Surprise Marriages

Marriage can be a mysterious thing. What drives people to get together? There have been several visualizations examining marriage dynamics through the lens of occupations. I want to explore this topic from a different perspective. Instead of simply visualizing who a plumber or a CEO is most likely to marry, I want to see if there are any factors driving these kinds of pairings.

Let's begin by looking at the most common matchups between occupations.

Most Common Marriages

Below are the jobs husbands and wives hold for the top 25 most frequently observed pairings. Rather than comparing specific job titles between spouses, let's instead examine the popularity of those jobs. Circles farther to the left indicate that the husband or wife works in a more popular job among workers of their gender. (Since men and women tend to work in different occupations, it's important to look at job frequency between the two sexes separately.)

As we can see, almost all of these couples have a spouse employed in a highly popular occupation. Therefore, seeing so many marriages between these workers would be expected as there are simply lots of people who have these jobs.

Just looking at marriages based on their absolute frequency doesn't seem to shed much light on why people tend to marry one another. The most commonly observed matches are due to the fact that some jobs employ a much higher share of the population than others. Thus we should try to identify marriages that are particularly surprising rather than just very common.

Most Surprising Marriages

One way of thinking about which marriages are surprising is by asking which pairings occur more often than we would expect. For this, I am going to apply the concept of Bayesian surprise. In a perfectly random world, the likelihood that a man marries a woman who works as a secretary, for example, would be equal to the percentage of all working women who are employed as secretaries. If 5% of women are secretaries then we would expect 5% of marriages to be to a female secretary. Is this in fact what we see?

The plots below compares the percent of spouses in each occupational pairing who hold a particular job with the share of the overall population of workers employed in that profession. Pairings that fall on the dotted line are ones where the share of spouses is equal to the share of all workers with that occupation.

It seems the frequency of most matchups is roughly similar to what we would have expected. But there are some pairings that occur at a much higher rate - we might consider these to be "surprising" marriages. Let's take a deeper look at these couples.

This row represents husbands who are chief executives or legislators.

Each row represents a different job for the husband.

This column represents wives who are chief executives or legislators.

Each column represents a different job for the wife.

Since marriages where both the husband and wife are chief executives or legislators happens more often than we would expect, the square for that occupational combination is filled in. The color means both spouses work in the field of management.

Each filled-in square indicates that marriages between these occupations occur more frequently than we would expect. Spouses who work in the same field are filled with a color representing that field. Grey squares mean the spouses work in different fields.

As we can see, many of these matches are between spouses who have the same job or work in the same field. This seems to suggest that people choose to marry either those who have similar interests to themselves or who they encounter frequently at the office.

Another interesting pattern that jumps out is that of the remaining surprise marriages not between spouses in the same field, there are a few professions that seem to be chosen as spouses particularly frequently. For women, elementary and middle school teachers tend to marry at a higher rate than their share of the population would suggest while for men, it is miscellaneous managers who are particularly likely to get married.

This leads to one last interesting observation about the dynamics of marriage - which occupations are more likely to marry and which are more likely to stay single?

Occupations Most and Least Likely to Marry

It appears that not all workers are equally likely to get married. Certain jobs such as teachers and secretaries for women or software developers and mechanics for men seem to make for particularly appealing spouses while other professions such as cashiers, home health aides, and maintenance workers account for a smaller share of spouses than their numbers would suggest.

Below is a look at some of the most popular jobs for women and men that get married at noticeably higher or lower rates.

Female Workers Most and Least Likely to Marry

Share of female workers One occupational pairing

Male Workers Most and Least Likely to Marry

Share of male workers One occupational pairing

Roughly speaking, it appears that workers in more white-collar professions tend to get married at higher rates. This would support studies showing how marriage has become increasingly associated with socioeconomic status.

Marriage certainly can't be reduced to numbers and no statistical analysis can explain what causes people to choose one another but looking at data on the occupations of married couples can help uncover some interesting patterns. At the end of the day, marriage is a very complex decision that is influenced by many factors.

Appendix

Find out who people are most likely to marry by occupation:

is a

Most likely to marry:

Note: Pairings with fewer than 10 observations were excluded.

Data Sources: 2014 American Community Survey, Bureau of Labor Statistics Employment Projections 2012-13, Current Population Survey 2015, Occupational Employment Statistics Survey 2015.

© 2017. Designed by fleecealeece.