What causes people to migrate to cities?

Why do few rural youth move, even when cities pay more?

There are enormous wage gaps between urban and rural areas in lower-income countries, yet every year only a small number of people move – even temporarily – to try to earn a better living (Lagakos 2020). A common explanation is that rural households face liquidity constraints: they cannot afford the initial transport, housing, or job search costs in the city. But there is another possible friction: potential migrants may not know what city jobs really pay. For those without social contacts in the city, it is especially difficult to receive support and information before and after they migrate.

In recent research on Kenya (Barnett-Howell, Baseler, Ginn, and Gordeev 2025), we identified five rural counties with high rates of underemployment among young adults. We then implemented a randomised intervention designed to directly tackle the information and network frictions that may be suppressing migration.

Providing rural Kenyans with information about the city

Households were randomly assigned, at the village level, to one of three interventions:

1. Information: Enumerators visited the household and delivered a brochure with simple statistics about income, employment rates, and prices in Nairobi, and read a script explaining it with built-in time for questions. Simple information treatments like this can increase migration when beliefs about urban wages are low (Baseler 2023).
2. Mentor: In addition to the information brochure and script, households were offered the opportunity to be connected to a mentor living in Nairobi. We recruited mentors from neighbourhoods with high concentrations of migrants, focusing on people with experience in the occupations migrants want to work in. Mentors made themselves available for phone calls or text messages for approximately one month and agreed to meet new migrants in Nairobi.
3. Group sessions: Rather than deliver information door-to-door, we invited households to a meeting with others in their village, during which they received the same information brochure and script. A group discussion followed, designed to encourage information exchange and help migrants coordinate their trips.

There was also a pure control group that received no intervention.

Importantly, we did not provide cash, transport vouchers, or job offers. The goal was to isolate whether addressing information gaps – or offering a low-cost social contact in the city – was enough to help new migrants.

Rural Kenyans were pessimistic about Nairobi earnings

In an initial survey, we found that rural respondents systematically underestimated Nairobi incomes (Figure 1). On average, people perceive a roughly 30–50% premium for workers in Nairobi compared to workers of the same age, gender, and education level in smaller towns in their home county. In representative survey data, those premiums range from 50 to 200% for most groups.

Figure 1: Information gaps in rural Kenya about Nairobi income

Notes: Income premiums between Nairobi and small, local towns by county and demographic group. Hollow symbols show the average perception about incomes in Nairobi relative to the average perception about small-town incomes; filled symbols show the corresponding actual income premium measured in representative household survey data.

This gap matters. If you think Nairobi only offers marginally better pay than you can find locally, then migration does not look worth the challenges and risk. But if you learn that entry-level security guards, shop assistants, and drivers are often earning multiples of rural day wages, that calculation changes.

Improved information increased migration and income

We track outcomes one year after the intervention. In all three arms (Information, Mentor, and Group), households were about two percentage points more likely to have migrated to Nairobi after the treatment. That’s around an 18% effect size. 

The Information and Mentor arms increased household income by about 9% (Figure 2); on a broader measure of well-being, the Mentor addition looks even better than Information alone. Those are big economic impacts, given the small absolute change in the number of migrating households. We estimate that the return to migration is roughly a 150% boost to household income, and that positive spillovers onto non-migrating households comprise a significant share of total economic gains. These spillovers do not appear in learning or migration outcomes for untreated households; instead, they show up as local business creation and hiring as migrants remit or bring home higher urban earnings. From the perspective of the rural economy, migrant remittances look like the cash transfers studied in Egger et al. (2022).

Figure 2: Treatment impacts on measures of household income

Notes: Household income includes wages from casual and formal work and business profits for current and former household members and is measured over the month preceding the survey. Price-adjusted income is deflated by a CPI based on where the income was spent. Amenity-adjusted income assigns an income value to migrants equal to the income they state would make them indifferent between living in Nairobi and their hometown. P-values compare treatment coefficients to Group.

Were these income gains offset by higher prices in the city? No, adjusting for a CPI specific to our study region doesn’t affect our estimates. What about differences in quality of life between Nairobi and rural villages? Apparently not – when we asked migrants to compare their overall quality of life in Nairobi and their hometown, around 80% said Nairobi was strictly better (Figure 3). This finding echoes Gollin et al. (2021): across sub-Saharan Africa, amenities like clean air and improved utilities are more often found in cities.

Figure 3: Most migrants report preferring life in Nairobi

Source: Phone surveys with migrants living in, or having returned from, Nairobi.

Overall, our findings add to a growing body of work showing that migration interventions in low-income settings can help people make better decisions about where to live and work (Bryan et al. 2014, Akram et al. 2017, Bazzi et al. 2021, Baseler 2023).

Be wary of group dissemination

Our Group arm increased migration by at least as much as individually provisioned information and mentorship. But it yielded no measurable economic gains: impacts on income are close to zero and statistically significantly lower than those in the other two experimental arms.

What happened? Our data shows that households without prior migration experience were less engaged in group meetings. Households with more migration experience dominated those discussions, and were subsequently the most likely to migrate. In contrast, door-to-door information and mentorship drew in more inexperienced migrants (Figure 4). This isn’t the first time that group-based delivery has worsened outcomes: Chandrasekhar et al. (2022) and Banerjee et al. (2024) find similar results in non-migration contexts.

Figure 4: Information and Mentor treatments draw in inexperienced migrants

Notes: Selection coefficients estimated from regressions of pre-treatment characteristics on treatment indicators among migrating households. Information and Mentor changed the composition of migrants towards households with less migration experience at baseline, and who were less likely to be planning on migrating at the time of the intervention. In contrast, migrants from the Group arm look similar to control-group migrants.

We were surprised to find higher returns to helping the inexperienced migrate: more experienced migrants should have better options once they arrive in Nairobi. But, as we show formally in our work, the sort of person who benefits more from an intervention is not necessarily the same sort who benefits more in the status quo. Migrating is harder for those who haven’t done it before: they tend to be poorer, less well-educated, less acquainted with urban life, and know fewer people who could help them in Nairobi. Thus, many inexperienced households will not migrate even if they face high potential gains. In contrast, it’s relatively easy for experienced households to migrate again, meaning that those with high potential gains will migrate with or without our intervention.

Policy implications for urban migration

Three points follow for policy:

1. Providing information can be a valuable labour market intervention. Simply offering convincing statistics changes beliefs about cities and increases migration and income. The cost of an intervention like this is extremely low compared to many public programmes with similar aims.
2. In lower-income countries, major cities nearly always have large populations of rural-urban migrants. Those migrants can act as a public good: they hold credible, current, hyper-local knowledge about jobs, housing, safety, and lifestyle. Many are happy to be connected to rural youth and share their experience and knowledge.
3. Individual visits can reach people that traditional programmes miss. In our research, the highest gains were observed among those who initially didn’t want to migrate. These people are less likely to seek out migration information on their own, but our findings suggest that they do benefit from door-to-door visits.

References

Akram, A A, S Chowdhury, and A M Mobarak (2017), “Effects of emigration on rural labor markets,” Unpublished manuscript.

Banerjee, A, E Breza, A G Chandrasekhar, and B Golub (2024), “When less is more: Experimental evidence on information delivery during India’s demonetisation,” Review of Economic Studies, 91: 1884–1922.

Barnett-Howell, Z, T Baseler, T Ginn, and S Gordeev (2025), “Reaching the novice or nudging the expert? Networks, information, and the experimental returns to migration,” Unpublished manuscript.

Baseler, T (2023), “Hidden income and the perceived returns to migration,” American Economic Journal: Applied Economics, 15: 321–352.

Bazzi, S, L Cameron, S G Schaner, and F Witoelar (2021), “Information, intermediaries, and international migration,” Unpublished manuscript.

Bryan, G, S Chowdhury, and A M Mobarak (2014), “Underinvestment in a profitable technology: The case of seasonal migration in Bangladesh,” Econometrica, 82: 1671–1748.

Chandrasekhar, A G, E Duflo, M Kremer, J F Pugliese, J Robinson, and F Schilbach (2022), “Blue spoons: Sparking communication about appropriate technology use,” Unpublished manuscript.

Gollin, D, M Kirchberger, and D Lagakos (2021), “Do urban wage premia reflect lower amenities? Evidence from Africa,” Journal of Urban Economics, 121: 103301.

Egger, D, J Haushofer, E Miguel, P Niehaus, and M Walker (2022), “General equilibrium effects of cash transfers: Experimental evidence from Kenya,” Econometrica, 90: 2603–2643.

Lagakos, D (2020), “Urban–rural gaps in the developing world: Does internal migration offer opportunities?” Journal of Economic Perspectives, 34: 174–192.