Wages Linked to Length of Service Discriminatory? An Economic Perspective

Is a compensation system that links pay to length of service discriminatory? The UK Court of Appeal thinks so, according to the October 20, 2009 ruling regarding Christine Wilson's allegations.

Ms. Wilson, an inspector with the Health and Safety Executive, claimed her pay agreement with the body was unfair because it was linked to length of service. She claimed the agreement was 'prejudiced against females' who broke their continuous service record with pregnancy. The Court of Appeal agreed and ruled that linking pay to length of service can disadvantage women who take time out to raise children. However, the Court also ruled that employers do not have to justify compensation systems that link pay and length of service unless there is evidence that doing do creates inequalities.

In his news article (here), John Lynes quotes Susie Uppal, director of legal enforcement at the Equalities and Human Rights Commissions (EHRC). Ms. Uppal said

Women should not be disadvantaged in the workforce because they take time out for maternity leave or to meet caring responsibilities.
According to Ms. Uppal, the lack of flexible working options and 'direct discrimination' exacerbate the gender inequality in the workplace.

Ms. Wilson's solicitor, Emma Hawksworth of Russell Jones & Walker, stated "employers need to give careful thought as to whether extra years in the job really do lead to better performance and if so for how long" (article on Russell Jones & Walker website).

Without realizing it, Mr. Hawksworth raises an interesting economics question: what does the relationship between employee productivity and length of service look like?

According to economic theory, the relationship is nonlinear. In other words, it's curved.

For example, assume that an employee begins a new position. Initially, small increases in length of service will lead to large increases in productivity. The new employee will accustom herself to the culture of the organization, learn how to use the organization's computer and information systems, develop a keen understanding of what is expected from her in her new position, etc. All of this "learning" happens relatively quickly. As a result of this "rapid learning", the employee becomes far more productive in a shorter period of time. During this "rapid learning" phase, the employee is in the lower left portion of the green block in the graph. The employee's productivity is increasing at a rate faster than that of length of service. For every one unit increase in length of service, the employee's productivity is increasing by more than one unit.

There comes a point, however, where this rapid productivity growth slows down. The employee understands her assignments well, knows what's expected of her, and is able to perform all of her job responsibilities at a satisfactory level. While productivity is still increasing, it's not increasing as fast as as it was, and the relationship between productivity and length of service is not as extreme as in the previous phase. This is shown in the lower left portion of the red area in the above graph. During this phase, productivity and length of service tend to 'equalize', meaning that a one unit increase in length of service yields a one unit increase in productivity.

Eventually, the employee's productivity begins to flatten out. There are a variety of explanations for this, coming from such fields as industrial organization, sociology, psychology, and neurology. The upshot is that productivity tends to flatten out and can be seen in the upper right portion of the red area in the above graph. For every one unit of increase in length of service, productivity increases by less than one unit. There may be a point beyond which productivity ceases to increase, regardless of how much length of service increases.

The question of whether a compensation system linked to length of service is 'discriminatory' is an interesting question. Consider, for example, a compensation system that gives every employee a five percent increase in pay every year, without exception. One could make the argument that rewarding employees on the basis of length of service increases early in their career under-rewards younger workers (whose productivity increases may be rising faster than length of service, meriting an increase of more than five percent) and over-rewards older workers (whose productivity increases may not be keeping pace with length of service increases, meriting an increase of less than five percent).

One could also make the argument that separation from the workforce for childbirth and child rearing occurs (for most women) during a portion of their careers when their productivity is anticipated to be increasing at the  fastest rate. Length of service-based compensation systems here would be accounting for the "gaps" in productivity due to absences from the labor force. From a purely economic perspective, this is what wages are designed to do - to reward employees for productivity.

It should be noted that women, while perhaps the largest group affected by this issue, are not the only group affected. Increasingly, men are taking leaves of absence for child care and elder care purposes. Individuals who experience an extended illness or injury are also affected by compensation systems linked to length of service.

The issues raised by the UK Court of Appeal poses some very interesting legal and economic questions. It's likely that these questions will not be confined to the UK; I expect that the United States court system will be addressing these same issues in the near future.