 thank you so yeah my name is Takwanisa and I'm going to present this paper by Shifa and colleagues so I'm here just presenting on there on their behalf okay okay like the previous presenters we we are comparing our country individual country story about inequality to the weed companion database and how the to how they come back so the outline of the presentation is yeah just like previous presenters we starting providing an introduction and highlighting the objectives and then we tell this inequality story about South Africa in our case and then we come back that to the weed database and then conclude on the comparison so as been highlighted before the main problem is it is a challenge to combat existing like several surveys from even from the same country even surveys that fall in the same series so this again is the same problem for South Africa as well so the objective of this exercise is to explore in detail the methodological challenges that are found in the household surveys the various household surveys from South Africa and in this case we are focusing on surveys from 1993 where the first representative household survey was conducted the PSLSD that you see presented in sections to come so using micro data to examine mostly income inequality then we compare these estimates with the weed companion estimates as well so what do we know about inequality in general in this by way of background if we look at the graph to the left showing the distribution of inequality across several countries as you can see the the countries highlighted in there shows that most country inequality is high among countries in the sub-Saharan Africa and South Africa is topping that list but when we compare inequality of access to resources we see a different story where other countries emerge like Mozambique and in EJ they have high inequality to access to resources when you come based out Africa which is on the on the lower end in that comparison so different inequality measures they tell a different story then this is these are some results from some existing work that we're doing within ASA showing inequality here from this any income inequality from 2006 to 2015 and we see we can observe a certain drop in the inequality from 2006 to 2009 and it becomes stable but when we compare the expenditure share decels for for the same years it's looking at 2006 and in 2015 we see they follow the the same distribution so for this particular case they say it is difficult to tell what is causing the sudden drop in inequality when you compare with other measures we also they also look at access measures of access to basic services so in this case they are saying in other than water you see electricity improved sanitation in internet there is a ketchup over time these are comparing differentials between rural and urban and over time they are all catching up kind of highlighting that they there is a reduction in equality even by raw ebb and differentials so in this table they show some of the terraces that we were used in the weed to to extract in inequality are estimates from the weed but in their case they started from 1993 where they believe the terraces are a national representative and as well when you compare to other data says that or estimates that were used in the weed that starts from the 1960s okay so in this case they show some yeah some data says for example some data says that were not included in the weed but they can also allow you to to estimate inequality I'm not sure yeah if I need to highlight anything I hope to remember that okay so yeah this is the the narrative from the weed estimates where the data sources are the PSLSD from 1993 in the 1996 census 2000 IE as the income and expenditure survey the 2001 census and again the IE as in 2005 and then with national income dynamic study estimates from 2008 to 2017 basically what they are showing that okay so from from this narrative basically they showing the same pattern right but in their case in 1996 they used a another survey but one of the issues they highlighted is in 2000 they using the IE as as well as in 2000 as well as in 2000 and 2005 right and the results show that even inequality using the same series of surveys the inequality increased between the over that period but whilst you're looking at it in general they have the same pattern they follow the same pattern so when you look at the full distribution or the full estimates from the 60s as used in the weed companion it is it is difficult to believe like this trend that is decreasing from the 1960s up to 1990s then they start increasing so which is the reason why they choose starting from 1993 using the PSP LSD survey okay so here we're comparing two sets of results the needs estimates and PLSG results in this case the surveys where the instrument are sort of comparable so the needs and the PLSD instrument is comparable and then the IE as in the LC as as well is comparable these are estimates from Jenny estimates from income and when you compare the pattern again is expenditure where we observe roughly from 93 when you're looking at income inequality a general decline from 68.8 to 66.3 and again other than the AI they the 1995 one we kind of have the same story where income inequality is also falling over time okay then in this case this is comparing the the weed standardized genie coefficients and the weed original these are the genome coefficients that extracted from previous surveys and then the gray one is showing the ASA estimates these are estimates that are committed by the ASA team for the South African case so just to highlight that the difference between these two is that they compared two years where for example you have the same estimate we have estimates of the genie and if they overlap and there's a difference then they will relief to the other one so that is comparable over time okay so other than these in this case they use the 99 the weed uses the the 1996 census but in this case they they remove the 1996 census and used one of the surveys and now I can't remember the exact survey but other than this part where they compare the PSLSD and as you go forward you see that the the estimates are comparable but the reason is that they are using the preferred weed estimates of net income per capita when you compare in that period okay so yeah in conclusion as highlighted before the weed database is very helpful what do we know about South Africa is that inequality is has remained high even yes with different with variations in in year-to-year comparison but some of these differences are due to methodological differences that were implemented across different surveys over time like instruments change to improve them and when you compare the patterns are not comparable over time so in conclusion as well they said it is not yet clear what causes this year to year variations as more research is needed to investigate these differences and yeah we they cannot draw like a conclusion about how the narrative from weed compares to each of the country estimates as the adjustment done sometime some of it it raised some of the estimates way higher than what is computed from some of the country estimates when you compare at the different surveys thank you