The Gender Pay Gap in the UK: Evidence from the UKHLS

Executive summary
This report provides an analysis of the main predictors of the gender pay gap (the gap between men and women’s average hourly earnings) using the latest waves of the British Household Panel Survey (BHPS) and the United Kingdom Household Longitudinal Survey (UKHLS) relating to 2014/2015. We also provide additional tests using 2009/2010 data. The findings delivered in the report can be considered alongside the findings of our earlier research on the topic (Olsen et al. 2010a, b), with a similar research strategy and research method applied.
The report contributes to ongoing efforts to monitor the Gender Pay Gap (GPG) in an effort to approach parity in pay for men and women. ONS estimates suggest a median gender pay gap of 19.3% in 2015 for the UK as a whole, in its analysis of the Annual Survey of Hours and Earnings (ASHE) data. This report uses alternative data sources: the British Household Panel Survey (BHPS) and the United Kingdom Household Longitudinal Survey (UKHLS), which provide additional information necessary for our statistical analyses. All figures within this report use these data unless otherwise stated. This has led to some divergence in our estimates of the GPG than those generated using ASHE. Further differences can be found in our estimates of the GPG at the mean, whilst the ONS tends to examine the GPG at the median1. ONS estimates of the gender pay gap at the median suggest a larger pay gap than their estimates at the mean (ONS, 2016a).
A review of our sample statistics underscores the ongoing differences between women and men in their working strategies. Our sample statistics show that 81% of part-time workers were women in 2014/2015 and that 56% of full-time workers were men. So while an increasingly large proportion of women are in full-time employment, part-time employment, in particular, is deeply gendered. We also find that women with children continue to pursue part-time employment over full-time employment. While 38% of our female sample has children overall, the percentage rises to 51% for those who work part-time. There is no similar effect for men.
Turning our attention to the job characteristics of those in part-time and full-time employment, we note similar percentages of women and men in temporary contracts overall, and that the percentages of those on temporary contracts are higher for workers in part-time posts. The discrepancy is also greatest for male part-time workers: 8% of women who work part-time are on temporary contracts compared to 18% of men who work part-time (Table 2).
Key to our analysis is the effect of labour market and working history on wages and, therefore, the gender pay gap. We find that men in full-time employment have longer full-time work-histories than women in full-time employment, 17.8 years compared to 13.2 years, and that men in full-time jobs tend to have had little to no exposure to part-time employment nor to unpaid care work in their work history. We also note the strong difference in work history between men in full-time jobs and those in part-time jobs: men in part-time jobs have almost ten years less full-time work history than men in full-time jobs, and considerable prior experience of part-time work.
We establish a decline in the gender pay gap; using data from 2004-2007 Olsen et al. (2010a) found a mean pay gap of 19%. In this report, using more recent data, we establish a pay gap of 14.4% in 2009/2010 and a pay gap of 13.4% in 2014/2015. While the most recent declines are welcome they also need to be understood within the context of declining real wages.
Table 3 presents calculations of gross hourly pay by gender and working-time status for 2014/2015. It reveals an average hourly pay of £10.47 for women in the UK and of £12.09 for men in 2014/2015, and a gross wage difference per hour of £1.62.
In this report, we decompose the predictors of the gender pay gap by key covariates, referring to those that increase the gender pay gap as ‘drivers’ and those that decrease it as ‘protective factors’. We established the following drivers of the pay gap in the UK using 2014/2015 data: The biggest drivers of the gender pay gap in 2014/2015 concern male-female differentials in labour market history, accounting for 56% of the drivers of the gender pay gap. We find that women earn £0.91 less per hour compared to men because they have fewer years of full-time work in their work history and because they have more years of unpaid care work in their work history compared to men. The next biggest factor concerns unobserved components of the gender pay gap, which includes all observed and unobserved characteristics systematically associated with being female, which accounts for 35% of the drivers behind the pay gap. Here we can say that women earn £0.57 less per hour than men because of these unobserved factors (Figure 3). While we cannot definitively say what these observed and unobserved factors are it is likely to be a combination of discriminatory behaviour against women and ongoing differentials in gendered behaviour between men and women. Pay differentials arising from industrial sector and occupational segregation are the next biggest drivers, accounting for 29% and 19% respectively.
We establish some significant changes in the protective features of the gender pay gap in the UK using 2014/2015 data. While institutional features, including public sector employment, continue to be protective against the pay gap, contrary to our earlier report we have found part-time employment to be a protective factor of the gender pay gap. We attribute this change, firstly, to the rise in male part-time employment which is of poor quality, and secondly to increased proportions of
female ‘retention part-time workers’
2. So, while previous research has found that many women have had to occupationally downgrade in pursuit of reduced-hours posts (Connolly and Gregory, 2008), recent changes in policy may be limiting such flows to lower-calibre positions. Though many part-time jobs continue to be of poor quality - part-time jobs are less likely to be permanent or unionised - there are also many part-time workers employed in the public sector, a sector typically regarded to have preferential working conditions over others. For our sample, we find 39% of women in part-time jobs work in the public sector compared to 21% of women in full-time jobs. We find a pay premium for public sector employment when compared to private sector employment in the wage regressions estimated for this report, and have found a similar dynamic in our earlier research (Olsen et al 2010a). In addition to compositional differences as a result of the right to request flexible working, there has also been an increase in the proportion of men employed in part-time jobs since the recent recession, from 9.7% in 2007 to 11.9% in 2015 (OECD, 2016: 227-228), and this increase in male part-time employment has coincided with an increase in the percentage of men who are involuntarily employed in such posts.