Earnings Biases in the UK Regional Accounts: Some Economic Policy and Research Implications

Summary:

The paper discusses a number of possible explanations for this discrepancy: changes in commuting patterns; the increasing prevalence of part-time work; the move towards sub-contracted employment that is not covered by the PAYE scheme; changes in the relative fortunes of different kinds of workers; differences between social security and occupational pension contributions by region.

While these explanations sound plausible, at best they can explain only a small part of the anomalous behaviour of the Regional Accounts data. Instead, the paper suggests that the method of construction of the Regional Accounts may be the culprit. For example, between 1982 and 1990, the Inland Revenue were unable to locate the region of residence for around 12 per cent of the sample of tax records that are used to calculate the regional allocation of income from employment. Only if the proportion of missing records by region and their earnings were uncorrelated with regional income differences would this not lead to a bias in the Regional Accounts. Between 1990 and 1995, the proportion of unallocated records fell to 1 per cent. It is no coincidence that the Regional Accounts, unlike other sources, show a widening South East differential in these years.

The historical unreliability of the Regional Accounts has two possible implications for economic research. First, the effect of income on consumption may have been under-estimated while the effect of housing wealth, which rose disproportionately in the South East, may have been over-estimated. Second, the apparent stability of measures of the regional dispersion of income may also be misleading. New Earnings Survey data show a near doubling of the standard devation of log regional earnings between 1981 and 1990, and a consequently dramatic relative fall in the earnings of regions other than the South East.

One implication of the downward bias in South East income levels in the 1980s is that income in poorer regions such as Wales and the North or North East regions will have been biased up. It will have misinformed public debate and public policy towards the English regions, Wales, Scotland and Northern Ireland. This should be of serious concern in a democracy.

Regional Accounts data also play an important rôle in the allocation of European Union Structural Funds. In order to qualify for Objective 1 status, an NUTS II region (in the case of the United Kingdom, typically smaller than standard regions and larger than counties) must have a level of GDP per capita less than 75 per cent of the EU average in a three year qualifying period which is typically four years before the relevant funding period. For the 2000-2006 period, the newly created Welsh NUTS II region of 'West Wales and the Valleys' has qualified for Objective 1 status. It is possible that if the biases in the Regional Accounts in the 1980s had not occurred, a Welsh region might have qualified for Structural Funds in either the 1989-1993 or 1994-1999 funding rounds. To indicate the funding levels at stake, Merseyside, which did qualify for Objective 1 status in 1994-1999, received around £85 per person per year. Had 'West Wales and the Valleys' been eligible for 1994-1999, with a population of approximately 1.5m, it might have expected EU funding to the tune of around £130m.

Abstract:

Keywords:

JEL Classifications:

Acknowledgements:

Download this paper
Back