In the late 1990s, the city of Albuquerque had a problem it needed to reduce the amount of overtime for garbage truck drivers. As is typical, city leaders turned to incentives to solve the problem. They allowed workers to go home if they finished their routes early and still get paid for 8 hours. Overtime decreased but audits revealed several problems. Drivers were driving over the legal weight limit because they didn't want to take time to go to the dump. They were speeding and having more accidents. They weren't picking up all the trash on their route and they weren't making the necessary repairs to their trucks. To quote the Albuquerque Journal, "the unintended results of the incentive program could be an increase in safety risks, cost of operations, legal liabilities and customer dissatisfaction."1 This is probably not the outcome the city had in mind.
There are countless stories like this one, some much more high-profile—like the financial crisis or the Atlanta teacher cheating scandal—that illustrate the dangers of incentives and performance-based pay. Despite stories like these, performance-based pay is not on its death-bed, quite the contrary. It is widely viewed as the "go-to" strategy for motivation, performance, and organizational change. After reflecting on my own experience living with and studying the effectiveness of these programs, I believe they are fraught with peril and should be scaled back or eliminated in most organizations. I also believe there is little chance of this happening in the near-term. Why? There are several reasons.
Everyone has one. At last check, 91 percent of companies link employees' pay to their performance.2 My hunch is the other 9 percent can't use these programs, but wish they could. It doesn't seem plausible that everyone is wrong...does it?
Problems aren't fatal. Companies don't go out of business because their performance-based pay programs aren't working. The problems they cause are annoying, more a "pebble in the shoe" of organizations, not a "broken leg". We also think these problems are solvable, the result of a poor design (not enough money to differentiate adequately) or poor implementation (supervisors don't have the skills or the courage to differentiate). It is also easy for leaders to rationalize the "noise" that accompanies these programs. They create winners and losers and of course the losers will complain. It is impossible to please everyone.
Employees want them. How else would they want their pay determined? Management scholar Herb Meyer captured this sentiment over 40 years ago: "It seems logical to base pay upon performance. In fact, the principle of merit pay is so logical, that it seems almost ludicrous to criticize it."3 In fact, research shows we are biased to assume everyone is extrinsically motivated; we are just giving employees what they want and expect.4 Most people think they are above average anyway and assume these programs will benefit them
We assume the principles are sound. We are prisoners of our own flawed assumptions, beliefs and mental models. We learned in basic psychology and economics classes that people are motivated by money and tying rewards to individual performance will improve results for people and organizations. In fact, organizational change today has become a matter of simply "getting the incentives right". The depth of our convictions keeps us "spinning the wheel", searching for the right combination of incentives. Chris Argyris called this "single-loop" learning. We try something and if it doesn't work, we try something different. We need more "double-loop" learning in organizations. When things don't work we step back and examine the beliefs, values and assumptions underlying our tactics. Are there more effective ways to motivate employees and improve performance?
It turns out our beliefs and assumptions betray us. Despite what many of us believe and what many writers tell us, performance-based pay turns out not to be broadly beneficial for employees or organizations, nor it is seen as widely effective in other contexts—healthcare, education and public sector management for example. In organizations these programs work in a narrow range of circumstances: Simple jobs; jobs where performance can be measured quantitatively; jobs that are boring or noxious; and jobs where paying attention and working harder can improve performance. These programs have not been broadly successful in improving performance, reducing turnover or improving creativity and innovation. And even when they do work, they can have undesirable side effects and unintended consequences like increased competition, reduced intrinsic motivation, and other bad behavior like that witnessed in Albuquerque.
I would argue we get the performance we do from our people and organizations despite the performance-based pay systems we have in place, not because of them. What's worse is we may be creating and perpetuating this problem ourselves. A study by Stephen Burks and his colleagues showed they could take perfectly cooperative bicycle messengers and turn them into competitive egoists by implementing a performance-based pay program.5 Michael Lewis strongly reinforced this point when he placed the blame for the financial crisis squarely on incentives, suggesting these programs have changed the values of people working on Wall Street.6 We may very well be teaching people to be self-interested and extrinsically motivated, which in turn reinforces the need for more performance-based pay programs.
There are alternatives. We could base bonuses on team, organizational, and company performance (e.g. gain sharing and profit sharing programs). We can base salary increases on market factors. We can develop more differentiated job structures that give employees more opportunities for promotion and to increase their pay over time. Finally, we can tap into other ways to motivate employees, focusing on needs like purpose, mastery, autonomy and belonging. These solutions require us to abandon our existing beliefs and assumptions. They are not as easy as throwing money at people but they will probably be more effective in the long run. Imagine all the things you could do if you weren't always tinkering with your performance-based pay programs?
1 Ludwick, J. (2004). How to collect overtime in under 8 hours. Albuquerque Journal, January 30th. http://www.abqjournal.com/news/metro/139316metro01-30-04.htm
2 Institute for Corporate Productivity (i4cp) (2011). Tying pay to performance. Research report.
3 Meyer, H. (1975). The pay-for-performance dilemma. Organizational Dynamics, 3, 39-50.
4 Heath, C. (1999). On the social psychology of agency relationships: Lay theories of motivation overemphasize extrinsic incentives. Organizational Behavior and Human Decision Processes, 78, 25-62.
Beer, M., & Katz, N. (2003). Do incentives work? The perceptions of a world-wide sample of senior executives. Human Resource Planning, 26, 3, 30-44.
5 Burks, S., Carpenter, J., & Goette, L. (2006). Performance pay and the erosion of worker cooperation: Field experimental evidence. IZA Discussion Paper # 2013.
6 Bloomberg Business Week (2013). Michael Lewis on the next crisis. http://www.businessweek.com/articles/2013-09-12/michael-lewis-on-the-next-crisis