The average yearly pay for a CEO of a FTSE 100 company is 4.3 million compared to 26,500 for the average UK worker. What are the implications of such an unequal distribution of wealth? Fat cats and unequal wage distribution By Benjamin Radoc Postgraduate, Second Prize In theory, wage represents the marginal contribution of labour to total output. A very high wage implies exceptional talent commonly observed amongst athletes and other celebrities, whilst a not-so-high pay signals not-so-exceptional skill. This pattern in compensation has also been observed amongst employees in listed companies: there is a wide wage disparity between executives and other employees. The gap has widened over the years and could not plausibly be explained by the market for skill. In this essay, I describe the pattern in executive compensation and its impact on wage inequality and workers effort level. Are chief executive officers (CEOs) paid too much? CEO earnings over the past years increased 14% annually, bringing the average yearly pay for a CEO of a FTSE 100 company at 4.3 million. The pace in the growth of CEO compensation and the 9% annual expansion in corporate profits do not closely match, but the pattern appears to approximate the pay-for-performance structure (Barty and Jones 2012). Under this scheme, actual compensation is aligned with pre-determined performance metrics so that executive bonuses serve as shareholders rewards to management for the company s financial achievements. Executive compensation consists of a fixed component (non-performance-based pay such as salary, pension and perquisites) and a variable component (performancebased compensation in the form of stock options, bonuses, deferred bonuses and long-term incentive payouts [LTIP]). Figure 1 shows that whilst the salary component has been relatively steady over the years, additional compensation in the form of bonuses, deferred bonuses and LTIP have risen dramatically.
Figure 1. Average Remuneration of FTSE 100 CEOs Sources: Barty and Jones (2012); The Manifest/MM&K Executive Director Total Remuneration Survey 2011 Gordon (2014) argued that the pattern in executive compensation can be explained by changes in tax policy and the weakness in corporate regulation, more than by stock market conditions. Given a high income tax rate and lower tax rate on capital gains, stock options provided a good alternative to reduce taxable income. Therefore, whilst the salary component barely increased over the years, CEO compensation has been significantly padded by other forms of remuneration that are subject to lower taxation. What is the rationale behind generous bonuses and payouts? Market for skills has been offered as an explanation for CEOs high earnings. This pattern in pay is prevalent in both financial and non-financial firms and potentially signals quality of skill amongst CEOs. However, Bivens and Mishel (2013) pointed out that the sharp difference in CEO compensation vis-à-vis other highly paid workers in the same industry has widened over the years. Market for skills cannot plausibly explain this gap: what extraordinary set of skills could a CEO possibly possess to explain any discontinuity in the distribution of skills, that none of the other executives in the organisation come close? Bivens and Mishel (2013) argued that high CEO pay reflects rents rather than executives individual contribution to the firm s total productivity. In this context, rent refers to income received in excess of what was needed to induce the person to supply labour and capital to these respective markets (p.62). Lower total CEO compensation would not have deterred executives from taking on the job, so that the excess compensation could have been redistributed elsewhere where incentives may have been improved (e.g. better incentives for non-managerial employees to encourage higher productivity). The presence of rents implies that very highly paid CEOs are not paid-for-performance, but arises from institutional arrangements that accommodate rent-seeking. For example, in setting CEO compensation, an arbitrary peer group is constructed by the board s compensation committee as basis to benchmark top executive salaries. Since CEOs are perceived as above average, the salary above the mean of that peer group is offered to the CEO.
CEOs ought to be rewarded for the marginal value they put in that allows the firm achieve its corporate targets. However, CEO compensation has been observed to be positively correlated with overall stock market performance. In a bull market, executives have the incentive to exercise their stock options that provide them the opportunity to purchase shares below the market price. Whilst options awarded to executives are designed to align shareholder interests with management incentives, it appears that CEOs are paid for luck and not necessarily paid for their performance (Mishel and Davis 2014). As counter-measures, a clawback provision and longer deferral of bonus payouts could create symmetry in CEO compensation in periods of success or failure, and may effectively address the principal-agent problem between executives and shareholders. A clawback precludes an executive from benefitting from past profits if certain corporate performance indicators are not met. This potentially deters a CEO from taking excessive risks and provides an incentive for other senior executives to challenge corporate strategy especially if the clawback also applies to them. Meanwhile, a longer deferral of at least five years with a straight-line vesting of benefits means that a significant amount of deferred bonuses are available for clawback in case of corporate failure. Also, a longer bonus deferral of five years approximates the length of a business cycle and helps reduce the impact of luck in CEO compensation (Barty and Jones 2012). Impact on overall wage inequality Mishel and Davis (2014) showed that the high level of CEO compensation has spilledover to other executives and managers. Growth in executive compensation has been observed for CEOs and other highly-paid executives. This spill-over is reflected in Figure 2. Whilst the real hourly income of workers in the 50 th percentile rose by 101% from 1975 to 2013, the wage of employees in the 98 th and 99 th percentile climbed more quickly by 167% and 189%, respectively. Meanwhile, the ratio of per hour compensation of workers in the 99 th percentile vis-à-vis median employees rose from 3.2 in 1975 to 4.6 in 2013, which clearly indicates that the gap in compensation has widened significantly over time. Figure 2. Real hourly earnings growth (excluding overtime) for full-time UK employees, 1975-2013
% earnings percentile Data source: http://www.ons.gov.uk/ons/rel/lmac/uk-wages-over-the-past-four- decades/2014/rep---uk-wages-over-the-past-four-decades.html#tab-uk-wages- Over-the-Past-Four-Decades 450 400 350 300 250 200 150 100 50 0 Figure 3. CEO-to-worker compensation ratio in the 1965 1973 1978 1989 1995 2000 2007 2009 2010 2011 2012 2013 Data source: Mishel and Davis (2014) The trend in the disparity in compensation between CEOs and typical workers in America is depicted in Figure 3. Total compensation shown here includes salary, bonuses, stock grants, and LTIP amongst the top 350 US firms ranked by sales from 1965 to 2013. The difference in compensation declined during the recent economic crisis, the gap resumed its incline at the onset of economic recovery. As economic recovery set in, CEO total compensation increased but the wages of typical workers (i.e. private-sector production/nonsupervisory employees) in the US have remained at 2009 levels. CEO compensation also rose faster than the income of other high-wage earners, whilst the wage gap between university degree holders and high school graduates did not widen as much (Mishel and Davis 2014). This difference in the compensation pattern is inconsistent with the market for skills story. Given the small number of CEOs relative to the entire labour force, does excessive executive compensation significantly affect overall income inequality? Joyce and Sibieta (2013) showed that in any given year, excluding the top income earners in the
calculations lowers the Gini coefficient. 1 For instance, had the top 1% of income earners (including CEOs) been excluded from the population, the Gini coefficient in 2010-2011 would have been lower at 0.30 vis-à-vis actual Gini of 0.34. Figure 4: Great Britain s Gini coefficient excluding various top income groups Source: Joyce and Sibieta 2013, page 184. Note: The Gini coefficient was calculated using incomes gross of housing costs The overcompensation of CEO does not only represent rents that could have been redistributed to other parts of the economy, but has precluded economies from achieving a more equal income distribution. These outcomes clearly provide a compelling case for taxation reform (to achieve redistribution objectives) and better controls in corporate governance (to reduce rents). Does wage disparity affect workers effort levels? Gordon (2014) argued that the stark inequality in compensation between top executives and median workers is a symbolic marker of social norms in our winnertake-all economy (p.1). Whilst a winner-take-all structure may be designed to attract high-quality talent, it may deter the participation of other members in the population. In a winner-take-all pay structure, incentives follow a step function that offers significantly higher payoffs to a winner relative to the next best performer, and then offers lower incentives for all other participants (Sheremeta, Masters, and Cason 2012). This mimics the pay structure in an organisation with a very highly paid CEO (the winner), well-paid managers (the next best performers), and rank-and-file employees (other players). Cason, Masters, Sheremeta (2010) compared the impact of entry and participation in a winner-take-all tournament with a single fixed prize vis-à-vis a proportional- 1 The Gini coefficient ranges from zero (perfect equality) to one (perfect inequality) and measures the dispersion in income within a group.
payment structure where the total pie is divided amongst contestants based on their contribution to the total outcome. Despite skill heterogeneity amongst players, proportional payoff (relative to each player s contribution) did not discourage the entry of weaker participants. In addition, the payoff structure had no effect on the effort level of stronger participants. Meanwhile, Sheremeta, Masters, and Cason (2012) showed that relative to a lottery (random payment) and the proportional-payment structure, the simple winner-takeall contest generates the highest effort levels (i.e. higher costs on the part of players) which translate to very low net payoff to players. Whilst a winner-take-all environment where all participants act on best-effort mode benefits the organisation as whole, this may not be sustainable if compensation does not catch up with effort level. High performing individuals who feel underpaid relative to their responsibilities are likely to leave the organisation to seek employment elsewhere. There is caveat to the applicability of experimental results describe above. For instance, in a winner-take-all game, a reader may assume that all players who participate in the task aim for the top plum. However, in a corporate organisation, only a handful would actually desire a senior management post, more so for the CEO post. Also, the hierarchy in an organisation, particularly the difference in proximity between rank-and-file employees and managers, and rank-and-file employees and senior management, is not captured in the tournaments described. A typical employee may be privy to the difference in her level of responsibilities and the immediate supervisor. However, self-comparison vis-à-vis the CEO may be less likely given proximity and lack of previous interaction. Therefore, the gap in her compensation compared to the manager may matter more than the CEO s compensation. Nonetheless, given the spill-over of CEO compensation to other executives and managers (Mishel and Davis 2014), the implications of the results in winner-take-all tournaments remain relevant. Should CEO compensation be reduced? High rents in CEO compensation are inefficient. Rents do not only affect the incentive structure amongst employees within a corporation, but also unduly penalise shareholders. Whilst the intention of high CEO compensation is to attract high-quality talent, the mechanism does not adequately address the principal-agent problem between management and shareholders, and the call for reforms to CEO compensation is in order, whilst ensuring that changes do not discourage the entry of high-quality talent.
References Barty, James and Ben Jones (2012). Executive Compensation Rewards for Success Not Failure. Policy Exchange. http://www.policyexchange.org.uk/images/publications/executive%20compensatio n.pdf Bivens, Josh and Lawrence Mishel (2013). The Pay of Corporate Executives and Financial Professionals as Evidence of Rents in Top 1 Percent Incomes. Journal of Economic Perspectives, 27(3): 57-78. Cason, Timothy, William A. Masters, Roman M. Sheremeta (2010). Entry into winner take all and proportional-prize contests: An experimental study. Journal of Public Economics, 94: 604-611. Gordon, Colin (2014). Fatter Cats: Executive Pay and American Inequality. http://www.dissentmagazine.org/online_articles/fatter-cats-executive-pay-andamerican-inequality Joyce, Robert and Luke Sibieta (2013). An Assessment of Labour s record on income inequality and poverty. Oxford Review of Economic Policy. http://oxrep.oxfordjournals.org/content/29/1/178.short Mishel, Lawrence and Alyssa Davis (2014). CEO Pay Continues to Rise as Typical Workers Are Paid Less. http://www.epi.org/publication/ceo-pay-continues-to-rise/ Sheremeta, Roman M., William A. Masters, and Timothy N. Cason (2012). Winner- Take-All and Proportional Prize Contests: Theory and Experimental Results. Manuscript.