 Thank you, Chair, for giving me the opportunity to present this paper. Whose topic is migration, remittances, and remittances in Senegal? Effect on labor supply and human capital of household members left behind. Here is the outline of the discussion. We'll talk about the motivation of the paper. We'll talk about the literature. The focus of this paper, about the methodology, the results, and finally the policy implications. About the motivation of the paper, this says that international migration in Senegal has attracted increased attention from the government and development partners. Indeed, several migration-related institutions have been created. The main concern of the government is how to channel remittances' flows toward more productive circuits. Indeed, some studies, like that of the African Development Bank in 2008, indicate that in Senegal only 11% of families use remittances to fund productive remittances. If you look at the level of remittances in Senegal, you can see that Senegal is and you compare it with other African countries. Senegal is in the top 10 remittances in the South African countries. In the CFA zone, Senegal is the first recipient country in terms of volume of remittances. In 2003, the level of remittances was 1,652 million, which represents 11.2% of GDP. International migration in Senegal is basically motivated by the need to look for better conditions, the need to address the unemployment issue. We know as well that remittances have been found as useful and effective in reducing poverty. You have some reference here. Then it means that migration and remittances could potentially have a role, could potentially impact labor market participation and human capital development. We have two sides. We have a negative side and a positive side. Tertially, on a negative side, migration and remittances, which are a number of things of revenue could generate a state of dependence. We know the parasitism effect in the literature, which reduces the labor market participation of household members left behind. What about the positive effect? Migration could contribute to improve human capital development by helping household members left behind to increase expenditures on education and on health. What is the main message of this paper? This paper tried to investigate how migration and remittances affect labor market participation and how remittances affect human capital development in Senegal. I recently received a comment asking me why you integrate both migration and remittances and why you want to look at the effect of both variables on labor market participation. The reason is that depending on whether the migrants living abroad have or not a job, left-to-behand members might receive no remittances or return or high levels of remittances. So there is a high uncertainty in the connection between migration and remittances. They might not lead to the same result. This is why it is important to investigate both. In the literature, as you know, the effect of migration and remittances on labor market participation was found to be inconclusive. Some authors found positive effects while others found negative effects. What about the effect of remittances on human capital in the literature? Most of empirical literature on this issue have found a positive effect of migration on human capital. Only few papers have found a negative effect of migration on remittances on human capital. What is the focus of this paper? The effect of international migration on labor market participation for the case of Senegal has not been investigated. Look at this paper found on CC. They do not focus on labor market participation. Only the paper by Schumann, 2013, has explored the impact of remittances on labor market participation. But Schumann used a binary specification of labor market. But where are our papers trying to use a set of econometric models? About the effect of remittances on labor market participation, past studies did not make a differentiation of the level of remittances, making a decomposition of the level of remittances in order to understand how this effect labor market participation is important. As it might help know that the effect of remittances on labor market participation might depend on the level of remittances and not only the status of resaving or not remittances. And this issue has received little attention in the empirical literature. So we are going to use the World Bank's household and migration remittances household survey in 2009 to investigate this issue. Well, about the effect of remittances on human capital for the case of Senegal, we have not found a systematic econometric investigation of this issue for Senegal. Previous issues, as I said before, have focused on total consumption expenditures, like the paper by Jain and Jain in 2008. We are not going to focus on consumption expenditures, we are going to focus on expenditures on education and on health. And as well, we are going to make a differentiation of the level of remittances to see how both the level and the status of resaving remittances affect human capital. About the methodology, the first model is the effect of migration on labor market participation. We use three techniques. The first one is the probate model. Here you can see that E is what? E is the observed variable which indicates whether individual i is employed or not in the labor market. M is the migration-related variable that takes the value of one if individual i lives in a household with migrants and zero otherwise. X is the, we control for the individual and household characteristics as the household size, sex, age, marital status, etc. Z i is the potential covariate. According to Roth and Tiberty 2016, the literature on migration considers basically migration networks as one of the influential and observable variable. For example, Taylor at 2003, then these also consider the migration, define the migration, use the percent change of migrants to the total population as a proxy for migration networks, which is the covariate. The network variable, of course, is computed using the data set, the Senegalese migration remittances household survey by the World Bank. The second technique used is that in the same, in the model one is what is called the analogous switching probate model, because here both the dependent variable and the main independent variable of interest are dummy variable. So the ESP is switchable. The ESP considers two different regimes. In our case, we have a regime with migrants and a regime without migrants. T i is the criterion function which help determine which regime the Asian faces with migrants or without migrants. And here you have the binary outcome. We are going to estimate the correlation coefficient in these mistakes, which will help us assess the impact of migration on labor market participation. The third technique in model one is the propensity score matching approach. Here the outcome is the probability of participating through the labor market, while the treatment is the probability of migrating. The impact is assessed as the difference between what is observed or treated and what is not observed or not treated. About the model two, which is on the effect of remittances on labor market participation, we use as well a simple probate model. The first technique is defined as above. R is the log of per capita remittances here. As well as I said, we use various levels of remittances and we generate different dummies. You have dummy zero, which means that the household receives no remittances. Dummy one, the household receives more than 100,000 remittances. Dummy two, the household receives more than 200,000 remittances. Dummy three, the household receives more than 300,000 remittances. So we have tried to make a desegregation, a decomposition of the level of remittances. The second technique for the second model is the IV probate model, because the simple probate does not correct for under genetic problem, does not address under genetic problem. To address this issue, we use the IV probate model and as we say it, Z is the instrumental variable, which has a migration network defined as above. Second technique is the propensity score matching. The outcome is the probability of participating to the labor market, while the treatment is the probability of resaving remittances. And finally, the third model, in the third model, we use two techniques. The first one is OLS, X point is per capita expenditures on education and per capita expenditures on health, R is per capita remittances. We use as second technique the propensity score matching. So with regard to the descriptive results, what we have, we found that households with migrants participate less likely to participate in the labor market compared to household with migrants. If you look at the mean, you have here 0.58 and here 0.52. So households with migrants, they are less likely to participate in the labor market compared to households without migrants. We found as well that households with migrants have smaller total per capita expenditure compared to households without migrants. This implies that households with migrants are basically poor. We found as well that households with migrants have higher level of per capita expenditure on education and health compared to households with migrants. This is basically what we found from the descriptive statistics. Regarding the econometric results, this table is about the impact of migration on labor market participation. With the probit model here, you can see the negative and significant effect of migration on labor market participation. Here is the marginal impact. What does it mean? It means that being a household with migrants having a migrant in a household will lead to a 9.4 decline in labor market participation. This is the first technique. Regarding the endogenous switching probit model, you can see here that the war test is significant. This confirms the presence of endogeneity. This validates as well the selected instrumental variables. Then you can see as well that, as I said in the methodology, the correlation coefficient. Row zero is the correlation coefficient with the first regime with migrants. What does it mean? You can see that this correlation coefficient is negative but not significant. It means that a member of a household with migrants does not have a significantly different probability of participating to the labor market that a member of the household randomly selected from the sample. Row zero is the correlation coefficient for the second regime. It means the correlation coefficient for households without migrants. You can see that the coefficient is negative and statistically significant. It means that a member of a household without migrants has a higher probability of participating to the labor market than a member of a household randomly selected from the sample. If you compare the two results, you can see that households with migrants will have the lowest probability of participating to the labor market compared to households with migrants. Overall, the overall result is the same that we found with the probit model, like migration reduce labor market participation. The propensity score matching, you can see that with regard to the treatment effect, we don't have a significant effect here. Why we have significant effect with the not treated, the untreated? It means that households with migrants are less motivated to participate to the labor market while households without migrants are strongly motivated to participate to the labor market. This support as well as the negative signs between labor market participation and household. I have only five minutes so I have to run two minutes. Okay, time constraint. Here I found basically with all methods, I found a negative effect of remittances on labor market participation very quickly. This is the effect of remittances on expenditures, on education and on health as well. Here as well we found a positive effect of remittances on expenditures on education and on health with less and with the propensity score matching method. So basically with regard to policy implications, what we can say is that as migration and remittances reduce labor market participation, the question is whether migration should be, should continue to be considered as a developmental strategy. Maybe what the government and another policy issue is that another important implication is that as migration reduces labor market participation, it means that migration and remittances might not be viable solutions for employment, for household with migrants. So the government may provide maybe direct subsidies or social protection. Thank you for your attention.