Benchmarking Cash to an Employment Program in Rwanda

Sample:
1,848 underemployed youth
Chronologie:
2017 - 2021
Target group:
  • Job seekers
  • Women and girls
  • Youth
  • Self-employed and microentrepreneurs
Outcome of interest:
  • Earnings and income
  • Employment
  • Asset ownership
  • Consumption smoothing
  • Mental health
  • Savings/deposits
Intervention type:
  • Apprenticeships and on-the-job training
  • Business skills training
  • Cash transfers
  • Training
  • Soft skills
  • Hard skills
  • Unconditional cash transfers
AEA RCT registration number:
AEARCTR-0004388
Partners:

Sub-Saharan Africa has a young population, but youth unemployment rates in the region are high. In Rwanda, researchers conducted a randomized evaluation to compare the impact of Huguka Dukore, a youth employment and training program, to that of cash grants of an equivalent value on several economic outcomes. Huguka Dukore improved hours worked, assets, savings, and subjective well-being, while cost-equivalent cash transfers increased all these outcomes as well as consumption, income, and wealth. In the head-to-head costing comparison, cash proved more effective across a number of economic outcomes, while training outperformed cash only in the production of business knowledge. 

Policy issue

Sub-Saharan Africa has a population of over one billion people, half of whom will be under the age of 25 by 2050.1 A young population provides an opportunity to benefit from many productive years while bearing a limited burden of dependency from older generations; however, this may not be possible if young people are unable to find productive employment.  

Previous research on active labor market interventions, such as vocational training and wage subsidies, has shown some positive effects on employment and income, but the impacts tend to be small and variable across studies. On the other hand, research has found that recipients of unconditional cash transfers invest in durables, productive assets, and microenterprises, suggesting that cash may be a reasonable alternative in delivering economic livelihood assistance to youth. How does the impact of a youth employment and training program compare to the impact of providing cash grants of equal cost?

Context of the evaluation

In spite of gains in formal educational attainment, youth unemployment rates in sub-Saharan Africa remain high. In Rwanda, where this study took place, 40 percent of the population is between the ages of 14 and 30, but 65 percent of these youth are unemployed. 

Aiming to support 40,000 youth in Rwanda, the Huguka Dukore program is a five-year project (2017-2021) providing employment preparation, individual and cooperative youth microenterprise start-up support, and business development for existing microenterprises. The program consists of a number of separate modules that are taken serially over the course of a year. The first module, taken by all students, focuses on eight topics: personal development, interpersonal communication, work habits and conduct, leadership, health and safety at work, worker and employer rights and responsibilities, financial fitness, and exploring entrepreneurship. From there, students choose the additional modules and the sector of work in which they receive additional training. 

To be eligible for Huguka Dukore, participants must be between the ages of 16-30 with six to twelve years of education and come from households in the two lowest-income levels. In addition, Huguka Dukore specifically recruits women, youth with disabilities, and those who have been out of school for three consecutive years. In Rwamagana, Muhanga, and Nyamagabe districts, study participants were low-income, underemployed youth who expressed willingness to participate in a training program prior to the start of the intervention. Just under 60 percent of study participants were female, with an average age of 24 and 7.6 years of education. Thirty-two percent of participants resided in households categorized by the government of Rwanda as “extremely” poor, and participants on average had a yearly income of about US$190. Thirty-three percent were employed before the start of the intervention. 

young Rwandan woman carrying a sack on her head
Young woman in Rwanda.
Photo: Rostasedlacek | Shutterstock.com

Details of the intervention

Researchers conducted a randomized evaluation to benchmark the impact of Huguka Dukore to providing cash grants of an equivalent cost on employment status, time use, income, household consumption, and productive assets. To compare the cost-effectiveness of Huguka Dukore to cash transfers, as well as to better understand the potential complementarities between the different types of programs, the researchers undertook a cash benchmarking exercise, or the direct comparison of in-kind to cash transfer programs in a single setting. To do so, they worked with GiveDirectly, a US-based nonprofit organization that specializes in making unconditional household grants via mobile money.2  

Of the 2,275 individuals who attended an orientation meeting and signed up for Huguka Dukore, 1,848 youth enrolled in the study. Thirteen public lotteries were used to randomly assign the youth into four groups:

  1. Huguka Dukore: The 485 participants in this group were offered the Huguka Dukore program. 
  2. Cash grant: The 672 participants in this group were offered payments in two installments two months apart, with the first payment comprising 40 percent of the total to be paid. In each case, the amount transferred to beneficiaries subtracted GiveDirectly’s standard delivery cost from the total costs of the program. Cash grant recipients were randomly assigned to four different transfer amounts: 
    • One group received a transfer amount equivalent to the estimated total cost of providing Huguka Dukore to each beneficiary, or US$411.
    • Because researchers did not know the true per-person cost with certainty before the intervention, the second group received a transfer of approximately US$90 less than the first group (US$317), and the third group received a transfer of approximately US$90 more than the first group (US$503). 
    • The fourth group received transfers of US$750.   
  3. Cash grant and Huguka Dukore: The 203 participants in this group were offered both the Huguka Dukore program and cash grants of US$411.  
  4. Comparison: The 488 participants in this group were not offered any program.

The Huguka Dukore program was implemented from January to November 2018, and the cash transfers were delivered between May and July 2018. Researchers conducted surveys prior to the start of the program from December 2017 to February 2018 and follow-up surveys from July to August 2019 on employment, work hours, assets, income, consumption, wealth, and well-being—which was measured based on questions about happiness and life satisfaction. Researchers also planned to conduct additional follow-up surveys 36 months after the initial surveys to assess the longer-term impacts of the programs.

Results and policy lessons

Huguka Dukore was successful in improving productive hours worked, assets, savings, and subjective well-being. Cost-equivalent cash transfers improved all of these outcomes as well as consumption, income, and wealth. In the head-to-head costing comparison, cash proved superior across a number of economic outcomes, while training outperformed cash only in the production of business knowledge. All results reported were measured fifteen months after Huguka Dukore was offered and fourteen months after the cash transfers took place.

Employment and work hours: None of the programs had an impact on the employment rate. However, young people that received cash transfers were more likely to move from wage labor into self-employment, while Huguka Dukore beneficiaries became engaged in more off-farm wage labor. Huguka Dukore increased the number of productive hours worked in a week by approximately 3 hours, a 15 percent increase relative to the hours worked by the comparison group (18.6 hours). The US$411 cash transfer led to a 6.5 hour, or 35 percent, increase in the number of productive hours worked in a week, but no other transfer size or the combination of Huguka Dukore and cash transfers affected the number of hours worked.  

Productive assets, income, and consumption: Huguka Dukore had no impact on monthly income or on household consumption. Cash transfers, on the other hand, led to an approximate doubling of monthly income, with intermediate-sized transfers appearing to be the most effective, and an increase in household consumption of 20–36 percent. Even though Huguka Dukore involved no material transfers to participants, productive assets (e.g., tools, machinery, furniture) increased to be 154 percent higher than in the comparison group. Productive asset values quadrupled in the cash transfer group. The combination of Huguka Dukore and cash transfers had similar impacts on productive assets, income, and consumption to the intermediate-sized cash transfers, suggesting a lack of complementarity between cash and training. 

Well-being: Huguka Dukore, cash transfers, and the combination led to an increase in participants’ subjective well-being. The effects for participants in any group receiving cash were more than twice as large as those for participants in the Huguka Dukore group. 

Wealth, savings, and debt: Savings at least doubled across all program groups. Huguka Dukore had no impact on household non-land or livestock wealth. Among participants in any group that received cash transfers, household livestock wealth tripled for all cash transfer sizes, and net non-land wealth more than doubled for all transfer sizes except the US$317 transfer. Neither Huguka Dukore, cash transfers, nor the combination had any impact on household debt. 

Costing comparison: In the head-to-head costing comparison, youth benefited more from cash transfers across a range of indicators central to economic welfare. The only outcome in which Huguka Dukore outperformed cash was increasing business knowledge. 

Taken together, these results demonstrate that cash transfers had a larger impact on the indicators central to recipients’ economic welfare. Each intervention had a distinct set of benefits, and little was gained by providing them together. This suggests that there were no complementarities between the cash-transfer impacts and the skills provided by Huguka Dukore. Additionally, because the US$411 transfers appeared to be the most effective at improving outcomes, transfers larger than approximately US$400 had limited additional value in this context.