Using administrative annual earnings data from Canada, Sweden, and the United Kingdom, the analysis finds that recent increases in top earnings have led to substantial “swimming upstream” effects, therefore accounting for differential progress in the gender pay gap across time periods and a growing share of the gap unexplained by traditional factors such as education and experience.

Since the 1980s, many industrialised countries have seen stupendous increases in top income inequality. In the United States the top 10% share of income reached 50% in 2012, exceeding early 20th century levels of inequality. These increases are concentrated in the very top incomes.

Canada and the United Kingdom follow closely behind. Their top 1% income share increased from 10 percent in 1980 to 15 percent before the Great Recession. Increases in top income shares in Continental Europe were less pronounced, with the exception of Sweden. This country also experienced a doubling of the income share of the top 1%, albeit at lower levels than the United States, going from 5 percent in 1990 to 10 percent in 2007.

The focus in this research on Canada, Sweden, and the United Kingdom is motivated by these large increases in top income inequality.

Interestingly, countries with the largest increases in top income shares are also countries that have among the highest rates of female labour force participation. For more than 20 years, women have represented more than 45% of the labour force. These countries thus have a long standing tradition of high female labour force participation. One would think that this would have opened opportunities for women to advance up the career ladder.

Yet, the numbers celebrated in terms of the percentage of women on British boards (FTSE 100) in 2015 were around 25%. In Canada, that percentage was about 21% for companies on the S&P/TSX 60 index in 2014. In the same year in Sweden, women held 29% of board seats for companies in the OMX Stockholm 30 index.

The study examines the role of changes in labour force participation, changes in the persistence of top earnings, and changes in industry and age composition as explanatory factors for changes in the gender composition of top earners. Previous studies have noted the different rates of progress over time in women's shares in top income groups, without drawing implications for the overall gender pay gap.

What are the consequences of the under-representation of women in top jobs for the overall gender pay gap? Have recent increases in top incomes led to slower progress in the narrowing of the gender pay gap and a growing unexplained (by traditional factors) share? Because increases in inequality have in large part been attributed to technological change, reducing this trend would be counterproductive. 

The findings of the research indicate that the under-representation of women in top income groups accounts for a substantial share, often a majority share, of the gender pay gap in annual earnings in the three countries under study. In light of these findings, public policies and private sector initiatives are likely to be most effective if they can improve the under-representation of women among top earners. The paper discusses evidence on the effectiveness of various such policies, highlighting there mixed effects.

Read the full paper on ScienceDirect.