Hidden Income and the Perceived Returns to Migration
Last registered on October 10, 2018


Trial Information
General Information
Hidden Income and the Perceived Returns to Migration
Initial registration date
October 09, 2018
Last updated
October 10, 2018 8:45 PM EDT

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
Stanford University
Other Primary Investigator(s)
Additional Trial Information
On going
Start date
End date
Secondary IDs
Stanford IRB Protocol ID: 37631
The large income gap between urban and rural parts of many poor countries invites the question of why migration does not arbitrage this gap away. Using a sample of households in rural Kenya, I show that individuals on average underestimate urban wages considerably. Randomizing the provision of city-level labor market information to 464 households causes them to update their beliefs about the returns to migration and increases the rate of migration to Nairobi. Treated households earn more income and express a greater desire to send another migrant to Nairobi in the future. To understand how downward-biased beliefs can persist, I collect data on perceptions of migrant earnings from residents of origin villages and match them with data on actual migrant earnings. The data are consistent with strategic income misreporting by migrants: when migrants have a greater incentive to hide income from an origin resident, the origin resident underestimates the migrant's income to a greater degree, and reports a lower belief about his or her own potential return to migration. Finally, I experimentally shock rural individuals' beliefs about migrant earnings and show that this translates into higher beliefs about the returns to migration. Together, these results demonstrate that urban-rural information asymmetries can distort the migration decision and suppress migration to high-wage cities.
External Link(s)
Registration Citation
Baseler, Travis. 2018. "Hidden Income and the Perceived Returns to Migration." AEA RCT Registry. October 10. https://www.socialscienceregistry.org/trials/3101/history/35604
Experimental Details
This study includes two interventions.

Intervention 1 (Urban Labor Market Information): Randomized at the household level. Each treated household is given information sheets with labor market statistics for Nairobi, Kisumu, and Eldoret.

Intervention 2 (Migrant Remittances Information): Randomized at the household level. Treated households told the average share of income remitted by Nairobi migrants to origin (village) households.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
Intervention 1: Beliefs about the returns to migration, probability of migrating within the year following the intervention, income, self-reported welfare.

Intervention 2: Beliefs about average earnings in Nairobi (unconditional and for migrants), beliefs about the returns to migration, likelihood of migrating within the next year.
Primary Outcomes (explanation)
I have two measures of migration after the intervention for each of the following destination types: any destination, Nairobi, Kisumu, Eldoret, other urban destination, and other rural destination. The two measures are 1.) whether the household sent at least one migrant to that destination after the intervention, and 2.) the number of migrants the household sent to that destination after the intervention. Income will be measured as the sum of individual wage and enterprise income across family members (as defined by the household roster collected at baseline) plus estimated agricultural output (farm-gate value estimated by the household head).
Secondary Outcomes
Secondary Outcomes (end points)
Intervention 1: Remittances, savings, spending on food, investment, whether the household experienced a financial emergency in the past 3 months, whether the household is worried about their finances, and whether the household could cope with a financial shock of 2000 KES.

Intervention 2: Expected share of remittances for the marginal migrant from that household.
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
Households selected systematically from villages in Bungoma & Kakamega Counties.
Experimental Design Details
Not available
Randomization Method
Treatment status was assigned by a random number generator.
Randomization Unit
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
Intervention 1: 497 households. Intervention 2: 340 households.
Sample size: planned number of observations
Intervention 1: 497 households. Intervention 2: 340 households.
Sample size (or number of clusters) by treatment arms
Intervention 1: 249 control households, 248 treatment households. Intervention 2: 167 control households, 173 treatment households.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB Name
Stanford Institutional Review Board
IRB Approval Date
IRB Approval Number