The capability–behaviour gap: when leaders can, but don't

What the capability–behaviour gap is

Most leaders arriving in a development programme do not have a knowledge problem. They know they should delegate, that feedback must be prompt and specific, that a team under sustained load needs protecting. Many can do that – they have shown it in calmer periods, without a client breathing down their neck. And still, the people around them rarely experience it. That difference is the capability–behaviour gap – often labelled the knowing–doing gap, although knowing is rarely the problem.

DefinitionCapability–behaviour gap

The distance between the capability a leader is credited with – above all by themselves ("I could do this if I chose to"), often by others too – and how often supervisors, peers and direct reports actually observe the corresponding behaviour. The gap does not describe a missing capability, but capability that fails to turn into behaviour – it asks not the same question twice, but two different ones: "Could I?" and "Do they?"

Why a gap tells you more than a low score

A classic 360° report shows behaviour scores. A low delegation score is information – but incomplete. It does not say whether the leader cannot delegate or simply does not. Both look identical in the behaviour profile and need entirely different responses.

The second case is the expensive one. Anyone who owns a development budget knows the pattern: training leaders in what they can already do – good course ratings, unchanged behaviour. The gap shifts the question from "What is missing?" to "What is in the way?" – the question that makes a development conversation productive.

Four typical causes and how to tell them apart

When capability is present and behaviour absent, in our experience one of four causes is almost always behind it, often two at once. Each gives a different answer to the question "When do you do it?"

1. Pressure and triggers

The cause we encounter most often in consulting and professional-services firms. The capability is real, but tied to calm conditions. Under client pressure, near a deadline or in conflict, behaviour tips into an older, lower-energy pattern: control instead of delegation, harmony instead of clarity, withdrawal instead of presence – and the switch is not experienced as a decision. Tell-tale sign: others describe two versions of the same person. More in Leadership under pressure.

2. Beliefs

The leader can delegate – but believes quality only happens when they do the work themselves, or that letting go signals weakness. Such beliefs are rarely conscious and were useful once. Kegan and Lahey call the mechanism competing commitments: next to the stated goal ("delegate more") sits an unspoken one that wins when in doubt ("never let a mistake reach the client"). Tell-tale sign: not doing it is justified with arguments the leader finds plausible – see Beliefs vs. Behavior.

3. Energy and Well-being

Behaviour that requires capability costs energy: listening instead of instructing, holding back instead of stepping in, explaining instead of deciding. When the buffer is spent – months of sustained load, no recovery, no sense of meaning – the same leader falls back on whatever costs least. The gap is then a load issue, not a leadership one; more training will not close it. Tell-tale sign: the behaviour was still there a year ago.

4. Context and incentives

Sometimes the gap is rational. A partner measured on personal revenue has little reason to delegate client contact; someone held accountable for every delay will not let go of control. The organisation rewards something else. Tell-tale sign: the gap appears in several leaders of the same unit. No questionnaire measures this cause – it belongs in the conversation with HR and the executive team.

How the gap is measured – and why only across sources

The obvious approach would be to ask a leader twice – "Can you?" and "Do you?" – and take the difference. That measures only their own story: one person answering both questions from the same self-image. A reliable gap needs two sources.

A gap must be large enough to mean something. On a six-point scale, small differences arise from response style, the respondent's form on the day and the number of raters alone. LEADBeyond 360° therefore interprets a calibration gap only from ±0.75 points; smaller differences are treated as noise. The threshold is a product rule, not a research finding – and deliberately explicit, because research on self–other agreement has used inconsistent metrics and produced correspondingly inconsistent results (Fleenor et al. 2010). How to read self–other differences in general: Self-image and others' image: blind spots in a 360°.

How to debrief the gap

A capability–behaviour gap is a rewarding way into a conversation, because it starts with a strength: the leader can do something. The delicate part is the second half: "others don't see it." Kluger and DeNisi (1996) found that feedback improves performance on average, but more than a third of the interventions examined made it worse – and effectiveness declines as attention shifts from the task to the self. For the conversation: stay with the situation, not the character.

1. Start with the strength. Where is the behaviour observed – which situations, with whom? That establishes the capability is real.
2. Map the exceptions. When does it disappear? "When the client calls" (trigger), "when I can't afford the risk" (belief), "since the spring" (energy), "never with this partner" (context).
3. Examine the self-rating, don't attack it. What does it rest on? Block or calibration question – both are legitimate.
4. Bring in the other layers. Does a trigger pattern fit? Is Well-being critical? Which beliefs might be blocking? The report offers hypotheses, not answers.
5. Agree an experiment, not a goal. "Next project, I delegate work package X completely" beats "delegate more". The gap closes in situations, not resolutions.

What this means for leadership development

In a meta-analysis of 24 longitudinal studies, Smither, London and Reilly (2005) found that other-ratings improve only modestly on average after multisource feedback, and advise against expecting large improvement. Change becomes more likely when the leader perceives the need, believes change is feasible, sets goals and takes action – working with a coach, say, or discussing the feedback. A report alone changes little; the conversation about what keeps capability from becoming behaviour is what works.

For programme owners that yields a simple test: if your 360° delivers only behaviour scores, you cannot know whether a low score needs training or a different conversation. An instrument that sets capability, triggers, beliefs and Well-being beside observed behaviour answers it – not conclusively, but well enough to spend the budget in the right place.