 good afternoon. I'm going to go from these these kind of far kind of broader macro kind of global looks into something that according to previous speakers perhaps a rare event in terms of actually looking a little more deeply at the kind of sorry I don't have to repeat all that do I I'm sure I didn't notice it here. So to to to look at what's happening at the kind of the very southern tip of Africa seriously sub-Saharan in terms of perhaps something characterized as being rare in terms of actually looking at internally at at tax administration data in kinds of various bits of research. So I'm going to touch on kind of sort of zooming into our administrative data giving a little example of how as you kind of dig deeper you kind of see a few more things. I'm going to talk a little bit about a project in which you and you wider is involved in terms of piloting the use of of anonymized tax data in firm level research and I probably won't get time to get to the third point. So just to kind of give you a very sort of you know kind of sketch sketch the picture in South Africa the big picture here you have what our revenue is and as a percentage of GDP so to kind of link to the previous discussions what you'll find is that those percentages as a fraction of GDP you probably won't have seen before and that's because the statistics agency the end of last year rebased GDP. So GDP of course SN8 2008 went up kind of three to five percent in levels if you're sitting at about quarter quarter quarter of your quarter of your GDP being collected revenue you get a sizable change. So again referring to to the kind of the work that has been you know the working paper the working paper 19 that you showed I mean in fact if you look at that percentage actually it looks like South Africa fits into the grouping that is called upper middle income countries rather than sub-saharan in terms of that overall shape which again I guess goes to the point of saying there in terms of tax certainly when you actually look at that data you realize that there really is no average country all countries are actually rather different in terms of tax collection. And so again say if you look at this as a kind of a time series this if you're looking at the period here sorry these are fiscal years which are shifted by a quarter relative to the calendar. You know you see as you would in many countries you see a decline in actual revenue collected and of course a kind of a much bigger effect. Of course also one of the technical issues and when in terms you come to kind of looking at data is that of course the national counts figures get compiled in terms of an accrual basis and tax agencies certainly like us we work on cash so you yeah so you're also going to get some sort of anomalies in the day but that's that's a kind of big picture. So let's look at a little more deeply in terms of where that revenue comes from. So the major component sorry I realized I didn't translate this from South African terms into more international terms. The blue line the biggest the biggest component in the country unfortunately well known for inequality but the biggest component is from income tax on individuals. Followed by VAT in general and then kind of we saw in the early in the in the sort of early 2000s up to the up to the financial crisis very rapid increase in the growth on in corporate income tax and it still in fact has not recovered to that it was pre-crisis. So again if you're kind of looking at an international comparison I mean the tax on individuals and it is our largest component and looking at kind of sort of that nice country database that we now have kind of we sort of sit just on the cusp for the bottom sort of quarter in terms of proportions of taxes on individuals to total tax. Okay so then a little bit more to look at who actually pays personal income tax and I'm just going to have to give you sort of a scale. So South Africa we sitting with a population of just over 50 million working age population just on 34 million. The average number employed according to our household labor market survey is about 14 and a half million very poor labor absorption right and of that there's there's certain a certain number of people who employed informally but if you add formal and agricultural employment you come to 11 compare that with taxpayers who are actually registered for personal income tax and we had about 14 million. I need to stop there and say that in a country with with such high inequality there's obviously a question about looking at what you can actually learn about income from from tax returns because generally I mean certainly in South Africa there's a there's a threshold above which you pay tax and there's an even higher threshold above which you oblige to make tax returns. So in terms of your actual tax returns you only going to be getting the top end. However the administrative systems have changed and in fact we do get what are called tax certificates information on any remuneration that is given by a registered employer to anybody. So in fact close down to about well and half the people who receive tax certificates so we have their information are kind of what are below tax threshold in South Africa so we are starting to actually probe the income of anybody who has some sort of association with with the formal labor market. But if we look at those who actually filed for returns we're looking at five million people this grew rapidly in fact from 1995 just just post democracy through to recent figures and we yeah so we're sitting at about five million but saying then there's this there's this interesting thing if you kind of dig a little bit deeper so going into the into the micro data I actually took a fixed number rather than a fixed proportion because I'd never seen this before and it looked like a fun thing to do and it was and so we actually just looked at the top 10,000 so that's 10,000 out of 52 million and we actually find this is the profile of how their proportion of personal income tax changes over time so actually see this steady increase actually match along the lines of taxes on on companies during that period up to the up to the up to the kind of financial crisis and a tapering off at about 7% so 10,000 the 10,000 top earners we hope unless they're people who are evading us entirely I'm sure there are actually contributing 7% of the tax on on individuals to dig to dig a bit deeper of course this is where one can in principle go in and I don't have that slide that still has to be done in terms of what it really was that drove that rapid increase was it in fact things like was it capital gains was it was it actually income from business that's something that we need to dig a little deeper not everything is done so again so you think you know as you see I mean tax taxes has a very I mean tax for revenue collected has a very close relationship to how the economy behaves over time so we saw rapid growth in tax as a whole and in person and personal income tax over the period when there's rapid economic growth but at the same time there are a whole lot of other interesting things happening in terms of the tax policy space so particularly in personal income tax there was a there were a whole lot of tax broadening policy changes so tax and capital gains was actually introduced only in 2001 and at the same time various deductions that had been previously allowed were abolished or reduced so in fact you really were just getting a broadening of the tax base as a whole at the same time there was actually considerable tax relief provided the the marginal tax rate dropped considerably but even so it was mostly amongst the low earnings that the that that the tax relief was given so that also to some extent maybe that explains partly the steady increase of the the higher the higher earners in terms of their contribution to PIT. The other thing that that that happened in the sort of what it was it was phased in several ways in several stages but in the very late 80s and early 90s as a result of several tax commissions was that South Africa had had multiple tax rates I mean if you go back to the tax tables of the 80s it was kind of astonishing you had tax rates for individuals based on race on marital status sex yeah and and kind of some somewhat inexplicable in places frankly I mean just why but anyway those those were reduced and in particular it was actually in 91 that we actually had a change in terms of going to the the last of the the taxation separately of what used to be referred as married persons married women and unmarried persons okay that became just a single tax rate and so it's quite interesting to to then track of course you know a whole lot of other things happened in the country at the same time but what actually happened to the contribution of female taxpayers the taxable income and tax contribution so you can see in terms of the numbers of taxpayers in 1991 we were sitting at kind of below 20% it's actually gone up to 43 44% so quite a dramatic change the taxable income has not grown quite so fast but it there's also well in in terms of I guess it depends what you're looking at it in absolute relative terms but I mean initially taxable income was sitting way below 15% and it's now kind of a little bit over a third of the total and of course being a progressive taxation system the proportion of tax collected on income is is is is low but is also increased appreciably and to say this does actually kind of this is something that you can actually track if one looked I just use the population censuses to look at the proportion of women employed to total employment and that actually tracks fairly closely so it's quite interesting that it's not that it that that certainly the proportion of women within the formal tax system is actually growing at kind of almost the same rate as as in employment okay within again kind of splitting looking at women and men taxpayers we've also kind of looked at how the actual tax that they the proportion borne by the top 10% in the bottom 40% tracks over time so you can see amongst women you can deduce that income is actually far more equal than it is amongst men and in fact there's been a growth in in inequality of income and hence and hence tax so another kind of progressively as you dig a little deeper you kind of get more and more sort of insights as to what is going on in in the country okay and this so then I guess this introduces the topic to to mention which is really a project that we're engaged in with national Treasury and UN UNewider is really kind of the instigator you know the the what what kind of made this happen and so we have provided roughly five years of anonymized corporate income tax returns that returns tax certificates so these are mentioned earlier these the details of the kind of the pay as you earn for each employee does include pensioners as well and so on and this has been provided to national Treasury and it's been done in such a way that that the data can be linked over time and between tax types for the same corporate entity and so on although of course corporate entities are generally different things for different tax types but that's another issue and the data has then been a made being made available to researchers for analysis within a secure environment there's been a call for proposals and these were evaluated from both a policy and feasibility perspective and so the current projects you see underway are now various things really actually sorry it's also the customs information that proved that that's in there otherwise you wouldn't have had projects like this in terms of looking at at a firm level detail from the tax data looking at markups competitiveness export growth in versus the intermediate inputs and so on and of course then also estimating worker and job flows which in the context of South Africa with persistent high unemployment is obviously a subject of a lot of interest some of our I mean the researchers and the researchers in progress but of course they've been a whole lot of learnings about the process and I think you know if we go forward again well we'll do it better so the data sets are large this brings challenges get about 20 million tax certificates a year and a lot of researchers I think have been working with this data used to working with sample survey data not actually used to working with these kinds of volumes of course we knew this from the start but it has become abundantly clear that the way data sets are managed for administrative purposes is not of course exactly how you'd want them to manage from an analysis perspective and linked to that of course is kind of the big fixing issue of metadata and and we've been actually having to build as we've gone along that this really is in some sense is very much a kind of a pioneering project I think and one one begins to realize that tax administration is complex you kind of get used to it after you've been in it for a while but it's kind of pretty messy and there's always kind of devil in the details so you know in terms of now actually wanting to try and understand even for something like a firm who what is a firm who is an employer from tax data you know there's a little bit of thinking that actually needs to be done but basically from the preliminary work that was done it was fairly early on established that creating a standard panel data set from the base records would actually be very useful for researchers in order to enable comparability of independent analysis so next time round we'll certainly do a whole lot of these things better I'm going to gloss over because I've got my red card okay so but just saying that the geographical component of tax data is also being explored by municipalities who kind of very interested in obviously in sub municipal level economic indicators and so on so in conclusion to say digging deeper and longer I think the time series I hope that some of these examples have illustrated how valuable it is to look at data over time actually enables you to track changes in policy economic and social policies as well and there's a lot more that can be done looking at job level analysis income changes with job changes firm survival analysis a whole lot of other things and of course I mean the one the one caveat is that taxpayer confidentiality has to be maintained in this so that adds a few kind of complexities but thank you