Money or Ideas? A Field Experiment on Constraints to Entrepreneurship in Rural Pakistan By Xavier Gine and Ghazala Mansuri Comments by Elena Bardasi, March 30, 2011
Policy questions Is the intervention worth replicating/scaling-up? 1.Does the intervention have an impact? 2.What explains the impact (or lack of impact)? – Selection issues – targeting – Implementation issues – Sources of the impact 3.Cost effectiveness
Randomization, selection, and targeting Treatment and control equally made of: ½ COs male (a) 747 COs randomization ½ COs female ORIENTATION SESSION 61% of individuals (a) / 90% of businesses for 55% eligible for loans (2284) (4162 / 2532) Special meeting to conduct (b) BASELINE SURVEY BT offeredNo BT offered Total2080Total2080 Trained1252 Interested in training? Yes (about 2600)no trained not trained L offeredL not offered Total1142Total1142 Take-up211 Willing to apply? NoYes (713) obtainednot obtained Randomization?
Definition of the ‘treatment’ and implementation issues Does same treatment imply same implementation modalities? – Is a pair of male and female trainers the same treatment for men and women? – Do different sectors require the same business training? What is training?
Sources of the impacts Exploring heterogeneity What is the effect of the combination of business training and education? Interaction with agricultural activities (50% of men have a agribusiness vs. only 20% of women)
Gender effects Gender dimensions: – Household level analysis. To what extent do women actually manage their loans? (What do their husband do? Do they have enterprises?) – Do women need “more intensive interventions”? Lower opportunity costs What do women need to increase their sales?