To have and to hold (a job). The correlation of marriage and employment is puzzlingly strong.

You’d think getting married is relevant to your home life. You wouldn’t expect it to change your employment outcomes.

I mean, I’ve never done it, but I doubt you get back from your honeymoon buzzing with a desire to read and reply to all those emails.

And yet, the correlation between marriage and labour market outcomes is quite astounding. 

Unemployment

The unemployment rate for unmarried men is nearly four times higher than for married men (11.3% vs 3.1 per cent). For women, the ratio is over two (8.9% vs 4%).

The difference between married and unmarried makes the difference between men and women look small. 

Essentially, if you are a married man, you’re living in a labour market no different from the best parts of the 1970s, with 3 per cent unemployment!

Might this be a statistical artefact? It could come about because the young have poor employment outcomes, and are less to be married. Let’s have a look at an older age bracket.

Unemployment 35-44

The absolute levels of unemployment have fallen, especially for men. But the ratios of unemployment rates between married and unmarried are about the same: 4:1 for men and 2:1 for women.

The above graphs make it look like married people are all hard at work in the office. But the unemployment rate hides a big difference in participation rates.

There are two distinct clumps in this chart. Married men, who participate in the workforce at a rate of 95 per cent. And everyone else, who participate at around 75 per cent.

participation

The 80s were a time of rapid change for women. But since 1990, one of the biggest changes in the employment market has been unmarried men dropping out of the labour force. Their non-participation rate basically doubled from 10 per cent to 20 percent.

Given the unemployment graphs on the previous page, I’d be very surprised if the red line (married women) didn’t tick up over the green line in coming years.

Two mysteries remain.

1. Why is the difference between the married and unmarried so strong, and so consistent over time?

I have a few theories.

Perhaps the unemployed are busy proposing, but are rejected because they are unemployed?

Perhaps there are confounding variables, like good looks or intelligence, which are correlated with both earning power and marriageability.

Perhaps it’s not about the kind of people they are but the incentives they face:

Obviously marriage and children are correlated. Obviously children (who are cruelly forbidden by the law to earn the money to feed themselves) are expensive. Could it simply be the compulsion to put bread on the table that explains why married people are so rarely out of work?

2. Why is the labour force participation rate of unmarried men eroding?

Marriage is increasingly rare, and increasingly for the old.

Can that explain the fall in unmarried men’s attachment to the labour force?

Screen Shot 2014-08-15 at 10.54.35 am

Screen Shot 2014-08-15 at 11.02.39 am(They are also increasingly likely to have a non-religious ceremony, but I’m not sure that’s relevant)

celebrant

 

I’m not sure it does, and this makes me wonder if perhaps the “discouraged worker effect” might be true. All those unmarried men might once have worked in factories. Maybe they’re less able or inclined to take service sector jobs. 

There might also be an echo of higher immigration rates in the data. The overseas born have lower workforce participation rates. (chart source)

immi

 

Which looks like a nice simple story, until you fold it back in on itself and see that immigrants actually get married at a higher rate than their proportion in the population! (Number of marriages is on the vertical axis, so in total, this graph shows that at least 40 per cent of people getting married in Australia are overseas-born.)

marriage of immigrants