Good afternoon everybody. For those of you who don't know me, I'm Alan Blinder. I'm ex everything except professor at Princeton, which I still am. At my age, you're ex more things than you actually are. It is our collective honor, Princeton and the Rabinowitz Center and the Griswold Center, to welcome here to speak to us today Dr. Adriana Kugler, a member of the board of governors of the Federal Reserve system. In return for that, I promise not to ask her about the next FOMC meeting. So thank you for coming. Adriana became a member of the board in September 2023, filling out a term. As those of you who know the Fed know, and those of you who don't probably don't know, all Fed terms end January 31st. So to fill out an unexpired term which lasts until January 31st, 2026, becoming by dint of that the first member of the board of Hispanic origins, which is kind of amazing when you think of it. The Federal Reserve started in 1914, it's been a long time, and good that we have you on the board. Prior to that, Adriana was serving as the US executive director at the World Bank. Her regular job, so to speak, when she's at the regular job, I'm going to mention something else in a minute, is as a professor at Georgetown, where I note that in addition to being a professor of public policy and economics, she was the vice provost at Georgetown for faculty, which sounds to me like a euphemism for herding cats, which you don't have to do in your current job. Prior to that, she was the chief economist at the US Department of Labor from 2011 to 2013. She's a research associate of the National Bureau of Economic Research and the Center for the Study of Poverty and Inequality at Stanford, and a whole bunch of other things that I'm not going to take our precious time to read to you, other than to note that she has a BA in economics and political science from McGill in Canada. We all know about Canada now. And a PhD in economics from the University of California at Berkeley. No Princeton degrees, but we're definitely not holding that against you. Thank you for coming here. And as you can see from the title, she's going to speak to us about inflationary expectations. The floor is yours. Thank you so much.
Well, thank you so much, Alan, for that generous introduction and thank you to the Griswold and Julie Rabinowitz Centers for the opportunity to speak to you today. As someone who has worked on both the public sector and academia, I certainly applaud the common purpose of our centers in connecting researchers, policy makers and the private sector to get a common purpose on policy ideas that serve the public good. To that end, I think there are very, very few individuals who have done more as a teacher, as a researcher, as a government official, and certainly as a public figure than Alan Blinder. That includes educating the public. In the spring of 2022, as many wondered whether the Russian invasion of Ukraine would add to the factors that already were driving up inflation, Professor Blinder wrote in the Wall Street Journal that a more important factor would probably be the public's expectation of future inflation. And as I will relate today in my remarks, he was of course absolutely right. As in the past, inflation expectations have played a crucial role in the course of inflations since the spring of 2022 and I expect they will be an important factor in the Federal Reserve's ongoing effort to achieve sustainable inflation of 2%. For that reason, I would like to focus on inflation expectations today before discussing my outlook for the US economy and the implications for appropriate monetary policy. First, I will describe inflation expectations within the conceptual framework that many economists use to connect inflation with the broader economic activity known as the Phillips curve. Second, I will discuss the central importance of the stability of these expectations which have come to be known as the anchor of inflation expectations. And third, I will explain how firms and households as well form their inflation expectations and how these expectations affect economic decision making. Throughout I will make some references to historical experiences in inflation but I will certainly focus on the period since the pandemic. Economists have long recognized the connection between inflation and overall macroeconomic conditions. But it was in trying to explain this empirical relationship that the importance of inflation expectations was actually revealed. The foundation of this work was laid by New Zealand economist AW Phillips, a fascinating figure who was among other things a mechanical genius who built an early economic model operating by hydraulics rather than electronics. In contemplating the mechanics of the economy in 1958, Phillips set out to explain why nominal wage growth was slower when unemployment was higher and faster when unemployment was low. His and other subsequent research showed that a crucial factor was the utilization of resources such as labor and capital. Generally when firms use labor and capital very intensively, production costs tend to rise and firms have more scope to pass those cost increases as higher prices which in turn pushes up inflation across the economy. In contrast, when the level of utilization is low, costs tend to rise more slowly or even fall and firms have less scope for raising prices and thus pushing down inflation. This trade-off has been known as the Phillips curve. In this simple form, this trade-off implies that governments can achieve and maintain very low unemployment only if they allow inflation to rise above a certain level. In the latter 1960s, Milton Friedman and Edmund Phelps asserted that this orderly trade-off was only temporary and would ultimately break down because of the role of expectations and in particular inflation expectations. To use an example, while current production costs are important to factory owners setting prices, the owner will also consider future production cost, future levels of demand, and expectations for inflation throughout the economy. Likewise, workers will factor expectations of future economic conditions into their pay demands and banks will consider future inflation in deciding loan rates. Consumers whose purchases constitute some two-thirds of economic activity make decisions about whether to purchase something today with an idea of what it will cost in the future. All these decisions are influenced by expectations and this is the way in which expectations may shape inflation today. In turn, when we think about the Phillips curve and its trade-off nowadays, we account for the importance of expectations of different individuals throughout the economy. There are different measures of inflation expectations. Some come from surveys polling business owners, others from consumers, and yet others estimating expectations among bond investors based on the differences in yields between nominal and inflation-indexed securities. While most of my points today apply broadly to all of the measures of expectations I mentioned, my examples will come mainly from surveys of consumers and businesses. While there are questions which I will address about how well these surveys measure inflation expectations, I really closely monitor all of these measures because they complement market-based indicators of future inflation that are affected by dynamics intrinsic to financial markets such as changes in risk tolerance. Let me note that in addition to the way expectations of future inflation influence prices in the near term, there are economic mechanisms that link current inflation with past inflation, such as those that set wages and the terms of rental contracts. In these cases, adjustments in these terms are often benchmarked on past inflation as for example when workers and landlords seek to recoup losses from increases in general prices. So to cite one example, as the economy reopened after the pandemic, workers sought higher wages to compensate for the earlier wave of inflation in food and core goods, thus pushing up inflation especially in the services sector where labor comprises the largest share of the cost in this sector. And because rental agreements typically last 12 months or more, as you know, these yearly rentals that one gets, landlords faced a lot in adjusting rents to reflect escalation of inflation after the pandemic and sought to recoup also those losses when renewing leases. By looking at price changes in this way in the rear view mirror, some decision makers in the economy end up making inflation more persistent. That is important for me to consider as an economic policy maker because I must pay attention both to expectations of future inflation as well as persistence of current inflation. When we speak of expectations of future inflation, it is crucial to define the time horizon and different surveys conducted by the Federal Reserve and others ask about inflation all the way from one year to as many as 10 years. Surveys with a shorter horizon, such as the Michigan Surveys of Consumers question on inflation one year ahead, are heavily influenced by current inflation. Near-term inflation expectations tend to be volatile. They move up and down when, for example, energy prices increase or down when energy or some other volatile set of prices also decreases. These expectations are important because many economic decisions such as major consumer purchases and hiring and investment as well for firms focus on horizons of only a few years ahead. By contrast, inflation expectations over longer horizons such as the Michigan Survey's question on inflation between five and 10 years ahead say less about current conditions than about the trend for inflation for some time in the future. You can think about these longer term expectations as much less affected by forces that push inflation up or down in the short term or what we economists call shocks. Longer term inflation expectations tend to be less volatile, affected less, for example, by what oil or food prices may have done lately, and more by the stability of inflation over several years or even decades ahead. I mention these different time horizons because they matter in my job as a central banker. Expectations a year from now reflect short-term shocks in the economy as well as ongoing efforts from monetary policy makers to bring the economy back to the longer run state. Thus, while short-term expectations may indicate whether inflation is expected to move towards its target, they're not the best measure of monetary policy credibility. So keep that in mind when I mention some of those measures later on. Longer-term inflation expectations on the other hand should be much less influenced by short-term shocks to the economy and a change in those expectations has implications for the Federal Reserve's prospects for meeting its price stability goal. When these longer term expectations are reasonably low and unresponsive to short-term developments, we say they're anchored. It is not clear who first coined this term, but Federal Reserve Chair Ben Bernanke in 2007 gave a speech on inflation expectations in which he described anchored inflation as relatively insensitive to incoming data. So, how should we think about the process of anchoring and de-anchoring of inflation expectations? The dynamics of short and long-term inflation expectations shed light on this issue. If the public experiences a spell of inflation higher than their shorter run expectations, it will revise up the short-term run expectations to ensure that their near-term plans account for the changing economic environment. That's what happened after the pandemic when inflation based on personal consumption expenditures rose to a peak of 7.2% and one-year expectations rose to more than 5%. But longer term inflation expectations remained very well anchored with values ranging along those seen unobserved since 1995. I would contrast this experience with the previous bout of high inflation in the US from the 1970s to the early 1980s. Among other issues such as high energy prices and accommodative monetary policy, rising inflation and inflation expectations fed a cycle of escalating inflationary pressures. Inflation was high and very volatile over this period and that is reflected in the shorter and longer term inflation expectations that were also high and volatile during that period. Another important difference between these two episodes has to do with the performance of the Federal Reserve as opposed to the late 1960s and most of the 70s. Most recently, the Fed acted aggressively to tighten monetary policy, raising the federal funds rate more rapidly than in previous cycles and lowering inflation also more quickly than ever before. This came after 30 years of success in keeping inflation in check and the credibility earned by the Fed's inflation discipline surely helped keeping long-term expectations stable and two people here contributed to that. This shows that an important role of the central bank is to convince the public through actions and communications about its intention to shape economic conditions and to use its policy tools to bring inflation back to its target. By committing to keep inflation low in the future, central banks seek to influence expectations of future inflation, which in turn influence conditions now and also over time. The Fed's credibility in keeping inflation low and stable won over many decades kept longer-term inflation expectations stable and that contributed significantly to the Fed's success in reducing inflation while keeping the labor market also strong. Those are some of the basics about inflation expectations and how they influence the economy and the conduct of monetary policy. So next let me turn to what I want to say about the patterns we see in the survey measures of inflation expectations. What influences expectations and how inflation expectations are used by the public in their decision making. Fortunately, there is a rich body of economic research that has shed light on some of these questions and I will focus on the evidence for households and for firms. We can then take care of some lessons from these empirical patterns for monetary policy. So one important observation is that both short and long-term inflation expectations are often notably higher than actual inflation even after a period of very low inflation. That is, there is evidence that survey respondents often believe the inflation they have experienced is higher than it actually is. Another pattern is that there is a wide dispersion of views about both short-term and longer term inflation expectations reflecting obviously at least in part the dispersion of inflation in the consumer baskets of goods and services that are purchased by different people. Research also finds that some groups such as women and lower income households tend to have systematically higher inflation expectations. In addition to these variations in expectations, there is high uncertainty in forecasts of future inflation. When people are asked to assign probabilities to different forecasts of inflation, surveys report wide distributions in the likelihood of one versus another outcome. Finally, short-term inflation expectations tend to be correlated with both recently realized inflation as well as perceptions about recent inflation. These patterns tell policy makers that inflation expectations of households and firms are diffused and likely harder to influence by monetary policy relative to financial market participants and professional forecasters who follow news more closely. Still, expectations from business owners and workers ultimately inform firms' pricing decisions and costs and may even be more relevant for inflation outcomes. And thus, it is very important for policy makers to communicate clearly with the public about our intentions to bring inflation back to our target. So because inflation expectations are diffused and heavily influenced by recent experience, let's consider now the reasons for the dispersion in these expectations. Unsurprisingly, it starts with the considerable variation in the sources that the public uses to collect information about inflation. Households report that their main source of information are their own shopping experiences, including regular purchases such as groceries and gasoline. And these price changes in those goods and services are what affect inflation expectations. Also, it seems like inflation expectations of homeowners tend to respond to changes in mortgage rates because homeowners have more of an incentive to track changes in rates that might affect, for example, their prospects for loan refinancing. Another important source of information is energy bills with evidence also pointing to household inflation expectations being more sensitive to energy prices when inflation is higher. More generally, consumers and firms seem to pay more attention to news related to inflation when inflation is high. And this has been found as well in many other countries. While the unique experiences of survey respondents matter, these points point to inflation expectations being dependent on the state of the economy. Thus, we policy makers should account for different economic conditions when assessing the risks of a de-anchoring of inflation expectations. For instance, with fresh memories of the post-pandemic inflation and with recent surges in prices of some food items regularly purchased, inflation expectations of workers and firms may now be more sensitive to anticipated future price increases relative to the pre-pandemic period. Let me now turn to how households and businesses employ their information, their inflation expectations in their economic decision making with much of the evidence consistent with what would be expected given longstanding economic theory which is good for us economists to know. Starting with households, in addition to any influence on wages from past inflation, expectations of future inflation help shape demands for pay raises. Workers care about their inflation adjusted wages rather than nominal wages. And we see a positive correlation between inflation expectations from consumers and wage growth with a close co-movement during the recent inflationary bout. A complementary decision for the worker is to look for a new job that pays more, especially if the person envisions a low probability of getting a raise in the current job or the raise would likely not fully cover losses in real incomes from inflation. Indeed, measures of general wage growth are more sluggish relative to those of job switchers. Moreover, researchers also find evidence of higher job-to-job transitions for workers who have higher inflation expectations. So inflation expectations of workers are an important influence on nominal wage growth and an important indicator of inflationary pressures for us policy makers especially in terms of services inflation. Now let's consider how these expectations influence firms' decision making. As I discussed in the context of the Phillips curve, firms with higher inflation expectations would be expected to increase prices more and indeed researchers find causal evidence of this. During the recent period of high inflation, the fact that business owners' short-term expectations about cost or input prices rose only modestly and soon returned to levels close to 2% may suggest that firms' inflation expectations may not have been a strong source of inflationary pressures. Still, researchers at the Richmond Fed also found that during this period, business leaders incorporated more information about aggregate inflation measures in their own pricing decisions compared to times before the pandemic inflation surge. While researchers also find that business leaders pay less attention to inflation as it came down. This evidence points to the inflation expectations of businesses being sensitive to underlying inflationary dynamics and monetary policy makers should certainly remain attentive to this and I'll turn back to this in a minute. Now let me turn to the recent developments in inflation expectations, the current US economic outlook and their implications for monetary policy. In recent months, we have seen several measures of inflation expectations increase with both consumers and businesses reporting new and proposed tariffs as an important reason. So, the word of the day, tariffs. Among surveys looking one year ahead, there have been notable increases in the Michigan surveys of consumers, the Conference Board survey of consumers, the Atlanta Fed survey of businesses, the Philly Fed survey of professional forecasters, and the New York Fed consumer survey. For instance, this past Friday's release of longer-term inflation expectations from the Michigan survey was the highest since February of 1993. Additionally, the recent spike in short-term inflation expectations appear to be mostly anticipatory as one can infer from the divergence between falling inflation perceptions, what consumers think price increases have been in the past year, and climbing short run inflation expectations. This anticipatory nature of the recent increase in short run expectations may allow for price pressures through a second channel. That is, businesses may feel a greater ability to pass along higher cost to consumers when they come from external factors out of their own control. Indeed, firms are already reporting not only higher costs but also expectations of higher cost according to some surveys such as the one conducted by the Atlanta Fed along with other manufacturing surveys as well. For now, I take some comfort from the much smaller increases in longer term expectations as measured by the Philly Fed survey of professional forecasters as well as the stability of longer-term measures on what we call inflation compensation which is based on yields from nominal and inflation-indexed treasury securities. As in past episodes when inflation expectations increased, uncertainty about future inflation seems to have also gone up as measured by the disagreement between the 75th and 25th percentiles of the distribution of individual respondents from the Michigan survey. Simultaneously in recent months, we have also seen measures of economic policy uncertainty increase and there is evidence that policy uncertainty and inflation uncertainty correlate with each other. One possibility is that policy uncertainty may be contributing to the rise in inflation expectations as well as to uncertainty about future inflation. Still it is hard to say at this point and I will certainly keep monitoring these developments very closely. So let me now turn from developments in expectations to realized inflation. After the substantial decline in inflation from its peak in 2022, recent disinflation has been slower and the latest data indicates that progress towards the Fed's 2% goal may have actually stalled. Core PCE inflation was 2.8% in the 12 months ending in February, which puts us back at the same level as we saw in the last quarter of 2024. The best news from February came from housing services inflation, which has come down steadily for at least a year to a 12 month rate of 4.3% coming down from 8.3% in April of 2023, but it's still quite a bit above what it was pre-pandemic at 2.5%. For the rest of the inflation categories, the news has been less positive. Core goods inflation, which had been negative for most of 2024, increased to 4% relative to a year before. February likely also marked an upward shift in market-based services inflation. While I do not discount price pressures in non-market services which do remain elevated upwards of 5%, the acceleration in market-based services in February from an estimated 3.1% to 3.5% is also not welcome given that this category often provides a better signal of inflationary pressures across all services. On the other side of the FOMC's dual mandate, employment continues to grow at a moderate pace and the overall labor market has remained resilient through February. The net 151,000 jobs added last month was not too far from the 177,000 average of the previous six months. And while the unemployment rate ticked up to 4.1% and labor force participation moved down to 62.4%, other labor market indicators suggest continued moderation in the labor market, but not significant weakening. Given the recent lack of progress on inflation, recent increases in inflation expectations and upside risks associated with announced and prospective policy changes, I strongly supported the FOMC's decision at our March meeting to maintain the target range for the federal funds rate at 4.25% to 4.5%. I will support maintaining the current policy rate for as long as these upside risks to inflation continue while economic activity and employment remain stable going forward. As always, I will continue to carefully assess incoming data, the evolving outlook, and any changes as we speak to the balance of risks that we face. So, thank you very much.
So thank you for the very interesting talk and especially for the reminder which a lot of us professional economists need, that expected inflation is not just one symbol in an equation implying by dint of that that it's the same for everybody, right? That it's not the same for everybody and that it moves for a variety of reasons which are also not the same for everybody. I mean, it's a thought that probably we should have more in our economic modeling than we actually did. So, thank you for that. And in line with that, I want to ask you about a group that you didn't
Mention, people on the Federal Reserve who don't always have the same expectation. And in particular, to sharpen the question just slightly, you started your talk by going back to Phillips, the Phillips curve. Many of us, I'll include myself in this, but it's not an idiosyncratic view, think the Phillips curve kind of disappeared from the face of the earth with the turn of the century or about that. So the question – you don't have to agree or disagree with that, but you may if you wish to – the question is about forecasting without a reliable Phillips curve, because the Federal Reserve is forced to forecast expected inflation to do monetary policy.
So yeah, that is something that some have questioned. I have to say as you know, back in 2016, Cher Yellen gave a speech on the expanded Phillips curve. And just literally a few weeks ago at the Wington Lecture at Georgetown – where I'm from – I gave a lecture on an expanded Phillips curve that includes the B to ratio, the vacancies to unemployment ratio, instead of the unemployment gap, that includes also supply side disruptions and supply side uncertainty. And that seemed to perform actually quite well. And honestly, the staff at the Fed does a really good job working on this expanded Phillips curve which seems to work quite well. Now you may be on the steep part of the Phillips curve, you may be on the flatter part. It's not uniform – this relationship does change over time. As you know, between the two recessions that Phillips curve seemed to be flatter, now we seem to have been on the steeper part. But I think it works, and used appropriately and expanded in the right way with the right data, it actually tells us something about the relationship between inflation and unemployment or slack. As I said, I replace the unemployment gap with the V to U ratio, and that measure of slack seems to perform a lot better.
You feel it's a tolerable – there's no perfect forecaster – but it's a tolerable, in the sense of usable, tool for forecasting?
Absolutely. It is informational.
On the difference, you emphasized at several points the difference between the Michigan survey, which is of ordinary people, and the market expectation from the implied breakeven rate from the TIPS market. Lately they're very different. I wonder if you would say something about what you think is behind that difference, and in particular, is some of it tariffs – to use the T-word again, which you used before? I'm thinking in terms of the ordinary people asked about in Michigan that now tariffs are coming.
Yeah, it's very interesting to see the responses to these surveys, both from consumers and also from businesses. In these surveys that are not prompted, they're volunteer answers, and about 50% of those in the Michigan survey said these higher expectations of inflation came from tariffs. In the business surveys as well – the ISM surveys, for example, the Institute of Supply Management, or the Dallas Fed survey of businesses – they also prompt this answer about potential increase in prices moving forward being related to tariffs. So that comes across in a range of surveys, not one but several. And it comes without the survey prompting for this answer, but it's volunteered information by those who participate in the survey. It seems to be related to what may be coming. That's why I said there seems to be an anticipatory reaction. But by the same token, if you look at the ISM, especially the manufacturing survey, that showed there are increases in costs already. So input prices, material prices, right? Those are already being felt by many businesses, and that in itself is increasing prices. The other part may be anticipatory.
So why aren't Wall Street traders reflecting that in the TIPS market?
I think there are many other factors that may affect the pricing of securities and raise premia, and they may be thinking even further ahead to what may be coming in terms of economic growth. So different factors affect the two measures.
So there's a broader phenomenon of which this may be a case, which people talk about more on the real side of the economy than on the inflation side, but I think this is an example where it's on the inflation side: developments are showing up in what is often called soft data – a lot of it surveys – without showing up in hard data. The inflation rate is an example of that. It has leveled off, not gone up, unless you want to count a tenth. But there's an expectation shown in some survey data that it will go up. And you see a lot of this in surveys of future real economic activity. I'm wondering how you square that as a forecaster.
I know people like to talk about soft data and hard data. I myself like to talk about traditional and non-traditional data. I don't think of it as necessarily being soft. I gave a speech about it less than a year ago at NABE, the National Association of Business Economists, talking about these two sources of data. I use both all the way. I count on the official data, the traditional data, which is usually available at much lower frequency and is highly vetted, tends to be representative. Non-traditional data, on the other hand, tends to come from businesses, from the private sector, from users. It's unlikely to be representative, but it's usually available in real time and at higher frequency, which is why I like to use it as well – because we know what's happening today, not what happened one month ago, one quarter ago, or even sometimes one year ago when you count on traditional data. So I like to use them combined. In terms of prices, I have to say that's one area where the two coincide. You're hearing from manufacturing surveys – the PMI – that purchases are more costly, costs are already going up, they're already charging higher prices. And then in the official data, it's true that overall headline core you may not see that yet, but goods inflation on the other hand really went up. It used to be minus 0.4% a year ago, it's now 0.4%. So it has substantially changed. And again, I mentioned market-based services inflation – that's another area where we're starting to see a reversal. So I'm paying attention to both soft data and hard data. I agree with you, there are other areas where we don't know as much. For example, in terms of real activity, we get these non-traditional data on retail sales. We had seen a bit of a softening, a slowdown. And yet, last night, we got data on auto sales and that showed a spike again, likely anticipatory to those tariffs that were announced today of 25%. So people rushed to the showrooms and got their cars before facing that. It could be that by the time we get our first quarter GDP data, we'll see a combination of some front-loading but then some reflection of caution from both households and firms in terms of investment and consumption, because of the uncertainty of what may be coming ahead. That is harder to read at this point, harder to make out clearly a way in which the non-traditional data can tell us about what we're likely to see for the first quarter – most likely some softening relative to 2.5% at the end of last year. But how soft, given that you also see front-loading by some households, is less clear.
You emphasize the advantage of non-traditional that you get data sooner. Presumably a trade-off between accuracy and speed.
Absolutely. That's exactly right, Alan. And the other thing is, with some of these non-traditional data, you don't have the historical trends. You cannot compare all the way back seeing how things behaved back then. So there is certainly a trade-off.
I want to shift gears only slightly to the coming, or already here, supply shocks that we're getting from tariff policy and immigration, which is affecting the labor force. A lot of the underlying growth in the last couple of years on the supply side of the economy, potential GDP, has come from immigration, both legal and illegal. That's almost certainly going to slow down quite a bit, probably already has – there are data lags. The tariffs are going way higher. I didn't hear what our dear president had to say in the last few minutes, but at last check he was going above McKinley, above Spoon and Holly in tariffs. So these are both potentially sizable supply shocks – stagflationary shocks when you put them together. There has developed over the years a – calling it a tradition is too strong – something like central banks supposed to look through supply shocks on the grounds that these are things that happen to you unhappily often and then go away, so you don't want to react by tightening monetary policy just in time for them to go away, and cause a deeper slump than necessary. I'm wondering how you're thinking about this received wisdom and applying it to what's going on prospectively very soon.
