Search Frictions in the Labor Market: The Effect of an Inclusive Job Matching Platform

Study Context

Job search costs in low-income countries, particularly in sub-Saharan Africa, can consume a significant portion of jobseekers’ resources, whereas firms face high screening costs in finding suitable candidates. Existing job-matching platforms often cater to educated jobseekers who seek white-collar jobs, operating on the assumption that they have internet access and are technology savvy. This excludes many young jobseekers in a different segment of the labor market with limited access to internet and digital skills – adversely affecting marginalized groups and deepening existing inequalities in accessing labor market opportunities. 

Researchers piloted and evaluated an innovative Text4Jobs platform, designed to be inclusive and enable job seekers to receive real-time job matches tailored to their preferences, while also addressing employers’ challenges of finding suitable candidates. In the platform, job seekers input their preferences (e.g. location, salary, and job type), and employers provide details about vacancies. An algorithm matches jobseekers with vacancies and alerts both via text notifications.

Study Design

Researchers implemented a randomized control trial in Accra, Ghana, targeting job seekers and firms. 300 job seekers were randomly assigned into one of three groups: 1) a group receiving job match notifications via text messages; 2) a group receiving text message notifications and bi-weekly reminders to apply for jobs; or 3) a group receiving no intervention. Moreover, qualitative data from interviews with jobseekers and firms will be used to refine the platform’s features and adapt future interventions.

Results and Policy Lessons 

Preliminary results show that the Text4Jobs platform improved job matching by increasing interview attendance and job notifications. While application and offer rates rose slightly, they were not statistically significant. The platform worked especially well for more experienced jobseekers, though education mismatches suggest a need for algorithm refinement. These early findings highlight the promise of low-tech tools to reduce job search frictions in low-connectivity settings.