Intrahousehold Gender Inequality, Downside Risk and Technology Adoption: The effect of increased labor productivity on income sharing in Mozambique

How to cite

Jones, Rachel. 2024. "Intrahousehold Gender Inequality, Downside Risk and Technology Adoption: The effect of increased labor productivity on income sharing in Mozambique." AEA RCT Registry. AEARCTR-0014775. https://doi.org/10.1257/rct.14775-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Rachel Jones*

* Primary investigator

Registry ID
AEARCTR-0014775
DOI
https://doi.org/10.1257/rct.14775-1.0
First registered
Last updated
Status
In development
Trial dates
2024-11-11 – 2025-10-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/14775

Abstract

The objective of the experiment is to test whether the intrahousehold distribution of cash crop income has the potential to constrain agricultural technology adoption among Mozambiquan maize farmers. In cross-sectional survey data of rural Mozambican households, I document a rule of thumb whereby husband-to-wife income transfers as a function of variable maize yields are flat in good years, and decrease in bad years. In a non-unitary, separate spheres household in which labor is non-contractible, income transfers indirectly incentivize the wife's effort in cash crop production. I show analytically that a productivity-increasing technology can exacerbate the misalignment of incentives between husbands and wives if the institution of the intrahousehold transfer fails to evolve.

In the first round of an incentivized game played in couples, wives choose to allocate their labor across their husbands' cash crop and a risk-free market activity; husbands draw an agricultural yield from a probability distribution that is conditional on the wife's labor decision; and husbands choose how to distribute their crop income. The first round of the game permits elicitation of an equilibrium labor allocation and transfer function under a low-return technology that is calibrated to resemble local varieties of maize.

In a second round, we introduce a high productivity maize variety and allow couples to choose labor and transfers. This round permits me to study the extent to which husbands adjust their transfers to wives to induce an increase in wives' maize labor.

Intervention

In the game setting, couples will determine how to share maize income. As a function of the division of income, the wife will decide how to allocate her labor between maize and an alternative activity for which she earns a market wage. The wife’s labor decision will affect the distribution of potential maize yields and game payouts. In the second part of the game, participants will experience an exogenous increase in the returns to agricultural labor. This shock to agricultural productivity enables us to study whether husbands adjust income sharing rules in order to increase their spouse’s agricultural labor in the game setting.

Experimental design

The game will target a subset of households that participated in previous rounds of a household survey, and recruit new households who have not participated in the household survey. We intend to play the game with 560 households. We will recruit 280 of these couples from households that participated in previous rounds of the household survey. We will refer to this subset of participants as in-sample couples. The other 280 participating couples will be recruited from households who have not participated in in the household survey.

The 560 couples (28 couples in each of 20 communities) who receive the game treatment will be exposed to two game components, each of which constitutes a different treatment. A “within” comparison of the couples’ behavior across game components will yield a sample size of 1,120 observations, with 560 couples per treatment condition.

In addition, we can study the effect of exposure to the game on out-of-game behavior by comparing the 280 households who were randomized to the game treatment and will participate in future rounds of the household survey to the 280 households that were randomized to the control group of the game treatment.

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
The unit of randomization for the within comparison of participant behavior across game treatment rounds is the individual. The unit of randomization for the study of the effects of game play on out-of-game behavior is the community.
Clusters
Within comparison of participant behavior across game treatment rounds: 1. Effects of game play on out-of-game behavior: 40
Observations
Within comparison of participant behavior across game treatment rounds: 1120. Effects of game play on out-of-game behavior: 560
Arms
Within comparison of participant behavior across game treatment rounds: 560. Effects of game play on out-of-game behavior: 230
Primary outcomes
Number of days wife allocates to maize production in the game; Meticais (Mozambican currency) of maize income that husbands commit to share with their wife given each potential maize yield
Secondary outcomes
Total hours of wife's labor in maize production (non-game outcome); total income transfers from husband to wife from December 2024 to June 2025

Cited by

No citation data available for this registration yet.

Keywords