 Minun tietysti tänään on taustattomisesta kehittämisestä järjestelmistä. Tämä on tärkeää, että Toni, joka on täällä, kysyy minulle, ottaa yksinkertaisuuteen ja yksinkertaisuuteen ja myös yksinkertaisuuteen ja yksinkertaisuuteen. Mennään yksinkertaisuuteen yksinkertaisuuteen. Joten minun jälkeen tänään on tämän takaisin. Miten vaikeudet ja selviöta perustumisesta kehittämisestä näytä virueen jekin saya, joka on yritys umpliäekonomisoituja vuodesta tai vuodesta. Tässä on umpliäekonominen yksinkertaisuuteen. Isä pyrkää routinegoikeusin järjestelmistä on suurin järjesteliäkin, ja monet small positive analysis on empirical work, serious econometric work on trying to estimate the causal impact of changes in, for example, taxes, tax policies on economic behavior, things like labor supply taxable income and savings. And then the idea is that how combining this background we here at VIDA could come up with a research proposal that would serve the purposes of getting useful policy relevant knowledge from regarding developing countries. So do I outline to my presentation now is that I'm going to first very briefly just remind you on what is this modern public economics approach to tax analysis. And I have a few slides about the current stance of tax revenues in developing countries utilizing the great data sets that was presented by the ICTD colleagues here. And after that I'm going to speak about the VIDA tax research. And unfortunately I'm now speaking more about our plans rather than actual results. So it's a scary thought that the talk is video and then somebody can keep me accountable for what I said after a couple of years to see if he actually delivered on anything we promised to do. So yeah, yeah, yeah. So if I'm not delivering it, stone is fall. That's obviously the answer. So let me start from the modern public economics approach to tax analysis. So also called as optimal tax analysis. So here the starting point is that the basic goal of the government is to collect revenues to finance public goods. And this should be done in a way that minimizes the tax distortions the tax system creates. And also at the same time the tax and the social protection system should achieve whatever the distribution goals of the government has. So in principle this gives rise to the basic efficiency equity trade off. And we need to have a balance between first of all minimizing the distortions and then achieving the distributional goals because achieving some of the distributional goals could actually entail tax systems that creates distortions on for example things like employment. One feature which I would like to emphasize regarding this modern analysis of taxis is that we don't make any artificial distinction between taxis and transfers. So if you like we can also see part of the expenditure status taxis and transfers being negative taxis. So the in essence the optimal tax models when optimal redistribution is analyzed covers both. Obviously also taxis have a corrective role but I skip that. Now on the empirical work on the impacts of taxation. Much of the current research actually emphasizes a notion called elasticity of taxable income or ETI which is defined as a percentage change in the taxable income TI whenever the one minus to marginal tax rate or the retention rate increases by one percent. So the benefit of this approach is that unlike some of the more traditional work it captures also other potential margins of responses. How the taxpayers might respond to changes in taxes. They can relate to increases in effort or also issues like avoidance animation. They are all captured by this elasticity. For the developed countries for the industrialized countries in a sense what are deemed as acceptable estimates of the ETI combines typically two things. One is the use of this large taxpayer register data and the second one is some sort of an economic tax policy reform. It generates variation across taxpayers which is fairly exogenous. So economists talk about credible identification strategies in these cases. Now when we go to taxis in developing countries this is the next small bit of my talk. This is an updated figure based which also appeared in the working paper you referred to but this is based now on the government revenue database which was released this year. And this is the revenues so this includes a total tax revenues including social security contributions for these country groups. And you can see that while there is clearly an increase even in the low income countries in the tax take the main thing is that they are still very much lacking behind in terms of the taxes out of the GDP on the current rich countries. But the point is that when we go back far enough in history countries which are currently rich also collected very little in terms of tax revenues. So this is a figure which I borrowed from a handbook article by Wesleyan Persson on this. And then you can see that the taxes share of GDP around the level where this tax ratio is nowadays in the developing countries. So a couple of observations based on this. So perhaps the main question in taxis and development is to how do countries go from raising approximately 10 percent of their revenues in taxes to something like 35-40 percent. So in other words how do they obtain more tax capacity. So to me because I'm trained in this optimal tax tradition this is an interesting contrast. In the poor countries we speak about that they should tax more. In the rich countries like Scandinavia some people think that we are taxing too much. And we are very worried about the distortions taxes create. So this begs the question that where is the tipping point. So when do the countries start to actually tax in a sense too much depending on at least some of us opinions. The second observation I would like to point out that how should countries design the tax and transfer policies as a system when they move. When they start to actually go from the 10 percent towards the 40 percent. I skip the remark on the field experiments because that was already covered in a mixed off for example. I now go to the part where I say a few words about our current research agenda regarding taxis and social protection. So there we have four main themes. The first is motivated by this by this need and desire to see the tax and social protection system as an entity. And one tool to actually do this is to build what are called tax benefit micro simulation models. I come back to them in a minute what they are. The second thing what we try to do is motivated by the fact that we currently as was pointed out by me already for here. We currently know too little on the actual impacts of taxes on economic behavior in developing countries. And then we try to use tax authorities register data for providing more evidence on these things. Then one issue is obviously improving the data access also macro tax variables. So we are committed to maintaining and conducting research using the ICT government revenue database. And we also have a political economy angle. This is the last bullet point here. We have a research program looking at some of the political hurdles countries may have when they build up the tax capacity and social protection system. There's a forthcoming symposium on these on political economy aspects to be held early next year. In this talk I'm going to focus on yes thank you for the for two issues. One is the tax benefit micro simulation models and the second one is the is the using register data to provide estimates basically on the ETI hopefully. I should mention that we also aim to conduct more theory work on on on on on designing cash transfer systems on the normative analysis that in the Merleys tradition. And also we would like to hopefully carry out some also work on the international tax aspects providing evidence on the extent of income shifting depending on on how much countries actually actually tax the multinational. So it is just trying to uncover this elasticity of income shifting. Okay so with a bit of you finish minutes I have I would like to say something about tax benefit micro simulation model building plans that we have. So so these models are combined to things so they build on a representative household level data on income and expenditures and then they also have a detailed coding of the tax and benefit legislation in the country. The word simulation refers to the fact that the user can can design alternative type of policies and then simulate the impacts on on on income distribution. The overall government revenues and for things like focusing incentives partial income tax rates or participation tax rates. The point is that these models are in everyday use in developed countries. And also I mean researchers find them useful because if you want to know the marginal tax rate we don't have any any other alternative than calculate these using my micro simulation models. And obviously these are essential for the developing countries once they try to build once they try to go from scattered isolated social protection programs to actual systems. Because then you need to also have an idea on the system wide impact of the policies plus an idea how to finance all those and attacks benefit micro simulation model is a tool to understand. Yet very few countries actually developing countries have access to such tools. So what we try to do is to is in a joint undertaking with the EuroMod team at the University of Essex and an institute called SASPRI which is a Cape Town based institute is to develop these models for a set of developing countries. Why EuroMod? Well EuroMod is both a model a micro simulation model for the for the European countries but it's also a statistical software for building micro simulation models and it improves and it's also creating comparability of policies across countries. One final point in two minutes so and this is about the use of revenue administration data in for tax research. We actually have very little published work on the impacts of taxes using credible identification strategies. The only paper I mean some were going on obviously by ICT the researchers on this but the only published papers are two papers on Pakistan. But it's clear that I mean Pakistan is not necessarily a relevant case for all developing countries so we need to know much more. So there we would like to do something similar what we currently are doing in the South African case and in also also other countries not only in Africa but also elsewhere. And it's a great opportunity for somebody like me to work to work using the South African data. So I'm at the moment I'm involved in a research project together with the person from the South African Treasury and two researchers to my fellow researchers from Finland. We are utilizing this South African register data to try to estimate the responsiveness of the corporate income tax base on the corporate tax system. And the tax rates they have in the South African case. So the idea is to is to is to utilize the fact that the in South Africa the small business income tax is linearly progressive. So that creates king points and the firms have an incentive to locate just below the king points. And because Kari is a ruthless mediator here I can say I would have these wonderful slides but I need to skip them so you can only blame Kari. So so but there's some wonderful work being conducted on on these and it's hopefully give then policy relevant evidence on the extent of elasticity of the corporate income tax base to tax rates. In in the South African context. So the conclusion is that the we are happy to participate in the field experiments and I think more of those should be done. But in a sense I would also step one. I would also take one step back and go to the world's these more traditional topics of building tax benefit micro simulation models and try to estimate the in a sense the the the impacts of the existing tax system on tax behavior behavior tax payer behavior and I stop there.