Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=

# Monetary-Fiscal Interactions: HANK vs RANK

Authors:   
Discussant: None  
Video: https://www.youtube.com/watch?v=5Oaz55JP2tk&t=254s


## Talk (00:04:14 – 01:05:32)

[00:04:14] [applause] Thanks a lot for the introduction.  
[00:04:21] I can only mess it up after that. So maybe I should step down. I will disappoint you. So uh we had a very interesting panel that was also a bit depressing about reality. So I will  
[00:04:34] bring you back to the comfort zone of uh theoretical models where things can be a little bit more less gloomy. Okay. So what I'm going to be talking is about physical matter interaction which was  
[00:04:48] the topic of many of uh the papers we saw today. uh uh I will be basing this on uh joint work with Chen Lean from Berkeley and Christian Wolf uh uh uh  
[00:04:59] from MIT and uh um sort of two two disclaimers before I jump into it. one is I'll try to to give you both a bit of a  
[00:05:12] methodological message and an applied message and maybe it's bad to try to do message two messages but I would ask you do that so I'll try to deliver the methodological method is going to be to clarify what we mean by fiscal dominance  
[00:05:26] in models and maybe that helps us understand reality okay and fiscal dominance can be very different in a sense Wallace model with flex enterprises can mean something very different than Ukian model and I  
[00:05:40] want to clarify a bit that the applied message will be about uh if you want a form of fiscal irresponsibility or fiscal inaction slow adjustment in  
[00:05:55] fiscal policy over the business cycle why it could be good. So that may be an antilimatic topic message to try to convey uh uh when we're thinking about you know what's going on in reality but I want to separate I'm going to be talking about  
[00:06:10] the business cycle not about trends and steady state my focus is on the business cycle okay so okay I'm not sure I seem to be ah here  
[00:06:23] it goes okay I want you to focus on the following two questions now to start become more complete they are very related somewhat different. The first question is how much deficits contribute to inflation and through which  
[00:06:37] mechanism. The second is what kind of fiscal framework the central bank prefers should prefer a contra model what kind of fiscal policy aids or  
[00:06:50] interferes with the central bank's dual mandate to stabilize output and inflation. Should the central bank prefer something like an active fiscal policy or a fiscal policy that adjusts taxes and government spending fast  
[00:07:04] enough or a fiscal policy that fails to do that? Okay, these are the two questions that I want to address [snorts] and we can address these questions in different kinds of models.  
[00:07:17] can address them in flexible models where real economic activity is independent of monetary policy such as such in the wall or we can address them within the new ken framework. So I'm going to commit on the new kunen  
[00:07:31] framework which means I'm committing about shortrun dynamics. We're not talking about the long running here, but within the new Ken framework, there are still I don't know why I'm also facing trouble  
[00:07:44] with this. Okay, there are still two approaches within the new firmware to address these questions. wants to study fiscal  
[00:07:58] monetary interactions within the representative agent UK model and to argue that within that model there is in a regime of fiscal dominance and in that regime of fiscal dominance that's how  
[00:08:10] fiscal policy matters. So what I'm going to argue is that in this and that way of thinking about fiscal matter interactions is delicate, fragile and in a way that I will makes clear  
[00:08:24] nonsensible. So I will try to move you away from this traditional approach of thinking about fiscal mant interactions and I will try to push you still within the new framework but in the direction  
[00:08:36] of heterogeneous changes models and by heterogenous newen models. My emphasis is going to be not so much on the heterogenity which is important but primarily on the fact that because of  
[00:08:51] incomplete markets liquidity constraints and other frictions households are non-recardian. We're going to have nonrecardian effects on consumption behavior. That's a mechanism non-recardian effects on consumption behavior  
[00:09:04] evidence on stimulus checks and so on. We have ample empirical evidence on nonreal effects on consumption behavior.  
[00:09:13] So I want to understand how fiscal policy affects a demand output and inflation through that mechanism and how that gives you a different way to think about fiscal and monetary interactions and how it gives you different answers  
[00:09:26] to those two questions. So first I think the technology is not helping me. uh uh first as I said I'm going to revisit rank with fiscal dominance and I want to remind you a little bit the focus on the question you  
[00:09:41] know what kind of fiscal framework aids the central bank the textbook answer you'll find in Woodford and Gali and others is the following that the central bank prefers a fiscal policy  
[00:09:53] that adjust taxes fast enough so that fiscal policy can uh uh uh make sure of debt sustainability and then The monetary authority is free to regulate output and inflation. Conversely, slow  
[00:10:08] fiscal adjustment is bad. Why? Because it opens the door to fiscal dominance in which case the monetary authority loses control of output inflation. The latter are instead driven by fiscal deficits.  
[00:10:22] So that's the conventional way to think about fiscal monetary interaction. Slow fiscal adjustment is bad. The central bank loses control of output and inflation.  
[00:10:34] I will argue that this way of thinking about the problem is subject to an obvious and uh uh discomforting paradox. What is the paradox? Within the new case model,  
[00:10:46] inflation is pinned down by real output and real spending. So for fiscal deficits to drive inflation, it has to be that fiscal deficits drive real spending, which means Ricardian  
[00:11:01] equivalence has to fail. But you're in a starting model where households are Ricardian and you're requiring Ricardian equivalence to fail. There's something paradoxical about it and I  
[00:11:14] will explain that paradox by saying that within that conventional approach fiscal mo dominance amounts to a purely self-fulfilling belief a self-fulfilling prophecy that's exceedingly fragile in a  
[00:11:27] way that I will explain and once you refine away this pathology the following is true within the standard new kenan model fiscal policy is relevant period  
[00:11:40] monetary policy is dominant even if the tailor principle fails. And finally, the transitional approach to physical mandate interactions is out. Now, if you  
[00:11:53] don't like those conclusions, you need to move away from this model.  
[00:12:00] How we're going to move away from this model? There are many ways to move away.  
[00:12:03] I will move in a specific way. As I said, the way I will try to move >> [laughter] [panting] >> I don't know if I'm doing something  
[00:12:18] wrong. >> The lower part.  
[00:12:25] >> No, that's okay. I I'll give it I'll keep giving it a try. But it's this right.  
[00:12:37] >> This left and right. Yes. Okay. It's not just me.  
[00:12:44] >> No, it's not you. >> Okay.  
[00:12:47] >> It's not working. So, say next slide. My colleague.  
[00:12:50] >> Okay. Okay. [snorts] Uh, so >> I check the battery.  
[00:12:55] >> Okay. Perfect. Thank you. Okay. So, I will move away from the standard representative agent newen model in the direction of dropping the assumption of a representative agent and allowing households to being nonrecardian. So the mechanism is going to be different.  
[00:13:09] Households are nonrecardant and that naturally allows deficits to uh stimulate aggra demand and thereby output inflation.  
[00:13:20] Basic intuition old fashioned logic if you push the tax burden to future generations or if you relax borrowing concerns for households that will stimulate spending output and the like.  
[00:13:32] Okay, that's a mechanism that as I will explain is theoretically more robust and it's also as we'll discuss supported by empirical microvari and then what are the lessons that uh uh come out from this approach that relates to two papers  
[00:13:46] I've written with my co-authors the first lesson is that through this different mechanism fiscal def can be quite inflationary and in a sense that we make precise in our first paper is  
[00:14:00] that despite the mechanism has been very different than in the fiscal theory of price level or than fiscal dividend. You may end up getting predictions that remind you fiscal dominance that reminds you of the fiscal theory of price level.  
[00:14:14] So highly inflational deficits even in situations that monetary policy is dominant. So it's no more about which authority is dominant. Maybe we should abandon that language. It's more about  
[00:14:27] both policies naturally drive demand [snorts] and then it's a question of how much accommodative monetary policies to fiscal policy or not but it's not about equivalent selection. Okay. And that  
[00:14:40] mechan is going to be disciplined by empirical evidence. The second lesson that I will try to focus in today's talk is going to be go to the second question that I raised before. Is slow fiscal  
[00:14:54] adjustment good or bad? And I will argue that perhaps contrary to conventional wisdom that the settlement bank should welcome slow or even no fiscal adjustment over the business cycle over the business cycle for two reasons. One  
[00:15:08] is that backloading fiscal adjustment is going to help stabilize the economy against a demand socks. It will sort of amplify and give you a dynamic version of the automatic stabilizer that will  
[00:15:23] help you stabilize the economy and will ease the job of the central bank. And second, in the case of supply shocks, what will be happening is as I will explain is that such slow fiscal adjustment will help endogulously  
[00:15:36] minimize tax distortions, minimize tax wages which will improve is the inflation output tradeoff in the case of Cospo.  
[00:15:46] So my bottom line applied message is going to be again what may look like fiscal responsibility not adjusting taxes fast enough following the  
[00:15:58] recessions that is a good thing. Yeah. So road map I will uh start with a framework that is going to be rich enough to nest both the no the  
[00:16:11] representative a newen model and a simplified version of heterogen and newen models the essence as I said that we'll pick from heterogen models is going to be the presence of non-recadian  
[00:16:25] consumption whatever I'm going to tell you in the simple model extends to state-of-the-art quantitative heterog and new models and I will show you that at the end. I will first deconstruct fiscal dominance in rank. Once I do  
[00:16:38] that, I will study fiscal mant interactions in hank and try to deliver the second lesson. Ah, and now it's working perfect. Okay, the building blocks of the model a supply get demand policy a supply is  
[00:16:53] going to be a Philips curve. [sighs] The Philips sc I'm going to use I want you to be thinking in general some mapping from the path of output to the path of inflation plus some disturbance some cost push socks okay h for some of the  
[00:17:06] theory we're going to be using a simple static Philips cave just to give you clean answers in the quantitative explorations we'll be using either the standard or the hybrid philips kev whatever the data want okay and as I said before elementary observation that  
[00:17:21] I want you to repeat for this class of models not necess necessary for reality or other models is that I'm going to be looking like many of us do in models where inflation is pinned down by Philips C. So it's pinned down by real  
[00:17:35] spend aggregate demand a is going demand is going to be a very simple extension of the representative newian model it's going to be textbook new kenian model  
[00:17:48] meets blansar overlapping generations of households with survival probability omega when omega equal one households live forever we're going to nest the  
[00:18:01] representative age and newen When omega is less than one, households die because they die they are become nonrecardian somebody else may pay the tax burden in  
[00:18:16] the future. Okay that version of finite horizons finite lives is just a proxy in our view for the effect of boring constraints in complete markets incomplete markets break regarding  
[00:18:29] equivalence similar to oil. Okay. And I'm showing here the aate consumption function because that's where a lot of the economics are going to be. So what this aate consumption function tells you how we obtain this you know individual  
[00:18:43] optimal consumption log linearized around the steady state aggregate across all households of different generations and the like. It has a component okay it has a component that is like Freriedman's permanent income  
[00:18:58] hypothesis. If you look at the black, not the gray stuff, it's simply the consumption is a multiple the margin propensity to consume out of what? Your financial assets and the discounted  
[00:19:11] present value of your income net of your taxes. Okay, that's what you would you know get from Freriedman econ 101. Okay, the h the gray terms on the right are the effects of interest rates and the  
[00:19:26] demand shock, a discount rate shock. Ignore the last part. Focus on the first. When omega equal one, we have literally permanent income infinite horizon permanent income hypothesis. In that case, the MPC is one minus beta.  
[00:19:40] It's 2%. Okay. And future disposable income is discounted at the rate beta which is the interest rate faced by the government. So households have very  
[00:19:51] little MPC and discount the future with extremely low discount rate. If you are in rank now as you move from rank to OLG what happens is households because omega  
[00:20:04] is less than one they have higher MPC and they discount the future more heavily than the discount rate literally this is because here households have finite horizons but within complete  
[00:20:17] markets again liquidity constraints give you high MPC liquidity constraints make people more forward less forwardlooking so in that sense we're mimicking key properties from a general class of H models. Okay.  
[00:20:31] Now you can think about consumption in terms of the consumption function. But if you use market clearing to replace C with Y and private assets with government debt  
[00:20:46] and rewrite this equation in recursive form with a little bit of algebra, you can show that you can write this in the bottom equation which is like the oiler equation or the dynamic is equation you have in the standard Newian model except  
[00:21:01] you get an extra term an extra term that depends on the quantity of public debt.  
[00:21:06] if and only if omega is less than one. If and only if you have finite horizons and that captures the worth or liquidity effect of government of fiscal policy on demand. Okay. So I will be talking about the effects of fiscal policy that come  
[00:21:21] through this red term fiscal policy baseline I want you to think to simplify your life risk-f free one period real bonds as in borrow h  
[00:21:34] later or in the papers nominal debt or long-term debt government budget you know debt accumulates depends on what on what is the interest rate which is better invest how much debt you have what surplus you run I'm writing And  
[00:21:48] here the surplus inclusive of interest rate payments. Okay, so surplus is taxes minus government spending and an adjustment for interest rate payments.  
[00:21:57] And I'm going to assume the following fiscal rule, a fiscal rule that equates surpluses to a proportion of income to y times income. And I want you to think of  
[00:22:10] this as the automatic stabilizer. If we have more real economic activity, if the economy is booming, automatically you are getting higher tax revenue and more surpluses.  
[00:22:23] On top of that, you may intentionally adjust the tax schedule to raise taxes if you have a lot of debt. The second term captures precisely that kind of  
[00:22:35] fiscal adjustment. If Taudi is large that means you are hiking taxes a lot when you have a lot of debt or you're hiking them fast. So a high value of TA means fast fiscal adjustment. A low  
[00:22:50] value of TAD means slow fiscal adjustment. Taud equals zero means no fiscal adjustment.  
[00:22:57] So what I'm going to be doing is understanding how the economy the equilibrium behaves for different monetary policies as we vary to okay that's going to be the key comparative static of interest if I have slow or  
[00:23:12] fast fiscal adjustment what happens to the business cycle what happens to monetary policy and so on okay equilibrium they usually think mandary policy sometimes we describe monetary  
[00:23:25] policy as a tail rule for interest rates in the nuc framework effectively by controlling the nominal interest rate.  
[00:23:33] The central bank can control the real rate. I want to be thinking of the central bank here directly regulating the real rate of interest at least in the short run. Okay, there are details how you can implement  
[00:23:46] this with a you know with a certain rule for the nominal interest rate but as I said I'm going to be thinking the central bank controlling the real interest rate and I will be asking two questions closely related the first is how did the  
[00:24:01] speed of fiscal adjustment matters for given monetary policy in the sense of a given path for real interest rates [snorts] and then I will be asking how it matters for the optimal monetary  
[00:24:13] policy when the central bank chooses the real interest rate and its contingency to socks so as to stabilize the economy or to do its best in stabilizing the economy. Okay.  
[00:24:26] So let's understand how TAD matters, how the speed of fiscal adjustment matters in rank versus hunk and I want to zero in on what I said before on within this model fiscal  
[00:24:41] policy can matter only if you somehow break Ricardian equivalence. Okay to do that what I'm going to do now is the following. Again I will suppose Omega Quran we have Ricardian households and  
[00:24:54] for simplicity we can generalize. I want you to focus on a monetary policy that perfectly stabilize the real the real interest rate to steady state value which here means in terms of log  
[00:25:06] deviations to zero and I don't want you to get confused with the oiler equation things like that let me look at the consumption function I'm doing freedman okay as basic as it is. So I look at the consumption function and I see the consumption  
[00:25:20] function. I'm going to rewrite it. Uh again I wish I could show you but the first equation that simply tells you consumption is the sum of two components. One minus beta Z plus one minus beta the discounted present value  
[00:25:34] of income. So the second part is permanent income as we learned from Friedman. And what is this mysterious Z?  
[00:25:41] It's not very mysterious. G is the private assets I have as a household one pocket minus the discounted present value of taxes that I owe to the government which is the obligation I  
[00:25:54] have in the other pocket. Okay. Now I'm a student of borrow. So when Barrow taught me the equivalence I understood as follows. Rational households should  
[00:26:08] understand that whatever they have in the one pocket cancels out with what they have in the other pocket. Now this is a property of rational expectations equilibrium. So it requires that people understand equilibrium. Okay. But rational expectations equilibrium is  
[00:26:21] what we use most often not me myself in all my research but in this class of models that's the benchmark. So if agents have rational expectations they should understand that in equilibrium  
[00:26:34] private assets equal government debt equal present discounted value of assets. So they should understand that this Z cancels us out.  
[00:26:43] So in equilibrium consumption has to equal permanent income and in equilibrium fiscal policy is nowhere to be seen in this equation.  
[00:26:52] So now let me ask you if the government decides to run more fiscal deficits today what should happen to spending when Robert Barro taught me told me nothing because regardless understand  
[00:27:06] this is a zero so they shouldn't change their spending okay I see it in this equation but then I look at all the new kins and literature that assumes fiscal dominance in representative agent model and I see  
[00:27:20] equilibria where in response response to h fiscal deficits there's more output and more inflation.  
[00:27:29] How is this happening? The resolution is still in this equation and the resolution is as follows that in the new model income is demand determined.  
[00:27:42] So this opens the door to a purely self-fulfilling loop. If you all spend more forever after, then our permanent income is higher and then we can all spend more forever after. Okay, now  
[00:27:55] emphasizing this self-fulfilling loop. That's what it takes within the model to sustain fiscal dominance. Okay, now I love multiple equilibria in general, but that particular multiplicity I'm very  
[00:28:10] allergic to. Okay? Because it requires through this lens, not everybody will agree with this prism, but through this prism, it requires consumers to coordinate on spending more when there  
[00:28:23] are high fiscal deficits simply because that's what it takes for the government budget to be satisfied. Okay, that's my interpretation of what fiscal dominance means in the representative education  
[00:28:36] model. That's no claim about what fiscal dominance may mean in the real world.  
[00:28:42] Okay, I'm not talking about how fiscal policy matters in the real world, but within these models, that's how it matters. As I said, this sounds suspicious in my in my view. And it turns out that is very fragile in a  
[00:28:55] sense that I have formulated make clear in other work. But I will tell you one simple way to see this fragility.  
[00:29:05] If consumers expect the economy to return to steady state, not asytoically as we usually route our models, but in finite horizon,  
[00:29:18] but in any finite horizon. Under this refinement, you cannot have physical dominance. The reason is simple. You can see it from the oiler  
[00:29:30] equation. If y at some final period is back to steady state, you can iterate the oiler equation and get that y has to be in steady state forever and can only be affected by changes in real interest and not by fiscal policy. Okay. So if  
[00:29:44] you use this refinement that beliefs of real economic activity at long enough horizons are anchored to the steady state then you get an equ unique equilibrium regardless of whether the  
[00:29:57] tailor principle is satisfied or not. a unique equipment where fiscal policy doesn't matter and unique equipment where monetary policy is dominant even if the Taylor principle holds. Okay. So  
[00:30:10] my take away from that is that you know I don't want to talk about fiscal dominance in the representative engine model. I want to go away from it to understand how fiscal policy affects  
[00:30:24] inflation demand and the like. Okay. How does ta mean hank? I can go back to the consumption factor and again it's very simple to see what's going on.  
[00:30:34] Consumption depends again on something like this Z and the discount to present the value of income. But there's a big difference when people look at what they have in their own pocket. They have private assets. That's public debt. What  
[00:30:48] is on the other pocket is the discounted present value of their taxes now discounting these taxes at a higher rate than the interest rate of the government. Why? Because either again I'm going to die and somebody else will  
[00:31:01] pay the taxes or because boring constants effectively introduce extra discounted of the future. Okay. So now if you look at this CT as you see clear is the discounted present value of taxes  
[00:31:13] under the interest rate minus the discounted present value of taxes with the extra discounted by the households.  
[00:31:22] As long as omega is less than one, as long as households are nonreardian, this term is non zero and has the following very natural property, which is important for what I'm going to talk now.  
[00:31:35] If you push tax heights further in the future because households discount them more and more, you are going to stimulate demand more and more today. So the same size of deficit or stimulus  
[00:31:48] check today will have a more stimulating effect in the economy if you backload the fiscal adjustment. That's the key mechanism. Okay, once you focus on this mechanism, there  
[00:32:01] are two lessons that come out relatively quickly. Okay, I want to say relatively quickly because we work hard to get the lessons, but I think they're intuitive. Okay, the first one is of  
[00:32:14] course now because households are recalian deficits contribute to demand higher demand will mean both more output and more inflation. So def can be inflation everybody understands that that's not our contribution. Okay. But  
[00:32:28] what we have shown in this in in a paper is that to the extent that these deficits come with very backloading of taxes which sounds like they may start looking at such unfunded deficits in the sense  
[00:32:42] you know you run a deficit and you don't plan to raise taxes in the short run.  
[00:32:46] Who knows what happens in the long run but you don't raise tax in the short run. then you may start getting through this classical non-recordardian mechanism sufficiently high inflation pressure that can be quantitatively  
[00:33:00] close to what we would have gotten in rank with fiscal dominance.  
[00:33:06] So what I just told you is like I may have almost negated my previous message but I think this is still a lesson. I told you that the mechanism that was there before in the literature is flawed.  
[00:33:18] But I also told you that some of the applied lessons of this literature in terms of inflationary pressures or fiscal deficit or fiscal origins of uh inflation could be safe. Okay. [snorts] So I think that's part of our our  
[00:33:32] lesson. The applied lessons okay lessons about lessons. some of the applied literature that has focused for example uh uh in this room they have thought  
[00:33:46] more about in the data how much inflation may how much fiscal deficit may have contributed to inflation they we may have used those models that had this problem but perhaps if those models  
[00:33:58] were proxy models with non-recordardian consumers maybe the answers they got in terms of the structural interpretation of the data was not that wrong. Okay.  
[00:34:08] However, this new approach of looking into this comes with both more empirical discipline because now how big are those inflation pressures are within the model  
[00:34:22] is not a question of untestable equilibrium selection. It's a question about measurable things such as how fast fiscal adjustment is and how accommodative monetary policy is in the sense of whether we raise real interest  
[00:34:37] rates or not in response to government sp deficits.  
[00:34:41] >> [clears throat] >> The second lesson that I want to use my remaining half hour to look at is uh uh okay once we think about this non-recardian mechanism  
[00:34:54] should the central bank prefer fast fiscal adjustment or slow fiscal adjustment over the business cycle and as anticipated my argu my going to try to make the argument that perhaps you  
[00:35:09] know exante maybe counter But exposed I think it was very intuitive that fiscal inaction in the sense of very slow adjustment or even no adjustment is good  
[00:35:21] for the dual mandate of the central bank for two reasons. It will help stabilize both output and inflation against demand socks and will ease tax distortions and  
[00:35:35] the inflation output trade off in the presence of cost socks. So what I'm going to do and I'll let you take a you know a break and let myself take a break for a moment you know sort of I to  
[00:35:49] finish this digression that I had in terms of the representative agent new ken model and the question of monetary versus fiscal dominance I'm going to focus now on nonrecardian keyn economies I'm going to sidestep  
[00:36:07] completely this equilibrium selection uh um debate and all those perplexing and controversial issues and I'm going to focus on this very basic mechanism non-recondant behavior and how it  
[00:36:20] affects optimal policy. Okay, that's the paper uh we're writing right now on fiscal in action as monetary support as support for the job of the central bank. Okay.  
[00:36:34] So the description of the private sector demand and supply and of fiscal policy exactly as I did before. What I'm going to add here is a description of optimal monetary policy. And to talk about optimal monetary policy I need to tell  
[00:36:49] you what is the objective of the central bank. So the objective of the central bank is going to be the one you see here. It has two very traditional components. the output gap square and  
[00:37:02] inflation uh square. So the central bank is trying to minimize uh uh both output gap fluctuations and inflation fluctuations. That's the standard dual mandate and a little bit non-standard  
[00:37:16] somewhat non-standard is the last term that I'm also adding the square of R of the interest rate. Okay, what I want to capture that what I want to capture from this is  
[00:37:30] something that is absolutely true in reality but absent in the textbook micro foundation of the social of the wealth of the central bank objective and to capture the idea that the central bank  
[00:37:42] does like quick and large movements in interest rates because they want to be more gradual and more slow and more reserved on how they move interest rates for a number of reasons that you know  
[00:37:57] better than me. Financial stability, uncertainty about what are the socks hitting the economy, uncertain about the transmission of monetary policy, erh uh the model we use about the world. I don't want to think of this as other  
[00:38:11] constraints to monetary policy such as the GLB in some situations. Okay. So all I'm doing there is to capture the idea that the central bank in the real world unlike our textbook models does not have  
[00:38:25] free h control of real interest rates and through that a demand. So the central bank cannot easily get the complete first best of zero output gap  
[00:38:37] and zero inflation when even if you have demand. you know can lead the economy in this direction but cannot get the first best okay that's all I'm doing the normative question is how the optimal  
[00:38:52] policy the path of our star the associated output and inflation and the resulting loss basically the total volatility and output inflation and  
[00:39:04] interest rates vary with ta and the intermediate step for to Normative question is the positive question. If I were to fix monetary policy, what would be the effect of  
[00:39:17] slower fiscal adjustment? So first I will explain you for given monetary policy what slower fiscal adjustment does and that will help you understand whether this helps or interferes with  
[00:39:31] the job of the central bank. So starting with a positive with a positive question with a question you know for giving interest states first a lema that tells you how you can  
[00:39:45] understand equilibrium or represent equilibrium output in this economy what the lema tells you and from here on I'm going to be thinking represent the  
[00:39:57] economy under perfect for sight so I want to be thinking that today there are some socks that hit the economy today.  
[00:40:05] But maybe we're also getting perfect foresight news about demand socks or about monetary policy in the future. And I'm going to represent output as an IRF with respect to this kind of socks.  
[00:40:20] That's what equation one tells you. tells you output in period as a linear combination of the expected interest rate and demand shock. E is the discount  
[00:40:33] rate. R is the interest rate in all future periods. Okay, this calligraphic Y are the IRF coefficients that of course you need to  
[00:40:47] solve through the model. Okay. Now any hunk model will give you some coefficient like this. You can solve the IC then the product and cross find the solution. You'll get something like this. Our G model gives you a closed  
[00:41:02] form or retractable characterization of these coefficients. Okay. So now I'm going to tell you what happens to these coefficients as I change the physical frameworks as I make the physical adjustments slower. And that's going to  
[00:41:16] be our key positive result. Holding the path of interest rate fixed slower fiscal adjustment lower to D stabilizes dynamically the economy in  
[00:41:30] the following dual sense. All these IRF coefficients increase as you lower to D. So to interpret the sign of this increase I want to think you  
[00:41:43] know if I raise interest rates or I raise discount rates then you should be expecting output to go down. So you should be expecting this calligraphic y coefficients to be negative numbers.  
[00:41:56] Okay it tells you basically higher interest rates lower output. So these are negative numbers. So when I tell you that lower ta slower fiscal adjustment increases these numbers it tells you that you you're getting less of a  
[00:42:11] recession as I will argue actually they make even cross some of them and you get something positive I'll explain that okay but in that sense you are stabilizing the entire IRF of the economy vav both  
[00:42:23] current and news about the future the second part tells you that if you look at the cumulative IRF >> [sighs] >> again because we're thinking it's like a recessionary shock that's negative the  
[00:42:38] cumulative IRF but as you lower to D you make this number closer to zero it increases and in the limit of no physical adjustment you're going to make  
[00:42:52] this number zero. So by backloading fiscal adjustment enough you can perfectly stabilize the NPV of output. You can make sure that whatever  
[00:43:05] recession you get today is perfectly offset by a boom in the future. Okay. So in that sense slow fiscal adjustment can be very powerful way to stabilize the economy at least in an NPV sense. Okay.  
[00:43:19] That's the statement. I will explain a little bit more the intuition in a second. But first I want to tell you the following. I want to keep that in mind.  
[00:43:28] These are the IRF coefficients to the demand sock. So I told you the good news that slow fiscal adjustment stabilize the economy against demand socks.  
[00:43:39] [snorts] But demand sock and interested socks is the same thing.  
[00:43:44] So the flip side of this result is that slow fiscal adjustment also stabilizes the economy against monetary policy actions. So slow fiscal adjustment reduces the effectiveness of monetary  
[00:43:57] policy. Now that may sound bad but in total because you are going to be at this you know the only reason that you are reducing the effectiveness of monetary policy is that you're reducing  
[00:44:12] to start with the effect of the demand sock. So in total it would turn out that the central bank is better off by this uh despite the apparent reduction in the effectiveness of monetary policy. So let me explain this what's going on how this  
[00:44:26] stabilization works. Okay. So I'm going to use not math because it's a bit more complicated but some illustrations.  
[00:44:35] So I want you to fix ideas. Think about an A1 demand sock that will create a recession in the economy. Okay. And I want to start with a situation where the adjustment the fiscal adjustment is  
[00:44:49] fast. So ta equal one you get a recession and the recession translates automatically to a deficit because your tax revenue is going down.  
[00:45:02] So of course debt will spike which is what you see on the right hand side. Now what is happening the fiscal authority is raising taxes a lot to try to bring  
[00:45:13] debt down. If the sock had been IID, what you would have seen is debt spike for one period and then it went down immediately because you raised taxes like crazy in the next period because  
[00:45:26] the recession is persistent. You know, you get a recession, you hike taxes, but then you're hit again by recession. So debt will will still remain above steady state, but will go back to steady state quite fast with the  
[00:45:40] same speed as the economy recovers. Okay, that's the benchmark with to equal one. So starting from this situation with fast fiscal adjustment, let me understand what happens to the economy.  
[00:45:52] I if I lower to less than immediate fiscal adjustment and I'm going to get the the slightly less dark line which is uh maybe it's harder to see here, but let me lower it  
[00:46:07] out even more and then you start seeing it more clearly. the impulse response of the economy is going up. That's the first part of my proposition. You are getting a smaller recession.  
[00:46:21] What's happening at the same time? Well, you're not adjusting taxes that quick.  
[00:46:25] Not surprising, debt initially is going up. Okay? But debt is going back eventually. It's going back for two reasons. One is that eventually you're raising taxes. The other is that  
[00:46:39] because you backload the taxes, you're stimulating demand more for the non-record that I told you before. Now these nonrecardian houses and they start, okay, there's a deficit now which effectively  
[00:46:53] as if you are giving me back some money. Okay? And you are not going to take that money away from me right away. You are going to take them away from me much later. Okay? So the stimulating effect of aggre is going to be larger when we  
[00:47:07] backload taxes. That's why you're seeing output being depressed less.  
[00:47:14] But since output is depressed less that means that contributes to debt going back to steady state even if you had not raised the taxes. That goes back to our self- financing mechanism of how  
[00:47:26] backloading taxes helps stabilize debt without the need for significant tax adjustment.  
[00:47:32] Now what is happening if you are reducing to D even more and more you go to zero then as you can see in terms of the impulse response of output you are getting a recessionary loan that is more much more modest than would have  
[00:47:47] gotten otherwise and you are actually getting a boom later on.  
[00:47:51] What is this boom? How is that happening?  
[00:47:54] Well, if you are not raising taxes fast enough, you allow some public debt to be out there in the market after the recession and shocks has gone away. That means you  
[00:48:09] have increased liquidity in the private sector even after the exogenous shock has gone away. So in this persistent increase in private liquidity that will explain why you end up getting you have relaxed boring concession for the  
[00:48:23] households even after the liquidity sock is gone that explains why you are getting a boom after the initialization.  
[00:48:31] Okay and as I told you if you were to integrate that line you are going to get exactly zero. I'm not going to explain why exactly you get zero, but I want you to understand that slow fiscal adjustment will give you a boom later on  
[00:48:44] that will help moderate the recession today.  
[00:48:49] Something also to note from that is the following that has the flavor of forward guidance.  
[00:48:57] What do we do with forward guidance monetary policy? you promise to stimulate the economy in the future and that helps you stimulate the economy today at least in the theory whether it works in reality it's a big question okay so another way to interpret this I  
[00:49:11] just told you is the following that slow fiscal adjustment in it implements a sort of forward guidance on the fiscal side instead of the monetary side and it works because households are nonrearding okay you are  
[00:49:26] creating fiscal stimulus effectively in the future Okay, now I'll skip the limit.  
[00:49:34] What happens with that was what happens with given interest rate with given monetary policy. So now let me explain what is the implication for optimal monetary policy and I think now you you know you  
[00:49:48] must almost guess the result. Okay. [snorts] So I'm going to argue that the strategy for proving the result is you look at the envelope condition that gives you how the the optimal loss of the central bank  
[00:50:05] depends on uh the speed of fiscal adjustment. Now this naturally depends on two objects. The optimal allocation and the sensitivity of output to T to D.  
[00:50:17] We already characterize what's the sensitivity of output of the impulse response of output to to D. That was the previous result. The properties of the optimal policies relatively easy to understand from what we already know. So you put things together and you can  
[00:50:32] understand what happens to the CB loss as a function of the speed of fiscal adjustment and the result is as follows that for oops there's a typo here the first line should be for demand socks  
[00:50:46] okay for demand socks I'm telling you that the loss increases with taudi which means the loss is minimized when ta equals zero so the settle bank prefers very slow fiscal adjustment ment in the  
[00:51:00] case of demand socks and conversely for supply socks is the opposite for supply socks here the central bank prefers fast fiscal adjustment what's the intuition for this basic on  
[00:51:14] what I told you before the stabilizing effect of low fiscal adjustment on F output is true regardless of whether you have demand or supply socks in the case of demand socks the central  
[00:51:28] bank is trying to stabilize output and inflation and the slow fiscal adjustment helps the central bank. What happens with supply socks? If you have a cost per shock, the  
[00:51:42] central bank is not trying to stabilize output. In fact, with a cost per shock, the central bank, the optimal monetary post to destabilize output in order to stabilize inflation. If you have a you know an inflationary cost per sock the  
[00:51:57] optimal monetary policy is to reduce demand reduce output so as to lean against the cost per sock. So the central bank in the case of supply socks is trying to destabilize output. Now I told you slow  
[00:52:11] fiscal adjustment stabilize output but now slow fiscal adjustment in the case of supply sock goes in the opposite direction of what the central bank wants. So that's going is going to be bad in the case of supply socks where it was good in the case of  
[00:52:26] demand socks. And this is an illustration now of what happens to the impulse response of output and debt and the total [clears throat] demand wedge in the case of demand socks. Basically you know even though monetary policy  
[00:52:39] adjusting here what you see is the following. If tout is close to zero that's the light gray lines the recession in the economy is weaker  
[00:52:53] comes together with some boom later on that's the fiscal for lines that I told you now if you also look at the demand wedge the demand you know the sock is always the same so when you look at the IRF of the demand wedge which the sum of the sock and the interested you're  
[00:53:08] really seeing what happens to monetary policy and the lesson there is the following You know you have a negative demand so sorry this is a yes you have a negative you know contractual demand shock.  
[00:53:20] Optimal monetary policy fights this by lowering interest rates both now and in the future but those movements are costly or difficult for the central bank. What the slow fiscal  
[00:53:33] adjustment does is yes it reduces the effect in monetary policy as I told you before but it gives you for free more stabilization. So now the central bank can achieve more output stability and  
[00:53:46] more inflation stability even with less action inside. So there's less need to move interest rates as aggressively as before. Okay.  
[00:53:56] Conversely, in the case of supply shocks, slow fiscal adjustment makes reduces the recession that the central bank is trying to create in the case of  
[00:54:11] inflation cost that not only increases inflation. Okay, so you're getting worse outcomes in terms of inflation which we were trying to fight. So slow fiscal adjustment maybe for the postcoid sock maybe that's  
[00:54:25] bad I don't know but for co socks that's where it's bad okay not only that if you now look what happens to interest rate policy you will see that the following interesting pattern  
[00:54:39] so if I start from very fast fiscal adjustment I and I make it a bit slower you know the fiscal policy works against the central bank initially the central bank tries to offset  
[00:54:51] the bad fiscal behavior by moving interest rates more. But if fiscal adjustment is so slow, then output is so well stabilized by the fiscal authority that the central bank  
[00:55:05] gives up. It's demoralized and say okay it's pointless for me to try to stabilize outward because the effectiveness of monetary policy now is low. So you are getting that even the optimal monetary policy response is now smaller when fiscal adjustment is low.  
[00:55:19] So bottom line, I've already said this two times, so I'm going to repeat three to make sure it registers. Okay, if I think about whether lower to D helps or  
[00:55:31] not, the description the the the lesson so far is the phone. This is describing think of the solid lines. The blue line tells you where the loss looks like as a fun fact of the speed of fiscal adjustment in the case of demand socks.  
[00:55:45] The red in the case of supply socks. For supply socks, you want fast fiscal adjustment to equal one. For demand socks, you prefer slow fiscal adjustment. The dash line  
[00:55:58] tells you how all this objective changes, how this state of change if you change to y, which is the size of the automatic stabilizer. So if you have a stronger automatic stabilizer, the usual  
[00:56:11] automatic stabilizer, that's going to be good in the case of demand socks and bad otherwise. Okay, so far I promise in the beginning tell you that slow fiscal adjustment is good  
[00:56:24] always. What I told you now is slow fiscal adjustment is good for demand socks, bad for supply. That's an important lesson. Okay, and I want you to keep it. And that's okay. But now I'm going to give you an extra lesson that  
[00:56:38] will tit in the direction of slow fiscal adjustment even for demand even for supply.  
[00:56:45] So far when I thought about fiscal adjustment the tax hikes were not showing up in the Philips curve they were only affecting demand. That meant that effectively I was assuming that all the tax adjustment  
[00:56:58] is non-distortionary. However, if tax hikes are distortionary, you know, they they hurt the labor supply, private incentive, the like, that will manifest as a tax wedge in the  
[00:57:12] Philips Kev effectively as an endogenous cost per shock proportion of the tax adjustment.  
[00:57:19] So now let me move to this situation and adjust the Philips Kev again. Maybe you have the forward looking or the backward looking components. But the key thing is now I'm going to add this tax wedge in the Philips curve and I want now to  
[00:57:32] focus what is the extra h mechanism brought from the supply side now. Okay. And given time constraints I'm not even going to go through the math but I will just give you the basic  
[00:57:46] intuition which is the following. I told you before that by regardless of whether you have a demand or supply shock, a a a slow physical adjustment  
[00:57:59] helps stabilize the net present value of output.  
[00:58:06] But if you stabilize the changes in the present value of output, you are also stabilizing the net discounted present value of tax revenue. Because if the net present  
[00:58:20] value of output doesn't move much, tax revenue in net present value is not going to move much.  
[00:58:27] So with fast fiscal adjustment, you are going to have big net present value of tax hikes because you're not getting tax revenue for free and you need to substitute for it with tax  
[00:58:40] hikes. Backloading those stocks highs by stabilizing the present value of output also stabilizes the present value of tax revenue which means that you can get into temporal government budget  
[00:58:53] balancing without tax heights. So by backloading tax hikes you make sure that the net present value of tax distortions doesn't move in the economy. Okay. And that makes those tax wages to be small  
[00:59:05] indulgencely thanks to the non-recardian behavior in the channel I described before.  
[00:59:11] So fast fiscal adjustment is going to be bad because it will create this indogenous cost for socks. Slow fiscal adjustment will minimize this endogenous cost for socks and is the trade-off between inflation and output even in the  
[00:59:25] case of supply shock. That's why now even in the case of supply shocks it may be that you prefer a slow fiscal adjustment. So to sum up the theory and then I'll conclude with quantitative application.  
[00:59:39] The theory has told you so far has given you reasons why slow physical adjustment may be good but didn't tell you always it's good okay it gave you some conditions for it to be good the conditions was the first one I didn't  
[00:59:52] explain enough but you need sufficient non-recordardian behavior of course and a sufficient physical automatic stabilizer but you know it's more likely to see slow physical adjustment being preferred if demand socks are the main  
[01:00:06] driver of the business cycle And if the inflation operation of tax distortions is sufficiently high. Okay.  
[01:00:14] Now is the business cycle driven primarily by demand or supply socks? We have guidance from the empirical literature about that. I think most of us would argue that it's not TFP socks that drive most of the business but some kind of demand disturbances that are not  
[01:00:29] fully offset with by monetary policy. The second what's the inflationary pressure of distort of tax distortions?  
[01:00:39] I don't know evidence about that but we can use standard estimates of the elasticity of free labor supply and so on to calibrate the model. So what we're going to do in our quantitative application is the following. First  
[01:00:51] we're going to the application with apologies is going to be in the US not in Europe. But uh uh what we're going to do in our richer model is the following on the demand side.  
[01:01:04] So far I use a stripped down hack model that focus on the non-recardian aspect but assume the way heterogenity. So we're going to reach a a we are going to write a richer model which will  
[01:01:17] accommodate heterogenity both in MPCs and in wealth accumulate in in wealth holdings. Okay. particular we're going to have agents who are relatively poor and h have relatively high MPCs and  
[01:01:31] agents who are relatively rich and have relative lower MPCs and that will approximate very well what you get from uh quantitative hunk models on the supply side we are going to use a realistic hybrid nuc Philips curve what  
[01:01:46] the data like calibrated the more recent evidence about the Philips curve on the fiscal block you know tow Why we're going to calibrate on what is the average rate of taxation debt to GDP?  
[01:02:00] We're thinking here about debt held by the private sector in the economy standard things and the maturity structure because we'll have long-term debt. What about socks? Socks will take two approach. You could think about writing a maximum likelihood estimation  
[01:02:14] of this model. We're going to follow a different approach which we think is a bit more transparent.  
[01:02:20] The demand block and the supply block. The key parameters are going to be a disciplined by evidence outside the US time series. Okay. In particular, the  
[01:02:32] demand block we're going to calibrate it to evidence about interal marginal propensity to consume and evidence about wealth heterogenity. The supply block from what I told you about identified  
[01:02:45] Philips curves. Okay, that leaves us with only one thing to have identified in the data. the socks.  
[01:02:53] So for that we'll follow two approaches. One by my co author Christianos his cos not the rest of my co authors where simply takes the following simple idea if we not if we know the rest of the parameters of the model and we only  
[01:03:08] don't know the socks we can identify the socks basically by the world de composition of the data. So you can use the world decomposition of the data together with the the primitive  
[01:03:20] parameters that I told you before to create counterfacts about how slow fiscal adjustment will matter without putting parametric restrictions on whether the demand and supply so AR1 or MA or whatever. Okay. And another  
[01:03:35] simpler version is from my work you know you can represent the bulk of the business cycle as some sort of a principal component that looks like a demand. So if you do this and you  
[01:03:48] ask what is the CB loss basically the sum of output volatility inflation volatility as a function of TAD you are going to get this figure for the US data. Focus first on the gray  
[01:04:02] line. The green line is with distortion at axis and with the wall de composition of the data. And you see here that the optimal you know the preferred is extremely close to zero. The other two  
[01:04:15] lines give you two you know variance. The red line is what if you had had lumpsum tax dis tax tax adjustment instead of fiscal. The blue is if all the business cycle was demand socks  
[01:04:29] instead of some mixture of demand and supply socks. The picture is the same.  
[01:04:34] So I'll conclude with here uh uh I gave you some intuition of why with non-recardian consumption behavior a slow fiscal adjustment  
[01:04:47] uh uh slow fiscal adjustment or fiscal inaction may be welcome over the business cycle. uh obvious caveats uh whatever I said uses the new kinsen model presumes output is demand  
[01:05:01] determined so it's about the business cycle nothing about structural trends steady state and the like uh and I I'll leave it here thank you [applause] so so that was a tour to force um of uh  
[01:05:22] covering such a wide range and uh and more or less an hour of continuous uh presenting is uh especially after a transatlantic flight is quite something.  

## Q&A (01:05:32 – 01:20:56)

[01:05:32] So uh I think to give Mario a little bit of a breather, we'll collect a number of questions before going back for any answers. Uh so uh Oscar here in the middle. Oscar, thank you.  
[01:05:46] >> Thank you. [snorts] Thank you very much for this excellent presentation. So two uh brief comments. one on the on the semantics remark that you made at the at the beginning of your talk on fiscal dominance. I think I would fully agree with you. I mean the channel you are  
[01:06:00] exploring I think is is fully in line with the normal prescription that any central bank model would would uh probably produce. I mean if if I ask any  
[01:06:12] ECB model about the likely outcome on inflation of more public spending that is at least partially finance with depth the empirical answer would be more inflation which is not about fiscal dominance it's about fiscal relevance  
[01:06:26] for macroeconomic outcomes including inflation and probably through the channels in part that you are analyzing here through non-recordardian channels so I I would agree it's not about fiscal dominance it's about the relevance of fiscal outcomes for for inflation  
[01:06:40] that in regular work here at the ACB I guess the same is true for any central bank in the world are absolutely central to our uh uh routinary analysis. Now more fundamentally I mean uh if I were  
[01:06:53] to focus on the which which I think is the most relevant quadrant of your analysis for us for the euro area which is one in which uh supply shocks are probably more relevant. This was one of the takeaways  
[01:07:08] of our analytical work in the last strategy review. We should probably be prepared for more frequent and more intense supply shocks, negative shocks perhaps. Uh and in face of frictions uh  
[01:07:20] you highlighted tax uh tax distortions. Uh my conjecture here is that perhaps um [snorts] a temporarily negative to a temporarily negative reaction to depth  
[01:07:35] that is contrastical fiscal policy could be part of an optimal program. No I in mind I have an example and I would like you to to tell me whether it makes sense or not. In 22,  
[01:07:47] following the huge increase in um energy prices, many European governments spent public money in putting some limits to the increase in energy prices and this facilitated to some extent the life of the central banks because total  
[01:08:02] inflation was more limited than otherwise and this in the presence of nominal rigidities in particular a strong wage rigidity in some countries and let's think about Germany it takes like three years for some wages to  
[01:08:17] adjust. This implied that the erosion of real purchasing power was lower thanks to this uh fiscal fiscal programs. So at the end of the day the persistence of the shock and so on was in terms of  
[01:08:30] inflation was lower than in the absence of this uh uh counter fiscal policy which I think it it facilitated to some extent the reaction of the central bankers. it allow us to be a little bit  
[01:08:43] more more smooth in in terms of the management of the interest rates and and so on. So would it make sense to relax the assumption that to this should be uh positive and consider possibly  
[01:08:55] temporarily necessary uh uh negative uh negative values for that uh for that key parameters. Thank you.  
[01:09:02] >> Thank you Rodrigo Oscar. Give to Leo. >> Um thank you. Super interesting uh the the presentation three three comments.  
[01:09:12] One is very related uh to the previous one but the first one is whether you have testable implications from here to differentiate from fiscal theory of the price level at  
[01:09:22] the end for in the data. Um second is whether I interpret correctly the numerical exercise at the end. Uh a  
[01:09:33] fiscal rule around a structural or cyclically adjusted deficit is the quote unquote the optimal one for for from the eyes of the of of the central bank. That  
[01:09:47] that I I guess is the the implication. And the [snorts] third question is if you had to think about a reaction function now from the government how you would would think about that we  
[01:10:00] will have a a similar loss function. Um what if they have discretion fully discretion what happened in that in that case? Um very interesting the point of  
[01:10:13] the negative uh tow I guess also there are some fiscal policies that are not directly demand policies only indirectly like subsidies no a fuel subsidy is a something that is not fully  
[01:10:27] captured here. So perhaps I don't know that opens also the the tool the toolkit. Okay, thank you Arreste and then Leo. Okay, >> so so thank you Marios.  
[01:10:41] Two clarifying questions. What on the first one is on the reference you made to the New Kenzian literature on the you know the the fiscal requirements for price stability. Uh I thought that you know one way to understand you know the  
[01:10:54] the Woodford and Gali the textbooks is is that it's convenient for the central bank. Well, the central bank operations are easier if the intertemporal budget constraint of the government is  
[01:11:07] satisfied by the government and and from this perspective you are imposing the same condition. Now I understand you're going beyond that but uh I I just wanted to make sure that this was again a clarifying question and the second one  
[01:11:22] also on the clarifying side is you emphasized a lot the the need the you know the key property this uh this lack of regardian equivalents in in what you're doing is this the only is it a sufficient condition so I was just  
[01:11:37] wondering for example we had this morning another paper with a two agent type of um you tank type of model. Would would this also hold in that kind of model?  
[01:11:47] >> Thank you, Leo. Then then Jessica at the back there.  
[01:11:57] >> Um yeah, I mean um uh thanks a lot. This was a super super rich presentation. Of course, very very well presented. Maybe just also  
[01:12:09] more on the clarifying side. Um the real interest rate you know and you talked about this a lot. I mean there's a preference for you to keep this over the cycle rather don't move it much keep  
[01:12:22] it constant but you and this made some of some of the expositions very clear but when you did the optimal monetary policy uh problem second part of your pres  
[01:12:34] presentation maybe you could make things clearer if you add to this maybe you do it somewhere in the simulation also you know the optimization problem without constraining the real interest rate. And  
[01:12:48] I say so because if you want to link finding maybe maybe two things that you have a new Keynesian setups with the tailor rule operating there there of course you know the strength of the real interest rate channel Taylor coefficient  
[01:13:01] larger than one less than one is very important kind of for the for the properties you see and this essentially you take out at the beginning. So this makes this could help maybe to to to to compare what you get with other papers  
[01:13:15] in the literature. >> Okay, thank you Jessica. Then then class.  
[01:13:21] >> So great presentation and the rank part I love and I'm all in on. And in terms of the Hank part, I guess there seems to be a couple assumptions that are doing a lot of work. one is the borrowing  
[01:13:35] constraint and the other is the um R squar term in the central bank rule. And I wonder if just an another interpretation of the data and maybe this is similar to the point about the  
[01:13:47] fiscal theory of the price level is you know a um fiscal authority that's been captured by special interests and is spending more than it should. If that is just an that's just a maybe an equally  
[01:14:00] good explan explanation of the data with very different normative implications.  
[01:14:05] >> Very good class. >> Thank you. Thank you for the wonderful presentation. You could really follow even without looking at the formulas as so well you explained. Uh my question is similar to follow following on to  
[01:14:19] Oscars. If you think about the US, there was a huge demand shock covid and fiscal policy expan very expansionary and then after and then inflation went up at that point in time. Uh fiscal authorities in  
[01:14:33] the US were very slow moving back to I mean there was no really consolidation which meant that inflation go went up and up to almost 10%. So in this period what would your optimal monetary policy  
[01:14:46] say? there was little of of an unemployment increase after say 2122 when when when fiscal policy could have stabilized more. So was it optimal for mon policy when inflation went to 8 9% that fiscal was so reluctant to really  
[01:15:01] consolidate or was at that point in time the object objective function of my policy all on output and very little on inflation. Thanks.  
[01:15:10] >> Okay. So you've accumulated a a mix of comments and questions. I I don't expect you to to you know uh fully address everything because we need to leave some of this for the for the reception and and the dinner conversation but but if  
[01:15:24] you want to take a couple of minutes. >> Perfect. Thank you very much and uh thanks for all the interesting questions. Uh uh uh a lot of homework also for me to think some quick answers. uh um completely  
[01:15:42] agree that we should separate fiscal relevance from fiscal dominance and and and I think part of the problem is sometimes we use the same word to talk about very different things and that's where theory is useful to clarify things  
[01:15:55] and that's what I try to do primarily to clarify how things matter uh in terms of whether it can be optimal in some situations to have like even ta negative I agree you know first of all you know  
[01:16:08] if you have uh uh uh it's not so much for the the energy example and that goes back to another question you know I didn't have subsidies to particular sectors but if you start having subsidies that's a way to directly  
[01:16:22] counteract the cost per sock uh uh so that was not the analysis and that you know there's a clear you know reason why you would like to subsidize costs when the cost go up okay and that was not in  
[01:16:35] my analysis okay uh um but you can think of ti negative proxy for that testable implication I want to to to to get into that okay it depends on whether we think about implication in terms of only macro  
[01:16:49] data or micro data for macro data yes there are differences with nonrecardian households they if you give them a deficit or a transfer they're hungry they're going to  
[01:17:02] spend it very fast so what you should expect is a boom in spending out inflation that is very front-loaded and short-lived.  
[01:17:12] Instead, if you are within the fiscal the price level, those households are really trying to smooth consumption forever. So there those fiscal socks generate very persistent movements in spending and inflation. Now I let you  
[01:17:27] you know I don't know the po the postcoid experience was inflation spiked but then went down. Who knows what happened in reality? But I'm telling you that's a testable difference at the macro data at the at the  
[01:17:40] macro level. Micro it's trivial consumers are not PIH households. Once they have you have nonicardian households we can just forget the whole debate about non-ESB equilibrium  
[01:17:54] selection talk about this standard classical mechanism and then we don't have to worry about it. It's true that this as I said this classical mechanism can give you outcomes that resemble FTPL at the macro level. Okay. But now you  
[01:18:08] don't have to you know obsess about whether John coin is right about the model or why I'm right. It's like we can forget all this controversy and focus on what is interesting which is outside  
[01:18:21] from the equivalent. I'm happy to talk talk more about that at dinner.  
[01:18:31] I want to emphasize the following. Again, this is like perspective, but for me that perspective is very important.  
[01:18:36] If house was a Ricardian, they don't think about the budget concern of the government. It's out. It doesn't matter. Okay. So then the government will either have to satisfy somehow the budget constraint. So fiscal  
[01:18:51] policy has to be active or then we cannot have an equilibrium. But what I learned from that is that if I'm in rank, active physical policy is not a meaningful assumption. Okay, tank versus hank,  
[01:19:06] tank is a very special degenerate version, so it has some issues.  
[01:19:11] But if you're in a generic hunk, whatever I told you goes through h R square h you know if we didn't have the R square let me be clear about the following with supply socks everything I  
[01:19:24] said would have gone through because with supply socks the central bank is not getting the first best it's in the second best already. Okay, with demand socks if you are in the textbook new kinen model and you don't have some  
[01:19:38] friction in monetary policy then the central bank gets the perfect first best. Okay. So in that case then your the monary authority is stabilizing demand no matter what the fiscal authority is doing and then fiscal  
[01:19:51] policy becomes irrelevant. Okay. But that's a knife edge property of really getting the first best. Anything that will keep the central bank optimum away from the first best will let the mechanism happen. So the R square and actually whether it's R square or the  
[01:20:06] nominal interest rate square or the nominal interest rate today versus yesterday the change in all of that will give you the same answer uh uh bo postcoid uh uh now I forgot what  
[01:20:21] the question was maybe that's a good moment to stop >> very good so so with that I think we've had a an excellent day let me thank all of the uh the speakers everyone who was energetic in asking questions uh all the  
[01:20:35] discussments um and uh you know this is of course just one day out of two so with that I'm going to close the live stream uh and then then I'm going to ask the uh organizers to to fill us in on the uh  
[01:20:48] the details of the reception and dinner so thank you everyone Thank you. Thank [applause] you.