Have outcomes improved?

Let’s look at three areas where disclosures have been introduced and outcomes can be easily quantified: executive pay, pay ratios, and the gender pay gap (which is not on the chopping block under the current consultation, but is subject to a similar theory of change² and touches on similar issues of inequality and fairness).

Executive Pay

Below is the evolution of average FTSE 100 CEO pay over the past 50 years or so, adjusted for inflation, up to 2022.³ The High Pay Centre’s 2026 report tracks more recent pay, which shows pay is now close to its 2016 peak once more.⁴

Source: Pepper (2024), CEO pay in the United Kingdom, 1968-2022

Annual advisory votes took effect from 2003; binding triennial votes on remuneration policy took effect from 2014. The advisory votes seem to have caused an explosion rather than contraction in CEO wages; the binding votes may have contributed to a temporary pullback which has since rebounded (COVID-19 also played a role, likely a larger one).

If investors’ (or policymakers’) intent was to reduce or restrain executive pay via shareholder votes, the policy clearly would have been an abject failure. However, this was never an explicit priority of either group; rather, the stated intention of these votes was to incentivise executives to deliver value for shareholders, and put an end to unmerited high pay awards. In other words, pay-for-performance.

On this front, investors might point to a variety of successes - most executive plans now have short- and long-term incentive plans with a variety of metrics tied to performance targets and clawback provisions. Executives are also subject to minimum shareholding requirements.

I would like to point out two problems with the idea that Say on Pay has delivered pay-for-performance.⁵

The first is what all of this “incentivisation” has cost. The explosion in CEO pay quantum is directly attributable to Say on Pay. Incentive plans and share grants were on top of rather than replacing existing compensation. CEOs demanded higher gross pay to account for the added risk, complexity and delayed payouts. Public disclosure of pay plans led to CEO pay being benchmarked against peers, leading to a neverending cycle of the below median CEOs’ pay being increased to catch up with their higher paid peers (whose pay would then fall below the median).

These increases vastly outpace the performance these executives have delivered (see chart below). Thus it feels inaccurate, or missing the forest for the trees, to claim that what Say on Pay has achieved is “pay-for-performance” - rather, they’ve increased pay’s sensitivity to performance, which isn’t quite the same thing. The worst-performing CEOs today receive a lot more than the top performers in 2004. If you were pissed off in 2004 about executives receiving awards beyond what their performance justified, you should be even angrier today.

Source: Pepper (2024), CEO pay in the United Kingdom, 1968-2022

Then there’s the indirect costs. These votes have swelled the ranks and demanded the resources of an industry of compensation consultants, proxy advisors and investors’ governance staff. The rising quantum has also further exacerbated inequality, stoked public rage and demotivated employees.

The second problem with this story is that it’s not even clear that all of this was necessary or effective in actually incentivising CEOs. A recent survey by Edmans, Gosling and Jenter found that both investors and boards believe a CEO’s intrinsic motivation to do a good job and their personal reputation are much more important drivers for their efforts than their pay packages. Many of the motivating aspects of modern CEO pay packages (such as getting paid more than peers) would only be missed because of Say on Pay in the first place. And the FTSE 100 index has not seen dramatic performance improvements since Say on Pay was introduced.

Given the number of metrics and targets a CEO is subject to under these plans at any given moment (from an example chosen at random, I counted at least 23 metrics or targets which the CEO is currently subject to⁶), it’s hard to say how effective they can be at motivating behaviour. On the other hand, some have also found incentive plans increase risks of illegality and crowd out prosocial behaviour. It is then no wonder that investors like NBIM have called for simpler plans. That, coupled with stronger monitoring by boards and investors, could have achieved superior results without all the headaches and ££££.

Say on Pay has had the most time of the measures discussed here to achieve its stated goals, and shareholders have been granted extensive rights to shape outcomes in their desired direction. The results have not been exemplary.

Pay Ratios

For all of the attention and ink spilled over executive compensation (to which I have reluctantly just added), it can be a bit of a distraction. From a public interest angle, executive compensation is a rough proxy invoked in conversations and commiserations over inequality, and who shares in the benefits of wealth that was collectively created.⁷ In recognition of that, the UK government introduced a requirement in 2018 for companies to report against the ratio of CEO pay to a company’s median employee (and employees at the 25th and 75th percentiles).

Fortunately, compared with Say on Pay, there is less ambiguity about the goal of ratios: bring them down. While pay ratios dropped substantially in 2020, this was a COVID-19 blip, reversed the following year; ratios stayed largely flat through 2024. Not captured in the below graph is early data from 2025, which suggests ratios are rising and may reach record levels: the median FTSE 100 CEO to lower quartile employee ratio was 127:1; the median employee ratio was 95:1.

Source: High Pay Centre (2025), Five Years of pay ratio disclosures

Gender Pay Gap

Gender Pay Gap reporting commenced in 2018 for companies with over 250 employees. Again, its stated aim was clear: the gender pay gap was closing too slowly, and the government wanted to accelerate progress.

The evidence here is mixed. Below is ONS tracking of the gender pay gap for the past 30 years.

Source: Annual Survey of Hours and Earnings (ASHE) from the Office for National Statistics

There appears to be a substantial drop occurring in 2020, with the gap falling from 17.4% to 14.9%. It could be tempting to ascribe that to the impact of COVID-19 and furloughs, but according to the ONS, “furloughing had a small impact on the gender pay gap, and the majority of the reduction in the gap is because of underlying changes in pay.”⁸

Attributing this fall to the influence of GPG reporting, though, seems even more unlikely. As you can see from the below chart, companies which were subject to reporting requirements (250+ employees) actually saw less reduction in their pay gap than those who weren’t.

Source: Office for National Statistics – Annual Survey of Hours and Earnings (ASHE)

The study with the most positive evidence of GPG reporting efficacy found that it narrowed the gap at firms just over the 250-employee threshold by 19% in the 2 years after reporting was introduced. This was achieved by reducing men’s pay growth rather than improving women’s pay. Given the size of companies where this effect was observed, the improvement may have been driven by internal forces triggered by revelation of the gap, rather than external pressure.⁹ Indeed, FTSE 100 companies - which are subject to greater public scrutiny as well as investor stewardship - have a larger GPG than the UK’s average (15% vs 12.8% in 2025).

* * *

Perhaps more worrying than the fact that these approaches have not delivered against their stated goals, is a failure to learn from their shortcomings. The unconvincing evidence on the impact of gender pay gap reporting has not deterred the creation in September of a new campaign for ethnic pay gap reporting. And then there was the campaign to introduce annual “Say on Climate” votes, without seemingly any sense of irony of the failures of the Say on Pay votes the concept was based upon.¹⁰ This habit of defending or even extending measures which have failed to deliver results or accountability is endemic.

Of course, a simple analysis such as this does not account for many nuances - for example, could outcomes without these disclosures have been even worse? More importantly, are these even appropriate measures against which to measure progress? They all gesture at the underlying systemic issue of inequality, though they are mostly crude proxies.¹¹

To varying degrees for each of these measures, part of the reason for the poor outcomes has been a weak or unclear theory of change underpinning them. Successive governments have focused on the easy work of surfacing inequalities through disclosure, rather than considering what are the drivers and who has the incentives and power to change them.¹²

Another lesson from this though is that when it comes to tackling the challenges of inequality and corporate accountability, investors may not be a very reliable partner.