Solidarity Research Institute

A brief critical analysis of the 12th Commission for Employment Equity report

September 2012

Introduction

The 12th report by the Commission for Employment Equity (CEE) was released on 11 September 2012 (see here - PDF). This report is based on the data submitted to the CEE by large employers (150 or more employees) only. It contains data for 2011.

As in previous years, the CEE has presented a picture of very slow racial transformation with regard to the progress of employment equity and affirmative action in South Africa, especially with regard to the upper levels of employment. In her speech at the presentation of the report, the Minister of Labour, Ms Neliswe Mildred Oliphant invoked a passage from the Freedom Charter that reads: "All laws which discriminate on grounds of race, colour or belief shall be repealed." It is truly ironic that these words were quoted in a speech extolling the virtues of the Employment Equity Act, a piece of legislation that once again discriminates according to race.

The approach followed by the CEE is once again a dogmatic adherence to the racial composition of the economically active population (EAP) of South Africa as the only yardstick for measuring the success of employment equity, while any and all other factors that may influence employment equity or the measurement thereof are excluded. This approach is not only extremely one-sided, it is also not aligned with Section 42 of the Employment Equity Act.

Furthermore, certain problems with the methodology followed by the CEE have remained unchanged. While there seems to have been some improvement, the report is not free of basic errors.[1]

This brief critical analysis elaborates on some of the points that have been mentioned.

The pace of racial transformation

In the introduction to the report, the chairperson of the CEE, Ms Mpho Nkeli mentions the following:

"Whites and males will continue to dominate in the middle-to-upper levels for the next 127 years as long as employers are caught up with the vicious cycle of continuing to employ people with mainly the same race and gender profile that just exited their organisations."[2]

This "127 years" figure was also used by Ms Nkeli in 2011, where she was quoted in Business Report[3] that, at the present rate, it would take 127 years for top management to represent the demographics of the EAP [economically active population] where 77 percent were black.

Firstly, note that where the "127 years" was in 2011 used to refer to top management level only, it has now suddenly conveniently been expanded to include "middle-to-upper levels". This is extremely misleading, apart from the fact that the "127 years" figure as used in 2011 was already incorrect and misleading in its own right when it was used in 2011. For an institution such as the CEE, which is supposed to provide accurate figures to inform policy makers, to play fast and loose with figures in this way is inexcusable.

In 2011, Solidarity pointed out that this quote was inaccurate on two counts, firstly, the reference to 77% was incorrect, as it could either be 73,6% (EAP for "African" people) or 87,9% (EAP for the generic "black" group), but not 77%. Secondly, while the methodology for deriving the "127 years" figure was unclear, it seemed to be based on a misinterpretation of the CEE's own figures. 

When the CEE's own figures from as far back as 2000 were taken into account, it emerged that from 2000 to 2010, white people at the top management level had decreased by an average of 1,44 percentage points per year, which, when compared to what the CEE regards as being a "representative" level for white people, would mean that if the pace of change remained the same, white "representivity" would be achieved at the top management level in 42 years, not 127.[4]

This is still an overstatement, as the assumption that the pace of change would remain at 1,44 percentage points per annum is tenuous in the face of dynamic factors like different and changing age profiles of the different population groups, as well as the changing dynamics of educational attainment and experience. These factors can be expected to contribute strongly to the acceleration of the pace of change, making 42 years also a gross overestimation.

The rate of decline for white men (the only non-designated group) at top management has also been even faster than for the white population group as a whole at this level.

The fact that the "127 years" figure has been trotted out once again this year, is disappointing. It is not based on figures from this year's report, which begs the question: Why keep repeating the research if you are going to continue using figures from previous years?

Furthermore, the latest CEE report includes data on a total of 5 174 860 employees. Of this total, only 19 352 were in top management. This is about 0,4% of the total. Only 10 679 of the top managers in this report are white men. If they had only a 6,6% share of top management, equal to their share of the EAP, there would have been 1 277 white men at top management level.

This means that fewer than 9 500 positions would open, as it were, for "racial transformation". This is an extremely small number of positions in the entire labour force for the CEE to be so concerned about. Even if the 31 332 white men out of the 71 415 people at senior management level were added too, it's still a minute portion of the total labour market.

The following table shows the changes in the percentage share of white men at the different levels of employment from the CEE's data from 2001 to 2011.[5]

The general downward trend can be observed in all categories. The "skilled" group is particularly important, as this is where the bulk of white men are employed. In the latest CEE report, 44,7% of white men covered by the report were employed at this level.

Comparing 2003 and 2011, shows a decrease of 8,6 percentage points in eight years, which amounts to a 39,3% decrease in the percentage share of white men at this level of employment in eight years. If one instead takes 2007 as a starting point, as the CEE does in its latest report, the change is 7 percentage points and 34,5% in only four years.[6]

White men may still be, when EAP is used as the only yardstick, "overrepresented" at the skilled level, but this has been very rapidly changing, if the CEE reports are to be believed. A decrease of more than one third in their percentage share at this level in only four years represents very significant change, despite what the CEE may believe.

Another way to illustrate this change would be to consider racial breakdown of the CEE's figures where all the levels of permanent employment are consolidated. The earliest available figures that are roughly comparable are from 2003.

This table shows that where white people in 2003 accounted for 22,9% of all employees covered by the CEE's report, in 2007 they accounted for only 19,1% and in 2011 for only 14,7%. This is also indicative of a large shift in employment when analysed by race.

This table also shows that the CEE's oft-repeated mantra that white women (in 2011 and this year also including Indians) "have benefited more from affirmative action" is, when the big picture is considered, false. White women's percentage share of the total has decreased from 9,8% in 2003 to 8,3% in 2007 and 6,5% in 2011. The decrease was less drastic for the Indian group, from 5,0% to 4,6% to 4,4%, but it is still a decrease.

Note that the group with the largest increase was "African" women, from making up only 18% of the people covered by the CEE report in 2003, they have rocketed up to 27,4%, putting all other groups in the shade. This has not only been due to "African" women entering the labour force at unskilled or semi-skilled levels. The rate of increase of their representation has been fairly similar over all the levels of employment. For example, "African" women's share of senior management almost doubled from 2003 to 2011, from 4,0% to 7,3%. While the absolute percentages are still low, the rate of increase has been very rapid.

By now it should be clear that by only considering the top levels of employment, which represent a very small proportion of the total labour market, and by using EAP as the only yardstick, the CEE misrepresents and misinterprets the information contained in its own reports.

The role of education

While the CEE does not include any figures or comments on the distribution of educational qualifications in the South African population, it does in a few instances, refer to the need for skills development and training.[7] While this is very much a muted theme in the CEE's report, it should actually be the main theme of the entire report.

Even the Employment Equity Act, in section 42(a)(ii), stipulates that one of the factors that must be considered when compliance with the act is being assessed, is "the pool of suitably qualified persons from designated groups from which the employer may reasonably be expected to promote or appoint employees."[8]

In reality, this is a complex exercise, as suitability of a particular person's educational qualifications and experience for a particular job has to be considered on a case by case basis. However, by considering the overall situation with regard to the spread of educational qualifications in South Africa, a large part of the reason for why employment cannot realistically reflect the total economically active population at every job level in South Africa at this point in time.  

The above graph may provide some clues as to why white people (and to some extent Indian people) still account for proportions of the upper levels of employment that are so much higher than their share of the EAP.

The graph shows that, according to the 2007 Community Survey, the most recent large-scale survey among South Africans conducted by Statistics South Africa, white people, in that year accounted for nearly 50 percent of the population aged 20 and older with a degree or post graduate diploma as their highest qualification. Indian people held a 6,5% share, which is also significantly in excess of their share of the EAP. At higher levels, from a honours degree upwards, white people's share increases to 58% and Indian people's share increases to 7,4%.

In a situation where more than half the share of qualifications at degree level and higher is held by one group, it can surely not be surprising that that same group accounts for the lion's share of employment at the higher levels of the labour market? This is not due to some sort of malicious racial discrimination plot by businesses, but it simply reflects the reality that educational qualifications are not evenly spread among the different population groups.

Consider a hypothetical situation where there had not been any discriminatory policies in the history of South Africa, but where white people, through some random chance, had ended up with the majority share of educational qualifications. In such a situation, if white people did not "dominate" the higher levels of the labour market, people would be confounded by such a strange state of affairs.

And yet, in reality in South Africa, in its report, the CEE does not mention even the possibility that the difference in educational attainment among the different race groups as a large part of the explanation for why white people hold such large shares of the employment at the higher levels of the labour market.

The same data can also be presented in a different way:

This graph clearly shows that the educational attainment profile of the white population and, to a lesser extent, the Indian population is significantly better than the black African and Coloured groups.

These proportions have been (and are continuing to) gradually changing over the last two decades or so. However, this process is gradual. The large educational deficit among especially the black African and Coloured groups takes time to change. It is an unfortunate reality that it is very unlikely that a 50 year old person with less than matric will, in his or her working lifetime, manage to matriculate or gain an equivalent qualification and even more unlikely that he or she will progress to degree level. However, this person remains part of the labour market (and the EAP), but is extremely unlikely to attain a position at the upper levels.

No figures on work experience at different levels are available, but it is reasonable to suggest that most black people have less work experience at higher levels than their white colleagues of the same ages. To a great extent, this is a result of the policies of the past. While these policies were unjust, it is unfortunately not possible to make up for lost experience overnight.

Of course, there certainly are exceptions to the situations outlined in the above paragraphs. There are black people who did and do manage to triumph over adversity and attain very good qualifications and experience, or who manage to surmount these obstacles through sheer business acumen and talent in the absence of formal qualifications. Such cases are, unfortunately, still a minority of the black people in South Africa (as among any group of people in any society in the world). If it was easy for anyone to become a senior manager in a business without the relevant qualifications, universities would have been out of business long ago.

More problems with EAP as yardstick

The aim of this section is to briefly touch on some of the reasons why EAP in and of itself cannot be used as a precise measurement of what each race group's "representivity" should actually be. This is not an exhaustive list, but only explains some of the most salient points.

The EAP is derived as part of the Quarterly Labour Force Survey (QLFS) that is regularly undertaken by Statistics South Africa. This is a sample survey, which uses estimates of the total population to estimate the relevant figures for the country as a whole. The estimates of the total population that are used, use the 2001 census figures as a base population, with various other data sources and assumptions factored in to estimate the current population in every specific quarter.

The fact that the base population that is used is from a decade ago naturally creates challenges with regard to accuracy of the current estimates. For this reason, the EAP figures cannot be regarded as a perfect representation of reality for each quarter.

Furthermore, the QLFS makes absolutely no distinction with regard to citizenship or any other such factor that has a bearing on whether a person is part of the "designated group" or not. The "black African" category in the QLFS, for example, includes any "black African" immigrant that has entered South Africa after 1994. In terms of the definitions used for the "designated group", such a person is to be considered a "foreign national" and should not be included in the "African" category. However, the yardstick that is used, EAP, includes such a person in the "African" category.

Because the QLFS racial definitions and the definitions used for determining whether a person is part of the "designated group" or not are different, it further limits the degree of accuracy with which the EAP figures can be applied.

The way in which the EAP figures are used also do not pay any regard to the differences in the age structures of the various population groups. For example, the "African" population of South Africa has proportionally much more young people than the "white" population. For a job that requires, for example, a three year qualification and 10 years of experience, only people of at the very least 31 years old can be considered. Using total EAP in isolation for judging "representativeness" for such a job, is improper, as the actual potential pool of people that can be employed has a larger proportion of white people and fewer "Africans" as the required age for the job rises.

The misinterpretation of turnover figures

It is unfortunate that this section has to be included in this brief critique of the CEE's report once again, as Solidarity has pointed this mistake out before, both in 2010 and in 2011. This has even been pointed out during a meeting with the CEE, but the problem still has not been rectified.

As has already been mentioned, the CEE uses the racial composition of the EAP as the absolute and only yardstick for measuring all of their figures. If a certain figure for white people is higher than the white share of EAP, the CEE has repeatedly stated that this is extremely negative and requires strict legislative intervention to correct.

Throughout the report, this is the approach that is followed with respect to the figures on recruitment, promotions and training.

For example, at the senior management level[9], the CEE notes in its latest report, with respect to the figures in the following table: "Whites and males are still dominating in terms of representation at this level. This scenario is likely to remain as long as Whites continue to dominate opportunities pertaining to recruitment, promotion and skills development opportunities."

If one were to compare these figures on recruitment and promotion to EAP only, the conclusion reached by the CEE seems to be valid. It is, however, an incorrect comparison to make.

The proper comparison is between the proportion for the specific group that is currently employed and the proportions of recruitment and promotions. If the recruitment or promotion figures for white men are higher than the proportion that is currently employed, it indicates that any individual white man at that level is actually less likely to be promoted or recruited than any individual black person or white woman. The figures in the table above therefore actually point to a gradual erosion of the "white male dominance" at senior management level.

To illustrate this principle, the following hypothetical example can be considered. Take a situation where a company has 1 000 employees at senior management level. Of these 1 000 employees, 700 are white and 300 are black. Now the company promotes 100 employees. Of those that are promoted, 60 are white and 40 are black.

The CEE would look at this and say "White people are more likely to be promoted than black people." or "White people are getting more than their share of promotions." According to the CEE, in this example, where 60% of promotions went to white people, "white dominance" would be further entrenched. However, this is incorrect.

Actually, in this example, each individual black person of the 300 is more likely to be promoted than each of the 700 white people. 40 promotions out of 300 means that 13,3% of the black people were promoted, while only 8,6% of the white people were promoted. As long as the percentages of promotion and recruitment for white people stay at a lower level than their current representation at that level, it actually gradually erodes white representation.

This situation is exactly what one sees in the figures from the 12^th CEE report, as would be expected, when looking at the decreases in white representation over the past years.

The absurdity of the CEE's approach can be illustrated by a further simple example:

In a specific company, there are 35 black people and 65 white people at a specific level. In one year, the company decides to promote all 35 black people. If they were to stick to the EAP-proportions, this would mean that they could only promote five white people in that year, to achieve (roughly) an 88/12 split. This would keep the CEE happy and the net effect of the change would be that 40 employees are promoted.

However, if the company instead decided to still promote all 35 black employees, and also promoted 35 white employees instead of only five white employees, the split would change from 88/12 to 50/50. This would make the CEE very unhappy, but what has actually happened? Was this a negative development for any black people? No, they were all promoted in either example. The only difference was to the white employees, who, in the second situation were positively influenced without disadvantaging anyone. Still, according to the CEE's approach, the first situation would be preferred.

If one were to, as the CEE consistently does, use EAP as the only yardstick by which to judge the fairness and desirability of the two alternatives above, the first one would be judged fair and equitable, even though the second alternative treated all the black employees exactly the same, but increased the net positive effect on employees.

This also speaks to the effect of the dogmatic adherence to EAP at an individual level. The promotion of 35 black and five white employees would be judged to be fair according to EAP at a group level. Say, however, that there were actually ten white employees who were equally deserving of a promotion because of their hard work and dedication? Five of them would have to be told that they could not be promoted, because by promoting them, group-level equity (as determined by a bureaucrat) would be negatively affected. This is the kind of unfairness at individual level that is created by a dogmatically numerical approach to affirmative action as advanced by the CEE.[10]

There is another important aspect to employee turnover that the CEE seems to fail to grasp. The CEE's approach to employee turnover supposes that every white person who leaves an employer, will not find employment another employer, or will/should only find employment only at a lower level than before.[11]

In reality, especially at the higher levels, this is not what usually happens. It is rare for someone to voluntarily leave a management position if he or she does not already have another job at the same or higher level lined up.[12] Therefore, most of the CEE's figures on "terminations" of employment and of recruitment of new employees are probably double-counting the same individuals. It is therefore not reasonable to suggest that it is simply racism by employers that prevents them from filling each position that is vacated by a white person with a black person. The turnover of employees takes place precisely because the employees vacating their jobs know that they can be employed at the same or an improved level at another employer - they already have the new job "lined up" before resigning. If this were not the case - if, for example, white employees were told that if they resigned, they could not be employed at another company at the same level - the number of people vacating their jobs would be significantly lower.[13]

Summary and conclusion

As in previous years, by misinterpreting and misrepresenting its own data and omitting relevant factors, the CEE has once again presented a skewed of racial transformation in the South African labour market.

While this critique of the latest CEE report is not exhaustive, it has mentioned some of the most salient points - summarised below:

1. The chairperson of the CEE's reference to a period of 127 years before whites and males will stop "dominating in the middle-to-upper levels" is incorrect and misleading.

2. The report's inclusion of historical data only as far back as 2007 creates a skewed picture of the changes that have actually taken place over a longer period.

3. The CEE makes no mention of the fact that the data from different years isn't strictly comparable, due to the methodology that is employed for data collection.

4. There is an excessive focus on the upper management levels, which constitute only a very small part of total employment.

5. The CEE does not mention the fact that educational attainment varies among the different population groups and the role that this plays in explaining why white (and Indian) people have such relatively large shares of employment at the higher levels.

6. EAP itself is, for several reasons, a flawed yardstick to use when measuring "representivity" and can actually only be used as a broad guideline.

7. The CEE uses the wrong yardstick when interpreting its figures on recruitment and promotions. It uses the EAP figures for comparison, while the current representation figures at each specific level of employment should actually be used. When properly regarding these figures, the conclusion is that white men's percentage share of employment at high levels is gradually decreasing, exactly the opposite of what the CEE finds.

It is extremely disappointing and unfortunate that the CEE persists in creating confusion and misperceptions among the South African public and among policy makers by misrepresenting its own data.

Footnotes:

[1] For example, in the first table in the appendix to the report, a figure of 11 470 is given for coloured women at semi-skilled level, while it should actually be 117 470.

[2] Emphasis added by SRI, not present in the original.

[3] Donwald Pressly, 2011. Firms ‘need job equity fines'. Business Report, 4 August 2011. Accessed online on 4 August 2011 at http://www.iol.co.za/business/business-news/firms-need-job-equity-fines-1.1111509

[4] For more detail, see the Solidarity Research Institute's document from 2011, entitled A brief critical analysis of the 11th Commission for Employment Equity Report.

[5] It must be kept in mind that due to the methodology that the CEE uses to collect data, and methodological differences over the different years, none of the years are actually strictly comparable to each other. No other comprehensive data are available, however. Also keep in mind that the figures from 2001 are even less comparable to the other years, as the format of the 2001 report limited the level of detail of data that can be gleaned from it.

[6] Even though white women are part of the designated group, the CEE has often lamented the fact that they too remain "overrepresented". The decrease for white women at the skilled level (where 45% of white women are employed according to the latest report) from 2003 to 2011 is 5,6 percentage points (34,4%) and from 2007 to 2011 it is 4,6 percentage points (30,1%). In percentage terms nearly as drastic as the changes for white men.

[7] For example on page 3 of the 12th CEE Report.

[8] Subsections (iii) and (iv) also refer to the economic and financial position of the employer and employers in the same sector. This serves as an acknowledgement of the fact that implementing employment equity can come at a cost to the company. The CEE also obliquely acknowledged this in its 11^th report, where Ms Nkeli said "Tough trading conditions and the negative effects of the recession must not deter people from putting pedal (sic) in order to drive and accelerate transformation." If it was not possible that implementation of employment equity could be costly, why would the possibility that a recession might deter people from implementing EE be mentioned?

[9] This is not limited to the senior management level - the CEE uses the same incorrect interpretation at every level in the report. Senior management is merely used as one example here.

[10] This situation is similar to that which Solidarity is contesting in court on behalf of members of the SAPS who are the most deserving of promotions, but who are not promoted in order to adhere to numerical targets of affirmative action.

[11] Ms Mpho Nkeli refers to the opposite of this as a "vicious cycle of continuing to employ people with mainly the same race and gender profile".

[12] Exceptions would of course be permanent retirement, emigration or death.

[13] The racial composition of the supply of skills, as outlined in section 3 of this document, also plays a role, of course.

Issued by the Solidarity Research Institute, September 13 2012

Click here to sign up to receive our free daily headline email newsletter