Much has been said both for and against using incentives to drive business results—be it revenue and profit, or market share. Despite this, there is surprisingly little research, particularly in Asia, to guide practitioners on how different pay models affect decisions that lead to better company performance. CEOs, Remuneration Committees (RemCos), HR executives and executive compensation consultants have been searching for the ‘Holy Grail’ of pay-for-performance for a long time. And yet, many questions remain unanswered. For instance: Do various types of Long-Term Incentives (LTIs) affect the achievement of business results differently? By LTIs, we are referring to incentives for performance over periods longer than one year, which are usually linked to awards in shares. Which type of LTI provides the greatest incentive to performance? Does changing the pay mix (that is, the relative weight of fixed pay, performance bonus and LTIs as a percentage of total pay) yield different results regarding performance? Which pay mix provides the greatest incentive to performance?
Besides the traditional pay-for-performance model, there are three other pay-for-performance alternatives: Tournament, Membership and Bonding models. Are there differences in how these models affect performance? Which pay-for-performance model provides the greatest incentive to improve business results? Are there differences in performance if the incentives provided are team-based or individual-based, or if they are offered in combination? And which approach—team, individual or a combination—provides the greatest incentive to improve performance?
There are four different perspectives when addressing pay-for-performance in organisations:
• CEOs are more concerned about fairness and recognition.
• HR heads look mostly into market competitiveness and retention.
• Executive pay consultants are concerned with design features and buy-in.
Executive compensation consultants are often tasked by RemCos of corporate boards to advise on developing an approach to pay their company executives to achieve better business performance. However, when it comes to honing in on a pay philosophy, these same board members fall back on the ‘evidence’ of market data, which shows how and how much other companies of similar size or from the same industry pay their executives, but does not provide information on whether their competitors’ pay plans are effective in driving results.
For RemCos, pay should not vary so much from year to year. And yet, an executive pay consultant in Hong Kong summarised this state of affairs when he said, “Shareholders have their favourite plan types; some like specific metrics, or specific vehicles and formulae. They tend to follow similar designs to competitor proxies.” He went on to add, “RemCos don’t always have the knowledge to make these decisions.”1
At best, paying the same as your competitors yields the same degree effective vehicles for delivering pay, calculating competitive levels of pay at various levels of performance, ensuring compliance with regulations, and staying close to existing parameters to avoid conflicts at the Annual General Meeting of shareholders.
Meanwhile, there are often other considerations besides pay-for-performance that are frequently ignored. A Singapore RemCo chair explained it this way, “Pay [for performance] is not the only thing, retention is also anissue. The CEO might say, ‘If you don’t pay me at the 90th percentile I won’t work so hard,’ to which we would answer, ‘What part of your job will you not do?’” To the CEOs, the view is starkly different. One of them put it this way: “[Pay] can act as a de-motivator so [best to] aim for satisfaction/motivation as much as you can.” The HR executives focus on pay more to attract and retain executives. One of them indicated, “We want to retain first, then motivate to perform. This has to be balanced with not encouraging risky behaviour.” HR executives typically rely on existing market benchmarks, some of which are of better quality than others, in order to tell their RemCos and bosses that they do not over/under pay.
In support of the idea that retention is a big part of executive pay plan design and implementation, researchers found that when a CEO leaves a firm, compensation for those left behind increases by an average of 46 percent.2 From the perspective of executive pay advisors, it is unclear if incentive pay plans work as intended. One of them expressed doubts in terms of trying to create homogeneous plans to motivate what is, in effect, a heterogeneous population. “‘Designing for the herd’ is the safest route both as a consultant and as a RemCo chair.”
Long-term versus short-term incentives
Asian companies constantly debate about the role that LTIs should play in the overall employee pay mix, alongside base salary and annual bonuses. This debate is spurred by the various viewpoints from the different stakeholders involved. For instance, boards have an interest in maximising return on pay and minimising risk to their investment. This can be achieved by limiting the amount of fixed pay (which is a cost regardless of performance) and tying as much pay as possible to financial KPIs (such as total shareholder return), particularly over longer periods of time. In this way, if the shareholder wins, he is willing to share of motivation of the company’s executives vis-a-vis the competition, rather than gaining any competitive advantage. Instead of searching for a different approach that might lead to better performance, efforts are only channelled into determining key performance indicators (KPIs), deciding the targets and the timeframe for their achievement, providing cash flow/tax with management, but if he is losing, management is losing as well. A Hong Kong RemCo chair stated it this way, “Without short-term there is no long-term, as decisions made today have an impact two or three years from now. We must balance the two, but it’s very difficult. Share price is often beyond management control.” Consultants agree that there is a need to see if the plan that was implemented moved the long-term financial needle beyond the ‘rising tide’ of the market and industry.
Often, management observes that share price movement (the main component behind total shareholder return) is not always linked to the performance of the company, as industry and market factors play a role as well. For CEOs, pay plans should be, “aligned with the creation of shareholder value, but in things [you] can control. If you provide consistent performance, the market should reward you in time. But it is ‘over time’…”
This divergence in viewpoints sets up a classic confrontation in the boardroom, and consultants are often called in to provide a way to mediate this dispute (refer to Figure 1).
Individual versus team incentives
Companies also often struggle to find a balance between individual incentives and team incentives as a motivator of performance. Here, again, there is a divergence of viewpoints. Whereas for the CEO it is straightforward, as the company’s performance is also his or her personal performance, as soon as you go down one level, distinctions start to emerge between an individual’s performance and the overall team performance. In the words of a global CEO headquartered in Asia, “The more senior, the more team [based], but there needs to be a mix.” However, standard practice differs from this notion: At the executive level there are often no team incentives except for when the overall annual bonus pool is based on company results for short-term incentives, and increases in share price are considered an all-of-company result, for long-term incentives.
Both of these approaches have detractors as well that hold the view that there is “no real way to differentiate performance” and thus maintain that team results should be balanced with individual performance. The corollary to this line of reasoning is that balance is important as, in their view, having only individual incentives could mean too much ‘kill all’ behaviour, and too many team incentives may foster freeloaders.
For RemCos, the view is that only the company’s performance counts, and the entire senior management is there to ensure overall results are achieved. And yet, while agreeing with the company-based bonus pool, a Singapore RemCo Chair provided this caveat, “The pool is company-based and then [we] need to carve it out by individual. But, all individual Balanced Score Cards have a big ‘fudge factor’ and, in fact, all RemCos fudge a little bit.”
For individual executives, the view is often that if they have performed their duties and achieved their individual results, this should be rewarded regardless of the company’s outcomes. HR and compensation professionals wrestle with this problem constantly, as executives with great performance but low overall pay are at risk of leaving the company and joining its competitors.
Boards, and, more recently, regulators, worry about the effects of different long-term vehicles as drivers of performance. In the minds of many, linking pay to performance has led management teams to seek strategies involving high risks and high potential returns that have, at times, resulted in very negative effects. Asian RemCo Chairs fear that executives only see upside and entitlement with very little clawback. One RemCo Chair illustrated this point by recalling a recent instance involving GE/Volkswagen: “The guys that caused the problem all collected their bonuses and the current ones are now punished. This is a huge issue of pay-for-performance. But what is the alternative?” As a result, Boards worry a lot about shareholder activism, defensibility and governance, so they tend to take fewer risks on how to pay executives.
Finally, HR heads all over the world search for the compensation plan that will allow their companies to find a pay-for-performance model that leads to higher attraction, retention and motivation of executives, and to hopefully achieve better business results. In an ideal set of circumstances, HR would like to come up with pay programmes that meet the objectives of the shareholders and executives. However, quite often, HR is under pressure to find and keep executives, and not so much for delivering performance, which is deemed to be the responsibility of the executives and the CEO. Thus, HR executives have an inherent bias to side with arguments for ‘market data-based’ compensation to discharge their attract-retain-motivate duties. They argue that there is competition for talent, and firms need to pay to attract and retain the talent they seek. Nevertheless, an executive pay consultant warns: “No one [has ever] benchmarked themselves to a position of market leadership.”
Need for a research-based approach
The discrepancies in positions and approaches to the near-dogma of pay-for-performance are troubling. There is a definite need in Asia to apply scientific rigour to the question of how best to design executive pay programmes to determine if they work at all, and which approach leads to better results. In short, there is a need to move from ‘market-based’ to ‘research-based’ advice.
Academic literature has an abundance of evidence on pay-for-performance. However, the evidence is inconclusive on two counts: The first is the debate on whether money is an extrinsic motivator at all, or if individuals are intrinsically motivated. The second unresolved element in literature is the impact that different pay mechanisms have on performance, if any.
Agency theory, expectancy theory and equity theory all would support the view that incentive pay can motivate improved performance. However, other researchers have found the data to be inconclusive, and even contradictory.3 This can be interpreted as evidence that, at best, incentive pay only works to increase performance under specific circumstances, such as repetitive tasks where working consistently and faster can lead to better results. Proponents of this view would then argue that incentives are not effective in driving performance based on decision-making. In the words of an HR executive, “No amount of incentive pay can make people smarter!” Proponents of intrinsic motivation argue that incentives can even be counterproductive, which is in direct opposition to proponents of these theories.
To complicate matters further, prospect theory—which states that people make decisions based on the potential value of losses and gains rather than final outcomes, and these are evaluated with certain heuristics which bias the decisions—supports the idea that incentives impact performance in the shape of an ‘S’ curve.4 That is to say, the further ahead on results, the less likely executives would be to exert additional effort in pursuing better results. Conversely, the further behind executives are on results, the more likely they will take big risks driven by incentive pay. Finally, while there is a general consistency in the literature regarding team incentives working to drive performance, there are very few data points comparing individual incentives to team incentives as drivers of performance.
Following research by Ederer & Manso, we created an experiment in which 510 subjects were tasked with running a business for 20 periods (iterations).5 For each period, subjects had to decide on the Four “Ps” (Placement, Packaging, Product and Price) to maximise profits. In making each of their 20 choices, subjects could fine-tune the current operation (as inherited from the prior manager) by adjusting each of the four parameters, or they could devise a new strategy by making decisions on each of the four parameters. Subjects were paid under 15 different pay conditions to determine which, if any, made any difference to their performance. Additionally, the willingness of participants to modify their strategies was also observed.
The results of our study revealed that differences in pay conditions do not significantly impact performance. Equally, no single approach (e.g., different long-term incentive vehicles, varying pay mixes, any of the three alternative pay-for-performance approaches, one of the team incentive designs) has a significant performance advantage either. However, we do find that, jointly, the team approaches we designed do significantly outperform all the individual approaches.
An analysis of the patterns of responses showed that nearly half of all participants stopped seeking more profitable alternatives when they were ahead in the game. Perhaps more importantly, a higher percentage of participants changed their strategy or ‘took a risk’ when behind. More surprising is the fact that nearly 20 percent of all participants did both during the span of the 20 periods of the game in the experiment. That is to say, the same person took a risk and was also conservative, depending on where their results were at any point in the game. The data also suggests that the higher the level of profit attained, and the earlier the run in the game to achieve that profit, the higher the probability of a prospect theory effect on the risk aversion side. But, for those falling behind, they began ditching their current formula and developing new strategies.
Further analysis of the data reveals that, for all individual conditions, the level of profit at which individuals risk when behind or stop when ahead is equivalent. These results suggest that the perceived risk is not inherent in any of the pay conditions. The implication of this result is that being risk-prone or risk-averse can be construed to be circumstantial rather than strictly a personality trait. In the words of an executive pay consultant, “the good thing, and the bad thing, about incentives, is that they drive behaviour!” In this case, risking when behind but also being conservative when ahead differs from what is normally expected, that incentives always drive behaviour in the same way.
Other interesting findings derived from the experiment were that LTI approaches were more efficient when looking at the period 20 results, whereas the three performance models’ approaches were more efficient when considering average results during the 20 periods. This result has implications when designing executive incentive plans. Perhaps, if there is no added performance increase that can be expected from any plan design (except for team incentives), then simple plans should be more the norm. Moreover, LTI-based plans will be cheaper to run (providing ‘more bang for the buck’) and, therefore, preferable to less efficient and, as demonstrated, equally effective alternatives.
Applying the research to address pay-for-performance
The current approaches to executive incentive plan design have left stakeholders on all sides concerned, confused and, at times, disillusioned with the results. On the one hand, practitioners worry that incentives may lead executives to be highly motivated but still reach poor outcomes. On the other hand, the perception is that results are not appropriately linked to the payout. onsequently, the ability of incentives to motivate performance is questioned. One of the CEOs we interviewed expressed it this way, “Incentives create too much angst. Having a competitive [pay] package is key, with only plus/minus 10 percent variation for individual performance. We pay for attraction and retention, and hope for business results. We adjust the pay levels to ensure that turnover [of staff] is within what we expect.” A RemCo Chair added: “Pay will not make an inventor invent more, but if he does, will he get his fair share? Pay-for-performance is not what it is geared up to be; it should be more about a sense of fairness.”
Our study provides empirically derived data on the effectiveness of often-used pay schemes, as a means to achieving higher company performance. One actionable outcome is that target setting has a greater impact on results than plan design. In line with prospect theory, as well as goal setting theory,6 stretch targets yield better results than average targets. Another implication of our study is that executive incentive plan design should focus on team incentives, as this approach is clearly superior to incentivise the achievement of business outcomes. The results also suggest that team incentive plans should be based on shared goals for maximum impact on business performance. How the team incentives are weighed seems less important in this respect. Our study also indicated that LTI plan designs could have a slight edge over other models, due to their higher ROI. Yet, perhaps LTI plans can be best used in a team context, or for executive retention purposes. A final implication of the study is that no amount of tweaking on any of the other elements of pay plan design seems to have any impact on performance. Thus, practitioners can focus on the elements of design mentioned above and worry less about the rest.
The study is limited insofar as it deals with relatively small amounts of money as fees, when compared to larger amounts of rewards in corporate executive pay packages. Furthermore, our LTI ‘share’ plans do not convey stock ownership, or a sense of ‘discounted expected future cash flows’, which company shares may convey. At any rate, we believe that the way the fee mechanisms were set up in our experiment adequately emulate the way LTI programmes behave in practice, since, for small companies at least, they tend to be based on a revenue-multiplier (or sometimes EBIT or EBITDA) formula, akin to our design. A variation would be to study if the conditions hold under different revenue-multiple scenarios to emulate differences in business values by industry. Another limitation of the current work is that the experimental design was not best suited for testing prospect theory. In an ideal setting, the game should be designed so that, for some subjects, early results would always be negative, and for others it would always be positive. That way, we could study the behaviour of all subjects who were presented with the same positive, or negative, stimuli, so that we can better compare the results. Perhaps additional research could address this issue.
Our findings are of use to RemCos, executive pay consultants, CEOs and HR professionals, in designing incentive programmes. One international executive pay consultant, who saw the results of the experiment, had this to say, “Given these results, I will provide greater weight towards team incentives [in future plan designs]. What gets measured gets done, but [the results indicate] it is not related to the amount or the plan design, but to the team goals. I would still split pay mix 1/3, 1/3, 1/3 but what will change is the team metric component. And I would not do away with share plan designs, but ensure the awards are based on team metrics.” Another consultant in the region suggested that, given these results, a more informed approach would be to use individual performance only for base pay increases and promotions, and make all incentives team-based. An HR executive in Hong Kong added: “[Individual incentives] are like adding sugar to the candy floss machine; it makes things go haywire! The best option is to simplify the plans.”
These executives, among others, are seeing the benefits of a premeditated approach to addressing pay-for-performance. It starts with recognising the limitations of current pay plan design and then developing a pragmatic pay programme that includes market data, governance requirements, and pay package competitiveness. Finally, these incentive plans will need to be tweaked to achieve a blend of results—from better performance to higher retention and adequate pay governance.
Dr. Fermin Diez
is an adjunct faculty at Singapore Management University
Levine, B., O’Neill, C., 2011. Abandoning Pay-for-Performance Myths in Favor of Evidence. WorldatWork Journal, First Quarter. Pink, D. H., 2009. Drive: The Surprising Truth About What Motivates Us. New York: Riverhead Books.
Vroom, V., 1964. Work and motivation. New York, NY: Wiley.
1. Extracted from author’s PhD thesis titled, “Pay for Performance: What Type of Pay Scheme is Best for Achieving Business Results?” 2018.
2. Gao, H., Luo, J., and Tang, T., 2015. Effects of Managerial Labor Market on Executive Compensation: Evidence from Job-hopping. Journal of Accounting and Economics, 59 (2015): 203-220.
3. Guzzo, R. A., Jette, R. D., Katzell, R. A., 1985. The Effects of Psychologically Based Intervention Programs on Worker Productivity: A Meta-Analysis. Personnel Psychology, 38(2): 275-91. Jensen, M. C. & Murphy. K. J., 1990. Performance Pay and Top-Management Incentives. Journal of Political Economy, 98(2): 225-264.
4. Kahneman, D. & Tversky, A., 1979. Prospect Theory: An Analysis of Decision under Risk. Econometrica, 47 (2): 263.
5. Ederer, F., Manso, G., 2013. Is Pay for Performance Detrimental to Innovation? Management Science, 59(7): 1496-1513.
6. Locke, E. A., Latham, G. P., 2002. Building a Practically Useful Theory of Goal Setting and Task Motivation: A 35-year odyssey. American Psychologist, 57 (9): 705–717.