I am often surprised at the level of distortion that occurs when the conclusion of a statistical study is rewritten and edited for consumption by the wider news reading public. A case in point is the recent study by Dr. Luis Angeles, “Do Children Make Us Happier?” This is a statistical analysis of a life satisfaction survey that covered British adults from 1996 onwards. It found that children had little to no effect on life satisfaction for the population as a whole, however if divided into different groups then children tended to increase the satisfaction levels of married couples. However this is a statistical average – a fact that is often overlooked when the headlines get written:

There is certainly a lot of confirmation bias amongst journalists and newspaper editors. Those with a large family are especially prone to see only what fits in with their beliefs. If they had read the study in detail they might have noticed that children had an even greater effect on raising the life satisfaction of widows. Which leads one to wonder why no one came up with a headline that read: “The key to happiness is kids and a dead husband”?

There is so much temptation to infer direct causality, but even the author of the original research cautions against this when he concludes with:

One is tempted to advance, on the contrary, that children make people happier under the “right conditions”. We do not mean this as a moralistic defense of marriage. Instead, by right conditions we have in mind the time in life when people feel that they are ready, or at least willing, to enter parenthood. This time can come at very different moments for different persons, but a likely signal of its approach may well be the act of marriage.

This is an example of a latent variable effect. Angeles did attempt to explain this to the mainstream media:

Dr Angeles, director of the university’s Centre for Development Studies, speculated that the reason for the difference in happiness levels lay in whether children were planned or not.

He said: “People who have decided to get married probably have in mind that they want to have children.

“But in the case of unmarried parents, children might not be expected.”

So children tend to make parents happier if they were planned for, and married couples tend to be experienced practitioners of planned parenthood. It is not a surprise that married couples (who themselves may have spent a few years living together unwed) tend to be more experienced with the use of contraception than unmarried cohabiting couples. Experience grows with length of time spent cohabiting, married or not. But no survey is going to ask parents if their children were expected or unplanned. Marital status merely serves as an imperfect proxy for this.

Furthermore there is an element of survivorship bias when one realises that the married group excludes the divorced. Weren’t they once married? It is unlikely that their children were conceived after the marriage was dissolved. (We’re assuming that the remarried would fall into the married group.) So any previously married couples that broke up after a fall in life satisfaction don’t get included in the statistical average for the married group. In the study, children exhibit a very wide variance on the divorced group that included both negative and positive ranges.

And so the conclusion we should have made, yet is almost never reported in the news, is that children make couples happy if they are well prepared to be parents. Who knew?

The childfree should stand proud as this is what they’ve been saying all along: Only have kids if you really want them, are prepared for the parenting role, and all the responsibilities, demands on time and resources that entails.

Advertisements