In light of the criticism's leveled at Benjamin and Kochin's 1979 article in the 1982 Journal of Political Economy, assess how successfully Benjamin and Kochin defended their statistical approach in their 1982 rejoinder.

Benjamin and Kochin also attempt to defend their ‘failure to explicitly estimate the effects of benefit disallowances on the unemployment rate,’ despite actually agreeing with Cross.  They criticise him for not attempting to estimate such effects and explains that ‘he does not do so for the same reason that we refrained from doing so: disallowances principally reflect confusion, not stringency.’ Although Benjamin and Kochin seek to explain that ‘it is possible to devise measures of stringency,’ a defence based on an agreement with the critic means that the extent of a successful defence has to be limited.

Although Benjamin and Kochin provided some reliable arguments against Cross’ criticism, the accuracy of the belief that the system was becoming more liberalised must be taken into serious consideration when analysing high unemployment in interwar Britain.

Benjamin and Kochin also attempt to defend their statistical model against critics Ormerod and Worswick, who question whether it was sensible to embark on the road of Benjamin and Kochin’s equation (1) at all.’  They are critical of the regression of unemployment on a measure of the benefits-to-wages ratio, that so-called replacement ratio, and on a measure of deviations of output from trend over the period 1920-38.’  Ormerod and Worswick ‘examine the statistical validity of the equation’ as ‘the estimated coefficient of B/W is very sensitive to small changes in the sample period’ and due to the lack of ‘the robustness of the results when a time trend is added.’  In summary they refer to ‘the spurious nature of the results obtained by Benjamin and Kochin over the 1920-1938 period’.

Benjamin and Kochin seem to criticise Ormerod and Worswick rather than directly defending their own statistical model but by nullifying Ormerod and Worswick’s alterations, they can attempt to defend their own model.  Benjamin and Kochin believe that ‘none of the time periods listed by Ormerod and Kochin in table 3 is significantly different for the whole period.’  Also, they comment that ‘the chief effect of adding the time trend is to increase the standard error of the coefficients of all the other independent variables – the classic result of introducing an irrelevant variable into a regression equation.’

Furthermore Benjamin and Kochin defend their statistical model as despite making alterations due to observations from Ormerod and Worswick, they explain that the changes actually ‘imply a somewhat higher estimate of the effect of the interwar insurance system.’  Thus, they manage to defend the overall conclusions of their statistical approach but only to a limited extent do

they successfully defend their statistical approach against Ormerod and Worswick.

Benjamin and Kochin’s statistical approach also needs defending against Michael Collins as he states that ‘nothing in Benjamin and Kochin’s regression equation directly allows for sectoral differences.’  He refers to the ‘serious omission of the absence of any independent variable which explicitly captures the effects of startling sectoral or regional disparities which existed.’  Collins does not see that Benjamin and Kochin’s main independent variable, (B/W), was important for all major groups within the labour force. 

Indeed his own data produced to show sectoral differences provides evidence that Benjamin and Kochin’s ‘regression results – seem to have only limited application,’ as ‘their model is far too narrowly based.’  The historical argument he makes is also an important point that ‘the implied policy recommendation of the anti-Keynesian analysis employed by Benjamin and Kochin – seems to fly in the face of historical evidence.’

Benjamin and Kochin themselves attempt to include an economy wide measure and state that ‘ the principle statistical effects are to (i) reduce the estimated coefficient of the industry-specific measure and (ii) to reduce the extent of serial correlation.’  However it does not seem too successful to state that only at ‘face value’ are ‘these results strikingly consistent with the notion that the insurance system had important effects during this period.’

Benjamin and Kochin are successful only to a small extent in their response to Collins in that Benjamin and Kochin defend their statistical approach by criticising Collins’ statistical approach of accounting for sectoral differences and by developing their own methods, instead of defending the reliability of their original approach. Thus, there was in fact a flaw in their original statistical approach.  

 

Even with the development of their approach to take into account the affect of sectoral differences, Benjamin and Kochin admit that ‘these industry specific regressions are, thus, somewhat less useful than our original time-series estimates.’  They do however state that they ‘offer new evidence on the effects of unemployment insurance,’ but this is not a direct defence of their original approach.

Although Benjamin and Kochin do indeed respond to Collins’ proposition for them to adopt a ‘more explicitly cross-sectional analysis’, their new equation does not lead to a successful defence of their statistical approach.

Metcalf, Nickell and Floros briefly refer to fact that they also do not view the Benjamin and Kochin replacement rate as valid but their main argument is instead a historical argument undermining the whole concept of Benjamin and Kochin’s statistical approach.  Metcalf et al. outright reject the assertion that unemployment insurance relative to wages was the cause of high unemployment in interwar Britain as their own evidence of the post-war period ‘demonstrate that the assertion is completely false.’

Metcalf, Nickell and Floros argue that the Post-War insurance system was much more generous than in the interwar period and due to the fact that unemployment in the postwar period was much lower, Benjamin and Kochin’s theory is contradicted. Benjamin and Kochin however highlight that their theory is based on evidence that ‘during the twentieth century, unemployment has been low in Britain when unemployment benefits have been unattractive, and unemployment has been high when benefits have been attractive.’  A substantial defence indeed but this does not address the issue of Post-War unemployment.  Benjamin and Kochin however, still feel confident that their defence is successful as they conclude simply, ‘that our study is correct.’

Benjamin and Kochin criticise Metcalf et al. for missing the ‘striking difference between the immediate and postwar period and so highlight that unemployment was not necessarily low for the whole ‘post war’ period.  Benjamin and Kochin also state that ‘their series are biased by their failure to account for the high wages of the postwar entrants to the insurance system.’ 

Importantly however Benjamin and Kochin discredit Metcalf et al. by explaining that since OXO systems were the single most important way in which the insurance system was exploited in the interwar period, we conclude that the interwar system was indeed unique and that its administrative features were generous.’ 

Benjamin and Kochin have had to respond to an unprecedented amount of criticism, which seriously calls their theory into question.  In their 1982 rejoinder, Benjamin and Kochin are confident in their ability to successfully defend their statistical approach, but in reality this success is limited.

        

---

Bibliography

Pat Hudson, History by numbers: An introduction to Quantitative approaches (New York, Oxford University Press Inc., 2000) pp. 254 – 55.

Daniel K. Benjamin; Lewis A. Kochin, “Searching for an Explanation of Unemployment in Interwar Britain”, The Journal of Political Economy, 87, 3 (Jun., 1979), pp. 441-478.

Daniel K. Benjamin; Lewis A. Kochin, “Unemployment and Unemployment benefits in Twentieth-Century Britain: A reply to our critics”, The Journal Of Political Economy, 90, 2 (Apr., 1982), pp. 410 – 436.

David Metcalf; Stephen J. Nickell; Nicos Floros, “Still searching for an Explanation of Unemployment in Interwar Britain”, The Journal of Political Economy, 90, 2, (Apr., 1982), pp. 386-399.

Michael Collins, “Unemployment in Interwar Britain: Still searching for an explanation”, The Journal of Political Economy, 90, 2 (Apr., 1982), pp. 369-379.

P. A. Ormerod; G.D.N. Worswick, “Unemployment in Interwar Britain, The Journal of Political Economy, 90, 2, (Apr., 1982), pp. 400-409.

Rodney Cross, “How much voluntary unemployment in Interwar Britain?” The Journal of Political Economy, 90, 2, (Apr., 1982), pp. 380-385.  

---