How linked are available benefits and unemployment levels?
“We have been encouraging working-age people to have children and not work, when we should be enabling working-age people to work and have children. So it’s time we asked some serious questions about the signals we send out through the benefits system.” (David Cameron)
“The problem then is that many men [sic] with families find themselves as well off, or in some instances better off ‘on the dole’ than they would be if they were working. The surprising thing is that, in spite of this, many of them actually prefer to work…” (Pilgrim Trust)
“it is a far cry from recognising the influence of these factors to Benjamin and Kochin’s sweeping claim” (Ormerod and Worswick, 1982)
So here we have David Cameron launching a full-on attack on the benefits system. This is significant for a number of reasons – firstly, the age-related housing benefit cap is an awful policy; secondly, as Jackie Ashley and Patrick Wintour ruminate over, this seems like a marked deviation from the coalition line; thirdly, some of the language I’m seeing is the clearest statement yet from Cameron that he is as wedded to a highly specific political theory about unemployment as suspected. I want to pick up on this last point for now. An article was written in 1979 which apparently had some considerable traction with the Tory group. I want to revisit this now because I think they’ve never grasped, even though it took mere months for academia to rubbish that article, that it’s fatally flawed.
Conservative thinkers have long been highly fond of an article by D. Benjamin and L. Kochin, entitled “Searching for an explanation of unemployment in interwar Britain”. This highly influential and widely debated article sets out the clearest articulation of a common conservative position – that unemployment originates from, or is at least exacerbated by, the existence of a generous welfare system. More specifically, Benjamin and Kochin’s claim is that there is a demonstrable correlation between unemployment data and wage:benefit ratios – such that an effective way to reduce unemployment would be to reduce available benefits. They believe thus that in that period, “the army of the unemployed… was largely a volunteer army.” This is still hugely popular as a theory. Whether that’s right or wrong is a debate that’s still going on to this day, but what we can say with some certainty is that the original paper by Benjamin and Kochin has big holes in it. I think this is an important point to make because people still casually cite their conclusion as if it’s an encyclopaedic fact, when it’s far from so.
In brief, then, I want to list some of the key points which arise from the expanse of available literature on why Benjamin and Kochin’s theory just doesn’t stand up. Note that the paper refers to the “interwar period”, taking data from the 20s and 30s.
Firstly, the thesis is grounded in what’s actually an unlikely assumption – that the decision to find employment or live off benefits is a relatively discretionary choice based on a kind of cost-benefit analysis. As Cross demonstrates, “genuinely seeking work” and means-based tests were increasingly influential throughout the period. Unemployment has also been shown to be a terrifying prospect which can induce and exacerbate depression and anxiety. We should not assume that benefit-dependence is a simple, rational choice. Even if it were, this would itself still be problematic for Benjamin and Kochin, as only 2% of adult men could receive benefits greater than 80% of their last working income. (In other words, if “should I get a job or not?” was actually analysed, free choice they describe, the likely answer would be “yes, I should”.)
Secondly, the fact that unemployment benefits were at similar levels throughout the post-war period, marked by radically low unemployment, should cast great doubt on Benjamin and Kochin’s theory.
Third, and crucially, the evidence employed by Benjamin and Kochin invites scepticism. They base their theory on three claims;
(1) that female unemployment declined when benefit entitlement for women was reduced,
(2) that juveniles (for whom low benefits were available) showed above-average employment,
(3) a regression-based analysis which they claim demonstrates a clear correlation between benefit-to-wage ratios and unemployment levels.
The first two can be easily dismissed – that the number of people in any group claiming unemployment benefits declines when regulations become stricter is evidence of nothing except the immediate impact of those regulations: we can draw no conclusions about the actual impact on female employment (those people may have just consequently starved). Further, it has been demonstrated that juvenile employment was a special case, characterised by short-term employment exploiting lax juvenile employment regulations.
With regard to their regression design, their use of a benefit-to-wage ratio is unconvincing, as a clear correlation is only evident using the highly selective and seemingly arbitrary values which they employ. When real values are substituted, the statistical relationship becomes much less clear. Furthermore, the use of a logarithm of net national product balanced against the trend value of the logarithm for the period is not convincing as a means of discounting such a broadly important factor as aggregate demand; they give no explanation which guarantees the effectiveness of this model, and it might be open to questions about delays between changes in demand and effects on unemployment.
Finally, Benjamin and Kochin’s use of regression data to explain what they claim is an individual phenomenon at a collective level is questionable and may lead to misinterpretation. Hatton demonstrates that unskilled workers were more likely to see unemployment due to structural reasons. Incidentally, this same group are also subject to lower wages and thus a higher benefit to wage ratio. At a collective level, this effect would deceptively appear to link the benefit-to-wage ratio to unemployment.
It doesn’t exactly trip off the tongue, but next time you hear Mr Cameron saying things which just assume that the unemployed are voluntarily so, you know what to say.
It’s worth adding that there are a lot of other things wrong with Cameron saying things like the quote at the top of this post. It’s offensive and propagates divisive myths about the unemployed. Much could be written about that, but I really wanted to focus here on the claim itself that unemployment is a consequence of high benefit provisions. Of course, the methodological flaws in Benjamin and Kochin’s thesis do not mean that it is definitely untrue that benefits affect unemployment. Doubtless, in some cases, it is very true that they do. The significance is that their article is some of the most important work done to try to prove something rather more doubtable – that it’s a big, widespread, highly influential factor. Given it’s flaws, that proof is still very much missing. Someone let the Prime Minister know?
If you’re interested and have the time, then I really recommend a look at some of the material I’ve listed below. There’s plenty more available, but this is a summary. The best place to go if you want more on the Benjamin and Kochin thesis is the symposium in volume 90 (number 2) of the Journal of Political Economy.
Benjamin, D. and Kochin, L. “Searching for an explanation of unemployment in Interwar Britain”, Journal of Political Economy: 87(3) (1979), pp 441-478
Broadberry, S. “The Emergence of Mass Unemployment: Explaining Macroeconomic Trends in Britain during the Trans-World War I Period”, Economic History Review 43(2) (1990), pp 271-282
Collins, M. “Unemployment in Interwar Britain”, The Journal of Political Economy 90(2) (1982), pp 369-379
Cross, R. “How much voluntary unemployment in Interwar Britain?”, The Journal of Political Economy 90(2) (1982), pp 380-385
Daunton, M. Wealth and Welfare (OUP: Oxford, 2007)
Eichengreen, B. “Unemployment in Interwar Britain”, in Digby et al, New Directions in Economic and Social History vol II (Macmillan: Basingstoke, 1992)
——. “The British Economy Between the Wars” in Floud and Johnson (eds.), The Cambridge Economic History of Modern Britain: Vol II (CUP: Cambridge, 2004)
Glynn, S. and Booth, A. “The Emergence of Mass Unemployment: Some Questions of Precision”, Economic History Review 45(4) (1992), pp 731-738
Hatton, T. “Institutional Change and Wage Rigidity in the UK”, Oxford Review of Economic Policy 4(1) (1988), pp 74-86
——. “Unemployment and the Labour Market, 1870-1939” in Floud and Johnson (eds.), The Cambridge Economic History of Modern Britain: Vol II (CUP: Cambridge, 2004)
Metcalf, D., Nickell, S. and Floros, N. “Still searching for an explanation of unemployment in interwar Britain”, The Journal of Political Economy 90(2), pp 386-399
Ormerod, P. and Worswick, G. “Unemployment in Interwar Britain”, The Journal of Political Economy 90(2) (1982), pp 400-409
Shamir, B. “Self-Esteem and the Psychological Impact of Unemployment”, Social Psychology Quarterly 49(1) (1986), pp 61-72