 So today I'm going to talk about a little bit about the effect of international remittances on informal labor conditions, so working conditions for informal, for individuals with informal jobs. So this is going to be the main topic in this presentation. Now, a little bit to understand the context that I'm going to be talking about. We know that informal type of jobs have been one of the main types of jobs in developing countries. So there is enough evidence showing all these, but mostly most of these, of these particular countries, we see the number of people who report being informal as a share of total employment, what we find is that this proportion has remained in constant or has increased. So we know exactly this, but then when we start analyzing this type of jobs, what we are going to find is that actually these particular types of employment, they report to having worse working conditions. So they report having lower earnings, they report long working hours, lack of social protections. So when we refer usually to social protections, what we are going to be talking here is about individuals who report do not do in pension contribution or having a lack of health insurance. So in order to like create a little bit of the context here in particular in this conference that we are here in Africa, so I'm going to show a little bit about how is the sub-Saharan region relative to Latin America. So I'm going to be working with data from Colombia, so it's going to be like, this is going to help us a little bit to understand the context. So what we find actually is that if we look for the sub-Saharan Africa, what we find is that by 2014, the share of on account and paid workers is around 76 percent, and we find something that has come in to become one of the challenges for this particular type of employment, which is young workers. So we find that eight in ten young workers were in today's category of self-employment. We find that self-employment accounted for 53 percent of non-agriculture unemployment. So usually when we start analyzing these particular categories of informal jobs, we are going to find that most of these people usually report being self-employed. Now if we jump a little bit to the case of Colombia, we will find that despite Colombia being a middle-income high country, there is not too many differences in terms of informal jobs. So what happened with Colombia in particular is after the 90s when the economy opens, there was an increase in this type of jobs. So the share of informal workers in Colombia is about 60 percent, and something that we are going to define here because actually we go to international labor organizations, we are not going to find a particular definition of informal jobs. So we are going to account as informal workers those who are in fears with five or fewer employees, who have some paid jobs, who are domestic workers, who report being self-employed, or who are business owners of fears with five employees or less. And as a robustness check, what I did here, I'm not going to show those results, but actually I get the same results, is the definition of informal labor if there is a lack of social protection or health insurance. Now the worry, again, we come back to the job and the use of new workers and the reason this paper focuses on them, is that six of every 10 new jobs available to you are actually informal, and this has going to be important consequences in future periods. Now one of the main concerns about international policy is that informal workers report working around 47 hours during the week for the case of Columbia, and 60% of them report having near health insurance or no pension contributions. So what we are going to be interested about is to see if international remittances have the capacity to improve working conditions for these particular workers. So when we analyze a little bit about Columbia, basically we have experienced, or the country has experienced three big wave of migrations, in particular when these processes started started going to in the regional, in the region. So a lot of people went to Venezuela during the 80s and 90s, and in particular during the oil boom, but then the preference of migrants went to outside the region. And in particular, people started going to the United States and in Spain at the end of the 90s. So what we find is that actually these three countries, the United States, Spain and Venezuela, come for more than 70% of the migrants' destinations. During the end of the 90s, we find that international remittances come for 1.6 billion, and in 2008 it has an exponential growth of 4.4 billion. So there was this rapid growth of remittances, and in particular these international income flows, they come from the United States and Spain. So we have seen here that international remittances and even internal remittances had the capacity to relax the budget constraint of the household, so we are going to explore if actually they can also change labor effort, the likelihood of having health insurance. So we are going to use one of the largest surveys in Colombia, which is the Great Household Survey, and we are going to focus in the periods of 2008 and 2010. It is important to take into account that this is a repeated cross-section data, so it is not a panel. The key parameters that we are going to use is that I'm going to focus in other labor participation, in particular at the intensive market. So we are going to be analyzing the impact of an hours' work, and the reason to do this is that there has been previous research focusing on labor participation, and we can find that actually employment outcomes doesn't change, but the amount of labor effort actually does. So this is going to be one of the reasons I'm going to be focusing in this particular margin. And we are going to analyze health insurance coverage, and in particular using the question I have the individual report having health insurance or not. So it is true that it may be possible that some individuals report not having health insurance, but they are covered because another member of the family have health insurance, but in this case we know, I mean, we can check both questions and see if they really have or they don't. And finally, well, the way, of course, we know that remittances have this big problem of indigeneity, in particular we have two sources of indigeneity, which is the omitted variables and the reverse causality, and in order to deal with this, we are going to use the historical migration rates, which has been used by the literate to be formed, but we are also going to use the macroeconomic shocks of the host countries where these migrants go. The reason to do this is actually if we only use the historical migration rates, then we are going to ignore all the unobserved regional characteristics, and if we try to use regional fix effects, well, we will kill the instrument, so we need to find a way to be able to control for these unobserved characteristics. Once we interact with the unemployment shocks, we will be able to use the region fix effects, but still we need to control for regional confounders of these regional trends that actually can be driving our results, so we are going to use also regional variables, so it is a common practice that people use historical migration rates and only use regional variables, but actually you are still having unobserved effects, so it is important to have the region fix effects. Now, to understand the instrument, as I say, most of these remittances come from Spain and the United States. We are going to use the grade recession unemployment shocks, basically, which happened during the 2007 and the 2009 period, so we had the grade recession by the United States, we contracted by 5 percent response, and we had the greatest Spanish depression also in 2008 until, well, even now, so something that we need to figure out here, or basically, that we find is that was the Spanish economy after a long period contracted by 3.7 percent response. What we are going to use as I was talking is this unemployment rate, and when we analyzed the United States and in Spain, when the United States reported an increase in unemployment rate by 86 percent, and in the case of Spain, at 125 percent, which is something interesting that you see additional surveys, what you find that most of these people actually report sending less money during these periods of time than what they did before the crisis. In order to be careful with the instrument and with this unemployment rate, what we are going to use is deviations from the average unemployment rate before the crisis. So this is going to give us some more energy in terms of unemployment rate. Now, it will say a little bit about how it has been the behavior of remittances in Colombia. So we have this exponential increase after the 1990s after these extensive migrations. Even after 2008, when the financial crisis happened, there was a slowdown of remittances. So there have been several literate to also explain, like, market conditions actually explain these particular slowdown on remittances, and we are going to exploit this and the regional heterogeneity in migration preference to construct or build our instrument. So in order to also make sure that we are capturing what we want to capture and we don't have, like, other unobserved effects there or a confunded effects. So what I did here was to divide the migration intensity by regions in four quartiles and just try to see what happened with the remittances and recipients in each period taking into account the unemployment rates in the host country. So taking into account the unemployment rate in Venezuela, the United States, expect. So what we find, actually, is that, yeah, these regions who experienced historical higher migration before actually are the ones who report receiving more remittances and are going to be the ones who are most exposed to this variation, these fluctuations generated by the unemployment shocks. So in the paper, you can sit in line, there is more tables, I'm trying to show that this is an effect really coming from these macroeconomic shocks, but here I just post because of the time just this graph. So let me talk a little bit about the empirical study. So what I did basically in order to build the first stage was to use these deviations from the unemployment rates, but I also weigh these macroeconomic shocks by the share of migrants of each of the regions. So we have here the different shares from region R to destination J, and we multiply these by the destination unemployment shock, which is the deviation I was explaining before. So when we run this first stage, we can run like a two-stay list square for the case of the hours worker and a linear probability model for the case of the likelihood of having hail insurance. So we control for some other observables who have been using the literatures, so we include individual characteristics as gender, age, year of education, relationship with the household head, which was one of the main control that people use since 2007, and then we start including other different controls, the percentage of members in the household younger than 60 are old and older than 65, which actually are part of the members of the household who are outside the labor markets and they are like a burden. So we control for the regional controls actually that we include to be careful with these regional trends and confund the factors of the regions what the genicoefficient and the regional domestic product grow. And in order to make this a little bit more conservative, we also use time on fixed effects, in particular because we are pulling all these repeated cross-sections from these particular years. So why exactly the period of 2008 and 2010, so this came from the literate remittances in particular for the Dean Jump paper, actually this strategy is very close to this particular strategy which basically used exchange rates as an instrument, we are using historical migration rates, but the idea is basically the same and what we want to try to avoid here is to not capture the endogenous adjustment that people can make during this period of the crisis. So in order to do this, this strategy will make sense if actually it happens close to the period when the crisis occurs. So according to what we have, the period of unemployment rate before they start to stabilize, it was 2008, 2010. So we are going to have basically 284, 361 observation of informal workers, so this is a large labor survey, so we are going to exploit this. Now let's talk a little bit about the main resource here, something that we are going to do, or at least that I do, sorry, is that I exploit the effect of international remittances for those individuals who report having an informal job. So if they have less than five employees, if they were in pay, if they were self-employed, but I also did self-employment by separate. So if we see self-employment, what we find is that some individuals will be self-employed because they volunteer of this type of job, they will sacrifice the benefits of a formal employment to get more flexibility. But there is other people who is forced to this kind of employment. So in order to try to see, I mean we can not exactly disaggregate this in self-employment, but we know that these particular individuals have certain characteristics that are different to their other types of job. So the good news that we find here is that international remittances have the capacity of giving individuals the ability to buy time away from the informal labor effort. So we find that there is a reduction in hours, we find that 25% increase on the mean of remittances actually reducing six hours and 10 hours per individual with informal jobs and for those who are self-employed. Now if you see the F-study study actually is kind of weak. So in order to make sure that our estimates are not biased by weak instruments, we report Anderson and Rubin tests and we find that actually, as you can see the series is not going to be here. So they are not biased by this problem. And this in turn has been shown or has been proved that actually robots do have a certain condition with instruments. And we also explore, well the other good news is that remittances also have the capacity to increase the likelihood that these individuals have health insurance. So again actually we find a very robust results doing it with informal jobs, self-employment sorry and the other definition of informal jobs. So it's a very robust result, we find that 25% increase of the mean of remittances actually increasing 25% response the likelihood of having health insurance. Again we do the Anderson and Rubin test to check that our results are not biased by weak instruments. Now here we have some interesting results, I will show you first the results and then I will explain a little bit of the discussion of why this is happening. So we find that actually international remittances have the capacity of reduce or at least woman use this money to buy time away from the labor market. But if you see actually the impact of men is positive. And if you see like previous literature on remittances actually what you will find is that it's the other effect. So this is something that is interesting that we need to understand what is the context behind and the patterns of migration in Colombia. Now in terms of health I also find that if we took into account all the informal jobs we find that the impact is more pronounced by men and actually we only took self-employment we see that there is not any effect of remittances in the case of women. And actually it's marginally small in this particular case in the case where we take into account all types of informal jobs. Now why is this happening right? So different to what is common in other previous countries or in countries like in Africa or in Asia like Albania which people go to Greece or in Egypt migration patterns in Colombia have been dominated by women. So when we take this into account and the fact that there is strong barriers to the formal sector actually these men who are left behind if there is a lot of effect of remittances which is having a migrant member if this is a lot of effect of said the income effect then what we will see is that actually men have to go to the labor market and if they find these kind of barriers then what is going to happen is that they are going to go to informal labor. Now another part of this still can be that actually since men are the ones who are left behind now they need to figure it out how to balance employment and their employment and balancing caring of the children who are also left behind. So if they prefer to go to informal labor or to sacrifice the benefits of formal employment in order to get this flexibility to take care about their members left behind then we will find also an increase in the case of then we will find an explanation of why this is happening in this positive side which is actually discussed by this seminar paper on remittances. Now the further men benefit the most from hell insurance actually from international remittances in terms of hell insurance is a good news. We know that actually informal labor have really bad consequence in hell status and actually it is not only in the head of the people who is the informal worker but it also the family members of that particular member and we find that actually healthcare access decrease the probability of depression which is one of the main illness that people with informal jobs suffer. So the use of preventive services actually can help to improve the working conditions of these populations. Now in order to start concluding this as I say one of the most vulnerable populations in terms of informal types of jobs is individuals that are less than 25 years old. So in the case of Colombia we find that six of everything so in Latin America six of everything new jobs available to the you are informal and we find that the problem here is knowing if they can go to work or stay at home and doing drugs or doing other things whatever it is the case here is that the people these these individuals who are starting with this type of jobs actually suffer a lack of other types of benefit that they could that they could use to improve their well-being like for example most of these people have lack of access to formal credit and most of these people actually are more likely to remain informal so this is this is something that actually government and policy in Latin America has been focused on trying to decrease the amount of new jobs available who are informal. Now what we find is that these international remittances have the capacity to reduce labor effort for these new workers but it doesn't have the capacity to improve like the likelihood of having head insurance and this may sense in terms that if you think about it they will be able they will be more likely to sacrifice some of these social benefits to gain some experience right so in the other case if we analyze people who is older where they go to the self-employment for example because they want some flexibility so they actually will they will be more likely to sacrifice this and use the remittances to get to get head insurance. So in order that my time my time is over so let me conclude about this basically what we find is that remittances have the capacity to relax liquidity constraints and using these remittances people or individuals with type of informal jobs actually can increase their well-being how they can access to health insurance and they can also reduce the amount of time in labor effort. So the other important results here is that actually international remittances have the capacity to compliment these efforts that the governments are doing to reduce formal labor employment for the youth. Thank you so much.