Preferences over consumption of child support-eligible children

How to cite

Ong, Pinchuan. 2022. "Preferences over consumption of child support-eligible children." AEA RCT Registry. AEARCTR-0009988. https://doi.org/10.1257/rct.9988-1.7

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

Pinchuan Ong*

* Primary investigator

Registry ID
AEARCTR-0009988
DOI
https://doi.org/10.1257/rct.9988-1.7
First registered
Last updated
Status
In development
Trial dates
2023-09-01 – 2023-09-30
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/9988

Abstract

A non-negligible fraction of men in the US are non-custodial fathers, and pay child support to maintain their children who do not live in their households. This study investigates the preferences of these fathers over the consumption of the children that they pay support for. We field a vignette survey experiment to a sample of such fathers obtained through an online survey panel. The vignettes randomize the consumption amounts allocated to these children and their current family, to allow estimation of preferences over how much these children consume vis-a-vis their current families.

Intervention

Vignette experiment showing child support-paying fathers scenarios that require spending on item categories.

Experimental design

Fathers who have to pay child support are shown ten scenarios. In each scenario, a father is told that the child support-eligible child needs to spend money on a particular item category (e.g. medicine), and that someone in his current family does as well. He is presented with two alternatives: to spend a certain amount on the child and to spend a certain amount on the family member. The total amount spent is the same across the two alternatives, but the proportion between child and family member is randomized. In all scenarios, the father is told that the child's mother will handle the money, analogous to how child support payments officially go to the mother before the child is of age and not the child. For fathers with children in their current family (e.g. if they remarried), we additionally field two branches. In the first branch, we specify "a family member", while in the second branch, we specify this current child.

Design details

Randomization method
Randomization done by computer
Randomization unit
Randomization is at the alternative level (within-scenario, and within-individual). Clustering is at the individual level. For fathers who have children in their current family, we randomize at the individual level to show slightly different sets of questions.
Clusters
100
Observations
100 clusters * 10 scenarios * 2 alternatives = 2000 observations
Arms
The proportion of spending (a continuous variable) going to child versus family is drawn from a normal distribution in this experimental design. Taking the alternative with a lower proportion going to the child support-eligible child as being the control arm, we have 1000 observations in control and 1000 in treatment.
Primary outcomes
Choice of which alternative the father chooses (total of 10 choices)
Secondary outcomes
Whether the primary outcome is different when we ask the father to compare with a child in his current family

Cited by

No citation data available for this registration yet.

Keywords