 Thank you very much for the invitation. It's a real pleasure to be here presenting this joint work with Jeff Nussiant. The first thing I would like to say is that the title is quite telling and let me just give an excited debate about data quality and inequality yesterday. I will start to open it for disclaimer. The title as I said is quite telling. I mean, the paper tries to look at what explains the reasons for the rigidity of labor regulation, not only across countries but over time. And we look at the roles of structural reforms as potential determinants of employment regulation rigidity and their impact, the impact of these regulations on growth and inequality. Then inequality will be on the correct side of the regressions according to what we heard yesterday in this paper. Then what is this paper about? I mean, you're going to be talking broadly. The emphasis here will be on less on inequality and more on this public policies. And we're going to be looking at the labor market institutions, which is obviously a very important topic, and in particular we're going to be looking at one of them. We're not going to be looking at different interactions between very important labor market institutions. We're going to be focusing on one. And the one we're going to be looking at are the laws and regulations. I mean, we're going to be looking at employment protection. I said something about the ceiling site, looking at this EPL index. And the ceiling site is essentially, well, if you care about reform, if you observe changes in this index, that would be an approximation to reform. And of course, if you observe, no change in this index is that would be a good hint that these rigidities are indeed important, or at least long-lasting. Then that's what the paper is about. I mean, can we come up with measures of the rigidity of these labor market regulations over time in the past? And try to explain them, and then go back and look at the implications of these changes in terms of both economic growth and income inequality. Hence, why this paper is in this conference. The contribution of the paper, if any, tries to be twofold. I mean, it's an empirical and a policy contribution. And the policy contribution, I think it's quite clear. I mean, the kind of empirical or academic contribution departs from kind of a caricature of this literature. And I kind of characterize it as basically kind of three features. Most of this literature is confined to the post-1995 period. When not confined to the post-1995 period, it focuses on two regions, which are OECD in Latin America. And third, I mean, this is the third feature I already mentioned, it focuses on levels as opposed to changes. It just occurred to me yesterday that this, if I describe the literature on income distribution, it's possibly a not very unfair caricature. But in any case, I'm talking about EPL, not about income distribution here. Then, the two main objectives. One is to extend the absolutely seminal paper by Botero et al. in 2004 that come up with this index for one point in time. And second, to assess whether or not our effort, our exercise, makes any sense. What's the worth of this exercise? Let me just spell out, because I don't assume that this audience is familiar with this work, at least not at the level of the detail required to follow what we're doing, because we're going to follow this very closely. This is one of the various papers by Djankov Laporta, Lopez de Ceylonis, and Schleifer. This is one of the quality of label one that came out in the quarterly journal in 2004. I'm sure you saw a number of them. This is kind of the theoretical philosophical basis to the doing business project at the World Bank. And this particular paper looks at label, label regulations. It focuses on 85 countries for year 1997. What we do is essentially we use exactly the same methodology. Extend it first for that particular time period to a considerably larger number of countries. They originally have 85. We go to 145, 150 countries, which is almost doubling the sample. And that would have, of course, implications to what kind of... As you can imagine, the countries that are being added, the very bottom of the income per capita distribution that increases the variability, the variance in terms of per capita income levels in this sample, and then enhance changes slightly the potential explanations of these changes. Then what we do is try to extend it backwards all the way to 1945. But the real action starts after 1906, and that's why we cut it out. Objective two, of course, is to assess this new measure. We call it LUNRIC, which is an index of the rigidity of label market regulations. And we do that in a number of ways. We try to replicate the Botero et al. result. They put forward three competing theories, three competing explanations. For those of you familiar with this line of work, the one that comes on top is legal origins. According to not only this paper, but a number of them, the origins of the country legal system plays a determinant role, a number of institutional features. I'll talk a little bit more about it later, but it's the contrast between civil English and French lot traditions where the action is in this particular story. The important thing to say here is that they find legal origin plays a very, very important role. And just the addition of all these countries at the bottom of the per capita income distribution, you can imagine bringing the role of, quite naturally, increases the power of what they call efficiency theory, which is level of development. That's one change that is not very surprising, but follows from the exercise. Then this is the first assessment we provide. The second one, well, now we have this measure over time. Then we see what happens in terms of levels and changes. Then how much inertia is there? Legal theories, legal origins still play a fundamental role in explaining levels over time or changes over time. We find that they do still play a role, but it's a much more nuanced story that we can put forward. We examine changes instead of levels. And because it's a more nuanced story, we need to search for other determinants. In some cases, per capita income disappears. In some cases, there's very little support for legal origins. Then what we do is we embark on a search for potential other determinants. I mean, in a very humble way, trying to assess whether or not this construction exercise, this data construction exercise makes any sense. There's no intention of providing a final theory or final explanation for these changes, but it's just to see whether or not they make some sense. And of course, at the very end, we do also look at the effects. Now we can endogenize the CPL changes. You couldn't before because you have one point in time whether or not they have any effect on growth and economic inequality. I'm going to give you the conclusions because I still have a zillion slides to go, and I'm not going to finish, obviously. They find that legal origins are very powerful explanation. We find that over time, particularly in terms of changes, if you care about labor market reform, legal origins play a much more less strong role. There is some evidence for legal origins, some evidence for the importance of the level of per capita income, but what we find that's quite interesting is that other structural reforms, we look at a number of other, we look at crisis, we look at structural factors, et cetera, and we find that other structural reforms play kind of an interesting role. In particular, trade liberalization seems to have a negative impact on rigidity of labor market regulation, and we have an opposite effect played by financial liberalization, which if you're trying to design reform packages, it's kind of an interesting insight. This is all done on five-year averages for 140 countries. A policymaker shouldn't take this as any sort of guidance, but it highlights the importance of these interactions between reforms and designing packages. The last one, I think, which would be the big interest to this audience, but the relief is that we do provide evidence for what we call the Freemont conjecture, which essentially says that this labor market, rigidity of labor market regulations have very little, it's very ambiguous effect in terms of growth, but they do indeed reduce income inequality and we do find support for that over time in a panel setting, which is obviously something that, if you don't have this set of measures, cannot be done. Anyway, the outline of the talk is introduction and motivation. We took up 15. Yes, excellent. Then I have the real talk that goes, how did we construct land rig? What does it look like? How good is it, this assessment? You know the results already and there are conclusions, maybe this afternoon I can get to that point, conclusions and next steps. I'll fly over these slides now because I presume you know what I'm going. I'm going to measure land rig in three easy steps into one checks and another one is balances. Then we take the Botero methodology and we combine it with a compendium of labor laws that the ILO has generated in the last five or six years. This is Botero, Botero employment law index has four components. I don't have time to go into them in greater detail, but I mean they can be mapped very straightforward manner in this data set. This is the website of the data set, of the not lacks data set and what it is essentially a compendium of labor laws. Okay, then we have different countries and the ILO very kindly classify each law in different areas. For example Australia, we have 2,506 laws since 1945 and they are classified over time in each one of different subjects. The list of countries as you can see is endless. The list of categories is also endless. An important thing for us is that they map into the four elements, into the four components of this index. We have labor contracts, part-time versus full-time contracts is one of the categories. This is for our junior, it's a random example. Cost of including hours of work, I'm sorry. I can't help you, I'm sorry, apologies. Right, hours of work, it's another category and then we have cost of firing is another one and dismissal. We're going to look only at these four boxes, 450 countries. You could kind of do something more refined but this is two guys and 12 research assistants. Okay, I was expecting zero. Okay, all right, plenty of time now, it's going to slow down now. First step then, we start with this EPL. The first thing we do is we know the result. They have like us a very detailed web appendix with a list of these laws and it's like an economics exam now. Imagine, discuss, pick a D, right, it's question one and somebody gives you, okay, this is the assay, this is an A, this is a B, this is a C and this is a D. This is a very kind of candid assessment of this literature and I hope you're getting the gist of it. I mean, there is no way you're going to come up with indexes that try to reflect anything important about labor market rigidity that do not involve a degree of judgment, okay. There are aspects of it that can be easily quantifiable. They are usually the less policy relevant ones, okay. And this is an issue, I mean, if you are very sanguine about data that is simply unavoidable in this literature, okay. Then if you don't particularly sympathetic to the view, I mean, you better stick to exchange rates or something a bit more objective if you want, okay. Then what we do is essentially, that's what we have. That's the starting point, okay. We know the laws from Botero. We know the grades that they give to these laws from Botero and we add to that the same laws for that five year period from Netflix, okay. And then we just essentially, we come up with a key, an answer key, okay. Once you have the answer key, we know exactly how it maps, we extend, okay, right. I mean, we bring in new exams to mark, which are, you know, the 85 to 145 countries, okay. And then we go back in time, all right, and mark all the exams all the way back to 1960, okay. This is exactly what we do, okay. You know, that's the second step. I mean, first we extend the marking if you want, generating the sentences, okay. And then third step, we just go back. And because these laws and regulations change slowly over time and because, I mean, you know, we want to see, you know, the growth aspect of it, I mean, we're going to look at these five year averages. I mean, we're going to kind of block them in these five year boxes, okay. And then we have, you know, three steps and five steps, actually. One is checks and the other one is balances. I mean, the checks, I mean, we compared, you know, our index with, you know, the ones for DOECD in Latin America. We generate 2000 and 2004 because, relax again, 2004 because of, it's the first vintage World Bank doing business, which is still very much, you know, in this tradition. I mean, I don't know how much you follow that work, but it has changed quite a lot after 2006. And when you compare, I mean, you know, discrepancies are literally minimal, okay. I mean, we're quite happy to see that we can, you know, I mean, I don't know if we're doing something right, but, you know, we're not doing anything different from everybody else. And balances, I mean, it's, you know, Jeff is very keen on Middle East. You know, he's kind of searching the web and really bringing in history to check, to try to check whether or not this indicators make any sense. And I mean, we're still quite happy with them. I'm going to talk about, you know, the impact of these balances in a little bit, okay. How does it look like? Because, I mean, you can imagine that, you know, there's no action at all, right. I mean, these are laws. Then, you know, they change very slowly. You know, the way, you know, how flexible or how rigid the implications of these laws are also kind of, you know, very, very slowly change. I mean, you can imagine that well, after 1991, everybody moves up, but, you know, there's very little variation, okay. This is the first, an important comparison. And it's important for a couple of reasons. First, because in the Botero paper, I mean, it takes, I believe, five or six pages in QJA, okay. Portugal is obviously a French legal-origin system. New Zealand is obviously, or not the typical case of English legal-origin, you know, legal system. And they, you know, get a lot of mileage, finally. Thank you so much. I get a lot of mileage out of this difference, okay. You know, Portugal is much more rigid label market than New Zealand, and you can say, well, you know, does it matter to look at overtime? Well, if you do look overtime, and if you choose, you know, how the point of comparison, you can, you know, draw a very different set of conclusions if you look at the beginning of the sixes, then if you look at 2004. There are a number of individual studies of, we don't know for Portugal, but for New Zealand, we do know individual lawyer efforts to measure, and the behavior of the index is very much like ours, then that's the checks and balances bit, okay. Then, you know, we have a bunch of other graphs, I mean, we have 145 countries, and we generate these very easily, but you can see there's movement, okay. There's action. This is China, India, and Brazil. You're not going to see it very clearly either. And there are a number of very interesting patterns, and legal-origins, when you start to look at these changes over time, you know, you know, it's less clear whether or not it's that type of pattern emerges, okay. These are the Botero and Etol results. We essentially replicate them. I'll just summarize what we found very quickly. I mean, I have a number of slides, as you can see, and I'm not going to be talking about them. Then there's some indication that it's going to go very fast. There was this, you know, that trade liberalization as a negative impact on labor market reforms and the other effect by financial reforms. And then, oops, oops, okay. Then we tried to, you know, replicate the original results. These are the original results. I'm sorry. This is cross-section, pulled OLS, and you can see it's pretty much, you know, follows it very closely. And then, you know, a couple of sample splits where, you know, there are some interesting differences that I don't have any time to talk about. Then we start to look at changes, and you can see suddenly all the significance from the legal origin starts to disappear, okay. And that's, I mean, what prompted us to look for other determinants. Then we go on and on looking for different determinants and don't find much other than those structural reforms. Okay, I mean, look at different types of crisis. It's going to go very fast. You know, look at different political crisis and factors. And the slides are going to be online. That's why I'm flying. And then we have some evidence on the Freeman conjecture in particular. Okay, it's not moving. It's not my fault. In the sense that on the left-hand side, the first three columns, I mean, you can see that long rig has a negative and significant impact on income inequality. Caveatria, this is weed, not sweet. It's weed true. I mean, then when we send the extended abstract, we're planning to use a number of different measures of income inequality. But I mean, this is the one we have right now. And in the columns four to six, you have the results for growth. I mean, you can see it's ambiguous. Okay, then for this audience, I mean, you can of course say, oh, we knew this all alone. We haven't seen evidence that, you know, suggests that there is a lot of us or, you know, some strong support for the Freeman conjecture in a cross-country panel setting. And I think this is, we find it interesting, but that obviously just us. Then just summarize if I can. Okay, all right. I really appreciate it. It's extremely generous. Then I think the paper offers a new index of labor market rigidity, legislation rigidity. I mean, it's totally the yure. I mean, I don't think, you know, I don't want anybody to have a different idea. It's a panel since 1960. You know, you can see quite a lot of interesting action. I mean, revisit the original explanations, potential competing explanations for these changes and they find slightly different results from before for reasons that I think are quite clear. I mean, not surprising. And there are two kind of additions to that list of findings. The first is regarding the importance of different reforms in terms of important implications in terms of design of reform packages and the other this evidence supporting the Fremont conjecture. Thank you very much. Apologies for running both.