Fatal Flaw?

Michael O'Connor
Feb 25, 2016 · 5 min read

The Telegraph today (24 February 2016) has an outraged article by Dr Adam Perkins on how

My work to improve the welfare state has provoked screams of outrage and posturing, who’re wilfully trying to stand in the way of scientific progress

I’m not sure who the who in who’re could be: is it the screams who are trying to stand in the way of progress (though see update below)? But let’s put that down to a tired sub-editor and move on to Dr Perkins’s point …

The central argument in my book is that the welfare state is eroding the economic and social prospects of the nation by increasing the proportion of individuals in the population who possess the employment-resistant personality profile.

Now despite the increasing specialisation of diagnostics of personality disorders (and for those who like that sort of thing I’d recommend ‘Fatal Flaws: navigating destructive relationships with people with disorders of personality and character’ by the eminent Prof. Yudosfsky), I don’t think an ‘employment-resistant’ type has yet been identified.

Whether such a type exists or not might be a reasonable matter for debate and one on which opinion might differ, but Dr Perkins is very clear that there is a sound evidence base for his work, saying in the article that

UK data suggest that for every 3 per cent rise in benefit generosity there is approximately a 1 per cent rise in the number of children born to claimants

This however is from a paper by Brewer et al for the Institute of Fiscal Studies that made no distinction between employed and unemployed people, and clearly noted that the primary driver of increasing benefit generosity they observed was in-work benefits, WFTC standing for Working Families Tax Credits.

Unfortunately, and especially unfortunately in a context of hope for ‘scientific progress’ Dr Perkins original research that does focus on associating fertility with unemployment contains its own fatal flaw.

In coming to this conclusion Dr Perkins correctly reports data published by the Office for National Statistics (ONS) about the number of households containing at least one child under the age of 16 and correctly calculates the average number of such children in these households according to the ONS categorisation of households as Working, Mixed or Workless.

However, from these data, nothing can be said about whether ‘unemployed’ people — which is the term used by Dr Perkins — have more children on average than other groups of people. There are both conceptual and data reasons for this.

Conceptually, the ONS definition of a workless household encompasses all households without anyone working. This does not mean that they are all unemployed in the ordinary sense of not working while being capable of work and available for work. It will also cover for example people who cannot work by reason of disability (whether temporarily or permanently) or who are temporarily out of the labour force because they have stopped work while looking after a young child but have every intention of returning to work when the child is older. This further illustrates that the ‘workless’ household might not have been workless yesterday or might not be tomorrow. The presence of a child in such a family today neither means that the child was born in a workless household, nor that the child will be brought up in a workless household, let alone born to or brought up by parents who are unemployed. So comparing the average number of children in households in which no one is presently working to the average number of children in households with someone presently working does not provide any information about the relative propensity of unemployed people to have children.

From the point of view of the data used, it is obvious that the number of children people have cannot be calculated from the number of children presently in their household under the age of 16. A family with children aged 4, 14 and 17 has three children, not two children. Any attempt at estimating whether one group of people is likely to comprise larger families than another group must count all of the children that they have. The ONS data does not allow that comparison. It seems clear that the numbers of children under 16 in households categorised as Working, Mixed and Workless is not a proxy for total children because the distribution of households within these categories differs if all dependent children are counted rather than just those under 16. The picture is further complicated by the fact that not all dependent children live with both birth parents, and of course older children may have left home.

So even leaving aside the conceptual issue, and presuming a impossible static state in which households could not move between the categories of Working, Mixed and Workless, it is simply not possible to calculate anything about the relative propensity of people in ‘workless’ households to bear children. Because this variable cannot be established from the data, so no comparison can be made between this and the generosity (or perceived generosity) of benefits available. For this reason alone there is no basis, let alone a scientific basis, for the conclusion reached by Dr Perkins on these data that benefit generosity drives child-bearing among unemployed people.

Whether an employment-resistant personality type exists, and whether such a personality can actually be created by a welfare state are other matters entirely that go a bit beyond my self-imposed ‘mainly welfare and tax’ remit...

Update Feb 25: I obviously wasn’t the only one to wonder about the ‘who’ was in

My work to improve the welfare state has provoked screams of outrage and posturing, who’re wilfully trying to stand in the way of scientific progress

as this intro has since been amended in the Telegraph, the work no longer provoking screams but merely eliciting outrage and from wilful students and reading

My work to improve the welfare state has elicited outrage and posturing from students who are wilfully trying to stand in the way of scientific progress

As I’m not a student, and I’m very even-tempered, I hope that Dr Perkins will accept my contribution as neither outrage nor posturing …

Further Update Feb 26

My original post assumed a knowledge of the background that the few I thought likely to read it would have already, but as it is gaining a much wider readership I’ve added a screenshot of Dr Perkins specific assertion about children born to unemployed people and added attribution and some explanation to the Brewer quote.