Beliefs, Reputation, and Barriers to Entry in Online Labor Markets  (Job Market Paper)

with Hamna Ahmed and Zunia Saif Tirmazee

World Bank Development Impact Blog | AEA RCT Registry

Hundreds of millions of workers in developing countries seek digital jobs online but face entry barriers without an established reputation. Novice workers could offset this by lowering initial wages, yet few do. Our baseline survey points to two explanations: workers believe employers interpret low wages as low quality signals and are uncertain about their own abilities. We conduct two field experiments on a leading global freelancing platform to examine how these beliefs shape worker outcomes. In the demand-side experiment, we randomize wage offers by novice workers to 703 jobs and find that workers misperceive employer behavior: low wage offers significantly improve their hiring prospects. In the supply-side experiment with 481 novice workers, we randomly provide them with accurate information about employer responses and their performance. Correcting workers’ beliefs increases their willingness to lower wages. Consistent with reputation models, effects are driven by high-ability novices with high returns to reputation once these frictions are removed. Simulations show that without external intervention, worker learning about employer responses is slow and costly. Our findings highlight that information interventions can help workers in developing countries overcome reputation barriers and accelerate talent discovery in online labor markets.

Intra-Household Incentive Design: An Experiment on Parent-Child Decision Dynamics in Pakistan 

with Hamna Ahmed, Zunia Saif Tirmazee, and Emma Zhang

Chicago Booth Review | AEA RCT Registry

How should we design and target incentives for skills investment in young adults who live with their parents? We study the role of intra-household payment and information targeting on the effectiveness of a conditional cash transfer (CCT) program using a randomized control trial. The program aimed to boost completion of a digital skills training program among young females in urban Pakistan. Fixing the incentive size and daughters’ knowledge about it, we cross-randomized (1) the payment split between parents and daughters and (2) whether parents received information about the daughters’ incentive. Under asymmetric information about the CCT, incentivizing parents increased training completion by 103% compared to incentivizing daughters, driven by increased parental nudging. When both parents and daughters knew about the CCT, completion rates did not vary by the incentive split. Two-year follow-up reveals that young women who completed the training programs are more likely to be working and demonstrate more extensive computer use. Our results suggest that in this parent-child context, incomplete information sharing is the main barrier to the optimal incentive targeting for skill acquisition. 

Intermediation by Superstars: Evidence from Online Labor Markets - Pilot completed 

Do Wedding Vendors Discriminate? Experimental Evidence from the Wedding Industry in Pakistan - Pilot ongoing

with Sher Afghan Asad and Sarah Shaukat

Estimating Global Poverty in Stata: The povcalnet command

with R. Andres Castaneda Aguilar, Christoph Lakner, Espen Beer Prydz, Jorge Soler Lopez, and Qinghua Zhao. 

Global Poverty Monitoring Technical Note No. 9, World Bank.