Joel Suss, Marilena Angeli and Peter Eckley
Diversity has risen up the agendas of businesses, regulators, and governments in recent years. How diverse are the upper echelons of banks and building societies in the UK? We answer this question in a recent paper using a unique data on the most senior employees for the last 20 years.
The data
Arguments for increasing diversity within banks usually stem from at least one of two perspectives: equity and efficiency. Our paper addresses both of these perspectives. First, we establish the facts and trends of diversity in UK banks over the last two decades. Second, we examine whether more diverse banks have better outcomes.
We bring together information on all approved control and senior manager functions (2001–20) – ie all senior manager, executive and oversight roles which require regulatory approval. We build a data set of the identity characteristics of senior managers for gender, age and nationality diversity in all PRA-regulated banks over the last two decades (total of 181).
The good, the bad and the ugly
The most interesting trends are in relation to gender diversity, so we will focus on these findings here.
The good news is that, overall, gender diversity has improved in UK banks over the last two decades. The bad news is that this is an improvement upon a very low base and that progress has been very slow. Chart 1 shows this trend, plotting the proportion of women in senior positions per quarter. Just under 10% of positions were filled by women at the end of 2001, while this figure is now just below 20%. Otherwise stated, just one in five senior positions are held by women in UK banks.
Chart 1: Proportion of senior positions filled by women, 2001–20
Things get uglier when we decompose the trend by role. Here we find evidence of a ‘glass ceiling’ operating in UK banks – the most senior roles have seen the least and slowest progress, whereas the more junior positions have seen faster improvement. Chart 2 shows the trends for four categories of functions: CEOs; executive; oversight (non-executive directors); and risk/compliance/audit personnel. The relatively junior risk/compliance/audit positions increased at the fastest relative pace and CEOs at the slowest relative pace. Indeed, just under 10% of CEOs were women by the end of the sample time period. The breaks in the trend lines by function are due to changes in the regulatory regime during our period of study, which led to changes in some of the populations captured.
Lastly, we show that close to 20% of all banks did not have any female representation in authorised positions at the end of 2020, which is down from approximately 40% in 2001.
Chart 2: Evidence of a ‘glass ceiling’ in UK banks – proportion of women by function
Does diversity affect bank outcomes?
This is a tough question to answer with the data we have. In an ideal research set-up, we would randomly allocate diversity across banks. That way we would not have to worry about all the other differences between banks because they would balance across those that were ‘treated’ with diversity and those that weren’t due to randomisation. The real world is messier – it is very hard to account for all the differences that might exist between firms to isolate the causal effect of diversity.
We therefore try to approximate an experimental design using Coarsened Exact Matching (CEM). This approach creates a synthetic treatment and control group by finding banks that are exactly or approximately the same across all dimensions we can measure (and deemed relevant) while differing only in terms of diversity. Using this approach, we find that gender diversity is associated with improved financial performance (measured by the return on assets ratio) and reduced riskiness (measured by the Z-score distance to default measure). For example, we find that a one standard deviation increase in the percentage of females in authorised positions (an increase of 12 percentage points) is associated with an expected reduction in risk of failure by around 9 percentage points.
While the CEM approach provides evidence that diversity is associated with better outcomes (ie better returns and lower riskiness), when we examine whether changes in diversity over time within firms affects outcomes (employing general method of moments regression models), we find a null result.
Our findings are therefore mixed. While we do the best we can to identify the causal relationship between identity diversity and firm outcomes given the data available to us, these findings come with a reliability warning. In the absence of some form of natural experiment, we are unable to control for all relevant factors which may affect both diversity and outcomes, and so we stop short of claiming that these findings are causal in nature.
What next?
While our results are mixed and come with caveats around causality, taken together they suggest the Prudential Regulation Authority (PRA) is justified in encouraging greater diversity in UK banks, both from an equity and efficiency perspective.
There is much more to investigate. One avenue for future research is to expand the range of diversity variables available for analysis to those that are potentially more closely related to cognitive diversity, for example diversity of educational and professional experience. This aspect is important because identity diversity might matter for outcomes only insofar as it leads to cognitive diversity.
The limitations of our data have several lessons in terms of the reporter requirements the PRA places on regulated firms. To better understand the facts of diversity and its relationship with firm outcomes, more data on demographic characteristics are needed. Notably, the PRA do not currently gather data on important demographic characteristics, eg ethnicity of individuals in regulated firms. A recent Discussion Paper, however, sets out possible ways to remediate this, and whether metrics for inclusions would also be helpful.
Joel Suss works in the Bank’s Advanced Analytics Division, Marilena Angeli works in the Bank’s Structural Economics Division and Peter Eckley works in the Bank’s Data Strategy Implementation Division.
If you want to get in touch, please email us at bankunderground@bankofengland.co.uk or leave a comment below.
Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.
