Happy employees, better performance?

What the science says when you look

At LearnTec, I was there to talk about culture; but the topic i came away thinking baout, among others, was happiness.

These things happen at conferences. You set off to discuss one topic and within an hour you’re three conversations deep into something else, watching ideas migrate between the coffee station and the breakout rooms. Culture talk bleeds into adjacent territory; one of the places it bleeds into, reliably, is wellbeing. And from wellbeing, almost inevitably, you arrive at the same article of faith: happy employees perform better.

I’d come off the back of a talk where I’d argued, with what I hoped was sufficient force, that we should be more sceptical of the things we believe because they feel right or align with what we already think we know.

Where had this one come from?

In thinking about it, I realised that not only am I not thinking about this enough, but the more I think about it, the more “happy employees perform better” has a lot of red flags.

So I went looking.

What follows is what I found after a couple of weeks of reading. The short version: the claim isn’t wrong, but it’s nothing like as solid as we treat it. The longer version is the rest of this article.

What we mean when we say “happy”

The first problem, is that there isn’t a single thing called happiness in the research. There are at least four overlapping things, and the studies that get pooled into the “happy employees perform better” narrative are often measuring different ones.

Hedonic wellbeing is the pleasure-based kind. It’s about feeling good, having positive emotions, and not having too many negative ones. The word comes from the Greek hedone, meaning pleasure. When you tick a survey box saying you felt “cheerful” or “enthusiastic” in the past week, you’re reporting hedonic wellbeing. The most widely used measures here are the Satisfaction with Life Scale (Diener et al., 1985) and the Positive and Negative Affect Schedule, usually shortened to PANAS (Watson, Clark and Tellegen, 1988).

Eudaimonic wellbeing is the meaning-based kind. The word comes from Aristotle’s eudaimonia, sometimes translated as “human flourishing.” It’s about living well: having purpose, growing as a person, exercising autonomy, doing things you find meaningful. You can have high eudaimonic wellbeing whilst feeling exhausted, frustrated, or stressed. Anyone who’s written a book, or trained for something hard will recognise the distinction. The standard measure is Carol Ryff’s six-factor model of psychological wellbeing (Ryff, 1989).

Job satisfaction is how you feel about your job specifically. Not your life, not your mood today, but your job, considered as a whole. It’s an evaluation: a settled view formed over time. The largest body of evidence on workplace happiness uses job satisfaction as its central construct.

Work engagement is being absorbed in your work, energised by it, and committed to it. In the academic literature it has three components, vigour, dedication, and absorption, and is measured via the Utrecht Work Engagement Scale (Schaufeli and Bakker, 2004). In the commercial world it’s been hollowed out into something Gallup measures using its Q12 instrument, which is a different beast entirely.

When someone tells you happy employees perform better, you should ask which kind of happy.

Hedonic feelings in the moment?

A settled sense that your job is alright?

A deep sense of meaning?

Being absorbed in flow?

These aren’t the same thing and they don’t behave the same way, and the evidence supporting them differs in important respects. Lumping them together gives you a confident-sounding claim built on shaky aggregation.

There’s a name for this kind of conceptual fudge. The organisational psychologists Bill Macey and Benjamin Schneider call it the jangle fallacy (Macey and Schneider, 2008): mistaking different constructs that go by similar names for the same thing. It’s particularly rife in the engagement literature, where the academic construct, the commercial construct, and the everyday-language construct often get treated as the same when they aren’t.

Bear this in mind as we dive into the evidence, as this will be an ongoing problem.

What the evidence shows

There are three pieces of research that do most of the heavy lifting here.

The first is a meta-analysis by Timothy Judge and colleagues (Judge et al., 2001), which pooled 312 studies covering more than 54,000 people. This is the closest thing we have to a settled baseline answer on the job satisfaction and job performance question. The correlation they found was around 0.30.

A correlation of 0.30 means job satisfaction explains about nine percent of the variation in job performance.

The second piece of research is experimental (Oswald, Proto and Sgroi, 2015). The team wanted to know whether you could cause a productivity gain by lifting someone’s mood, not just observe that the two correlated. So, they showed some participants a short comedy film clip to put them in a good mood, others got nothing, and then everyone did a paid task. The cheerful group were about twelve percent more productive.

This is useful evidence, because it’s experimental. The mood change came first, the productivity change came after, and the assignment was random. It’s a short laboratory task done on a piece-rate, which is to say, the more you produced, the more you earned. That’s not the same as a normal job. It doesn’t tell us what happens to a knowledge worker on a Thursday afternoon in the third hour of teams calls.

The third piece is more recent (Bellet, De Neve and Ward, 2024). A team studied nearly 1,800 call centre workers over six months. The workers reported their happiness each week using a simple emoji scale, and the researchers linked that to actual sales data.

A one-point increase in happiness, on a scale of zero to ten, was associated with about a twelve percent increase in productivity, measured by sales. For complex tasks, like negotiations or contract renewals, the figure rose to around twenty percent. The happier workers weren’t working longer hours; they were converting more calls into sales.

There are two things to really consider here.

The first is that the effect was concentrated in complex work. On simple, repetitive calls, the happiness boost mattered far less. This fits with Judge’s meta-analysis, which also found that the satisfaction-performance link is stronger in complex jobs. If your job is largely cognitive and involves judgement, negotiation, or persuasion, happiness probably does help. If it’s largely procedural, the effect is smaller.

The second is that this is one company, one sector, one country. It’s the best field evidence we have. It’s not proof of a generalisable idea.

Taken together, these three pieces of research tell a coherent story: there is a modest, relationship between happiness and performance, it’s larger in complex work than in simple work, and the effect sizes in well-designed studies hover around twelve percent in specific settings. Not nothing, but not transformative either.

Note: There is a huge amount of research into happiness and its effects on human beings. Here I’ve tried to focus on studies that either specifically relate to, or have been leaned on, by the business world to justify claims about happiness’s impact on performance.

I would not claim that this short piece provides a full synthesis of all the research out there on happiness and its impact on people and their work. Just on the reading that I have completed.

The mythology of happiness

Most of the popular case for “happy employees perform better” doesn’t rest on the evidence I’ve just summarised. It rests on a handful of stories that have been propagating for so long they’ve taken on the texture of facts. But do they hold up under scrutiny?

The first is the Hawthorne studies. You’ll find these cited all over the place as the foundational evidence that worker happiness, or even worker attention, drives productivity. Lighting was adjusted at a factory in the 1920s; productivity went up; productivity went up again when lighting was adjusted back down; and so the “Hawthorne effect” was born.

Except it wasn’t. In 2011, the economists Steven Levitt and John List recovered the original experimental data, which had been thought lost (Levitt and List, 2011). What they found was that the dramatic productivity patterns that everyone refers to weren’t in the data. The story had been polished and embellished over decades of retelling until it bore little relationship to what the experiments had shown. The Hawthorne studies are not evidence that happy workers perform better, because the productivity pattern they’re famous for isn’t supported by the original numbers.

Note: If you’re unfamiliar, the Hawthorne effect was stated as a psychological phenomenon where people temporarily change or improve their behavior simply because they know they are being observed.

Attempts to replicate the study have regularly failed, and this is no longer generally considered to be good or reliable science.

The second is the happiness pie chart, popularised by the positive psychologist Sonja Lyubomirsky and colleagues. It says that fifty percent of your happiness is determined by your genes, ten percent by your circumstances, and forty percent is under your conscious control through “intentional activity.”

It’s also not supported by the evidence (Brown and Rohrer, 2019). The original calculations don’t hold up. The authors themselves have since revised the model and conceded the original figures were, in their words, “approximate” and “speculative” (Sheldon and Lyubomirsky, 2021). Which is one way of saying they made the numbers up.

The third is the Losada line or “critical positivity ratio,” the claim that flourishing teams maintain a 2.9013-to-one ratio of positive to negative interactions. It was based on borrowed equations from fluid dynamics, applied to human emotions, and presented with four decimal places of precision. As you know, more decimal places always means greater accuracy…

In 2013, the journal American Psychologist effectively retracted the mathematical model after it was challenged and disproved by a paper showing the underlying equations had been misapplied (Brown, Sokal and Friedman, 2013). The ratio was, in the authors’ phrase, “entirely unfounded.”

None of this means there’s nothing to the happiness-and-performance idea. The evidence I covered earlier still holds. But it does mean that a substantial portion of what gets cited in support of the claim is no longer credible, and hasn’t been for years. If you’re presenting the case for workplace happiness using Hawthorne, the happiness pie, or the Losada ratio, you’re presenting folklore rather than the rather more solid evidence that is readily available to support your point.

Are we thinking about this the wrong way round?

So at this point, it very much looks like there is a strong argument to say we should be doing what we can to make employees happier at work in order to increase performance. It may not be a massive transformational improvement, but incremental gains are important, especially when increasing happiness need not cost much, if anything. All of this means it’s time to throw a bit of a spanner in the works…

What if we’re thinking about this all wrong?

In 2001, the psychologist Roy Baumeister and colleagues published a paper called “Bad Is Stronger Than Good” (Baumeister et al., 2001). It pulled together findings from across psychology to show: bad events have a larger psychological impact than good events of equivalent magnitude.

• A criticism stings more than a compliment soothes.

• A betrayal damages a relationship more than an equivalent kindness repairs it.

• Negative information gets processed more deeply, remembered longer, and acted on more strongly than positive information.

If you apply this to the workplace, and everything we’ve discussed so far, we start to face a rather more interesting question. The asymmetry suggests that the real productivity lever might not be making people happier; it might be addressing what’s making people unhappy. Or, in other words, rather than happy employees being more performant, is it actually the case that unhappy employees are significantly less performant?

A study using British workplace data (Bryson, Forth and Stokes, 2017) found that job satisfaction, the settled, evaluative kind, predicted workplace performance. But transient positive feelings at work, the kind you’d see on a daily mood survey, didn’t. Which is to say, what mattered was the absence of dissatisfaction, more than the presence of cheer.

The burnout literature, summarised by Christina Maslach and Michael Leiter (Maslach and Leiter, 2016), points in the same direction. The organisational costs of burnout, including poor performance, errors, absenteeism, turnover, and a higher risk of mental illness, are substantial and well-documented. The causes of burnout are also well-documented: workload mismatch, lack of control, insufficient reward, breakdown of community, perceived unfairness, and value conflict.

These aren’t problems you fix with a wellness app or a chief happiness officer. They’re problems you fix by changing how work is structured and how people are managed.

So when we say happy employees perform better, what we might be picking up on, partially, is the inverse: that unhappy employees perform worse, and that the cost of unhappiness is much larger than the gain from happiness. It’s a less marketable framing, it doesn’t fit on a poster, but it points to where the evidence leads, which is towards the unglamorous business of job design, workload, fairness, autonomy, and decent management.

What this means in practice

To my fellow L&D people, you’ve probably been part of conversations where the happy-employees-perform-better claim was used to justify a programme, an intervention, or a piece of spend. Some of those interventions are good, and some are wellness theatre. The evidence gives us a way to tell the difference. Here’s what I’d take from it.

Stop citing the founding myths. If you find yourself reaching for Hawthorne, the happiness pie, or the Losada ratio to make a point, find different evidence. These don’t support what they’re being asked to support, and citing them undermines the credibility of everything else you’re saying. There’s plenty of better-grounded research to draw on, including the three studies I’ve highlighted in this article.

Be precise about what kind of happy you mean. If you’re proposing an intervention, ask which construct it targets, whether that’s hedonic in-the-moment cheer, eudaimonic sense of meaning, job satisfaction, or engagement. Different constructs respond to different inputs, and the evidence on what works varies. Conflating them in your business case is how you end up with a yoga programme being used to address a burnout problem caused by chronic overwork.

Audit your dissatisfiers before adding happiness initiatives. The Maslach and Leiter work on burnout gives you a six-item list to start with:

• workload,

• control,

• reward,

• community,

• fairness,

• values.

If any of these are seriously broken in your organisation, you’ll get more performance gain from fixing them than from adding anything new. This is unglamorous and involves harder conversations than launching a programme. But it gives you a much higher chance of delivering a measurable, positive performance improvement.

Treat vendor data as marketing, not science. Gallup’s engagement research, and the parade of similar research from wellness vendors and HR tech providers, has commercial interests baked into it. The peer-reviewed Gallup meta-analysis (Harter, Schmidt and Hayes, 2002) exists and is real, but it uses Gallup’s own proprietary instrument, on Gallup’s own client base, with Gallup employees as authors. That’s not to say that it has no value, but it’s not the same as independent science, and shouldn’t be considered of equal merit.

Be careful with the productivity figures. The headline “twelve percent productivity gain” comes from specific studies in specific contexts. If you’re quoting it in a business case for a software platform aimed at office workers in a complex hybrid environment, you’re stretching the evidence. Honest framing protects your credibility for the long run, and the long run is where most of us are trying to build a career.

Don’t mandate positivity. The evidence on emotional labour, traced back to the sociologist Arlie Hochschild (Hochschild, 1983), shows that requiring people to display positive emotions they don’t feel produces emotional exhaustion and burnout, the opposite of what most happiness interventions are aiming for. Forced cheer is worse than no intervention at all.

Design jobs, don’t manage moods. The clearest practical implication of the evidence is that the things which most reliably support both happiness and performance are structural:

• autonomy,

• manageable workload,

• fair treatment,

• social support,

• meaningful work,

• competent management.

These are job design choices. They’re not in our gift alone, but we can advocate for them, surface them through the work we do on management capability, and stop being used as a substitute for them.

Where this leaves us

The claim that happy employees perform better isn’t wrong. There’s a modest relationship, with some good causal evidence behind it. But the way the claim is used in our industry is far more confident than the evidence supports, leans on stories that have been discredited for years, and points us towards the wrong interventions.

There’s a relationship between certain kinds of workplace happiness and certain kinds of performance, the size of the relationship is modest, it’s larger in complex work than in simple work, and the most powerful lever is probably reducing unhappiness rather than adding happiness.

Job design beats mood management.

Decent treatment beats wellness initiatives.

Fixing the broken things outperforms adding new things.

That’s not as marketable as “happy employees are twenty percent more productive.” It is, as far as I can tell, what the evidence says.

---

References

Baumeister, R.F., Bratslavsky, E., Finkenauer, C. and Vohs, K.D. (2001) ‘Bad is stronger than good’, Review of General Psychology, 5(4), pp. 323-370.

Bellet, C.S., De Neve, J.-E. and Ward, G. (2024) ‘Does employee happiness have an impact on productivity?’, Management Science, 70(3).

Brown, N.J.L. and Rohrer, J.M. (2019) ‘Easy as (happiness) pie? A critical evaluation of a popular model of the determinants of well-being’, Journal of Happiness Studies, 21, pp. 1285-1301.

Brown, N.J.L., Sokal, A.D. and Friedman, H.L. (2013) ‘The complex dynamics of wishful thinking: The critical positivity ratio’, American Psychologist, 68(9), pp. 801-813.

Bryson, A., Forth, J. and Stokes, L. (2017) ‘Does employees’ subjective well-being affect workplace performance?’, Human Relations, 70(8), pp. 1017-1037.

Diener, E., Emmons, R.A., Larsen, R.J. and Griffin, S. (1985) ‘The Satisfaction With Life Scale’, Journal of Personality Assessment, 49(1), pp. 71-75.

Harter, J.K., Schmidt, F.L. and Hayes, T.L. (2002) ‘Business-unit-level relationship between employee satisfaction, employee engagement, and business outcomes: A meta-analysis’, Journal of Applied Psychology, 87(2), pp. 268-279.

Hochschild, A.R. (1983) The Managed Heart: Commercialization of Human Feeling. Berkeley: University of California Press. (PAID)

Judge, T.A., Thoresen, C.J., Bono, J.E. and Patton, G.K. (2001) ‘The job satisfaction-job performance relationship: A qualitative and quantitative review’, Psychological Bulletin, 127(3), pp. 376-407.

Levitt, S.D. and List, J.A. (2011) ‘Was there really a Hawthorne effect at the Hawthorne plant? An analysis of the original illumination experiments’, American Economic Journal: Applied Economics, 3(1), pp. 224-238. (PAID)

Macey, W.H. and Schneider, B. (2008) ‘The meaning of employee engagement’, Industrial and Organizational Psychology, 1(1), pp. 3-30.

Maslach, C. and Leiter, M.P. (2016) ‘Understanding the burnout experience: recent research and its implications for psychiatry’, World Psychiatry, 15(2), pp. 103-111.

Oswald, A.J., Proto, E. and Sgroi, D. (2015) ‘Happiness and productivity’, Journal of Labor Economics, 33(4), pp. 789-822.

Ryff, C.D. (1989) ‘Happiness is everything, or is it? Explorations on the meaning of psychological well-being’, Journal of Personality and Social Psychology, 57(6), pp. 1069-1081.

Schaufeli, W.B. and Bakker, A.B. (2004) UWES - Utrecht Work Engagement Scale: Preliminary Manual. Utrecht: Occupational Health Psychology Unit, Utrecht University.

Sheldon, K.M. and Lyubomirsky, S. (2021) ‘Revisiting the sustainable happiness model and pie chart: Can happiness be successfully pursued?’, Journal of Positive Psychology, 16(2), pp. 145-154.

Watson, D., Clark, L.A. and Tellegen, A. (1988) ‘Development and validation of brief measures of positive and negative affect: The PANAS scales’, Journal of Personality and Social Psychology, 54(6), pp. 1063-1070.