Pay for performance sounds good, until you think about it. I added a comment to an item on HRM Today this week suggesting “pay for performance” may not be the best way to manage people’s compensation. This is a complex area and sometimes what seems simply intuitive turns out to be a poor approach under closer examination.
Pay for performance has been debated for decades, but it was W. Edwards Deming who demonstrated that the performance of the individual is highly dependent on factors external to them. The systems by which the work is done as well as the expectations of the employee, management, and the organizational culture are the biggest determinants of individual performance, and the worker’s ability to influence most of these factors in most cases is small.
In a collaborative and enabling culture the individual may be able to affect those factors, but will rarely have enough influence on them to justify individual performance as a sufficient basis for compensation. Compensation based on organizational performance where the contribution of the individual to those results is fairly clear can be a better motivator, but this still depends on those external factors and the management savvy required to be aware of and manage them. Savvy, like common sense, isn’t common, unfortunately, which is why pay for performance does poorly in most organizations that try it.
On the downside, pay for performance can put employees (and sometimes departments or divisions) at odds with each other, especially in tough times where budgets for pay increases and other rewards are limited, and this can compromise overall efficiency and results. More detailed analyses have been written, so I won’t go farther, but for these reasons I disagree with #5 “pay for performance” as a good approach in most circumstances.
Evaluation of the worker is difficult, and creates issues. The biggest problems with “pay for performance” are the need for evaluation of the employee and the fact that merit bonuses will usually come from a fixed budget. First of all, the evaluation process is easily skewed, or can be seen to be skewed, by the subjectivity of the individuals involved, personal or organization-related agendas, and a variety of other factors that have nothing to do with the performance of the individual. There is no way I have seen or know of, of ensuring objectivity in such a system. The simple measurement of the work itself is usually insufficient, as even a count of parts produced, for example, will be affected by wear and tear on the equipment used, late component deliveries, holidays, company celebrations or meetings, just to name a few factors that will affect output. With more complicated work the challenge of how to measure performance is even greater, and most evaluation systems become increasingly subjective. Various schemes have been tried, such as “360 evaluations” that include not only supervisors but peers, “customers and suppliers”, and they seem better, but still have significant risks and imperfections.
Standard business practices such as budgeting cause systemic problems with pay for performance schemes. Secondly, since pay increases are almost always tied to a budget, this can put employees in competition with each other. I can also remember cases where an organization was having tough times and the employees, called on to put in extra hours and effort, performed admirably almost to a person only to find that because of the fixed budget for increases they had to each take an average rating in their record and a mediocre increase, a fact which made everyone angry. The supervisors themselves would have received poor evaluations if they went over-budget. In the end, in that case, pay for performance made everyone unhappy and less committed to the organization, certainly not the desired result.
Bonus programs, while often used, have significant shortcomings. It may be that pay based on results of the organization as a whole, often implemented in Western businesses as an annual bonus program, can work effectively. Unfortunately a bonus can be seen eventually by workers as a regular part of their pay, and many will fail to see how their individual efforts contribute to it, detracting from its potential to motivate better performance.
The simplest approach may be the least costly, and work as well as any other. Deming and others who examined this problem in the past found that pay based on time spent on the job was, for most positions, a sufficient and less-potentially damaging approach. It leaves the success of the organization as a factor if management regulates pay according to overall profitability of the company, for example, but avoids costly and problematic evaluation systems.
The handling of pay issues is important for companies, but it is tough. While workers should be paid a fair and competitive wage, and this must be determined with an eye towards retaining them while maintaining the profitability of the organization, this area is as complicated and ridden with pitfalls as any other area of human relations. Since most organizations have practically no ability to assess their real costs around such matters (this is a very difficult undertaking), and the cost of a less-than-optimal pay determining system can be huge in terms of organizational efficiency, it seems the best approach is probably the simplest and lowest cost: pay appropriate for the position and competitive for the industry, modified by time spent on the job.
As always, I solicit your comments. Thanks for reading – Tim