Relative performance evaluation, sabotage, and collusion

Matthew Bloomfield, Catarina Marvão, Giancarlo Spagnolo 09 October 2020



Since the late 1990s, firms such as Yahoo, Continental Airlines, and Delta have benchmarked their performance against that of a reference group of firms, using their performance as informationally valuable signals. By including relative performance evaluation (RPE) measures in executive compensation packages, firms are able to strip away the effects of common shocks, thus making managerial incentives more precise and effective (e.g. Lazear and Rosen 1981, Holmstrom 1982). According to agency theory, the use of RPE in managerial compensation should be widespread, but Bizjak et al. (2018), for example, show that only around 57% of S&P 500 firms use some sort of RPE incentive, and that a large majority of firms do not use any RPE. This has come to be known as the ‘RPE puzzle’.

Indeed, RPE may have some unintended consequences. One of these is the potential for costly ‘sabotage’: managers may be willing to take actions that hurt the reference group’s performance so as to inflate their own relative performance, even if at significant cost to their own absolute performance (e.g. Gibbons and Murphy 1990, Chowdhury and Gurtler 2015). This RPE-induced sabotage has been shown to play an important role in many contexts, from corporate promotions (e.g. Chen 2003) to higher education (e.g. Royal and Guskey 2014) and sports (e.g. Del Corral et al. 2010).

Relative performance evaluation in CEOs’ pay packages

The potential for RPE-induced sabotage may also be an important consideration in the context of CEO incentives. In this context, costly sabotage would likely take the form of overly aggressive product market strategies, such as sub-optimally low prices, extreme output volumes, or excessive advertising spending. When the potential for costly sabotage is significant, shareholders may choose to withhold RPE incentives, even at the expense of deprecated informational efficiency (and therefore impaired risk-sharing). This could in part explain why RPE is used so unfrequently for managerial compensation. The literature, however, provides no compelling evidence that firms avoid using RPE in their CEOs' pay plans because of the potential for costly sabotage. 

A new study

In a new paper (Bloomfield et al. 2020), we examine whether the potential for RPE-induced sabotage is an important consideration for CEO incentives. We exploit cartel participation as a source of variation in the potential for costly sabotage. In explicit collusive arrangements, otherwise rival firms collectively agree to soften product market strategies. We posit – and show formally – that if the cartel is stable, it removes concerns about sabotage by constraining managers’ ability to choose overly aggressive market strategies, making RPE more palatable to shareholders. 

To test this hypothesis we take advantage of SEC’s 2006 requirement that public companies disclose RPE use and provide statements including the performance benchmarks and reference group.

A visual inspection of the relation between (detected) cartel participation and the use of RPE clearly shows that cartel members are about 60% more likely to use RPE than non-cartel members. This effect is confirmed in our analysis, which also unveils that it is driven by more concentrated product markets, where the potential for RPE-induced sabotage is larger: in industries of above-median concentration, cartel firms are roughly 130% more likely to use RPE than non-cartel members.

We then examine whether, among RPE users, cartel firms construct more economically similar peer groups, and we find that cartel members do indeed select more similar peers (in terms of 2-digit SIC and similar product offerings) than do non-cartel members. These results are consistent with the notion that firms are cognizant of the trade-off between risk-sharing and costly sabotage, and that cartel membership allows firms to focus more on the risk-sharing aspect of peer selection.

We further show that firms frequently drop RPE from their CEOs' incentive plans when their cartel membership is terminated by such an enforcement action (see Figure 1). This pattern is once again driven by firms in concentrated markets.

Figure 1

Our results also indicate that, among non-cartel firms, RPE is positively and significantly associated with measures of product market aggression (sales volume, total costs, spending-to-sales, and advertising expenditures). This evidence is consistent with the interpretation that RPE induces more aggressive behaviour (i.e. costly sabotage), and that cartel membership is effective at curtailing these destructive actions.

Finally, we test the alternative hypothesis that it is the excessive competition induced by RPEs that leads managers to form cartels, rather than cartels allowing for a sabotage-free use of RPE. This reverse causality hypothesis is rejected by our data: RPE adoption does not predict cartel membership, while cartel membership does predict RPE adoption.

Collectively, our evidence provides support for the notion that costly sabotage is an important deterrent to firms' use of RPE, and that explicit collusion mitigates this possibility, thereby facilitating more efficient risk-sharing between shareholders and executives.


Bizjak, J, S Kalpathy, Z F Li and B Young (2018), “The role of peer firm selection in explicit relative performance awards”, SSRN Working Paper No. 2833309.

Bloomfield, M, C Marvão and G Spagnolo (2020), “Relative Performance Evaluation, Sabotage and Collusion”, CEPR Discussion Paper 15115.

Chen, K P (2003), “Sabotage in promotion tournaments”, Journal of Law, Economics, and Organization 19 (1): 119-140.

Chowdhury, S M and O Gurtler (2015), “Sabotage in contests: a survey”, Public Choice 164 (1-2)- 135-155.

Del Corral, J, J Prieto-Rodriguez and R. Simmons (2010), “The effect of incentives on sabotage: The case of Spanish football”, Journal of Sports Economics 11 (3): 243-260.

Gibbons, R and K J Murphy (1990), “Relative performance evaluation for chief executive officers”, ILR Review 43 (3):30{S.

Holmstrom, B (1982), “Moral hazard in teams”, The Bell Journal of Economics 324-340.

Lazear, E P and S Rosen (1981), “Rank-order tournaments as optimum labor contracts”, Journal of Political Economy 89 (5): 841-864

Royal, K D and T R Guskey (2014), “The perils of prescribed grade distributions: What every medical educator should know”, Journal of Contemporary Medical Education 2 (4): 240.



Topics:  Competition policy Labour markets

Tags:  RPE, cartel, Collusion, risk-sharing, sabotage, CEO pay, performance-related pay

Assistant Professor of Accounting, The Wharton School, University of Pennsylvania

Lecturer, Technological University Dublin; Research Affiliate, Stockholm School of Economics

Professor of Economics SITE - Stockholm School of Economics and University of Rome Tor Vergata, Research Fellow EIEF and CEPR


CEPR Policy Research