances of meeting desirable applicants show that firms may fail to optimise recruitment efforts. Abebe et al. (2021b) randomise application incentives, which are calibrated to cover the transport and the time-cost of application, to job seekers interested in working at a large firm. Treated job seekers are more likely to apply. Importantly, the incentive induces high-ability, but disadvantaged, job seekers to apply more than in the control group. Moreover, employers do not anticipate this result, expecting the incentive to attract worse applicants, suggesting that employers may fail to optimise their recruitment strategies. 

Relatedly, Abebe et al. (2022) find that a randomised job fair intervention in Ethiopia leads employers to change search strategies and effort. After attending job fairs, treated employers post more vacancies, particularly for professional positions, and are more likely to use external job boards for vacancy posting. The authors contend that attending the job fairs corrects initial overoptimism of firms about the ability and experience of highly educated workers and reduces the reservation quality of workers the firm is willing to hire. Yet, despite these changes to search behaviour, there is limited change in hiring at treated firms.

Policy take-aways

Overall, while there is some evidence to suggest that firms do not optimise recruitment strategies and efforts, likely due to limited experimentation with different recruitment methods, the evidence does not currently establish that this limits the quantity or likelihood of hiring. However, it could still be that the lack of experimentation with recruitment technologies may influence the quality of hired workers, as suggested by Abebe et al. (2021b), and worker turnover. More work is needed to understand the implications of hiring costs in different settings.

For example, some educational programmes offer matching services for graduating students. Banerjee and Chiplunkar (2023) study the behaviour of placement officers at vocational training institutions across multiple Indian states. They show that placement officers have systematically inaccurate information about students’ preferences, that giving them this information changes which students they recommend for interviews at specific jobs, and that this can increase some proxies for subsequent match quality.

References 

Abebe, G, A S Caria, and E Ortiz-Ospina (2021b), "The selection of talent: Experimental and structural evidence from Ethiopia," American Economic Review, 111(6): 1757–1806.

Abebe, G, S Caria, M Fafchamps, P Falco, S Franklin, S Quinn, and F Shilpi (2022), "Matching frictions and distorted beliefs: Evidence from a job fair experiment," Working Paper.

Abel, M, R Burger, and P Piraino (2020), "The Value of Reference Letters: Experimental Evidence from South Africa," American Economic Journal: Applied Economics, 12(3): 40–71.

Banerjee, A, and G Chiplunkar (2023), "How Important Are Matching Frictions in The Labor Market? Experimental and Non-experimental Evidence from a Large Indian Firm," Manuscript, MIT.

Beaman, L A (2012), "Social networks and the dynamics of labour market outcomes: Evidence from refugees resettled in the US," The Review of Economic Studies, 79(1): 128-161.

Beaman, L, N Keleher, and J Magruder (2018), "Do job networks disadvantage women? Evidence from a recruitment experiment in Malawi," Journal of Labor Economics, 36(1): 121-157.

Beaman, L, and J Magruder (2012), "Who gets the job referral? Evidence from a social networks experiment," American Economic Review, 102(7): 3574-3593.

Beam, E, and S Quimbo (2023), "The Impact of Short-Term Employment for Low-Income Youth: Experimental Evidence from the Philippines," The Review of Economics and Statistics, 105(6): 1379–1393.

Beam, E A, J Hyman, and C Theoharides (2020), "The Relative Returns to Education, Experience, and Attractiveness for Young Workers," Economic Development and Cultural Change, 68(2): 391-428.

Burks, S V, B Cowgill, M Hoffman, and M Housman (2015), "The value of hiring through employee referrals," Quarterly Journal of Economics, 130(2): 805-839.

Caria, A S, and P Falco (2022), "Skeptical Employers: Experimental Evidence on Biased Beliefs Constraining Firm Growth," Review of Economics and Statistics, 1-45.

Chandrasekhar, A G, M Morten, and A Peter (2020), “Network-Based Hiring: Local Benefits; Global Costs,” Working Paper 26806, National Bureau of Economic Research.

Chiplunkar, G, E M Kelley, and G Lane (2023), “Competitive Job Seekers: When Sharing Less Leaves Firms at a Loss,” Working Paper.

De Mel, S, D McKenzie, and C Woodruff (2019), “Labor Drops: Experimental Evidence on the Return to Additional Labor in Microenterprises,” American Economic Journal: Applied Economics, 11(1): 202-235.

Fernando, A N, N Singh, and G Tourek (2023), "Hiring frictions and the promise of online job portals: Evidence from India," American Economic Review: Insights.

Friebel, G, M Heinz, M Hoffman, and N Zubanov (2023), "What do employee referral programmes do?" Journal of Political Economy, 131(3): 633-686.

Godlonton, S (2019), “Employment Exposure: Employment and Wage Effects in Urban Malawi,” Economic Development and Cultural Change, 68(2): 471-506.

Groh, M, N Krishnan, D McKenzie, and T Vishwanath (2016), “Do wage subsidies provide a stepping-stone to employment for recent college graduates? Evidence from a randomized experiment in Jordan,” Review of Economics and Statistics, 98(3): 488–502.

Hardy, M, and J McCasland (2023), "Are small firms labor constrained? experimental evidence from Ghana," American Economic Journal: Applied Economics, 15(2): 253-284.

Heath, R (2018), “Why do firms hire using referrals? Evidence from Bangladeshi garment factories,” Journal of Political Economy, 126(4): 1691-1746.

Hensel, L, T Tekleselassie, and M Witte (2022), "Formalized employee search and labor demand," Working Paper.

Ioannides, Y M, and L D Loury (2004), “Job Information Networks, Neighborhood Effects, and Inequality,” Journal of Economic Literature, 42: 1056–1093.

Loiacano, F, and M Silva-Vargas (2023), “Matching With The Right Attitude: The Effect of Matching Firms With Refugee Workers,” Working Paper.

Magruder, J R (2010), "Intergenerational Networks, Unemployment, and Persistent Inequality in South Africa," American Economic Journal: Applied Economics, 2(1): 62-85.

Osman, A, J D Speer, and A Weaver (2022), "Connections, Referrals, and Hiring Outcomes: Evidence from an Egyptian Establishment Survey," Journal of Economic Behavior and Organization, 204: 342-355.

Pallais, A, and E G Sands (2016), "Why the referential treatment? Evidence from field experiments on referrals," Journal of Political Economy, 124(6): 1793-1828.

Wang, S Y (2013), “Marriage networks, nepotism, and labor market outcomes in China,” American Economic Journal: Applied Economics, 5(3): 91-112.