What the data actually shows
People appear to judge others on two main dimensions, and warmth tends to lead. The warmth-competence model developed by Fiske, Cuddy and Glick finds that when we size someone up we ask, roughly in this order, whether they mean us well (warmth) and whether they can act on it (competence). Warmth — trustworthiness, good intent — is typically processed first and weighted heavily, because historically it mattered most to know whether a stranger was friend or threat before judging how capable they were.
Building on that, Amy Cuddy has argued that leaders who establish warmth first earn more trust and, through it, more genuine influence than those who try to lead with toughness or competence alone. The claim is that projecting strength before warmth can read as cold and even threatening, which makes people defensive rather than cooperative — whereas warmth first opens the door for competence to actually land. This is a strong, well-argued position rather than a settled law, but it is consistent with the broader two-dimension research.
The team-level evidence points the same way. Amy Edmondson's research on psychological safety — the shared sense that you can speak up, admit a mistake, or raise a problem without being punished or humiliated — links it to learning and performance, and it was prominent in Google's internal study of what made its teams effective. Fear is more or less the opposite of psychological safety: it teaches people to stay quiet, hide errors, and tell the leader what they want to hear, which is precisely the information environment in which teams underperform.
Why this feels different from how it actually is
Fear feels effective because its results are immediate and visible. A frightened team moves fast, hits the deadline, and does not push back — so in the short run a fear-driven leader can look highly competent. The costs, by contrast, are slow and largely invisible: the problem nobody flagged, the mistake nobody admitted until it was expensive, the talented person who quietly disengaged or left. The benefits arrive now and the bill arrives later, which makes fear look better than it is.
The 'feared versus liked' framing also taps a stubborn intuition that warmth and toughness are opposites — that to be respected you must be a bit hard, and that being liked means being soft and easily ignored. The two-dimension research cuts against this: warmth and competence are separate axes, not ends of one line, so a leader can be both genuinely warm and genuinely demanding. Being liked and being respected are not the trade-off they are often assumed to be.
And there is a confusion between fear and respect. They can look similar from the outside — people defer, people comply — but they produce very different behaviour when the leader is not watching. Respect tends to survive bad news and invite honesty; fear tends to produce silence and self-protection. Because the surface looks the same, it is easy for a leader to mistake the compliance fear buys for the trust respect earns.
Fear can get people to do what they are told, but it tends to cost you the honesty and trust you most need from them.
What the research says to do about it
The most consistent guidance is to establish warmth before leaning on competence or authority. Cuddy's argument is that signalling genuine good intent first — that you are on people's side and can be trusted — is what allows your competence and your demands to be received rather than resisted. That does not mean being permissive; it means people need to believe you mean well before your toughness reads as standards rather than threat.
Deliberately building psychological safety is the team-level version of the same idea. Edmondson's work suggests leaders can foster it by responding to mistakes and bad news with curiosity rather than punishment, inviting dissent and questions explicitly, and admitting their own fallibility — all of which signal that speaking up is safe. The payoff is a team that surfaces problems early, which is hard to overstate the value of.
Pairing warmth with real competence and clear, high standards is what the two-dimension research actually points to — not warmth instead of demands. The combination people respect most is a leader who plainly has their back and plainly expects a lot. Warmth without competence reads as nice but ineffective; competence without warmth reads as capable but cold. The aim is both at once.
What the research says does not help
Leading with toughness or intimidation to establish authority tends to backfire on the very thing it is trying to build. The evidence suggests projecting strength before warmth often reads as cold or threatening and makes people defensive, which buys compliance at the cost of trust, candor, and the willingness to flag problems — the things that actually drive team performance.
Chasing likability for its own sake does not help either, and is a common misreading of the warmth research. Warmth in this literature means trustworthiness and good intent, not avoiding hard conversations or trying to be popular. A leader who is pleasant but will not hold a standard or deliver difficult feedback scores high on warmth and low on competence — which the research says people find likeable but do not deeply respect.
Treating fear as a sustainable strategy because it 'works' confuses short-term compliance with effectiveness. Fear does produce obedience, exactly as Machiavelli suggested — but obedience is not candor, initiative, or honesty, and the team-performance research consistently points to those latter qualities, not mere compliance, as what separates the best teams from the rest.
Being liked and being respected are not the trade-off they are often assumed to be.
What this looks like in real life
The team that hits the deadline but hides the problem
A fear-driven leader gets fast, visible results: the frightened team moves quickly and does not push back. The costs are slow and invisible — the problem nobody flagged, the mistake nobody admitted until it was expensive, the talented person who quietly disengaged. Fear teaches people to stay quiet and tell the leader what they want to hear, which is precisely the information environment in which teams underperform.
Warm and demanding at once
A leader plainly has their team's back and plainly expects a lot. Because warmth and competence are separate dimensions rather than ends of one line, this is not a contradiction — it is the combination people respect most. Warmth without competence reads as nice but ineffective; competence without warmth reads as capable but cold. Establishing genuine good intent first is what lets the demands land as standards rather than threat.
Real numbers in context
This is an area where the honest move is to flag that the findings are robust but mostly directional rather than precisely quantified. The warmth-competence model (Fiske, Cuddy and Glick) and Edmondson's psychological-safety research are well-established in social and organisational psychology, and warmth being judged first is one of the more reliable findings about social perception. But translating 'warmth beats fear for leaders' into a single number overstates what the evidence can carry.
It is also worth being fair to the foil. Machiavelli's claim that fear is more dependable than love is not simply wrong — fear genuinely does produce reliable short-term compliance, which is part of why the approach persists. What the modern research adds is the part of the ledger he could not measure: the slow erosion of trust, candor and psychological safety that fear tends to cause, and the link between those qualities and how teams perform over time. The verdict is a lean, not a certainty: respected, warm-and-competent leadership generally outperforms fear.