 Thank you for the first speaker. You mentioned about the Muslim and Christian about happiness. So I don't know how about your country in Ghana, the population of the percentage of inhabitants of the Muslims and Christians, and the Muslims, whether they belong to poverty or richer, I don't know. And then how would you come to the conclusion that the Muslim is happier than Christian? Because of religion or because of other, would you elaborate? The next question? Okay. Sorry. Perhaps, okay. Robustness tests for your variables, for your results. Yeah, for the second presentation also, if I may. Were there any control sort of about the villages talking to each other about the experience of watching the documentaries? And I'm interested in whether the tests have been done individually as in the, the villages go, see the movies all together in one group, those who had the tickets. Or did they do that individually as in times where they chose? And thank you, everyone. Thank you. The next, any other question? Okay. Please wait for my question. Maybe I have, yes, minor comments on the first two. Maybe one of, like, as a follow up, maybe the majority of your population are Christians, I think. If the majority of your people are Hanapis in the econometrics itself, I don't know. But one very important thing, especially for rural society, there are many variables which might affect actually the degree of your happiness, even time. You know, human mood is quite variable sometimes. And anyway, for example, age, some people will tell you a few aspirations, they tell you that I'm waiting to die. I'm waiting for days, you know. That, maybe, include gender plus age, for example. Now, some people might have children and they do see some potentiality in their children. Because if you have, because, yeah, children are, especially rural societies, children are your investments. Even if you are of not much, you're not good, probably you might see something in your children and be happy about it. Maybe type of domestic conflict. Whether they, even if you are married, maybe you are having a conflict with your spouse. So I think there are these kinds of things which probably not mainly captured there, but which a bit worrisome, you know. The subject is quite complex because the degree of happiness might, with individuals, even might change from time to time. That is a first paper. The second paper, I think, I was thinking, what if I go and do the same thing? And try out on a other place, and then I got worried. So that, I have to read that paper very well, but it is quite impressive, and what if I do that? Maybe one of the things I start to think for myself is that people might, psychological, re-respond to my treatment. Because I treat them today. Next time I won't ask for something. If they do not give me some sort of good aspiration, they may feel embarrassed. At least they have to give me some requirement for credit or something like everything. Or some people might think, okay, especially in a dependent rural communities where now foreign everything is there, probably if I, some people might think, okay, they might respond. Psychological is that, probably if I need for credit and everything, something might be coming. I don't know how to control this, probably after the paper in detail, but I may be quite naive for this question, but that is the way I start to think. Maybe if my treatment itself might even make psychologically, you know, respond to my treatment, that is the worry I was having to myself when I was, yeah, that's the thing. Thank you. Okay, thank you very much. Any other question? I think my question is on the, was it the second one about the Ethiopian case? Okay, I wanted to find out, were the documentaries shown at the same time to the 18 villages at the same time or they were done one at a time? Basically in the context of the village, there's the possibility of reporting what happened, what we did this evening from the people that came from the city, that type of thing, in the context of comparing with those that were not given or were not shown the documentaries. Again, the other aspect that I wanted to find out from you is the content of the documentaries. For example, were they in the form of the people that were in the documentaries, advising the communities to have higher hopes or aspirations or were they just showing their life, their history of what they've gone through or what was the nature of the documentary itself? Thank you. Okay, thank you very much. Any other question or maybe we? Okay, okay. My question to Mariano, you present a result with OLS regression which so you assume that satisfaction is cardinal, but don't you think that it should be rather ordinal and you could have used the ordinal probit or ordinal logic rather than OLS? I respond to this round of questions and then we can come back for the next round. So who would like to go first? Okay, please go ahead. Okay, so thank you for the question on, for example, interaction variables. I actually tested a few, I did a few interactions. For example, trying to see whether the degree of, you know, probability would vary if you interact variable like say operating a farm plot with say a non-farm activity. But I've concentrated a bit more on the variables that were showing signs of some positivity but I think it's proper to also take into consideration some of the other demographic variables and interact them to see how they respond to that. So it's a good point and so good observation, I appreciate it. Ghana actually is according to the recent census that has been, that was done in 2010, the Christian population is still in a majority with close to about 70% of the population being Christian and the selection of households to benefit from their cash transfer is really not based on any of those variables. It's purely based on the, of course, application of the poverty line and those who fall within that and even at the community level beyond because it is selection of the poor, the poorest of the poor. And so people within the community are also used to identify people in their opinion, really poor. And it doesn't just end there, they actually kind of match that with the data to be sure that they are getting the right people. And so it's not really, that's why I said that we're not too sure what the Muslim households might be doing, but it could be a certain way of life and knowing that it's a community related way of taking care of oneself. If I know someone is poor, how we go about things might be slightly different. Again, like I said, the result there was not that robust and so I didn't want to dwell so much on it, but I just thought that there was some direction that we need to pay attention to. Thank you. Okay, thank you. Yes, please go ahead. Thank you very much for all the good questions. I will try to answer them as they came. The first one links with the design of the experiment, in particular how the shows where the screenings happen. So there are three groups, as I said. One is the documentary group or the treatment group. The second is the placebo group and the third is the control group. So the first always screening happens with the documentary group, while those who are going to attend the placebo group are gathering separately. Once the screening is completed, the control group would be interviewed while the placebo group watch. So the two do not have any contact and at the same time, the control group, which is not invited to any of these screenings, were being interviewed at all at the same time. We didn't screen at the same time everywhere that is technically impossible because we have 64 villages and 16 screening sites. So we divide them into 16 screening sites and we're going, showing the movies and the documentaries and the survey. And we are not worried about spillover effects, so to say, because of only this, because the villages are relatively remote to one another. There was a peer effects. We have two measures of peers. The first is we asked every respondent to identify four of their best friends and we recorded who they were. And we also have, and this is used in a certain way I will describe very shortly, and then we varied the intensity of what we call the village treatment. We varied the number of people who are invited to the screening of either the placebo or the documentary in each village. So some villages are placebo-intense villages because 18 more household nurses in principle 36 individuals are invited to the placebo screening. These were not interviewed, but they were invited. In another village, we have 18 households invited to the documentary screening and again that would be a treatment-intense village. We divided the villages equally into these two. So we have 32 treatments-intensive villages and a equal number of placebo-intensive villages. And so we have interactive terms. So using this village treatment, we test whether peers affect because the more people there are subject to these screenings, the greater the likelihood of spillover. So we have that. We did not get significant impact. And instead we also measure peer effect in terms of the number of best friends who are invited to the placebo or invited to the documentary. And we detect some significant effect of peers, treated peers on expenditure on children's schooling. Correlates of gender, correlates of aspirations, expectations like gender, age, education exist and local economy characteristics. These are controlled in our, all in our ANCOVA measures. So yes, these influences exist, but we control, we control for them. I suppose you are thinking in terms of the Horthone effect. So, I mean, there is no 100% way of eliminating that. But the fact that we use placebo, we have randomized both at the village and the local, at the household level that all these documentaries are from successful people what achieved, what they have achieved without support from NGOs or government and they are, some are farming and others are a little bit of a petty trading and so on. So, and in all this, we have no promise to support it or to provide any support. So just that the Horthone effect or similar effects may be limited. Again, related to the contents, they narrate their life. They don't advise, they, they don't advise. They do not provide any specific advice on being successful. So they just describe their life. Thank you very much. Maybe the, okay, Maria. Yeah, just a technical question. The response scale for both the life satisfaction and the life evaluation was in a one to 10 scale. So you could argue that this is cardinal, easy to interpret in a cardinal way, but it's correct. You can assume that this is a categorical response scale and then you can use order probit and we have done that. But it's more difficult to explain the results and to interpret because you need to define a response category and a representative person and talk in terms of marginal effects. But the general results, usually when you work with subjective wellbeing data, even if you treat the variable as cardinal or as ordinal, the main results sustain. So that depends on the audience whether you use cardinal or ordinal. Okay. Thank you very much. Maybe before we go for the next round of questions, I would just like to ask a little bit on how generalizable are these results because you chose six months to go back and evaluate. So if you had chosen perhaps a year, would we have seen any effects? And then you also selected a number of characteristics or variables when you were doing your regression analysis. So if you had changed the choice set, would it have also had an impact? So in other words, did your regression suffer from some omitted variable biases? So just think about that as we have some questions from the audience as well. Any questions? Any other questions? Thank you very much. This is a kind of a thought. We are talking about poverty and happiness, but actually happiness is nothing to do with poverty. I was in Indonesia five years, most of them are very poor, but they are very, very happy every day because they know that they have fertile land. If they plant today banana, they will come banana tomorrow. So this is the expectation. I stayed in Kuwait, the richest country in the world for five years. They are Muslim, inshallah. But at the same time, they are most unhappy people because they cannot go further. So depending, this is what you mentioned about poverty with income, but there are many, many other things like inheritance, intangible cultural heritage, background, tradition, all these things are also very, very important depending on which area, which circumstances. Thank you very much. Any other questions? Thank you for your presentations. I have a question for the third presenter. What do you consider the policy implications of the relative income, relative poverty that you measured? Thank you. Any other questions? Okay, there is one at the back. Thank you for your presentation. When I was doing my research work, my supervisor used to say that there are a few differences between life satisfaction and subjective well-being or happiness. He asked me to figure out that I'm surprised concerning the third presentation of the two middle, second and third, that they are the same subjective well-being and life satisfaction. So I don't know if you can help me to understand the few differences my supervisor was talking about. I don't know if you get my point. Thank you very much. Any other questions? Before we give the opportunity to... Okay, so maybe we give the opportunities to the speakers to respond. Thank you very much. So the chair's question, I think, relates to external validity in more broadly. There is no perfect way of ensuring generalizability. What we tried to have done is to focus on key dimensions that are important in most places, income, standard of living, social status and children's education, and also try to link them with, again, factors that are considered important. Then you have the standard rules of running a randomized control trial, so on and so forth, which, in addition to enhancing internet validity, will have some importance, some credence to generalizability. But as I said, this is quite... The message, perhaps, is about the approach rather than the specific results or the specific estimates that we talked about. There's also a limited variable bias. What we tried, of course, to do is that, again, there's no perfect way of dealing with that. But we started by, from a list of the usual suspects, link with aspirations, tests for that, and those that come out significant are controlled. So that is the defense. Thank you very much. Okay, so the first comment was a comment, basically indicating that there are a lot more factors that potentially could influence the happiness level, so there's need to throw in some more variables if the data allows. Like I indicated before, this is work in progress, and so we're going to throw a bit more variables, and, of course, with all these comments, factor them in to make it a better one. The question on the difference between life satisfaction and subjective well-being and happiness, the data I am using for my piece of work does not have the breakdown, but based on the literature, we're using this interchangeably to mean the same thing to enable us to do what we're doing. So it's purely the respondents or the household's own self-assessment, which is purely subjective as to what, whether or not they think they're happy, and so that is what we're using. So we're actually using it interchangeably in our literature. Thank you. Regarding policy implications, I think Jacques presented one which is to start understanding poverty not as an absolute line, but as a line that depends also on the context. And, of course, this will have policy implications for what some people call identifying the poor. I prefer to use the word classifying people as poor because that's what we do. We do not identify poor people. There are people, not poor or rich, and the methods we use classify some people as poor or as non-poor. And then we change the methodology and that implies changing those who we consider. Actually, when we say we consider them poor, what we are really saying is we consider them in need of some social attention. That's basically social problems. But there is something very important here, and that's that happiness, life satisfaction. There are differences, but we have studied the difference between happiness and life satisfaction. Happiness has a larger, effective content than life satisfaction. Basically, that's the difference. Life satisfaction or happiness is also a source of motivation. People are motivated and people act when they marry, when they change jobs, when they move to another city or country, migrate. They are pursuing happiness. So motivations are important because sometimes we think that people, the beneficiaries of social programs are there like bricks and we can do whatever we want with social policy. We don't expect the reaction. We think that what we do is good consider that as good and they will accept that. Understanding their happiness and understanding that this is relative allows us to understand phenomena such as in Latin America pro-market reforms. Economists say they are very good because income is rising. But we find that people are frustrated and they are voting for new parties, even for leftist parties. They reject the reforms that according to the indicators we assume that they are good. So I think this allows us to understand also how people will react what is important to them and to design policy according to what is important to them, not to us. Another point which is important here is to look for policies that focus on absolute impacts. There are things like leisure. Leisure is not relative. The well-being you get from leisure is absolute. It doesn't matter whether your colleagues have more or less leisure free time. So that kind of policy rather than focusing on rising income, you can expand the options of instruments and make of other instruments that are more absolute than income part of your strategy in policymaking. Thank you very much. May I say that if you define a utility function where my happiness or well-being depends on my absolute income and the relative income, then the lining assumption is that I have very good knowledge about what the average income of the reference group is or some distribution of that income. Now, if you empiricize this type of model in a situation where someone may not even be familiar with what the average income of people in the same age group, the same baby belonging to the same gender group is, then aren't we making a very strong assumption here? Well, because of lack of data we took as reference income the median of the median income, but it doesn't exclude it in future research. We may take as reference income something which is maybe which replies your question like people who live in your neighborhood so you know how to relate to them or people who work with you. The only problem I see with this is that you're going to have a poverty line which will depend on each individual. The intention I think is good to get close, but I think from a policy point of view this would not be applicable I think. Okay, please. Go ahead. So, given your true axioms, the results follow. Why did you choose those two axioms? How do you motivate them? Maybe I missed it. How do you motivate them? The intuition that you know what happened when your income and the reference income increased by the same amount then what we assumed is that utility will also increase but not the utility, it's not in dollar terms so we have to assume that it increases because the impact of the reference income doesn't change because my income and the reference income actually and the reference income increased by the same amount so for example in the first approach a difference doesn't change but my income increased so utility has to increase, okay? Now, so this is why proportionate? Now I didn't do the axiomatic derivation so I'm not sure I can answer 100% your question I think these are axiom which you find in the whole axiomatic literature in inequality or poverty measurement you have this kind of axiom so the question I think the question is more important whether you like you know you can have for example a measure whatever poverty inequality you propose a measure you look at the properties okay? People or many people prefer you know to do the reverse that you define a certain amount of axiom and you look for the measure which answer all these axioms it is certainly cleaner from a mathematical point of view now sometimes the axiom selected you have some background idea and what the measure should be and sometimes the choice I am frank with you although I'm not an axiomatic specialized axiomatization but I work with people who don't axiomatization sometimes my discussion with them was that some of the axioms they use I've been selected I am frank with you I've been selected because they give you the results you want to have so it is even if you take if you take for example the past breaking paper by Serge Coleman inequality measurement in 1966 which was published in 1969 where he derived his measure of inequality it is clear that some of the axiom they were selected because it gives a convenient result so I rather be frank with you okay thank you very much the presenters for giving us this excellent presentations if you all agree with me please join me to give them a round of applause for an excellent drawing