The case that measurement design is the root cause
V. F. Ridgway’s 1956 research, one of the earliest systematic examinations of this problem, found that single, multiple and composite performance measures can all produce undesirable consequences for overall performance, and documented concrete examples: employment interviewers evaluated on interview volume conducted rapid interviews and placed few applicants; investigators given a monthly case quota selected easy cases toward month-end. Ridgway’s conclusion was pointed: “quantitative measures of performance are tools… but research indicates that indiscriminate use and undue confidence and reliance in them result from insufficient knowledge of the full effects and consequences.”
Steven Kerr’s 1975 research extended this into the specific territory of incentive design, examining reward systems across a manufacturing company, an insurance firm, and settings including politics and medicine, and finding a consistent pattern: organisations that reward one behaviour while stating they hope for a different one should expect people to pursue the rewarded behaviour, because, in Kerr’s words, “most organisms seek information concerning what activities are rewarded, and then seek to do… those things, often to the virtual exclusion of activities not rewarded.” On this account, persistent underperformance despite apparently reasonable metrics is a sign that the metrics themselves are measuring, and therefore incentivising, something other than what leadership actually wants.
The practical version of this case: if performance isn’t improving despite good-looking numbers, examine what is actually being measured and rewarded — formally and informally — before assuming the strategy or the people are the problem.
A further consideration specific to this position: measurement problems are often self-concealing in a way decision-rights problems are not. A flawed metric that is being optimised against will, by construction, show improving performance on that metric — which is precisely what makes it hard to catch from the inside. Leadership reviewing a dashboard of improving numbers has little structural reason to suspect the numbers themselves are the problem, which is part of why Ridgway’s original 1956 warning about undue confidence in measurement remains as relevant now as it was then.
The case that decision rights are the root cause
Paul Rogers and Marcia Blenko’s Bain-based research offers a different diagnosis. Their survey of executives at more than 350 global companies found that only around 15% reported having an organisation that helps the business outperform competitors, and identified unclear decision rights — not unclear metrics — as the primary bottleneck, stalling at points such as global-versus-local authority, centre-versus-business-unit tension, and function-versus-function disputes. On this account, teams can be measuring the right things and still underperform collectively if it is unclear who actually has the authority to make the calls that would translate good measurement into better outcomes.
Donald Sull and Rebecca Homkes’s research reinforces this from the execution side: their finding that only 9% of managers could reliably count on colleagues in other functions and units suggests that even well-measured, well-intentioned teams can fail to convert individual performance into organisational performance if the cross-unit coordination and decision authority needed to act on what the metrics show is itself unclear or contested.
The practical version of this case: if performance isn’t improving despite good-looking numbers, examine whether anyone actually has clear authority to act on what those numbers show, before assuming the metrics themselves are the problem.
A further consideration specific to this position: decision-rights problems tend to be more visible to the people experiencing them directly than measurement problems are, but less visible to leadership reviewing aggregate results. A team that knows precisely which decision it cannot make without an escalation that routinely stalls experiences the bottleneck daily; a leadership team reviewing quarterly performance sees only the downstream effect — results that underperform expectations — without an obvious signal pointing back to the specific point where authority is unclear.
Why the disagreement persists — and what it means for a leader
The disagreement persists because both diagnoses are supported by substantial, credible research, and because the two problems — flawed measurement and unclear decision rights — can produce an identical symptom from the outside: metrics that look fine, performance that doesn’t improve. Neither research tradition claims the other’s mechanism is false. Ridgway and Kerr’s argument does not depend on decision rights being clear; Rogers and Blenko’s argument does not depend on measurement being well-designed. Both problems can be present at once, and in practice, for a growing organisation, they frequently are — which is part of why this question resists a clean resolution.
What the evidence does not support is a default assumption in either direction. An organisation that assumes its measurement system must be flawed, and redesigns its metrics without examining decision rights, may improve its measures while leaving the actual bottleneck — who has authority to act — untouched. An organisation that assumes decision rights are the problem, and reorganises reporting lines without examining what is actually being measured and rewarded, risks the same outcome in reverse. The more defensible starting point, per both bodies of research, is to test which pattern is actually present — teams optimising visibly for the measured target at the expense of the stated goal (Ridgway/Kerr), or teams with good individual metrics unable to agree who has authority to act on them (Rogers & Blenko/Sull & Homkes) — rather than assuming either cause by default.
For a growing Singapore organisation specifically, the stakes of misdiagnosing this are raised by the pace and regional complexity many Singapore-based organisations operate under: a wrong diagnosis costs more time to correct in an environment where market conditions are already moving faster than a standard review cycle, as related research on regional leadership alignment in this series discusses.
There is a further, compounding cost specific to misdiagnosis in either direction: each redesign attempt — of metrics, or of reporting lines — consumes organisational attention and goodwill, and a second redesign attempted shortly after a first one failed to fix the underlying problem tends to be met with more scepticism and less genuine engagement than the first attempt received. An organisation that has already tried “fix the metrics” and seen no improvement, and now tries “fix the decision rights” without having established that the diagnosis has actually changed rather than simply moved to the next plausible-sounding explanation, risks a pattern where each subsequent intervention lands with less credibility than the one before it — regardless of whether the second diagnosis happens to be the correct one.
Where this shows up
The questions here are set in Singapore. More Perspectives for this market are on the Singapore page, and the Markets index shows the other business hubs.
Questions worth considering
- When a team’s metrics look good but organisational performance doesn’t follow, is the more visible pattern teams optimising for the measured target, or teams unable to agree who has authority to act?
- Has the organisation ever redesigned its metrics without first checking whether decision rights were the actual constraint — or vice versa?
- What would distinguish, in this organisation specifically, a measurement problem from a decision-rights problem, given that both can produce the same surface symptom?
Your team is facing similar challenges?
Subscribe to The Riddle Ledger
Perspectives are sent as they are published, with no promotional email, and you can unsubscribe at any time.