Article Summary
Introduction
Political polarization in the U.S. has been described as extending beyond disagreement over issues, and into dislike and avoidance of politically dissimilar others. This paper focuses on romantic relationships, which previous work have shown to be highly important for overall life satisfaction, and highly sorted by political orientation. Huber and Malhotra specifically take on the question of whether people preferentially choose to date politically similar others, or whether people often end up in romantic relationships with a politically similar partner through another mechanism (people become more similar to their partner over time, are geographically constrained to mostly others with similar politics, or choose on other factors that are correlated with political ideology). Further, they ask whether people choose partners who are similar on partisan identity, political ideology, or level of engagement with politics. Knowing whether people choose partners based on politics or simply are more likely to end up with a partner with similar politics provides important causal clarity, and speaks to the extent of political polarization within the country.
Analytical Approach
The first study showed each participant 10 experimentally manipulated dating profiles. To increase external validity the pictures, texts, and usernames were taken from public dating profiles. Experimenters then manipulated the profiles’ politics, as either “conservative,” “moderate,” “liberal,” “not interested in politics,” or blank. They also accounted for profile age, education, race, and religion in their analyses.
The second study analyzed real-world behavior on a dating site, which offered higher external validity but less experimental control. On this site, users can provide information about their politics, which is not displayed on their profile, but which can be used as a filtering criteria in a search. These questions assessed political identity, issue positions, and level of engagement. They then measured the rates at which users matched on these political questions when a) someone initiated contact, b) someone responded to the initial message, and c) a message and a reply together, all compared to the prevalence of political matches in a given zip code.
Main Findings
In the experimental study, shared politics (or shared lack of political interest) increased participants’ interest in initiating contact by 6-8%. Shared ideology increased interest in long-term dating by 11%. This provides causal evidence that people are more likely to choose to date others with similar politics. In the analysis of dating site behavior, they find that party identification, common ideology, and level of engagement are all associated with increased likelihood of contact. These associations persist when controlling for the desirability of the partner and concordance on non-political attributes.
Implications
This paper provides a strong argument for what the authors call “political homophily” – that people actively choose partners with similar political attitudes. This suggests that married couples are more likely to share political beliefs due to active choice, rather than through secondary mechanisms like convergence of beliefs after marriage. Here we see evidence that people prioritize potential partners’ politics at the stage of partner choice, and that politics affects who people see as long-term mates. Sorting by politics in partner choice has the possibility to exacerbate political polarization in the county, both by ideology and between highly-engaged and unengaged citizens.
Questions left unanswered
The largest open question from this work is how partner choice in online settings compares to in-person dating, where political attitudes may not be as neatly categorized or readily displayed. Additionally, it is not clear whether the analyses used definitely rule out sorting on “non-political” factors, as users may be sorting on “non-political” factors which are not directly assessed by the dating site, such as beliefs about gender roles.
Methods and Analysis
Was the study and its analyses pre-registered?: No
Did the study rely on proxy variables to measure polarization?: No
Were standard p-value thresholds used (p<.05 or 95% Confidence Intervals that don’t overlap zero)?: No
- Largest p-value presented as significant: 0.1
Were correlational results interpreted with causal language?: No
Limitations / Weaknesses
This study did an admirable job assembling causal evidence on an important social phenomenon that is quite difficult to disentangle from other explanations. However, it would be worth addressing more deeply what people interpret from the party label or political ideology of their partner. Do people choose similar partners because they want to be with someone with the same political beliefs, or because political ideology is a useful heuristic for understanding the moral worldview of someone else? It is not clear that the analyses presented rule out this type of “non-political” preference, though the authors argue that they do.
Open Data & Analyses
Does the article make the replication data publicly available?: No
Does the article make the replication analysis scripts publicly available?: No
Article Citation
Huber, G. A., & Malhotra, N. (2017). Political homophily in social relationships: Evidence from online dating behavior. The Journal of Politics, 79(1), 269-283.
Bibtex
1
2
3
4
5
6
7
8
9
10
@article{huber2017political,
title={Political homophily in social relationships: Evidence from online dating behavior},
author={Huber, Gregory A and Malhotra, Neil},
journal={The Journal of Politics},
volume={79},
number={1},
pages={269--283},
year={2017},
publisher={University of Chicago Press Chicago, IL}
}