Traditionally, Budget analysis consists of calculation of the "overnight" effects on particular families. Our analysis departs from this by using survey data, representative of the actual population, which allows us to assess the overall distributional effects. We capture the diverse range of living situations and individual characteristics in the UK today, in the correct proportions. In comparing the impact of two policy regimes it is usual to compare the new policy with a version of the old, indexed forward by a year to account for inflation. Thus the counter-factual is that benefit levels and tax thresholds would have been uprated for changes in the price level.2 However, the recent trend for announcing changes several years in advance of their implementation, makes the comparison more complicated. Should we compare the actual systems, this year and next, even though some of next yearÃs changes have been known for some time? Or is it more interesting to analyse the impact of the new measures that are announced, regardless of the date at which they will take effect? In this note we do each in turn.