Making sense of divorce rates: It’s all about the fish!

It’s that time of year again when the Office for National Statistics produce their annual round-up of divorce statistics. And we’ve made our own contribution to the debate with a new report that you can read here.

Hidden away behind the overall numbers are the thousands of individual couples with their own very personal stories.

For some couples, divorce is a disaster. The dream of happily ever after has turned into an unwanted nightmare. For others, divorce rubber stamps the seemingly inevitable reality that their relationship had drifted far from its romantic beginnings. For more, it appears to offer the opportunity for a fresh start free from the pain of a bad experience. Even if some children do better as a result, the average experience for the children involved is invariably negative. But that’s a topic for another day.

Whatever the personal stories, everybody has a view on divorce. So the media love churning out stories about divorce, whether it’s on the way up or on the way down, and why.

I’m interested in divorce stats because they tell us about trends in married life, how long couples stay together. More accurately, they tell us long the couples who split up have stayed together.

Unfortunately the big problem with the way divorce stats are presented makes it hard, if not impossible, to interpret trends. The result can be strangely conflicting observations, such as “divorces up on recession” and “divorces down on recession”.

(Actually, they are both wrong. Boom and bust don’t seem to have an overall impact on divorce one way or another. You can read another of our reports on this here).

Here’s a visual analogy of why divorce rates are so hard to interpret.

Imagine a giant swimming pool full of fish. Some of the fish are young and have just been put into the pool. Others have been swimming around in the pool for years. At one end, in place of a diving board, is a big tube through which newcomers pour into the pool. At the bottom are two big plugs through which fish exit the pool.

Marriage and divorce statistics are a bit like this pool. Each fish represents one married couple. The total number of fish equates to the total population of married couples. Newlywed couples are the fish entering the pool. Divorcing couples are the fish exiting the pool via one plug. Mortality accounts for the other plug.

The way divorce rates for each year are reported equates to what happens to the pool over the course of a year. We are told how many fish enter the pool – ie number of marriages – and how many fish exit the pool – ie number of divorces – during that year. Because we know how many fish there are in the pool overall – ie married population – we can work out the “rate” at which fish leave the pool – ie divorce rate. We also know other things about the departing fish, such as how old they are – ie the age at which couples divorce.

This fish analogy immediately helps clear up once common misconception. It shows why “divorce rates” have relatively little to do with “marriage rates”. For starters, twice as many fish enter the pool as leave it. So even with marriage rates down, the pool is still being replenished. But more importantly, the “divorce rate” we want to know depends on the total number of fish already in the pool and not just how many enter it in one particular year.

So the “divorce rate” that gets reported in our newspapers is really the proportion of all of the fish in the pool who disappear through the plug. I call this the “year of divorce” method of reporting divorce rates. It gives us an idea of flow but makes it hard to interpret trends from year to year. This is because the population of fish in the pool changes from year to year. It still doesn’t stop social commentators from being very creative in their explanations!

So how to work out what’s really going on? We need to segregate all of the fish into lots of different pools. Instead of staring bemused at one giant pool full of fish comprising all different ages and stages, I have built a model where every individual pool comprises one year’s intake of fish.

Keeping records of how many of the original fish leave each pool each year tells me a lot more about the trends in divorce rates for different stages of marriage.

So for example, I have one pool comprising the 351,329 couples who married in 1963, the first year for which the data is good enough to do this. Because I’ve got data on how many of these couples leave the pool each year until 2012, the most recent year for which I have data, I can tell you all about the divorce rates for couples who married in 1963 over their first 49 years of marriage. So far 98,783 couples from this pool have divorced, making their overall divorce rate 28% so far. Very few more will do so.

My most recent pool, the 247,890 couples who married in 2011, have only been married one year. Not surprisingly, only 49 of them divorced during 2012. So I can tell you that their first year divorce rate is just 0.02%. I can also tell you that this divorce rate fluctuated up and down since 1963 between about 0.01% and 0.04%. This confirms what one might suspect happens in the first year of marriage. They have always been extremely rare.

This “year of marriage” method, tracking the progress of couples who marry in each particular year, gives us a clear indication of trend that we simply don’t get from the “year of divorce” method.

Since Sir Paul Coleridge founded Marriage Foundation less than two years ago, I’ve produced a stream of reports about divorce using this “year of marriage” method.

Aside from confirming things that I might have suspected, I’ve also learned some very surprising things that I genuinely didn’t know before I put my “year of marriage” model together.

Most remarkable of these is that all of the change in divorce rates since the 1960s has taken place in the first five to ten years of marriage. In the early years, divorce rates depend on how healthy couples are when they begin married life, just as they would be for the first few years as a young fish in the pool.

The flipside is that once couples pass their tenth anniversary, their risk of divorce is identical to every other couple who has got married since the 1960s. In other words, despite all of the tremendous social and technological changes of the last fifty years, the nature of established marriage has stayed the same.

The useful conclusion from this analysis of divorce is that strengthening marriage can only really happen during the first ten years. After that, it doesn’t seem to matter in which year you got married. Married couples are married couples just as fish are fish!

2 thoughts on “Making sense of divorce rates: It’s all about the fish!

  1. Comment from Disqus, John: Harry, how might changes in the length of cohabitation before marriage influence the divorce rate in the early years of marriage or the peak risk period? Can you sort the ‘fish’ by previous relationship status? Have those who cohabited effectively already survived some of the early years?

  2. Comment from Disqus, Harry Benson: Some people think that it’s the total length of the relationship that matters: living together as unmarried and then married merge into one another. So if couples are living together for longer before they get married, then maybe early divorce is in reality the end of a relationship that has lasted ten years or more.

    i don’t think this argument holds. For starters, the very act of marriage involves a decision that clarifies the plan for the future and makes being married feel different. That decision can affect how people behave.

    But also all the stats since the 1960s show that years 3 to 6 are the riskiest for a marriage. So whether couples have lived together for shorter or for longer doesn’t seem to have affected the timing of this year 3 to 6 effect.

    The chart below shows this well.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s