 So, now I'm going to talk about the cross-country evidence, about the growth, employment and poverty nexus. I'm going to focus on the evidence, on the data, taking from the beginning to the end. So we are going to have 16 data points in each of the claims that I'm going to make now, because we are only considering the analyzed changes between beginning to end for each one of the 16 countries that we are analyzing. So I want to focus on three topics that we talk about in our cross-country chapter. The first one is the grow employment and grow poverty nexus. The second one is the role of macroeconomic variables other than economic growth on improving labor market indicators, and the third one is the employment and poverty nexus. So Mariana talked about that we found that in the majority of the country there was economic growth and there was a general improvement in labor market indicators. So the next question is how these two variables are related to each other. Are countries that experience higher rates of economic growth also experience larger improvements in the labor market indicators? And what we conclude is that there is a relationship with economic growth, but the relationship tends to be weak. So let me now show you some evidence to support that point. So in this graph what we have in the vertical axis is the percentage of labor market indicators that improve, and in the vertical axis is the rate of economic growth. So we can see that the slope of the curve is positive indicating that in general country with larger rate of economic growth also experience a larger share of labor market indicators moving in the welfare improvement direction. In the extremes we can see that the two countries that experience the larger rate of economic growth which was Peru and Panama experienced almost an increase of almost an improvement of almost 100% of the labor market indicators and in the other extreme the two countries with the slowest rate of economic growth which was Mexico and El Salvador experienced a much more mixed experience in terms of labor market indicators. And in the middle the countries with moderate rate of economic growth are very, very heterogeneous. So Honduras and Bolivia for example have similar rates of economic growth but very different experience in terms of the percentage of labor market indicators that improve. As a result the R-square of this relationship is very weak is only 0.1 and this is true for these indicators that try to summarize all the information but this is also true if we look at each one of the labor market indicators one at a time. So this is for example the relationship with main labor earnings. In this case the R-square is even slower but still we have the positive slope and we have a larger generation in the middle. So that's a pattern that is repeated for almost all the labor market indicators. This for example is the case of unemployment. We also see the same pattern and this is a case for labor market indicators for which the relationship was tight. So countries with that experience larger rate of economic growth also tend to experience a larger increase in the share of registered workers and our interpretation of this result is not that the share of registered workers produce economic growth but it's that countries that manage to have a higher rate of economic growth has the political space to implement reforms that can be translated into an increase in the share of registered employment. But definitely it's an interesting result and the share of registered worker the change in the share of registered worker is not related to changes in other labor market indicators and that may be something to study further in the future. And this is also a very interesting relation. So our countries that higher rate of economic growth managed to reduce poverty to have a larger reduction in poverty and we found that in general yes the slope is negative but the relationship is also very weak. The R-square is only 0.05. So the main conclusion from this section is that okay countries with larger rate of economic growth in general has a larger improvements in terms of labor market indicators but the relationship is not very tight. So as Mariana said before economic growth seems to be important but it's not the only thing that seems to be important for the region. So the question is what other factors could be there that are important right. So that's the next thing that we analyze in our cross country study and here I present an evidence of some of the things that we look at. So we look at the initial GDP thinking that maybe countries that initially were richer or poorer may have a different experience because they can have convergence or divergent patterns and we found no relationship at all with the initial level of GDP. We also look at the initial level of the labor market indicators because maybe if you start for example with a higher unemployment rate it's easier to go down than if you start with a very lower unemployment rate and we found that there are some evidence of some convergence patterns of that kind. So countries that started with a larger unemployment rate or with a larger poverty tend to have to be the countries that experience the larger reductions and that's a very positive story. And finally we look at some other macroeconomic variables to see if the changes in these variables may be related to the rate of the improvements in labor market indicators and we found some macroeconomic variables associated with improvements in labor market indicators but we don't find a unique configuration of macroeconomic variables that could explain our findings and let me now concentrate on the evidence on the last two bullets. So what we did in the case of the macroeconomic variables we took each of them, each of these macroeconomic variables and we estimate the correlation between the changes in these variables and the changes in each one of the labor market indicators that we analyzed. And here in this table what I show you is for the number of labor market indicators for which we found that the correlation was tight and also we found, I show you the direction of these correlations. So for example the changes in the stock of public debt increases in the stock of public debt are related to worsening in 12 of the labor market indicators. On the contrary increases in the share of export are correlated to improvements in nine of the labor market indicators. So the main takeaway of this table here, this graph, is that some macroeconomic variables are related to seems to be related to improvements in labor market conditions. Now in this slide what we have is for two of the macroeconomic variables we have the relationship between their changes and improve the percentage of labor market indicators that improve. And I want you to concentrate the attention first on the right panel which is the percentage of export and what we see there is that in general countries that experience a larger increase in export also has a better performance in terms of labor market indicators. But we found Brazil for example that has not changed in the percentage of export and still has an improvement of 100% of the labor market indicators. And the other extreme we have Bolivia that Mariana just talked about and also with a large increase in export has a 100% of increase in labor market indicators. And if you look at the left panel and you look at service you find that for example Argentina, Bolivia and Peru were concentrated into increase the share of industry and has a consequence of decreasing the share of service and they also have a very positive story in terms of labor market indicators but in the other extreme you have Costa Rica, Paraguay and Panama with also positive story but reducing the share of export. And the conclusion by looking at this graph is similar to what we hear in the panel, in the first panel yesterday that one size doesn't fit all in the sense that different countries follow different development strategies and by following these different development strategies they were equally successful or similarly successful in terms of labor market indicators. So there is not a silver bullet, there is not a unique recipe of development for all the countries in Latin America. So now let me concentrate on the last of the point that I'm going to talk about which is the cross-country evidence on the employment and poverty nexus. So we saw that increases in the rate of economic growth are not related to better performance in labor market indicators but what about the relationship between the employment and earning indicators and changes in poverty? If our countries that have larger improvements in labor market indicators have larger reductions on poverty that was the question that we want to answer and what we found is that yes, that there is a very tight relationship between improvements in labor market indicators, improvement in employment and earning conditions and the changes in poverty, the reductions in poverty. This table here shows the correlationship between reductions in poverty and changes in each one of the labor market indicators that we analyzed besides poverty. Each one of these numbers represents the movement in the welfare improvement direction which means that reductions on poverty tends to be correlated with, for example, increases in mean earnings tends to be correlated to decreases in unemployment and then to be correlated with decreases in inequality and these correlations are quite high and tend to be correlated with improvements in the job mix of the economy. And now let me show you how does it look in the data. So for example, this is the correlationship between reductions on poverty and increases in mean labor earnings and this correlation is very tight, as tight as the correlation can be. Even my point of view, the R-square is almost 0.8 and all the point lies very close to the line. And this is not obvious because, for example, as Gary was explaining before, in the case of the United States, we have that mean labor earnings has increased but poverty has not been reduced. So this relationship is not automatic, it's very interesting that we are finding this in Latin America. This is another relationship related to the job mix. So improvements in the job mix, in this case, expressed by the share of salary workers are also correlated to decreases in poverty but the relationship is a little bit weaker. So let me conclude with just by saying that despite we don't found a very tight relationship between the rate of economic growth and decreases in poverty, we found a very tight relationship between improvements in labor market indicators and reductions in poverty.