Crossing the Finish Line: Intervening in a Critical Period for Zambian Women's Educational Investment

The intervention is cash assistance for university students.

The intervention is an unconditional cash transfer. Participants will be randomly selected selected to received 1000 Zambian Kwacha each month (approx 40 USD) for 4 months.

2024-09-16

2024-12-16

Risky sexual behaviors/transactional sex; educational outcomes

Risky sexual behaviors: engaging in condomless sex (measured every two weeks), reporting not using a condom because a partner did not want to (measured every two weeks), age gap with sexual partners (measured every two weeks), sex with partners with whom women do not feel safe.
Transactional sex: Large gifts (e.g., paying tuition or housing) by sexual partners, amount of transfers from sexual partners
Educational outcomes: self reported enrollment information, grade point average (if possible), administrative data on enrollment/dropout, grades, and on-time completion (if possible)
For dimensions of sexual behavior and education, average effect sizes can be used to create aggregate measures across all specific outcomes.

Mental health, cognitive load, amount of debt, time use

Mental health: Questions from the PHQ-4
Cognitive load: Assessed using Raven's matrices
Time Use: time spent studying/attending classes, in income generating activities, etc.

Approximately 50% of participants will be randomly assigned to receive cash assistance after 1 to 2 baseline surveys.

Participants will enroll in the study for a period of approximately 8 months and take surveys every two weeks for this period. They will complete 2 baseline surveys prior to being told whether or not they have been randomly selected for cash assistance. Selected students will be called and told about their treatment status and the timing of the treatment. Students will receive 1,000 Zambian Kwacha monthly for 4 months. After this, surveys will continue for another 4 months. This design will allow the PIs to (a) include baseline controls from the first 2 surveys in their analysis and stratify the random sampling, (b) observe effects during the cash transfers (4 months), and (c) separately observe if effects persist after the cash transfers (remaining months). The PI team will also seek to obtain administrative educational data from the National University of Zambia to supplement the measurement of educational outcomes in the self-reported survey data.

For our analysis, we plan to control for baseline values of the outcome variables (when available) and the stratification groups. To improve power and balance, additional controls can also be selected using the double-lasso methodology.

Due to our interest in transactional sex and financial disadvantage, we will also investigate heterogeneous treatment effects for those who are sexually active and on bursary (our stratification variables).

Stratified randomization will be done in office by a computer. Randomization will be stratified on sexual activity (has never had sex, sexually active but not in last two weeks, and sex in the last two weeks) and whether a student is on financial aid ("bursary").

Individual

No

N/A

1,000 students

500 students per treatment