On the Folly of Rewarding A, While Hoping for B

TL;DR: Steve Kerr’s famous paper shows how organisations undermine their stated goals by rewarding the behaviours they claim to discourage.  This has direct implications for risk management: risk teams should recognise the explicit and implicit incentives that influence how frameworks are designed and how challenge is delivered, because only by making those pressures visible can they decide when standard approaches are useful and when they need to be adapted or abandoned.

“Whether dealing with monkeys, rats, or human beings, it is hardly controversial to state that most organisms seek information concerning what activities are rewarded, and then seek to do (or at least pretend to do) those things, often to the virtual exclusion of activities not rewarded.”(i)

The views and opinions expressed on this account are my own and do not reflect the official policy or position of my employer.  Any content provided is for informational purposes only and should not be considered or relied upon as professional advice.

2026 marks the 50^th anniversary of Steve Kerr’s classic management paper, “On the Folly of Rewarding A, While Hoping for B”.(i)  Kerr’s core argument was that organisations often say they want one behaviour whilst rewarding another.

From a risk perspective, the most obvious application of his ideas is that poor incentive systems can weaken risk management processes. Failures in controls, risk appetite and strategy are often symptoms of incentives that encourage behaviours that do not align with business objectives. I explored this idea in an earlier post (Individual vs Organisational incentives) where I noted that incentives do not need to be financial, or even formal, to shape behaviour.

In this post, I want to apply Kerr’s ideas to risk teams themselves. Before doing that, I'll briefly summarise the paper.

On the Folly of Rewarding A, While Hoping for B

Kerr’s central idea is that dysfunctional employee behaviour is not always the employee’s fault: management often creates the conditions for that behaviour. Many management systems assume that individual incentives naturally align with organisational goals, and that employees understand strategy in the same way senior leaders do. Those assumptions can make organisations underplay the impact of poorly designed incentive schemes.

Good incentive schemes support clear organisational objectives. The problem is that many organisations start with objectives that are vague, inconsistent, or poorly designed. Vague goals are often easy to agree with because they are broad enough for everyone to support. For example, few employees would object to BP’s current slogan, “Reimagining energy for people and our planet”, but many would struggle to explain what it means in practical terms.

There is an important nuance here. Vague objectives may be meaningless on their own, but changing them can be meaningful. BP’s earlier slogan, “Beyond Petroleum”, was also vague, but removing it suggested a clear move away from green technologies. Similarly, Google’s removal of its “don’t be evil” motto prompted a stronger reaction from employees than the motto’s introduction ever did.

Kerr described vague goals as “high acceptance, low quality”. He contrasted these with more meaningful “operative goals”, which are higher quality but often less widely accepted. Operative goals can create friction because they force choices, expose trade-offs, and limit flexibility. Because too much friction can weaken management authority and create conflict, leadership teams may avoid setting too many specific goals themselves and instead delegate them down the organisation.

Even where organisations have operative goals that are widely accepted by the business, incentives systems can still become corrupted. Kerr identified four main ways in which this can occur:

1. Fascination with an “objective” criterion: Organisations often want incentive schemes to feel fair, clear and objective. However, not every valuable behaviour can be assessed objectively. If an organisation focuses too heavily on what is easy to measure, it may reward the wrong things. Kerr’s warning was that organisations should measure what matters, not simply what is easiest to measure.(ii)

2. Overemphasis on highly visible behaviours: Once organisations decide incentives should be objective, they often try to align them to measurable and visible behaviours. This can be effective for some roles, but in others it encourages people to focus on visible outputs rather than the more complex behaviours that actually create value. Kerr later grouped this point with objectivity, but I think it remains useful to treat this separately, especially when considering James Thompson’s work on interdependence (see below).

3. Hypocrisy: A more appropriate word might be “dishonesty” or “fogginess”. Kerr’s point was that leaders sometimes avoid stating their real goals because they do not think they will attract wider support. This corrupts incentives because the stated goals of the organisation do not match its real priorities. The opposite can also happen: organisations may dishonestly claim to want to achieve a goal, but actively design incentives that discourage employees from making it happen.

4. Morality or equity rather than efficiency: Sometimes moral or fairness considerations override efficiency. Kerr’s example compared US soldiers in the Second World War vs those in Vietnam. In WW2, soldiers were generally in service until the war ended, so their interests were more closely aligned with army initiatives to win the war. In Vietnam, soldiers were able to go home once they had survived their tours, misaligning soldiers’ interests (minimising personal risk) with those of the US military (winning the war quickly).   A system that aligned soldiers’ incentives more closely with military efficiency might have been more effective, but it may not have been morally or politically acceptable given the unequal burden placed on a smaller group of soldiers.

Kerr acknowledged that people are not solely motivated by rewards and punishments. Status, identity and professional values also impact individual decision-making and prioritisation (see Individual vs Organisational incentives). However, if people behave well despite the incentive system, the organisation is lucky rather than well designed. A formal reward system should reinforce the behaviour the organisation wants; it should not be an obstacle employees have to overcome.(i)

Three Types of Interdependence

Kerr was heavily influenced by James Thompson, especially his work on interdependence(iii). Thompson described three broad types of interdependence, moving from simpler to more complex:

1. Pooled interdependence: Group performance is the sum of individual performance. The Ryder Cup is a useful example: each golfer has limited ability to influence the performance of other team members. In this case, rewarding individual performance should clearly incentivise better team performance.

2. Sequential interdependence: Team performance depends on individuals acting in sequence, where one person’s performance affects the next. A relay team is a good example. You can measure individual hand-offs and overall team performance, and both affect the overall team performance.

3. Reciprocal interdependence: Team performance depends on collaboration. Football and rugby teams are good examples. You can reward individual outcomes, such as goals or tries, but if those rewards are too strong they can damage team performance. For example, a player may shoot from a difficult angle, incentivised by a goal bonus, when a pass would create a better likelihood of the team scoring.

Industrial Manufacturer Case Study

Kerr described a US industrial manufacturer whose senior leaders believed junior employees were behaving in dysfunctional ways. An independent survey asked employees whether specific behaviours were likely to be approved or disapproved of by managers.

The results showed that the behaviours senior management disliked were the same behaviours lower-level employees believed were rewarded. For example, many employees thought that agreeing with the boss, staying on good terms with everyone, and avoiding risk were more likely to bring approval than disapproval. Only at more senior levels was there evidence that the informal reward system support the behaviours management said it wanted.

Overall, the same "tendencies toward conservatism and apple-polishing at the lower levels" which divisional management had complained about during the interviews were those claimed by subordinates to be the most rational course of action in light of the existing reward system. Management apparently was not getting the behaviours it was hoping for, but it certainly was getting the behaviours it was perceived by subordinates to be rewarding. (i)

Kerr’s ideas applied to risk management teams

The incentives discussed below are not necessarily bad. In many cases, they are understandable and even necessary. For example, risk teams should meet regulatory requirements and avoid becoming the cause of a fine. The problem arises when these incentives become too dominant. A risk team that focuses too heavily on evidencing regulatory compliance may give less attention to analysis that would support an immediate, material business decision.

How incentives shape risk framework design

Meeting regulatory requirements: In financial firms, risk teams often derive their clearest legitimacy from regulation. This can incentivise them to prioritise work that produces formal evidence of compliance, when they should be incentivised to prioritise work that leads to better decision-useful analysis.

• Prioritising formal processes that regulators expect to see, while pushing more valuable informal work into “off-cycle” periods.

• Using a single management approach for all risk types because it is easier to evidence and explain.

• Aligning risk activity with reporting and governance cycles rather than the timing of important decisions.

Prioritising simplification and consistency: Board members and decision-makers are busy and have a lot to read. This can incentivise risk teams to simplify risk reporting, underplay complexities, and exacerbate silo-thinking.

• Simplification and consistency are valuable, but they can also restrict the risk team’s toolset.

• Risk registers often simplify ownership by assigning a risk to one individual. This fits the PRA SMF framework, but it does not always reflect how risk is actually created or managed.

• Risk registers tend to be anchored to existing power structures, oversight groups, leadership roles and management functions. When risks cut across silos, registers may struggle to capture the uncertainty or test the processes used to manage it.

Prioritising industry conformity: Non-Executive Board members often sit on several boards and may have clear expectations about what risk management should look like. Regulators can also spend more time seeking to understand non-standard frameworks. This incentivises risk teams to default to "industry-standard" approaches, raises the cost of innovation, and increases the personal risk for risk teams who try to do innovate.

How incentives shape risk management action

Prioritising what is visible: Risk teams rarely own decision-relevant data. As a result, they are incentivised to conduct risk analysis that relies on already accepted, visible business outputs, such as control processes, management information and decision analysis, as the basis for risk analysis. This can create blind spots where important parts of decision-making and uncertainty management are informal or invisible.

• Risk frameworks often align to processes rather than decisions. Where controls capture decision-making, they may be “soft”, such as confirming that a decision was made or reviewed by the right committee, or only indirectly related to decision quality.

• Risk frameworks tend to prioritise what can be measured, quantified, standardised or controlled.  This often reflects how risk owners are assessed by the wider business.

Maintaining legitimacy: Risk teams usually have limited direct decision-making authority. Their influence depends on maintaining legitimacy with decision-makers and being seen as useful, proportionate and relevant [see Social License & Risk Management]. This can incentivise risk teams to maintain their legitimacy over providing meaningful challenge to support material decisions.

• Risk teams can prioritise challenging decisions where risk involvement is expected, while avoiding areas where there is no clear expectation of involvement.

• When a review produces several recommendations, risk teams may prioritise those most likely to be seen by the business as material, even where this does not align with risk’s opinion.

• Risk teams may provide stronger challenge to business areas with less organisational power.  In insurance firms, for example, risk teams may challenge operational functions more robustly than underwriting functions.

Avoiding repeat mistakes: Risk teams are rarely blamed the first time an operational failure occurs. However, if the same kind of event happens again, they may be seen as complicit, even where earlier analysis reasonably supported inaction.  This can incentivise risk teams to promote a risk-averse, rather than risk-proportionate, culture;

• Pushing the business to introduce new controls and bureaucracy after a loss event.

• Encouraging the business to reduce risk appetite after a loss event.

• Giving more attention to events that have previously affected the firm, or firms where members of the risk team have worked before.

What should risk teams do about it?

There is no silver bullet. The incentives created by organisations, industries and regulators will not disappear, and risk teams cannot ignore them. But naming those incentives is useful. It helps risk teams:

• Understand how their frameworks are shaped, and where they may fall short,

• Identify where framework limitations may become problematic,

• Decide when the default cadence of risk activity needs to change.

• Explain to business stakeholders why risk frameworks and activities sometimes need to be adapted.

---

I hope this blog sparks ideas and discussion. If you found it interesting, please share or connect with me on LinkedIn to contribute or provide feedback!

---

(i) Kerr, Steven. “On the Folly of Rewarding A, While Hoping for B.” The Academy of Management Journal, vol. 18, no. 4, 1975, pp. 769–83. JSTOR, https://doi.org/10.2307/255378.

(ii) Wright, Thomas & Hollwitz, John & Stackman, Richard & Groat, Arthur & Baack, Sally & Shay, Jeffrey. (2017). 40 Years (and Counting): Steve Kerr Reflections on the “Folly”. Journal of Management Inquiry. 27. 105649261771266. 10.1177/1056492617712664.

(iii) Organizations in Action . By James D. Thompson. (New York: McGraw-Hill, 1967)