 Thanks to all of you for joining the session and thanks for the two previous speakers for doing an excellent job and already zooming in in very detail into two of the most known Contra cases that have affirmative action policies in place And actually something that we're going to do the project together with Rachel Gisselquist is we are working on a database that tries to give an overview of what kind of affirmative action policies have been implemented around the world and Rachel is basically the brain behind this even though I'm presenting But she did an excellent job in bringing all of this together and it has been great working on this It's the first time we are presenting it. It's still work in progress. So any Comments are very welcome So where we basically started from is observing that there's a large literature on affirmative action But most of this is heavily concentrated on few country programs South Africa being one of them Malaysia being one of them the US India but actually there's way more countries around the world that have this kind of policies in place and What we have been doing is Looking at the most known country cases and basically from those drawing conclusions about whether the same type of policies Could be working elsewhere without really knowing what is the scope of policies that are there and how have those been operating? So this basically means that we have a very weak basis of understanding What is the actual impact of affirmative action policies and basically as a first step in this project? What we have been trying to do is to get this overview on the one hand What is known in the literature? What can we get from this and then how can we supplement this with this database? So basically we've been working on two different outputs here And I'm mainly going to talk about the database But also give you some glimpses of the work on the systematic literature review that we've been doing in parallel At the moment in the policy database we have 45 countries coded And they prove what we are trying to provide is information on the implementation of affirmative action policy in general And then along five main policy domains in which affirmative action policies have been implemented Which is education, public sector employment, private sector employment, political representation and then any other And in the literature review basically what we have been doing is we have evidence on a total of 195 Studies which include 182 case studies covering 27 different countries and then 13 comparative works Oops Okay So this is the data structure of the database We start with a roster that basically allows that also to merge our database with other databases like the VDEM the Where UN databases, population databases and so on so we have information There on country name at the different identifiers, the population shares and we also merge in an ethnic fractionalization index Then on the just general affirmative action policies we look a bit at the origins of affirmative action in the country and The main thing we try to capture is different types of controversies and we're going to look at this a bit Then as I said we look at this different policy Domains we look at what are the target groups of the policies against what it's origin Have there been any amendments as a policy still in place and we try to bring in evaluation variables and both this controversy and evaluation We had two research assistants working on those and it's basically them scoping the literature scoping available data on like literature on articles on news reports and trying to capture a broader Countrywide debates around those affirmative action policies. So how are those perceived in the public domain and Then the last set of variables basically captures specific policy documents kind of in providing the link to those and For all countries that are coded we have a fact sheet which explains the coding basically So for researchers who are interested in working with the database You can have the fact sheet on the side to make it transparent how the coding was done So in total we have 413 variables at the moment in the data set Including the identifiers so a total of 386 variables capturing different characteristics of affirmative action And what follows we're going to look at each of these in detail Of course not so we're going to zoom into a few of them I can only give you a broad overview and I guess you want to be done by the Asimoglu lectures this evening so on the country coverage, so this is a map of our Literature review just to show you a bit the concentration of the evidence that we have been seeing so we started with a total of 400 4,000 Publications of which these just below 200 met our inclusion criteria Which where we were looking for studies that look at the impact of affirmative action Which look at policies the target ethnically defined groups, so we didn't look at policies the target for example gender only And what we observed is even though we have there's almost 200 studies Covering 27 countries about 70% of the evidence is made up by just four countries So that are the most frequently studied and those are the US India Brazil and Malaysia So we do have quite a set of evidence, but it's heavily concentrated Which was one of the key observations that we had from the literature review and this Threatens to bias our perceptions of affirmative action Moreover, even though we have 27 countries, which isn't bad This is still way less than the actual number of countries that have some type of affirmative action policy in place And this is basically the gap that we are trying to fill with the database to even give an idea of what is the total universe of cases that we should be looking at so So when we started this work about a year girl We started the coding with the ten countries that we called our priority one countries basically Which are the usual suspects kind of the one that you find the most often discussed in the literature So we started coding those and something that we observed like I hear plotted ethnic fractionalization index is the countries are very different right you have some with a lot of scattered ethnic groups and then you have some that are more concentrated and we had this There's sometimes affirmative action policies target my door minority groups sometimes a target majority groups and It's quite diverse in the setting So it's not necessarily that affirmative action policies only adopted in certain contexts But they are basically across the board And since then we have been extending the scope those were the priority to list here That we added we have Britain as the first country that we actually decided this policy It's an anti-discrimination policy, but we wouldn't classify it as affirmative action and think something that Was quite a challenge for us And I think we also saw this a bit in the first presentation is what policies actually classify as affirmative action Where does equal opportunity and non-discrimination stop and where does affirmative action start basically? What does classify as preferential treatment? And this is kind of a lot of back-and-forth discussion that we had in the coding basically, but we tried to really Limited to policies that give a clear preference to one group that has been experience marginalization in the country's history and then Priority three. This is basically the coverage that we've reached at the moment. So you have the 45 total or 46 no 45 total countries that are Coded that have affirmative action policies plus the number where we actually couldn't find an affirmative action policy and ideally we want to further scope further extensive scope To to cover a wider range of countries and To also be able to have comparison between countries that have and that don't have affirmative action policies So for now, we mainly focus on those to get those coded But to actually say something about what maybe country characteristics make it more likely that you adopt one Compared to that you don't We would also need to make sure for all the countries that are not among those where we expect them to see or where we've seen Or we know already that they are to make sure has there been any type of policy like this or not And it's gonna still take us some time to do this But we hope to launch at least this version one that has already rich set of information within the next months so just some Some early insights on by policy domain So those are the countries where we found affirmative action policies in Education and we try to differentiate between policies that use some type of more hard quota With is those that have some kind of softer policies in place and if you look on the right-hand side So we have out of the 45 Case studies that we have coded around 29 So just above 60 percent had policies in education So those are among the most frequently adopted type of affirmative action policies according to what we see in the database up to date Also, we see about 60 percent of those have actually quota policies. So those are rather frequent in education and Most of those target groups defined by race or color or some type of Edna regional classification So examples would for example be policies for the population of African descent in a number of Latin American Countries that have been adopted for example in Brazil, Colombia, Costa Rica, Ecuador and Uruguay And many of those being adopted quite recently actually. So there's an expansion of the sort of policies And other kind of measures that are less Quota driven are for example scholarships financial aid and those we've been seen more for indigenous group actually So rather smaller groups that receive more kind of targeted Type of policies that that's something we've been seeing for example in Australia, Canada, Chile, Fiji, Indonesia and Taiwan Looking at public sector employment This finds us in a similar number of countries But we see that the quota the share of countries that adopt strict quota policies is lower So here it's more often that is a kind of more softer policy measure Again, mostly I then adopted for groups defined by race or color or region And also here we see that non quota policies actually most often occur for race or color So you have a more of a policy mix in this area basically Looking at private sector employment We see yet a stronger shift towards non quota measures So we only have about 20% of the cases showing strict kind of what we reasonably could identify as quota and A lower number of countries having adopted those policies So most of the policies are in the public sector basically both in education and in employment, but there is Also more this kind of incentive policies in the private sector. So this can be something like for example in the US it's more of a equal opportunity non-discrimination kind of policy But firms are for example requested to submit affirmative action plans or something like this So it's more softer targets that are supposed to increase diversity of them of the labor force And Actually interestingly we saw where quota were adopted those refer indigenous populations, which are normally a minority So you would you would see those and This we saw in Australia for example, where there were jobs that were specifically created for indigenous applicants and We also classified the Malaysian case here And then we have what we haven't yet talked about in the previous Presentations is basically affirmative action policies in political representation. So this is usually you have Reserved seats in parliament for example for certain ethnic groups And this is usually a type of quota policy. So as I just said is normally reserved reserved seats For example adopted in Bolivia, Chile, Fiji, Indonesia, Jordan, New Zealand, Peru, Taiwan and Venezuela So there you basically try to ensure the political representation of minority groups often defined by indigeneity and You make sure that they have representatives in the in the legislative And then well any other It's usually a non quota policies that we see here and what we Coded here for example policies in social and public Housing what we see in Albanian Singapore or also in for Roma in a number of Eastern European countries We also classified here indigenous land rights giving preferential access to land and For example the exemption from the one child policy in China So basically in any other domain trying to to improve the position of minority groups I think we already heard in the first presentation that usually the adoption of affirmative action policies is linked to some type of of Events that acted as a catalyst for those policies and we tried to to capture those by in different domains So what we see for the largest share of cases that we have coded is that the adoption of an affirmative action policy Was actually linked to new constitutions. So often this was adopted as gaining independence of the country We also see in a number of cases about 30% that they were actually linked to some type of major violent Conflict or event that could be a civil war. It could be a riot which we had also seen in the first presentation So quite often there's strong public pressure or this Yes, some type of extreme social tension That's kind of boiling and at that point the affirmative action policies are implemented in order to address So horizontal inequalities that have often been there for a long time often been rooted in the historic Context of the country being rooted in structural discrimination, for example regarding the controversies Basically for almost all the cases that we could if you could find some type of controversy Which is probably not surprising like policies are controversy any type of policy in most cases Again as I said we coded those like our research assistants a screened news articles Is that a trying to figure out? How are those policies usually? Portrait and what are the concerns that are being raised by different groups and who which groups are raising those concerns Just here is a split by policy domain so most of those For countries that have different policies in place not necessarily all are as controversial basically We saw policies in education being the most dominant in the public discourse of being discussed Controversially why for example policies in the private sector which also don't come with quotas and in political representation Tended to be a bit less controversially discussed If you look at what The first one is okay. Did this we asked do those controversies actually lead to some type of protest or other kind of violent events we saw that Just in about 13 to 16 percent of the cases we could associate those controversies with some type of violent event in the country But more often they were in about 60 percent of the cases They were related to some type of protest or civic action that emerged after the policy was in place So this is not any more linked to the adoption of the policy, but rather after it had already been in place and most of the time those are actually linked to Claims by the target groups so it's quite often that groups that are the supposed beneficiaries of the policies are Actually not comfortable with the way the policies are implemented So it might be that they feel it's more That is not implemented properly that it's Kind of more in place to show something is being done But not actually giving them the representation they were hoping for for example if you think about policies in political representation Just having a representative in parliament doesn't necessarily mean that your concerns are getting the attention you were hoping for We also see that in about 30 to 40 percent of the cases This controversies are brought up by people who are not the beneficiaries of the policies So you might have other minority groups who are being left out, but also for example I think both of these we see in the United States for example You can have the majority group or the non-marginalized groups It says it's not always divided like this that are feeling that it's an unfair treatment That's being experienced. So for example in the US case in the education policies You had Asians as a minority group who were kind of losing out due to the affirmative action policies and you have the white population who was also raising claims about unfairness of the policy So the last aspect that we tried to code is basically different evaluations So we looked for is there an official government evaluation of the policy which in many cases we couldn't find but where we could find it We see that even those were not necessarily always positive kind of so we see the largest share of positive Evaluations in that area so by the government itself, but quite often even those came to the conclusion Okay, they have been kind of a mixed effect. So there's improvement in some areas. There's still a lot to be done and The least favorable Evaluation we saw in the public discourse, which is basically coming back to those controversies. We were looking at at the beginning so usually affirmative action policies are discussed extremely controversially and I have a rather negative image in the way it's portrayed in the news and media for example Which might also be partly due to the nature of news generally giving maybe it's hard to see positive News or to have somebody just lobbying in favor of a policy that's already in place like it's maybe natural to see that Just for comparisons, those are the coding results from the literature Reviews so here we tried for each of the studies that we coded to classify the the effect into positive negative Mixed or insignificant and for the effects on the target groups We distinguish between first and second order effects. So by first order effect. You basically mean The immediate outcome of the policy as it was intended So for example, if you have a policy and educational admission the first-order effect would be the effect on marginalized students that are Being admitted to university and naturally those effects are usually positive So this kind of first stage usually works Affirmative action policies tend to increase the representation of the target groups However, what we look at in the second-order effect is more like, okay What does it mean for performance and for kind of more longer-term? Outcomes and here's something that we observed is that the effects are more mixed So for example, this would be actually educational attainment. It could be the labor market performance It could be poverty outcomes and here we find quite Still dominantly positive, but also very mixed results for the target groups Just as a last one So this is the effect on other marginalized groups, which has been quite rarely studied on the literature and the effect on other Non-marginalized groups and those are largely negative. So you have these replacement effects basically So this is not just a thing and that's happened in the public In the public debate, but there's actually also evidence that there is those replacement effects Okay, so just to summarize as I guess I might have time Something that we observed is that much of the affirmative action literature focuses on a small set of country cases Basically affirmative action policies are always controversially Debated one main concern that has been raised is basically tokenism and implementation failures But also claims by non-beneficiary groups And just on the next step So we have two papers coming out of the global literature review One is a global paper and one where we zoom into the evidence that's available on the two most frequently studied country cases Which are the US and India? We're currently trying to Compile the first cut of the database. So we're doing some final quality checks and hope to release a version one sometime in the next month and Yeah, then continue adding additional country cases. Thank you