Over the past decade, CSAE researchers have been investigating one of the defining challenges of African urbanisation: why, in rapidly growing cities full of young workers and expanding firms, so many people remain unemployed or underemployed, or trapped in jobs poorly matched to their skills. 

From Addis Ababa to Johannesburg, this body of work has shown that the problem is often not simply a shortage of jobs, nor merely a shortage of skills. Instead, labour markets are shaped by search and matching frictions, the barriers that prevent workers and firms from finding one another efficiently, understanding each other accurately, and making good decisions. 

Across a series of studies, CSAE scholars have revealed that urban labour markets are frequently obstructed not only by distance and cost, but also by missing information, distorted beliefs, and misdirected search. Together, this research has put a spotlight on youth unemployment in African cities not as a static problem of scarcity, but as a dynamic challenge of coordination.

As African cities have expanded, they have brought workers and firms physically closer together than ever before. Yet proximity alone has not guaranteed employment. 

In ‘Location, Search Costs and Youth Unemployment: Experimental Evidence from Transport Subsidies’ (The Economic Journal (Volume 128, Issue 614 pp. 2353-2379), Simon Franklin (former Dphil student at the CSAE), examined whether where you live within a city can shape whether you find work. Working in Addis Ababa, Franklin showed that job search itself was prohibitively expensive. Small transport grants significantly increased search intensity and improved access to stable, formal-sector jobs.

The findings demonstrated that cash constraints and urban geography matter profoundly. Even in one city, distance from opportunity could exclude workers from better employment, meaning that labour market exclusion could stem from search costs, not just a lack of vacancies. 

A related strand of experimental research shows that these frictions are not confined to transport alone. In ‘An Adaptive Targeted Field Experiment: Job Search Assistance for Refugees in Jordan’ (Journal of the European Economic Association, 2024), Stefano Caria, Grant Gordon, Maximilian Kasy, Simon Quinn, Soha Osman Shami, and Alexander Teytelboym studied interventions designed to support Syrian refugees and local jobseekers in Jordan. The paper finds that financial support for job search increased search intensity and improved employment outcomes, particularly when assistance was adaptively targeted towards those most likely to benefit. Together with Franklin’s work, the findings suggest that labour market inefficiencies often arise not simply because jobs are unavailable, but because workers face substantial informational and financial barriers in accessing them. 

But if reducing transport barriers helps, why are labour markets still so inefficient even when workers can search more widely?

In ‘Anonymity or Distance? Job Search and Labour Market Exclusion in a Growing African City’ (Review of Economic Studies, 2021), Girum Abebe, Stefano Caria, Marcel Fafchamps, Simon Quinn and collaborators expanded this analysis beyond physical barriers to examine how difficult it was for job seekers to demonstrate their abilities to employers. 

Comparing transport subsidies with job application workshops, designed to equip jobseekers with credible signals of their abilities, in Addis Ababa, they found that both interventions improved short-run formal employment, but only the job application workshops produced lasting gains. Four years later, participants in the workshop had higher earnings, longer employment duration, and greater job satisfaction. This implied that in many urban labour markets, workers’ lacked the skills to apply for jobs and ways of providing information on their skills to employers. Job application workshops changed trajectories far more durably than merely helping them travel farther. This added a conceptual shift for the field where urban labour markets were not only fragmented by transport costs, they were also fragmented by anonymity and lack of credible information.

CSAE research also illuminated the employer perspective. In ‘Job Search and Hiring with Limited Information about Workseekers’ Skills’ (American Economic Review, 2022), Eliana Carranza, Robert Garlick, Kate Orkin, and Neil Rankin, studied the effects of providing credible, shareable skill assessments on both sides of the labour market. They study an intervention assessing young jobseekers’ skills, and helping both workseekers and firms to learn the assessment results by providing workseekers with a report certifying their skills. This increases workseekers’ earnings and employment. They show that effects occur because both firms and workseeks respond to the information. Firms respond by being more likely to interview workers who provide signals with their applications. Workers respond by directing their search to jobs that suit their skills. Conceptually, this work shifted the field in highlighting that both firms and workers face an information problem. Returns may be higher on interventions providing information to both sides of the market, such as certification, than one-sided information, such as assessments by firms of job applicants’ skills or career counselling by job search assistance programmes.

If workers struggle to reach firms, firms may also struggle to identify talent. In ‘The Selection of Talent: Experimental and Structural Evidence from Ethiopia’ (American Economic Review, 2021), Girum Abebe, Stefano Caria and Esteban Ortiz-Ospina showed that application costs disproportionately deter high-ability candidates. Firms therefore, may not simply choose from the best workers available, but from a distorted applicant pool shaped by frictions. By incentivising applications, researchers found firms attracted more talented candidates. Search frictions were not simply slowing hiring—they were distorting talent allocation itself. 

More recent work ‘Formalised Employee Search and Labor Demand’ by Lukas Hensel, Tsegay Tekleselassie, and Marc Witte extended this logic to small and medium enterprises in Ethiopia. Reducing the cost of formal vacancy advertising encouraged firms to create more ambitious white-collar roles, yet many struggled to fill them, especially when unfamiliar with formal recruitment channels. This highlighted the barrier that institutions and habits matter. Even when firms expand search, weak familiarity with formal systems can constrain hiring

A particularly innovative contribution of CSAE’s labour market research has been to show that frictions do more than slow matching, they shape beliefs. 

Building on earlier work, ‘Jobseekers’ Beliefs about Comparative Advantage and (Mis)Directed Search’ (revise and resubmit, American Journal of Applied Economics) by Andrea Kiss, Rob Garlick, Kate Orkin, and Lukas Hensel showed that skill information also redirected workers toward jobs aligned with their comparative advantage. Jobseekers were often searching hard, but sometimes for the wrong jobs. Search effort alone, this work suggests, is insufficient if search is misdirected. Workers who received better information about their skills applied more to jobs which better suited their skills and increased their earnings.

Beliefs about the labour market extend beyond workers’ beliefs about their skills to firms’ and workers’ beliefs about the labour market. In ‘Matching Frictions and Distorted Beliefs: Evidence from a Job Fair Experiment’ (Economic Journal, 2025) Abebe, Caria, Fafchamps, Franklin, Quinn and Shilpi found that job fairs to introduce firms and workers generated surprisingly few direct hires. They did correct overoptimistic beliefs held by both workers and firms.

Before the intervention, both sides often misunderstood market realities, leading to poor search strategies. Workers overestimated their chances, while firms misunderstood applicant pools. By exposing participants to real conditions, the intervention shifted expectations and improved employment outcomes for less educated workers. This was a subtle but important finding. Labour market frictions do not simply create barriers, they can also preserve false assumptions that worsen decision-making.

Together, this research moves beyond simplistic narratives of unemployment as either “not enough jobs” or “not enough skills.” Instead, it presents cities as search ecosystems where frictions at multiple stages, from discovering vacancies, to applying, to signaling, to screening, shape outcomes. 

The policy implications are powerful. Effective labour market interventions need not always create jobs directly. Sometimes they can unlock employment by improving the functioning of search itself. Low-cost interventions like transport subsidies, application workshops, skill certification, recruitment support, and job fairs can meaningfully improve labour markets when they target the right friction. In ongoing work, CSAE researchers are now trying to better understand the implications of these frictions in general equilibrium. 

This cluster of research stands as an example of the CSAE’s endeavour to combine rigorous field experimentation with deep structural questions to address pressing economic challenges.