![](https://dl6pgk4f88hky.cloudfront.net/2021/06/gettyimages-86347603-1038x778.jpg)
In 2015, Dr Adam Perkins, a psychologist at King’s College London, published his book The Welfare Trait, and gained considerable attention as a result. The book argues that “the welfare state is gradually warping the personality profile of the population so that more people in each generation are resistant to employment… the welfare state means more children are being born to welfare claimants than to employed citizens.” Unsurprisingly, this is somehow the fault of “the intelligentsia”.
Equally unsurprisingly, Dr Perkins attracted support from those in the media whose prejudices this confirmed. Toby Young, in an article headlined ‘Tell the truth about benefit claimants and the left shuts you down’ described his findings as “bleedin’ obvious”. Jenni Russell of the Times praised the academic for his bravery and wrote: “With an average of 2.5 children in workless households, compared to 1.5 in working ones, he thinks the welfare state has become a production line for personality traits that reduce the motivation to work.” the Telegraph and the Sun were also happy to give him a platform.
One wonders (sadly, I doubt it) if any of the journalists responsible will be remotely embarrassed by reading a quite astonishing “corrigendum” (or correction), which appeared recently in the Journal of Personality and Individual Differences (PAID). It deals with one of Dr Perkins’ papers, which directly underpins the argument in his book. The paper, which was published in 2013, concludes – on the basis of the most astonishingly flimsy, indirect and, it turns out, flatly erroneous evidence – that “the human personality profile in developed countries such as the USA may be gradually evolving by natural selection towards lower levels of personality traits that predispose an individual to be a solid citizen.” The correction, which you can read in full here, identified at least seven errors. These included both the misreporting of results and the use of obviously inappropriate statistical techniques. The errors were pointed out two years ago by Dr Andy Fugard, but it’s taken the journal this long to correct them.
This correction should come as no surprise to anyone who read Perkins’ original blog summarising his argument that “each generation will contain proportionately more children who are exposed to disadvantage”, which in turn increases the frequency of “dysfunctional personality characteristics”. This blog too contained an elementary error. As I pointed out at the time, it just isn’t the case that a disproportionate number of children are born into workless households. In fact, only around one child in ten lives in a workless household. Separately, one of the authors whose work Perkins cited, Professor Mike Brewer of the University of Essex, was sufficiently irritated that he wrote a detailed blog explaining the errors. (Perkins maintains that the results of the paper remain statistically significant and told Times Higher Education that aside from these results, “we still have a range of evidence showing that there are significant links between personality, employability and reproduction”.)
But the more interesting point here is the way in which Perkins was able to influence both the editors of a respectable academic journal and journalists who really should have known better. The system of peer reviewing was introduced for a reason. There is no excuse for accepting and publishing a paper which contains multiple obvious errors, and which uses statistical techniques in a way which would not pass muster in an undergraduate thesis. Elsevier, which publishes PAID, has some hard questions to answer.
With journalists, however, the ignorance is even less excusable. Anybody who knew the first thing about the welfare state could have explained that Perkins thesis was worse than half-baked, that the numbers Russell was quoting were plain wrong, and there was no need to take him seriously. Instead, she chose to paint him as somehow being a martyr to left-wing academic correctness. In this case, “correctness” is indeed the right word.
It’s also worth noting that Russell was happy to vouch that “Perkins is no right-wing ideologue…he is far from being a racist or a eugenicist, as his more ludicrous critics have claimed.” Perkins has since argued that Trump’s ban on Muslim immigration was justified on “human capital grounds”, and appeared on the show of the notorious alt-right commentator Tara McCarthy, now banned from YouTube, during which he engaged in a cosy chat about his theories.
Were the journalists just being lazy? Or did they want to believe Perkins’ arguments? I don’t know, but there is a broader project afoot that argues that unequal outcomes – educational or income inequality, in particular – are driven primarily by genetics, and that those who don’t subscribe to that view are somehow “anti-science”. Toby Young, again, is at the forefront here, as when he recently argued that schools couldn’t do much to help less able pupils or to reduce educational inequalities. Like Perkins, his lack of understanding of basic statistical concepts meant he got the logic completely wrong: this tells you little or nothing about what schools could, if they were run and funded better, do to improve pupil outcomes.
Now, this isn’t my field, so I’m not going to say anything about the real – and exceptionally complex – science (as opposed to the statistics). And I have no doubt that the vast majority of actual scientists working in this field have the best intentions, and understand the data they are working with. But what does appear obvious is that some journal editors and journalists appear to be quite happy to validate and promote obviously junk science when it supports their ideological position. The tale of Dr Perkins should be a salutary lesson.
Jonathan Portes is a professor of economics and public policy at King’s College, London