 Thanks for the very generous introduction, and for having me at this conference. It's a great pleasure to be a part of it. In this talk, I will try to give you a synthesis of my recent work with various wonderful co-authors on the interplay of monetary policy, fiscal policy, and inflation expectations, and hopefully contribute to current policy debates about what we should do about inflation and what we should not do about inflation. Now, to give you just a little bit of a background, obviously you all know that the COVID crisis has been stressful at all levels. But if you're in charge of macroeconomic stabilization, this was a particularly challenging task. I don't know how central banks could cope with this. But in any case, it's very clear that the central banks fired every gun. They had their disposal interest rates are very low, rounds of quantitative easy and lots of forward guidance, a lot of interventions to make sure that the financial markets are functioning properly. The policy response, unlike the recession, was also very active, very aggressive. It's true in Europe, but it's especially true in the US, where we have trillions and trillions of dollars of fiscal support flowing into the economy. And when people look at this, some of them say, maybe we're doing too much of a good thing. Maybe we over-stimulate the economy. And as a result, we can create inflation that we have not seen from the 1970s. Double digit inflation, where we have high inflation, high inflation expectations, this chronic leashes cycle, when we have high inflation because inflation expectations are high, and inflation expectations are high because inflation is high. And so a lot of this analysis is based on the idea that once we incur inflation expectations, it's going to create a major problem for central banks because stopping this high inflation is going to be very costly in terms of our food. And this is the last thing we need to do when the economy is still very weak and we need to help the recovery, get steam, and happen as quickly as possible. Now, this inflation expectation is getting a lot of attention recently in part thanks to Jeremy Rod. But it's also true more generally that at least in the banking, central banking circles, there is a very clear understanding that inflation expectations are very important. And yet we have lots and lots of open questions about what inflation expectations are and how we can use them for policy and what kind of responses we should see from inflation expectations when we do certain types of policies. Now, to summarize the state of knowledge, let me give you three quotes from former Fed chairs. This was Alan Greenspan in the early 90s. He says inflation expectations is a key variable, but we don't know what it is exactly, but it doesn't mean that it is not real. So he kind of says, look when you don't understand inflation expectations, where have they come from? And it's obviously very important for policy. And some years later, Ben Bernanke gives another major speech about the importance of inflation expectations. Again, he says we don't have complete answers. It's a great practical importance. 10 years later, Janet Yellen gives another major speech and she says it's very important. We need to know more for understanding microeconomic dynamics and also for making policy. We want to use inflation expectations for policy to manage expectations as a tool. Now, it's about time Jay Powell gives a speech about the importance of inflation expectations, but recently he said that inflation expectations are terribly important and he is watching them all the time. Now, in any case, what should be clear here is that we want to know more about inflation expectations because the important for macroeconomic dynamics and for policy and we don't have complete answers. So we need to do more research here. And to some extent, these quotes are a bit in fear to how much progress we have made since the 1970s. We have a lot of progress in measuring inflation expectations. We have a lot of progress in understanding inflation expectations. So a lot of work has been done. And one of the striking findings in this work is that when we talk about inflation expectations, we should appreciate that we have many players in the economy and they have very different inflation expectations. And this is important for how we should be thinking about policy and how much we should expect from our policies in terms of macroeconomic impacts. Let me start with professional forecasters, this is easier. They have incurred inflation expectations. So you look at the US, the blue line is one year ahead inflation forecast for professional forecasters. They consistently predict inflation to be very close to 2%. If you look at longer horizons, it's even more so. So their inflation expectations are really anchored at 2%. And this is despite the fact that inflation over the last 10 years has been consistently below the 2% inflation target. So they believe that the Fed is credible in delivering on its inflation target. And the same is true for the Euro area, even though inflation has been consistently below 2%, you look at longer horizons and professional forecasters, financial markets, they all believe that the ECB can deliver on the 2% inflation target and we shouldn't have 2% inflation in the longer run. And so to some extent it's surprising that now when we have a little bit above average inflation that everybody is so much panicking, it's great we're compensated for this missed inflation targets in the past. But in any case, what I want to make clear here is that when you look at professional forecasters, their inflation expectations are anchored. And this is a good sign. It tells us that the rational players in the economy don't see any inflation on the horizon. And so their interpretation of what is happening with inflation now is that it's a transitory shock. And we should obviously be concerned about this, but it's not any synchronic anything that we had like anything else we had in the 1970s. Now, so this is one type of players, the rational players in the economy, but they are not the most important part of the economy. We have to think about the general public, the households, the firms, because these agents, these players make consumption decisions, they make employment decisions, they make pricing decisions. And so we have to understand their inflation expectations as well. And this is where since start get interesting, mainly because we see massive departures from full information, rational expectations at the micro level when you look at households and firms. They look different. And I would like to highlight four dimensions of those differences. The first one is that unlike professional forecasters, households and firms may have different interpretations of events that happen in their economy. You know, professional forecasters may think we have a demand-driven recession. Professional forecasters may think it's a demand-driven recession. Households may think it's a supply-driven recession. And this has important implications for policy. We can also have differences in sensitivity to changes in salient prices. So I'll show you some results. The professional forecasters are basically insensitive to changes in energy prices, so very insensitive. But you look at households and firms and these players are very, very sensitive. And again, this is important for current policy because as I'll show, there is big divergence and expectations for these different types of players. And we need to understand what we should do about this. And energy prices is going to be a key element in explaining these differences. Another point I want to make is that unlike professional forecasters and financial markets, households and firms tend to be relatively inattentive to monetary policy and fiscal policy. And again, the reason why it's important is because if you have concerns that all these policy interventions that we have seen in the recent years are going to directly trigger a rise in inflation expectations, you know, people will read the news and understand that, you know, this is happening, that kind of scenario seems unlikely because most people are unaware of monetary policy and fiscal policy, or at least they don't understand the implications of these policies. And finally, you know, you may say, well, you know, this inflation expectations of households and firms, maybe noisy and people not really use them for anything in their lives. That's not true. You can create exogenous variation in inflation expectations for these players. And A, you can document that you can change these expectations. And B, you can document that people act on these changes and expectations. They change their consumption, they change their portfolio choices, they change their behavior. And these changes are not necessarily consistent with what we have now mainstream models that rely on full information rational expectations. Let me now elaborate on each of these points and hopefully convince you that these are important dimensions and we should keep them in mind when we think about optimal policy and also how much policy can deliver to us in terms of macroeconomic stabilization. This figure here on the first point about differences interpretations shows you the dynamics of inflation expectations and now cost of GDP for professional forecasters in the US. So you start here, people expect 2% inflation and they expect some healthy growth in the US economy. Then as the COVID crisis unfolds, these expectations and perceptions start to be really, really going down. There is a major downgrade in the perceptions of the gross rate of GDP and also a big decline in expected inflation. And this is very much consistent with the view that the COVID-19 crisis was a demand-driven recession and there is some type of Phillips curve that gives us this movement of the variables. Now, on the other hand, you look at households and you see a very different picture. First, you start with a different level of inflation expectations. It's a separate story, but it's a fact that is interested to note. But more importantly for us is that as the crisis unfolds, people downgrade their perceptions of current business conditions, right? So we're going down, but instead of lowering inflation expectations as professional horror customers did, households are moving in this direction. They raise their inflation expectations, which tells us that the way they interpret the COVID-19 crisis is different. It's consistent with the inflationary view of inflation. They have high unemployment and we have high inflation. Now, this is not just about COVID crisis. You can look at other times and you're going to see more or less the same picture. For example, you look at the information that we have in the service professional horror customers in the US, remove house, professional horror customer, fixed effects, time-fixed effects. So all you should see is that if you pick a professional horror customer who projects above average inflation, somewhere here, this professional horror customer is also going to predict above average gross rate of output. So we have this positive relationship. And again, it's consistent with the view that recessions are mostly driven by demand shocks and there is some notion of the Phillips curve out there. On the other hand, you look at households. This is the Michigan Service of Consumers. Again, you do the same exercise. You take out fixed effects for households. You take fixed effects for time. You pick a household who predicts above average inflation somewhere here. This household is going to predict the low average gross rate of output. So again, this is telling you that households have this really strong staglational review of the world. They associate inflation with high unemployment, which is in sharp contrast with what we see for professional horror customers. Now, why is that? To answer that question, we need to see more data. First, this is something specific to the US. So this is a more general pattern. And this figure here shows you the results for professional horror customers, not just for the US, but also for the Euro-Aryan countries, Germany, Italy, Spain, and New Zealand here for your reference. And again, here in all countries, we see a strong positive relationship. So pretty much everywhere in the advanced world, we see that professional horror customers have the same pattern that was so for the US. There is a positive relationship between inflation and the gross rate of output. You look at the households, and what we saw for the US, this very sharp negative relationship, is also going to be true in the Euro-Aryan countries. And by the way, I would like to thank Dimitris Gergarakos for producing these figures for the Euro-Aryan countries. He used a new survey of households run by the ACB. And I think it's a terrific investment, just in general, in terms of measuring inflation expectations. And I think it's also great for science because we have now this consistent measure of inflation expectations across different countries. And it's a very rich source of information for research. Anyway, so this is professional forecasters, upward sloping, this is households, downward sloping. So people think inflation is associated with high unemployment. And this is very different for professional forecasters. Now, why is this important? Well, it's important because if we have high unemployment as we had during the COVID crisis, we can have elevated inflation expectations because people think it's a bad state of the world. As the economy improves, when we have a strong recovery and people have more employment, unemployment declines. We should also see that inflation expectations may be lower through this channel. So if the economy is on track, if we leave Delta variant behind us and the vaccinations proceed at the high rate and we control the spread of the COVID virus, it gives me reasons to think that this increase in inflationary expectations on the part of households and firms is going to be a transitory one. So that's an important dimension, let's skip this slide. Now then, high sensitivity to salient prices. Okay, so as I said, this is important because professional forecasters, they look at a lot of data. They try to aggregate this data in a consistent manner and households don't have these levels of sophistication. They tend to look at simple kind of rule of thumb prices, statistics that are going to inform them about the state of inflation in the economy. And energy prices, as I said, is one of the most important salient prices out there. This figure here shows the time series of expected inflation, the blue line in the US and the Michigan serve consumers. The black line is the price of gasoline, the US. And I keep showing this figure over and over in various conferences and meetings to highlight that inflation expectations in the US for households very much driven by the prices of gasoline. You know, it's true at low frequency and there are the cycles up and down, but also it's true at high frequencies. For example, there is this months when there is a spike in the price of gasoline. This is also the months when we have a huge spike, huge increase in expected inflation. So it's also true at high frequencies. And you know, this is historical relationship. You can look at more recent data. Again, there is a lot of movement here. You can look at the COVID crisis and you see that again, this doesn't prove that there is a causal relationship between the prices of gasoline and the expected inflation. But it's striking that this two cents are moving together so strongly, okay? So we see a big increase in expected inflation for households. You look at professional forecasters, there is barely any movement, barely any movement. Now, why is this important? Well, it's important because, you know, these guys are telling us the professional forecasters that we're not going to have a lot of inflation. The shock is transitory. Look at households and firms, they seem to have much higher inflation expectations and there is a big divergence between these expectations. Now, I believe that this is coming from the dynamics of energy prices. In the US, it may be the price of gasoline. In the US, I'm sorry, in Europe, it may be a different price, maybe a price of natural gas or utility bills or something else. But if this increases and energy prices are going to be very persistent, this may be a source of problems for us because this will be a force that can keep inflation expectations of households and firms at an elevated level for a long time. I know energy prices are not directly under the control of the central banks, but, you know, maybe other storages can do something about this. Well, there's also striking, as I said, that professional forecasters don't think anything is happening here in terms of inflation. We don't have a lot of inflation or pressure. And at least in the US, it's consistent with the idea that we have a lot of under-employment in the economy. You look at the employment to population ratio in the US now, it is roughly at the level that we had at the trove of the Great Recession. So with so much under-employment or non-employment, we likely have strong disinflationary pressures in the economy. So at least on the part of labor, the cost pressures are not necessarily very strong. And we see that wages, unlike inflationary expectations, are not going up at the rate that will be unsustainable or highly inflationary. In any case, what I want to highlight here is that we have those different players and the price of energy is a source of risk, because this is something that can elevate inflation expectations for households and firms for quite some time. And unless energy prices are worth back to levels where we saw before the COVID crisis, well, we can have an issue here. Now, another point I want to make is that there is a slow attention to monetary and fiscal policy. I will focus just on monetary policy for now, but the same is true for fiscal policy. And let me illustrate this point with two figures. We're in a survey of chief executives in the US and asked them to tell us about the inflation target of the Fed. We have 30 people who said, I don't know, then another 20% said something that is not interpretable. It was an open-ended question and only 20-25% of CEOs gave you the correct answer. Okay, so this is striking that some of the more informed players in the economy, the guys who set prices, managed lots of people, set wages in many ways. They tend to be relatively uninformed. You look at households and their picture is even more dim. Lots of people give you unrealistic responses. The Fed is not targeting 10% inflation or anything about 10%. Only a relatively small fraction of people gives you the correct answer. And you can interpret this as a sign of problems, but I think it's really a sign of success. The Fed and the ECB have been so successful in delivering low and stable inflation for so many years that most people don't care about inflation or the central bank. How much difference does it make for an average American or an average European where the inflation is 2% 2.1 or 0.9? For most people, it doesn't matter. But one important policy implication is that if we have this policy announcements about forward guidance, quantitative easing, we know we're going to have a strong reaction of rational players, professional, forward casters, financial markets. But for most households and firms, it's not going to ring the bell. And so they believe that this kind of policy announcements have the potential to trigger an increase in inflation expectations. I think it's an unlikely scenario. And it may be useful because by delivering low and stable inflation for so many years, the central bankers develop this credibility capital, the stress capital that they can spend now and help the economists rule this various tools without triggering inflation expectations. For me, this is a sign of inflation expectations being incurred, that is they're insensitive to changes in policy moves that may raise inflation expectations very, very strongly. Now, finally, how much difference does it make? Maybe what will happen is that households raise inflation expectations for themselves and this is going to create a boom in spending, which is going to create even more overheating in their economy. And so that's going to raise prices even further, raise inflation expectations and we can continue this loop and end up in a very bad scenario. Now, to answer that question, we can think of a sort of experiment here. We have consumers spending on the left-hand side and maybe portfolio choices, it may be any other choice made by households. Then you have inflation expectations of this edge and it's a bunch of controls in the narrative. And so in theory, it should be simply run a regression and this coefficient B is going to tell you how sensitive this choice here, consumers spending is going to be expected inflation, is going to be expected inflation. Obviously, the main issue here is in the donating. Expectations and choices formed simultaneously and we can run this regression but this coefficient is not going to mean anything for you. And this is where the science is going to be important and quite appropriately for this conference, bridging science and practice. You can run a randomized control trial, this is going to be your practice and can break this in the genetic issues and have some structural or causal interpretations of this coefficients B. How do you do this? Well, typically it involves several stages. You measure priors, you ask people what inflation is going to be. Then you're going to randomly assign them the households into treatment and control groups. People in the treatment group are going to receive some information. I'm going to give you some examples. You can measure presteriors, how people react to the provided expectations. This is going to tell us something about our ability to manage inflation expectations. And generally we see very strong responses to this information treatments. This tells us that management of inflation expectations is a potentially very powerful tool. Then you track those people over time. For example, you record their consumption and at some point you compare consumption choices for people in the control group and for people in the treatment group. And if there is any systematic difference, it has to be a causal effect. That is you manipulate it in a statistical sense. So you don't fool anybody. You just change the beliefs of those people. And you want to see if there is a change in the consumption response. There is a difference here. You know it has to come from here. And the only reason why we can see this kind of response is because there was a treatment applied to households at this stage. Now what kind of treatment do you can give? You can tell people about the inflation target and inflation forecast, forward guidance. You can do the same thing with fiscal policy. You can inform people about current deficits, future deficits, current level of debt. See how people react to this. And this is important especially if you believe in sense like fiscal seer of price level where we know that it is fiscal policies that drives inflation. So monetary policy directly. So I'll show you some examples of what will happen if we have this as a source of variation. Now, now this treatment should work. If you believe in full information rational expectations because information that is provided here is publicly available. You know, you go to the ECB website and you're going to see very clearly the inflation target. You can go to the FAD website and you're going to see that target very clearly. All of this information is within clicks away, right? So it's very easy to reach this information. But somehow people choose not to pay attention to this. And again, as I said, maybe it's a sign of success. In any case, let's go back to our sort experiment where we want to understand how consumption responds to changes in inflation expectations if we try to change this inflation expectations from forward guidance. So we'll tell people something about the future course of policy. They're going to revise their expected inflation. We want to see the change their consumer spending as a result of that. Now, this is what you'll find for the US but you can see something similar for, for example, for the Dutch households. You raise inflation expectations, you will stimulate people to buy more durable goods, food, a little bit of furniture, this kind of stuff. But this effect is going to be very shortly. Within a few months, it's gone. On the other hand, you raise inflation expectations and people are much more reluctant to purchase durable goods. Cars, houses, appliances, electronics. And the reason why it's important is if you have this kind of scenario, then remember I told you about the spiral that you raise inflation expectations, people are going to consume more, this is going to create overheating in the economy and that's going to raise inflation and then it's going to feed into higher inflation expectations. If you have this kind of response, it's hard to see that this cycle, this loop is going to be present and active in the economy. People withdraw their consumer spending and they take high inflation as a bad sign in the future. It's not the time to spend. It's time to set aside reserves of war chest that they can use in times of crisis when, for example, unemployment is low. And as a result, the policies aim to specifically raise inflation or inflation expectations can be far, at least on the part of households because they may interpret this future inflation in the wrong way. They can think it's a sign of bad sense to come rather than a sign of good sense to come. So we have to be careful there. And as I said, this is not just true for the US. You can also look at, for example, Dutch households and you're going to see something similar. Now, fiscal policy. Okay, so this is another important dimension. I know it's not directly under control of the ECB or the Fed, but this is something to keep in mind because the level of public debt has exploded in the recent years. And we have not seen this level, levels of debt since World War II. And we know historically, when we have this much debt, public debt, there is always the tension to inflate it away. And we know, for example, after World War II in the US, inflation was running really high for the first three years. So it is 3% cumulative way. And in part, this was motivated by the desire of the government to lower its debt burden by inflating the way it's debt. So this can happen. And obviously we have to be concerned that households don't look at these numbers and say, well, the debts are really, really high. And so we expect the Fed or the ECB to print money to pay for this. How can you establish this more formally? Again, going back to the sides. We run in our city. And we did this in the US in December, 2018. So it was very forward-looking experiment. We split households in the US into the control group and various treatment groups. These two treatment groups were told about the current deficit and also the current level of debt to see how people react to that. Also had treatment groups where we tell people about future level of government debt. And also inform people about the interest rate the government pays on its debt and also how much inflation we're going to have going forward. So then what we want to see is how people are spawned in terms of their inflation expectations through these interventions. So if you have, households could take that, which is true that US debt is high, but not as high as we have in reality and would surprise them in a negative way and tell them, actually it's a lot higher than you think. How would they respond to that? And what I'm striking is that if you tell people about the current level of debt and you look at their short-term inflation expectations one year ahead, the current fiscal variables are not really bothering people. They don't think that what is happening now is a sign of any kind of problem in terms of inflation next year. It's also true for longer inflation forecasts. So 10 year ahead horizon, telling people that we have a lot more debt now or we have really bigger deficits than you think, this is not raising their inflation expectations. But telling people that we will have a lot more debt in the future, this is what is raising their inflation expectations. And all this response is statistically significant in part because we have a huge sample so we can estimate this coefficients precisely. But in an economic terms, this number is not huge. For example, you take this as a base, roughly 4 percentage points here, you move it by 25 basis points. So there is a response, but it's not as big as predicted. For example, by the fiscal year of the price level. And so we shouldn't be worried that government debts projected in the future are going to be creating huge problems in terms of raising inflation expectations. They are raising inflation expectations, but not as much as one may think. And again, maybe this is a sign of success that the central banks are credible and the general public doesn't think that the government will use the printing press to pay for its debts. Now, where does this live in terms of firms? Firms are somewhere between households and firms and not as informed as professional forecasters, but they're not as ill informed as households that tend to have better forecasters more generally. Similar to households, they appear to be relatively inattentive to aggregate inflation monetary policy. Again, maybe it's a sign of success. Their responses to expectations can vary depending on the circumstances. For example, you can create scenarios where you tell people that we have forward guidance, we raise inflation expectations, and this is going to lead firms to raise prices as desired, we don't want to have a deflation. But that's more nuanced. And it seems that the biggest source of concerns when firms think about how they set prices is really cost pressures and what the competitors are doing. And at least in the US, it seems that there is a lot of slack in the labor market. So it's unlikely to come from employment, unemployment kind of trade-offs and inflation. But the deficits, we have shortages, we have for various parts in the economy. This may be something that we should keep an eye on because this may force firms to change their prices in a big way and kind of again, trigger this spiral of high inflation expectations and high inflation at the same time. So to conclude, back to the 70s, now when I look at my research over the last few years, I think it's unlikely. We have a lot of slack in the economy, a lot of under-employment in the US. Maybe it's to a smaller extent in Europe, but in the US it's very clear we have lots and lots of under-employment. The expectations of rational players are incurred, so those people don't see anything on the horizon. This is important because they set interest rates for households, for example, and this is an important propagation mechanism. For households, they have high inflation expectations, but this is mostly coming from energy prices to the extent the shocks are transitory. Maybe next year when we're going to have super expensive natural gas prices in Europe, maybe when I'm going to have very high prices of gasoline in the US next year. Also as the economy improves, these people will be less concerned about unemployment and so by extension, we'll be less concerned about high inflation because they think inflation and unemployment go hand in hand. Also if we are concerned that we have this super expansionary fiscal policy that this is something that can directly trigger inflation expectations that doesn't seem to be there and also the policy communication is different. Instead of saying, we need to raise inflation, it's more about we need to help the economy, we need to have full employment and I think it's very important in terms of conveying the right message to the people achieving the desired outcomes without triggering high inflation or high inflationary expectations. Obviously there are some risks, rising commodity prices, I already mentioned this, looming debt, that's a problem. And also this, you know, shortages that we see in manufacturing, that's a problem and I hope that we're going to resolve these issues. This issue and this issue to some extent and have a credible plan, how we're going to control the debt in the future. All of these steps can mitigate these risks and again result in a transitory blip in inflation rather than something systematic and chronic. Let me stop here and thank you again for having me at the conference. Thank you, thank you very much for this very encompassing and eye-opening, I would say presentation for us working in central banks.