 Good morning everybody. I'm delighted to welcome you to this sixth Biennial ECB conference on fiscal policy and and your governance Unfortunately, France Mets couldn't make it. So I was right now asked to chair this session My name is Etore Dorucci and I had the fiscal policy division here at ECB This conference is actually a well-established conference For which over the years We have been hosting very fruitful dialogue between central bankers academics policy makers market participants and Which has promoted the new and influential research on fiscal policy And on the reform of MU architecture We plan to do the same also in the coming two days The organizers who I would like to to thank You can see some of them over there Jacobo is there demos Leo And others So the organizers have really put together I would say a very nice program rich program Which includes eight academic papers a keynote lecture by Olivier Blanchard this afternoon Which will be chaired by our board member is a best novel and in a dinner speech by Victor Gaspar A policy panel that will be chaired by Philippe Lane and will include Sylvia Ardanya Royal Bedsma, Valdez Don Vrostis and Jérôme Zettelmaia Today we will have two sessions of academic papers Respectively one on fiscal policies in monetary unions and the financing of fiscal deficits And the other one on the design of fiscal rules and fiscal Consolidations tomorrow the two sessions will cover macroeconomic effects of fiscal policy and monetary fiscal interactions in a monetary union Before starting with the first paper, let me remind you a few housekeeping rules Each paper will have 25 minutes for the presentation 15 minutes for the discussion and 10 minutes for the general discussion and Answer from the speaker Since the conference is a tight schedule For the discussion and Q&A we will give priority to the questions from the floor Coffee break will be on the terrace on the third floor just in front at the end of this corridor and On internet you can find QR codes over there or outside the room. So If you want to connect, you know, how to do that So without further ado, let me now give the floor to Eugenia or If Ramon hasn't arrived yet, right? Okay, then in this case Welcome also to Christian Wolf Let me maybe just briefly introduce the two presenters so Christian will From the MIT will present a paper on can deficit finance themselves co-authored with George Marius Ankelatos and Chen Lian and the discussion is going to be David David a debauchery from university university Fabra How Eugenia from the Gonzales Aquado from the to lose school of economics will present a paper co-authored with the raffer burial Patrick Kihoi and Elena Pastorino the discussant is going to be Ramon Marimon from the University Pompeo Fabra. So Christian the floor the floor is yours. Okay. Good. Okay, perfect. Thanks. Thanks for the invite. Thanks for having me I'm presenting this paper with Mario Sanchen on whether fiscal deficits can finance themselves. Okay, and this This works perfect. Okay. Good. So here's the setting of the paper When a world where r is bigger than g and there's a fiscal deficit right now say because the fiscal authority send out transfers to households The question we're going to be asking is because within our bigger-than-g world something will need to give to finance that initial deficit How is it going to happen? So this is a this is a theory paper and Does it no yes There you go, maybe a bit of a lag. Okay, good. Hey, so it's a theory paper We're going to be answering that question in a in an environment with the following and two two key ingredients So number one on the private sector non-policy side of the model There's going to be number one non-Ricardian households So that's going to mean if there's an initial deficit because households are non-Ricardian It's going to influence demand and number two output is at least gonna be partially demand determined because prices are sticky Okay, good. Then on the policy side the policy mix is going to be what I'll refer to as conventional in the following sense Number one the fiscal authority is going to be Ricardian in the classical sense meaning there's a deficit today And it does stand ready to hike taxes just maybe not immediately but down the line to balance the books and number two The monetary authority will not be accommodating that initial deficit real rates will not be declining the Taylor principle will be satisfied Okay, now in that environment, we're going to get at least our show This is a bit unfortunate Like sorry guys, okay Like is it gonna be like that throughout there'll be a bit of an issue. Okay. I hope it's um, yeah Okay, okay good, I'll try to anticipate and like click in advance. Let's see how let's see how that goes Okay So what we're gonna get is kind of because of these ingredients what's gonna happen is there's a deficit today demand will go up That will mean in general equilibrium Because there's a demand boom Output will go up so the tax base will go up and as a result of that even without the fiscal authority doing anything Tax revenue is going to be going up at least a little bit So that will contribute to the financing of the initial deficit Number two if government debt is nominal and the boom also induces an increase in prices inflation Then the real value of government debt is going to be eroded I'll refer to both of those as self financing in the sense that both of these Kind of lead to financing of the initial deficit not via the government actively adjusting something but via equilibrium responses of prices and quantities Okay, so far so standard that this is what's gonna happen in this environment now the Okay, good, but this was quick So the kind of the core result of the paper is now going to be that and of what I'm gonna be varying is the horizon of the Fiscal adjustment the horizon which the government says I'm gonna be hiking taxes to balance the books And the core result is going to be that number one as you delay the horizon of financing more and more These two self financing margins output going up so tax revenue going up and prices going up So the real value of government debt being eroded will become more and more important And you're gonna be limiting to the case of full self financing. So as I as the government Delay my adjustment more and more Indogenously prices and quantities will respond in exactly the right way to fully finance the initial deficit Okay, and the strength of nominal rigidity So just be governing the split between the two if prices are fully rigid it all comes via the output tax-based expansion The more flexible prices are the more important the initial spike in inflation is going to be okay. Good. So that's the outline of the talk Environment core result intuition and then a bit of quantification at the end good So let me let me jump right in unless there any questions and feel free to interrupt me. Okay. Good So here's the environment In a second Let me already sketch the environment because it's kind of it's going to be relatively vanilla as I've kind of described it here It's going to be essentially like a small scale kind of standard New Keynesian New Keynesian environment I think on the supply side and on the policy side the key twist is going to be on the household side How I'm going to make households be cardian and how that is going to lead to kind of these demand effects initially Kicking off and how that's going to be interacting with the horizon of fiscal adjustment. Okay, so Two blocks non-policy block and policy block. Let me begin with the non-policy block and in particular the aggregate demands spending relation so there's going to be a unit continuum of Overlapping generations of households that's going to be my kind of my reduced-form stand-in for non-recardian non-recardianist With a period to period survival probability of omega It's between zero and one if omega was equal to one guys I just infinitely live this would be the usual representative agent permanent income Ricardian consumption consumption savings block I'm going to be looking at omega less than one and you can in reduced-form here interpret one minus omega as the probability of Borrowing constraints binding from period to period. It's just going to effectively shorten the planning the spending horizon of households These guys are going to be Investing as usual and actually fair annuities So I'm going to have that that this confector is beta beta times one plus our bar the steady-state real rate of interest is one So our bar is bigger than zero. So something will need to give to finance any initial fiscal deficit Now out of this consumption savings problem of my overlapping generations block of households I'm going to get the following aggregate consumption Relationship and that relationship if the economics of it are key for anything I'm going to say today So I want to take my time going through it because that's what's going to be driving my main results so What we see here is Consumption on the left-hand side as a function of current household wealth dt as well as current and expected future income and tax payments Y&t the real interest rate term are on the right-hand side I'm going to return to later on for now we can largely ignore this That's why it's grayed out now Let's suppose first of all for a second that omega was equal to one So we'd have the usual permanent income model in that case What's your MPC where your MPC is just if you have money you can you consume the annuity value of that So it's one minus beta or equivalent the R bar over one plus R bar So you have some wealth dt today your MPC would be one minus omega very small now as omega gets smaller So we have non-recardian households as we know the MPC is going to get elevated in a literal sense because these guys expect to die So they're front-load their consumption in a less literal sense because they expect boring constraints to bind Okay, good. So that's the first key feature that will make my results take the elevated MPC Then there's this so that's what you kind of see in the left-hand side the one First time on the right-hand side the one minus beta times omega then out of current wealth dt And then we have the income term so current and future income yt less current and future tax payments tt Okay now usually in the permanent income model We just be discounting that at the steady state we are rate of interest So that's why you have the beta beta squared beta cubed and so on now What's the permanent the what are the binding borrowing constraints going to give us they're going to give us additional discounting So we're going to see that any future stream of income and any future stream of tax payments It's going to be discounted incrementally more. That's where the omega is showing up again The little interpretation is your consumption today doesn't respond as much to your income tomorrow because you expect Perhaps to die in a less literal interpretation because you run into binding borrowing constraints between today and tomorrow Okay, good. So these two key features of demand the elevated MPC and the discounting I've got to be what's driving my core results throughout the entire paper Rather than the literal olg micro foundation as and later as i'm going to go quantitative and connect to the hand literature It'll be those two key features that will continue to drive my results just with this stuff here I can easily prove stuff. Okay, good. So that's the demand block now the rest Kind of the supply set I can go through relatively quickly if the technology cooperates. Let's see There you go. So these households supply labor as usual They're going to be the usual nominal rigidities among the sticky price retailers Taxes are also for the main part of the talk because not going to interfere with my main results Going to be lump sum just to make things very easy and out of all of that. We're going to get the vanilla textbook New canes in philips kerf relating inflation today to output today expected inflation tomorrow Okay, so these are my two my first two relations the demand block equation one and the supply relationship equation two Okay, now to complete the environment. Let me go through the policy side Um, so there's there's a monetary authority. There's a fiscal authority Okay, good. So there's going to be two sets of relationships that we begin with the monetary authority The monetary authority will be setting the nominal rate on one period nominally risk free bonds I'll denote that nominal rate by it and for the main part of the talk I look at the following simple kind of rule Where the nominal rate is set so that the real rate or the expected real rate is equal to phi times output So phi you can interpret as if it's positive the monetary authority is leaning against any fiscally induced boom If it's negative it's accommodating For the main part of my talk, I look at just to make the math very clean and simple I look at phi equal to zero. So an acyclic real rate. That's what the monetary authority is doing So no accommodation, but also no particularly aggressive leaning against I'll relax that at the end Okay Now key part the fiscal policy because that's what i'm going to be varying this kind of key parameter the horizon of fiscal adjustment So the fiscal authority issues nominal debt bt In real terms, this is what its budget concern is going to be looking like Real government debt going into tomorrow is just debt today dt less tax revenue tt Humiliated up at the real rate of interest if real rates are away from steady state Say if they're higher then this increases the debt burden that's the second term you see And finally because government debt is one period nominally risk free if there's surprise inflation That will erode the real value of government debt So we see here. How is any deficit going to be financed? Well, there are two main avenues for me the blue ones its taxes and its surprise inflation Now, let's dig a little bit deeper into how taxes are determined because that's again the key thing i'm going to be varying Okay, there are going to be three terms. Let me begin at the at the end epsilon t That's kind of the experiment i'm looking at a shock a lump sum transfer to households So enters negatively so surprise taxes are going to be somewhat lower for a period That's going to be increasing the deficit. That's my experiment the shock And whose propagation i'm going to be studying and then there are two additional terms The middle one is the tax base channel from the introduction if output goes up Then mechanically even without the fiscal authority doing anything tax revenue goes up by some coefficient tau y So in the data think of that as like so the average kind of tax take and transfer response like 0.3 or something A dollar of output improves the fiscal balance by 30 cents Okay, good And then there's a final term which is the actual fiscal adjustment tau d It's going to be between zero and one if it's one and fiscal adjustment is very quick There's a deficit today. I immediately hike taxes to balance the books The smaller tau d is going to be the more delayed fiscal adjustment is going to be the weaker it is Okay, and to make it even clearer what i mean by delayed fiscal adjustment I'll also be looking at the following variant rule kind of a non-marcovian Time dependent generalization of of the rule equation five that you see here, which should come down there in a second There you go No, come on Which okay, there you go Which just makes this tau d coefficient time dependent and says there's my initial deficit my initial shock at date zero Then for h periods tau d will be zero There's no fiscal adjustment whatsoever and then after h periods where h is something i'll be varying tau d is equal to one Which just kind of in english means i have a deficit today I don't do anything for h periods and after h periods i'll tax So ever much is needed to bring my debt back to steady state So a very clear example of a ricardian rule just a delayed a delayed adjusted ricardian rule Just one quick question. How much time do I still have left like if um you still have um 40 minutes 40 minutes, okay, perfect because I didn't quite keep track of when we started Okay, good now one more slide on the environment and then i'll get to the to the headline results So what i've sketched now equations one to five under my the assumptions i'll be maintaining throughout that omega is less than one So households are at least marginally non-ricardian and number two tau y is bigger than positive bigger than zero So this tax-based channel is operative This economy i've just described equations one to five has a unique bounded equilibrium That's the equilibrium i'll be studying and i'll be focusing on one particular property of that equilibrium one particular object Whose properties i want to characterize um There you go. So my question is how are fiscal deficits? How is this initial shock epsilon epsilon t this lump sum transfer shock to households? How is that financed? Okay, I can just look at the government budget constraint iterate that thing forward What i'm going to see mechanically is the initial deficit can be financed in one of three ways The first one kind of the first term on the right hand side is the kind of conventional fiscal adjustment The government with this tau d term actively responds to a higher deficit to increase taxes to balance the books Okay The other two terms are going to be what i'll refer to as self financing number one In general equilibrium the initial deficit may be leading to an increase in output That will mechanically increase tax revenue That's my output self financing term and number two that will may also lead to inflation surprise inflation The value is real government debt. So that's the price self financing term Okay Now i'll refer to the overall share of these self financing terms as new So something that's between zero and one and I want to characterize how that share of self financing the relative importance of those terms Is going to be changing as I change the horizon of fiscal adjustment. So as I vary tau d or as I vary h Okay, so here's now the core theoretical result of the paper What i'm going to do is i'm going to just formally state the result Then give you kind of just Communicate the content of the result via a figure and then we're going to go through the economic intuition of it What makes the result tick and then at the end we're going to have the quantification Okay, good So same assumptions as before omega less than one omega less than one and tau y strictly positive Then the self financing share new as the following two main properties Number one is monotonicity. It is increasing in the delay of fiscal adjustment So as I lower tau d or as I increase h the importance of these two self financing channels price and quantities is going to be increasing Okay So more comes via those relative to the actual unequilibrium fiscal adjustment the government initially promised to do Number two is a limit result As I increase the horizon of fiscal adjustment more and more in the sense of I increase h or I lower tau d New is going to be converging to one Okay, so all financing turns out to be along the equilibrium transition path self financing So say h is reasonably large the government said I would stand ready to balance the books h periods from now But on equilibrium it doesn't turn out to be necessary Turns out to be the case that the boom the initial deficit induced Is of exactly the size required to finance that initial deficit. Okay And just two more little properties of this So number one is so along this equilibrium transition path government debt will just on its own Gradually return back to steady state and number two nominal rigidities Then just govern the split between quite prices and quantities In particular the more rigid prices are the more it comes via output And that just mechanically means if prices are fully rigid the cumulative output multiplier needs to be one over the tax rate Just as a matter of arithmetic, right? good fortunately, it's Okay, now it's coming in. Okay See if I had known this I wouldn't have put in so many pauses in the slides. This is kind of unfortunate I'm already dreading the next slide, which has like many things popping in and out. Oh, well, uh, okay Trying to move ahead. See I don't know. Am I supposed like does it have if I point this somewhere? Uh, no, no, it doesn't. Okay. Good fine It's fine. Okay. Good. All right. So what am I showing here is a visual representation of the results So left and middle panel are going to be impulse responses of output and government debt to this initial fiscal deficit Chalk epsilon zero under different assumptions on the fiscal adjustment horizon h Okay, that's what the left and middle panels show us the right panel will show us this self financing share new And it's split into prices and quantities as a function of the adjustment horizon h on the x axis To give you a quick preview for this slide like just to contextualize it, right? This is supposed to just visually illustrate the contents of the result Don't take them the magnitudes at this point seriously neither speed of convergence nor split between prices and quantities Okay Now what I'm doing here and these are the things that will kind of gradually pop in which is the decision I'm now regretting is I'm increasing h. Okay. I'm increasing the horizon which I'm doing the fiscal adjustment So what do you see here as I'm increasing h? So the first lines were h is equal to zero which is immediate fiscal adjustment Which is kind of schizophrenic. I give you money today at tax immediately Of course, nothing happens and the self financing share zero as I increase h. I'm going to get a transitory boom That's what you see here So government debt is then elevated up until date h and then by definition of my rule goes back to zero So to steady state at date h Okay Now what happens as I increase h you see as things ideally slowly pop in you see a more and more persistent output boom on the left hand side In the middle you see government debt on its own gradually going back to steady state And on the right hand side you see as I'm increasing h this importance of self financing this Scala new it's going to be increasing and it's going to be converging to one Okay So kind of just the message here is that as the government commits to doing the adjustment later and later Instead already the adjustment will be happening along the equilibrium transition path through prices and quantities adjusting Okay, good. So that's the core content of the result illustrated visually So things Supposed to be one more thing popping in um There you go. That's what I just said. Okay comes via prices and quantities. Good So now I want to communicate the core economics the economic intuition for the result That's kind of the key thing I want to communicate and then if we still have some time left I'll do a little bit quantitative stuff theoretical extensions and so on Okay, so here's what makes the result tick and why non-recardian household behavior interacts with delays and fiscal adjustment in that particular way So let me start by already pulling in a bunch of things in advance Okay, so I want to communicate the intuition as it turns out this kind of infinite horizon model I just sketched has very precise and formal analogs and a kind of dinky seeming excessively simplistic two-period essentially static Keynesian cross model Okay, that allows our tax-based channel So here's what happens in this kind of mickey mouse model and then I'll try to convince you this has analogies in our full environment So consider two-period Keynesian cross model Where in period zero the government gives money to households and says I'll tax you at date one if needed to balance the books Households at date zero are fully myopic. They see the money today They don't respond to the tax site tomorrow. Those are my strong assumptions I'm imposing for now an output is fully demand determined at date zero usual Keynesian cross Then what's output at date zero? Well, I give you the transfer. How much do you spend the mpc times the transfer? That's how much demand has increased. What's the Keynesian cross amplification? Usually it's five minutes. Usually it's one over one minus the mpc But now I have this tax-based channel So it's one over one minus the mpc times one minus the tax rate because some dollars go to the government So what's the self financing share? It's just the output response times tau y because I in this rigid price model. I only have the tax-based channel Okay, good. Now. What do we see number one? This channel is increasing in the mpc the strength of this self financing channel Obviously because the demand boom is bigger and number two It's one if and only if the mpc is one right because then on the right hand side The numerator becomes tau y the denominator becomes one minus one minus tau y So also tau y the tau y is cancelled. It's one. Okay, good So what does that have to do with the environment? I just described Well, the dynamic economy has a as precise analogs to the static Keynesian cross in the following sense Let's begin with what I call partial equilibrium. So prior to prices and quantities adjusting In that case the government says I give you a dollar today And I tax you a dollar times the accumulated interest age periods from now. How are you as a house? So going to respond. Well, if you're alive at the beginning, you see the one dollar you get You think the tax in the far future someone else will need to pay Okay, so you see the dollar at the beginning and you spend it and you spend it pretty quickly because you again literally expect to exit the economy at some point Okay Conversely for those people a lot alive around date age. They see the the tax hike so they spend less So what do you see prior to prices and quantities adjusting is the figure on the on the next slide You see this. So what you see here is demand over time some time on the x-axis And the amount of spending on the y-axis So what do we see demand goes up at the beginning by how much? Well, in that present value of one dollar because I gave you one dollar and you spend it kind of quickly Which is a little bit like our static Keynesian cross you call me an NPC of one Just now period zero is the short run And then what happens in the long run when the tax hike comes online Well, there's a bust a spending bust of minus one dollar in that present value terms because the government isn't creating money out of nothing It says I give you a dollar I take away a dollar in that present value terms So there's a spending bust later on if I had ricardian households This would just be a line of all zeros because these are infinitely lived guys With my olg or hank households. It's a one at the beginning and a minus one at the end Okay, good. Now. What happens in general equilibrium? Well For now, let's suppose output is fully demand determined. So demand has gone up at the beginning So income has gone up whose income income of the people alive at the beginning So when do they spend it at the beginning? Okay What happens with this additional output? Well, some of it goes back to them. How y goes back to the government? Okay, so what do you get is a general equilibrium multiplier that will bid up to one over the tax rate at the beginning So that's again the mpc of one in our static Keynesian cross where again period zero is not the short run Everything gets spent in the short run. So you're bidding up to a cumulative multiplier of one over the tax rate So then what happens in g e in this figure? I increased this initial boom from one to one over the tax rate So then how much tax revenue have I generated by like date 50 or something? Well one So then do I need the tax hike at date h? Well, no, I don't okay. That was my initial fiscal rate said hike tax is if needed So that's the core intuition. So if you got this you understand why in our olg environment It works out and why in a hang type environment you would get exactly the same because mpc's are elevated at the beginning And spending is front loaded because of that Additional discounting embedded in the consumption function those two key features were everything that's driving my results Okay, good. Now I have like a what two minutes left. Yes. Yeah, okay. Good. So let me Let me just give you kind of a very quick overview of what the remainder of the paper is doing If I can there you go So kind of the paper has a discussion of practical relevance in a dual sense Number one It kind of extends the theory in multiple directions and here's just kind of a quick sketch on the fiscal side That distortionary tax and government purchases don't really change much Should be relatively transparent and talk about this more later if you're interested on the monetary side That's kind of a little bit more interesting if the monetary authority accommodates So phi is less than zero then all of this kind of Keynesian boom stuff happens even more quickly So the result goes without change if the monetary authority leans against the boom Then the boom is delayed you may still get this self financing to play out if the monetary authority is not too aggressive Perhaps the most interesting implication for this audience is that if monetary policy is very very aggressive Then for a bounded equilibrium to exist Fiscal adjustment actually needs to be quick enough Usually that in the textbook we just say as long as fiscal policy is passive So at some point there's fiscal adjustment we can forget about the fiscal side Okay, that's why we just have our usual three equations here We're saying if monetary policy is very aggressive then fiscal adjustment needs to be quick enough So the requirements of what passive fiscal policy means become higher So that there's another point we can maybe discuss in the q and a And then on the economic environment I've already clicked. I anticipate it will be slower than this Main point is You can introduce investment doesn't really change much or you switch to a full-blown quantitative bank model I can't prove stuff anymore. I can show it numerically Okay, final point here is just then in kind of a calibrated quantitative version of our environment What do we see here in thoughts responses of output and inflation to a one-off fiscal deficit? Under different empirically relevant assumptions on the speed of fiscal adjustment Which is now my the tau d policy that we can take from the data How large tau d is and what you see on the right hand side These three dots are kind of three numbers of how quick fiscal adjustment has been in the data from prior work You see with those numbers you get relatively meaningful self financing shares So the fiscal boom induced by an initial deficit That's only gradually finance is already large enough to quite meaningfully contribute to its own financing in this environment Via the induced output boom and the increase in inflation Okay out of time. So let me just wrap up with the main thoughts. I want to leave you guys with So kind of the core theoretical insight of the paper is In kind of an environment that people nowadays very much used to think about fiscal policy and its propagation If in that environment you have a deficit and you delay adjustment They are very strong forces to the financing instead coming via price and quantity adjustments along the equilibrium transition path And that kind of in our view has some interesting theoretical and practical implications Theoretically it relates somewhat to the to the fdpl literature But here it's grounded in a very classical failure of recording equivalence Okay on the household side. So you don't need to enter into the debates of Can the government commit to never adjusting right which leads to this discontinuity in the fdpl literature Here is just it smooths things out as you delay your adjustment all of these forces are going to be there But we also very much shine the light on adjustment coming via the output tax base expanding rather than via prices final thing in practical terms Self-sustaining self financing fiscal stimulus Maybe at least somewhat less implausible than commonly believed in an environment with the key features I've stressed So you want to think that monetary authority is not too aggressive is somewhat accommodating the initial boom And output is at least partially or even largely demand determined So you get things via the tax base going up and not so much via inflation Sometimes these assumptions will be satisfied. Sometimes they would not be okay. Good. That's all I have. Thank you Thank you very much christian for this very interesting presentation So not only fiscal adjustment was delayed but also the point that yes, yes Okay, I give now the floor to davide All right Well, first of all, I want to thank the organizers for giving me the opportunity To be here today and discuss this very interesting and thought-provoking paper. So There is a long lasting debate in the fiscal policy literature about how to finance fiscal stimulus whether it should be self Whether it should be tax financed debt financed and this paper explores a third option The one that we are all hoping for that is whether it could be self financed Um All right, so it asks two questions Is it possible in theory and then is it possible in in practice and the answer to both questions In this paper is yes Let me describe start describing what is the mechanism So the key idea is to think about an environment with two key ingredients One is the presence of nominal rigidity So that makes demand the demand channel particularly active in stimulating economic activity And the second channel is the presence of household with finite lives Or liquidity constraints. So the main idea is that If we today stimulate Provide a fiscal stimulus, but we increase we finance it with taxes somehow in the future Because of finite lives or liquidity constraints people will not fully internalize that taxes will increase in the future And as a result, they don't save Some of their income to to pay for the future tax increase So this is illustrated in an olg framework With liquidity constraints, but the same logic would also be correct if we think about environments with You know some behavioral Assumption like bounded rationality finite or landing horizon and so on the logic will be will be the same So the key policy implication is that if we push fiscal adjustment Farter enough into the future this will create Because of these myopia or you know Finite lives or liquidity constraints it will generate a boom in the present All right Now the second question is plausible in practice and again as I said the answer is yes the authors provided quantification in a Model that is an olg model or also in the appendix. They have a fully fledged heterogeneous agent new kinesian model and where they find that The degree of self financing is substantial In it could be substantially in our in our economy And of course this is a very thought-provoking result. So let me see if I well here The my discussion is going to be organized in the following way So first I will try to summarize briefly the main mechanism again to provide You know a context and discuss the the results of this paper in the context of the literature And then I'm going to raise two Main comments that is well I agree that in theory it is possible to have full self financing But I'm going to argue that there could be alternatives way to get self financing that do not depend on the delay of fiscal adjustments And the second is that if we take at least literally the result that there is a substantial Degree of self financing in our economies. This opens up potentially some empirical puzzles And I will try to offer some possible explanation even though of course It's a more of a conjecture than than I saw a result At this stage, uh, so let me get started by looking a bit at the main mechanism. So Let's see if I can do it. Okay here So, uh, the starting point is a very simple government budget equation that holds true in a variety of context And that we all know so the evolution of that that depends on interest rates and the primary deficit Here, uh, the key one Important element is to realize that what you see on the right hand side with the letter t is the tax revenues Can always be thought as the product of two elements One is the tax base that we call y and the other is the Implicit tax rate that we call tau. So you can think we can always think that as a tax revenues has been decomposed into these two elements and then So if we just play a bit with this budget constraint, we take We fully differentiated the budget constraint and we iterate forward We impose that debt has to go back to steady state in the in the in the long run at least in the trend the terms Then we have a very simple decomposition that however is very useful. So on the left hand side We have the total fiscal stimulus that we plan to implement between today and the future That is what we need to finance then these, uh, uh, fiscal stimulus can be financed through three Components the one that is the The focus of this paper is the first one that is called tax base because of the fiscal stimulus We may stimulate economic activity and as a result as a self financing part But of course then there is a component that is due to monetary policy that can affect the Real interest rate that is the second term and the last term that is fiscal policy That is the fiscal adjustment through adjusting the tax rate Uh, now, uh, what we are mainly interested in as I said is the tax base Uh, and in fact this paper shuts completely down almost completely down monetary policy Assuming that the real interest rate is kept constant apart from the very initial change In inflation that affects the initial interest rate So let's focus on the tax base But then if we focus on the tax base, we clearly see that the share of self financing is intimately related To the fiscal multiplier the share of the tax The financing share of the tax base over the total is just how the tax rate multiplied by the fiscal multiplier or the cumulative fiscal multiplier all right, so If you want to have a full self financing and abstracting from the initial debt erosion that I don't I mean We can define it as part of self financing or not. I prefer not to I think of inflation as a tax So let's say let's focus on the tax base This implies that to get full self financing We need a multiplier that is bigger than one over tau That of course is bigger than one Now an important aspect That to understand the contribution of the paper is to realize that this multiplier fiscal multiplier is not a number that is constant regardless of economic conditions regardless of a variety Of structural features of the economy But importantly this fiscal multiplier and this is the main point of the paper depends on fiscal policy So depending on how quickly we adjust the tax rate the the last term in the composition If we delay fiscal policy to the future, we are getting a higher fiscal multiplier cumulative fiscal multiplier So well, of course, this logic would hold true in any theory that is able to generate a sufficiently high fiscal multiplier Okay, this is one particular mechanism, but there could be others So let me briefly summarize some theories where we get a high fiscal multiplier Well, one is in real business cycle models as for instance, we see in baxter ranking We have a multiplier much bigger than one for persistent shocks to investment or public investment say or a reduction Well to distortionary taxes Well multipliers as we know are larger than one in new kinesian models It's easier to find the multipliers bigger than one in new kinesian models This could be true for instance, even in the baseline as free equation new kinesian model If we think for instance that monetary policy is sufficiently accommodative If the central banker reduces the real interest rate in response to a fiscal stimulus Then we can get an arbitrarily large fiscal multiplier Okay, and this has a point made by woodford in 2011 or christiano nike and vonna and rebello For instance in the concept of the zero lower bound when it becomes binding the real interest rates fall So therefore multiplier are are much larger A second strand of the literature is the therogenius agent model Where the multiplier could be larger than one because of liquidity constraints or finite lives That is essentially the point of this paper that this could be done in papers with two agents A fully fledged hank models as i'm mentioning in these slides But is in this paper Is about olg or perpetual youth models blanchard and yari And actually this is also a point that is already made By i don't know ranking in scalera in a You know in a italian journal published in an italian journal In 1995 and more recently in a more empirical and quantitative work by a two economist Well former of the bank of spain basso and and rascay and rackety Having said this this paper has a clear contribution relative to the literature Because it's not only shows that the multiplier is could be large But it actually look at the conditions for full self financing and it provides a quantitative assessment So it's a distinct contribution that and this is why I think this paper Is important and I encourage you to read All right So from a practical point of view this paper provides supports for delaying fiscal adjustments and one point I want to make is that Well Fine, but this is not necessary And for instance, I can we can think about two examples where a self find a full self financing Of course without delaying fiscal adjustments the first example And actually I will just focus on one example that is a baseline euknesian model Well, there is a minor finance fiscal stimulus But the same holds true also in a two agents euknesian economy Where even though there is no delay in fiscal adjustment or actually it is enough to have a one period As long as the fiscal deficit is not financed immediately. It's delayed by one period. We could already have a full Self financing as long as there is a high enough share of in two mouth households again So let me go to the first example a money finance fiscal stimulus So this is a very simple three equation euknesian model On the of a non-policy block on the policy block We have the govern magic constraint that I shown you earlier Augmented by a signal that is generated by creating money that is the term in red And then I have a very standard passive fiscal rule Taxes are adjusted to stabilize debt But I have a monetary rule where the quantity of money is adjusted precisely to keep debt constant Okay, so this is an active fiscal rule It doesn't have anything to do with the fiscal theory of the price level But is a sufficiently accommodative monetary policy, okay That as a result would lower the real interest rate So this is called the money finance fiscal fiscal Stimulus and then if you put it, you know If you look at these pictures and you actually focus on the Picture on the very on the very right of the slide You see that even though it is called a money finance fiscal stimulus These these policies generate a large increase in the tax base. That is the blue part And if we look at this picture, we see that the share of self finance of self financing due to the increase in the monetary Base is about 50 percent Or you know 60 percent So this a bit causing two questions, you know, it's a bit of a terminology issue It's called money finance fiscal stimulus because of the accommodation of monetary policy But actually a large part is a self financed according to the terminology introduced by by this paper I made as I said, I we could make the same example Through a similar case with a two agent eukenesian model, but I will skip it in the interest of time Uh The second comment is about the empirical relevance. So The paper shows that if we calibrate the model matching the mpc's and then we have a slow fiscal adjustment the self financing share could even be 95 percent Okay, and that is mostly driven by the tax base in the inflation erosion is not very relevant But then it implies that the model according to the logic that I've been exposing that the fiscal multiplier in in this model must be Bigger than one over tau tau is set to 0.0 0.3 30 percent it means a multiplier larger than three Well in the room in the room, we have many influential Explonents of the literature on the on the fiscal multipliers My view is that the macro estimates of fiscal multiplier are quite smaller According to the survey or Amy is they are around one Of course, there is lots of variation depending on the type of expenditures and the variety of other issues But still having a multiplier bigger than the three is Is not easy to see in reality So two possible reasons why the multipliers are are much lower in reality than in the present model one is Is it is a bit borrowing the logic from the depth literature or the political economy literature? If we delay fiscal adjustment, this may create a credibility problem of fiscal policies If you create a credibility problem, we will raise interest rates and these are higher interest rates may actually Draw have a counteracting effect on the fiscal multiplier Now the way I like to interpret the result of this paper in this sense is that The paper is providing a rationale for smoothing taxation over time rather than for delaying fiscal adjustments If we smooth taxation over time that is shown to be the optimal policy Typically in this political economy lack of commitment literature Then this paper shows that there is an additional reason of why we should smooth taxation that is we may increase Increase the fiscal multiplier I will be reluctant in saying that the best recipe is to delay fiscal adjustment into the future over a certain horizon because of the It may create credibility problems The second is a is a while why the while is a typical argument related to why the multipliers could be much smaller In reality is that in this model the labor market is kept simple for tractability purposes in reality for the fiscal multiplier to be Operative in them in the kinesia says we need slack in the labor market and Arguably the slack in the labor market, you know, if we don't have enough last year's slack in the labor market The multiplier is much smaller than Then these numbers that were proposed and this is there is evidence including my own work But many other papers have pointed these out So that multipliers much smaller than one when when we have when we are not When unemployment is not very high and that is probably relevant for the current debate So I will conclude here nice and thought provoking paper shows conditions under which deficits could be fully financed them fully self financed and provide some support for delay in fiscal adjustments and My comments is that I'm skeptical about, you know, taking literally at least this Implication It might not be necessary to delay fiscal adjustment to reach high level of self financing If we take it literally it may also have negative consequences But as I said, perhaps the key takeaway is that smoothing Smoothing the fiscal adjustment over time is it, you know, as a as an extra kick in our economies Once we take into account these multiplier multiplier effects Needless to say we need more evidence on the effects of the late fiscal adjustments before we We actually take these recommendations seriously into account. Thanks. Thank you very much Yes, I would now open the floor for 10 minutes Questions and remarks so that then you can Rope up at the end and Okay Take everything together. Okay If it's fine with you. Yeah, sure. Okay So Argentina has been trying to do this for the last 10 years and eventually, you know Because I think it's going to the point. What do you think is the? The problem in there is probably what David was saying that the eventually, you know, the credibility destroys the The fiscal multiplier, right Okay. Yeah, so to to speak to that Let me so for a second stick within the confines of the model and how I think I can relate it to that experience So note like one of my one of one of the key assumptions in how I was presenting the results is The ability in my case via the central bank to fix or have not increased too much the real cost of borrowing on the government side Right. That is a crucial throughout maintained assumption that that real rate either remains fixed or in my generalization That does not increase too much Clearly in the case of Argentina either because just in general borrowing is coming from abroad or because of Now outside of the model fiscal credibility related reasons You may think that real cost of borrowing is going to be going up with a fiscal expansion And in that case, I think what's actually much more relevant is what I emphasized in my in my extension at the end Of kind of broader monetary fiscal interactions the converse implication of If I as the monetary authority want to do the Usual thing of a following a strict inflation targeting regime what I need in that kind of environment Is a fiscal authority that I mean we always say it needs to play along in the sense of balancing the budget Balancing the books down the line here in this kind of environment It needs to be it needs to happen quick enough and that's kind of the core economics of what's going on in these kinds Of environments is what you saw in this figure the separation between the short run and the long run These binding liquidity constraints finite lives whatever they they induce and the question is what kind of what's the economics to come out of This if you say I can fix the real cost of borrowing Then one of the economics coming out is this multipliers converging to one over the tax rate abstracting from the price response If you cannot fix this then the other thing is you actually need to adjust the deficit quick enough because otherwise You cannot be controlling the demand side via the monetary authority as you would have liked to Yep Oh and maybe just maybe just on that on that side because then So if I may kind of connect this to reply to one of the things in the in the discussion at the end the the empirical relevance which very much referred to as we also do in the paper to Empirical evidence that's mostly from the us where you think And of maybe there at least for these transitory fiscal expansions that the vl literature looks at It's actually relevant assumption to say that the real cost of borrowing is not responding Responding that much now if you kind of dig deeper into those fiscal policy var's that's actually already what you're seeing So nominal rates don't tend to respond that much inflation also doesn't tend to respond that much So in that in that sense our assumptions there seem to be satisfied Number two you kind of also see in those samples surprises don't respond too much So maybe you're traveling along a flat phillips curve. So then what is it? Why are we not seeing those particularly large multipliers? Well, if you kind of look at the actual along the equilibrium transition path in data fiscal adjustments Following the spending expansions that say Valerie rainy has looked at they are reasonably quickly tax financed There there's some differences in how deficit finance things are like immediately the deficit goes up But then taxes adjust pretty quickly Like what we're saying here is you just have it over like 10 years or something things go down So kind of one provocative take that I could make on the existing the literature is that we haven't quite run that experiment yet Of having a transitory spending expansion and then not having a fiscal spending Like a planned fiscal adjustment within the next five years or so by delaying a little bit further out Yeah To keep with timing Take another question from uh Lero and then maybe you can also Yeah, I mean just just a quick follow up. I mean, but let me first say I think this paper is super Interesting because it is so provocative, right? So much my main question would have been it's humans. It is right I mean that full financing is possible full self financing is possible And the more you delay adjustment the higher the chances are I mean we could essentially forget about the panels this afternoon in a sense, right? So the question is, you know How robust is this and I mean many things have already been addressed so may One entry policy I think somehow needs to be better understood maybe even bringing it to the point what is optimal here, but but More on the technical side assume you have Because of monetary policy, for example leaning against something you have a point where full self financing Full self financing is not any longer possible Then I think it would be interesting really ought to dig to dig deeper on the fiscal side for as for example Assume distortionary taxes and my knowledge of blanchiari models is, you know That then you easily can have situations where you lose a single steady state multiple steady states can emerge You have laffer curves determinacy issues maybe at stake. So All these are things I think all these are aspects that could matter then just to understand better what needs to be done to To get rid of this result so to speak Okay, so just just a couple of quick thoughts in in reply So number one one thing I do want to stress is that and I try to communicate this in my Presentation that this is a is an entirely positive paper saying in a Class of environments that for better or worse recently sent for banks academics and so on have used to think about the effects of fiscal policy There is this kind of time horizon dependence as you're doing as you're delaying You yourself making the adjustments something else will be doing the adjustment and this also then goes back to Davidas point we refer to like self financing as this I guess somewhat in my view For purpose of our paper unfortunate Normative connotation of this necessarily being something good here It's more something else will give along the equilibrium transition path this could be quantities responding and We made clear under the the assumption that which is going to be the case It could also be prices responding and you may very well interpret the recent us experience in particular in those slides If you think there was a large initial fiscal expansion, not a particularly at least initially aggressive monetary response to it And the capacity of the economy was such that the adjustment then didn't happen I only partially happened via quantities, but in a good chunk happened via prices Okay, so you can always get that part and that would kind of then relate to the fdpl literature saying We don't need to enter these theoretical debates of is there infinite delay in fiscal adjustment If it's already just a little bit delayed you're already going to get Things happening via inflation, which is kind of normatively undesirable If you think that's how the adjustment is going to happen and not not via quantities Okay, so that that kind of just on a on a first part Second part distortion taxation. I kind of briefly got into that That's not going to change a huge amount at the end of the day will somewhat change the split between prices and quantities Not going to be doing much more than that. We can discuss and be a little bit more afterwards The reason I'm actually not particularly deep for why that's the case Final thing I forgot I wanted to reply to one more thing Sorry, yeah blanking on the third part of my reply. I hope that partially got at least to some of the things you had in mind Yeah, okay Oh, yeah So just kind of on the the kinds of environments that allow for this And this may be also kind of relating a bit to your question of how we can break things And I want to relate to david is so that david is a little bit about what kind of other environments can generate this kind of Self-financing as as outcomes. So number one. He mentioned behavioral stuff myopia at the beginning there I want to be clear that is actually not quite sufficient. It gives you half of what we need It gives you no response at the beginning to the future far delayed tax cycle But remember I needed two things in my figure I needed the discounting of the future tax cycle But I also needed the money you get at the beginning to be spent quickly enough in the short run Behavioral on its own is not quite enough. It only gives you the first one doesn't give you the second one And then on the other extensions, but certainly yeah, the the money supply moving that the senior rich stuff very much So what we wanted to communicate is in our kind of environment even without any monetary accommodation the strict sense of fixed real rates There is this kind of time dependence Delay in fiscal financing giving you that as interacted by the with the discounting on the household side So that's just theoretical distinction and final point because I know a bunch of people are working on this class of models So davida didn't get to talk about the the tank stuff But there is a very important in this case your conceptual distinction between what hank or olg and tank give you So in tank the self financing and the large multipliers there with a fixed real rate That can happen if and only if So for those who are in the literature Expenders and savers are differentially exposed to the fluctuations in income say because one of them gets dividends payments The others don't if you have uniform incidents of everyone This result doesn't get off the ground in a tank model cumulative multipliers are always one for fiscal spending or zero for transfers You need this kind of hank or olg stuff to give you this timing logic that I sketched in my key figure So that's actually an important in my view qualitative difference between hank type environments and tank environments Yep Thank you. Thank you very much and we now move from to another very topical issue This cathedralism in modern unions, okay, I already introduced you I'm seeking because from from home. They cannot see That's right. Yeah, no problems From what? Okay, and the clicker Yeah, I have a lot of poses Okay, uh So first of all, thanks to the organizers for putting the paper in such a nice conference So what we do in this paper is um To us how should policy choices be delegated between central and fiscal authorities And related to that some argue that we should know over which matters several local tribunals Are to have jurisdiction And in which authority should be centralized and as you see this goes back to Aristotle So this to show that this question has been around for a very very long time But still we have no subtle answer For example, tabellini argues in the context of the european union That one of the most pressing Questions is what task should the EU have and which ones should be left to the member states This answer is also important Uh for many countries not just the european union Some work has argued that complex rules for fiscal federalism in some latin american countries are responsible for much of their full performance in policies So what are the answers that we have so far in the literature? We have on one branch an answer from a small macro literature on monetary unions That says that centralized authority with fiscal decision-making power is always weakly better Why is so well the assumption the assumption on this macro literature Is that absent any externalities central and local authorities are equally good so they can equally Tailored the policies in uh in these countries This implies that Even if we have only tiny externalities because centralized fiscal authority Internalizes this externality Then the conclusion would be that centralization is better The idea in this macro literature is that Suppose that we are having a a country in a union and this country increases its nominal debt That then it induces the monetary authority To a increase inflation so as to reduce the the real uh value of such debt Then if we're thinking of a decentralized fiscal authority this Uh decentralized authority Does not take into account the cost of inflation on others But the centralized fiscal authority will take into account these uh externalities Therefore it will spend less and therefore we will have lower inflation So basically the main takeaway from this macro literature is that there's no benefit to uh decentralized fiscal authority But on the contrary we also have a Large kind of large micro literature on fiscal federalism and the answer in that case Is that the local authority is better unless fiscal externalities are huge Why is that well in contrast to the macro literature the presumption in this literature is that Ups and externality local fiscal authorities decentralized fiscal authorities are much better So given that ups and externalities, this is the case. We would need substantial Externalities before centralization. It's better The idea of why local authorities are better Is that they are better at tailoring the policies to its local citizens We can think that they know better the preferences of the citizens so they can tailor this fiscal policies better and this goes back to the seminal work of odds So basically the main takeaway from this literature is that We would need we would have large benefits To decentralized authority because of this better knowledge of the thesis So what's our approach in in this paper? We want to isolate the circumstances under which centralization is better to this decentralization Are we contrast precisely these two forces? We have the informational benefit of the decentralization in the spirit of this fiscal federalism literature And we also have the externality benefit Of centralization in the spirit of the macro literature Just to say a few more words about these these two contrasting forces This will lead to a benefit of decentralization in the sense that the central authority Can only observe a noisy signal of the local preferences Now a natural question given this assumption would be why can the central authority easily listed its locality Tastes this just by simple by simple mechanism The most simple of all would be just asking these local authorities. What are the preferences of of their cities and right? but uh Here what we an example that we always put and we like especially in this in this time of the year is that Think of giving presents or receiving presents for the holidays even if Both the giver and the receiver are well-intentioned and care about the welfare of everybody There's evidence actually empirical evidence that there's a Value losses in this gift giving That it's up to 10 of their value in the case of a partner. So we can think of the partner as like a Local authority, but these losses go up to one third of their value If it's the odd that is giving you the present So we can think of this odd this analogy as the central authority that does not know very well the the preferences of of your of yourself So that's the that's the example that we put of why we think that this assumption is is it's reasonable So that would be the benefit of decentralization now for the benefit of centralization so that we obtain this trade-off We we have that the central fiscal authority Will internalize the inflationary cost of that so we will see this in a second just a brief overview of uh of what We do and another view of the results We have a companion paper that deals with a real version of the model So that just boils down to the initial Oats decentralization theorem such that apps and any externality local authorities do strictly better So this gives rise to a cut-off rule that if there are some class of fiscal externalities Central centralized authorities are better only if and only if the union is sufficiently large What we do in this paper is we take the the nominal sorry the monetary model with nominal debt as in the macro literature We obtain a generalized decentralization theorem such that If the monetary authority has commitment We will see what we mean by that in a second then local authorities are strictly better And we also obtain a cut-off rule in the sense that if the monetary authority does not have commitment to its policies then Centralized authorities Are better only if the union is sufficiently large So this is important that it contrasts with with the results in in existing macro literature And also I want to to point out that this has important implications if we think For example of the enlargement of the european union because it's all going to depend on how big The the the size of the union is So let's just jump in the model. So I present this very simple two period monetary union. I just want to mention that we are currently working on an infinitely version of the time version of the of the model So that we can get more dynamics of the model there's going to be high different regions or countries And they have a representative consumer and a local fiscal authority Now there's also a union wide central fiscal authority That goes about the welfare of the entire union. This is what we can call a fiscal union And fiscal authorities either if they are locker or central are in are in charge of choosing the level of nominal debt The timing is as follows in period one Governments finance some government spending with the nominal debt that they choose and in period two Governments have to repay the real value of their nominal debt and they have to repay it with distortionary labor taxes So it's a monetary union. Uh, sorry the monetary authority. They wanted church of uh, choosing inflation What's the trade off for the monetary authority? Well, there are going to be some cost of inflation that are reflected in the model as a decreasing productivity with inflation And benefits of inflation in the case of Commitment these benefits are not are none in the case of no commitment. We're going to have that exposed inflation Erodes the real value of this nominal debt so that what the governments have to pay It's less in terms of in real terms these lowers the distortions from from taxation. So that's the benefit of inflation The firm problem Uh, it's going to be useful to see precisely the inflation cost in the model So firms in each country have a given amount of money M. This is exogenous Uh parameter and they buy inputs x With this money These inputs are going to use to be used to enhance productivity So the firm problem is just Uh to Well, they have a linear technology in labor Multiplied by this a function of the of the inputs that the firm uses Subject to a sort of a cash in advance constraint So they have to use this money that they have to buy the inputs at a given Price and here it's where we see precisely the cost of inflation, right? Because if there's an increase in inflation the price of these inputs increases So given the exogenous amount of money that the Unconstant amount of money that the firms have the amount of inputs can be reduced And therefore these effects negatively the productivity. So this is the cost of inflation to see the benefits of Inflation it's useful to look at the government budget constraint There are two periods. So we have two government budget constraints In period one the government has to finance this level of Public goods g And to finance this amount of of public Good it uses some claim to bi dollars In period two to be repaid in period two at a given price price one over one plaza So that's the the first period government budget constraint Is to finance this amount of public debt. Sorry a public good with nominal debt In period two it has to repay this debt and it does so by issuing taxes So in the second period a government budget constraint We can see what is the source of benefits to inflation an increase in inflation reduces the real value of that So it reduces the taxes that it has to Levy to to pay for this for this real value of that And the consumer's problem to close this economy have Two ways of saving they can save using nominal debt From the government or they can store real assets k with some technology that has a fixed real rate of return In period one consumers receive some endowment Consume and save and get utility from the public goods That the government provides In period two they supply labor to the firms They consume and they get the returns on their savings and on labor in the form of of the competitive wages So basically the consumer problem is what you see in the slide In period one they get utility of consuming and of the from the from the public good And in period two they get utility from consumption in period two and these utility from supplying labor Now an important term in this equation is the theta parameter multiplying the utility over public goods For simplicity we assume in in the in this model that This parameter can only take two values So either a country has a high preferences For public good or a low preference for public goods And this is important because this is a Precisely what affects the the the benefit of decentralization We assume that the local authority perfectly observes The these preferences of the country so the local fiscal authority knows Theta, however the central authority observes only the noisy signal over these preferences this is this s parameter In the slide With a given informativeness level informativeness level fee that ranges between one half and one Intuitively if fee is equal to one then we would have that the signal. It's perfectly informative So there would be no difference between the local central authority and the and the Central fiscal authority But if the fee is equal to one half then the signal Here I have to point out a type in the slide the signal is not informative at all So that the expected value of the of the theta for a given signal is just the The the average of the theta Okay, and of course for general fee we can use base rule to to to compute the expectation over the theta so this goes back to the to the To the interpretation that Local knows better about the preferences of their citizens just to what this is to Point out the consumer's problem solution. I just want to point out here that we can write the Labor supplied by the the curriculum labor supplied by the consumer as a function of taxes or Equally tax rate or equivalently as a tax revenue of the government and the productivity in the country okay, so we have now Gone over the agents on this On this economy now we are left with the monetary authority that remember it's the one setting the inflation of the union First consider the case that the monetary authority has commitment. This means that in terms of the timing First the monetary authority sets inflation that it's going to prevail in the second period And the it sets such inflation before any information or signal is realized Then all other agents move taking this inflation as given So because it has commitment and because of the lack of arbitrage between nominal debt and Real assets that the consumers used to save that we have the the the fishery equation Implies that inflation has no effect on the real return on nominal bonds So basically this is saying that in this economy where the monetary authority has commitment There would be still the cost of inflation because inflation decreases productivity But there are no benefits of increasing inflation So the basic result of this Simple model is that the optimal inflation rate is then Zero because of this cost And the decentralization theorem Immediately applies in that if signals are not perfectly informative Then locals always do better Therefore a decentralized regime will yield a higher Exante welfare than the fiscal union And the differences in the welfare between central and and and Decentralized regime increases as the information problem becomes worse And this it's again in contrast with the results in the standard macro literature But let's jump to the more interesting case that it's what happens if the monetary authority does not have commitment This means in terms of timing that the monetary authority will move in the second period After all the nominal debt decisions and the real savings decisions have been made So in period one preferences and signals are Are realized Consumers and governments choose their spending their nominal debt and savings And it's in period two when the monetary authority sees what's the this vector of Nominal debt for each country and this vector of real savings in each country And then it chooses inflation based on this state of the economy Given this inflation, then the government has to choose taxes or labor to pay for the real debt And consumers choose how much labor to to supply Now in contrast to the previous case here, we have Costs of inflations are going to be the same. There's this decreasing productivity But now the benefit of inflation comes from the real value of nominal debt decreasing with inflation And hence the distortions from Taxation and labor will also decrease Okay, so just to give the intuition of why The model delivers this indirect fiscal externality Suppose that utility is additively separable as we have shown before So that the optimal choice of inflation by the monetary authority Only depends on the State that it receives the nominal debt that it sees that all countries has issued And the real savings of all the agents in this economy. So basically it means that Monetary authority at the start of period two only cares about the utility in period two It does not care anymore about the preference for public goods of a country being high or being low because that's already gone in the first period So the monetary authority will maximize Equally weighted sum of the utilities in period two of each country in the union And this is going to generate this indirect fiscal externality in a decentralized regime Because basically the government of country i when it chooses its amount of nominal debt Only takes into account that by increasing its amount of nominal debt It's increasing inflation But the cost of this Inflation increase it only takes into account its own cost not the cost in the rest of the economy so that's precisely the point of the paper and Where the externality is coming from key to this externality is What the fiscal authority Anticipates that the monetary authority will do in the two different types of regime the centralized and the and in the decentralized So just for simplicity we consider here a linear utility function Then the problem of the monetary authority. It's just That equation that that we see in the slide so it's basically summing over the utility that it's now the production plus the The real savings returns minus the the utility of working So just to simplify matters we are going to use this notation that the f it's the part that it encodes the benefits and cost Of inflation for any given level of debt so basically if we assume that Preferences are perfectly correlated across countries assumption that we will relax later Then all countries are identical the centralized regime chooses a symmetric amount of debt for all countries and hence the fiscal authority this central fiscal authority anticipates that the monetary authority Optimal choice of inflation is just given by this first order condition at the bottom of the slide And this contrasts with the decentralized regime again because the decentralized regime is choosing inflation for country i It's taking what everybody else is doing as As a symmetric outcome. So they think that everybody else is doing b minus i Then the monetary authority faces an almost symmetric history And the first order condition changes with respect to to the one that we see in the centralized case So just to make this more clear When Central fiscal authorities choose its nominal debt It has to take into account if it increases its nominal debt how much inflation it's going to change So that's why it's useful to to look at the elasticity in this case If the the changing inflation that the central regime anticipates is given by this red formula However, the decentralized regime only internalizes if they change in inflation affecting its own country i So by this you can see in the green part contrast contrasting it with the red part That it only anticipates a change in inflation that it's a fraction one over i Where i is the number of countries in the union of what the central fiscal authority takes into account because it internalizes all the The effects in all the countries So this is the same But written in terms of elasticity the elasticity the debt elasticity of inflation in the decentralized regime It's a fraction one over i of the debt elasticity to inflation in a central regime And this is precisely capturing this fiscal indirect fiscal externality in this model Okay, let me skip this for the sake of time And let me just say What we find in this model So we find that for a class of distortions from inflation Embedded in this function the productivity function a We obtain a cut-off rule for optimal delegation such that for any given degree of the informativeness Of the preferences of of the countries Either there exists a finite cut-off in the number of countries i of fee such that a fiscal union is preferred If the number of countries in the union is bigger than this cut-off Or we can also end up in a result in which a decentralized regime is preferred for all i This would be the case if the if the information problem. It's bad enough So we have here three cases on the left panel We have that so this is plotting sorry the ex ante welfare Under a centralized regime the solid line and decentralized regime the dash line as a function of the number of countries in the union The centralized ex ante welfare It's constant because as we've seen the inflation change with uh with debt does not change with the number of countries But the externality problem gets worse As the number of countries increases right so because each country internalizes like a smaller part of the of the effect of inflation so In the in the left panel, we have that the centralized regime is basically always prefer unless The the the absurd case that the monetary union comprises only one country In the middle in the middle panel. We have that the decentralized is always prefer To a centralized regime and in the right panel It's like the most interesting case That we find a cut of it of the number of countries such that if the union is small enough Then this externality problem is not such a big issue and because of the informational benefits of local authorities A decentralized regime would be preferred But if the union is big enough, then the externality problem becomes Bigger and at the centralized regime would be prefer Moreover, we find also that the cutoff will decrease As it is intuitive with the informativeness of the signal. So I think I conclude or Um, yeah I mean I just to point out that I've shown you the case with perfectly correlated Preferences meaning that all the members of the union have the same preferences. This is obviously um Not the best case. So we also play with what happened with independent preferences Uh, algebra becomes a bit more complicated because you have then to take into account the combinatorics of all possibilities In the union, but the results still hold with more difficult algebra So we obtain the cutoff rules a gentleman the centralized Authority it's better for a large enough union And I think I sorry I will conclude now. So we have shown uh, how insights from the fiscal federalism literature Uh Changes the principle of delegation from the existing macro literature in that Optimal delegation does not just depend on whether externalities exist or not But there's this trade-off between externalities fiscal externalities and the natural advantage of local authorities um So there's this no one size fits all rule applies to delegation It's important to point out that this has implications for the design of a monetary union And for the analysis of enlargement policies So, yeah conclude here Thank you. Thank you, Eugenia for this very interesting presentation Um, I give now the floor to Ramon Ramon. You arrived a bit later. So I remind you that you have 15 minutes at this position Thanks Thanks for inviting me and thanks for having the chance to discuss this paper, which I think is very timing for the topic of the conference Do I need So there are basically two dimensions to fiscal federalism And which of course are not orthogonal one is the subsidiarity principle which is A basic principle for example in the design of european union And the other is the this problem of federal internal externalities, which is a fact So the question is I think it before I was going to tell my students that was my who should fix my bicycle Now it changed to who should fix my battery And in the old times That's what it was. You you got your cables and that's it and when I first arrived to chicago in the Call days people will take in the battery back at home at night. Now things have changed At one point I was doing this thing and made a small mistake with the car of my wife and it was very expensive So then I decided to move to have the local Mechanics doing it. That's a good mechanism because there's some competition and also some friendship but Things are getting differently now. So I have this problem with a new car and I took it the local mechanic and maybe he did the mistake Had to go to the central honda in florans, which is Not very close to home and it turns now. It's taking care by the central honda. I know japan or something but So here where we are This problem is that how are we going to say always as economies that Who should do it is who is more efficient? But as I said technologies change and that answer can be different in different solutions The focus has been in the literature as the hennie was pointing out particularly in fiscal failings about This idea that locals know better actually should be in 72 of 99 And There is another side of it which is to say well Better commitment and this is a long literature and about thinking for example the european union that the local monetary Policies were not so much Consistent and instead here this cb is much more time consistent and therefore that's how it goes but here then you then have this So-called problem of internal external Externalities which is a simply fact that if you have a joint good with some characteristics of public good The free rider problem becomes worse and worse as you get larger So those things are well understood So here's like I will call it two fold theorems One is this generalized decentralized theorem which says under subsidiarity principle You got the best thing is the decentralization they call it generalized because in Times the idea was that the centralized policies always have to be uniform They do a little better than that and they say no you can even have target, but maybe they are not well target and So that's a theorem no if the locals are better informed on citizens needs and aspirations Physical policies should be run by local governments and the welfare gains will be larger the more Informa Relative information they have is better Okay, and here the key element of the proof of that is just this thing playing with them for how informative The signal is and that one gives you everything the second theorem is that is the Free rider law large numbers theorem, which is a perversive theorem Which is as the number of locals increase the free rider problems aggravates And here the key proof of the of this theorem is about elasticities showing that The elasticity of the local action with respect to the center gets more and more inelastic So there is a corollary of course like these two is that then if you have two things coming Then if they coexist then should be some place on another a cutoff point On the level of multisibalities, maybe two So this is what the paper makes the point. It was basically the first part which now seems to be another paper Is more of a real economy Then without externality theorem one applies and with the fiscal externality Theorem two and the corollary apply and there is a little bit cooler externality because You like the goods Of every one. Okay, that of all the other municipalities or countries or whatever So as more countries they have the better you are Well, I am in Florence in the center in the Florence of Barcelona sometimes now it seems the other way around but anyway They make the point and the the interesting thing is that then this point is helpful To have the model, which is the monetary model, which I think is the contribution Where we just have seen It's clear if you and what you have seen now is you you have this timing From the perspective if you're a local government And you have here the pie in red You're clearly you don't want much of the pie for the consumer side And certainly if you look at the expenditures that the g and you look at the the nominal rate between the real and inflation At the end of story by raising inflation, you don't gain anything from that perspective and that's true If it is a single local government, it's also true if there is the whole central So it's all about timing commitment and externalities And ways we have seen So inflation cannot be below one So therefore That's question is if it's decided first before all the decisions and it's not changed So which means you have full commitment is obvious that pie is equal to one Either is yours decentralized or decentralized and then of course, there is a problem of information third and one applies However, if we reverse the timing then And pies decided that after all the decisions are made about savings and there was of that Then they are very close to be in a situation of limited commitment And then we have the externality And and as we have seen exactly The proof relies on this fact that if I look at the elasticity Of what of how much increased the debt over respect to the inflation And you would think about the symmetric case, but it's generalizes to the non-symmetric case then When you are decentralized Okay, is one over i the elasticity of the central So as simple as that as i goes to infinity, this elasticity just goes down to zero So then you have momentum two and the color and of course in general you're going to have that developers is worse under decentralization so The question is who is the center? Well, maybe so We want democracy, but Then there is a problem with democracy, which is that problem killed even more perverse When we have a strategic delegation So if I can change slightly the model here and put that Now you're working period one and then in period two there is some uncertainty And to make it simpler So we have different states and every there is a bad state only One of the countries or locals, okay at the time So then We're gonna have a version of this where it's a theorem two and in the corollary then Why because there is a world of imperfect resharing If you have in this model if you were on your own then you will use inflation exactly to adapt to this shock But when you are in the union, you're not gonna do that But most of the time and for everyone is not having anything So you should put it at zero But well at one by at one basically, but insist you will insist to the one period that this affects you And therefore what will happen? Who will you delegate to this central bank with representatives? You delegate the guy who takes the bacon home The most radical of possible guy even if we are all moderate So we have an obscure and representative paper which are in john's and it's obscure because I realize it only has one Citation and that means he represented it because most of the papers they get none of almost one. So But I think the first one has more Or to do So what have been the solutions to that? Okay, by this thing of bringing the bacon how he comes from the fact that Workers had this possibility to take some bacon from free and that was important. So you will take the guy with the larger hand there to get it so How is it solving in us? For example, otherwise you're gonna send the guy who is the To the congress who wants to take more staff home But then to counterbalance this maybe someone in the central government that which order That was in the previous century now the us is who knows beyond our theorems, okay But and then and then of course then you also have other things, okay They realize that they have to have budget constraints at the local level or nobody allowed and so on In monetary unions just as we have central bank independence with pleasure stability mandate On the fiscal side, we're going to have then and now fiscal compact and stability and growth impacts Which seems to go back towards with proposing uniform policies, okay, so The question is is you? locals getting larger Well, yes, we had the granada declaration. We're going to go to 35. Okay, so we're going to have more members and we are moving fast But beyond the number of members there are other ways that we are getting larger Why because now we're gonna have When we have some populist that formerly were all the skeptics, but now They upon themselves or they upon them to bring them home the bacon the other day Meloni said this guy draggy Oh, I did they took pictures with McCron and all these guys, but he did nothing for Italy Okay And then but also at the local level we might have more nationalistic policies or claims Which are along the same lines So we can think of our new european parliament of 60 official languages We're going to have catlin and and basking at least in the soon And I don't think it's a problem If you have artificial intelligence, it only requires three thousand six hundred per wise simultaneous translations But it's a problem if then we have this question of strategic delegating The most nationalistic guys to bring the bacon home Furthermore, the list of public goods is getting larger and they have this issue like We should solve the problem of my card getting those most of the problems You can do very little locally And actually coming here realize how fast europe is expanding From home florence. I live in florence and marcellona to here. He has taken me 23 hours Not with my garbage. So you don't have it but with Lufthansa amazing. I mean before it was much shorter much smaller. I don't know so Can you bet do better? Well, I think we do we can We are not just a two-period You neither the central bank is So I think we should have an european stability fund Where we can have Also long-term state contingent kind of specific fiscal contracts, which are specific to the countries It's will allow that you don't have permanent transfers They can provide it's a way to get resharing and proper contraceptive legal fiscal policies And you don't need to interfere much with the net markets And so that's too much self-referential But those are papers to explain about these things. The other was talking about these things here But the issue is That You might have some need of independence Because I think it's a good idea that we spend and maybe the commission becomes an executive The committee or whatever But you need to have some independence before The same way that you needed to put the power from more political interference interferences so Just remember however that the center may have its own problems. I'm gonna get the car until next year. I think Okay, thank you Thank you very much. So open maybe I can't resist asking a question myself. I wonder What you are just saying How this conceptual scheme and your model Could be adjusted to reflect the fiscal design in the monetary union Going a bit in the direction of you know next generation You where you have neither a decentralized nor A centralized fiscal regime, but rather a different type of interaction Between the central and the local authorities whereby The central authority You know besides on the funding strategy and the prioritization of the public goods to be To be attained in this case for instance the green digital transition or Social and economic resilience Whereas the local authorities keep the ability to exploit the natural advantage of Being better able to adapt policies And to local characteristics and and preferences So under this scheme basically the the magnitude of the fiscal externalities Across countries and the quality of the information The central and the local level about The citizens preferences Do not determine a threshold Between the centralized and centralized regimes, but rather An optimal trade-off, you know between central And local decisions while keeping a fundamental continuity among them Yeah, then Do you prefer to how do you prefer I can answer after no as you as you like. Okay, then we see Leo I mean just a quick comment, you know on What is meant here by decentralized and centralized? Maybe, you know the conscious Is Comes out here as being too stark in the chari kyu model to the extent I remember and this is close to what you do All that is needed to address the externalities is to have agreement on a fiscal rule Which restrains borrowing behavior if you design this in a clever way the externality is gone So there's no need to go for a full-fledged fiscal union or something like this below this rule Countries are free to choose instruments taxes spending as long as this particular rule is Is is respected. So Yes, so this is Another variation on Ramon marimons who is who is the center comment So so the model assumes that the central authority Uh maximizes the welfare of citizens of the union therefore internalizes externalities But if I look at how decisions are made at the eu Uh, they're usually made by majority rule or qualified majority or unanimity sometimes So I'd like in this little church to see some kind of analysis of how these Voting arrangements or actual decision-making arrangements at the eu map into this idea of internalizing externalities. I'm not sure that in fact externalities are internalized To a sufficient extent with the the setup we have now Okay, so going back Going back to your question. Um We started actually thinking of an economy in which Uh There are two types of public goods. So you can think of parks and uh tanks No kind of goods and uh some economies Some countries prefer to to spend more on one and some countries in others So I think I was thinking of this related to your question because I think one possible way to address What you were referring to is to have like these two kinds of public goods in which in one It is very important the information About the local preferences and in the other one, maybe the central authority has a better management of it So I think that's a way that we can address this issue Um I presume that similar like kind of results in terms of this kind of result would apply maybe with more like Yeah, non-linear it is in the in the type of cutoff um Then um Ah related to the fiscal rule in chari kicho. I completely agree. So if if you put like some fiscal rule, for example in the level on the A cap on the level of of the nominal debt, uh, you can address these These distortions from from the externalities, but now the question would be that in order to set I I guess in order to set this Uh fiscal rule the central authority should know something about the preferences of the countries again So we still have this informational problem. So so How high this cap is to be it's still like a question that would Create some distortion between the central and the local authority And regarding the the last question that yeah, I think it's it's it's very interesting like to have like this voting arrangement The one way that we could deal with this is that so far we assume like some Monetary authority or the essential sorry the central fiscal authority having equal weights across across countries Maybe we can play around with these weights such that we can Deal with some sort of of voting decisions Who has more power for example in the EU or things like that And uh, yeah related to to Ramon discussion. Um, I find very interesting the this strategic delegation issue and It's something that we have to explore more um But uh, so this paper for sure it extracts from all the wish sharing issues actually For simplicity, we just assume that all the countries are the same in all the ways Of course, there are no shocks. So we abstract completely from this This very large wish sharing literature Related to what how the EU is changing and whether it's getting bigger or not something that we are working on is Again, we're assuming this equally weighted kind of preferences for the for the Authorities, but of course, uh, even if we have this equally weighted um assumption It depends a lot on the size of the countries and in particular the uh, how fast they are growing For the countries that are Getting into the european union. No, so usually the countries that are Being added to the european union are probably more fast growing So this implies that there's going to be a faster increase in nominal debt because they need to do resources to grow And this can create even more distortion. So these are things that we can take into consideration and and we are planning We're planning to do Uh, yeah, and I think I'm not forgetting anything