Review of Seasonal Adjustments
20 December 2004
By Dr. Gordon Hughes et al.
In this report, the authors review the seasonal adjustments applied to data on social security benefit claimant numbers used within the formulae for determining Revenue Support Grants for Local Authorities. This analysis was to support the Office of the Deputy Prime Minister (ODPM) in setting Revenue Support Grants for Local Authorities that more fairly reflected local needs.
The numbers of social security benefit claimants are currently recorded at a local authority level for all claimants in August each year and for a 5% scan of claimants in each quarter of each year. The August counts are used as a measure of annual local authority needs in the revenue formulae and are seasonally adjusted to take account of the differences between August counts and annual averages across local authorities.
NERA's research team identified differences in the seasonal pattern of claimant numbers across local authorities using the quarterly data. The team also examined which characteristics of local authorities explain seasonal differences (e.g., whether or not the authority is in a coastal area, has a university, etc.). NERA's econometric methodology involved examining the time-series properties of the data using X12-ARIMA filter analysis and examining the variance of the seasonal effects in respect of local authority characteristics using panel data analysis. The econometric estimates were subjected to rigorous statistical tests for significance to ensure that local authorities were not unduly favoured or harmed by the presence of sampling errors. The results from this analysis provided the basis for an updated set of seasonal adjustments for ODPM to apply in future local authority funding determinations.
