so good morning everyone it is a pleasure and an honor for me to introduce today's debate with gillette levy and i'd like to say a few words very briefly about her and about what we're going to discuss today i think that we could define gillette as a brilliant young economist with four marx and then she got her doctorate in princeton and now she is written at the london school of economics she already has a very rich bibliography she has written on many prestigious international magazines and if you enter her site and you look at the list of publications well it is really an impressive and extremely interesting list of publications and this is one of the reasons why she has been invited now if you do access her site many of united levy's works focus on political choices based on the analysis of economic models and has developed economic models to analyze political choices and decisions and then she has applied them to analysis of society but she'll discuss it in a detail in her presentation and she has the ability to translate into a simpler language more complex problems and i'm sure that she is going to be able to do it very effectively so i just wish to add three words on this application indeed at the very focus of theoretical analysis and empirical implications there is the following idea society comprises a very large number of heterogeneous individuals society is the result of heterogeneity now diversity can be measured in many different ways and one of the uh most important aspects of these differences is the quantitative difference between various groups differences depending on age level of wealth ethnic cultural religious origins and so on and so forth and this kind of diversity can give rise to different preferences different decisions and this has an impact on the way in which society tries to as it were reconcile different uh these differences the problem of different pressures from various groups uh the issues of taxation how society decides to tax its citizens and how the resources collected by taxation are then redistributed now results of these decisions vary largely depending on the society's composition of the society's makeup and so without further ado i'd like to give gillette the floor for his presentation and i'd like to invite the audience to switch off mobile phones before we start thank you very much gillette thank you very much um thank you for this current presentation paula and thank you for inviting me here it's a beautiful place toronto and i've enjoyed the festival so far the title of this talk i prefer to stand up so the title of this talk is uh diversity and redistribution and i hope you will be able to see the the slides okay okay okay so let me start and hopefully we'll see the the slides soon on the screen so the title of this talk is diversity and redistribution it is based on work which i've done uh both by myself and together with raquel fernandez at new york university who attended this festival last year and orianna bandura from lse who gives a talk this time at the department of economics so the okay it's almost there so the big question is when we have diversity in society is this going to lead to more or less public goods provision of public goods such as health or education we know that people in society differ from one another okay people differ in income some are rich some are poor people differ in age so that may mean that they want different things right the young want education perhaps whereas the old they had enough okay people differ in their ethnicity or religion which may lead to a preference for some faith schools or language programs or special education programs people different geographical locations so not all of them if you live in toronto you don't want another hospital to be built in rome it's not going to serve you okay so people differ in what they need and what they want and what is the effect of diversity in society on the government activities for example on income redistribution is the size of income distribution going to be more or less when we have such diversity what is the effect of diversity on the size of government the total tax revenues and what is the effect on provision of beside income redistribution other public goods okay such as health um education and and so on okay infrastructure is another example so typically we think the way we think about diversity we divide it into two main groups we when we think about diversity in income okay we tend to think that this will result in pressures for income redistribution or redistribution via general goods that everyone likes okay so it is typically the case that when we have and let me exaggerate we have poor and rich the poor will pressure for income redistribution from the rich towards themselves okay this must sound familiar when we think about other forms of diversity such as ethnic or religious diversity this also typically would lead to pressures for redistribution in particular in the way via targeted or specific goods that this group wants okay so uh religious organizations will pressure for religious education or for some ethnic groups may want their own language to be taught at state schools okay so they will want their own particular goods again geographical diversity in some region people want to have transfers to their own region only they care about their region mainly okay so besides diversity in income other types of diversity would also lead to pressures for this type of provision of goods okay but it's it is not a very good idea to think about these two types of diversity in isolation societies are diverse diverse both in income and in other dimensions such as ethnicity geography religion etc okay so we have we cannot just say yes there will be pressures from these people and pressures from that people we have to think about the complex um dimensions of diversity in society together in other words how do income and preference heterogeneity interact okay or perhaps a question which is uh sort of a burning question nowadays how does multi-ethnicity or diversity in ethnicity or religion affect the conflict between the rich and the poor okay is it going to be the case that the poor will get what they want in a homogeneous society in which everyone wants the same or are they going to get more if there are divisions among them okay so this is the type of question that are going to ask and let's be more specific so if it is the case that the poor individuals do not agree on how to divide resources how does this affect their ability to press for income redistribution okay or in you can see it from the other way around how does this affect the ability maybe of the rich to take advantage of the poor or you can ask how does income inequality the division between the rich and the poor affect the ability of the specific interest groups to receive transfers and attract resources for themselves are they going to be more successful when we have rich and poor and income inequality or are these narrow special groups going to be more successful when everyone has the same income okay so this is the type of questions that i'm posing and you can think of several answers okay so one possible answer is that conflicting preferences of a resource allocation okay these divisions among the poor could work against their general class interest okay and that would mean that they would be less able to press for income redistribution but it could also be the other way around and other theories would say that if society is so divided into narrow special groups then maybe the rich they get fed up with all these requests for this type of good on that type of good they would simply if perhaps it will be cheaper for them to align themselves with the general interest of the poor as this may uh imply a lower tax burden so you can think of many answers but the advantage of economic theory is that we can construct a simple model of reality okay to account for of course we cannot account for all the complex dimensions of reality but a simplified version of reality which is what we call a model will allow us to think about these questions in a in in a simple but perhaps refreshing way okay so i'm going to present now uh a very simple what we call a theoretical model and then i will try and test the predictions of the model with the actual empirical uh information that we have the data that we have okay so what i want to combine in the model is both economics and politics okay so if you just think from an economy's point of view and you don't take into consideration the political decision-making i think we're going to lack something so i want to account for both economics and and politics but i want to keep it simple so the economics the economics is going to be very simple and in order to discuss the questions that i've posed i'm going to think of a society in which we just have three groups of people okay this is obviously a simplification i'm going to think about the general poor in society and i'm going to think another group will be the poor who have specific interests so they could be thought about as a different ethnicity or a different religion um they have their own specific interest but they are also relatively poor and then i'm going to think about the rich group okay and as typical to many countries the rich are going to be in a minority so the poor overall constitute a majority however the poor are divided okay there is the general poor and there are those with the special interests now as a society what do we need to decide about we need to decide about the rate of tax so we're taxing income and that is going to be in a sense the size of government and we have to decide how to spend it okay it's very simple and what we can we do is the tax revenues we can either use it for income redistribution okay so everyone will get some share of the tax revenues which essentially means that the rich are subsidizing uh income redistribution because they will pay more in tax okay but we can also finance uh the specific good okay the good that only this interest group enjoys okay so this particular special school or this language program in a school that only this particular ethnicity is is enjoying okay so these are the two types of decisions that we have to make what is the size of government and what is the government doing with the money now what do the people want well i'm going to um again sort of make a very simplified description of reality the poor want as maximum taxation as possible and then they want to get it to get everything via income redistribution they they enjoy income or a general good let's say okay the poor with the special interest they also want maximum taxation because this is something which the rich subsidize they want some income redistribution perhaps but they also want to get some of their own they want the government to subsidize some of their own good they want the government to provide them with their special good okay and the rich being the ones who subsidize everything prefer no redistribution whatsoever okay so these are the sort of preferences that this is what you're born with if you belong to one of these groups that's what you want in this simple society with these three groups if you think about it essentially the general poor are the ones that could win a majority of votes so they can gain at least one other group to vote with them okay why is that the case well let's say you know the poor they offer to society what do they offer they offer maximum taxation and income redistribution well if someone else tries to challenge them okay he will not be successful so if for example the special interest poor challenge them well the rich would still vote with the poor the general poor okay because the rich prefer the general pool to the special interest poor both want to take as much tax from them as possible but at least they get some via in they get something back via income redistribution okay so they are closer to the general pool then to the special interest poor which are different from them in more than one dimension right they like the special good and they also want a lot of redistribution the general poor just differ from them in one dimension so the poor don't want any money wasted on goods that they don't the rich sorry don't want any money wasted on goods that they don't like so given that any poor would want maximum distribution they prefer the general poor to the special pool also if the rich would challenge the general poor in an election they would not get more votes than from the rich voters because the special infraspool prefer the general pool to the rich at least they get some income redistribution they don't get their own good neither from the poor nor from the rich but from the general poor at least at least they get some income redistribution so the general poor are some sort of they will be the winner in such a society if we don't think if we don't think in too much details about the political process we would think ah the poor win right they get they can expropriate they're rich they get maximum taxation and this society gets a lot of income redistribution the special groups get nothing okay but now i want to think more cl more in details about the political process okay and what do we do in the political process we form coalitions okay we form coalitions at many levels okay at the local level at the government level we have parties we have alliances all the time we form coalitions and these coalitions are formed from different groups different factions go together unite together in one coalition okay so this parties you can think of parties or coalitions more generally coalitions of voters or alliances of parties or parties so they're composed typically of different factions and what do they do they enforce some compromise between these different groups right so um for example in the in the uk the party meets once a year to discuss the party manifesto this is a compromise between the different groups for example the labour party is composed of the trade unions and some other groups and they have to vote together on the party manifesto each group gets some voting power and somehow they reach a compromise after a long bargaining process they there is some compromise between the groups okay so what you can do in a party you can if you unite together okay you can reach some compromise and you can offer this compromise you can offer this agreement to the voters okay essentially this is most of the political processes that you think about are built of you know compromise at the very initial level uh in the party let's say and then several parties go together and form another coalition and they have to compromise and so on until we reach the final compromise at the end which is what the voters want okay so let me add to the model the political angle okay and i want to think about these people these groups of voters rich poor poor special interest i want to think about them as forming parties and forming coalitions and and i want my model to to tell me i want the model to tell me what is sort of the stable political outcome of this process okay and when is it going to be stable well politics is stable if each each party can offers the best policy that it can given what the other parties are doing okay so you cannot improve uh the welfare of your voters okay you you're doing the best you can okay and also the party itself is stable in the sense it offers a compromise such that no groups want to split the party okay so no one want to leave the party and run independently let's say okay so when we think about politics in this way we can the model can allow us to predict what parties will be stable that is who goes with who okay who goes with who what is the policy that they offer who wins the election what is the final political outcome okay and again i'm keeping it uh i'm keeping it simple and then what the model what this simple model predicts um in terms of which coalitions form is a very simple coalition of the rich and the special interest poor they join together in one party and what do they do they find the compromise policy okay that both of them prefer compared with the policy advocated by the poor of maximum taxation and no special good at all okay so if they unite okay they can find something some policy which they both prefer it's an agreement between them it's a compromise between what both of them like and both of them prefer this policy to what would happen if they split the party right we know that if they split the party the poor the general poor will win okay so let me explain a bit more about what is this policy well it's not going to be maximum taxation the tax rate is going to be reduced okay and some of the tax revenues will finance the special good okay now why is it good why is it better for both these factions to have this compromise if the tax level in this compromise is low enough then it is better for the rich true some of it goes to a good that they don't enjoy like perhaps these fed schools but at least the tax burden is relatively low okay why is it good for the special poor yes they suffer a bit because the total size of the government is lower but at least they get some of the good that they like okay so this sort of compromise which reduces the size of government but targets some of the taxation to the special interest group is better for both the rich and this subsection of the poor compared with the policy of maximum taxation and no special good okay so this simple explanation could be a reason to a phenomena that we observe in many democracies that is sometimes called why the poor do not expropriate the rich the poor are typically a majority in society there is a democratic vote and still in no country there is maximum taxation there could be other reason why there are no maximum taxation but um this is another explanation to why the poor do not expropriate the rich the rich take advantage of the division among the poor they collude with another with a subsection of the poor that wants something else and manage to reduce the size the size of government so this very very simple model okay let me now perhaps over interpret it perhaps can allow us to think about um you know how left and right so now i'm perhaps giving a bit too much to to the simple model but go with me bear with me for uh for a minute so perhaps we can think about the left that is the general poor that won't be government and perhaps they are anti-clerical they don't want religious education if i take the example of religion and perhaps the right are the fiscally conservative they want small government because the rich are there but also they may want subsidized religious education because this is the coalition of the rich and the special and the special interests poor okay so i think in several countries you can find such examples so in the us the republicans are composed of the fiscally conservative and at least nowadays the christian right in india there is the bjp party of the rich and fundamentalist islam so these are coalitions of fiscally conservative with some special interest and these issues at least according to this model combine not because the fiscally conservative are more religious they combine because of the economic and political reasons that go together here okay i did not think about the riches people who like religion or do not like religion they they care about their money and that's it okay but because of political and economic reason they collude they go together uh in one party with those who like some specific targeted provision of public good which could be for example religious provision or at some sort of provision for particular ethnicity okay so the model tells us that if we combine together diversity in income and ethnicity we need both of them here right if we only have diversity in income just rich and poor the poor win for sure there are a majority so we need both of them here and it implies that we're going to have smaller governments so less total redistribution but it also implies that redistribution is tilted from the general interest from income redistribution or from some general good to the specific uh to the specific special interest that are represented in the coalition okay so what is the conclusion that we can take from the model if we want to sort of test it with the data that we have in reality the conclusion is that outcomes in diverse democracies okay if i think about democracies with multi-ethnicities with a lot of geographical diversity in regional governments that can pressure for their own redistribution in such diverse democracies the outcomes will be closer okay compared with homogeneous democracies maybe sweden okay compared with homogeneous democracies outcomes in diverse democracies should be closer to the preferences of the rich the elite the reach and the preferences of the special interest groups okay so if i want to test what my model is saying this is the conclusion that i'm taking for the model i can i can look at different democracies the when we have more diversity we should have political outcomes that are closer to the preferences of the rich and of the special poor groups okay so i want to i want to look at the data and see what the data says okay so i'm not looking across countries because there's so many differences across countries i'm looking at one big country so i'm looking at policy outcomes of local governments in indonesia okay it is one of the world's most ethnically diverse countries there are few hundreds of different ethnicities most of them are muslim but they practice different types of islam and they have different languages okay so indeed these different groups want their own goods okay they they their pressures for language different languages programs besides java at schools and so on okay but also so these local governments make a lot of decisions about how much goods to provide okay and these local governments also differ from one another okay by customer law some of these these are villages but villages with you know 20 30 000 people okay so these are big villages some of these villages are controlled by the they lit by the rich in the village by the rich people in the village and some of these ins the decisions are actually taken democratically in community meetings okay so i don't know how many who is attending this community meeting but potentially in this community meeting there could be coalitions could be created or created behind the scenes and then the voters do what their leaders tell them okay so i have ethnic diversity and i have some village which are in a sense oligarchic the the rich control them so i can see what the rich like okay and i have some villages which are democratic and some of them are going to be homogeneous democratic some of them are going to be more heterogeneous okay so i can compare these three types of villages and so they decide for example about the number of schools per capita the number of health posts investments in roads sewerage etc okay so what can i do i can compare i can examine whether there is a difference between oligarchic villages and democracies in the provision of these goods okay so first of all this allows me to show that there are difference in preferences between rich and poor the rich do not want redistribution or perhaps want some goods and not others and the main point that we want to check is whether the difference between democracies and oligarchies is decreasing in the level of ethnic diversity in the village okay so if we think the democracies give more weight to the poor from the model we know that this weight should decrease if society is more diverse okay so the difference between democracies and oligarchies in terms of the provision of these goods the general goods should there should be less of the general goods these are general goods that everyone enjoys okay like income redistribution there should be less of these general goods the more the democracies are diverse okay and there should also be more provision towards the special the minority groups okay which which i will get back to okay so this is the test that i'm doing and here are the results first of all i'm going to compare between oligarchy so controlled by the rich and homogeneous democracies democracies with very little diversity so maybe the java group which is typically the majority composes of 80 or 90 percent of the population okay so comparing between homogeneous democracies and oligarchies allows me to see that indeed the rich and the poor want different things okay and the poor indeed get some the poor get some voice in in the democratic political system otherwise no democracy will give them anything okay so what we find here is that indeed in the types of goods that you would think the poor would like the state to provide the general goods like increment distribution so health and education there is more health and education in democratic villages okay and you can look into many parameters so for example school fees are lower uh the teacher pupil ratio is higher so everything works in the direction of better and more provision of these general goods that everyone enjoys health and education okay security the other way around there is less neighborhood security in democratic villages and that's also very reasonable who needs security those that have the big houses okay so the rich like security this is what they like they don't need the state to provide health and education they can buy their own in the private market so when we move to democracies we have less neighborhood security infrastructure no difference everyone likes sewage in the same way or sorry dislike sewage in the same way electricity piped water there is no difference between the rich and the poor okay so this goes in the direction that we think in homogeneous democracies indeed the poor get the voice they manage to implement their preferences okay but in atrogenous democracies the outcomes are much closer to oligarchies okay so first in terms of the general goods in terms of the general goods so health and education there is much lower provision in heterogeneous democratic villages than in homogeneous ones okay there is much lower provision i give you an example here that the number of health clinics per 1000 inhabitants is 12 higher in homogeneous democracies okay so and that's a lot in these types of villages okay so when we move to heterogeneous democracies we're getting closer to what the rich want okay security also there is much higher provision of security in heterogeneous democratic villages than in homogeneous ones okay you can think of many reasons why but it's clear that this is what the rich want we're going we know this because this is what they've done when they were in power when the village is an oligarchic village they had a lot of security so this is what the rich want and in a heterogeneous democratic village this is what they're getting they're getting much more security infrastructure there was no difference between rich and poor in the first place so in homogeneous democracies and in oligarchies they all wanted the same also we don't find the effect of ethnic diversity on the provision of infrastructure okay so that's reassuring so we know that the outcomes go in heterogeneous democracies towards what the rich want but our model also says that the outcomes should go in the direction of what the special interest poor want okay and these were all general goods these are not specific goods to any ethnicity or religion but what we can look at we can look at targeted income transfers okay so there are income transfers that can be that are targeted to your ethnicity so some people because people know what ethnicity you belong some income transfers are targeted to particular ethnicities so we find that indeed in heterogeneous villages the outcomes do not only cater more to the elite but also to the special interest we see that the poor ethnic minorities the non-java groups are much more likely to get transfers in heterogeneous villages okay in homogeneous villages they they are still ethnic minorities there are just very few of them okay they're i'm saying homogeneous but of course there are some there are always some ethnic minorities they hardly get any income transfers once the village becomes heterogeneous they get much more income transfers so the findings consist are consistent with the assumption that the poor are divided okay the elite can rule by forming a coalition with the ethnic minority okay and the coalition chooses policies that give more weight to the preferences of the rich and the ethnic minority at the expense of the poor majority okay so the rich manage to take advantage of the divisions among the poor and get what they want and by doing so they also help the special interest group okay you could think of other explanation to why you know why um why in the u.s now let me perhaps speak more broadly across countries why in the u.s which is very diverse we have we don't have much of a welfare state whereas in europe which is less diverse we do have a very good social welfare system why are african countries which are again very diverse do not provide public goods and countries like states like scandinavian countries which are very homogeneous do provide a lot of public goods so you can think of other explanations for example let me think about alternative explanations perhaps simply let's focus on the first line perhaps perhaps individuals do not like consumption by other groups okay perhaps you know you could be i mean the simplest simple uh you could start from racism or pop we don't like like they are transfers going to other groups we preferred the transfers would go to us okay so perhaps we simply want to limit what other types of people consume okay or it could be that individuals have different preferences and i'm afraid that others will make decisions on my behalf okay so i want to limit the size of government because i don't want others to choose their goods okay i know that if in these diverse societies it's likely that someone who has who's different from me will make a decision i want a small government because i don't want the money to be wasted on someone else someone else okay so these two alternative explanations which do not take into consideration diversity in income okay will give us the conclusion that that all public goods should be lower okay all public goods should be lower in heterogeneous democratic villages okay because i simply don't want i want small government i don't want others to decide for me i don't want others to consume so this this explanations will give us this sort of conclusion that all public goods should be lower okay but we don't see that in the data for example infrastructure is not lower as investment in roads are not lower in heterogeneous democratic villages so it's not because people simply don't like others to consume okay another alternative explanation is that individuals do not like interacting with others right so you can think about you know in mixed neighborhoods sometimes people are not happy sending their children to a school with people from other ethnicities because they think oh they will you know we want them to learn our values and our language and so we're not that happy about consuming education let's say with other ethnicities okay because again we want to we want to maximize the time that our child stays with people from the same ethnicity so he knows our values or language let's say okay so that explanation will tell us that for goods that people interact with one another like in education we should see lower provision in heterogeneous villages okay but this is not what the data tells us because we know we see in the data for example that there is going to be lower provision of health facilities in heterogeneous villages nobody interacts when they stand in the queue for the nurse okay i mean this is not about values or so so we don't find support in the data for such an explanation because um because we see that other goods in which there is no interaction between ethnicities there is lower provision okay so at least in indonesia the reason that diversity affects provision of good could be a combination of economics and political reasons people have different preferences and the political process in which coalitions form give an advantage to the elite and to the special interest okay so let me conclude so i've asked how does ethnic or religious diversity affect the conflict between the rich and the poor we've given a an explanation according to which the rich can collude with these special interest groups to reduce the size of the government and to target some of the transfers to the special interest group and we've provided some evidence from indonesian local governments so the title of this festival is democracy and markets and putting markets aside a big question that i think a lot of the discussion will be in the next few days is do democracies affect economic policies okay is it important to be in a democracy are we going to get different economic policies in a democracy and the world bank and a lot of other institutions call for decentralization of decision-making and democratization and the idea is that democracies would give voice to the poor so the poor would get more of what they want when we are a democracy so we've concluded that democracy matters a lot in ethnically homogeneous societies but it doesn't may not have much of an effect or at least as much as we would expect in atrogenous in heterogeneous uh societies okay so if the world bank tries to democratize so sometimes the condition the money that they give by a democratization process okay so if the world bank goes to indonesia for example to these villages and demands a democratization process okay the la the message is that it is not enough to give voice to the poor there should be other institutions that would compensate for the fact that even in a democracy when there is diversity of ethnicity for example and as the rich can take advantage of the division among the poor even in democracies the poor would not get voice if we do not allow them to change or affect the political system in some other way so we need more institutions in order to allow for democracy to have its its power okay thank you very much thank you very much for your very interesting and clear presentation now i think we can open the debate please raise your hands and you can ask questions both in italian and in english because there is going to be simultaneous translation please introduce yourselves just say who you are and what what your name is so i've seen a couple of hands here good morning my name is ludovica i study economic science at the university and i have two quick questions first of all on your conclusions the last point in your conclusions can it be seen as a call to the development of increasingly local localistic forms of government question mark second question do the predictions of your model change if we add and the multiplication of diverse specific interest groups that is to say if we have different ethnic groups and then you have the local ethnic groups which prevail on immigrant ethnic groups okay thank you for uh for this question um i think so regard regarding your first question could this be a recommendation to increasing uh localistic local governments um i've talked about in a sense from the point of view of the general poor only about the downside of diversity maybe of ethnicity but there is also a lot of upside okay so for example uh we have different knowledge and there is research about so this was all about consumption but there is research about production uh which tells us that uh diverse countries are also more productive okay so this was only the consumption side so i i think it's not a good idea to take this conclusion and say okay let's go of course we are all different so maybe we should go to the individual level and surely there is as we say in economics there is a trade-off because the bigger is the group that we make decision the more taxes we raise and again the the more productive we can be because each of us brings different knowledge different ideas um and so on so this is with regard to your first question the second question it's a very good question so i've made a simple example simple model there is only one ethnicity what happens when we let's say take this ethnic group and we chop it to many many many narrow special groups okay so there is a bit of research on that it starts being more complex because it may depend on the types of goods that we give these groups so for example if it becomes more costly to provide the same total amount of goods okay when when we have more and more special groups because we have to design the special features of these goods for example to fit it to the different ethnicity then the conclusions that i've described sure i mean they will continue to hold even if you increase from 1 to 2 to 10 to 20 maybe but at some point it indeed may be the case that at some point it becomes too costly in the coalition it becomes too costly for the rich in the coalition to hold all these many many groups together okay so it could be that you increase diversity the conclusion hold we have more targeted redistribution less total redistribution but at some point the coalition breaks down and it's just too costly for the rich to hold them in the coalition but we don't have any examples of countries like that yet uh but so we don't see enough diversity to to create a situation in which the poor get what they want but we should probably look better i have a couple of questions here in the front row so my first question is what's the crucial assumption behind the idea that the emerging coalition is actually the one between the rich and and the special interest pause i mean the alternative one is between the two types of paws so i read this model as a as a coalition a forming coalition kind of model and i was wondering about what's behind the conclusion of it the the second question is um well you've talked about two dimensional fiscal policies that are really interesting dimension of of public expenditure and type of expenditure but of course deficits and debt are also uh interesting and there's some evidence that hidden heterogeneous coalitions actually give rise to a higher level of debt so is your model fit to that kind of extension so first the first question uh it's a very important question so why did i say that the only stable coalition is the rich and the special interest poor so think think about the poor coalition that unites all the poor subgroups and the general poor is the special interest poor but i also said that it is important that these coalitions will be stable and no one would want to split the party and and try and run independently i mean we know a lot about party splitting new parties forming i mean it happens all the time so perhaps politics is not stable but we we do try and predict what what could happen in special situation in in this stable situation sorry so if you think about this poor coalition it is not stable in the sense that the general poor okay have an interest to split this coalition because when there are no coalitions at all they can win by themselves okay they are the sort of focal group in society everybody likes them better compared with the third group okay so they have in a sense a lot of power but this power plays against them they cannot credibly commit to stay in the coalition and give something to the special interest board especially it was poor know that they will want you know next election before the election after the election they will split the party and they will afterwards they will they are the ones that will be chosen in the election if there are no other coalitions the coalitions of the rich and the special pool no one wants to split the party if they split the party again they get the general poor policies which is maximum taxation and no special goods okay so this is the crucial assumption that if you split the party you have to think you you don't just split the party you split the party and think what will happen afterwards okay if you split the party in this model there will be the poor party there will be no parties and and and hence the general poor would win so so they cannot take part in a stable coalition i have to say i've not thought about the policies of deficits deficits and debts i think it could relate to even further dimensions of diversity that we have not discussed one of them could be for example age diversity so who is going to pay this debt these are different generations we also have different preferences another uh another reason that this could arise is that we know that the the person that will be the people that will be in power next time perhaps could be from another ethnicity and hence um and hence we don't want to leave them any reserve to the contrary we want to tie their hands they're not going to be able to spend any resources on the goods that they like so this has been this has been explanations that were suggested but i'm not accounting for that another question second row yes i have two questions for you i have two questions for you first of all have you actually applied this model or do you think it is applicable at state level not at the level of local communities and do you think this model could be applicable uh where the poor that whether a new general poor well the poor are structurally heterogeneous so unlike indonesia oh then think of another possible situation uh where no one wants infrastructures the poor want security and the rich want high levels of education and human capital let's call this country italy just for an example then what would happen if we apply this model to italy okay so thank you um at the state level there i i did not apply this model uh specifically but there is research on the state level to see so the research so far at the state level doesn't look at the special targeted goods so i cannot apply the model directly they only look at the size of government okay so indeed when you look across states across countries the size of government is lower in more heterogeneous and more diverse countries okay but again they didn't look at at um the model would say that the redistribution is also tilted from the the general good to the special good and and this we don't have i think i mean it's very hard to do analysis across countries because there's so many different things so this is why i was happy about this data from indonesia because we're still in the same country and within the same region we have many many villages so all the other things that i've not talked about like you know the type of income inequality that we have are democracies richer than uh perhaps democracies are richer than all legal cases perhaps that account for the differences all these things are sort of equal in in these indonesian regions and of course they're not equal across states so it's very hard to say things across states but the general conclusion that diversity means less public goods this is true across states uh and people have looked at that if there is no general poor there is always a group which is sort of similar to the general poor in the sense that you know they like their special good their list they the least okay so you can think about them as the poor group which sort of caters uh to the rich in the best way or caters to other groups in the best way so it's not general poor but it's sort of the most modest special interest group you can think about it in this way so it doesn't have to be completely general but the poor that like the group that likes uh they like when they think about income redistribution versus their own good so they like most compared to other poor groups income redistribution versus their own good so they like their own good the least in the relative terms so it's true that i should have thought about applying this model to italy before coming here but i think the the general conclusion should should stay the same that you should see you know a coalition between rich and special groups so perhaps um you know perhaps if the rich what they like here is high level of education and social capital and if perhaps a group i don't know what are the specific groups that are you know currently on the agenda but let's go a big a bit back in history perhaps you know catholic education is something that uh a special interest group would have wanted so perhaps this could have been uh perhaps you know christian democrats i mean perhaps this sort of type of coalition of you know human capital and uh clerical interest or catholic education could have been a coalition that um that could have survived but it's true i should have studied my italian parties but is better any other questions first the person here and then the third row my name is federico i'm a journalist and how about countries where diversity is rapidly increasing via immigration do you think the new special interest groups could be in a coalition with all the interests of rich minorities i think so this model one of the core assumptions is that you are politically represented okay so to be in this coalition you must have an established machinery of politicians to represent you or even to do the deals behind the scene to create this voting coalition so i think you know typically new immigrants that come to the country are not politically represented and it takes time for them to establish themselves and be in a coalition and perhaps also people can so that's you know in the short term it's probably not the case and also perhaps you know perhaps people can foresee what would happen and again put more institutions in place to either limit political representation or create a situation in which the general poor get get what they want okay thank you very much my name is francesco i have just got my degree and i have just become unemployed and um i'm curious about one point the model you have illustrated seems to indicate that for the rich it is extremely interesting to have the poor divided not united poor now have you ever seen a case in which the rich have actually created themselves a special interest group in order to be able to become a majority by aligning themselves with this creative special interest group so yeah that's a very interesting question so so it's true that it's uh the model sort of indicates that divide and conquer policies is sort of a good it's a good policy for the rich there is an interesting story if you go back if you go back in history to to the united states at around 1860 slavery was not an issue okay so people had slaves and nobody talked about it there were the democratic party and the republican party which was called the wig then also the democratic party represented the agricultural interest okay both in the north and in the south and the republican party represented the commercial the business uh the business interest but the republican party was in the minority okay so what did lincoln do that is a famous piece of speech by lincoln and abram lincoln and what he did was suddenly to raise the slavery issue to to be very salient okay so he was sort of of course people talked about it but it was not it was not a big issue okay before lincoln started it was a republican and he started he raised this issue so what did this do it created a situation in which the north democrats which had to be against slavery okay being liberals could not cooperate in one party anymore in the southern democrats with the southern democrats which were pro-slavery so he introduced this diversity okay which meant that actually people in the north voted for the republican uh party um which uh the rich the north rich which used to be democrats had to vote for the republican party and lincoln won okay the republican party won the election so they've introduced strategically i mean strategically made this second dimension an important dimension on which you vote in the election and manage to get you know i would not call the north democrats special interest group but they managed to get the anti-slavery northerns who were agricultural they had land and were supposed to vote for the democrats they they they moved with the to vote for the republicans so they created this coalition in a sense um you can think about um you can think about belgium and the language issues division of parties there the parties there according to the language issues they've managed to sustain themselves up to some point and then in the 1960s then sort of things started started breaking uh breaking down so you can look at these cases case studies in a sense and think you know was it created beforehand afterwards because there were some groups that were pushing to create of course the diversity always exist but is the diversity politically represented that could be by strategic manipulation of some of the groups in society yeah thank you that was an interesting question further questions i have a question myself but if there are questions this is the time to ask you have the floor then i will ask my question derives from the curiosity raised by a number of things i read from gillette levy in 2004 the author considered the model of political choice and one of the key elements in that was the difference in numerical terms within a society between different age groups young people and the elderly and a specific analysis was made for the expenditure in terms of education so the model in that specific case considered implications that exist or that you find when you have a big numerical difference in terms of different age groups young people and older people so i got curious in a way because i thought immediately of the italian case because we have a lacking democratization process in terms of representation and we have an aging society with an increasing number of old people and so thinking of public expenditure in that respect and reading about your theoretical model and the implications i thought immediately of the italian case because italy i think really represents very well the choices made in this country which favor bulkly the elderly we have basically no family policy which would be important for young people so in this contribution today we're focused on diversity and ethnicity but it was important for me also to spend a few words about the 2004 work which focuses on the demographic differences in our society and their impact on political choices because again we think immediately of italy where we have this imbalance in terms of age groups which gets even worse due to the implications of that would you like to expand a bit expand a bit on that so thank you yeah um so indeed i think you know you can think about this model and think about this uh special interest group as as the older part of the population which of course they care about their children and the grandchildren but they don't care about the child the grandchildren as the parents care about their grandchildren or at least you know for them it's more important to invest in pension plans than to invest in in education primary education or secondary education it's clear um even though they may be you know altruistic to some level um they have other needs and and then again you can think about the situation in which perhaps the rich people would uh if it is cheaper for the rich people to um to go with uh the to colluding coalition with uh the older part of the population and if it means lower tax burden from their point of view okay then we're going to see if this is the coalition again it doesn't have to be explicit in a party okay it can be that this is how voters vote together so the older part of the population together with the richer parts of the population uh if they vote together then we're going to see less resources put in in education okay or it could be the other way around if it is uh if if if there is only a minority consuming education perhaps if they're not going to be as many children it could be that in the future it will be cheaper for the rich actually to provide education because they get these votes and it doesn't cost much because not many people consume education so perhaps in the future it could be the case that the rich would collude more with the younger parts of the population as they realized that this is the minority they can collude with that's it