Periodic incomes policies have by now become a feature of most Western economies. It has also become clear that there is a trade-off between the features of an incomes policy that make it initially acceptable and the features that make it ultimately effective. For example, most British incomes policies of the 1970s have provided higher permissible proportionate wage increases for low-wage than for high-wage workers. As we argue in Section II below, the primary advantage (and perhaps cause) of such policies is that they can more readily secure the ex ante support of a majority of workers for any specified overall increase in the wage bill or inflation target. If they are effective, however, such policies also result in the narrowing of wage differentials, and this results in new problems. Some economists consider such wage equalization the inevitable cost that an economy must bear for the sake of an effective incomes policy. Others believe the arbitrary narrowing of wage differentials creates industrial relations problems that ultimately cause the incomes policy to fail. In either case, it is important to know how much equalization has been caused by the actual incomes policies and how this compares with what would have been expected from strictly adhered-to policies. In consequence, the first (and major) part of this paper contains a simple analysis of the movement of differentials in Britain in the last decade. Most incomes policies have been explicitly equalizing, but how much equalization occurred, compared with the equalization that would have occurred if the policies had been explicitly adhered to? We show that the 1973 formula of ?1 plus 4 per cent was associated with the full equalizing effect it implied, but the 1975 ?6 a week was largely submerged in other changes and was associated with less than a third of the equalization implied by the formula. Indeed, such mild equalization as occurred between 1975 and 1977 may be explained partly by the lower inflation rate, which reduced that part of the inequality of instantaneous earnings that is caused by different groups of workers settling at different times of year. Against this, higher unemployment may have been tending to increase inequality at the same time. In 1974 threshold payments were associated with an effect of about one-half of their predicted level. The main force that offset the equalizing nature of the threshold formula and the ?6 a week was, we believe, the pressure of employers bidding up the relative earnings of the more skilled workers. Having looked at the overall picture, we then examine more closely which groups have gained and lost from recent changes in relative wages, and what equalization has occurred within the different groups. The groups are defined by occupation and by collective agreement. Our general conclusion is that, however finely the labour force is subdivided, equalization between groups has played a very small role in the overall equalization and the main changes