"Britain has lower unemployment than Germany" (1997)

[I noticed this media fallacy in 1997. Numbers of unemployed were used to compare unemployment rates between two countries with different-sized populations. Most reporters failed to spot this basic error. The following text is from an article I wrote for In Business magazine – Brian Dean]

The much-reported rise, earlier this year [1997], in the number of German unemployed, might have presented an opportunity to raise questions about mass unemployment in developed countries. Unfortunately, with the British general election looming, many UK politicians were more interested in presenting the German unemployed figures as evidence of the economic failure of European social policy – they quickly proclaimed the economic success of Britain in having "only half as many unemployed as Germany".

When one looks at the rates of unemployment in Britain and Germany, rather than comparing the total numbers of unemployed, a different picture emerges. In Britain, 20 per cent of households are without a job – more than in Germany. It is misleading to compare numbers of unemployed in different sized countries. Germany has a population approximately 1.5 times that of Britain, and one must allow for this in any comparisons.

At the beginning of 1997 the official jobless figure for the UK was 1.9 million compared to 4.3 million in Germany. So, even allowing for the larger German population, Germany still looked worse off. This appears to contradict the above data showing Britain with a higher rate (20%) of unemployment than Germany. Unfortunately such contradictions abound, due to the fact that there are several different ways of calculating the jobless tally.

The official monthly jobless figure counts only those claiming Jobseeker’s Allowance. This count of claimants gives the lowest measure of unemployment, and is widely mistrusted. In 1995, the Royal Statistical Society called for a new monthly count to be derived from the Labour Force Survey (LFS), which provides an unemployment count based on internationally agreed definitions of what constitutes an unemployed jobseeker. Although the LFS measure of unemployment is higher than the official claimant count, it still excludes people on the margins of the job market, who want jobs but are discouraged from "actively" seeking work, as they feel there are no jobs available.

The Employment Policy Institute (EPI), an independent monitoring organisation, provides a measure of unemployment which includes these "discouraged" jobseekers – it simply counts those who are unemployed and want a job. Amazingly the current UK count of such people is well over 4 million (more than double the official claimant count). The recent church report, Unemployment and the Future of Work, endorsed the figure of 4.5 million unemployed in the UK (as measured in mid-1996).

Another measure of unemployment provided by the EPI counts households where no adult has a job (excluding student households and those with retired householders). Although most counts of unemployed individuals reflect the fall in unemployment registered by the claimant count, there has been no corresponding reduction in the number of jobless households. As mentioned above, one in five British households are without a job – 20 years ago the figure was less than one in ten. This measure is probably a more accurate indicator of "social distress" than the individual counts and, significantly, Britain rates worse than Germany in this respect.

John Philpott, director of the EPI, has remarked that politicians should consider why the number of jobless households remains so high, even after four years of recovery in the labour market. Perhaps they should also reconsider the fundamental relationship between economic growth and unemployment in technologically advanced countries.

[2007 Postscript – Prior to being elected in 1997, New Labour promised to use a more accurate (ie much higher) count of unemployment than the "claimant count". But, too date, the "claimant count" is still being used as the official count – ie the figure released to the media]