The gender gap in disposable income in the UK and the role of local market dynamics: a multifocal peek

How much does gender shape people’s disposable income from various sources? And does where you live further influence the distribution of income between men and women? To explore these questions, we analysed more than three million individual consumer current account records of UK adults (18+ years of age) provided by NatWest Group (NWG). Specifically, we sought to measure the finance gap between men and women across salaries, benefits, total income, and expenditure. We situate these individuals in postcode areas derived from customers’ address details as registered with their bank to explore whether an individual’s regional context matters. Below, we present a descriptive picture of what the numbers say.

We start by comparing men and women in the full sample. Women constitute 52.4% of the sample, a slight female oversample relative to the 51.6% female representation in the adult population. Unsurprisingly the income averages split in the directions established research suggests, demonstrating a clear gender gap in individual income and expenditure.

In Table 1, we can see the salary gap is the widest of the various finance categories. Here, we employed the take-home salary as received in the bank account and averaged it over the full adult population, including employees, and adults not receiving a salary such as the unemployed, pensioners and economically inactive people. Note that only 69% of the sample has a non-zero salary income, hence the rather low average of £11,755 annually or around a £1,000 a month. This figure indicates what financial wellbeing would look like for individuals and families in the UK if people were relying only on a salary.

Looking at the gender pay gap, we find that the average woman takes home about 84 pence for every pound received by the average man, a 16% shortfall. As expected, total income narrows the gap to roughly 10%, and a big reason for this is benefits, excluding pensions that are observed separately in the data. At around £2,000 annually or £165 a month, average benefits pay can be seen to be relatively low. This figure demonstrates the indicative level of universal non-means-tested financial support individuals in the UK could receive if present disbursements were redistributed evenly. However, comparing benefits pay by gender, we see a female to male ratio of 1.84:1 which favours women. In substantive terms, this translates to around £100 per month more in benefits for the average woman in the UK compared to the average male. Figures for expenditure, capturing moneys released from the customers’ current accounts, follows total income almost exactly. This suggests that typically, cash outflow broadly mirrors the inflow within the same period.

Given the well-known geographic inequality in the UK, we explored whether where people live influences income dynamics and particularly the gender income gap story. Regional economies vary in how thick or thin they are. We observed aspects of this in two specific ways. First, we drew on data from UK Finance to measure lending density among small firms in given postcode areas. This is captured as the total outstanding volume of SME lending (in Millions of GBP) per 100,000 residents in the respective postcode areas.

Second, we also observed firm density using the number of of VAT and/or PAYE registered enterprises per 100,000 residents in a given postcode area. These two variables capture how vibrant and competitive the local business environment is. Differences in these contextual factors can be expected to have implications for income generation and distribution both across regions and between genders within a given region.

Table 2 reports the correlations between given income/ expenditure categories at the individual level and the two measures of regional market thickness. While the coefficients are small, the directions are insightful. We find that denser regions, with relatively more firms and more lending, are associated with slightly higher salaries, incomes, and spending, but lower benefits income. Here, thicker markets will arguably absorb more locals into employment and higher pay overall, reducing benefits dependency. Further, more competitive markets with greater firm and lending densities can also be expected to have a lower gender pay gap as more women will access greater, and thus more equitable, employment opportunities locally. Our analysis finds indicative support for this dynamic in the data. Specifically, looking at those receiving a salary, greater regional firm and lending densities are associated with a narrower gender pay gap.

Recall, however, that women draw relatively more from benefits compared to men and take home relatively less in salaries. How then does regional market thickness influence how these two (and other) incomes sources net off on aggregate? Our analysis suggests that despite the narrower gender salary gap, the benefit-suppressing effect of thicker markets may actually exacerbate the total gender income gap on aggregate. Salaried women n thicker markets may be in receipt of slightly higher salaries but may fare comparatively worse on average relative to their peers in thinner markets if being in thicker markets means lower access to benefits and other non-salary income. As such, a vibrant local economy may not lift all income types for all residents equally and there may be a greater penalty for women on aggregate.

Before reading too much into this, it is worth being clear about some important limitations of our exploratory analysis. First, Table 1 shows a set of averages, so tells us that the typical woman earns less in salaries and leans harder on benefits. This however hides the spread behind the averages. Table 2 also presents a set of simple pairwise correlations. With reported coefficients below 0.3, conventional rules of thumbs would consider the relationships estimated to be relatively weak even as they may be statistically significant. Additionally, prima facie associations may show apparent patterns in the data yet fail to surface the underlying drivers of the relationships.

Thus, it is possible that women that self-select into thick market postcodes, presumably in urban areas, have distinct characteristics associated with lower access to benefits and non-salary incomes. For example, recourse to public funds is much reduced for young or immigrant women with no children that are more likely to reside in thick towns and cities. Further, men in thick markets may have greater opportunities than women to accrue non-salary income, such as investment income. Taken together, it is thus possible that the narrower gender salary gap in thick market postcode areas may ultimately be dominated by the much wider gender gap in other income sources.

Overall, it would appear that the distribution of income is highly uneven and complex with the various sources of income, gender and contextual factors, such as local firm and lending density, all variously significant. Our forthcoming full report explores gender and regional variables in greater detail, although much remains unexplained. Future research, particularly drawing on disposable income in people’s bank accounts and wallets, and exploring other factors and mechanisms, is thus needed to understand issues around income distribution and financial wellbeing in society more granularly. This will help advise policy on how more equitable outcomes can be facilitated.

Samuel Mwaura, Lecturer in Entrepreneurship & Innovation

Xiaoting (Sam) Song, Doctoral Researcher in Management (Decision Analytics)