Unemployment was an important feature of Britain during the interwar years and a conditioning influence on economy and society. Indeed, unemployment is a defining characteristic of the period. The aim in what follows is to examine the extent, nature and causes of the unemployment problem which Britain faced in the 1920s and 1930s.
All unemployment statistics rest upon arbitrary assumptions and imperfect information. It is important, therefore, to know the basis and limitations of quoted statistics (Garside 1980). Official figures before 1914 were compiled by the Board of Trade from returns made by certain trade unions which paid unemployment benefits to their members. These returns reflect a biased labour force sample of mainly skilled workers, but from industries which were prone to fluctuation. They indicate an average level of 4 to 5 per cent unemployment between 1855 and 1914, fluctuating from 2 to 10 per cent over the trade cycle. Beveridge suggested that the average rate of unemployment before 1914 was rather higher than these figures indicate (Beveridge 1960:73) but it is impossible to know how much higher. It is clear that in the pre-1914 labour market there was a good deal of casual hiring and underemployment which is not captured in the unemployment statistics and which probably continued after 1918, providing a further reminder of the limitations of official statistics (Whiteside and Gillespie 1991). Employment and unemployment do not fall neatly into distinct categories and there may be a range of multiform circumstances between full-time work and complete unemployment.
With the advent of the NIS (National Insurance Scheme) in 1911, and its extension in 1916 and 1920 to 60 per cent of the work-force, new statistics became available and formed the basis for an official series. This was based upon spot monthly tallies (and, therefore, does not give accurate