1.  It is important that the data used to monitor the Treasury's performance are reliable, accessible and accurate, and we take care to specify the various data series against which PSA and SDA performance can be assessed.

  2.  Seasonal adjustment removes the seasonal component from a time series, leaving the trend and irregular components. This has the advantage that, by removing the distortions caused by predictable seasonal effects, it allows month-on-month or quarter-by-quarter comparisons to be made. Seasonally unadjusted data, in contrast, allow reliable assessment only by comparison to the same period the previous year, meaning that a seasonally adjusted series will, in many cases, be preferable.

  3.  The specification of SDA B1 ("to achieve a continued reduction in the number of unemployed people over the age of 18 over the economic cycle") means, however, that in this instance we have had to use a seasonally unadjusted data series. As Welfare to Work strategy is devolved in Northern Ireland, we use data for Great Britain rather than for the United Kingdom as a whole. The Office for National Statistics (ONS) does not seasonally adjust unemployment data broken down by age for Great Britain.

  4.  As, however, both PSA 6 ("to increase employment over the economic cycle") and SDA B1 assess performance over the economic cycle as a whole rather than at any particular point on the calendar, it makes little difference whether the measures used are seasonally adjusted or seasonally unadjusted. Across the cycle as a whole, both data approaches will lead to the same conclusion.

22 September 2003