Leadership and the Productivity Gap
Why leadership capability, not technology, is the overlooked driver of organisational performance.
Download the Briefing (PDF)
The lever hiding in plain sight
The global productivity conversation is narrowly framed. Economists and commentators measure productivity as an output ratio and then search for levers to move it, and technology dominates the search. The evidence points elsewhere. Technology is not moving the ratio in any reliable way, and the most consequential driver of productivity is going largely undiscussed: the competence and judgement of the people leading the work. This briefing sets out why, and what to do about it.
- Standard productivity statistics cannot distinguish a manager who uses a team's time well from one who wastes it. The measurement itself has a blind spot.
- Australian productivity is in structural decline. The Reserve Bank has downgraded its long term growth assumption and now attributes the slowdown to structural, not cyclical, factors.
- AI has not moved national productivity statistics. The evidence points to an amplifier effect, magnifying existing capability rather than replacing it.
- A manager's leadership ability is roughly as important to team output as the team's total productive capacity, and the effect is driven by task competence, not personality.
- The global fall in engagement was driven by a collapse in manager engagement specifically, not by front line disengagement.
- A defined, learnable set of leadership competencies accounts for most of the variance the research attributes to manager quality.
- Leadership development that is applied and sustained produces a measurably different result to one off training, and the return can be tracked at the team level.
What productivity actually measures, and what it misses
Productivity, at its simplest, is a ratio: output divided by the inputs required to produce it. In practice, that gets reduced to one of two figures. Labour productivity is output per hour worked. Multifactor productivity is output per combined unit of labour and capital, essentially a broader version of the same idea that also accounts for equipment and infrastructure. Both are useful numbers. Both also hide something important.
The Australian Bureau of Statistics itself acknowledges that labour productivity "implicitly reflects the other factors of production, such as the contribution of capital and other factors affecting production such as technological change." In plain terms, the standard measure bundles the contribution of managerial quality, skill, decision making and organisational effectiveness into a single number. When that number falls, the policy response defaults to capital investment or technology adoption. The contribution of human capability rarely gets equivalent scrutiny, because the statistic was never built to isolate it.
Deloitte's 2023 research makes the point sharply. Employees spend an average of 32% of their time on work that gives the appearance of productivity without adding to it, and while 60% of executives track hours and email volume as productivity proxies, only 15% of employees believe that kind of tracking actually helps efficiency. The metric has drifted away from the behaviour it is meant to represent.
There is a more refined ABS measure, quality adjusted labour input, which attempts to account for differences in workforce education and experience. By this measure, labour productivity fell 0.7% in 2024 to 25, compared with 0.2% on the standard hours worked basis. The gap suggests that improvements in workforce qualifications were quietly masking a worse underlying picture. But even this more sophisticated measure only captures credentials. It does not capture applied competence: the difference between a technically qualified manager operating at a mediocre level and one operating with genuine skill. Both look identical in the data. That is the blind spot this briefing is about.
Australia's problem is real, local and structural
National productivity fell 0.5% in 2024 to 25, against a 20 year average of 0.4% annual growth. That is not a small miss. It is a move in the wrong direction. Labour productivity fell 0.2% over the same period and now sits at roughly its 2016 level, close to a decade with no real net gain.
The Reserve Bank, in its August 2025 Statement on Monetary Policy, cut its medium term labour productivity growth assumption from 1.0% to 0.7% a year. That is a meaningful admission. What had previously been treated as a temporary soft patch is now described as structural. The RBA points to declining business and labour market dynamism, slower diffusion of better working practices, declining competition, and slowing growth in workforce skills. That last factor, skills and human capital, sits on the RBA's own list of causes and rarely features in the public conversation about fixing the problem.
The Parliamentary Library's April 2026 research brief describes the slowdown as a sustained threat to long term economic prosperity, and the Centre for Independent Studies frames it as a growth slump undermining national living standards. None of this is fringe commentary. It is the consistent view of the institutions responsible for tracking the number.
None of the structural factors on the RBA's list sit inside a single organisation's control. Business dynamism, national skills growth and competition policy are economy wide forces. The quality of the people leading your teams is not. That distinction is why the rest of this briefing focuses where it does.
The technology dividend has not arrived
The dominant narrative right now positions AI as the fix for sluggish productivity. The OECD's 2024 report on AI notes the technology has real potential to lift productivity, but that the evidence to date is concentrated in innovation metrics rather than broad national productivity figures. The gap between promise and proof is wide, and it has a precedent.
In 1987, economist Robert Solow observed that the computer age was visible everywhere except in the productivity statistics. It became known as the Solow Paradox. A 2025 empirical study across OECD countries from 2000 to 2024 found the same pattern now applies to AI: rapid technological progress is not automatically translating into greater national productivity or economic growth. McKinsey's February 2026 analysis of the United Kingdom found labour productivity growth of just 1.1% over the year to Q3 2025, with no detectable AI effect in the aggregate figures. Separate 2026 research on the AI productivity paradox reached a similar conclusion: individual level productivity gains from AI tools are not translating into organisational or national outcomes.
AI is an amplifier, not a substitute. It magnifies whatever capability is already there. It does not create capability that was not there to begin with.
None of this makes AI irrelevant. A mediocre decision maker with AI assistance makes faster mediocre decisions. A capable leader with AI extends their judgement further and faster. The commercial implication is sequencing. Organisations rolling out AI tools this year should build the decision making and prioritisation capability of the leaders who will use those tools first, or at the same time. Skip that step and the organisation is paying for the amplifier with nothing worth amplifying.
Same title, completely different results
Most conversations about leadership treat it as binary. A person either is a manager or is not. That framing misses the point. Picture five people with the same title, the same size team, the same resources. All five, technically, play tennis. One plays socially on a Saturday morning. Another has represented their state. If they competed, the scoreline would not be close, and nobody watching would call it a fair contest, despite both being tennis players. Leadership works the same way. The title tells you almost nothing about the level someone is actually operating at.
A 2026 study in the Quarterly Journal of Economics, conducted by researchers at the University of Gothenburg, put a number on that gap. Using randomised controlled experiments, managers were assigned to multiple different teams while team member skill was held constant, allowing the researchers to isolate the manager's own contribution to performance. The finding: a manager's overall leadership ability is roughly as important to team performance as the total productive capacity of the employees. In a related field study of managers at a large retail chain, moving from an average quality manager to a good quality manager was associated with a 25% increase in annual sales.
A manager's leadership ability is roughly as important to team performance as the total productive capacity of the employees.
What predicted a good manager was not personality, charisma or self confidence. The strongest predictors were measures closely tied to the actual tasks of the job: applied competence, not presence. A separate 2023 study published in the Review of Economic Studies, combining granular production data with manager surveys across 120 production lines in India, found substantial productivity variation both across teams doing overlapping work and within the same team over a production run, and linked that variation directly to supervisor quality. Managerial attention and control turned out to matter more to output than either cognitive skill or tenure. A 2023 IZA discussion paper adds a further dimension: the effect of a genuinely high quality manager is not confined to the period of direct supervision. It leaves a lasting, measurable mark on the career trajectories of the people who reported to them.
The CIPD's 2023 systematic review of people managers confirms the same relationship from a different angle, finding a direct and robust link between manager behaviour and employee engagement, wellbeing and performance. Yet only 44% of managers globally have received any formal management training. Most people playing this game have never had a lesson.
This is arguably the single most commercially useful fact in the entire productivity conversation, and it is barely discussed outside academic economics. If the gap between an average and a good manager is worth a quarter of annual output, then moving a population of managers from average to good is a growth strategy that needs no capital expenditure and no technology purchase. And because the predictor is competence rather than personality, it is a development problem, not a hiring problem. It needs the people already in the room to be properly developed.
The engagement multiplier, and where it starts
Engagement is usually treated as a wellbeing metric. It should be read as a productivity metric, because Gallup's meta analysis consistently finds that roughly 70% of the variance in team engagement is attributable to the manager.
Global engagement fell to 21% in 2024, tying the lowest level recorded since the pandemic, at an estimated cost of US$438 billion in lost productivity in that single year. The most recent Gallup data, released in April 2026, shows engagement has fallen further still to 20%, with the annual cost now estimated at US$10 trillion, roughly 9% of world GDP. Gallup's own modelling suggests that if workplaces globally reached best practice engagement levels, an additional US$9.6 trillion could be unlocked. Australia sits below the global average, with only around 20% of employees engaged and four out of five either indifferent or actively disengaged.
Look past the headline figure and the real story is about where the decline came from. Manager engagement fell from 30% to 27% globally, the sharpest drop of any group measured, while individual contributor engagement held steady. The collapse came from the middle of organisations, not the front line.
That matters because middle managers occupy the most leveraged position in the system. They sit between strategic intent and operational execution, and their leadership quality moves in both directions from there. Research in the Journal of Business Economics confirms two distinct pathways: a direct bypass effect, where middle manager behaviour influences employees two levels below them, and a cascading effect, where middle manager leadership shapes the behaviour of first line supervisors, who then shape employee performance. Both pathways are empirically confirmed, which means fixing the middle moves the needle twice, not once.
A multi year Italian panel study across all sectors, published in the journal Industrial Relations, found that formal off the job training for middle managers has a significant effect on total factor productivity at the whole organisation level, not just on the individual manager's own output. The effect was stronger in larger firms, and different training methods produced meaningfully different results, which is a reminder that training quality and design matter as much as training volume. A 2025 case study in the energy sector confirms the same bridging role in practice: middle managers translate strategy into ground level action, and the quality of that translation shapes everything downstream of it.
The practical implication is where to intervene first. Because the effect is a cascade, the fix does not need to reach every leader in the organisation at once. Direct development toward the middle layer specifically, with the same calibre of investment typically reserved for senior executives, rather than the generic supervisor training most middle managers received on the way up.
The competencies that actually move the needle
Leadership capability is not a personality trait and it is not charisma. The research points consistently to a specific, learnable cluster of skills that drive team level productivity.
Decision Quality and Prioritisation
There is a real difference between managing time and prioritising tasks. Time management is about the calendar. Prioritisation is about protecting the highest value work and having the discipline to say no to the rest. Most teams default to activity because it feels safer than judgement. Leaders who can make and hold a genuine trade off decision under pressure create a productive environment. Leaders who cannot create only the appearance of one.
Meeting Design and Facilitation
Atlassian's 2024 global research, covering 5,000 knowledge workers including 1,000 in Australia, found meetings in Australian workplaces are ineffective 76% of the time, and identified meetings as the single biggest barrier stopping employees from completing their core work, ahead of unclear goals or lack of motivation. Almost two thirds of meetings surveyed never stated their intended goal, and nearly three quarters ended by scheduling another meeting.
Stakeholder Management and Lateral Influence
A leader's productivity is not confined to their direct reports. The cascade research above confirms influence moves sideways and upward as much as it moves down. Leaders who cannot operate without formal authority create drag on the wider system, even where their own team is performing well.
Feedback and Coaching Orientation
Managers who operate as coaches rather than controllers produce measurably different outcomes. Data on coaching trained managers shows engagement gains in the leader themselves of up to 22%, team engagement gains of up to 18%, and 20 to 28% improvements in team performance metrics. When managers receive role specific training and consistent support, their own reported wellbeing rises from 28% to 50%.
None of these require a personality transplant. They require deliberate, applied practice, which is exactly what most leadership training does not provide.
The macroeconomic case, not just the organisational one
This is not only a workplace argument. The OECD's 2024 working paper on human capital and firm productivity establishes that firms at the productivity frontier do not just rely more heavily on high skill occupations, they hire the most skilled available workers within each occupation. Differences in workforce skill composition explain close to one fifth of firm level productivity dispersion, and young firms with a high quality workforce are more likely to grow rapidly. Skills and human capital are structural determinants of productivity, not background noise.
The Reserve Bank's own August 2025 statement explicitly lists slowing growth in skills and human capital in the labour force as one of the potential causes of Australia's productivity slowdown, citing a companion OECD paper titled From Decline to Revival: Policies to Unlock Human Capital and Productivity. Human capital development is being treated, at the central bank level, as a macroeconomic policy lever rather than a human resources conversation.
The implication for individual organisations is direct. In an environment where aggregate productivity is declining, where technology is not yet delivering the promised gains, and where disengagement is rising, the organisations that invest systematically in leadership capability are positioned to gain relative ground while their competitors wait for a policy response or a technology cycle that has not yet arrived.
Training is not development
Attending training and developing capability are not the same activity. The Italian panel research cited in Part Five confirms that different training methods produce materially different effects on productivity. A one day workshop and a generic off the shelf framework are both training. Neither, on their own, is development.
What actually shifts team level output is coaching applied to a leader's live context, structured feedback on real behaviour rather than theoretical concepts, and enough sustained time for it to change how someone operates, not just what they know. This has been CoachStation's operating model since 2010, and the evidence sits in the relationships, not the marketing. The Toyota Finance Australia engagement began as two days of consulting in December 2012 and is still running more than a decade later. Wabtec's engagement has continued across leaders moving between roles and locations, precisely because the development travels with the person rather than being tied to a single event.
There is also a measurement argument worth making plainly, because the claim that leadership development returns cannot be measured is convenient but wrong. The research in Part Four confirms productivity varies significantly within organisations, between teams doing comparable work, based purely on the quality of the person leading them. Any organisation with reasonable data on team output, engagement or retention already has the raw material to track this properly: establish a baseline before development begins, set specific competency targets tied to productivity rather than generic leadership qualities, and review leader behaviour and team output together, not leader satisfaction alone. That is a measurement design, not a leap of faith.
Reflect and Act
For Individuals
If I am honest, which end of that tennis court am I standing on right now, and in which parts of my role?
Where do I default to activity because it feels safer than making a genuine prioritisation call?
When did I last receive structured feedback on my actual behaviour, rather than a general performance conversation?
For Team Leaders
Do I know, with any precision, what my team's productivity gap would be worth if I closed it?
Which of my meetings would survive being cancelled, and what does that tell me about how I am running them?
Am I coaching my people or controlling them, and would they answer that question the same way I would?
For Organisations
Where is our leadership development spend going: one off training events, or sustained, applied coaching?
Do we track team level productivity against leadership capability, or only against headcount and technology investment?
If manager quality is worth as much as team capacity, what does our current investment ratio between the two actually look like?
A practical starting point
If the argument in this briefing holds, here is a straightforward place to begin, without a major programme or significant budget commitment.
-
Find the Gap (Weeks 1 to 2)
Identify two or three teams doing comparable work with different managers. Look honestly at the variance in output, engagement or retention between them. This is usually where the tennis analogy stops being theoretical and starts being visible in your own numbers.
-
Name the Competencies (Weeks 3 to 4)
Rather than a generic leadership audit, assess your managers specifically against the four competencies in Part Six: prioritisation, meeting facilitation, stakeholder influence and feedback quality. These are observable, not personality based, and the gaps are usually obvious once you look for them specifically.
-
Invest Where the Leverage Is (Ongoing)
Direct development toward the middle of the organisation first, not the top. This is where the research shows the cascade effect originates, and where most organisations invest the least relative to the leverage available.
The organisations that separate themselves in a low productivity, low engagement environment will not be the ones waiting for the AI dividend to arrive. They will be the ones already building the leadership capability that determines whether any tool, any strategy or any team ever reaches its actual capacity.
You recognise the pattern. Let's find your gap.
If the ideas in this briefing are relevant to your organisation, book a discovery conversation. No pitch. No pressure. A direct conversation about where you are and what is worth doing next.
Book Your Free Discovery Conversation