Inequality & performance

Inequality hurts. That's the message of this new paper from Bruno Frey and colleagues. They looked at how US basketball players and German bundesliga footballers' performance is affected by pay inequalities. And they found that players who earn less than their team-mates tend to perform worse, even controlling for ability. Low pay relative to one's peers reduces one's performance, even though rising pay across the board improves it.
This has implications outside of sport. It suggests that a firm that pays a big bonus to a "star" performer - or just gives him a pay rise to try and keep him - might see its overall performance worsen, as its less well paid employees work less well in response. This effect could outweigh the tendency for the well-paid employee to respond positively to the higher incentive.
What's not so clear, though, is the precise mechanism. Do relatively poorly paid players consciously under-perform out of resentment? Or is it instead that pay is a way for employers to convey expectations? Low pay signals that a boss doesn't expect his employee to perform especially well, and people have a way of living up or down to expectations?
Whatever the mechanism, this suggests that conventional thinking might not be the whole truth. It's commonly thought that people are poorly paid because they are unskilled. But could there be an element of reverse causality here - that people are unskilled because they are low paid?

Comments

[kinglear]

I think the incentive bit is the important part . If you all start out on the bottom rung ( OK I know people hire in superstars but just say..) one will emerge at the head. He gets paid more. It just might be he is better at what he does than others. That being the case, the also rans are never going to catch up.

[Rob Knight]

There is probably considerable subtlety here. It's not hard to imagine an alternative scenario in which the high pay of the superstar spurs others on to match his achievements and thus earn the same amount. This is the standard assumption that most economists would make, as far as I am aware.

Perhaps this works differently for footballers though? At their level of income, does an extra 20 grand a week make all that much difference if they're already on 50 grand? Conversely, the 'work hard and you could get what he has' incentive might function better at low pay levels, where the difference made by even a small pay rise is considerable.

[Simon]

Watch the performance of the LA Galaxy this year. Because of salary cap everybody except Becks is on peanuts. Prospects - not too good.

[Kevin Carson]

I've seen a pretty conscious connection in my work life. At the hospital where I work, orderlies went three years without even a COLA raise. I heard more than one person say things like "I earn my pay, but ten bucks doesn't buy as much work this year as it did last." I've also seen orderlies systematically fail to swipe barcodes to charge supplies to patients, and let the hospital eat the cost. "Let them take it out of what they saved cancelling our PTOs [paid time off] last December." (By the way, exactly four months after they cancelled PTOs, supposedly to stave off bankruptcy, management announced it had leased a corporate suite in the skybox at the local baseball stadium). Corporate management would be amazed at how much better the numbers looked if the workforce didn't utterly loathe them.

[jameshigham]

It suggests that a firm that pays a big bonus to a "star" performer - or just gives him a pay rise to try and keep him - might see its overall performance worsen...

Oh yes, undoubtedly, particularly if it is just one or two star players and then if their attitude to teammates is somewhat summary.

[Bruce]

Most players (and workers) know they are never going to be star players. So the incentive factor in seeing someone else get a big rise is next to nil, or even negative as none of us like to be reminded of that fact. Football teams, like most companies, require teamwork to succeed. Few stars will get very far without solid supporting performances from their colleages. And this is more so for some sports (and businesses) than others - where a team is only as strong as it's weakest link. Successful football teams will spend as much time looking for weak links in the opposition as they do dealing with their opponents star players.

[reason]

Incentives have nothing to do with payments to star players, it is quasi-rent. As for the other players, I wonder how the performance relative to ability is measured. It could be simply that the difference in payment affects the co-ordination between the "ordinary" (really merely very talented) players and the stars. That is, if the difference in payment is too great all the play will tend to concentrate on the well paid star to the detriment of the other players.

[kio]

"It's commonly thought that people are poorly paid because they are unskilled."

Might be my rewording allows to see the process from the other side.
It's commonly thought that people are poorly paid because they are CONSIDERED BY OTHER PEOPLE unskilled.

I have a quantitative justification of this approach explaining personal income distribution in the USA.

[a very public sociologist]

Don't repeated studies demonstrate that income differentials arising from performance-related pay show an overall drop in productivity and staff morale?

[buy propecia]

As for the other players, I wonder how the performance relative to ability is measured. It could be simply that the difference in payment affects the co-ordination between the "ordinary" (really merely very talented) players and the stars.