Arrest the income gap bandit

Take a holistic approach to measuring income inequality.

Income inequality is still the elephant in the room. It is by far the biggest time bandit on Remuneration committees today and will continue to be for the foreseeable future.

The South African economy has historically been afflicted by income and wealth inequality. Given the income inequality and socioeconomic conditions in South Africa, it is not surprising that income inequality within companies has been the topic of much debate. The instrument most commonly used when measuring income inequality within a company is the Wage Gap.

There are many different angles to the wage gap and many different methods of calculating it. The aim of this article is to demystify these angles and methods, outlining the key features of each, and provide discussion to assist with the selection of the most appropriate measure.

Wage Gap

Although there have been a number of different methodologies used to calculate a “Wage Gap”, the methodology used by the United States Securities Exchange Commission is calculated as:

Wage gap calculated by USSEC

A, B and C band employees covers unskilled, semi-skilled (operational) and skilled (advanced operational) – stopping just short of managers and known collectively as the general workforce. This calculation provides the user with the ratio of how many times higher the CEO’s pay is than the median worker’s pay. The benefit of using this methodology is that it is simple to compute and easily understood. There are however weaknesses in the Wage Gap methodology, which include:

By simply halving the CEO’s pay, the ratio will halve even if there has been no gain to the worker’s welfare; and
The grade of the “median employee” is unknown and therefore an organisation with a high number of lower level employees is more likely to have a higher wage gap.

There are a number of measures of income inequality which can be calculated, all with their own benefits and limitations.

10 -10 Ratio

The 10-10 Ratio is often used as a crude measure of inequality within macroeconomics when all the data needed in order to calculate a Gini Coefficient is not available.

Calculate a Gini Coefficient

The benefits of this methodology include:

– It’s easy to calculate; and
– It’s more inclusive than the standard Wage Gap in terms of how the numerator is calculated.

The disadvantage:

– It doesn’t provide information regarding the other 80% of the sample.

Gini Coefficient

This measure is traditionally used in macroeconomics and measures the inequality within a distribution. In layman’s terms, it measures the area between the “Line of Equality” (when everyone earns exactly the same) and the Lorenz Curve which is the cumulative distribution of all incomes in the sample, ranked from lowest to highest. The greater the distance (area) between the Line of Equality and the Lorenz Curve, the more unequal the income distribution is.

The Line of Equality and the Lorenz Curve

Benefits include:

– It makes use of the entire population.

Associated weaknesses are:

– A small sample may result in biased results;
– It can be computationally difficult to calculate;
– It does not provide a measure of inequality within sub-sections of the population; and
– It counts the unemployed in the calculation, which is one of the reasons the Gini coefficient is so high in South Africa.

Coefficient of Variation

The Coefficient of Variation provides a measure of inequality between sub-sections. In statistics, it is defined as a normalised measure of dispersion within a sample and is calculated as:

The Coefficient of Variation

Benefits include:

– It can be used to calculate inequality within each grade or other sub samples;
– It can be used to measure macro and micro levels of inequality; and
– It is easier to calculate than other methodologies which analyse inequality within sub-samples.

The disadvantages are:

– A large enough sample within each sub-sample is required; and
– If the distribution of the data is far from a normal distribution, it can produce biased results.

The Coefficient of Variation has the potential to provide the most holistic view of income inequality of all the measures discussed. It is clear that each of these methodologies has its own strengths and weaknesses and, as a result, a “multi-measure” rather than a “single measure” approach is required in order to obtain a holistic view of income inequality.


The Wage Gap provides a good starting point for addressing income inequality within an organisation but it does not provide a holistic view. Several alternative measures of income inequality have been discussed, each with its own strengths and weaknesses – none of which can provide an accurate holistic view in isolation. This leads one to conclude that in order to obtain a holistic view of income inequality within an organisation, a number of inequality measures should be used in conjunction in order to obtain the true picture of income inequality.

To fix this is not easy. If one halved the CEO’s pay of every single company and distributed it to the works, they would each get only a few Rand each. We need to think about structural methods to address this, one of which is to give employees shares in the companies they work for. However, the government and DTI need to assist companies to spread the wealth and companies should be given tax incentives to do so.

Bryden Morton has a B.Com (Hons) Economics and is a Data Manager at 21st Century Pay Solutions Group and Chris Blair is the CEO of 21st Century Pay Solutions Group, Chris has a B.Sc in Chemical Engineering. Dr Mark Bussin is the Executive Chairperson at 21st Century Pay Solutions Group, a Professor at University of Johannesburg, Professor Extraordinaire at North West University, Chairperson and member of various boards and remuneration committees, immediate past President and EXCO member of South African Reward Association, and a former Commissioner in the Office of the Presidency.

This article appeared in the July 2015 issue of HR Future magazine.

Read Previous

Fair does not necessarily mean equal

Read Next

Reclaim relevance and drive value