The persistence of unemployment at the local area level: evidence from the US and the UK

Abstract

In the 1990s and 2000s, unemployment was seen, both by academic labour market economists and policymakers, as a short-run disequilibrium phenomenon. Policy was aimed at increasing the ‘flexibility’ of the labour market, at removing obstacles to the workings of the market, which would ostensibly restore equilibrium in the labour market. In this article, I examine the correlations over time of relative unemployment rates at the detailed disaggregated level of both US counties and UK local authority areas, using the 1990–2010 period. The United States and to some extent the United Kingdom are held up as examples of the more ‘flexible’ labour markets to which other countries should aspire. But even over a period of 20 years, there is strong persistence in relative unemployment rates at local area levels in both countries, and especially the United Kingdom. This result extends to counties and local authority areas within individual states and regions. Local areas with high (low) unemployment in 1990 are likely to have high (low) unemployment in 2010.

Keywords:

• unemployment
• persistence
• local area

JEL Classification:

• J64
• J68
• R12

Notes

¹ Data for Puerto Rico is also available, but this is excluded from the analysis. The county of Yakutat, Alaska, is excluded due to the lack of data in 1990.

² Except for Northern Ireland.

³ We exclude the City of London. It is a most unusual area, having employment of over 250 000 but a tiny resident population of only some 5000 in total. The Isle of Scilly is also excluded due to the lack of data.

⁴ Full results available from the author.

⁵ Connecticut, Delaware, DC, Hawaii, New Hampshire, Rhode Island.

⁶ Wales and Scotland are described as ‘regions’ for these purposes.