Senior Economist at the Nevin Economic Research Institute (NERI), Lisa Wilson, highlights a lack of consensus on how the gender pay gap should be estimated and outlines her belief that the excuse of complexity should not limit policy action.
The question that is usually to the forefront of public and policy discussions of the gender pay gap is ‘what size is it?’ or in other words, ‘how unequal are the earnings of males and females?’. Less frequently asked and reflected upon, are the more basic questions of what is the gender pay gap? And how do we measure it?
These questions form the basis of a recently published NERI study looking at the gender pay gap across the island of Ireland titled, How Unequal? The unadjusted gender pay gap in Northern Ireland and the Republic of Ireland.
There is a general air in public discussion around the gender pay gap that ‘everyone knows what the gender pay gap is’ and that the only point of contestation is whether or not it is large enough to be an issue which should be of significant policy concern. In following how this discussion plays out on social media platforms, or even on webinars dedicated to the issue, this debate translates back across the screen as being played out amongst five rival opponents. Each are staunch that their position on the issue is one of absolute fact.
The first is one which holds that the gap in pay between males and females is so large that it merits being considered as one of our most urgent social issues. The second is the opposite point of view holds that the gender pay gap is completely made up. This position was captured well recently by an online poll by TIME magazine of over 9,000 Americans which found that close to half of male respondents see any estimates of the gender pay gap as representative of ‘fake news’ rather than a legitimate issue. The third is one which holds that the only information of interest when it comes to the gender pay gap is the extent to which any differences in pay are owing to gender-based discrimination, and so equate the gender pay gap with the concept of ‘unequal pay for equal work’. The fourth is a position which holds that even if the gender pay gap is real, it’s not an issue worthy of policy action, because we can explain the gap through the fact that males and females engage differently in the labour market. The fifth, is a combination of the other viewpoints, and posits that there is no real gender pay gap, but rather there is a motherhood pay gap. From this point of view, any discussions around the gender pay gap are misplaced because they are not accurately diagnosing the issue.
The NERI took interest in these differing viewpoints precisely because whilst on face value it appears as though these opposing viewpoints are born out of conclusions on estimates of the gender pay gap, what is more accurate is that they are a corollary of the fact that there is little to no consensus on how we should estimate the gender pay gap. And so, it is more accurately the case that dependent upon which estimate of the gender pay gap one is presented it is not at all that unreasonable that one would arrive at one of the various different abovementioned conclusions on the matter.
In this regard, because there is no consensus on how we should measure the gender pay gap and often little to no clarity given by those whom are presenting the estimates on just what exactly their measure is capturing, conversations around the issue are often muddied from the outset. Thus, when it comes to debates and discussions around the gender pay gap more often than not, the debate is being played out through different songs, albeit to the same tune.
And so, at the NERI we thought it worthwhile to go back to these very basic questions and sought to bring some clarity to the debate in terms of ‘what is the gender pay gap?’ and ‘how do we measure it?’.
“When it comes to debates and discussions around the gender pay gap more often than not, the debate is being played out through different songs, albeit to the same tune.”
The gender pay gap is a concept that is broader than the concept of ‘equal pay for equal work or work of equal value’ and should not be equated such. However, just because it is not about unequal pay for equal work does not mean that there is not merit in assessing the gender pay gap. That is, as well as capturing any differences in pay owing to discrimination the gender pay gap also captures the fact that: (a) different characteristics are rewarded differently in the labour market such as for example, different levels of educational attainment; and (b) males and females have different characteristics. And in broad terms, the gender pay gap gives an overview of how males and females are rewarded differently on the labour market.
There are numerous different ways to measure the gender pay gap and various different decisions need to be made in doing so. These differences in measures and the fact that there is little disclosure of this fact in estimates of the gender pay gap may go some way to explaining why some equate the issue with fake news. The gender pay gap is oft depicted as a discrete single, easy-to-understand concept, and so these disparities in estimates come across as unreliable and at the more extreme end made-up. Alas, that is untrue, but not being forthright about the limitations of one’s measures does create this perception.
There are two overarching ways to estimate the gender pay gap. When the gender pay gap is calculated by simply comparing the pay of all males to that of all females the estimate is known as the ‘unadjusted’ or ‘raw’ gender pay gap. Calculated in this way, the gender pay gap does not take into account all of the different factors or characteristics of males and females that may play a role in explaining the differences, such as for example differences in education, occupation, hours worked etc.
On the contrary, when the gender pay gap is calculated after accounting for underlying differences in characteristics the estimate is known as the ‘adjusted’ gender pay gap. Those of the viewpoint that the gender pay gap is not an issue worthy of policy action, because we can explain the gap herald the adjusted gender pay gap as the measure of prestige. However, just because we can explain some of the gap does that make the gap any less real?
When seeking to estimate the gender pay gap, the earnings period for which the earnings data covers (hourly, weekly, annual earnings) also matters, a lot. In this regard, whilst it is most common to see the gap based on the difference in hourly earnings we need to remain cognisant of the fact that when a woman applies for a mortgage or a car loan, she is not asked about her hourly income. The income statistic that affects whether or not she gets the loan, and indeed what kind of life she is able to afford, is her annual income.
Furthermore, when looking at differences between men’s and women’s pay there is a choice as to which point of the earnings distribution we seek to compare and whether estimates of the gap in earnings are based upon differences in mean or median earnings. Depending on which measure we use we tend to get somewhat different estimates of the size of the pay gap.
Northern Ireland’s statistical body NISRA tends to prefer the median, the Republic of Ireland’s statistical body the CSO tends to prefer the mean. Because there are generally fewer women in higher-paying roles than men, the gender pay gap as measured by mean earnings arguably better captures how the structure of the labour market affects gendered differences in pay.
There is a tendency in discussions of the gender pay gap to prefer exact ambiguity. However, policy action in the area will only have limited success while we accept the excuse of complexity.