Perceived and Elicited Valuation of Partner Attributes in the Marriage Market

How to cite

Inoue, Chihiro; Ishihata, Yusuke; Mugiyama, Ryota; Sugimoto, Haruki; Uchikoshi, Fumiya. 2026. "Perceived and Elicited Valuation of Partner Attributes in the Marriage Market." AEA RCT Registry. AEARCTR-0018695. https://doi.org/10.1257/rct.18695-1.0

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

Chihiro Inoue* , Yusuke Ishihata , Ryota Mugiyama , Haruki Sugimoto , Fumiya Uchikoshi

* Primary investigator

Registry ID
AEARCTR-0018695
DOI
https://doi.org/10.1257/rct.18695-1.0
First registered
Last updated
Status
In development
Trial dates
2026-10-01 – 2026-10-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/18695

Abstract

This study examines how men and women evaluate potential marriage partners and whether they hold systematically biased beliefs about how they themselves are evaluated in the marriage market. In particular, the study examines the marriage-market value of two attributes that are central to household formation but are less aligned with conventional gendered partner-role expectations: men’s contributions to housework and childcare, and women’s earnings.

Using an online survey experiment, we elicit respondents’ evaluations of hypothetical potential opposite-sex partners with experimentally varied characteristics. In addition, we measure respondents’ beliefs about how members of the opposite sex evaluate hypothetical partners of the same sex as the respondents, and collect respondents’ own characteristics. This design allows us to construct predicted measures of respondents’ value as potential partners, based on their observed characteristics and estimated preference parameters.

The primary outcomes are respondents’ stated evaluations of hypothetical opposite-sex partners and their beliefs about how hypothetical opposite-sex individuals evaluate hypothetical own-sex partners. The study tests whether men underestimate the value that potential partners place on men’s housework and childcare time and whether women underestimate the value that potential partners place on women’s income. We can also compare the respondents' perceived own value as potential partners relative to evaluations by the opposite sex, based on their observed characteristics and estimated preference parameters. The results will provide evidence on gendered misperceptions in the marriage market and their potential role in shaping partner search, matching, and household formation.

Experimental design

We conduct an online survey experiment with unmarried adult respondents aged 25–29 in Japan. Respondents evaluate experimentally varied hypothetical opposite-sex profiles and report their interest in considering each person as a potential marriage partner and willingness to meet the partner. Respondents also predict how hypothetical opposite-sex individuals would evaluate hypothetical own-sex profiles. Thus, male respondents evaluate female profiles and predict how each female profile would evaluate male profiles, while female respondents evaluate male profiles and predict how each male profile would evaluate female profiles.

Profiles vary randomly along characteristics relevant for partner choice, including income, housework and childcare time, and other demographic and socioeconomic attributes. The design allows us to estimate the marriage-market value of men’s housework and childcare time and women’s income, and to compare these evaluations with respondents’ beliefs about opposite-sex evaluations. These comparisons are used to study whether men and women systematically misperceive the value of these attributes, as well as their predicted value as potential partners.

Design details

Randomization method
Randomization is implemented by a computer as part of the online survey platform. Partner profile attributes are randomly assigned by the survey software according to a pre-specified randomization procedure.
Randomization unit
Randomization occurs at the respondent-profile level.
Clusters
1000 unmarried men and 1000 unmarried women aged 25-29
Observations
Each male respondent will evaluate 10 female profiles, so the number of male-to-female evaluation observations will be 10 times the number of male respondents. In addition, each male respondent will predict evaluations by a hypothetical woman of 10 male profiles. The analogous design applies to female respondents.
Arms
The experiment uses randomized profile attributes. Each respondent evaluates multiple profiles, and the randomized attributes vary across profiles, and there is no pure control group. Across the full sample, profile observations are designed to be approximately evenly distributed across the pre-specified combinations or levels of randomized attributes.
Primary outcomes
Respondents’ own evaluations of opposite-sex profiles Respondents’ beliefs about opposite-sex evaluations of own-sex profiles Predicted partner-value measures under estimated opposite-sex evaluations and belief-based evaluations

Cited by

No citation data available for this registration yet.

Keywords