How do municipal-level data help us problematise the relationship between age, gender and unemployment in South Africa?
Depending on the breadth of one’s definition, unemployment in South Africa at the end of the fourth quarter of 2020 stood anywhere between 32.5% and 42.6%.¹ In comparison, the 2020 global average for unemployment is roughly 6.5%. While this particular problem is felt by the country as a whole, parts of the population bear a larger burden. Take it from members of Codebridge Youth hubs, whose deliberations led OpenUp to produce the following visualisation based on data from Youth Explorer:
What the data demonstrate are the many layers of disparities that are buried within the headline unemployment rate. For a start, as the visualisation illustrates so vividly, unemployment rates vary drastically by age: the average unemployment rate at the national level was 29.7% in 2011, whereas youth unemployment, that is for persons between the ages of 15 and 24, sat at 52.4%. This may be of concern given that the youth make up to close to 20% of the population.
A large gender gap is also evident: for the population as a whole, the average female unemployment rate was 34%, whereas the corresponding figure for males was 25%. This gap was even larger within the youth as a group, with female unemployment standing at 58% on average for the country as a whole, whereas male unemployment, while still high, was significantly lower at 47%. What’s more, according to the CEO of Harambee Youth Employment Accelerator, the COVID-19 pandemic and associated deleterious economic impacts appears to have exacerbated this already large gap, with similar trends evident among the population as a whole.
What interests us too, and something which seems otherwise under-explored, is the significant variance of employment figures at the municipal level. As the above visualisation shows, for example, some municipalities — including Phumelela, Kgetlengrivier and Tokologo — exhibit female youth unemployment rates at least twice as high as corresponding rates for males. Elsewhere, however — in places such as Cederberg and Indaka — almost no gender gap exists at all, with male and female youth unemployment being virtually the same. In Ntabankulu, King Sabata Dalindyebo and Engcobo, moreover, female youth unemployment is even lower than male youth unemployment.
What this suggests to us is that local, municipal-level factors matter a great deal. At OpenUp, we believe that data are not necessarily the answer — they can also become the question. We are going to delve deeper into these questions within the very communities in which these challenges arise. In other words, we’ll use this data to launch into a new phase of exploring, which will entail a listening more to the community partners we already have. As we currently do not have good answers to what are clearly complex questions, we aim to dig deeper into these factors over the coming months with a view to focusing on the lived experiences of young people within select municipalities.
What does education have to do with it all?
The data also pointed us to something that seems somewhat anomalous at a glance. While less female youth are ‘employed’ than their male counterparts, female youth are more likely to achieve particular levels of education. One Statistical release points out that enrollment rates across all educational types are higher for female youth. At the national level, for youth between the ages of 20 and 24, 54% of females have at least a matric education compared to 47% of males.
But at the municipal level, disparities can be orders of magnitude higher. In the Ubuntu Municipality, for example, the percentage of female youth with at least a matric education is double that of male youth. All the same, female youth in Ubuntu are still about 1.5 times as likely to be unemployed compared to male youth. On the face of it, this seems like a paradox because conventional wisdom suggests that having a higher level of educational attainment is more likely to lead to a person being employed — that seems to be true for the youth as a population group as a whole: according to StatsSA, for example, while 55.2% of all 15- to 24-year-olds were unemployed at the end of the first quarter in 2019, the corresponding figure was 58.4% for those who had not completed matric (and even lower for those who have completed some form of tertiary education). The same trend was even more pronounced for 25- to 34-year-olds.
Our first set of questions accordingly go to the relationship between education, unemployment and gender. Why is it that in some municipalities female youth generally have higher levels of educational attainment than male youth and are still less likely to be employed?
How and when do we measure employment levels at the municipal level?
Our second set of questions go to how and how frequently we measure employment levels, especially at the municipal level.
The visualisation above relies on data from the 2011 Census. This is because there are significant gaps in the data on youth unemployment at the municipal level for subsequent years. As a general proposition, StatsSA is by law supposed to conduct a population census every five years. However, purportedly due to capacity constraints, censuses are currently only conducted every 10 years. StatsSA does, however, conduct what are known as ‘community surveys’ as a sort of midpoint check between censuses, but the microdata for the 2016 Community Survey does not include data on employment status, this despite the questionnaire used for the survey having included essentially the same question on employment as the 2011 Census questionnaire. This leaves us waiting for the 2021 Census results to see whether (and possibly to what extent) things have changed.
While there are other sources that contain unemployment information, the data are usually not available at the municipal level. StatsSA’s quarterly labour force surveys, for example, seem to collect data at the metropolitan municipality level, but not at the district and local municipality levels. Given that disaggregated data are of particular importance to understanding how broader social trends are playing out at the municipal level, our hope is to get a better understanding of why data are collected so infrequently, whether there are reliable alternative sources of data that are kept more current and how this situation can be improved as a general proposition.
What is the role of municipalities (and what should it be)?
Our third set of questions go to what municipalities could possibly do to help, both in terms of youth unemployment as a general proposition and, more specifically, in relation to bridging the gap between male and female youth unemployment.
In terms of section 4(2)(g) and (h) of the Local Government: Municipal Systems Act, municipalities have a duty to … ‘promote and undertake development in the municipality’ and ‘promote gender equity in the exercise of the municipality's executive and legislative authority’. One could accordingly easily argue that municipalities are obliged, at least within their powers and financial and other constraints, to address youth unemployment generally and the gender gap specifically. The question, though, is what specifically municipalities are currently doing about these issues. For example, are municipalities allocating portions of their budgets to addressing youth unemployment and/or the gender gap? Do they have specific programmes that seek to deal with these issues?
Another question of perhaps greater importance is what municipalities should be doing and what can be done to facilitate this. We should be weary of the fact that there are clear legal and financial constraints to what municipalities are able to achieve. As things currently stand, perhaps the national and provincial governments are better equipped from both legal authority and financial capability perspectives to address these issues.
On gender, for example, there are strongly-worded rhetorical commitments in favour of gender-responsive budgeting, such as the 2018 Declaration of the Gender Responsive Planning and Budgeting Summit, and tools such Vulekamali show that gender is taken into account in national budgets, albeit that the manner in which this is done does not yet amount to a comprehensive effort to adopt gender-responsive budgeting principles as a general proposition across different parts of the budget as a whole. Tools such as Municipal Money, moreover, tend to show that there is a lack of evidence that gender is taken into account in municipal budgets.
Yet, local factors clearly matter when it comes to youth unemployment, specifically in relation to its gender component. We are accordingly interested in the extent to which issues such as these should be addressed at the local level, especially given that the national and provincial governments are obliged, in terms of section 156 of the Constitution, to hand over traditionally national and provincial functions to a municipality if a given matter would most effectively be administered by that municipality. The question of what municipalities could — and perhaps, therefore, should — do, remains fairly open to the imagination.
What issues do the data raise for you?
What questions do the data raise about age, gender and unemployment in your municipality? Use the ‘Search for a municipality’ function in the visualisations to find out! Also check out Wazimap and Youth Explorer for more data on your municipality.
For further information, contact us at email@example.com.
¹ In terms of the ‘official definition’ it stood at 32.5%, whereas it stood at 42.6% in terms of the ‘expanded definition’. In the remainder of this post, the cited employment rates are those which fit the ‘official definition’. Expanded unemployment figures would be higher.