He Said/She Said: Testing Respondent Effects in Household Income Reporting

How to cite

Masselus, Lise; Fiala, Nathan. 2019. "He Said/She Said: Testing Respondent Effects in Household Income Reporting." AEA RCT Registry. AEARCTR-0005162. https://doi.org/10.1257/rct.5162-1.6

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

Lise Masselus* , Nathan Fiala

* Primary investigator

Registry ID
AEARCTR-0005162
DOI
https://doi.org/10.1257/rct.5162-1.6
First registered
Last updated
Status
Completed
Trial dates
2020-01-06 – 2022-06-30
Countries
Paraguay (), Uganda ()
Registry record
http://www.socialscienceregistry.org/trials/5162

Abstract

Obtaining reliable data on household and respondent income and understanding the quality of this data is of great importance to researchers and policy makers. Several papers have found evidence that household income is often misreported. However, the causes and consequences of intra-household reporting discrepancies remain poorly understood. Commonly, the most knowledgeable household member, often the household head, is interviewed in household surveys. Surveying only one household member could potentially lead to incomplete or erroneous information on the households’ aggregate income for several reasons. Household members could be oblivious of all income-generating activities of their spouses, for example because spouses intentionally hide income from one another. Incomplete pooling of information within the household could also result in difficulties to accurately report on other’s income, even if a household member is aware of all sources of income. These concerns can be overrun by assessing multiple household members separately. However, this comes at a logistical and monetary cost. This survey experiment aims to shed light on the discrepancies in intra-household reporting on aggregate income.

Experimental design

This study consists of a survey experiment where we randomize who in the household responds to income questions. The experiment is designed as a stratified clustered-RCT where villages are allocated to one of three treatment arms. Randomization will be stratified at the subcounty level.

Design details

Randomization method
Stratified clustered randomization done by a computer
Randomization unit
The unit of randomization is the village.
Clusters
200 clusters in Uganda 165 clusters in Paraguay
Observations
3000 households in Uganda 2000 households in Paraguay
Arms
Uganda: 100 clusters T1, 50 clusters T2 and 50 clusters T3 Paraguay: 110 clusters T1, 55 clusters T2
Primary outcomes
The primary outcome interest is aggregate household income
Secondary outcomes
As secondary outcomes, we will also look at income per household member, and income per economic activity as outcome variables.

Papers from this trial

  1. Fiala, Nathan and Masselus, Lise, (2022), Whom to ask? Testing respondent effects in household surveys, No 935, Ruhr Economic Papers, RWI - Leibniz-Institut für Wirtschaftsforschung, Ruhr-University Bochum, TU Dortmund University, University of Duisburg-Essen. https://www.rwi-essen.de/publikationen/wissenschaftlich/ruhr-economic-papers/detail/whom-to-ask-testing-respondent-effects-in-5169

Cited by

No citation data available for this registration yet.

Keywords