Going to the chapel and we’re going to apply for transfer payments

Or, three cheers for the age of entitlement

Jennifer Oriel has a theory. I don’t think it’s a very good one. Surveying the contemporary Australian political scene with one sub-Gibbonian eye, she finds that:

“the Australian Labor Party’s national conference last weekend illustrated a modern truism. The 21st-century Left does not empower the disadvantaged. It empowers disadvantage.”

Empowering disadvantage seems like a moderately silly political platform to me. How come those on the Left are so fond of poverty to the point where they are liable to entrench it where they find it? Well, you see, portions of the Left have historically been in favour of making it easy and financially viable for women in particular to exit from terrible, abusive marriages, and to stay out of such marriages in the first place. If you want to blame the hippies for successfully lowering marriage rates, then that’s something that the hippies might well be proud of.

Well, that’s all very well. But these things don’t come for free! Oriel is worried about a culture in which marriage is no longer the subject of universal hortations:

“Recent American research indicates that new, entrenched categories of social disadvantage have emerged that may be a byproduct of hard-Left revolutions against traditional institutions such as marriage and the family. The research has renewed calls for a compassionate conservative approach to equity, which promotes key conservative values exercised with individual responsibility and fiscal prudence as the path to good society.”

The gravamen of the offence certainly seems serious. But I’m afraid you’ll have to do a bit better than that to make me cry. ‘Recent American research’ is if anything just as divided on the sources of social disadvantage as it ever has been; some Americans may very well believe that ‘hard-Left revolutions’ have everything in the world to answer for, but the consensus is by no means total. Oriel, on the other hand, thinks that the selection of journal articles she has compiled is so compelling as to demand an Australian equivalent, a market demand which she then undertakes to supply.

“Using the Australian Bureau of Statistics’ socioeconomic index for areas, I compiled a list of the highest and lowest ranked locations…and compared data…the small sample I analysed did indicate religious composition and marriage rate may be correlated with socioeconomic status in Australia, especially the latter variable.”

There are several reasons I’m not all that surprised. The most obvious one is that the ABS’s ‘socio-economic index’ — actually four separate indices — is constructed from component variables, one of which is the percentage of people separated or divorced, and another of which is the percentage of one-parent families.

Department of Obvious Conclusions, Figure 1: correlation between marriage and divorce rates in State suburb areas. Source: ABS Census 2011. Both rates are calculated as the number of individuals married or divorced as a proportion of all people over 15.

I suspect that the proportion of people in a suburb who are married might be related to the proportion of people who are separated or divorced and the proportion of one-parent families in ways that are more mathematical than sociological. (Look to the left.)

A much better way of looking at the benefits of marriage would be to look at its impact on poverty or on income. We can do this easily enough at the aggregate level. (Oriel’s use of state suburb areas is interesting enough as far as it goes, but then it doesn’t go very far. Just because some areas with low marriage rates also have high rates of low income households doesn’t mean that the unmarried people in that area are the ones who have low income.) Below are the 2011 income distributions of households with children. The red distribution is for households with one parent, and the blue distribution is for households headed by a couple. Given the choice, as children are not, you’d probably pick the family with two parents.

Weekly household income distribution, single parents and married couples with children. Source: ABS Census 2011.

Marrying someone means pooling your economic resources together: it’s no wonder then that families with two adults have more economic resources than those that only have one adult. Of course, they have to share it between twice as many people as they did when they were single (more, if they have children). But there are serious economies of scale: if you move in with your girlfriend or boyfriend, you don’t need two beds, two ovens, and two showers like you did when you each lived in your separate apartments. You don’t need as much money to live the same lifestyle.

But this does not mean that the solution to poverty is marriage, a proposition that Oriel implies but never comes out and states. One excellent reason to be skeptical that it might be a solution is that there are other things that make people both more likely to be rich and more likely to get married. For just one thing, people’s wages tend to rise as they get older, and people are more likely to be married the older they get. (In 2011, around 6.3% of people aged 20 to 25 were married, compared to about 60% of those aged 35 to 39.) Implement some compassionately conservative government program that encourages people to get married, and you may very well encourage two people who otherwise wouldn’t to get hitched. But don’t be surprised if those two people don’t magically become rich.

For example, among people in their 30s during the 2011 census, 76.4% of those who were married had a Year 12 or equivalent qualification; that number was only 61.7% for those who were divorced, and 58.08% for those who were separated from a partner. If you think that there’s some kind of relationship between education and marriage, which do you think is more likely: that having a Year 12 education makes a 38-year-old woman more likely to get married, or that getting married makes the 38-year-old woman more likely to get a Year 12 education? Without getting into the copious literature on this subject, I think that, if relationship there is, the former is rather more likely than the latter.

Unconvincing though Oriel’s own analysis is, her treatment of other people’s work is, if anything, even shoddier. For example, in considering the role of marriage in subduing income inequality, she writes:

“A 2005 study by Chulhee Lee demonstrated that over the three decades from 1968 to 2000, changing family composition had a more depressing effect on the income of the lowest income groups in the US than labour market factors.”

This sounds very impressive. If only poor people had gotten married more, they wouldn’t have stayed so poor, right? Well, I looked up Chulhee Lee’s 2005 study, “Rising family income inequality in the United States, 1968–2000: impacts of changing labor supply, wages, and family structure”. In the paper, Lee examines the increase in income inequality between the poorest 10th of society and the middle of society. The increase is assigned to various contributing factors: changes in labour supply (that is, the number of hours worked), wage rates, the composition of income from different sources, the marriage rate, as well as changes in other income.

Contribution of various changes to increased inequality between the bottom decile and the middle from 1968–2000. Source: Chulhee Lee, “Rising family income inequality in the United States, 1968–2000: impacts of changing labor supply, wages, and family structure”, NBER Working Paper.

Oriel would have us believe that the contribution of ‘changing family composition’ was larger than ‘labour market factors’. In fact, as the graph to the left demonstrates, changes in the proportion of families with both a husband and a wife accounts for merely 11% of the increase in inequality between the bottom decile’s average income and the overall average income. The combined impact of changes in the quantity of labour supplied and the price at which it was supplied accounts for just over 44% of the increase in inequality.

Inequality is a subject that has received increasing attention from the Left, especially since the publication of Thomas Piketty’s Le capital au XXIe siècle in 2013 and its translation into English the year after. (Interestingly, one of Piketty’s earlier papers was on the impact of divorce on educational attainment: he found that it was the parental conflict that caused lower grades in children, not the actual divorce.) And though Oriel doesn’t seem to have reported Lee’s numbers very accurately, it is nonetheless true that a certain proportion of the increase in inequality from 1968 to 2000 could be blamed to some degree on the decline in marriage rates.

But that doesn’t mean that egalitarians ought to direct their attention to the marriage rate. (It’s not particularly clear how you solve a problem like a secular decline in the marriage rate, either. As Steve Randy Waldmann says, marriage promotion is a cargo cult; hence why, when rigorous examination of existing, and often quite expensive, marriage promotion schemes is undertaken, they don’t seem to have much of an impact.)

Concerns about inequality are much better remedied with a less oblique treatment: income redistribution. Think of this as the logical extension of, rather than an alternative to, the traditional economic function of marriage. Marriage was, among other things, the species’ first attempt at a redistributionist welfare state. By itself, it was small by comparison to welfare schemes that exist today, although marriages were sometimes much larger in than in contemporary Australia. (“And he [King Solomon] had seven hundred wives, princesses, and three hundred concubines: and his wives turned away his heart” — I Kings 11:3). Nonetheless, marriage facilitated the pooling of economic resources between two and sometimes more people, which reduced the risk of poverty for both adults and children. The need for redistribution was particularly acute for women: for most of human history for which we have reliable estimates, average wages for women were perhaps enough to feed and house one person, but never all that much more. This is illustrated for the case of England in the chart to the side.

Welfare ratio for males and females, 1260–1850 in England. The welfare ratio is calculated as the average nominal wage divided by a typical consumption basket. Source: Jane Humphries and Jacob Weisdorf, “The wages of women in England, 1260–1850”, http://www.economics.ox.ac.uk/materials/papers/13260/jhreplacement.pdf

Wage-earning families were therefore usually redistributionist: men would earn specie and give it in part to their wives and children. Women would redistribute to their husband and children the fruits of the non-market work they performed, as well as any wages they earned in the market. This pattern endures in many societies today. Historically, this was redistribution of a decidedly tyrannical flavour. Women were often forced, legally, to surrender to their husbands all rights to the property they owned and the income they earned. In England, as E. S. Turner pointed out in his witty little book on social reform, Roads to Ruin, it was for a long time entirely permissible under the Common Law for a man to support his mistress on the earnings of his wife, without any need for approval from the latter.

Fortunately, the law has changed. Women in the West enter into marriages today with much more power to control the joint resources of the family, although in many cases this power is still unjustly circumscribed. But the ability of a family to redistribute income is severely limited by the individual market wages that a small group of adults can command. The ability of a whole country to do so is much, much greater: hence the rise of the welfare state. In France, the government spends an amount equivalent to over 30% of the country’s annual production on social expenditure. On average, OECD countries spend around 12% of GDP on cash benefits alone.

This by no means obviates the need or desire for marriage. Historically, given that women earned much less than men in labour markets, it made economic sense for many of them to undertake non-market work. Fortunately, though, the vast gender wage gap that persisted throughout most of recorded human history has begun to recede, although it still has some way to go. As Betsey Stevenson and Justin Wolfers point out, this means that the complementarities in production that used to be part of the economic rationale for getting married have become less important, and consumption complementarities — in English, the things that are more enjoyable to do or to consume with someone else — will become increasingly important. Similarly, the increase in women’s market wages has meant that the redistributive function of marriage has waned.

Correlation between child poverty rates and marriage rates in the OECD, 2012. Source: OECDStat.

That the power of marriage to alleviate poverty is not as good as Oriel might like is particularly obvious when you look at other advanced countries. Child poverty rates are actually positively correlated with marriage rates in the OECD: the higher the proportion of people who get married, the higher the child poverty rate. Now, there are good reasons to suppose that this might not be a causal relationship either. But more sophisticated exercises also confirm that there’s not much of link between marriage rates and child poverty levels at a country level— a phenomenon for which there exists a very simple explanation. Countries with low marriage rates and high out-of-wedlock births are also countries with very generous social welfare schemes.

So if you’re interested in reducing income inequality, or eradicating child poverty, or even mitigating the socio-economic disadvantage upon which Jennifer Oriel rather cluelessly concentrates, don’t bother promoting marriage. It’s a dead end, and it ignores the unmistakably clear choices about marriage that men and women have made over the past four decades or so — choices that have been made both in rich societies that, if you like, ‘incentivise’ marriage with meagre welfare systems and in rich countries, like the Nordic ones, that do not. If people don’t want to get married or stay married at the rates they did in the 1950s, then the task ought not be to attempt to change those choices; it should be to build a generous social welfare system that ensures that children who grow up with one parent aren’t consigned to poverty because of it.

But if none of this is convincing to ‘compassionate conservatives’ like Oriel, then I can think of one thing that might increase marriage rates — an idea which Oriel, rather unfathomably, doesn’t like:

“Another quota for Labor women, a vote for gay marriage and a campaign against the yen reads like a Dr Feelgood script for a party that has lost its moral ­compass.”

Who knows what a ‘campaign against the yen’ means (and who knew that international currencies were so vital a component of one’s ‘moral compass’) but I wonder why, if marriage is such a great thing, Oriel thinks its acceptable to continue preventing two women or two men from contracting one.