Public work programs vs active labour market programs: evidence from Papua New Guinea

·1 hr 1 min·5 clips

Researchers used youth who missed job training to volunteer at a FIFA tournament as an accidental control group to measure program impact.

This Devpolicy Seminar features a presentation by PhD researcher Chris Hoy on an impact evaluation of Papua New Guinea's Urban Youth Employment Program (UYEP). Host Stephen Howes introduces Hoy's work, which forms part of his thesis on impact evaluations in PNG, a rare focus. Hoy begins by defining impact evaluation as measuring downstream outcomes against a counterfactual, crucial in a context where Port Moresby's formal employment has fallen 20% since 2013. He outlines two program types: public works programs paying for manual labor like cleaning, and active labour market programs (ALMPs) aiming to skill and match people with employers. Hoy notes a global meta-analysis by the World Bank's David McKenzie found ALMPs in developing countries create only about two extra jobs per 100 participants. The UYEP uniquely combines both approaches, offering basic training, employer matching, and a five-month wage subsidy. The evaluation's core question is the additional impact of the ALMP component beyond public works. To estimate the counterfactual, Hoy exploited a natural experiment: in one intake, eligible youth were diverted to volunteer at the FIFA Under-23 Women's World Cup instead of receiving ALMP training, creating a comparable control group. Pre-program data shows treatment and control groups were similar in education, test scores, and income distribution. Results indicate the ALMP component increased the likelihood of having a formal job by approximately 25 percentage points, with almost half the treatment group employed versus 14% of the control group. This effect was entirely driven by interns staying on with their host employers. The program also increased job-seeking behavior among those not retained. However, interest in further education declined, more sharply in the control group. A concerning finding is that unemployed participants in the ALMP component showed increased antisocial and criminal behavior compared to the control group, while employed participants showed decreases. Surveyed employers valued all three program components—free labor, job matching, and pre-training—with the subsidy being a key incentive. Hoy notes limitations, including potential displacement of other workers and the need to disentangle which components drive results. Future research will test adding cognitive behavioral therapy and examine equilibrium labor market effects. The presentation concludes with Q&A discussing gender, selection processes, and qualitative research.

As heard by us

An impact evaluation examines whether Papua New Guinea's largest employment program led to formal-sector work.

Papua New Guinea's Urban Youth Employment Program is treated here as a practical test of a wider policy question: whether public works and active labour market programs can move people into steadier employment.

Read the full review in PlayNext →

Why you'd press play

You want evidence on whether PNG's Urban Youth Employment Program helped move young people into formal-sector work.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Devpolicy Talks