Are UK labour markets polarising?

In recent years, there has been a growing acceptance of two related phenomena: routinisation and polarisation. However, whilst existing evidence has shown what has happened to occupational structure in the UK, little has been done to look at the effects these processes have had on the resulting wage distributions. In particular, there has not been a full consideration of the implications of within-group effects, where new employees in growing occupations have different productive characteristics compared to existing employees. This paper uses a new method, proposed by Firpo, Fortin and Lemieux (2009), which allows for the decomposition of all distributional statistics into the composition and wage effects of individual explanatory variables, much like the well-known Blinder-Oaxaca decomposition of the mean. Using the Family Expenditure Survey between 1987 and 2001, our results demonstrate the importance of polarisation for the changing distribution of wages in the UK. First, there are effects from the composition of occupations, however, these are only really noticeable at the top of the wage distribution, and in many cases are smaller than the effects of deunionisation and the expansion of further and higher education. Second, there is evidence to suggest that within-group effects are important and may dominate other wage effects that could cause the 'polarisation of wages'. The final distribution is one where top wages have grown more than middle wages, which in turn have grown more than bottom wages, leading to more inequality. However, routinisation and polarisation have played only a small part in these changes. Third, this methodology reveals a number of other interesting results. Declining gender pay gaps are observed at the bottom and middle of the distribution, but not at the top. Returns to experience have also increased in the middle and, particularly, at the top, suggesting the increasing value of on-the-job training, informal training and soft skills