Who is actually suited to working for themselves?

You have had the thought more than once now, usually on a Sunday. Not a business plan — just the feeling that you would rather answer to the work than to a person, and a quieter suspicion that you might not be the type. The type, as everyone describes it, is a risk-taker. And you are not especially one of those.

That last part is worth pausing on, because the research on it is much messier than the stereotype, and in one large study it points the other way entirely.

The short answer

Two different questions, two different answers. The traits that predict whether you’ll start working for yourself are reasonably well established. The traits that predict whether you’ll still be doing it in five years are not — the two best studies disagree outright.

On starting: founders score higher than managers on conscientiousness (d′ = .45), openness (.36) and emotional stability (.37), per Zhao and Seibert’s 23-study meta-analysis. But split conscientiousness in two and the whole effect sits in achievement motivation (d′ = .59) and none of it in dependability (d′ = .01). “Entrepreneurs are conscientious” really means achievement-striving, not tidy and dutiful.

On risk: the founder-as-risk-taker idea is genuinely contested. In Rauch and Frese’s meta-analysis, risk-taking is the weakest predictor in the whole list — r = .10 for both starting and succeeding. And one study of 1,261 nascent entrepreneurs found them more risk-averse than the general population.

What actually predicts becoming self-employed?

Achievement motivation, openness, and emotional stability — with an important caveat about who they’re measured against.

Zhao and Seibert compared entrepreneurs with managers, not with the general population. So the finding is about how founders differ from other business people, not from everyone. Their definition was also narrow: founder-owner-managers whose principal purpose is growth, which excludes most ordinary self-employment — the contractor, the consultant, the person with a trade.

The follow-up meta-analysis of around 60 studies separated intention from outcome and found four of the Big Five predicted both — conscientiousness, openness, extraversion and emotional stability — while agreeableness predicted neither. The combined effect was a multiple R of .36 for intentions and .31 for performance, which is to say the Big Five together explain roughly 13% of the variance in whether someone intends to do this. Real, and a long way from decisive.

The most useful detail in that paper is a split: risk propensity predicted intentions but was unrelated to performance. Being drawn to risk makes you more likely to start. It does not make you better at it.

Do I need to be a risk-taker?

No — and this is the part of the literature worth knowing about, because it contains a published argument that never resolved.

Stewart and Roth (2001) meta-analysed 14 samples and found entrepreneurs higher in risk propensity than managers, d′ = .36. Miner and Raju (2004) added fourteen studies the first review had omitted, all using projective rather than self-report measures, and reached the opposite conclusion — entrepreneurs looked more risk avoidant. Stewart and Roth replied with the finding that explains the whole dispute: the direction flips with the instrument. Self-report measures give d = 0.31; projective measures give d = −0.35.

Then there is the comparison nobody makes. Both of those pit founders against managers — themselves an unusually risk-tolerant group. Xu and Ruef (2004) compared 1,261 nascent entrepreneurs against the general population and found them significantly more risk-averse. Their explanation is that risk-averse people enter risky ventures because they want autonomy and identity fulfilment, not because they expect to win a bet.

The most recent large study cuts the difference honestly. Kerr and colleagues (2019) found entrepreneurs reported the highest risk tolerance of any group they measured — but on Big Five traits they were similar to ordinary employees apart from somewhat more openness and more neuroticism, which the authors say “aligns with earlier literature that has failed to consistently identify a strong entrepreneurial personality.” They also note the entrepreneur premium was weakest on the one behavioural measure — an actual lottery — rather than on self-description.

So: people who work for themselves describe themselves as risk-tolerant. When you ask them to gamble, the difference shrinks.

Will I last?

Here the honest answer is that the evidence is thin and conflicting, and you should be suspicious of anyone who tells you otherwise.

Ciavarella and colleagues (2004) found conscientiousness positively related to long-term venture survival — and, more interestingly, openness negatively related to it. The trait that helps you start may be the one that hurts you at lasting: the appetite for the new is not obviously an asset in year six of the same business.

But the better-powered study disagrees. Volery and Mattes (2022) tracked 1,621 self-employment stints through twelve waves of an Australian household panel and found conscientiousness, emotional stability and openness had no significant main effect on survival. Only extraversion (positive) and agreeableness (negative) had direct effects.

The defensible summary: conscientiousness — specifically achievement-striving — reliably predicts becoming self-employed. The evidence that it predicts surviving is mixed, and the largest longitudinal study finds no direct effect at all.

For base rates rather than traits: in the UK, 39.4% of businesses born in 2018 were still active five years later, per the ONS business demography release. US figures are similar in shape — about 80% survive one year and roughly half reach five, remarkably stable across cohorts. Note the ONS figure covers VAT/PAYE-registered businesses, which under-counts the smallest sole traders — likely including the kind of work you’re contemplating.

Would I actually be happier?

A little more satisfied with the work. No different on mental health. And it depends enormously on why you left.

The largest synthesis — Stephan, Rauch and Hatak (2023), 319 effect sizes across 94 studies and 82 countries — found self-employment associated with higher positive wellbeing overall, dr = 0.117, falling to 0.064 once corrected for publication bias. Work satisfaction was the strongest result at dr = 0.129. Life satisfaction was roughly half that, at 0.056 — which the authors read as entrepreneurs trading other life domains for satisfaction at work.

And the finding most people assume goes the other way: negative wellbeing showed no difference at all (dr = 0.000), across a sample of over 6.7 million. Stress-related mental health problems: no significant difference. Self-employment is not, on this evidence, worse for your mental health. Nor is it better.

The moderator is the thing to take away. Opportunity entrepreneurs — people who left because they saw something worth doing — scored higher than employees on positive wellbeing, work satisfaction and life satisfaction. Necessity entrepreneurs — people with no better option — scored lower than employees on positive wellbeing and life satisfaction, with no work-satisfaction advantage at all.

The authors are candid that the overall effect is small by statistical standards. The gap between opportunity and necessity is not.

So what would actually tell me?

Less than you’d like, and a few things clearly.

• Look at achievement striving, not conscientiousness overall. The predictive weight sits there, and dependability carries none of it.
• Stop using risk appetite as the test. It is the weakest predictor in the literature at r = .10, it predicts intention rather than performance, and its direction reverses depending on how you measure it.
• Notice which kind of move this is. Whether you are going towards something or away from something is, on current evidence, the single biggest wellbeing variable here — larger than any trait.
• Treat survival as a base rate, not a character test. Roughly six in ten UK businesses are gone within five years. Most of that is not personality.

The uncomfortable summary is that personality research can tell you something real about whether you’ll start and very little about whether you’ll succeed. If you are looking for permission, a test cannot give it. What it can do is tell you which parts of the job will cost you most — and that is worth knowing before you find out the expensive way.