 So I'm going to talk about the South African chapter. This paper was written together with our chair Haroun and two of our other colleagues listed here on the slide. So the main kind of idea of this chapter is to highlight some of the important trends in terms of demographics, labor market, employment, and wages in South Africa. So there are a few key trends that I'm going to highlight, but the presentation will focus on point number four, which is the role of the public sector in employment. So just to give you some background and context to the South African economy, we see that the slide shows the kind of real GDP growth rates in South Africa over time. And it's not necessarily one of the very fast growing emerging kind of market economies in Africa, but it has seen kind of moderate and consistent growth rates, particularly in the first decade after democracy. Since the kind of 2009 recession, growth, however, has been quite slow. And so in some senses, I think South Africa exhibits some characteristics of being in a middle income country trap. So this kind of moderate levels of economic growth, a high levels of unemployment, and very high levels of inequality. So it's kind of against this backdrop, then, that this paper tries to highlight some of the important trends that we think are affecting the South African labor market. In terms of the economic structure, this shows the kind of composition of GDP in South Africa. And we see that already in 1994, South Africa was a service-oriented economy. However, we had kind of substantial shares of mining and manufacturing contributing to output. By 2014, we see that the economy has increasingly then been driven by service sectors. So some of the fast-growing service sectors have been transport, storage, and communications, the financial sector, and wholesale and retail trade. Therefore, we can say that the economy is essentially a service-driven economy. So in terms of poverty and inequality, according to a kind of World Bank $2 a day poverty line, we do see that poverty has reduced since 1995 or 94. But it's kind of remained fairly high at about 26%. I guess one of the key kind of mediating factors in preventing this poverty rate from reducing further has been the very high levels of inequality. So as many of you may know about the South African society, the Gini coefficient is estimated somewhere around 0.65, so extremely high. And underpinning this kind of inequality is a labor market that's really been unable to absorb a large proportion of the labor force. So the unemployment rate using a very narrow definition is at about 25%. For the youth, this is substantially higher. And I guess the state has kind of stepped in to try and mitigate some of these challenges by implementing a comprehensive social welfare system. So there is research that shows that the social transfers in South Africa have been quite effective at reducing market inequality that would be even higher than reported here. So I'm gonna touch on this demographic transition and not go into too much detail. But essentially what we see in South Africa is that in the kind of late 1990s, early 2000s, there was a rapid growth in the size of the labor force, which was primarily driven by increasing participation rates as opposed to increases in the working age population. So therefore we've seen that unemployment growth has been unable to kind of keep up with the rapid growth in the labor force. So this is slightly different to many of the other African economies that we see and some that were presented here in that we do not expect to see rapid growth of the working age population in South Africa. In fact, the country is quite somewhere along its demographic transition. So estimates using the UN kind of population statistics show that the working age population is expected to remain at between 65 and 67% for the next kind of decade or two. These are just some estimates that our co-author did that shows essentially that South Africa is already kind of far along this period where they're expected to be a demographic dividend. So some kind of growth benefit from population change. South Africa is already in the portion of the curve where it's declining. So let's get into some of the key trends that we highlight then for the labor market. This table shows the main sectors in South Africa and the kind of absolute and relative growth in employment as well as employment shares. And the key takeaway here kind of highlighted by the colored boxes is that over this 10-year period that we look at, this date is also a little bit dated, but we see the primary, where the job losses are concentrated are in the primary sector. So agriculture and to some extent mining as well. The sources of employment growth in the South African economy then are primarily from the tertiary sector. So here we've highlighted the financial sector as well as what's called the community social and personal services, which then is primarily made up of the public sector. So again, in terms of the structural change, we see that according to output, employment is also being generated primarily by services. The second important trend has been the increasing use of temporary employment services. So this is also called labor brokers in South Africa. And these are employment services where workers are then contracted out to firms. So they're like third-party employers. And it's a little difficult to estimate this given the labor force surveys, but it's kind of hidden within a category within the financial sector, kind of an unclassified item. And we see that the TS employment as a proportion of financial sector employment has risen from about 27% to about 47% in more recent years. And as a proportion of total employment has also gone up to between six and 7%. So some of the main kind of job activities within this TS sector are things like protective services, farm hands, helpers and cleaners. So low and some medium-skilled type jobs. And essentially the benefit to firms is that they are able to circumvent some of the kind of labor regulations in South Africa. So the debate in South Africa is that this mechanism is used to overcome some of the labor market rigidities. However, when we've kind of looked at the indicators of labor market rigidity and compared them to other similar countries, we find that in terms of non-wage labor costs and indices relating to firing costs, South Africa is not really substantially higher than other middle income countries or upper middle income countries. One aspect of rigidity is perhaps in the difficulty of hiring workers. But again, it doesn't seem to be overwhelming evidence that the labor market is too rigid. So then the third important trend is related to skills-biased labor demand. So this fits in very well with the sectoral shifts, I think. Because again, what we highlight here then is if we kind of disaggregate the sexual employment by a skill type, we see that again the job losses within the primary sector are again concentrated in low-skilled and medium-skilled occupations. Where employment is being generated in the tertiary sector, it's then more in the high-skilled and medium-skilled type occupations. And this is particularly concerning in the South African context, given the kind of highly unequal education system in terms of quality. And this has been one of the legacies of the apartheid system and I think continues to be prevalent in South Africa today, where a large proportion of high school graduates perhaps do not have adequate level of skill in terms of quality of education. So then the main kind of focus of the paper moves on to the role of the public sector in employment. So one of the trends that we've highlighted is that in terms of the share of employment in South Africa, we see that the public sector employment has risen from about 14.5% in 2008 to about 17.5% in 2014. So it seems to be a source of employment growth in the South African labor market and also perhaps suggests that the state has stepped in, particularly in 2009 at a time of labor market crisis, to cushion the effects of the global recession. And post-2009, employment growth has been slow. So perhaps the state has continued to play this role since then. In terms of the composition of this increase in employment, this graph essentially shows that for the change, net change in public sector jobs in the post-2008 period, these are the occupations that kind of contributed to that change. The fastest growing public sector occupations are then quite a few low-skilled occupations. There's sweepers, cleaners, other types of helpers, as well as some medium-skilled occupations. So you have kind of police officers, primary education teachers, home-based care workers as well. So we're trying to kind of go a little deeper to figure out what are the differences then between the public and private sector employment. And this table compares worker characteristics between these two sectors. And when we refer to the private sector, it's the formal non-agricultural private sector. And essentially what we find then is that public sector workers are an average older. There's also a larger proportion of women that work in the public sector relative to the private sector. And there's a larger proportion of Africans working in the public sector relative to the private sector. And this is important because it also shows that the state has been able to transform its workforce at a faster rate in the private sector, as these are kind of two previously disadvantaged groups in South Africa. The public sector workers are also kind of have a higher average level of education measured by years of schooling. And they are substantially higher proportion of unionized workers in the public sector. In terms of the kind of occupational distribution, it seems that the public sector has a slightly more compressed occupational distribution, so not as many workers right at the top end and also not as many workers at the bottom kind of occupational levels. An important and interesting trend then within this whole context of rising public sector employment has also been a rapidly rising public sector union density. And what's also interesting here is that the private sector has shown kind of the completely opposite trend in South Africa. So the private sector union density has declined since 1997, whereas the public sector union density has increased. So kind of the latest estimates we have is about 70% of public sector workers belong to a union. So we want to then compare the wage distributions between the public and private sector to estimate a kind of wage premium. And this is a simple plot of the wage distributions between these two sectors. And it seems that on average, both median and mean wages in the public sector are higher than the private sector. If we then disaggregate these distributions by union status, we see that for non-unionized workers, it seems that this wage premium doesn't really exist. However, for unionized workers, they're still higher average wages in the public sector relative to the private sector. So in the interest of time, I'm not gonna explain in too much detail what we do, but it's a standard kind of two-stage heckman employment model to estimate the wage premium in the public sector. So in the first stage, we correct for selection into the labor market and then estimate a standard earnings function. And this is the main table of results from our analysis. And what we get out of this is then, in the first two specifications, we just merely control for union status. And what we see there then is that there is no average public sector wage premium. However, there is a positive and significant union premium. However, when you interact union status with either public and private sector employment, we see that for the unionized public sector worker, there is a positive and substantial wage premium. And this holds true when we separate out public sector employment into government and state-owned enterprises. In terms of, sorry, I should add that we've obviously controlled for all worker characteristics and done the usual kind of diagnostic tests. In terms of magnitude then, what we see is that in this interactive specification, for non-unionized workers is actually a wage penalty for working in the public sector of about 18.5% on average. But for unionized public sector workers, there's a positive and significant wage premium of about 21%. So therefore, when you control for TES employment, there's kind of no average public sector wage premium. However, belonging to a union then results in a significant and large public sector wage premium. And I think this is kind of something new and interesting that's perhaps been looked at a little more in recent years. We've essentially done the same thing across the wage distribution, running these quantile regressions. And we find that the premium for unionized workers is consistent across the wage distribution. And it's highest at the median worker. The wage penalty for non-unionized workers seems to also exist kind of across the wage distribution. So we've also then, I guess, from this try to model a labor market segmentation model, kind of using simple multinomial logic kind of function. And I think it confirms our result that there is this wage wedge between the unionized public sector worker and the rest of the kind of formal economy in South Africa. And the segmentation model also then confirms that they are distinct characteristics as distinctly different characteristics of workers in the public and the private sectors. So to conclude then, I think what we see in the South African economy and the labor market is that both output growth and employment growth is increasingly being generated by the services sector. There is also then a rise in the use of labor brokers. We also then highlight this kind of important rising share of public sector workers and employment along with this idea that for the unionized public sector worker there's a significant wage premium. So it starts to suggest that perhaps in the post 2000 era there's been a new labor market elite in South Africa, the unionized public sector employee. And I think for us this is particularly worrying in a context where there is unequal education system where firms are trying to avoid direct employment where manufacturing has essentially stagnated and the share of manufacturing and employment has declined. I think this new trend is particularly worrying. Haroon presented some updated work two days ago and I think it also confirms this idea that if you look at wage growth across the distribution you see a hollowing out in the middle of the distribution. So essentially there's some wage growth right at the bottom which is attributed to minimum wages and then wage growth right at the top and we think that part of this is the public sector wage premium. So yeah, I think there's some important and difficult problems ahead in the South African labor market.