Saturday, October 06, 2007

Unity on Marriage

Unity's post attacking the IDS-Cameron position on marriage consists of four arguments and one little aside about the Conservative Party.

Argument no. 1: LATs are benefit fraudsters and don't deserve to be rewarded.
Argument no. 2: The success of lesbian families undermines the importance to children of a father.
Argument no. 3: Selection bias is the cause of marriages being more stable than cohabitation.
Argument no. 4: Not all married couples have children.

Aside: The Conservatives are trying to pretend that they're a modern party but they're too old.

I hope if I've mischaracterised or misunderstood any of these arguments Unity will correct me but his was quite a long post so I've had to boil it down a little. I'll try to reintroduce some of the complexity in his ideas as we go along.

I'll deal with the aside first:

"Oh, and as a snarky aside, does anyone else think that Cameron’s efforts to put himself and the Tory Party across as young, fresh and modern were made to appear rather hollow every time the TV coverage - I did catch odd bits of it - cut away to from the speaker to show them addressing an audience that looked for all the world like a Darby and Joan club outing to an Old-Time Music Hall Show? That’s the trouble with Tory conferences, no matter how hard they try to project an image of modernity it all falls flat because most of the rank and file make the auditorium look like god’s waiting room."


I haven't seen enough of the other conferences to be able to make a comparison but Matthew Parris has. While he is a Conservative I think his days of blind partisan loyalty are over, if such loyalty ever really existed, and this analysis can be trusted in the absence of another estimate:

"Mind you, it’s not just me: these Tories really are getting younger. I conduct an informal grey-heads count at all three party conferences these days, and this year the oldest conference was indisputably the Liberal Democrats. Either the Tories are getting younger or the ones we used to see are now so ancient that they’ve stopped coming; but there’s no doubt that the Conservative representatives of 2007, if not in the first flush of youth, are at least young enough to find it worth tinting their roots."


All parties are ageing. My understanding is that fewer young people support the Conservatives but that within the 0.001% of each party's support that are activists (and might go to conference) the Conservatives actually have, if anything, more relatively young people. Look at university Conservative Associations - they're usually pretty large compared to the other major parties.

On 'Living Apart Together' Unity misrepresents Cameron. When he cited that particular phenomenon it wasn't an attempt to excuse the 'LATs'. Instead, the LATs are practical examples that suggest something is going wrong in the benefits system for some people (even if there is a mistake in Frank Field's particular calculations). You don't have to think that it is morally acceptable to cheat on benefits to think that a system where cheaters cheat by pretending to live apart must be poorly constructed.

The existence of the LATs suggests that a combination of the benefits system and potential punishment for benefit fraud (plus, for a precious few, a good old fashioned respect for the law) will be incentivising others to genuinely live apart. The 'reward' Cameron talks about isn't for those committing benefit fraud but for couples that honestly stay together. The hope is that such a reward would encourage others to behave that way.

Next Unity moves on to attack Iain Duncan Smith's report and its emphasis on marriage. I do agree with Unity that the Social Justice reports gets civil partnerships wrong and I don't think that they undermine heterosexual marriage. It seems quite possible they do exactly the opposite.

However, I don't think the success of lesbian couples is nearly as important to an analysis of marriage as Unity makes it out to be. The number of lesbian couples bringing up children is, I would expect, pretty small and those doing so probably have a lot else going for them. I am glad that they can do well but they are too small and unrepresentative a group to take meaningful lessons from that can be applied to the rest of society.

In particular, my simple knowledge of these things suggests that one of the lesbians will often take on a more masculine role within the family which may decrease the importance of the lack of an actual man. Single mothers cannot take both gender roles. Another, probably more important, difference could be that the children in a lesbian family will more often have a father who has not deserted and is still ready and able to play a helpful part in the child's life. A lesbian family is less likely to have been caused by a deserting father than a single parent family. Finally, for a lesbian couple to look after a child is still sufficiently rare that lesbians doing so will have known, when they decided to become lesbian parents, that they were taking on an exceptional challenge. This raises a selection bias of the sort that Unity makes much of later in his piece.

The success of lesbian families is not sufficient evidence to undermine the importance of a male presence in the family and even if it was absent fathers would still be a big problem. Absent fathers - the lack of a stable heterosexual relationship - are in most cases inseperable from single parent families. These face very real problems that Unity acknowledges.

Finally, Unity discusses selection bias and makes the case that it is the main cause of the superior stability of married families. The "social/behavioural changes over the last forty years" that he posits as another factor to explain's marriage's success alongside selection bias only seem relevant in that they might make the selection bias stronger over time if people are increasingly getting married because of genuine, rather than socially forced, commitment. As marriages were, in fact, more durable during the days of the 'shotgun wedding' it certainly doesn't suggest that marriage only encourages stability if it takes place for the right reasons.

Unity doesn't have any real evidence that "in reality most, if not all of the seeming statistical advantages of marriage over cohabitation, in terms of longevity of relationship and likelihood of a couple staying together, can be accounted for by self selection effects and social/behavioural changes over the last forty years."

The closest he comes is an Austrian example:

"Recognising’ marriage in the tax system through an additional tax allowance, effectively a financial incentive to marry, may well induce more couples to marry in the short term but over the medium to long term it is likely to prove counter-productive in the sense that any increase in the number getting married will be offset by a rise in the divorce rate. This is exactly what happened in Austria during the 1970s and early 80s, when the Austrian government responded to concerns about the country’s declining marriage rate by introducing a a modest cash incentive payable to couples entering their first marriage to assist with setting up home together.

This did have the effect of increasing, in the short term, the number of marriages - although only sufficiently to slow the overall decline in the marriage rate and not cause the trend to turn upwards, both on the introduction of the incentive and in the final few months (in 1982) between the announcement that it would be withdrawn and its actual withdrawal, as people married to take advantage on the financial incentive on offer. In between these two ’spikes’, the net effect of the incentive settled on the lowest age group (16-20), where the incentive was perceived to give rise to the greatest benefit by those taking up the government’s offer.

This would be all well and good were it not for the fact that the effect of propelling couples into marriage in order to obtain the incentive offered by the government led, predictably, to a significant rise in the divorce rate, especially amongst those marrying while under the age of 20 during the period in which the incentive was in effect."


It seems interesting but, just like the successful lesbian families, Unity builds a house of cards around it. That more marriages will lead to more divorces is almost a truism - "how many?" is the important question. Most under twenty relationships do not last very long. Did those who got married in response to the subsidy stay together longer, on average, than they would have without the subsidy? If we cannot work that out we cannot say anything about whether getting married has made their relationships more stable or not.

The problems with the Austrian system would seem to be due to a poor policy design. An ongoing tax credit like IDS has proposed seems more sensible than a lump sum subsidy. That will be there to encourage people to stay married, at the margin, as well as to encourage them to get married in the first place. It will also have less attraction to under twenties who are not starting a family and whose relationships we have less of a stake in as a society.

It's very hard to disentangle with the data quite why marriage is more stable than cohabitation. There would seem to be a series of quite plausible explanations of why encouraging marriage might encourage more stable relationships. Three examples:

1) By promoting marriage you send political signals (politics isn't all about direct policy levers) that your society values commitment. You can increasingly socialise an understanding that commitment, particularly to your family, is a good thing.

2) Ending a marriage is, ceteris paribus, a bigger step than ending a cohabitation. Relationships do sometimes end for transient reasons and both partners are, in the medium term, sometimes worse off for not staying together. While some relationships do need to end it is a good thing if people err on the side of staying together.

3) There is a public commitment at marriage that can bind people together. This is a hard one to explain but, at least to me, intuitively obvious when you see people getting married. The character of their commitment does change and become something more substantial.

In the absence of data sufficient to make the decision for us my judgement is that these two goods are more than worth what IDS and Cameron aim to spend on them.

Much of any spending on a married couples allowance will go to couples without children as Unity describes. I would actually prefer a "married with dependent children" allowance to better target the group where there is the biggest public benefit to stable relationships. However, the difference isn't as important as Unity makes out. Stable relationships before people have children will make children, when they do arrive, better off and might also encourage more children - something I would take as good although it's probably not worth starting the huge debate around demographics. Equally, parents are still important to their children after they have fled the nest. Parental support becomes less essential but remains very valuable.

All in all, while I'm glad Unity is engaging with this debate I don't think he has made the emphatic case against the IDS-Cameron agenda on marriage that he thinks he has.

1 comment:

Unity said...

I'll get back to you on much of this, Matt, but what I will say is that I don't see my arguments as a slam dunk, merely as putting up a challenge to the Cameron/IDS agenda on matters that merit a serious debate, in the hope of eliciting further debate.

In short, I think families are a much more complex thing in practice that they are as written in policies.