Arnold Kling  

Pay for Performance

Jackals in Retirement... Three Quotes from the Kinderga...

Bruno S. Frey and Margit Osterloh give a number of reasons to question the effectiveness of performance-based compensation systems. One of the reasons should be familiar to readers of this blog.

It would be naïve to assume that the persons subjected to variable pay-for-performance would accept the respective criteria in a passive way and fulfil their work accordingly. Rather, they spend much energy and time trying to manipulate these criteria in their favour. This is facilitated by the fact that employees often know the specific features of their work better than their superiors. The wage explosions observable in many sectors of the economy can at least partly be attributed to such manipulations, eg when managers are able to contract easily achievable performance goals.

When a remote authority sets incentives, people respond by manipulating the system. This fact is poorly understood by education reformers who are fond of pay-for-performance and national standards, by health care reformers who are fond of paying for quality, and by financial regulators. In fact, the quoted paragraph provides an excellent description of the financial regulatory process under risk-based capital. The banks spent much energy and time trying to manipulate the risk-based capital regulations in their favor. They got what they wanted, in terms of risky portfolios backed by little capital.

The Hayekian story here is that effective compensation practices require local knowledge and tacit knowledge. In a large company, you give a middle manager a fair amount of discretion in compensating his or her staff. If instead you try to implement an automated bonus system, you will get gamed.

Unfortunately, other people commenting on this article have stressed other issues that the authors raise with pay for performance. In particular, people are attracted to the behavioral economics stuff about non-monetary compensation. I think the Hayekian issue is the most important, and it has the most implications for (taking a more humble approach toward) public policy.

COMMENTS (11 to date)
Brock writes:

So is there a better alternative?

And isn't market action "pay for performance"? Isn't that how my relationship with my landscaper works? He mows, I pay. And also with my clients - I consult, they pay.

Mr. Econotarian writes:

With the exception of performance bonuses for salespeople, I've found most corporate performance incentives to be nearly useless. They rarely effectively reflect on individual employees, but are typically linked to company-wide or division-wise performance, which does not link well into individual performance.

Where they do reflect individual performance, they often do so "on a curve". Even if you, as a manager, feel that all your employees are doing an excellent job, you typically have to choose one to get "an A" and three to be "a B" and five to get "a C", etc.

Salespeople also game their bonus system a great deal as well. They tend to be recompensed for revenue, not profit, so they will sell things cheaply to expand revenue (or sell something actually impossible to deliver).

I think the art of proper compensation is an evolving one, and best worked on by the direct manager than anyone else.

Rick Hull writes:


Market action and rigid, "automated" performance scales both attempt to reward merit. Market action uses distributed knowledge, while performance metrics and weightings are a centralized, planned approach. By giving the middle manager more leeway, the knowledge and ability to incent is more distributed.

Shangwen writes:

I did an evaluation of P4P as an option for our health-care organization (both as cost-saving or quality-enhancer), and the finding was very negative. The reasons outlined by Frey and Osterloh were among them, but those are just the tip of the iceberg. Some others:

- Individuals have a natural peak point in their skill. Ability improves with diminishing returns, but bonus pay is arithmetic.

- Health care isn't as scientific as we think. We lack tools to measure effectiveness reliably in 80% of what we deliver, and even where we have tools we don't always use them.

- Even if you had the tools, the IT and administrative costs to implement them would be daunting, if the technology existed.

- People game the system in other ways, such as "cream-skimming", or turning away complex or chronically ill patients.

- Few systems have the ability to manipulate pay and prices as finely as would be needed to make it work.

- The best work and the biggest improvements are found in teams, not individuals. And calculating team bonuses is hard (and, where you have unions, illegal).

- A Hayekian reason: people who are institutionally encouraged to be hubristic really hate being evaluated.

I'm also thinking of an incident I reported here.

I think P4P, at least in health care, is craziness. If there is a big problem--high medical errors or a big wait list--then you call a meeting and say, "People, we have a problem we need to fix. Who has some thoughts?" What you don't do is spend a zillion dollars on some software that you think is going to measure everything even though you lack the tools. And the idea that more pay is going to work has been, I believe, long disproven both practically and empirically.

Dr. Kling, you or someone else like you needs to write a book about the wonderful economics of just plain muddling through.

Hugh writes:

I am a big fan of partnerships: at the end of the year the partners, who are also the owners of the business, divide up the year's profit.

If a partner wants to earn more, the firm has to make more profit. Excess risk taking is discouraged as the partner also has is equity at risk.

perfectlyGoodInk writes:

In a large company, you give a middle manager a fair amount of discretion in compensating his or her staff. If instead you try to implement an automated bonus system, you will get gamed.

Well, this isn't a perfect solution either. Anybody who has worked in the private sector for a while -- particularly in large companies -- knows that the name of the game is "playing politics." When your manager has a fair amount of discretion in compensation, your manager's perception of you and your performance becomes much more important than your actual performance.

Market competition checks this somewhat, as theoretically, managers who figure this out and take steps to be impartial ought to get better results. However, companies generally don't have the time luxury to be scientific in their decision making and isolate variables one by one (there are many variables that affect a team's performance), and they also never have a control group to compare against. The only comparison is the competition, and I've yet to see a market large and varied enough where two companies differ only in middle-management techniques (which is also one of the differences that is least visible to one's competitors anyway).

Also, the market incentive to solve a problem is greatly diminished when the market as a whole has failed to solve it, which is one way to explain why discrimination persisted far longer than theory would predict. And finally, there is also a pretty powerful counter-incentive for managers to be impartial: many managers like being treated nicely, and perhaps even consider this to be a perk of their power. After all, power corrupts is an axiom that also applies in the business world, and middle-managers, of course, face the exact same political game with their own bosses.

Manager discretion is no magic bullet. It can be gamed just like any other system. Instead, I think the best way to set compensation relies upon a mix of rules and discretion, so that they each can check the weaknesses of the other.

steve writes:

Hey Arnold, I just noticed the widely-unread Unchecked and Unbalanced is included in the new Ebsco ebooks subject set for economics.

Whitaker writes:

Why don't people read the original article in full? The full article provides a very different context that changes the pro-Hayekian slant drawn by this blog.

The first two paragraphs of the original article read:

"Scientific literature has extensively dealt with variable pay-for-performance. Despite the fact that serious problems linked to this approach have thus become obvious, many authors continue to support compensation according to predetermined performance criteria because they are committed to the traditional concept of the ’homo oeconomicus’.

Overall, there has been a marked change of opinion in academia (see for instance Bryson and Freeman 2008 on this site). The idea that people are solely self-interested and materially orientated has been thrown overboard by leading scholars. Empirical research, in particular experimental research, has shown that under suitable conditions human beings care for the wellbeing of other persons. Above all, they are not solely interested in material gains (see eg Frey and Osterloh 2002). Recognition by co-workers is greatly important. Many workers are intrinsically motivated, ie they perform work for its own sake because it is found challenging and worth undertaking. This applies not only to qualified employees but also to persons fulfilling simple tasks. They often are proud of their work and performance."

The article suggests three alternatives:

  • The employees have to be carefully selected.

  • Employees have to be paid a fixed compensation corresponding to their performance.

  • Awards can be used to enhance employees’ work motivation.

Arthur B. writes:

This is related to Goodhart's law's_law

Mark writes:

As a teacher who works under an automated bonus system ("Aspire" in Houston), I can confirm that the pay-for-performance has a very limited effect on teacher performance. Great teachers are great because they derive self-worth from rising to the challenge of educating kids. Bad teachers are bad because they game the system and skate by on doing as little as possible to keep their job.

I would love a bonus for being a superior teacher, but not because it would motivate me to work harder. It would however, be a way for a school to communicate to me what my value is, and differentiate the bad teachers from the good. I think it would lead to increased retention and attract better people into the field. To quote "Moneyball": It's not about the money, it's what the money says about you.

Teachers want to feel valued and appreciated. A tiered compensation structure for teachers who take on more work and have more of an impact would accomplish that. But it would require training and empowering principals to evaluate their teachers properly, rather than play politics.

Automated bonus systems are indeed bogus.

Mike Rulle writes:

As long as we look to strangers hundreds or thousands of miles away to solve micro local problems or seek improvements in micro local living standards, the more we are subject to absurd outcomes. Its always adverse selection---

I wonder why this happens in all advanced economies?

Comments for this entry have been closed
Return to top