Assortative Mating and Joint Income Dynamics of Couples

Project overview

This project uses rich tax register data from Denmark to explore empirical regularities in matching patterns of couples and to take high-resolution pictures of the labor income dynamics of males, females and the joint income dynamics of couples.

The goal of this project is to answer questions such as: 

• To what extent are spouses exposed to similar labor market risks? 
• What are the sources of this similarity (e.g., spouses work in similar careers, in the same local labor markets, and so on)? 
• In light of this correlation, how well can they insure against individual risks?

In order to understand the relative importance of several forces generating the observed empirical patterns of household formation and the link to income dynamics, the second part of the project plans to set up a quantitative structural model. This model can be used as a laboratory in which to explore how individuals react to changes in the degree of correlated risks, how the success of within-household insurance against individual risk interacts with the degree of sorting along various dimensions, and finally how changes in the tax and transfer system affect joint income dynamics.

Summary, output, and dissemination

Research summary

This project shows that changes to a person’s individual earnings vary across households. Negative changes to earnings are strongly correlated when a couple shares labor market characteristics. This means that in households where both spouses work in similar jobs, an income loss of the main earner translates into stronger consumption reductions than in households where spouses work in different jobs.

From the perspective of one household, the simple intuition is that individual risk is correlated stronger for households that share job properties (like sector or occupation). So, when households do not work in the same type of job, the probability of both partners experiencing a drop in income at the same time is lower and therefore when the main earner’s income drops. Resources of the household are not as badly affected as for households where both spouses face greater correlated risk.

The main results from this project are summarized as follows: 

1. Couples who have the same education, or work in the same occupation or industry, or are at the same firm, display a much stronger correlation of individual-level annual earnings changes. 

2. Differences across groups (“sorted vs. not sorted”) are accounted for by negative changes of main the earner, i.e., income losses of the main earner is more strongly correlated with income changes of their spouses when the spouse has similar labor market characteristics.

3. The differences across groups translate into how individual income changes translate into household-level consumption, for example, the households’ capability of consumption smoothing is affected by the degree of similarity of spouses in labor market characteristics.

4. It was documented that there is sorting by, i., education, ii., occupation, iii., industry, iv., firm of employment, where sorting means that there are more couples that share the same (education, occupation,…) than what would be observed under random matching of individuals given the marginal distributions of males and females. This amplifies the role of similarity for aggregate measures of insurance.

5. Divorce risk introduced additional income risk at the individual level.

Conferences, workshops and seminars

Conferences

• Presentation at Midwest Economics Association Virtual Meetings, March 23, 2021, virtual; Christopher Busch. Target group: Researchers
• Presentation at HHEI Alumni Conference at University of Minnesota, June 25, 2021; Rocío Madera (co-author). Target group: PhD students and researchers 
• Presentation at the Annual Meeting of the Society for Economic Dynamics (SED), July 1, 2021, virtual. Christopher Busch. Target group: Researchers 
• Presentation at 12th ifo conference on macroeconomics and survey data, ifo institute Munich, February 4, 2022; CB, target group: Researchers 
• Presentation at Midwest Macro at SMU Dallas, November 12, 2022; Rocío Madera. Target group: Researchers
• Presentation at SAEe, December 16, 2022; Rocío Madera. Target group: Researchers

Seminars 

• Presentation in a research seminar, Arizona State University, October 1, 2021; Rocío Madera. Target group: Researchers 
• Presentation in a research seminar, University of Copenhagen, November 26, 2021; Rocío Madera. Target group: Researchers 
• Presentation in a research seminar, Goethe University Frankfurt, December 2, 2021, virtual; Christopher Busch. Target group: Researchers 
• Presentation in a research seminar, New York FED, December 13, 2021, virtual; Christopher Busch .Target group: Researchers 
• Presentation in a research seminar, University of Southern California, February 17, 2022; Rocío Madera. Target group: Researchers  
• Presentation in a research seminar, University of Würzburg, May 31, 2022; Rocío Madera. Target group: 5 Research Grants on Socioeconomic Wellbeing researchers 
• Presentation in a research seminar, University of Munich, June 1, 2022; Rocío Madera. Target group: Researchers 

Workshops

• Presentation at DTMC workshop on interactions between labor and marriage markets, University of Aarhus, April 6, 2022; Christopher Busch. Target group: Researchers