Do markets inherently create bad incentives? Two recent experiments by Michele Belot and colleagues suggest so.
First, she got people to spend ten minutes identifying the country of origin of euro coins, with payment either as a fixed wage, piece-rate or as a competition with a prize for the winner. She found that productivity was 15% higher with piece-rates than a fixed wage, and 13% higher still with a pure competition. Competition, it seems, raises productivity. But it also increases cheating; subjects in ther competition over-reported the number of coins they identified much more than subjects in the other payment methods.
In a second experiment, she got subjects to make choices that affected others' payoffs. She found that in small groups where the effect of the choice on others was clear, people were more likely to behave altruistically than they were in larger groups where the harm to others - though just as large - was less salient.
All this suggests that whilst market forces have an upside - increased productivity - they also have a downside. They can encourage cheating and - insofar as market relations can be anonymous - they can crowd out altruism.
Although some people think that laboratory experiments in economics might not be always applicable to "real life", these findings chime in with real world experience. Strong competition between newspapers led to the phone-hacking scandal, whilst competition in banking has led to various episodes of mis-selling.
There is, though, a complication here. In that coin-identifying experiment, Ms Belot found that 94% of the cheating was done by just 19% of participants. This suggests that it is not the case that competition encourages everyone to cheat. For most people, social norms (or conscience) keep them honest. Instead, competitive incentives encourage misbehaviour by a minority.
The point here is not a simple one about markets and competition being good or bad. Instead, it's that they have ambiguous effects, and their desireability or not might be very sensitive to contexts, pay-offs and pre-existing social norms.