Job Performance

What is the relationship between age and job performance? The average age of people in the workforce is getting higher, with increasing numbers of middle-aged and older workers employed in many different jobs (Fullerton; Johnston and Packer). Thus, it is important to know whether job performance is higher or lower for older workers in comparison with younger workers. Most reviews of empirical research on this issue have concluded that although individual studies differ, averaging across available studies reveals virtually no relationship between age and job performance (McEvoy and Cascio; Rhodes; Salthouse and Maurer; Warr). The fact that there is no observable relationship is interesting to many people because it is known that age-related declines can occur in important mental and physical abilities (as documented elsewhere in this encyclopedia). If abilities that are important for performing work do decline with age, but job performance is not lower for older workers, this seems paradoxical.

When considering this issue, it is important to realize that two different but related questions can be asked about the job performance–age relationship. First, one could ask whether, at the same point in time, younger workers in a given job perform differently than older workers in the same job. This type of question is answered through cross-sectional research in which people of different ages are compared against each other (for example, people age twenty-five versus people age sixty). However, a second question that can be asked is whether workers who are twenty-five years old today will perform better or worse after thirty-five years in a particular job. That is, would performance in that job increase or decrease over time for the same people? This type of question is answered through longitudinal research, in which the same people are tracked across time to observe age-related changes in behavior. Perhaps for obvious reasons, research conducted on the age–job performance relation has thus far been largely cross-sectional (Warr). It is difficult and expensive to conduct research on the same people across many years, and of course people change occupations, switch companies, and take other actions that make this impractical. Therefore, there is not a great deal of data on whether job performance changes with age. There is mostly research to show whether there are differences across age groups in the performance of various jobs. And cumulative research shows that there are essentially no differences in observed performance as a function of worker age.

In reaching this conclusion, it is important to realize that, other than the possibility that there may truly be no relationship, there are several other possible reasons why there is no observed relationship between age and performance in research (McEvoy and Cascio; Rhodes; Salthouse and Maurer; Warr). First, in work organizations job performance is very important. Thus for a variety of reasons, people who are not performing well generally do not stay in a job. People may quit or find a job that they can perform better, rather than stay in a position that exceeds their capabilities. An older person who finds that he or she cannot perform the job anymore may decide to seek a different one. Also, those who cannot perform a job may be removed by the organization. They may be transferred into a different position that is more suited to their skills, or they may be terminated. The end result of either of these processes is that if older workers cannot perform well because of age-related changes in their skills, they probably will not be in the job for very long. Consequently research comparing older and younger workers in performance will not include those older workers who cannot perform and are no longer on the job. The end result will be no observed relationship between age and job performance.

A second possible reason why there is not an observed relationship between age of a worker and performance is that there may sometimes be nonequivalent responsibilities between people within the same job. That is, although the job title may be the same, the duties or responsibilities may be somewhat different across workers. If older or younger workers get different assignments (for example, easier or harder ones or assignments that rely on certain skills to different degrees), then making meaningful comparisons of the job performance of older and younger workers is difficult.

A third possible reason why there is no observed relation between job performance and age is that job performance measures often rely on supervisory ratings, which are imperfect measures of performance. To the extent that these ratings are not accurate measures of performance, conclusions about job performance may not be accurate. Various rating biases and errors can occur in performance appraisal. For example, if a supervisor wants to be lenient or give ratings that are higher than a person deserves (a leniency bias), then the performance measure will be biased upward. To the extent that this or other types of biases influence ratings, the measure of performance may not reveal true performance differences very accurately. If there are true age-related differences in performance, they may not be reflected in the measure.

A fourth possible reason why there is no observed relation between job performance and age is that the true relationship may be curvilinear rather than linear. Most research on the age– performance issue has tested for linear relationships, meaning that as age increases, performance consistently increases or decreases. However, if the relationship is actually curvilinear, this would mean the relationship has a different form than simply higher or lower with increasing age. For example, job performance might increase up to a certain age (e.g., fifty), then level off and begin to decline at later ages (at age sixty and over), becoming increasingly lower after that. If job performance is plotted on the y axis against age on the x axis, the relationship would look somewhat like an upside-down U. (This is only a hypothetical example to illustrate curvilinearity—research has not established this trend conclusively.) If the research has been done to detect linear relationships, and the relationships are actually curvilinear, then it is possible that the linear tests will show no relationship when in fact there are curvilinear ones.

A fifth possible reason why there is no observed relation between job performance and age is that job performance is almost always multidimensional, meaning the work requires more than one important type of behavior or skill for overall success. Thus if some workers focus their efforts on one part of a job and other workers focus their efforts on other parts of a job, their overall performance may be similar, even though they reached the performance level through allocating their efforts differently. If older workers' abilities or skills decline in one aspect of work, they may focus their efforts on another aspect to achieve the same overall level of success.

A sixth and very important possible reason why research shows no relationship between job performance and age has to do with job experience. Job experience usually leads to more job knowledge (Schmidt et al.). It makes sense that someone who has been around longer will acquire more knowledge that is relevant to the task(s) at hand. Research has also shown that job knowledge is a very important predictor of job performance (Schmidt et al.). This also makes sense: those who know more about how to perform a job tend to perform it better than those with less knowledge. Thus experience leads to more knowledge, and job knowledge leads to higher performance. Age is often positively related to job experience (older workers have been around longer, on the average). This means that older workers (with more experience) may have acquired more job-relevant knowledge than younger workers (with less experience). Therefore any age-related mental or physical declines in basic abilities needed to perform the work may be offset by greater job knowledge that goes along with greater experience. As a result older workers may be able to compensate for less raw ability through the use of more job knowledge. If this occurs, then observed relationships between job performance and age may be zero because the overall performance of younger and older workers is actually equivalent, but for different reasons. That is, younger workers are relying more on their raw ability and older workers are relying more on their accumulated knowledge.

Also, Warr suggests that in many jobs declining abilities can be compensated for by other factors, such as more experience, more knowledge, taking more time to complete demanding duties, and so forth. Further, there is little evidence that the type of job in question moderates or makes a difference in the size of the job performance– age relationship (McEvoy and Cascio). Thus it is possible that in many jobs, a decline in abilities can be offset by experience and knowledge.

However, this does not address the question of whether two workers of equal experience and unequal ages will perform similarly. That is, although age may be associated with experience in many settings, it may not be in all settings. Less research has addressed the question of whether an older worker and a younger one assigned to a brand-new job at the same time will perform similarly. In fact, much of the cumulative research done on the age–job performance relationship has not controlled for differences in experience across age. Thus the effects of age on performance by workers of equal experience but unequal age is not as well researched as the overall question of whether there is a general association between age and job performance. This is an important question because the nature of work has been changing, with an increasing number of workers being called upon to learn continuously, often changing career lines for various reasons (Hall and Mirvis). Increasingly the idea of someone working on the same job from age twenty-five through age sixty may be unlikely (Greller and Stroh). This raises the question of whether job experience can accumulate and have the same compensatory effects if workers are continuously facing changing circumstances.

To the extent that performing work successfully will increasingly involve learning and developing new skills, learning and development will be an important concern for aging workers. Some literature has noted that older workers may not participate in training and development experiences to the same extent as younger workers, and has made suggestions on how to increase motivation and participation in these activities (Maurer). However, there has not been as much research on these issues as there has been on the job performance–age issue. As rapid changes in technology and in business strategies demand new skills of workers at midlife and beyond in order to continue to perform their jobs, it is important that older workers (and younger ones) continue to be involved in learning and developing their skills at work. Further, there is potential for age discrimination in dealing with performance-relevant training and development of workers (Maurer and Rafuse). There could be a tendency to give preference to younger workers over older ones in assigning or encouraging training and development opportunities. Increasingly, as organizations manage older workers, this legal concern is a critical issue to deal with effectively for the sake of both the organizations and the workers. Some literature has suggested how to avoid these personnel and legal problems (Maurer and Rafuse).

Most research so far indicates that the age of a person, by itself, has little real meaning in explaining job performance. However, other than the possibility that there may truly be no relationship between age and performance, there are several other possible reasons why there is no observed relationship between age and performance in research. It is also important to note that age is only a number that is very loosely associated with many other more meaningful (and potentially job-relevant) changes and differences in people. Much research remains to be done on this general issue.