Does competition make people perform better than cooperation?

The belief

Many schools, workplaces and sports clubs run on a simple idea: rank people, reward the top, and everyone tries harder. Leaderboards, class rankings, sales contests and "employee of the month" all assume that beating someone else is the strongest motive we have.

Researchers have tested that idea for about a century, mostly in classrooms, because classrooms make it easy to set up the same task under different rules.

Three ways to set a goal

Psychologists David and Roger Johnson, at the University of Minnesota, sort any task into one of three structures:

Structure	How it works	Example
Cooperative	You succeed only if the group succeeds.	A team project marked as one piece of work.
Competitive	You succeed only if others fail.	Marking on a curve; only the top three get a prize.
Individualistic	Your success has nothing to do with anyone else's.	Pass if you reach 70%, whatever others score.

What the pooled evidence shows

In 1981 the Johnsons and colleagues pooled 122 studies, with 286 findings, comparing these structures in North American samples. Cooperation produced higher achievement and productivity than either competition or working alone.

In 2008 a larger review focused on early adolescents: 148 studies, more than 17,000 students, in 11 countries. Cooperative goals went with higher achievement and with better relationships between students. The two moved together: students who got on better also learned more.

A 1995 review by the same group pooled 46 problem-solving studies published between 1929 and 1993. Cooperative teams beat competing individuals on every type of problem: written and spoken, mathematical and practical, clearly defined and open-ended. The result held at every age and in weaker and stronger studies alike.

Two cautions. First, the Johnsons also train teachers in cooperative learning, so they are not neutral observers, and much of the research comes from their centre. Second, classroom studies measure classroom outcomes. They say less about, say, a sprint final. Critics such as Alfie Kohn, whose 1986 book No Contest reviewed the wider research, argue the pattern holds well beyond school; that remains an argument, not a settled finding.

Same benefit, different cost

A 2024 study at the University of Turin tested the question in a new way. People did a demanding attention task in virtual reality, either cooperating with a partner or competing against one, while researchers tracked heart rate and asked how stressed they felt.

• Cooperating with a weaker partner lifted performance.
• Competing against a stronger opponent lifted performance by a similar amount.
• Only competition raised stress and heart rate. The pattern held when the sessions ran longer.

So competition can work. It just charges a physical price that cooperation does not. The study ran in the lab over short sessions; what that price does over a school year or a career nobody has measured this way yet.

Where competition fits

The research does not say competition is always bad. The Turin study found it lifted performance, and games and sport can be fun when losing costs little. The evidence argues against competition as the default way to hand out grades, pay and respect, because the same gains come cheaper through cooperation.

What you can do

1. At work or school, check what the reward structure actually is. If one person's gain is another's loss, expect hoarding and stress.
2. Set goals the group reaches together, then check that each person contributed. Group work fails when one person does it all.
3. Mark against a standard, not against each other. A pass mark lets everyone succeed.
4. If you coach or teach and want a contest, keep the stakes low.
5. Watch your own body. If a contest keeps your heart racing, the Turin result suggests working with someone may get you the same gain for less strain.