50 Years of Leadership Development but Still No Measurable Improvement?
Vicky Russ is a leadership coach and former CEO with 20+ years’ experience across the public, health, and education sectors. She specialises in leadership, transformation, and helping leaders navigate complex change and life transitions.
Organisations spend heavily on developing their leaders and then send them back into structures that will not let them lead. The evidence is more uncomfortable than it first looks. Leadership training does work. It produces real learning, and yet when researchers ask the people being led whether anything actually changed, they notice less than anyone else involved. This article argues that the problem is not the quality of leadership development but its unit of analysis. We have spent fifty years refining the individual and almost no time changing the conditions that individual returns to, and in large, bureaucratic organisations, those conditions reliably win. It sets out what the research says really determines whether leadership learning survives contact with the job, why public sector organisations are hit hardest, and ten questions any leadership team should ask before commissioning another programme.
The Monday after training
She comes back from the programme lit up. Three days away from the service, up on the balcony, good facilitators, a room full of people wrestling with similar things, and a notebook with four pages of notes and actions in it. She has decided she is going to push decisions down to the people doing the work, say less in meetings and ask more, and stop being the bottleneck she has privately known she is for about two years.
Week one back, she is covering two vacancies, and the notebook stays in her bag. Week two, she tries the first thing on the list, devolving a small budget decision to her team leader, and is told by finance that it needs her sign-off because that is how the scheme of delegation is written. In week four, the restructure consultation opens. By week six, her own manager, who did not attend the programme and has not asked about it, sends back a decision she had delegated with a note asking why it did not come through her.
Six months on, nothing in the service looks different. The organisation concludes, reasonably enough, that the training did not work. It books a different provider for next year.
I have watched this happen from inside organisations, from the front of training rooms, and directly to myself. It is not a story about a weak leader or a poor programme. It is a story about what we think leadership development is for.
The evidence is more uncomfortable than “it does not work”
The easy version of this argument is that leadership training is a waste of money. That version is wrong, and the research says so clearly.
The Chartered Institute of Personnel and Development (CIPD) 2023 evidence review of leadership development found a moderate positive effect overall. The individual studies it draws together report effect sizes from around d = .43 for changes in behaviour up to δ = .76 overall, which in this field is respectable. People who go on good leadership programmes learn things. They can demonstrate that they have learned them. The training, considered as training, works.
Then the review reports something that should stop everyone who sells this for a living. When researchers measured whether the learning actually transferred into the workplace, the answer depended almost entirely on who they asked. Leaders rating themselves produced a transfer effect of d = .64. Their own managers put it at .53. Their peers at .26. The people they lead, the ones the whole exercise was supposed to benefit, put it at .13.
Read that sequence again, because the pattern in it is the article. The further you move from the person who went on the course, the less anybody can see that anything happened. The individual most convinced they have changed is the individual who attended. The people underneath them, who experience leadership rather than receive training in it, register almost nothing.
This is not new. The same review cites an analysis of managerial training effectiveness covering roughly fifty years, from 1952 to 2002, which found no improvement in measured effect across the entire period. Fifty years of better theory, better facilitation, better materials, better evaluation, and the needle did not move. When an industry refines its product for half a century, and the results hold still, the problem is not the product.
I should be straightforward about something here. I design and deliver leadership development. I lecture in Leadership at BPP University, and I sell leadership programmes to organisations. So this is an uncomfortable set of findings for me to put in writing, and you are entitled to ask why I am doing it. The answer is that I have watched too many capable people come back from good programmes into organisations that had not changed by a single degree and conclude privately that the failure was theirs.
What actually determines whether it sticks
If the variation is not mostly in the programme, where is it? The research is unusually direct about this, and it has been for years.
The strongest predictors of whether leadership learning transfers sit outside the training room altogether. They are, broadly, whether there is an opportunity to use the learning at all, whether the climate supports the new behaviour, whether constraints get in the way, and whether rewards and sanctions point in the same direction as the training. Meta-analytic work on transfer identifies work environment factors, including support, climate, and constraints, as central rather than peripheral. More recent work finds that the opportunity to apply what was learned is itself a significant predictor of whether it is applied, which sounds obvious written down and is almost never designed for.
One finding deserves to be printed and pinned above the desk of everyone who commissions training. Transfer is greatest when the trainee's own superior has also been trained, and when the rewards and sanctions in the work environment support the new behaviour. Send one person on a programme and their manager becomes the wall. Send the manager too, and the wall becomes a door.
The CIPD review's own conclusion is that leadership development should be strongly and systematically integrated into workplace climate and processes rather than run as an event. That is a quietly radical statement from a professional body, and the market has largely ignored it because events are easier to buy, easier to budget, and easier to put on a slide.
Why this is worse in large, bureaucratic organisations
Every organisation has some gap between what it trains and what it permits. In large bureaucracies, the gap becomes a canyon, and for a specific reason.
Bureaucracy, stripped to its mechanism, is agency replaced by process. Where trust is thin, a sign-off is added. Where a decision once went wrong, an approval stage is created. None of it is irrational, and almost none of it is ever removed, so over the years, the organisation accumulates a structure in which nobody below the top can change anything without permission from three people who were not there.
Now consider what leadership programmes teach. Make decisions closer to the work. Speak up when something is wrong. Experiment. Challenge upwards. Give your people autonomy. Every one of those behaviours is precisely what a sign-off culture is built to prevent. We are not merely failing to reinforce the training. We are running a far longer, far better-funded programme in the opposite direction, five days a week, and calling it governance.
The public sector shows the consequence most clearly. National Health Service (NHS) sickness absence in England stood at 5.0 percent in May 2026, up from 4.7 percent a year earlier. The Office for National Statistics puts the private sector at 1.7 percent and the public sector as a whole at 2.9 percent. These are different series and not a like-for-like comparison, but the direction and the size of the gap are not in doubt, and the gap has been there for as long as anyone has measured it.
Meanwhile, the structures themselves are being torn up. In November 2026, the National Audit Office reported that meeting running cost savings across integrated care boards requires around 8,000 posts to go, roughly a third of their staff, that £582 million was spent on redundancies in 2025 to 2026, and that the restructuring has damaged morale and left it unclear whether the capability needed has been retained. At the same time, NHS leaders are being measured harder than ever through the oversight framework and the published trust league tables.
So the sector is removing a third of a leadership layer, measuring the survivors more tightly, and continuing to send individuals on programmes designed to make them more autonomous. It is difficult to design a better experiment in how to produce frustrated leaders.
The 4 conditions, briefly
In an earlier article for this magazine, I set out the four conditions every culture either builds or breaks, whether people believe that acting will change anything, which is agency, whether they feel connected, whether they feel capable, and whether they find that the direction makes sense. Agency is not one of four equals. It is the ground the other three stand on, and it is the one that goes first.
Here is where that argument meets this one. Conventional leadership development builds capability in one person. It is good at it. But capability is one condition of four, and it is the only one a training room can reach. Agency, connection, and coherence are properties of the organisation, not of the individual, and no amount of development delivered to a person in a hotel conference suite will change them.
Which means we have been sending people away to improve the one condition that was probably not their problem, and returning them to the three that were.
From developing leaders to creating the conditions
The evolution this points to is not better content. It is a different unit of analysis. Five shifts follow directly from the research above, and none of them requires anyone to abandon what they already do well.
First, send people in groups, not alone. If transfer is greatest when the trainee's superior has also been trained, the implication for admissions is obvious and almost universally ignored. A leader who returns with two colleagues and a shared vocabulary has allies. A leader who returns alone has an anecdote.
Second, diagnose the conditions before you develop the people. You would not prescribe before examining. Establish which of the four conditions is weakest in this service, with evidence rather than assumption, and aim the development at that. Most programmes are bought on a hunch about what is wrong.
Third, make the work between sessions an organisational change rather than a personal reflection. Not “journal on your leadership style” but “remove one sign-off and report what happened”. Reflection changes the person. Removing a sign-off changes the conditions, and it also tests, cheaply and early, whether the organisation will actually permit what it has just paid to be taught.
Fourth, develop the layer above at the same time. This is the finding that most directly contradicts how leadership development is sold because it means the cohort cannot simply be whoever has budget and availability.
Fifth, measure the conditions, not the satisfaction. Happy sheets measure the event. If you must rely on self-report, remember that self-report is the measure that flatters the programme most and the people being led least. Ask the team. Measure before and after, against instruments somebody else validated.
I have seen this from 3 sides
I teach this material. I lecture in Leadership at BPP University on the Level 7 Senior Leader Apprenticeship, working with senior leaders from across sectors, and I watch what happens when they take an idea back to a service that has no room for it.
I have commissioned it. As a director in the public sector and as a charity chief executive, I bought leadership development for my people, and I did it the way almost everybody does, one person at a time, chosen for merit, sent somewhere good, and welcomed back into exactly the same structure they left. I did not think of that as a design flaw. I thought of it as development.
I have been on the receiving end of it, as a leader who came back changed to an organisation that had not, and who quietly concluded for a while that the problem must be me. That conclusion is the real cost of getting this wrong, and it is paid by the most committed people you have because they are the ones who tried.
10 questions to ask before you commission another programme
None of this means stop developing leaders. It means change what you are buying. These are the questions I would want answered before signing anything.
Who else from this organisation is going? If the answer is nobody, you are sending one person back to argue with a system.
Is their line manager going, or have they been? The research is unusually direct here. Transfer is greatest when the superior has been trained too. Otherwise, the manager becomes the wall.
What will they be allowed to do differently on the Monday? Name it before they go. If you cannot name anything, you have just answered a different and more important question.
Which of the four conditions is weakest here, and does this programme touch it? Most programmes build capability. If your problem is agency, you are about to make capable people more acutely aware of what they are not permitted to do.
What will you remove to make room for what they learn? A sign-off, an approval stage, a reporting line. Development without subtraction is just more to carry.
How will you know it worked, other than asking them? Self-report is the least reliable measure available and the one almost everyone uses because it is the one that tells us what we hoped.
Have you asked the people they lead? They notice least of all. That is either the most damning finding in this literature or the most useful, depending entirely on what you do next.
What happens to this if there is a restructure in month three? In the public sector at the moment, assume there will be. Learning that cannot survive reorganisation has a very short half-life.
Are you developing people, or being seen to develop them? An honest answer sometimes changes the budget line and occasionally changes the whole conversation.
If nothing in the organisation changes, what do you expect this to achieve? This is the question the rest of the article exists to make askable.
The question underneath all the others
Leadership development is not failing because it is badly made. On the evidence, it is made rather well. It is failing to show up where it matters because we keep aiming it at the only part of the system that was never really the obstacle.
The individual leader is the easiest thing in an organisation to change and the least likely to change anything on their own. The conditions are harder, slower, and less satisfying to work on, and they are the thing that stays behind when the people leave. Change the leader but not the conditions, and you have changed nothing. Change the conditions, and you will find you needed to change the leader far less than you thought.
So the question I would put to any board considering next year's leadership budget is not which provider or how many places. It is this. What is it about an ordinary Tuesday here that would stop a well-trained leader from leading well, and what are we willing to remove?
Great leaders are not in short supply. The conditions for great leadership are. Conditions, unlike people, are built on purpose, or they are built by accident. There is no third option.
This is the work I do with organisations through leadership development, the Conditions Audit, and coaching that builds the four conditions deliberately rather than hoping for them. If you would like to see how it is structured, I have a corporate brochure that sets it out in full. To request it or to book a free, no-obligation conversation, visit here or message me directly.
Read more from Victoria (Vicky) Russ
Victoria (Vicky) Russ, Managing Director & Leadership Coach
Vicky Russ is a leadership coach and former CEO with over 20 years’ experience across the public, health and education sectors. She specialises in leadership, transformation and helping leaders navigate complex change and life transitions. Her work focuses on supporting leaders to maintain performance while creating psychologically safe, high-performing cultures. Vicky also teaches Positive Psychology Leadership and brings a practical, human approach to even the most challenging situations. She is passionate about “protecting the protectors” and helping leaders move from pressure to purpose.
References:
Blume, B. D., Ford, J. K., Baldwin, T. T. and Huang, J. L. (2010). Transfer of training: a meta-analytic review. Journal of Management. As cited in CIPD (2023).
Chartered Institute of Personnel and Development (2023). Leadership Development: An Evidence Review. Scientific summary. Available at: cipd.org.
National Audit Office (2026). Restructuring of Integrated Care Boards, 4 November 2026. Available at: nao.org.uk.
NHS England (2026). NHS Sickness Absence Rates, May 2026. Available at: digital.nhs.uk.
Office for National Statistics (2025). Sickness Absence in the UK Labour Market. Available at: ons.gov.uk.
Powell, K. S. and Yalcin, S. (2010). Managerial training effectiveness: a meta-analysis 1952 to 2002. As cited in CIPD (2023).
Taylor, P. J., Russ-Eft, D. F. and Taylor, H. (2009). Transfer of management training from alternative perspectives. As cited in CIPD (2023).
Taylor, P. J., Russ-Eft, D. F. and Chan, D. W. L. (2005). A meta-analytic review of behaviour modelling training. As cited in CIPD (2023).
The King's Fund (2017). Caring to Change: How Compassionate Leadership Can Stimulate Innovation in Health Care. Available at: kingsfund.org.uk.
