 Mika, please. Thank you very much, Yooka, and I have to say it was surely after supervising you I decided to leave the academic world But that was completely uncorrelated events But it's a pleasure to thank you again for for inviting me. It's a great pleasure and honor to be here It's been a I think a terrific event I've learned a lot and I've been impressed by the passion with which many of our presenters have spoken I'm English, so I'm passion is not my strong point But a few things perhaps I'd like to get your views on Maybe a good as good a place as any for me to start is there was a member an interchange yesterday between François Bourguignon my colleague Andy Berg on on the fund and its Place in the issues we're discussing at this conference And I think it's certainly true that in terms of the kind of big macro picture issues of links between Inequality redistribution Backer of performance growth now the dimensions has come to the fore at the fund in in recent years, of course it at a more Practical hands-on Country level we've been closely involved in many of the issues. We're talking about for for a long time for for many years many decades and certainly the part of the fund where I work I work in the fiscal affairs department and A large part of our business is what we call technical assistance and as many of you will know one of the Many benefits of membership of the fund is that you can ask for technical advice on a range of topics but including on tax reforms spending reforms and So the fund provides a lot of very detailed advice to countries on these issues and have I has done For many for many years and in fact one of the reasons you'll see there's a 50 on my on my slide Which is because this year is actually the 50th anniversary of the fiscal affairs department and so we've been looking back in the archives and It turns out that when the fiscal affairs department was Founded there was actually quite significant opposition to to forming it from a number of countries because they took the view that well It's clearly the fund should care about you know big T total revenue should care about big G total spending Should care about D some measure of the difference between the two But in terms of the kind of more detailed issues of composition of taxes and spending clearly to get into that you had to touch on very sensitive issues of Distributional concerns fairness you're really going to the heart of Social and political issues and they really think the number of countries didn't think that was Appropriate, but of course those countries lost FAD was founded But of course they were quite right as soon as we do give advice on these technical issues of I don't know reforming income tax Changing tax administrations all these kind of things think about spending policies We're getting right into the issues that are at the heart of a heart of this conference Typically, of course, we don't just by way of context what we typically do of course is we don't Impose our own views. We don't say the tax system isn't progressive enough spending is not progressive enough We tend to kind of spell out options thinking it's really for national authorities to take decisions on these wider wider issues But it does that background doesn't lead me into a first or first of a several general points I wanted to make By way of background to the kind of puzzle I wanted to talk about So the first point is of course, well, you know, there's a big literature of big debate about Essentially the equity efficiency kind of frontier Whether it slopes up or slopes down slopes up over some range Maybe down over others and of course that's something to we know. There are many reasons why it might slope up We know the reasons why it might slope down And I've all kinds of shapes and of course one of the things that some of the things we've heard about at the Conferences trying to do is to get to some sense of what that you know, what what we're talking about in this kind of dimension I'm not really going to be talking about that I'm really focusing on much more the kind of thing that we get into in advising countries Which is well, we may not know if this frontier slips up or down, but at least we would like to be on it That is if we can actually improve efficiency at no equity cost so I'm sure that that's a good thing and The other way around as well if we can kind of move northeast on this picture I always get east and west wrong, but you know what I mean do better on both Then surely that's a good thing and essentially a lot of our advice. I think what we could think of in a kind of metaphorical sense We're not calculating these things in a precise sense, but in a kind of you know, poetical metaphorical sense I think we try to get countries onto the frontier We're going to be talking about as some of the parent puzzles we have in trying to to even do that The second point I think is a familiar one and Joel's mentioned it But I think it's again just always worth stressing And it'll be a theme again in what I'm going to talk about is again Though this is a session on taxes of course when we think about redistribution inequality role of government It's really only it's really the joint impact of taxes and spending that matters It doesn't it really makes very little sense to talk about one side of the budget of the government's budget constraint or even one tax one spending item in in isolation and of course It's an obvious point, but it's still Thing that you know recurs all the time in public debate people talk about this tax is progressive Well, that's a well-defined statement, but in some sense, you know Doesn't really matter doesn't really tell you anything unless you tell me something about what's going on on the spending side as well Again, we all know this but again, I think we're not you know, we should insist on this I think rather more than than we often do even if thinking about the effects of of these simple measures We have a redistribution. Let's sort of think about the whole thing and that's the point I'll come back to and I think I put there as I've saying that we are bad really taking this holistic view and when I say we I Mean probably all of the communities that Joel mentioned. I was trying to think which community I belong to. I'm not quite sure Mainly the community of Chelsea supporters, but that isn't really very helpful observation For today, but I think we here means everybody is certainly it includes people who contribute to public debate on these issues really just Just get this wrong and misleading And we here certainly includes the fund and I think some of our other international organizations I may for this time I can say a little bit more about that But I think we've been very bad at putting into practice the this holistic view of tax and and spending instruments spending it here I mean both the kind of cash transfers of the kind Joel mentioned but also sort of the various items of public spending The third point I wanted to make is really nothing to do with what I'm going to be talking about But it's just a point. I wanted to take the opportunity to mention which is incidents A lot of these discussions we have about redistribution effects of government policy We talk as if we knew what the incidents of all these tax and spending things really was in terms of you know How the effector who really bears the burden who really gets the benefit and I think it's worth remembering reminding ourselves now and Again in the same spirit of humbleness Humility that Tony Shorrock said yesterday that we know remarkably a little about the incidents of many public Sector instruments even if you take simple things like the VAT We really know very little about the impact of the VAT in terms of who bears the real incidents We have some sort of studies for you know restaurants and things like that but in terms of you know Evidence that meets the standards we expect these days really very little We're doing some work on the VAT at the funder which I could talk about but when even for these simple taxes We know very little and what about the more complicated ones when you think about tariffs. What is the incidence of tariffs? Well, you know, we know they're partly a consumer tax partly a production subsidy all kinds of Cascading going on because on intermediate products figuring out the incidence of that is pretty hard. What about the corporation tax? Well, we don't really know how much of the incidence of that is on owners How much of that is on labor we could certainly talk more about that and so quite often people say well I don't really know about tariffs I don't really know about corporate tax or throw those away But of course in many developing comes you just thrown away about 20% of all revenue of course many the best studies don't do this But even the best that is I think are not not entirely We can't we just can't be entirely confident on the incidence many things similarly on the spending side with a number of cash transfers as well and So that's that's kind of that was that was fairly gratuitous. I'm afraid that was a point I thought was somebody at some point ought to make But let me then come on to the particular kind of puzzle that I wanted to talk about maybe puzzle may not be the right word That's really just to get your attention It's kind of a frustration as much as anything in that there are kind of areas of advice that we give at the fund What that we've been giving for about 20 30 years that never seemed to actually stick never did very hard to get countries to Follow our advice and this is one of those areas that I want to sort of raise with you and see if you have any Brilliant ideas about it So the issue is well We know that lots of countries try to support the poor by various measures of kind of generalized price support They may subsidize particular goods they may tax other goods or services at particularly low rates and they often do this Rationalize it on while this is our way of protecting the poor. We're raising revenue, but we're protecting the poor by this kind of differentiation but of course we know and As a general lesson that this could be a very expensive way of helping the poor Why is that well the basic point is straightforward if you think about say the argument that says well Let's have a low rate on food as a way to protect the poor So basically you the observation is well the poor spend a larger proportion of their income on food than the rich That's certainly true But of course the rich may well spend and a larger absolute amount on food than the poor They almost certainly do when we define food in this in this very very general sense So what does that mean that means that if you're forgoing revenue by essentially having a low tax rate say on food Most of the cash benefit of that actually goes not to the poor but to but to the rich This is just a kind of a mathematical Necessity and there are plenty of examples of this and there are two I think that strike home particularly to us one of course it is whole issue about reduced rates of value added tax So here are some numbers for Mexico. These are a few years old. It's OECD data So basically Mexico food base has a zero has a zero VAT rate And so you can think about well compared to taxing food set the same rate as everything else This is kind of effectively is acting as a subsidy now. What is the how does the benefit of the subsidy look? Well This is the subsidy received by income group as a percent of their of their income So this slopes down which is basically saying, you know the poorest percentile The subsidy is worth more to them as a proportion of their income than it is to the topic top Decisive the 10th decide so this is picking up the fact that the poor spend a large proportion of their Income on food, but now if you ask well of the money that's being gone for gone by the government How much of that is going to each of these different deciles? So we think of this as now a spending measure We're kind of spending a hundred dollars say by forgoing by having this reduced rate Who's getting the benefit? Well, this is the benefit as a percent of income and you can see the ones who are Essential percent of the total subsidy sorry so you can see that if you take the bottom Decisive they're getting less than four percent of the of the value of the subsidy and the top quintile the top Decisive sorry is getting more than twenty percent. This means we're spending a hundred dollars four dollars of which is going to benefit the poorest group 20 percent twenty dollars of which is going to benefit the highest Decisive looks a very expensive way of helping the poorest the other sort of classic example Of course is energy subsidies and so this basically is some Numbers put together by colleagues on the distributional effects Essentially who gets the benefit of fuel subsidies and this is average for over five or six countries And I forgot to put the labels on which quintile is which but you can pretty much guess The big area is always the top quintile So if you look at gasoline that's saying sixty percent of the cash benefit is going to the top Quintile go all the way around to the blue one the blue one is the bottom the bottom lot are getting three percent So they're getting a hundred dollars being spent. They're getting three dollars Similar for the for the other fuels a kerosene. Of course, it's a little bit different, but I won't I won't particularly linger on on that So what is this saying? This is saying? These generalized price support measures very extensive still are Very expensive. They're a very expensive way of getting money to the poorest But of course that doesn't necessarily mean they're a bad idea Suppose, you know, you have a just a maximum objective. Suppose your objective is simply your concern is simply with the welfare of the poorest person Well, if this is the only way you can get some money to the poorest person, maybe that's sensible policy You know if you if the only way you can get any money to the to the bottom Decisely in that Mexican case was by spending a hundred dollars of which the bottom will get four dollars If that's the only thing you can do well, maybe that's a sensible way to do it Maybe that's just the price you have to pay to pursue your equity objectives But of course the real issue is then this leads to the to the real issue I mean that the argument so far doesn't say these things are a bad idea The question is really whether they're better targeted ways to support the poor and as a reminder I mean these we're talking about subsidies that actually Do matter for these groups they can be a significant proportion of their income I have an example here for for fuel that says, you know fuel changing fuel prices really does matter for the poor So we do have a there is a real issue of how we can buy better ways to support them So are there better ways of targeting and protecting the poor clearly for advanced economies? The answer is or should be at least yes Inter in given that the wide range of instruments they have so this is another example that we did for there for the Merleys review in the UK that you can UK also zero rates food So this basically this chart basically shows the loss by income group if we simply Charged food at the standard rate in the UK and you can see big losses to the to the lowest deciles But on the other hand if we we've all kinds of income related benefits and social support measures We can upgrade in the UK and one way of doing that for example I think what we did was just increase them at the bottom by 15% you get this pattern of gainers and losers So clearly that the people at the bottom on our net are gaining as a consequence of eliminating the zero rating of food and Reforming the structure of income related benefits instead Top groups are not doing so well of course And of course just to bear in mind what you're doing here is your rate You're you're essentially also raising some revenue because you're only you're only compensating people at the bottom You're not compensating people at the top. So the way this works in the UK I think the numbers are something like we eliminate zero rating at least and when we did the exercise that raises about 25 billion pounds Compensation costs you at 11 billion pounds So you're left with a net gain of 14 billion pounds. This seems kind of like a pretty good a pretty good idea Well, I'll come on to that But I think maybe also just Also leads me I think for a little bit of a nuance on on on Franco's argument yesterday that you know in in advanced countries We sort of at the end of the we need to kind of out of the box ideas But how to pursue our equity objectives more effectively I think in some cases there are still some ideas in the box that we haven't quite scraped up and Used I think there are still things that can be done It's not to say that there are other things I think property tax I think and so on is is underused in many countries But I do think again, this is just an aside It's certainly true that we should be thinking about About out of the box ideas and one set of issues I think just to think about because I don't sorry. This isn't a side but just to say I don't we've thought about these enough is What technology can do for us? We've thought a lot about what technology can do on the administration side We haven't thought much about what it can do on the policy side Just one example is we tax income on an annual basis. Why on earth do we do that? We could certainly move away from having annual income taxes We could tax over much more much more sophisticated Ways than that a lot of the things in the Glitches some of you may know the sort of new dynamic public finance which link Which use information on various aspects of your behavior look at your income But also your savings indicators of your wealth a lot of these things actually look far less outrageous as practical ideas And they did a while ago, but again, sorry. I digress So advanced economies we can certainly do better than these badly targeted measures What about emerging and developing economies? Well, they're of course, you know We have to face the fact that the exit the available instruments are much blunter in that sense the case for differentiation for price support of various kinds is likely to be rather stronger but The question really is well If we essentially what we then have to think about is how well targeted are the spending instruments We have available spending instruments a week, but they nevertheless have some targeting properties How well targeted do they have to be to be better than? These kind of generalized price support and here's the sort of condition that you come up with if you think about this Suppose again suppose our objective is simply a maximum one We just care about the poorest person and we're thinking about raising the tax on some item Let's call it food and spending the money on something or other What are the conditions under which this the poorest person will gain? Well, it turns out if we ignore behavioral effects and I can talk more about that It's a condition like this You simply have to compare the proportion of all food they consume with the thing at the bottom there Which is essentially how much of the additional how much of an additional dollar of public spending do they benefit from and this could Be either sort of you know health. It could be some public good or sort of classic public good You multiply that by the valuation they place on that dollar which maybe which of course may be Different from a dollar if this is something other than a cash transfer if this if this condition is met Then basically you do better for the poor by putting the tax on the food and using the spending measure So let's take a couple examples One is cash transfers in that case this thing this this lambda thing at the bottom is simply one because we're just talking about cash Well, then the cut that condition simply becomes that your consumption share is lower than your share in the total cash benefits if that condition is met you benefit by eliminating this reduced rate and Having the cash benefit instead and suppose you simply take the case where what we do with the cash is we pay out a Pole subsidy we give everybody the same amount well in that case this policy Raising the tax spending the proceeds has got to be better for the poorest so long as their consumption is below the average of this Good, which is a pretty weak which is a pretty weak condition And just to say this is a pole subsidy you may think well pole subset is very hard to do well Actually, no not necessarily and I'll come back to that the classic example here is Iran So Iran had these huge oil subsidy petroleum subsidies absolutely huge Essentially, then undertook a major exercise that involved paying up all subsidies same amount to everybody It's a huge thing. They had to set up ATMs all over the country And it worked maybe worked too well, but we'll come back to that later But it worked in the sense that once they did one reform of the subsidy Would of the of the subsidy to energy prices by the time they people had received their cash transfers They were asking well when is the next reduction coming because we're looking forward to getting some more cash Of course if there's some pro-poor element in these in these cash transfers, that's going to be even better than a pole subsidy What about in kind benefits? Well in kind benefits are a little bit more complicated because now we have to think about Individuals valuation of these goods, but we could have for example in my earlier expression now lambda is greater than one We value you know a dollar of health expenditure received is worth more than a dollar to us So and then we get into this issue of benefit incidents that is which I'm by no means expert on But here's one example for example for which is for India, which is looking at the essential distribution of Spent public spending on curative health care by by death cell So if you look like one time, so if you look at the bottom 10% They're getting they're getting 10% of the benefit the cash benefit of this kind of spending What does that mean that means they're going to do better By increasing tax financing this if they account for less than 10% of food consumption, which is pretty Quite plausible that they would be food. I'm saying food I'm thinking of kind of as a you know a metaphor for whatever the item is we're thinking of having a reduced rate on So what is the moral the moral seems to be the price subsidies can be so badly targeted That spending measures don't have to be very well targeted to be better even in kind of low-income countries There's some work on this in Ethiopia, but there's certainly more needed To bear this Conjecture out lots of qualifications here. We could have kind of inferior goods Just to jump to the last one, of course, there's always going to be in all these reforms There's always going to be someone who can't be protected that is very difficult to protect or to be able to ensure Benefits from these spending measures. That's just the kind of problem. We have an all kinds of policy You're there's almost inevitably going to be someone some group who's going to suffer so I'm trying to make a case that in many instances even in countries that have don't have very sophisticated spending instruments We can probably do better than these kind of generalized price subsidies including Reduced rates of tax on various things. So why on earth aren't these better targets what these better policies adopted? You see that the thought is we're not on the frontier that the suggestions with these policies They mean we're not on the equity efficiency frontier in Some kind of Bruce broad loose sense. Of course one argument is well Actually, you know, you've really ignored efficiency considerations here that when you do these kind of reforms You are going to play around with marginal effective tax rates if I take for example the the UK tax reform where I eliminated zero rating and then kind of Strengthened income support at lower income levels Mathematically it must be the case that over some income levels. We're actually increasing Marginal effective tax rates. That is we're making earning income less attractive That must be the case Because the wider question then is the kind of optimal tax issues that Joel gets in the Joel's referring to We have to think about whether you know the case whether that's going to have an adverse effect on work incentives Actually the empirical evidence on that on the food example in the UK is not very strong But such as it is it actually goes in the opposite direction says that these things should be taxed more heavily for those reasons What about in efficiency considerations in emerging and developing economies? Well, those I think are much less developed than they are for for advanced economies So we have this whole apparatus of well-developed optimal tax ideas to to guide us Well, if we put efficiency considerations aside then clearly we're left with sort of political economy arguments of various kinds What might those political economy political economy considerations be? Well, one is of course that well just as you know the poor aren't receiving much of the benefit the rich are getting a lot of Benefits the rich are the powerful. So they're the ones who have an interest in actually keeping these things going There's clearly I think in many cases some kind of distrust that when the government says I'm going to I'm going to raise the tax on This good, but don't worry because they're not going to compensate you on the spending side It may be natural for many people to be rather distrustful that that will actually happen or to believe that it will be sustained I'll come back to that. I think there are clearly cultural sensitivities on particular goods For the UK, I think it's not as I think about the UK. Why don't why do we still have zero rating on food? My conjecture would be that this actually probably goes back at least nearly 200 years Since the repeal of the Corn Laws taxing food has been political death In the UK the Conservative Party has split twice In the last 150 years both times on taxing food So there are sensitivities It could be that you know you could tell us sort of slightly over clever story to me over clever story that says well The way you signal as a politician That you really care about the poor is that you implement really very costly ways of helping them because that's the way you signal How much you're willing to to forgo to support them and less important these days? I think there are issues of stigma What might then be the ways ahead if we accept that this is you know, there's this kind of Potential to move closer to a frontier. How might we actually do that? Well, you might have well certainly we at the fund rather hope that crisis might help but if you look at For example, what countries did in advanced economies? You know was it you know, we might have had some naive hope that when the UK was thinking about its revenue needs They might think about doing something on the zero rating of food rather than playing around with the rates But of course in the end of the end of the day they played around with a rate rather than doing something on the base That's pretty much true in in most other advanced economists They didn't bite the bullet in terms of some of these rather pointless rate differentiations. They have clearly there are cases Which crisis help if you look at Egypt the last few weeks With quite dramatic action on subset is that clearly just comes on the fact that this is just fiscally Unstainable this is you know, I don't know what the last number was it's eight nine ten percentage points of GDP Going in substance. You can't use it at some point. You just can't carry on What other things might help well this whole of course we like to talk about transparency that what you need to do is make It clear to everybody that how expensive all this great differentiation is well, that's certainly true but I Think it's clearly not enough. You know if you think about the UK again It sort of tends to be my favorite example. There's no great mystery in the charts I gave you there's no great mystery in the fact that you could Eliminate zero rating and make the poor better off and have some more money It's not that no one understands this. It's simply some other reason. It doesn't it doesn't happen What about ear mocking ear mocking is the idea of Hypothecation which says well, let's actually try to say to kind of commit ourselves to overcome the distrust issue I mentioned let's have a commit ourselves to say well We're going to raise this revenue But we're going to kind of somehow tie our hands to make sure we only spend it on This particular stuff that you like which might be health education transfers whatever and sometimes that has been used by governments Ghana raised the vat rate from 10 to 15 by by ear marking Of course the trouble with ear marking In a sense either it really constrained there are two possibilities One it really constrains your spending on some items. It's really the marginal source of finance Well, if it's really constraining your spending You have to think what is that constraint one I want? Do I really want my education spending to depend on how much revenue I get from the vat this year? Maybe not on the other hand if it doesn't really constrain your spending if it's just Can some kind of trick that you kind of sell to people? It's not really going to constrain it because we have money is fungible will finance this stuff somewhere else Then this is completely then presenting in its ear marking is misleading non-transparent all these other things On the other hand you could say well really as a last resort if this is the only way you can do it Maybe it's a price worth paying but I think does also point to links between tax reforms also and Changes to the system of public financial management, which are designed to give people more faith in That money is going to be well spent Lessons from subsidy reform of course I have to give a small plug to the book produced by my colleagues at the fund which looks at essentially a whole series of episodes of Energy price reform efforts in in a range of countries both successful ones Unsuccessful ones comes out with a number of number of issues here the idea that you should have a comprehensive reform plan communications sequencing so on I Think the evidence is those those things all matter crises in some sense cases can also help But I think perhaps the last part I'd make is just and I'm not sure Maybe this is going to sound a bit more optimistic than I thought and again. I think it's picking up something Francois mentioned yesterday, which is that in many ways the possibilities of what we can do in particularly lower income Economies are actually changing quite rapidly if you think for example just about biometric cars biometric cars basically identify you That essentially in principle helps you out of the main problem of giving a poll subsidy the main problem with that has been identification potentially that becomes much more much more something one could think about just a Straightforward poll subsidies have a you know have a number of theoretical and political attractions We know that targeting by income or cruder indicators of needs, which we do in some You know, we've been doing it and it's been being done in in the US for a long time But we know it's also becoming much more feasible in lower income countries again, Egypt Is an example where as part of the removal of the breadcubs it is They essentially use smart cars that limit you to a certain number of loaves per day And the bit that you don't use you can actually spend as I understand it You can essentially use as as cash. So all these things are I think helping us See better how we can address some of these issues and maybe Bring ourselves closer to the front here, but so but I'm slightly torn that in one sense That makes me optimistic, but I sort of keep coming back to maybe maybe I'm too fixated by it But it's still come back to the UK example where everybody knows what a sensible policy is No question of feasibility, but it doesn't happen. So with that you've got thank you very much And thank you for your attention. Thank you