Skip to main content

tv   Former Fed Chairs Yellen Bernanke Others Discuss Inflation  CSPAN  October 3, 2019 9:45am-1:30pm EDT

9:45 am
washington, d.c. and around the country so you can make up your own mind. created by cable in 1979. this, thi c-span is brought to you by your local cable provider. c-span, your unfiltered view of government. former federal reserve chairs this morning will discuss the impact of inflation on monetary policy. the brookings institution is hosting this event in washington, d.c. >> good morning. i'd like to welcome you to the event today. the mission of our center is to improve the quality of both monetary and fiscal policies and in particular the public understanding of them. under the -- under the very able leadership of david wesle, we've worked to promote analytical rigger but also to make complex
9:46 am
questions relevant. today one of the most vexing questions for people who participant in markets, but that's a different part of the world, is that inflation has refused to rise, much less persist, at the 2% targets. key question is, why? while history will be the judge once the data has been interrogated, we don't have time for that, because today's policymakers and investors need to act and to do so require at least a rudimentary answer to this puzzle. today we have gathered some of the leading experts in the world. inflation didn't fall as much during the great recession as the models anticipated and
9:47 am
hasn't risen as much lately. is it dead or just dormant or has the death been much exaggerated. in my experience, wiped capitulation is the first sign of the bottom. that's kind of where we are right now. has monetary policy been so well anchored. have globalization and technology made raising prizes difficult? has the economy evolved in ways the models simply haven't been incorporated? and are we at an inflection point in a transition in ways we cannot yet do not know how to
9:48 am
measure. and finally should the fed incorporate answers to these questions as it pursues full employment and price stability. we'll start today with janet yellen. janet will set the stage and then as you can see in your agenda, i think everybody should have an agenda, we'll focus on a few of the leading explanations for the unanticipated behavior of inflation. after a distinguished career at the fed, janet is now a distinguished fellow here while many of us thought she -- the feds loss is our gain. janet, it is yours. [ applause ] >> thanks so much glen. it's my pleasure to join glen in welcoming you to the conference on inflation. my objective is to prepare the ground for today's parresentersy
9:49 am
presenting the puzzle we'll be discussing and highlighting why the answers matter for economic performance. the main puzzle pertaining to inflation is aptly summed up by the title of this conference, what's not up with inflation? inflation hasn't moved up through an expansion that now ranks as the nation's longest on record. unemployment declined from about 10% in 2009 to 3.7% today, a 50-year low. yet headline inflation over the last year according to the federal reverses preferred personal consumption expenditures or pec measure stood at only 1.4% in august. core inflation which excludes
9:50 am
food and emergency prices came in higher. but the current pace is close to that in 2009. contrast this experience with that during the long expansion of the 1960s, unemployment declined from 1961 to 3.6% in 1969. over the same period, headline pce inflation rose from just under 1% to roughly 5% and core inflation increased by a similar amount. the phillips curve has long served as the workhorse model of inflation, and it's used by most economists, including federal reserve staff, to analyze and forecast inflation trends. it posits that current inflation depends on the degree of labor market slack, on lagged inflation, and on a variety of supply shocks, including those
9:51 am
affecting the prices of food, energy, and other commodities and the value of the dollar, which affects import prices. various transitory factors, or noise, also affect inflation readings. the comparison i just described concerning the behavior of inflation as unemployment declined during the current expansion and that of the 1960s, illustrates two robust empirical findings. first, the slope of the phillips curve, a measure of the responsiveness of inflation to a decline in labor market slack, has diminished very significantly since the 1960s. in other words, the phillips curve appears to have become quite flat. and second, inflation has become much less persistent because the impact of lagged inflation on current inflation has declined considerably. this likely reflects the fact
9:52 am
that inflation expectations, which affect wage and price setting, now appear to be quite stable and unresponsive to variations in actual inflation. inflation expectations, in other words, have become well anchored. during the 1960s and 1970s, in contrast, a rise in actual inflation appeared to boost inflation expectations considerably. high responsiveness of inflation expectations to actual inflation works to boost the longer term responsive inflation to changes in labor market slack considerably. in fact, it creates the possibility that a temporary decline in unemployment below the natural rate where a transitory supply shock results in an increase in the rate of inflation, a so-called
9:53 am
acceleration of the phillips curve. an important question is just why the relationship between inflation and unemployment has become so attenuated. in the view of some observers, almost nonexistent. our presenters today will consider some possibilities. one possibility is that labor market slack is not appropriately measured by the civilian unemployment rate. perhaps broader measures of slack, including, for example, individuals involuntarily working part time, were some who were considered to be out of the labor force entirely, are rel vantd to wage and price inflation. the willingness of individuals considered to be out of the labor force to enter in response to strong opportunities may also restrain inflation. prime age labor force participation has declined considerably in the united states. it moved up during the
9:54 am
expansion, but even now remains roughly half a percent below its level just prior to the great recession. another possibility is that the flatness of the phillips curve reflects downward rigidity of nominal wages. firms are generally opposed to impose wage cuts on their employees. during the great recession, wage inflation fell very little. perhaps for this reason, both wage and price inflation were restrained during the recovery. the flatness of the phillips curve might also reflect global influences with trade slack and foreign economies in the prevalence of global supply chains attenuating the influence of domestic factors on inflation. conceivably, the flattening of the phillips curve reflects technological or market structure factors, such as the ease of price comparisons over
9:55 am
the internet resulting in declining pricing power for domestic firms. it's also possible that the flatness is a statistical mirage. the success of policymakers in achieving stable inflation, when supply shocks are pushing inflation up, could be masking a true relationship between inflation and unemployment that is stronger than it appears to e con ma trigss. finally, the success of policymakers in holding inflation low and stable in recent decades could explain the decline in persistence of inflation, which in effect flattens the slope of the longer run phillips curve. longer term inflation expectations, as measured, for example, by the survey of professional forecasters, have been remarkably stable in the vicinity of 2%. such well-anchored inflation
9:56 am
expectations may have been fostered by the fed's adoption of a numerical inflation objective of 2% in 2012, preceded by a long period during which the fed was committed to first bringing inflation down from unacceptably high levels and then maintaining inflation in the general vicinity of 2%. well, what difference does it make from a policy perspective if the phillips curve is flattened and if inflation has become less persistent? if the phillips curve is very flat and if inflation expectations are insensitive to fluctuations and actual inflation, the federal reserve may be able to run the economy high, yielding significant benefits to workers while imposing only minimal costs on society in terms of higher inflation. indeed, given that inflation has been so very low for so long,
9:57 am
it's averaged only 1.5% over the last decade. it may be necessary to have a hot labor market for inflation to move back to 2% on a stable basis. allowing the labor market to run hot could bring substantial benefits. as brookings economist arthur oaken observed in 1973 in an early volume of brookings papers on economic activity, a high pressure economy improves upward mobility, and we're seeing that in the current expansion. those who were leased advantaged in the labor market are experiencing the largest gains and wages in declines in unemployment. moreover, when firms find it hard to fire, they tend to lower qualifications and provide more training. we're seeing more partnerships
9:58 am
with community colleges and local governments to develop job market programs with promised employment to those who successfully complete them. in slack labor markets, those who check the box indicating a criminal record have little chance to find work. their resumes quickly end up in the circular file. in today's strong labor market, that's changed somewhat. such individuals are getting a chance to turn around their lives. it's possibility, albeit uncertain, that the skills and experiences these individuals are acquiring in today's tight labor market will yield longer-run benefits, enabling them to do better, even if the economy experiences another downturn. a significant benefit of better-anchored inflatiexpectat is that they enable the federal
9:59 am
reserve to all but ignore the inflationary impact of supply shocks in their conduct of monetary policy, mitigating the need for monetary policy to tighten when negative impacts on employment following adverse supply shocks. well-anchored inflation expectations, in effect, mitigate what might otherwise be painful conflicts between the fed's inflation unemployment objectives. indeed, over the last decade, u.s. monetary policy has barely responded to movements in inflation, driven by oil prices, the dollar, and other supply factors. for example, after a multiyear sequence of unanticipated oil price increases between 2004 and 2008, inflation reverted to its previous trend once oil prices
10:00 am
stabilized. similarly, inflation moved back up to 2% in 2018 after several years in which declines in oil prices and a marked appreciation of the dollar pushed inflation below the fed's target. in contrast, in 1973 and again in 1979, oil prices jumped markedly, producing sharp inflation spikes. in those episodes, inflation appeared to move permanently higher, even after oil prices settled down. it's important to point out that a flat phillips curve has a downside, which is that it raises the so-called sacrifice ratio. the sacrifice ratio measures the cost in terms of higher unemployment to lower inflation, should it rise too high. with flat phillips curve, it's necessary for monetary policy to create a good deal of slack in the labor market to return
10:01 am
inflation to levels consistent with price stability. and if the flatness of the phillips curve is a statistical mirage rather than a true change in the economy's structure, an attempt by policymakers to exploit it could push inflation up much more than they expect. finally, even if the phillips curve is quite flat over some range, it's conceivable that it could become a lot steeper if unemployment is pushed to very low levels. that is, it may be nonlinear at very low unemployment. there is some evidence of such nonlinearity, so it's a significant policy concern. another policy consideration pertains to inflation persistence. can we count on such low persistence going forward? or might inflation revert to its earlier behavior if policymakers
10:02 am
attempt to exploit the current benign inflation dynamics to keep unemployment very low or push it yet lower. if the decline in inflation persistence reflects increased credibility of monetary policy, this valuable asset could be squandered if monetary policy consistently allows inflation to deviate from its 2% objective. as in the 1970s, a willingness of the fed to tolerate inflation persistently above 2% could ultimately deanchor and boost inflation expectations. alternatively, and arguably the more relevant consideration today, a chronic failure of the fed to stably achieve its 2% target could deanchor inflation expectations on the downside, exacerbating the problems
10:03 am
associated with the effective lower bound on policy rates and creating significant deflationary risks. well, to conclude, our panelists will address questions that are vitally important for economic policy and performance in the years ahead, and i look forward to today's discussions. [ applause ] >> thank you, janet. thank you for joining us here this morning. so i'm pleased to present -- to start with the two presenters for our first panel on what's not up with inflation, and that's is it's all about monetary policy and it's all about inflation expectations. so we have fortunate to have two presenters first, professor of
10:04 am
economics. second, we have michael weber, associate professor at university of chicago, who has done very innovative, ill luminous research on how households form inflation expectations. without further, i'll have them start. >> thank you for having me. as janet yellen just said, in the past few years, several academics have observed that the phillips curve, the relation between inflation and measures of slack, has flattened or even disappeared. some studies even highlight that inflation appears to follow a process totally unrelated to
10:05 am
slack, such as employment. this led some academics to argue that the disconnect between inflation and measures of slack forces a big challenge to the monetary policy framework currently used by major central banks like the fed, which uses the phillips curve as a key building block. so in my work at the bank of england, we asked the question, what's the disconnect between inflation and measures of slack really posing a challenge to the monetary policy framework? our answer is no. on the contrary, the disconnect between inflation and unemployment or the output gap is actually what a new framework like the one used in central banks with the phillips curve and a central bank that has a dual mandate would actually predict. let me try to make my argument
10:06 am
in a simple version of the framework using the central banks. so a central banker has a dual mandate with achieving an inflation target and full employment or output at its potential. and this is represented here by this simple formula that says the central banker minimizes the deviations of inflation from its target, denoted here by pi-t, and deviations of its potential, x-t. the lam da letter here captures the relative way the central bank puts on output stabilization vis-a-vis the inflation target. so the central banker minimize deviations subject to a structural curve or aggregate constraint for the economy. the supply constraint is very
10:07 am
intuitive. it says when output increases relative to its potential, the gap increases. that puts upward pressure on wages and costs, cost in prices and inflation and causes inflation to increase. this phillips curve in blue here can also be shifted around by cost shocks, such as oil shocks or changes in caps that are collected in the term ut. and it can also be affected by changes in expectations of future inflation, represented in the first curve. now, this illusion to this optimization problem by the central bank is a targeting rule depicted in red, which is also very intuitive. it says when the central bank sees that inflation is set to rise above its target, she will
10:08 am
withdraw stimulus from the economy, causing the output gap to fall. this targeting rule imparts a negative relation between inflation and the output gap. so i plotted the two curves in this simple diagram. we see the upward slope in blue. and the downward slope in targeting rule depicted in red. now, in eke we liwe don't direc the targeting rule. we call it classical identification problem. this identification problem is made more difficult by the fact that the phillips curve can be shifted around by cost per shocks or changes in inflation expectations. so here an observer trying to
10:09 am
run a regression of inflation on the output gap will actually trace a wrong slope, will only see the response of monetary policy to cost per shock. so we'll be identifying the targeting rule rather than the phillips curve. so two lessons from here. if monetary policy successfully offsets all dmantemand shocks, all that's left is cost per shocks, which will trace the wrong slope. the second lesson is that in equilibrium here, inflation will inherit all the properties of the cost per shock, so effectively we'll see inflation following this process with the persistence of this process, but we won't see the correlation with slack that we would expect from a phillips curve. just a couple more remarks on
10:10 am
this. the model i just presented you has a very healthy phillips curve, by assumption, yet we cannot see an eke we librium. the argument i'm making here is likely different from the one on anchored inflation expectations reducing or weakening the reduced form correlation between inflation and slack. my argument says that independently of inflation expectations, the structural relation between inflation and slack will always be masked by the conduct of monetary policy. now, of course as janet yellen just said, if monetary policy affects inflation expectations, leading to an anchoring, that effect will also add or go in the same direction. so these two effects are in some sense complementary. let me go back to the
10:11 am
identification problem now. how can we see this phillips curve? so ideally, what we would like is to have a very stable phillips curve so it doesn't move around and have deviations from the optimal targeting rule, some shocks to that optimal targeting rule. in that case, if we have demand shocks that move us along the phillips curve that are not offset by monetary policy, we'll be able to trace a right slope. this is easy to see in this simple model, but it goes through in a more complex quantitative setting that central banks are using as well. so just to summarize, in order to actually see the phillips curve in the data, we need to control for supply shocks. that's not simple and it's not sufficient. many shocks with the economy are
10:12 am
part supply and part demand. and it's not sufficient because we need these demand shocks to move us along the supply relation or aggregate phillips curve and not offset by monetary policy or other policies. now, in practice economies, economists have used monetary policy shocks, deviations from optimal policy due to mistakes in forecasts or judgment, and that's allowed them to identify the phillips curve. now, interestingly, those monetary policy shocks, or fortunately, i should say, have become very small and infrequent since the '90s. so it's much more difficult to use them to identify the phillips curve. what we use in my work is desegregated data at the regional level in order to identify the phillips curve. the idea here is that monetary policy offsets aggregate demand shocks, shocks to the whole country, but does not offset
10:13 am
regional demand shocks. so from the perspective of a region, monetary policy is suboptimal, and the region might have an open output gap or higher unemployment than the average. so we use this idea to identify the slope of the phillips curve. here i'm showing the outcomes of regressions of u.s. metro area inflation on slack from 1990s onwards. this is a period of a very flat aggregate phillips curve. each regression here regresses core cpi inflation on the unemployment rate. so in unemployment space, the slope is negative. we control for inflation expectations and we control for lags in inflation.
10:14 am
now, if we don't use any other controls, the slope will be similar to the one we obtained with aggregate data. it speaks of a very flat phillips curve with a coefficient of 0.15. i would like you to focus your attention on the last column. here we control for time effect. so controlling for any aggregate supply or demand effect, including the effect of monetary policy at the aggregate level, and we also control for timing variant, regional specific effects. when we do that, the slope coefficient increases to nearly 0.4. so this is a very healthy phillips curve, type of numbers we tend to put in our models. and it's in line with intuition or the insight we get from the model that's at the aggregate level. it's very hard to see the
10:15 am
phillips curve because monetary policy is upsetting demand shocks. but we can see that by looking at regional, desegregated data for which monetary policy is not offsetting demand shocks. so i leave you with my summary. i'd like to just say a couple remarks. the reduce from correlation, or phillips curve, is a mix of supply and demand factors. monetary policy is one of them. for individual episodes of missing inflation or disinflation, such as the great recession, the theory and data point to a variety of factors. on the supply side, large cost per shocks, such as what we saw during the recession, energy prices were going up, productivity fell, and there was a large financial shock. all of them contributed to higher costs and also press down
10:16 am
on activity. as janet also mentioned, there are arguments about the measurement of slack, and many people have argued that other measures of slack, such as short-term unemployment, for example, didn't go up as much as total unemployment during the great recession. so maybe that one was overestimated at the time. i conclude here and i'm happy to take questions. thank you. [ applause ] >> thanks a lot for putting together this amazing event and having me today. what i want to start out with is what was alluded to.
10:17 am
typically when we do models for policy recommendation and policy analysis, we would assume there's a representative agent that has all available information. then whenever there's any new information, immediately update expectations and accordingly change savings and consumption plans. to get an idea on this paradigm, given we see large deviations often times in data, it's maybe a useful framework for policy analysis. we feel together with a representative example of 20,000 americans spread out across the whole country, just to sort of get an idea 06 whether individuals have a good understanding of basic policy paradigms. what we found, and this is maybe an interpretation, you see between 20% and 30% of the average american have an understanding that the fed aims to maybe achieve in longer periods of time inflation rates of about 2%. so people have an understanding
10:18 am
what an inflation target is. what you also see is there's this big chunk of the population, more than 40% in our example, that answered a number larger than 10%. some said as large as 50%. so of course now what does this mean? there's a benign interpretation of this finding. policymakers, many of whom are in the room, have done an amazing job over the last couple decades bringing inflation down, low and stable. so therefore, the average american maybe doesn't care about inflation expectations on a daily basis. but of course, there's actually not an innocuous interpretation. times such as now, there's only limited space to go lower on interest rates. for example, in the eurozone, we might run into con strastraints year on how much qe we can do. communication might be the only game in town. if the average person does not have a well understanding of what basic policy aims are, it seems hard that by moving to
10:19 am
more direct communication you can indeed achieve maybe big demand stimulation by moving inflation expectations. now, of course, motivated by this finding, we tried in some other work to figure out what other sources of information people would use when forming inflation expectations. to some extent to our surprise, the most predominant source in a survey of 50,000 americans was people's daily grocery shopping experience. people tend to focus on the price changes to form overall inflation expectations. to now directly document that in the data, we paired up with nielsen and 50,000 people were surveyed. we were able to observe the daily frequency the type of goods they purchased over the previous years as well as the prices they paid for those goods. so therefore, what you can do, you can construct similar to the
10:20 am
view of labor statistics a measure of household cpi, what is actually the past inflation at the shopping of individuals and pair it up to see past observed price changes for inflation expectations. what you see here to the far left, those are individuals that had the lowest observed inflation. so their inflations expectations are about 0.5 percentage points lower than the inflation expectations of individuals that had the highest past observed shopping bundle inflation. you might say, well, this is already kind of surprising giving what we see if nielsen. it's only about 20%, 25% of overall consumption expenditure. therefore, you might expect that maybe it is not that every price change is created equally, but maybe individuals might focus on specific goods when forming inflation expectations. that's actually indeed the case in the data. it's not like individuals would actually use expenditure shares in nielsen to weigh price
10:21 am
changes. instead, think about milk. for all of us in the room, mill success a tiny expenditure share. some maybe buy milk ten times a week, others only one time a week. it's indeed the frequency of shopping seems to be determining the price changes people actually tend to focus on. what does it mean for policy? of course there's a concern this might undermine policy credibility because if you think about the inflation measure many tend to focus on, core inflation, as an indicator for inflationary pressure, tends to actually strip off completely all those volatile price changes that instead the average american tend to focus on, forming inflation expectations. to get an idea, on the one hand what might help, there may be some observer characteristics but also motivated by this recent work that actually argue that maybe the reason why we see, let's say, a forward
10:22 am
guidance effect that promises to keep interest rates low for an extended period of time might not be as effective as our benchmark model might predict. there's been this idea that maybe cognitive frictions, maybe some kind of naive behavior could explaining those differences in policy effectiveness. what we did here, we teamed up with the bank of finland where we now for all finnish men from the military, we can observe a measure of cognitive abilities. then being a nordic country, we can link based on personal social security numbers this measure of iq in the survey to individual inflation expectations from the european commission consumer survey and also household balance sheet, income, and all of that, to get an idea, how well people are informing expectations, how well people are prescribing to the prescription of our basic models. if you go all the way to the far
10:23 am
left, those are the men in finland with the lowest measure of iq. to the far right, men in finland with the highest measure of iq. you see on the y axis, the mean absolute forecast for inflation. all i want you to take away from here, you know, there's vast differences in how well individuals are forecasting inflation by measured cognitive abilities. but it's also then directly translated to policy effectiveness. what we can show in the data is that actually the top 50% of the iq distribution, they pretty much behave as we would model them. whether the ecb lowers rates, they take out more loans. so they indeed as we would hope to behave. the bottom 50%, instead, does not react at all. i think we can convincingly show it's not due to financial constraints and other potential
10:24 am
alternative explanation. it indeed seems to be the case there's a limited understanding what policymakers try to achieve. of course, this then also raises potential concerns if you do not communicate with ordinary people, there might be a concern of redistribution. now to directly show you that there is indeed evidence that the complexity or simplicity of policies seem to be important for their effectiveness. i'm not just documents some micro survey data from germany. two policies that are somewhat different in how much basic understanding germans have to have to understand the policy transmissions. both of those policies rely on the consumer. what does it mean? you raise inflation expectations and hopefully people start consuming more. the left policy here, what we label in this figure unconventional fiscal policy, this was when chancellor merkel in 2005 was first time elected in germany. she announced immediately that she wants to raise consumer
10:25 am
taxes 14 months down the line from 16% to 19%. and what you see after the announcement and before the effectiveness of the raise in consumption taxes in the top panel, inflation expectations went up immediately. in the bottom left panel, you see that germans on average immediately went out and started consuming more. on the right panel, you see the first two announcements of mario draghi of unconventional monetary policy in the sum of 2013 and in january of 2014. when you look at household expectations and purchases, those two announcements had no effect on the average german but also in all different types of populations. so it looks like simplicity and complexity of policies matter for effectiveness. then of course maybe also this means that we might have to think harder on how we want to actually communicate policy so
10:26 am
that actually everyone tends to behave as we would hope. to close the circle and also a little bit end on a more benign or positive outlook, we actually want to go back to this initial paper in which we then try to do an information experiment or a so-called randomized control trial. what we do think about, i ask everyone here in the room, what do you think is inflation over the next 12 months? then just randomly group people into nine different subgroups. then each subgroup in random fashion gets a different piece of information relevant for monetary policy. subsequent to this information provision, i ask everyone again what his or her inflation expectations are. in case you are able to talk to people directly, does it affect individuals' expectations? what you see here in red, there's strong evidence that if i tell people the fed arguments an inflation rate of 2%, you see
10:27 am
this very strong reaction in inflation expectations. so it's indeed if you have a way of communication directly with ordinary people in the u.s., people do react to it. but the second thing you also see in blue, oftentimes one argument is, you know, we have no press conferences, the media reports about our policy actions, so therefore there's maybe no need of direct communication, but i could indirectly rely on the news media to convey my message. when we forced individuals to read a newspaper article about the most recent fomc meeting, you see that actually the reaction at least in our example was only half as strong. so on the one hand, i think the bigger challenge is to get people to read newspapers but even once you force them to read, they seem to systemically discard information. and let me actually end on maybe some of the conclusions of the paper. so there is hope that if you are able to directly communicate with individuals, they do react,
10:28 am
but simple messages seem to be important. one message might not be enough by repeated communication. thank you. [ applause ] >> so thank you for two terrific
10:29 am
presentations. i thought i would start with a few questions to let these authors elaborate even a little more on what they've just said because there's a lot of richness behind what they've said. we'll do that for a few minutes, then we will also take some audience questions following that. so if i could start with you, you know, pretty compelling evidence there's this underlying phillips curve alive and well once you control for the supply and demand shocks in your way. but we still have this puzzle of inflation below target in the united states for the last ten years. can you say something about whether -- do you think it's more attributable to cost shocks or demand shocks or, you know, your model assumes there's effective monetary policy, but we have this weakness in inflation over the last decade.
10:30 am
how would you try to parse out, if i could, from your model or just your judgment? >> okay. so one thing we have said right now is that -- let me speak for the uk. the phillips curve in terms of wage growth and unemployment, wage inflation, is very evident and very healthy. we have very low unemployment, as in the u.s., and the last reading of wage inflation has been 4% in annualized. so that's a very healthy phillips curve. the big puzzle is why this wage increase is not being passed through on to prices or not as strongly as we would have expected. so a couple of reasons. obviously wages are not the whole story. so there are other inputs into production, and it's possible
10:31 am
that some of them, you know, like commercial rents, are dragging down on inflation. another possibility is that lower markets are absorbing part of the cost increase from wages. so this wage increase is not being passed through to prices right now, but it's a matter of time. and another possibility is that productivity growth is higher than what we are measuring currently, particularly in retail. so the cost increase is not as high. so i think we need to investigate all these channels and probably the truth is, you know, a share of each of them. so in terms of wage space, which is a direct connection with unemployment, we're seeing it. i think also until the u.s., though to a slightly lower point
10:32 am
in the uk. my own sense is it's just a matter of time. >> okay. thank you. so turning to michael, some compelling evidence about household formation of inflation expectations. and that households may not be that well informed but can be sort of trained. so can you talk a little bit about whether house holds, does it matter if it's households or professional forecasters? inflation models are about firms as price setters. in the u.s., we have households and market or professional forecasters. does it matter? is it sufficient for market participants to understand? did you need households to have accurate inflation expectations? >> so i think you raise an important question. at a basic level, the type of models we typically would write down, they have consumers con
10:33 am
s suming and saving. they have firms setting prices. people typically don't have professional forecasters. they also do not -- rarely have financial markets in there. ultimately, at least through the lens of our model, what i would take away from that to actually understand how effective policies are in stimulating act reg -- aggregate demand, we have to understand whether those people react to those type of policies. of course, as you alluded to, unfortunately in the u.s., at least up to now, we do not have a good survey of firm expectations. so we can actually go back a little to other countries where we have good evidence. for example, there's a nice paper who feel their own survey in new zealand and what they found is that if you look at the expectations of firms, they're pretty much identical to expectations of households. instead f you would compare the expectations of firms to the
10:34 am
expectations of professional forecasters, you see this big gap. so therefore, to the extent we can extrapolate, this would tell me that even in the u.s., most likely i would expect firms to be similar to households relative to the maybe well-anchored expectations of professional forecasters. now, to actually go to the second point of your question about financial market participants, one concern there is, of course, that what you measure is neutral expectations. they might be so-called polluted by measures of risk. also, it's difficult to infer for movements in financial market expectations that it's due to the expectations component or from the risk premium component. so therefore, i think we can learn important things from financial markets, but i think ultimately, we have to better understand households and firms. >> interesting. one more thing.
10:35 am
i think it's important who you ask in the firm. people who are in charge of pay reviews and pay increases, they do know what cpi inflation or pce inflation is. >> my understanding is that actually they surveyed the c-suite of firms. i'm with you that at the end of the day, we have to also better understand who are the right people to ask. there's evidence that at least some people that might be important for firm decisions possibly have aless good information of overall inflation. >> so perhaps could both of you comment or either of you comment on how important -- so going forward, forward guidance might be a really important tool for central bankers in the u.s. and maybe elsewhere.
10:36 am
given the research on sort of monetary policy effectiveness and inflation expectations, what do central banks need to do? so to help make that more effective. and in terms of forming expectations, central bank independence. open question, so. >> i think as alluded to in the presentation, central banks have been really successful in guiding expectations. one way of seeing that is just monetary policy shocks that are really small. so they have been successful to manage expectations of one part of, let's say, the relevant players. for household and firms, maybe less so. so i guess in this respect, we can maybe learn things from other central banks. for example, on the one hand, we have now amazing initiatives -- let's say the new york fed consumer expectations, but they try to understand how
10:37 am
individuals form kpngss. the fed now following with the center for inflation. there are indeed initiatives. the bank of england has done amazing work along those lines. i think maybe we can learn a bit more from more exotic type of central banks. on the one hand, we are working a lot with the bank of finland. the most recent ecb decision was heavily commented on twitter by a board member of the bank of finland. he explained in plain finnish terms what the bank of finland, what the ecb did and what it implies for the average fin. so on the one hand, i think this was really successfully in trying to get the average fin on board. maybe we have to even look at more exotic type of central banks. i'm not sure who's aware of what the bank of jamaica is doing. the bank of jamaica has this really successful initiative of anchoring inflation expectations. how are they doing it? they actually recorded reggae songs with national reggae
10:38 am
heroes singing and making music about price stability, why you want to have an inflation target, why you don't want to have too high inflation, why you do not want to have deflation and what all of those things would mean. so maybe regular day gae is not popular in the u.s. maybe rap is more popular. we have sports stars that are national heroes. so this is now an on vabservati. i ask my haircutter at some point whether she thinks her brother, who is maybe 20, 21 years old, knows what the federal reserve is. she said no. do you think he knows what price stability means? no. do you think he cares about what lebron james says? yes. so do you think if lebron james would rap about price stability, he would listen? she said, whatever he says, he would care about. maybe this might be a little bit far out, but this could be things we could learn from other central banks. >> that's a lesson.
10:39 am
>> i like to add two things. so let me tell you about the case of the uk. we see huge differences depending object population. people farther away from the target are people with low income and low education. now, it's possible that this segment of the population have little impact on the key decisions because their relative income is low. however, there's a big challenge here more broadly to try to improve financial literacy in this group of the population so they can improve their decision making. i think in terms of quantitative income for the macroeconomy, probably is will not be large. you know, that's -- so that's the fair thing. we need to work on that. the bank of england has now
10:40 am
different layers of communication, trying to simplify the message for groups who have a lower understanding of the economy or economics. and i just wanted to add one thing. it's tricky with inflation. coming from argentina, everybody knows what inflation is. it's something that people -- it's very pressing and very important in their lives. so people are constantly paying attention to that. so one reason why in the u.s. and other advanced economies people have, you know, less informed about it is that it's not that relevant for their day-to-day lives. >> that was a favorable takeaway from one of your results. so we can take some questions. we'll take a few at a time. could you state your name maybe?
10:41 am
>> great papers. i have many questions. i'll just ask one. the slope of the aggregate phillips curve has decreased over time. if your hypothesis is correct, we should not see the same decrease for the phillips curve. so what do we see when we look at the estimated slope of the state phillips curve over time? >> yeah, unfortunately, we only have the data from the '90s onwards. i cannot answer that question. >> can we take a few? if that's okay, yes. please. state your name, please. go ahead. >> i wonder if your survey actually tells us that people in this country, even the 40% who are least knowledgeable, know more than some economists do.
10:42 am
to wit, the biggest checks that the average family writes each month are for housing, health care and insurance, and higher education. i can show you chart after chart that over the last ten years they have all risen much more than 2% or 3% a year. so are we measuring the wrong things when we really try -- or at least underweighting these topics when we're looking at inflation? >> thank you. >> question over there. and could you say your name and please stand. thank you. >> hello. i just want to follow up olivier's point. in a sense, if we're trying to get to what's up with inflation, we probably need to know what's changes. i think his point is right. if you estimate a wage equation post-2008, the unemployment rate actually doesn't enter into a wage equation.
10:43 am
my suspicion is it probably doesn't enter into this inflation equation either. presumably what we need to see to understand the question is what's changed rather than have a 1990 to 2017 equation. you mentioned the wage growth in the uk is 4%. that's true, but it's also the cpi has just dipped below 2. it's now 1.7. we have to wonder about the relation between wage growth and inflation, the two things now going in opposite directions. but i think we need to see the structural break. >> thank you. >> let me take olivier's. as you probably also know, what's dragging inflation in the uk at the moment is energy prices. so domestically generated inflation is going up, in part because of wage pressures. and the big drag is driven by fiscal shock. but olivier, coming back to you, we cannot do that, as i said, in
10:44 am
the regional level. what's interesting, and i think deserves further study, is that the wage phillips curve for the u.s. seems to be working very well, and that's the point. even over time t hasn't flattened as much. that's the point matthew wrongly has made in a very nice discussion. maybe more to see. it might be that the cost base has changed and in some sense wages matter less for the cpi or pce basket. but in wage spaces, it seems to be working well and the flattening is less clear. >> yeah, you raised many important points with your question. i think one way of thinking about it, i think the points you alluded to, those are price changes that are very salient, important in daily lives and easy to measure. instead, you know, if you think about an iphone, let's say it
10:45 am
had a price of $999 for the last 15 years, even though an iphone today is a very different animal than an iphone 15 years ago. so official statistics take this into account and do a quality adjustment, which actually of course we now are maybe a calculation of my personal inflation rate i would not take into account. there are many, many subtleties. i think we'll also talk later about some other things that enter exactly this discussion you were raising. i think it's definitely important that people tend to think about this big ticket things in their daily lives that are important but might ignore other things that might trend down. >> thank you. microphone, please. we'll take three more questions. if you could make them brief, please, so we have time. >> peter doyle. on jamaica, the videos that you refer to came out after
10:46 am
inflation had been lowered. so their particular role in what happened in jamaica is probably contradicted somewhat by the data. my question about inflation undershooting for long periods of time in the united states, to what extent do you think that's the possible explanation that the central bankers themselves do not believe in 2%? they take to thimemselves the judgment that it would be better for the united states and others to be below 2, and hence that's what they deliver. >> thank you. yes? >> thank you. i'd like to expand on a comment that was just made. that is that lower income people do see higher rates of inflation because of this quality adjustment thing. and this could be a large part of why so-called lower iq people see a higher -- have a higher expectation of inflation because
10:47 am
they're seeing higher inflation. and does it really make sense to measure inflation the way it is now? >> okay. a question over here. >> thank you. i noticed in your phillips curves' regressions, that there was no control for relative import price. i was wondering what the thinking was there. then for both of you, perhaps to comment whether import prices have -- play a role in the recent weakness in inflation. for example, in the data we notice a distinction between core services and core goods inflation in the u.s., both in cpi and pce, for about 2 1/2 decades. core goods inflation has been almost zero. at the same time, these are the components of inflation that have a higher import penetration
10:48 am
and are more globally competitive. >> do you want to talk about the -- >> yeah, taking the last one, obviously cpi import -- sorry, import inflation is important. we control for past cpi inflation as a way to capture those past pressures, but you could also add measures of import price inflation directly. it wouldn't change a lot the slope, but i think that would be the right thing to do if you have enough, you know, power to also capture those. we do control in the final one for aggregate effects, which in principle should affect the whole country equally. so to some extent, some of these effects might be upsold there. i won't comment on the undershooting on the fed, just to say that one possibility is that, you know, the economy could -- you know, the monetary
10:49 am
policy could be potentially more expansionary. it depends, again, on how much weight is put on achieving the target or not. so let me just stop there. >> maybe just a quick remark on finland. so i guess from the perspective of the u.s., we might actually call finland a socialist country, which means that actually there's not a strong correlation between iq and income. in microdata, the correlation is about 0.15. so i think that's a nice testing ground because it allows you to disentangle effective cognitive abilities from effect of income, education that is free, everyone can go to college, and so therefore, actually, i'm pretty confident that i can at least in that specific setting disentangle the effect of cognitive abilities from things you have raised. >> interesting. okay. on that, i think we thank our panelists. [ applause ]
10:50 am
we're going to move right into our second panel. it's all about technology and globalization. we are very please to have two really wonderful presentations. the first is by christine forbes, who will talk about globalization, the professor of management in economics at the sloan school in mit and was an external member of the monetary policy committee for the bank of england. our second presentation will be about technology. we have a presentation by alberto koballo, associate business professor at harvard business school and founder of this business prices project he
10:51 am
will tell us more about. >> okay. i've been asked to answer the question is it all about globalization. there's a lot of reasons you might think it is. a lot of reasons why changes in the global economy would be affected including advanced economies. increased wages in global economies would be a one off in pricing and inflation. increased trade integration would mechanically mean a higher degree of price indices are imported and therefore prices would be more related to changes in demand and global supply or emerging markets have a greater heft in economy and it increasingly drives shifts in economy prices and driving larger movements in economy and oil prices over the last decade and that increased volatility and commodity prices and energy
10:52 am
prices could feed through prices in advanced economies especially if effects are non-linear there is evidence of. there's a whole literature arguing global supply chains, the ease of shifting small parts of a production process to where it can be done most cheaply would affect pricing decisions by company and affect how exchange rates affect companies and increase bargaining power within companies because it's much easier to shift small parts of production where it can be done more cheaply elsewhere. lots of reason it might affect inflation. whether one off effects or affect the whole process or phillips curve relationship between domestic slack in prices. if it does, there's pretty important information for banks as janet yellin mentioned in her introduction. if global inflation has affected, i have some papers and
10:53 am
research i've been working on a couple of years presented right here at brookings a few weeks ago. where the literature is moving, globalization is important. some of what is going on in dynamics affects globalization. it's not the whole story. some factors we talk about today are also important and the answer is more nuance than i thought when i started working on this. yes, globalization seems increasingly important for cpi inflation and headline inflation and increasingly important when addressi addressingsicular ups a and downs and corn inflation and wage inflation. not to say it's not important. it's more important mostly for core inflation, and the cpi with
10:54 am
moderate effects for wage and core inflation. in the brookings paper i did on this, these results are supported by coming at these questions with different approaches, different inflation measures and different techniques and countries. i will give you a couple of the key results that support the same story. let me start with principle components. i have inflation data for 35 economies throughout the world and i take out the shared principle component of these shared inflation matters, basically how much inflation moves together in these advanced economies around the world. what you see not surprisingly, producer inflation, the brown line at the top around the world, and the high level components are not surprising they move together in different countries. interesting patterns are the other lines. cpi inflation now moves much more tightly around the world
10:55 am
than in the past. cpi inflation could be driven by global factors but wage and core factors seem to move by their own beat. there doesn't seem to be a global component driving wage and core inflation as you have increasingly seen for cpi inflation. a lot happening. the pattern in the data doesn't show you what's going on. let's go a little deeper. i will show you the results from the phillips analysis. i wasn't sure what i would find when i started. when you do the phillips curve for a cross section of countries you find pretty strong robust results. i will estimate the standard workhorse phillips curve model economists look at for large economies such as the u.s. estimate inflation is a function of inflation expectations, lag
10:56 am
inflation and domestic slack and i will use domestic slack, not just unemployment. i will do a simple augmentation of that model you also use for import prices, a variable for everything going on in the world. those are the standard workhorse models for standard domestic inflation. i will control for several different ways global inflation could affect inflation, changes in oil prices, changes in commodity prices, changes in exchange rates, changes in global slack, not just domestic slack and changes in use of global value changes. i will do one more augmentation to allow global barriers not have just one off shock inflation but affect this phillips curve relationship between slack and inflation. basically, can a country's exposure to imports explain the flattening of the phillips curve we talked about today. when i estimate those models for cpi inflation, i find the simple
10:57 am
domestic phillips curve variables all come insignificant pretty robust expected signs, higher expectations, higher lag mentation domestic slack correlated with higher inflation in a cross section of countries. this sort of framework works pretty well in cross section of companies and correlated with higher inflation. what's more interesting controlling with global variables. those col insignificant with the expected signs. higher oil prices, higher commodity prices, ex-energy, exchange rate depreciations, less world slack and less use of global value chains is significantly correlated with higher inflation. that suggests these global variables matter at least explaining cpi inflation. one more variance, when i also allow for exposure to trade or imports to affect the flattening of the phillips curve that comes through in this this explains
10:58 am
this significant also. it looks like it's quite important. to give you a concrete example, if you look at the phillips curve over the last decade relative before the crisis, the phillips curve flattened as we've seen in a lot of other work. import exposure explains half of the phillips curve. that shows globalization not on only immediate, but on domestic. i also ran these models with rolling regressions and calculated the errors. what if you estimate just a domestic model explaining inflation or add these domestic variables, how much does that affect our ability to understand inflation. what you want is smaller errors and the smaller the numbers the
10:59 am
better it works. you have the black lines estimating models with just domestic variables and the rest including global variables. including global variables reduces errors by over 12% on cpi inflation, on average. and particularly big reduction in errors on financial crisis incorporating what is going on in the rest of the world is important and more important 2012-2015, world slack plays a role in commodity prices explaining why inflation models didn't do so well. if you add the global variables in inflation models you can reduce the errors and improve the fit by 12%. not the whole explanation, far from perfect and still other models going on but it does make a difference. if you go back for core inflation and wage inflation,
11:00 am
the domestic model still works and domestic variable still important explaining inflation. global variable sometimes important and often prices explaining core inflation and much less robust and magnitudes smaller and less important than global variables explaining wage inflation. some flattening of phillips curve you estimate this for core inflation and not as much as wage inflation and increased exposure for core wage inflation. estimate the model and see how much does adding these global variables reduce errors in models to explain core inflation on the left and wage inflation on the right. it improves the models a little bit but not as much when you try to estimate cpi inflation. global variables only moderately reduce areas in models of core wage inflation. still helps to add them but you
11:01 am
won't explain any apparent puzzles over the last decade. i could stop there, and when i presented this paper here a few weeks ago, that's largely where i stopped. bern bernanke asked me a good question. global wages matter for inflation. how much and do they really matter for the u.s. basically. i have an answer for your question now so you won't ask me in five minutes. what i did is i took these models and used the estimated co-efficients and plugged in the actual variables for the u.s. and estimated how big an impact these global variables had on cpi inflation for the u.s. this is what you find. let's start with some of this puzzles we've been talking about today. during the period immediately during and after the global financial crisis, u.s. inflation was higher than many people expected.
11:02 am
what drove up u.s. inflation for that period? from 2010 on, the change rate propped up inflation by .1. oil prices and commodity prices were important in 2009, 2010, 2011 boosting up inflation during that period. global value chains by contracting weren't dragging down on inflation as much as before. that's part of the explanation why inflation was a bit stronger during the crisis. the biggest results come in for the period after the crisis, 20102015 window, when inflation was slower to recover in the u.s. than some of our models predicted. how much of that was global variables? let's suggest the dollar exchange rate was part of the story, .1, .15, .2. oil prices, commodity prices also were a temporary drag. i think the most interesting what hasn't received as much
11:03 am
attention is the bottom left. global value chain as they picked up, trade started to pick up more quickly 2010, 2011, 2012, emerging markets were reacting more quickly. that acted as an important drag on inflation and global slack acted as a drag on inflation. each effect by themselves is very spral, .1s and .twos. very small. but domestic slack a broader measure of domestic slack from unemployment but as the drag on domestic slack faded you still have pretty meaningful drags from global value change and global slack. that can explain some of the puzzle. not the whole story. the obvious question is what does that mean for now. my sample ends at the start of 2018. it suggests from 2018 forward, a lot of these drags that were holding down inflation and
11:04 am
presenting some of this puzzle aren't as potent anymore. exchange rate and i won't predict what happens to oil and commodity prices there but oil slack was dragging isn't dragging as much as global value change especially as trade tensions are flaring up and supply chains could be an important factor no longer keeping inflation down which hasn't been fully incorporated in most of our standard models. to tie up i have a whole set of results making modeling inflation out of the phillips cycle and you get very similar results to what i showed you today increasing my confidence in this set of results. to summarize the point is globalization is increasingly important for understanding cpi inflation in the cyclical short term movements in inflation. wage core the underlying trend
11:05 am
in inflation still largely a domestic process. globalization matters a bit and helps a little bit but not the answer to puzzles out there. as to which matters for cpi inflation. impo import exposure is important and explains cpi inflation, but what you're looking at, the time window is, the global variables do seem to be important as a hole but their specific importance depends on what variables and channels you're looking at. thank you. >> thank you. thank you for the organizers for
11:06 am
inviting me here. i tried to mimic christine' title and change the word at the end and decided to put amazon. i notice these days you put amazon at the end of the paper a lot of people pay attention. you can more generally think of these as online competition, something happening in the market that is affecting pricing. as you probably remember, the story of the amazon effect became quite interesting particularly around the time where we were seeing low inflation two years ago that seemed quite bustling. the argument was amazon is somehow putting pressure on the margins of some of these retailers and constraining their ability to increase prices even in the face of rising demand. there's been increasing interest in the topic of online competition. i personally, i should have
11:07 am
mentioned this, i don't work with online prices because 10 years ago we started to start collecting a massive amount of data from retailers that sell on line through the prices project we have at harvard. in a sense, our dataset should be exactly what you need to detect if online competition with amazon is affecting the behavior of large companies in the middle of this offline and online world. i decided to tackle this question. others have related research worked a lot on the characteristic of online only type of pricing and a tremendous flexibility has been detected in prices online. we have others focused on the measurement size, what this internet pricing does to the way we calculate pricing the goods for the keno paper.
11:08 am
i focused on is there an amazon effect and what is the focus in terms of dynamics. to give you a preview of my answer, this story about the shrinking margin, i think may be relevant. the truth is i cannot observe those margins. these companies will not give us that information. you can think of it essentially if it is putting some pressure it is a temporary pressure. there is so much margins can fall and when they're small you would have to see quite quick reaction to some changes. i think, if you want to take out something technology or online competition is affecting in inflation, it's far more important for us to focus on the way pricing behaviors of these companies have changed over the years. we have 10 years of data we can follow over time. one thing we have detected, two trends we detected being particularly relevant for the case of mon tied policy, one is
11:09 am
the frequency of price adjustment has increased dramatically. part of increasing algorithms and monitor what others are doing and trying to mimic their behavior and the second is uniformity in pricing across the u.s. many retailers have a single price. people have come to expect they will get the goods in a matter of days for free in theory in terms of shipping and shouldn't be any difference in boston or san francisco. that is making these companies price identically all over the u.s. you put those things together, what does that mean for inflation? i argue in this paper it's basically making these prices more reactive, more sensitive to, i called them here aggregate. but you can talk about local aggregates but less about the inventor and more about what's happening on a national level,
11:10 am
which can be related to increasing domestic type of national shock or global shock. i will show you in particular to argue some of the slowdown we had in 2017-18 was actually related to the types of shocks in gas prices. that has actually changed recently, and there is more inflation now. let me walk you through the results and document these facts. first, there's a tremendous amount of algorithmic pricing, at least in theory. companies are not very straightforward how this is happening. since 2014, we have been hearing stories of companies using algorithm to make pricing decisions. it makes sense. we're letting algorithms drive our cars, why wouldn't we let them make these pricing
11:11 am
decisions. these companies of a lot of data. it doesn't matter if it's one company or all companies doing it but they're monitoring each other constantly and the cost of doing that has decreased dramatically. some are monitoring algorithms and mimicking this behavior and the pricing algorithm is more-or-less the same. thinking how long prices send to stay constant has shifted dramatically across all configures of goods, particularly where you would expect online competition of amazon in recent years. household goods with just six months. i'm keeping contact with the sectors and retailers. we know it's not about the competition. in sectors where amazon does not seem to have much of an impact
11:12 am
you can see it's relatively stable. that may change the next few years. a tremendous amount of flexibility. the second point, it has increased over time. i try to link it specifically to companies that are linked to amazon and flexibility. the next is uniformity of pricing, a little bit bustling. because if they're using algorithms, why not give us the same prices at the same time. it gets complicated, not a technological constraint they face, i believe they face the fact that online prices make it tremendously transparent. you can see prices in different locations, and people have concerns if they see they're getting charged a different price for exactly the same good in the same location. there's a fairness concern. they will tell you we are very
11:13 am
worried about using algorithms to customize prices because we will break the choice of our consumers. there's an old story of amazon facing criticism in 2002 for trying to price discriminate and sell cds to different people at the same time and they promised never to change the prices based on demographics. technologically they can but in terms of transparent concerns there is the opposite, uniformity. i'm showing you a sample we collected data, for 100,000 coats and you can see the sheer identical prices is nearly perfect. it's higher in amazon than more traditional retailers but not that far behind. why does this matter? you put the two things together. i mentioned at the beginning,
11:14 am
high flexibility and these type of national pricing you would expect the prices to be reacting more to national. you can take gas prices and exchange rates, two simple ones we can monitor and you can see those in direct competition with amazon, thinking walmart's prices, they tend to change prices more often, tend to be more uniformly price and tend to have higher pass-through rate for gas prices for exchange rates as well and the sensitivity has increased in our sample during time as well. you may be thinking this is exactly opposite of what we are thinking to explain the bustle. they are adjusting so what does it explain? we are trying to identify what are the shocks we have experience in the last few years. in that sense, what i did was
11:15 am
try to show you how, first with online prices, in particular, you can actually find the volatility or movements in gas prices and exchanges explain the core prices. looking at the orange line, that's like an online core, equivalent with the u.s. core and data we get from these retailers. far more exposed to the amazon competition than you get in a cpi sample that has retailers and others upon what i circled, the period where it was great. you look at the timing, basically the general 2017 the inflation rate was falling was a time gas prices was falling and dollar appreciating. it started to turn exactly at
11:16 am
the point gas prices rose in january 2017 and the dollar appreciated for roughly one year. since then, you can think of these two forces balancing each other as we have appreciation of the dollar, relatively high gas prices. the understand and downs you see in this orange line are actually quite closely co-moving with gas prices. it shows you the same message, only shifted, has a lag, complicates the identification of this type of effects of these core shocks. to finalize because i'm out of time, my view is that even core prices today due to the technological changes and changes in pricing behavior is becoming less insulated from the shocks and we should be paying attention from them. it can go both ways, be more deflationary but if more shocks pick up we should be able to see more inflation in the short run.
11:17 am
if the focus is understanding these shocks we need to think more carefully what sort of variables affect these pricing algorithms and how the shocks are perceived by the customers. i'm doing research on the trade war. when i started i expected it to be quite high. at the retail level there is some limitation because of the way many retailers perceive how temporary it was going to be or easy to be to wait a little longer before implementing those. in any case, i have a last point i mentioned. i'm way out of time and happy to talk about that later. thank you.
11:18 am
thank you both for a really fascinating and incredibly clear presentations. i want to start with a clarification for you, kristin, not really clarification but interpretation. you're finding globalization important for explaining headline cpi and even flattening of the phillips curve really interesting but not for core and not for wages. so one of the mechanisms i thought it might flatten, there's no -- you can't ask for wage increases because you have foreign competition. what is the economic story that
11:19 am
explains that distinction? i know you don't know it from the data but what do you think? >> let me clarify. i still find some role for the global variables for core inflation. when you put them in they would be 3-8% or something. >> what about flattening? >> not 12%. >> it's about 10% of the flattening. it's there but not nearly as strong as for cpi. i think there is a puzzle there. some of it could just be we have had these huge sum of multiples and commodity prices seens a short term. back to our first panel, central banks may not respond to those because inflation expectations are well anchored. you don't get the reactions there. what also came up in the first panel, there is this puzzle where there has been -- the wage
11:20 am
curve is working really well. you haven't seen the same flattening and same effects of globalization. wages are going up in some countries but prices are not. that subjects, as said, global growth may be increasing more than we know. i don't know about that. or firms are reducing their margins. there was some excess profits in there that could be shrinking. that can't go on forever. if that's part of the story, we might see some effects come through later on. it may be a temporary effect. >> that brings to what's up with inflation, how much do globalization and technology explain it. to the extent they do explain it, is it because things have been happening that have masked this underlying phillips curve as in the first panel or because they fundamentally changed the structure of the inflation process and relationship between
11:21 am
gaps and inflation. i want to get both of your views on that with globalization. why don't we start with technology. >> it is masking the relationship. it's basically because i believe -- i do know the prices are quite reactive to some of the shocks, a matter of identifying what the shocks really are. we may be placing too much emphasis on the very short one what is happening with the u.s., which is understandable. i think we need to acknowledge if we carefully try to understand the characteristic that shock generates at the retail level, we're likely to understand better. i'm more in the camp of seeing these as a relationship that is there, just that we are not considering the right types of
11:22 am
shocks to see it. >> people don't have pricing power because they can't react. now, it's so toes price compare and so maybe that would change the slope? >> i think that's a concern. particularly in the short one. you think, how low can margins go and how sustainable, kristin said. we were looking at the effects of the trade war. you can think of this additional shock passed on to consumers. there are some stages where they can stop it. one would be if the exports from china change their prices. we do not find that. we find importers today pay a significant cost of the tariffs. we should observe a quick pass-through at the consumer
11:23 am
level. it depends how these retailers are internalizing their reaction and how long they can sustain reaction to margins. retailers suggest they are reducing some margins, but how long can they be maintained. as soon as these margins are low enough or convinced the shock is low enough we will see a significant pass-through in consumer prices. >> if i asked you to start with the puzzle, not enough deflation during the great recession and not enough inflation now. how much of the amazon effect can explain it, not just particular years. is that an important part or not an important part? is that a sense? >> kristin's analysis was a far more long term analysis. i am mostly focused on the last 4 or 5 years. i can tell you in that perspective, knowing some pricing competition is making versus quite reactive to the shocks, i can only explain or
11:24 am
argue there's not much of a bustle going on with inflation, when you take that into account. we tend to have this impression the pass-through rates of some shocks of retail prices is relatively low, we rename them and focus on core, i think we should be increasingly more focused and making the connection to commodity prices and other type of shocks. >> before i answer, i want to comment. i think there are two results, a nice complimentary in my analysis i didn't get into in much details you see movement in commodity prices over the past decade and a direct result on cpi you see in your results. i find sensitivity to prices to commodity price increases has increased that also fit with your story. how i justified that before
11:25 am
seeing your pape are the could be nom naitors. there's nice work by hamilton that shows larger oil prices lead by larger adjustments by companies and fits with many price models, bigger shocks prices feaster and accelerated between the shocks you're getting. a nice link between the papers. back to your question does globalization mean one off effects in globalization in the process? i find in cpi inflation it's both. globalization has big one off effects on the level of inflation. some could continue depending what happens to global change. but the slack for inflation and cpi inflation. core inflation, you still get some direct level effects, commodity prices, global slack had some more role and smaller effect on inflation and smaller still the direct level effects
11:26 am
of channels through slack. >> we talked about how much we can explain what happens in inflation, going forward with monetary pricing in the central bank. how important do these matter? more frequent price changes? should they be focusing on a different manner of inflation or moving around a bit more or different moncation in the structural change of the economy in terms of globalization. >> for what to look at, it's important to in corporate rate the exchange rate and gas prices and discussions of inflation expectations, also why the fed may want to pay more attention to headline than to core. i also wanted to make the argument that measurement can actually be changing because of this. we have statistical
11:27 am
methodologies based on a very different type of environment. as you can think of the frequency pricing is increasing, also the rotation is changing. we should be thinking about whether we're measuring well some of these inflation statistics. to something often overlooked to but to think inflation is better measured than other statistics. you can expect it to be playing role as well. >> my answer to that is pretty supportive. i would like to see especially in economics, there's still a tendency for economics who are in the u.s. to write a phillips curve and ignore the rest of the world and you're done, maybe oil prices. i would like to see economics doing simpler models to include more terms of what's going on in the rest of the world. it can matter. the gsd models do incorporate
11:28 am
the rest of the world some and some ways they can involve that and put more weight on it. the effect of the global variables does vary over time and you need to allow flexibility. some seem non-linear and seem to grow in certain periods and non-linear in certain periods. you have to have that growth built in. >> we will open up for questions. take three questions at a time. raise your hand, the mic will come to you stand state your hand where you're from. any questions? way in the back. >> michael redmond, xpx capital. i hear conflicting stories on both these issues. sometimes you hear we have increasingly comfortable mon noplistic companies who don't have much competition and others forcing margins and passing prices quickly. with globalization, even before
11:29 am
trump there was talk about the slack of globalization, that the hyper globalization era already ended if you look at the import of intermediate goods that flattened out before the financial crisis period. how do you square the competing stories and local economists and what that means for inflation? >> david beck -- sorry. david beckworth. this is maybe more for kristin, but on the globalization front, i know we're thinking about the u.s. here, low inflation of the u.s. if you look at the world and advanced economies, japan, europe, we have this struggle. it's also where we see low exhibits rates. at jackson hole you had a graph showing a rate where it's going down and emerging markets high and robust. we're also the part of this
11:30 am
world that provides the safes a set. i'm wondering if there's a safes a set story here advanced by the economies global demand liquidity shot that feeds into low inflation and advanced economies. >> why don't you take those two? >> let me start with the first one, different stories of globalsation, has it increased and trade on global supply chains. i thought global supply chains, easy. it's very hard to know what you mean by that. increased trade seems to correlate with gdp growth. really flesching out how much global supplies are increasing. i finally get how much trade there is for intermediate goods and crossing and not just one time. the measure you use in this paper, a number of different
11:31 am
components get together, at least in my head i think of as global supply chains. global supply increases before the financial crisis, collapsed, came back pretty quickly and now at high levels and started to decline recently and i don't have data for the last year. my guess is they would have declined more in the last area. that's how you get different effects of global supply and globalization over time contributing to the period after the crisis this mode of producing things more cheaply collapsed and came back and advanced economies contributed to lower inflation in 2016 and '17 and we're having a diminishing impact. that's the time series. safes a set story and lower interest rates, i think that's interrelated to some of this. one of the key points i tried to make at this jackson hole discussion you referred to was i
11:32 am
think we put far too much weight on this estimates of the neutral interest rate, margins of error, massive, depending how you estimate them. the concept makes sense but we're wringing our hands getting these exact estimates when we just can't. where i think the safes a set story does play more role that sometimes gets attention, when you do get some shocks, you get movements in and out of certain currencies, safe asset that drives exchange effects and seem to be more important than many people take into account in terms of what's happening in inflation, especially as trade increases and global supply chains and they can have pretty big effects on pricing. in the u.s. they say it doesn't matter as much but certainly not as much as other countries but i suggest it is there and does help on the margin explain some puzzles when inflation is a little higher or lower.
11:33 am
i'd like to see more inflation in that channel and less estimating this neutral large start. >> i completely agree with that last statement. maybe because i'm from argentina because everything i see with prices to my mind has to do with exchange rates. but certainly, i think my results suggest there is more than we typically assume. it is true there are conflicting stories about the margins, for example, margins shrinking margins and someone with huge margins have lower pricing and i was trying to distinguish between the traditional amazon effect, i think that's a short term story. 20 years ago people were concerned about walmart and how creations from china were creating a walmart effect. there can be short term deviations but i see it as a temporary effect.
11:34 am
changing the way pricing decisions are made, it's more of an inflationary environment and where costs pick up. that's why i believe that's where we should focus our emphasis on the impact of technology and its implications for money tarry policy. >> all right. [inaudible] i was surprised by your results when i heard the comment. amazon is not just a monopoly, also maybe those are the margins that are shrinking and can explain that. >> one over here. >> carl poser. center on capital and social
11:35 am
equity. if we had this conversation when the curve was originated and monetary system developed there would be a lot of talk about labor and bargaining power. headlines could affect the expectation of inflation and shut down the steel industry. we really don't see that anymore. now, you see -- if you go to mcdonald's and talk to people about should you get a minimum wage increase that's significant. they say, then, our hamburgers will cost more. are we seeing -- in other words, the flatness of the curve, is a lot of it because labor is now more of an independent variable or is it more of a dependent variable and not independent actor? is that a question, i guess? >> lately, what he asked, it's actually consistent with the paper of penelope goldberg. at the border, there's full pass-through. it's not the case the chinese
11:36 am
exporters have to significantly drop their prices, so the burden is on the u.s. there is a question whether that is being passed onto the consumers and that's where our paper adds to the picture. we find the pass-through. that's why i tried to make an emphasis we have to think about what it generates. the transmission of that shock will be different from a shock like gas prices which can feed directly into these pricing algorithms and quickly consumed at the consumer retail level. >> in response to your question, i think there are a lot of interesting things going on. when i was working at the bank of england and companies around the uk i was amazed how many companies say, we can't find enough workers. i say, pay them more. they say, no, our prices would be too high and we can't compete. it's again and again. interesting dynamic, some
11:37 am
globalization, increased competition from other companies. more workers are self-employed, part of the sharing economy, working for an uber and working part-time. in the analysis i did i did not just look at unemployment rate as matter of slack. i looked at hours worked relative to normal hours and some self-employed and some may want to be and some may choose that. and some part-time and those who dropped out of the labor force. i found having that broader measure significantly improved the fit of the phillips curve. those other dimensions of the labor market are very important. the reason i didn't talk about that much because we have the next panel to focus on that. i hope it will get at your questions in much more detail. >> thank you. we're out of time and thank you to this interesting panel. you have a 10 minute break. sorry.
11:38 am
we're having a short break right now. sorry.
11:39 am
11:40 am
a look now at some of today's live coverage across the c-span channels. former ambassador to the united states and special assistant to president obama, samantha power talks about policy. and republican ock casio cortez will talk about housing and access to federal benefits. c-span will have live coverage starting at 5:45 eastern and see both live online at or list within the free c-span radio app. to listen free at 8:00 p.m. eastern we are featuring
11:41 am
programs available every weekend. tonight, marking the 400th anniversary of slavery in the u.s. and four scholars discuss activism and institutions and how they have survived despite it. and listen every weekend on c-span3. saturday at 2:30 p.m. eastern, on american history tv, winston lord, author of "kissinger on kissinger, on president nixon's relationship with his secretary of state, henry kissinger. >> nixon had read kissinger's books. nixon wanted to dominate foreign policy when he was so well-versed and interested. he knew to do that he needed a very able thoughtful national security advisor. >> at 10:00. >> there was little to do except work.
11:42 am
life seemed rugged indeed. here in such a setting they prepared to search for oil. >> on real america, the 1948 film, desert adventure on the saudi arabian oil industry. and then american artifacts we preview the "votes for women" exhibit. >> victoria was well ahead of her time. she started her own business as a wall street banker with her sister. she advocated for free love, which means sex outside of marriage. >> at 6:30 p.m. eastern, author, sofia rosenfeld discusses her book. >> no one person, no one institution, no one sector, no king, priest, national research body, specific cast would get to call all the shots. >> explorer nation's past on american history tv every weekend on c-span3.
11:43 am
while the brookings institution on inflation and monetary policy is taking a break we will show you this morning's comments by former federal reserve chairman janet yellin. she outlined some issues the fed is dealing with. >> pertaining to inflation is aptly summed up by the title of this conference. what's not up with inflation? inflation hasn't moved up through an expansion that now ranks as the nation's longest on record. unemployment declined from about 10% in 2009 to 3.7% today, a 50 year low. yet headline inflation over the last year according to the federal reserves preferred personal consumption expenditures or pce measure stood at only 1.4% in august.
11:44 am
core inflation, which exsclyouced volatile food and energy prices came in a bit higher at 1.8%. the current pace is close to that in 2009. contrast this experience with that during the long expansion during the 1960s, unemployment declined from 6.7% in 1961 to 3.6% in 1969. over the same period, headline pce inflation rose from just under 1% to roughly 5% and core inflation increased by a similar amount. the phillips curve has long served as the workhorse model of inflation and it's used by most economists including federal reserve staff to analyze and forecast inflation trends. the current inflation depends on
11:45 am
the degree of labor market slack and lagged inflation and a variety of supply shocks, including those affecting the prices of food, energy and other commodities and the value of the dollar, which affect import prices. various transitory factors or noise also affect inflation readings. the comparison i just described concerning the behavior of inflation as unemployment declined during the opportunity expansion and that of the 1960s illustrates two robust empirical findings. first, the slope of the phillips curve measure of the responseness of inflation to a decline in labor market slack has diminished very significantly since the 1960s. in other words, the phillips curve appears to have become quite flat. second, inflation has become much less persistent because the
11:46 am
impact of lagged inflation on current inflation has declined considerably. this likely reflects the fact that inflation expectations which affect wage and price setting now appear to be quite stable and unresponsive to variations in actual inflation. inflation expectations, in other words, have become well anchored. during the 1960s and 1970s, in contrast, a rise in actual inflation appeared to boost inflation expectations considerably. high responsiveness of inflation expectations to actual inflation works to boost the longer term response of inflation to changes in labor markets slack considerably. in fact, it creates the possibility that -- >> so, i'm very excited to introduce our next panel. it's all about the labor market. we have two great panelists. the first we well hear from
11:47 am
kostya, a principal economist in prices and wages section in the division of research and statistics at the federal reserve board. to provide context, that's the section responsible for preparing the forecast of inflation and prices for the fomc. i have to admit that katia and i were long time colleagues. i want to assure any lawyers here none of our conversation should constitute my making a request of katia, the federal reserve system or fomc. our second panelist is jared bernstein, a senior panel at the budget on policy and been active many years in the labor market and inequality and policy both in the obama administration and in various think tanks in d.c. and you can read more about his thoughts on the economy.
11:48 am
i will turn the presentation over. >> thank you very much, stephanie. it's a pleasure to be here. how do i -- thank you. okay. so, in the spirit of making disclaimers, i should also make the disclaimer that the views expressed here and in the following discussion are my own. they don't necessarily reflect the views of my colleagues at the federal reserve board or research staff at the federal reserve system as a whole. with that said, i will start my presentation by illustrating some of the points that our former fed chair, janet yellin, made in her opening remarks. the fist chart is a chart of the four quarter changes in u.s. core inflation, as measured by
11:49 am
the pc price index since the 1970s. you can see that inflation dynamics have changed dramatically over the past half a century. over the past 25 years, inflation has been considerably lower than over the preceding 25 years. furthermore, over the past, it indicates calling inflation most relatively narrow range despite big swings in oil and other commodity prices, the great recession and unprecedented monetary policy actions. so, i think we can say that over the past indicates core inflation can be well characterized empirically along the stable long term trend. this fluctuations in turn can reflect changes in
11:50 am
philips documented a negative relationship between wage growth and the unemployment rate. soon thereafter economies confirmed this relationship for a lot of developed economies. now, labor costs are about 60% of the firm's production cost.
11:51 am
intuitively they should matter a lot for the pricing of firms. indeed in the mid 60s they pointed to a negative relationship between the unemployment rate and price inflation, which we refer to this negative relationship these days as the philips curve. this curve, this relationship has also changed a lot over the past half a serchcentury. the green line is the difference between the unemployment rate and the estimate of the unemployment which in this case is the congressional budget office estimate. i want to point out that in the early '90s and the early '80s when the unemployment rate moved
11:52 am
up above the natural rate of unemployment which means the green line went above the zero line, inflation declined markedly amarke markedly. now, if you look to the latest 2009 recession, you don't see such a decline in the inflation even though the unemployment rate went up. so inflation didn't come as much as it did in previous downturns and the models we were using back then could not explain the behavior of this inflation. so the failure of the models to explain this have a lot of
11:53 am
research and papers. robert gordon's 2013 paper was the philips curve is alive and well. but then in 2018, wrote about the debt of the philips curve blinded in 2018 is the philips curve and most recently the philips curve that or is it just hibernating. and i think there's a fairly broad consensus that the relationship between unemployment and inflation has changed, but i don't think economies agree whether inflation has become completely disconnected from activity. so today i will argue that, yes, the philips curve relationship is not the same as 20 years ago but it has not completely disappeared. and i think that the philips curve framework is still a
11:54 am
useful framework to think about. in the rest of my presentation i will provide an update of research i've done with one of my colleagues who is also here that has allowed us to look at the responses of different economic variables over time. so the model is a vector out to regression model whose coefficients are allowed to change over time, which allows us to look at responses. on the left panels we've plotted the responses of labor cost growth and on the right it's the responses of core inflation on unemployment cap shock. just forgot to mention we have four variables, a measure of inflation which is core market based inflation for the u.s., import prices, a measure of slack which is the unemployment
11:55 am
rate gap and wages or unit labor costs in this case. as you will see in the left panel, the response of labor costs to the unemployment cap shock have changed but not that much. at the same time, the response of core inflation to unemployment shock has changed a lot. this is the same thing you can observe in standard phillips curve models. even though the red line which is the latest response to an unemployment cap shock is very small. it's not nonexist tanent. it's still there. if you look at the decomposition of recent movements in inflation so that inflation is plotted as the black line, there's a
11:56 am
baseline forecast from this model which is the red dash line. and then the baseline model plus the effect of the structure of shocks in this case we'll look at the effect of the unemployment cap shock. you can see that the unemployment or slack in the economy is still an important factor for low frequency movements in inflation. since this is a discussion of labor markets and inflation, i would like to address one question that we very often get. are we measuring slack appropriately, is the unemployment cap an inadequate measure of slack. and these are all good questions. but going back to this picture, i want to point out this thing. the fact that we don't see a flattening of the wage phillips curve but we see a flattening of the price phillips curve means that it would be very hard to explain what we see on the right, but what we see on the
11:57 am
left, it's not coming entirely from the labor market developments. and there are other things going on. that's why we are not the only panel presenting today. [ laughter ] >> so against the background of all these long-term changes in inflation dynamics that i plotted on the first linslide, has become increasingly difficult to discern from a single factor. it can be obscured by other factors affecting inflation. so before i pass onto jared, i would like to reiterate two things. over the last couple of decades we've seen two major changes. one is the change in the trend which is now much more stable. the second is that we have a
11:58 am
lighter phillips curve. we don't perfectly understand the reasons behind those changes. we have hypotheses and a lot of research is going on now but i don't think we have a very good understanding of what's driving them, which leads me to the final point that from a policy perspective, it's worrisome how much do we think these trends will continue in the future, what will take us to move them away from there, and can we exploit the stability in a productive way. thank you. [ applause ] >> thanks very much and thanks for inviting me.
11:59 am
the center always comes up with the most interesting topics and gets, present company excluded, the most interesting people to talk about them. so this is a real great opportunity. i'm going to buzz through a number of points that have just been relentlessly made today so i don't feel i have to spend my precious time on them. the first she took us through in an extremely clear fashion. others have noted that the wage pc is steeper than the price pc. i think that's important for what i'm going to talk about in a somewhat different way although janet yellen references this in her talk, and that is the opportunity that the persistence of low or stable inflation in the flatter phillips curve determines for us in terms of achieving much tighter labor markets persistently and better yield to
12:00 pm
the people who are frequently left behind in slack periods. so that's going to be the core of my discussion for you today, the opportunity that these die ma' dynamics present to us. the labor market, we are here to try to talk about the role of the labor market. it actually explains an increasing share of the variance of prices, more so than i thought. we still have -- that doesn't undermine the opportunity that i mentioned. these gains, by the way, the gains that i'm going to talk about and show you that are engendered by this opportunity, they're not just about equalizing wage pressure. i've been working on this area and the benefits of full employment for a long period of time. i tend to emphasize how the
12:01 pm
elasticities are larger for lower paid workers. that's important. in that sense full employment is equalizing. what i'm going to talk about today is opportunities for people left behind on the labor supply side. the kind of conical story. this is just the unemployment gap plotted against a components mashup of five different wage theories. the two things i wanted to raise about this today is the first which is consistent with my theme which is that less page price pass-through which is the implication of this and something we've already heard about today, not zero, but less wage price pass-through further underscores the opportunity of a flat phillips curve to run lower for longer unemployment and tap the benefits that i'm going to show you in a minute.
12:02 pm
but i also think that this slide poses a bit of a challenge to those who want to argue that the regional phillips curve is perhaps a more important one or a more valid one or certainly a better identified one given the variance and the observations. there's a lot more variance in unemployment if you get below the sub national level. we have a wage curve that's decently identified from the national data and it's tapping the same variance in the national unemployment rate. i guess i'm very interested in the fact that the regional phillips curves tend to be steeper, the ones at the msa or the state level. but i'm not sure that seals the deal as to solving the mystery. because the national wage phillips curve remains alive and well. how much variance does the labor market explain anyway? this is a very, very simple
12:03 pm
exercise that i think is hopefully somewhat revealing in the sense that regress the expectations and you take out the residuals so you net out the part of the variance in inflation explained by expectations and you do a rolling regression on the residual using the u-gap series. i think so far today we've only used cbo's because that's only a mouse click away. so there's that. at any rate, then you start peeling off the r squareds from that. this slide shows you how much of the variance in inflation does the labor market or at least the unemployment gap explain. in recent years it hasn't explained very much at all, but it's started to climb up. this is very much consistent with katia's point that it's not everything but it's not nothing either. in fact, something like 12% of the variance is explained.
12:04 pm
therefore 12% of our time today should be spent talking about the labor market. these dynamics create an opportunity. yes, there's a flat phillips curve. but at least as important we've had decades of higher inequality, long-term real stagnation for middle and low wage workers, lots of people in places left behind and bargaining power deficits that are offset by full employment. i recently wrote a paper with keith bentley where we concluded that these change in dynamics should create a symmetry in the fed's reaction function to elevate the benefits of full employment and diminish the risks of inflationary pressures. i say that with recognition of janet's point that those risks haven't gone away and the flat phillips curve could be a problem on the other side if you're hit with persistent
12:05 pm
inflationary shocks. david miracle is an analyst from goldman sachs and he makes the same point. i only put that in there because he's -- you know, i'm sort of associated with the left side of the spectrum, so there's my point. here's an inflation analyst arguing that the tight labor market poses less risk today and policy makers can exploit the disinflationary expects to run the labor market hot as long as expectations stay anchored. that's really important and the key to my presentation. here is some new work that i just released this morning on my blog with keith bentley. it's going to become a full paper for the unemployment project. this gets to the benefits of full employment in ways that you probably haven't seen before. we look at all the different
12:06 pm
qui kwi quintiles. but what this shows is the blue line is the employment rate. it's very cyclical. these folks are very responsive to the cycle in terms of their labor supply crossing the extensive margin. and the protection theediction unemployment rate to predict the line. at some point you're saying one cyclical variable predicts another cyclical what i have on the right, i'm just going to focus on the circled numbers. we look at the share working in the bottom quintile. looking at the 1990s because it
12:07 pm
was such a high pressure labor market. there was a lot of other stuff going on in the 90s. all true. but there was also a really high pressure labor market. if you look at the earnings of low income workerin eers -- and includes zeros, this is if you have earnings or if you don't. they grew by 50% or 50 log points. because the three columns are multiplicative to the earnings colu column, basically the story is that half of that increase or almost half of that increase is crossing the extensive margin coming into the job market. for african-americans, their earnings and i'm including zeros so annual earnings including zeros more than doubled over this high pressure labor market and half of the increase was crossing the extensive margin,
12:08 pm
was increasing in the share of work. that shifts in reverse in the bottom panel when t. when the economy sniffles, more economically vulnerable folks catch pneumonia. in the downturn half of those income gains were reversed and three-fifths of that decline was people crossing the extensive margin the other way. so the cost to slack in the job market are particularly born in this group, but the benefits are quite pronounced. we show the most recent period of tight labor markets and while the gains aren't as dramatic as they were in the hot '90s labor market, they are comparable and economically significant and large. very important to these folks. now what i have here is just the
12:09 pm
unemployment gap and thanks to my excellent r.a. i was able to do a dynamic slide. here are just some recent headlines. these are headlines over the last few months. tight labor market, disability may not be a barrier. i thought i was done, open doored fdoor doors for convicts. lower income americans are increasingly job hopping to tap some wage increases. in conclusion, i want to thank the federal reserve for recognizing this opportunity. if you listen to the words of jay powell and other former fed officials here who said much of the same thing, you will hear him expressly making these connections. so i wanted to thank the fed for that, for tapping those benefits and i also want to thank the american people for their anchored expectations -- [ laughter ] >> -- which are key to all of
12:10 pm
this. thank you very much. [ applause ] >> okay. so i just want to thank our panelists again for those great presentations. i have a lot of questions. we probably won't get through them all today. but one of the things i am curious about is how you're interpreting what's currently going on in the economy. so i think you both asserted and we've heard other people this morning assert while the price phillips curve seems to have
12:11 pm
flattened out a lot, there's still some action in the wage phillips curve. but when i look at wage growth now, it looks to me that it's running at about the pace you would expect given productivity growth and inflation. so that doesn't actually seem to be suggestive of a hot labor market, which is how a lot of people are characterizing it. is that some at tenuatitenuatio curve? >> my presentation was more focused on the long-term changes, but it has been a constant question why isn't the inflation back to target, why aren't we at 2%. and if you remember, the chart i showed decomposing recent movements on inflation can explain why inflation was below
12:12 pm
the baseline forecast during the recession as the unemployment rate was high, it has gone back to zero so we don't get anything from slack at this point, which is understandable. but we are still not back to zero. what has happened, if i had time to shot another chart, i would have shown one that not only it's the structural effect of the unemployment rate but also from import prices. that could explain a lot of what happened in '15 through '17 and '18. so it doesn't get us all the way there. the third thing on which i didn't have time to focus was on the trend too much. if you go back to the first chart, you will see the trend and the forecast is not 2%. it's more something closer to
12:13 pm
1 3/4. even though the trend has been very stable, it's not been at 2%. that's my few explanations for the last few years. >> i think there are three explanations to your excellent question, because i think you're right. the first is that there's been a little stronger wage growth at the bottom. you're citing the average. this is consistent with my earlier point that the elasticities of wage growth relative to the tightening job market are stronger at the bottom of the scale. the second point is that probably maybe the most important is that probably the job market isn't really quite at full employment yet. that may sound somewhat controversial given how low unemployment is relative to these estimates of u star. the just based on the inflation data and the wage we can get as complicated as we want but the
12:14 pm
first simple realization is that we're probably not yet at full employment. the third thing is worker bargaining power has been so severely diminished that it's not going to take just low unemployment but low unemployment for a very long time. the share in the private sector is somewhere around 6 or 7%. >> okay. thanks. so another question i wanted to get at was to try to tease out a little bit more what you think is going on about wage pass-through into prices. some of our other speakers this morning have talked about it but looking at it from the perspective of the labor market as well, we still see some evidence of a wage phillips curve, much less evidence of a price phillips curve. prices have picked up a little bit. wages have been pretty flat. what are your explanations for why these changes in wages aren't translating more into price inflation?
12:15 pm
>> again, we'll build on the research we've done on the pass-through from wages to prices. because doing what we do at the board, this is a question we vrch get. we struggle to explain why we don't include wages on the right-hand side of our price phillips curve equations. we began looked at the pass-through from wages to prices over time. depending on what measure you use, if you use the more comprehensive compensation, you could see a steady decline over the years to the point where you don't see a pass-through from wages to prices. if you use the employment cost index, it's more stable but not very big. okay. now, this is not to say that wages don't matter for prices. they should matter. they are two-thirds of production costs. but i think what we found in this paper that if you plot in
12:16 pm
addition to the stochastic trend, if you do a similar one for unit labor costs, you will see they are both very stable. and in this stable stochastic trend environment it's very hard to identify movements in wages that are translating to prices. in fact, what you find is that year to year movements in price inflation can reflect slack and can reflect import prices or other idiosyncratic factors, but it's hard to identify movements from independent pass-through from independent wage movements to prices. >> i don't have a ton to say about this but two points. one is i really do find this relatively new literature on the
12:17 pm
impact of firm concentration within key industries to be relevant to this conversation, to this question. you see this in retail, in health care, in technology. i sort of grew up in an economics where mononpolistic firms took hold. you're seeing it on the price side. you're seeing it on the labor cost side. some good papers where firm concentration seems to be correlated with diminished labor share. so using your monopoly power to kind of screw workers instead of on the price side. you know, i obviously think that's problematic and i think the tight labor market is starting to push back on that even though labor share is still uncharacteristically low even with the revision that took it up a bit. it's still uncharacteristically
12:18 pm
low. i always tell this to my friends on the fed who will listen to me, important caveat. [ laughter ] >> that there is room for noninflationary wage gains through a rebalancing of factor shares, through catching up. that's not a slam dunk because firms will resist that and may try to pass it through on prices. we haven't seen a lot of that, as your question implies. >> i guess related to this, we've heard some talk this morning about whether we should be focusing on alternative measures of slack. the unemployment rate was never a comprehensive measure of slack. it was more a sufficient statistic for slack. i think this does raise the question of whether it no longer serves that role as a sufficient statistic for slack and where we need to be looking at a more comprehensive measure and we've getting a mixed signal about how much slack there is. >> i think that is a fine point and we should all try different measures in our model.
12:19 pm
it's kind of overblown because i recently did a blog on this where i just correlated the unemployment rate with all the other measures i could think of. and the correlations are really, really high. they're all above .9 for the most part. my friend john roberts has done some work on sort of throwing a bunch of these slack variables into the washing machine and seeing what kind of comes out. i believe the unemployment rate kind of dominates. so yes, sure, other measures, but i wouldn't spend a ton of time on that. >> i'll just reiterate the point from my presentation that i think it's a very valid question and certainly one we should be looking at. but the fact that we don't have the flattening in the wage phillips curve means there's something else going on in the price phillips curve that does not come from bargaining
12:20 pm
necessarily, bargaining power or the right measure of slack. >> great. so the last question i want to ask before i turn it over to the audience is, actually, katia, i'm going to pick on you a little bit. jared came out very strongly saying not much inflation, phillips curve seems flat, we have an opportunity here. you were sort of expressing i would say some concerns at the end of your talk. i was sort of curious, like do you think we have the space that jared asks, knowing you're not speaking for anyone in the federal reserve system, or do you think that the risks sort of dominate? >> since i'm not speaking for anybody, i would say how i spent my days at the fed and in the morning i worry we are never going to get to 2%. the underlying trend is lower. we have to do something. in the afternoons what if we
12:21 pm
overshoot. so i think i am worried that, for example, if you look at the michigan survey, expectations have moved lower. some would argue that's because people are just waking up and realizing inflation was never 3%. but that might be just a reaction to persistently low inflation that we've seen. it would be great to push those up, but only by three-tenths. if we can do that, that's great. my concern is how confident are we that we can -- again, i am worried. >> i don't want to create the misimpression that i'm operating from a, what, nothing to worry about kind of perspective on the threat of future inflationary pressures. i mean, i guess the thing i take
12:22 pm
from the conference so far and all of the reading i've done in recent years on this is that i don't think we have a great -- and i think i've heard janet say something similar. she'll correct me if i'm wrong. i just don't think we have a great understanding of what's driving inflation dynamics these days. that means for now based on the empirical evidence i and others have presented we can tap the benefits that i was stressing as being so critically important to tap given the hour of what's be fa befallen middle and low income folks. but when you don't know, what you don't know can help you, but it can also hurt you. before we get to the questions, we have three labor market analysts up here. i don't think you should just ask questions. i think you should have to answer. stephanie et al did a really important paper on the impact of tight labor markets recently.
12:23 pm
i think it was a brookings paper. one of the things you talked about was something very important that janet referenced in her comments. this is reverse historesis. what janet was referring to was the possibility that my slide showed all these people coming into the job market and yielding benefits. historically there's been a problem of last hired first fired. people get into the job market. they don't stay in the job market. the idea is that perhaps tight labor markets can improve the supply side of the economy by pulling people in in a lasting way. i really want to hear what you have to say about that possibility. >> it's true. so in this recent paper that i worked on we did find a little bit of evidence that particularly along the participation margin, the extensive margin that hot labor markets do pull people in and it has a bit of a persistent
12:24 pm
effect. i think our result was consistent with some of the micro literature. we use time series data, which i think have their limitations for looking at this issue. but it's consistent, i think, with the micro literature which has found some short-term effect that when people are pulled into the labor market and become employed, that they remain employed for a bit longer. so there are some lasting effects. as you said, this question is very crucial. if there were really a significant effect by running the economy hot we could really boost labor supply, that would reduce the tradeoff that the fed would face in setting interest rates. the one thing i've been thinking about is why aren't these effects bigger. i think we all have the intuition that once someone comes in they gain experience, they create networks that they should have a more long lasting tie to the labor market. part of the issue is what you said, that the economy largely
12:25 pm
operates as last hired first fired and people tend to be hired into very cyclical industries because those are the ones that are expanding employment a lot late in the cycle. and also then we don't run the economy hot for very long. the periods in which the unemployment rate is below the natural rate, they just aren't that long. i think if you look to some other experiences in history that aren't dependent on the business cycle, you actually see better evidence of hysteresis. that really significantly boosted women's labor supply. it was a permanent effect. lots of women left the labor force in the '50s but it changed the way families operates and changing the structure of work and institutions.
12:26 pm
there is a powerful case that his ree sis can matter. >> the macro economy helps them jump over those barriers and the macro economy goes south, we're going to have to do stuff to keep them in the labor market including job subsidiesubsidies training programs and things of that nature. >> let me see if we have time for a few questions from the audience. let me just gather a few. danny, i think i see your hand. >> i want to answer your question about what variables actually int lly enter wage equ. i've worked on this the last 30
12:27 pm
years. post 2008 there's one variable that works and nothing else does. the unemployment rate doesn't work. u-7 works. what u-7 is the thing that drives u-6. the reason it's a quite good variable which explains why there's no inflation is it hasn't been reverted in any country in the world. when you run a wage equation in u.s. post 2008 the unemployment rate doesn't enter -- sorry, jared, it doesn't -- but the underemployment rate does. that's the answer to your question. i have several papers on it. it's the only variable in the u.k., in the u.s. and in panels of european countries. so i don't buy your comment that the unmoiemployment rate -- >> i'm sympathetic to that. i'm a little embarrassed because we published one of your papers. >> i know you did.
12:28 pm
>> with great respect for you and your work and i think you're largely right, i do get a little nervous when economists glom onto one variable that they love. that often doesn't end well. [ inaudible question ] >> i just want to counter your question on why the phillips curve for wages is steeper than the price. central banks are targeting inflation so to the extent they're successful at it, inflation will be fairly constant and uncorrelated with everything including slack. now, you might wonder, okay, why are we slightly undershooting in the case of the fed, not the u.k. the fed is not targeting wage inflation and so that relation is freer to some extent. >> makes sense.
12:29 pm
>> so the lower income folks, they can't get unions to help them, so they go to the government. the government populist enough to really contemplate a minimum wage increase. the lowest income people arguably would be the most affected by inflation too because they spend a greater portion of their income of necessities. how high can we put the minimum wage without jeopardizing their security at the bottom fifth? >> yeah. i will speak to that but first let me point out one of the reasons i really enjoy hammering on this full employment, the benefits of full employment piece is that it is pretty nonpartisan. everybody should support that. it's not government, you know, necessarily doing stuff. the fed is in the mix, no question. but it's really markets working for low income people. you know, i guess from the evidence i've seen that you
12:30 pm
could go to $15 on a minimum wage which is something that's been proposed, but you'd really have to phase it in over a long period because there are places in the country where $15 an hour gets you to probably in the 40th percentile wage. you'd want to phase it in slowly. i think the research would support that idea. >> okay. i'm going to have to end the panel now. thanks again very much to our panelists. [ applause ]
12:31 pm
12:32 pm
>> thank you all for continuing the conversation. it's almost hard to believe there's anything left to say after those panels. so i was going to propose that we do this whole thing in wrap. [ laughter ] >> or maybe in finnish.
12:33 pm
[ laughter ] >> but i thought that was a little unfair given that i haven't given the panel advanced warning of that. >> that's right. >> so i'm joined here by olivier now at the peterson institute. loret loretta mester. and paul krugman, the economist at city university of new york and columnist for the "new york times." krugman is the guy who makes all journalists nervous because he seems to be more productive than the rest of us and that's kind of frightening given that he does all these other things on the side, but we ask discuss those things later. i wanted to start by asking each of the panelists what they took afwra t away from the conversation we had this morning. i want to start with olivier who
12:34 pm
makes the observation in the slide you see behind us that's been made about how the wage phillips curve behaves perhaps as one might have expected, but prices aren't rising. so the question is what the hell is going on? >> okay. so i had prepared a slide in anticipation. i think it has survived the previous three hours fairly well and coincides with the general message and tells us where we need to look and what implications it has. so the diagram is familiar except it's not in familiar form. it basically has the inverse of the unemployment rate and the inflation measured by the employment cost index. you could plot it in a scatter diagram but it's there. i think that's consistent with everything we've heard. now, what people have put on the
12:35 pm
right-hand side typically is the phillips curve, price phillips curve. what i've done instead is plot, again, the wage inflation using the employment cost index which is the blue line and the gdp price deflator. for the purposes of thinking about mark-ups, it's clearly a much better variable than the cpi. it's visually striking, which is that the wage inflation was kind of okay from the previous graph. that doesn't seem to be much of a relation between the gdp price deflator and the eci employment cost index. it's really here the disconnect. people have talked about the low pass-through. it's true but there is much more than that. there's a lot of variation that in the price index which is not coming from the cost index. so starting there, my reactions
12:36 pm
to today, i think the first puzzle is the slope of the relation between wage inflation and unemployment. and on this i think we're not clear as to whether the slope has really declined or not. my sense is that it has, but somebody argued that if you do it right, it really has remained more or less the same. i think it has decreased. the explanation is a very plausible candidate subject to kind of a test that i suggested. but it seems to work. the other is that maybe there has been a change in the bargaining structure on wages. we think of bargaining as splitting -- if the workers are already at the bottom end, there's nothing that can be done to decrease their wage. if everyone was paid the minimum wage, then we would find no effect of unemployment on wages.
12:37 pm
i think there's something like this happening but i don't think we solved that one and we need to do it. on the markup in the second graph, what is it, what is going on? i suspect measurement is a big part of it. the more you know about the gdp deflator, the more you worry. we've heard various explanations about pricing. i think some of them say, well, you know, some sectors really have to take the international prices given and, therefore, you would not see the kind of pass-through. that's not true for all sectors. i think here we need to go sectoral in the same way that was done for labor and try to understand it. again, i think there's still work to be done. on the past implications which is the theme of the panel, the stability of a wage phillips curve and the instability of the price phillips curve has a fairly big implication which i think hasn't been examined and should probably have been in the context of thinking about what
12:38 pm
the fed should do, which is to have a wage inflation target rather than a price inflation target. i mean, it clearly is much more related to labor market development. so from just an empirical point of view it seems like a better measure to actually look at. but from a normative point of view, if you take -- the markups reflect larger distortions, then actually it's a good idea to take a measure that doesn't have a markup in it, namely the wages, and basically focus on this. if you see wage inflation at 3% and you see productivity off at 1%, then you're home. even if gdp -- the cpi moves around. so we had talked at the dinner with janet and we said we would write a paper together. janet has been a bit busy, but i think that's still worth exploring. also telling people that the fed
12:39 pm
cares about wages and has wage inflation as a target is probably a plus in addition. >> does that mean you'd be tightening now because wages are rising? >> i think we are more or less at full employment. wage inflation is around 3 something. that's consistent with the price level correctly measured of about 2. yeah. i'm more on the hawkish side these days. >> oh my goodness. >> than jared. >> loretta, so i've noticed that in a number of your presentations over the recent months, you have a lot of things. if it's the case that nonmonetary structure factors are holding back measured inflation, that's a very different implication for policy than if it's an aggregate demand shortfall. i'm curious what sense you make of all this evidence we've heard this morning on what's driving inflation and the work that you and your people have done at the
12:40 pm
cleveland fed. what explanations do you buy for what we have seen? >> i grew up in baltimore. hl menkin is from baltimore. he said this saying that for every problem, there's a simple elegant solution that's wrong. i don't think there's a simple one-word answer to this. our inflation research center is doing a lot of work in trying to understand inflation dynamics. we have results that basically say if you look at the tru structural part of inflation that is related to the labor market, the cyclical asik acyclical part -- >> health care which is the big part of it, other parts that aren't really related, that the cyclical part is only 40% of inflation now. part of what's going on is that
12:41 pm
you have idiosyncratic factors that affect the labor market and affect the inflation rates. if you do just look at the cyclical part, it certainly is correlated with tightness in the labor market. i believe there is still this relationship and that we can use it to help predict inflation. that said, you have to recognize there are these acyclical factors affecting measured inflation. i think you want to be careful about trying to explain everything with what's going on in the labor market. that means that if you see a shock, it may take longer to get back to your inflation target than it would otherwise. from a policy and explaining communications point of view, you may want to be thinking about a range as opposed to a point target, because you have these shocks that are going to move inflation off your target even if your trend inflation rate is moving up. so there are implications for
12:42 pm
how you communicate your target. i think the results we saw today on inflation expectations were very provocative in the sense that it makes you think about a different way of communicating and perhaps a different way of thinking about the interplay between inflation expectations and your target. so from my point of view, you might have shocks that move inflation off your trend, but the trend underlying inflation rate is going to be driven by inflation expectations. so i agree with the remark that janet made in the beginning that the stability of inflation expectations is key to allowing you to be able to run your monetary policy. i think the results that talked about the amazon effect and the fact that pricing models are changing are very interesting because if it's true that prices are becoming more flexible, if you think about in a dsg model why is monetary policy able to
12:43 pm
cushion against these shocks on the real side, it's because of sticky prices. if prices become less sticky because there's more frequent price changes, that means monetary policy is less effective. or said differently, you have to move your monetary policy more to have the same effects. these changes, these underlying structural changes i think are more important than just how do you measure inflation. they may actually change the transmission mechanism of monetary policy. >> ben, are you convinced that something's changed in the structural side of unemployment? or is it just that for some accident of history we've had wise monetary policy makers and that everything's all the same, the only thing that's changed is the quality of monetary policy? [ laughter ] >> they're laughing but that's actually my opinion. i think the most important factor over the long haul has been changing not wiser monetary
12:44 pm
policy but a monetary policy focused on anchoring inflation expectations. in a paper they studied the dynamics of inflation using the unobserved components model. 30 or 40 years ago if there was a shock to inflation, a significant part of that shock was permanent. it would stay away from its initial point for a sustained period. whereas since the '90s, if there's a shock to inflation it's transitory and you go back to the underlying level. that's consistent with a world in which inflation expectations have been well anchored. shocks to inflation tend to be transitory as long as policy is consistent with that. that is, by the way, it's very clemen complimentary. if inflation expectations are well anchored policy can very well succeed at keeping inflation at target.
12:45 pm
and vice versa. i think that explains the broad changes in the inflation process. it probably helps at least to a first instance explain the most recent behavior, part of the reason inflation didn't fall so much in the recession is because inflation expectations were well anchored on the downside. they did fall to some extend which is probably part of the reason inflation has been slow to come up. having said that i think there are a lot of points made today. for example, the fed underestimated how far the labor market could be pushed. u star is probably lower than the fed thought a couple years ago. that gives some scope for further expansion in the labor market. some of the points that kristin made about the global factors that kept inflation from falling quite so much after the panic, i think, are relevant and the
12:46 pm
transmission of that to prices. in any sort period it's going to be a bunch of idiosyncratic factors and others that are relevant. over the longer period i i do think the change in the monetary policy is the most important thing but the other things are relevant too. >> paul, what sense do you make of all this? >> yeah. maybe the most important thing is that we actually basically understand the economy a lot less well than we thought we did, which has a huge bearing on policy. i actually wanted to say something about oh lilivier's proposal because i have thought something along the same lines and then kind of backed off it even though the economics seems to be totally right. here's the point. >> talking about targeting wages. >> targeting wages. there is one thing, one huge success story of sort of c
12:47 pm
conventional monetary analysis over the past ten years has been the distinction between core and headline inflation. i'm sure ben remembers this even better than i do. we remember 2010, 2011 when headline inflation up mostly because of oil prices. the fed was saying, calm down, it's core inflation that we should be looking at and was totally indicated in that. there's stuff going on even with the core inflation measure. if we think about conceptually what we mean by core, we mean stuff that is kind of sticky. the kcoriest of the core is in fact wage inflation. it makes a lot of sense to target wages, except i'm trying to imagine the situation of the chair of the fed saying we are raising interest rates because wages are rising too fast. [ laughter ] >> i think it's good economics but the political economy is just going to be disastrous.
12:48 pm
>> saying we are committed to make sure that wages increase no matter what at 3% a year, terrible. >> i think it's not just 40% of fins but a large portion of the population as a whole. >> ben, your body language suggests you're not ready to endorse the blanchard rule. >> i was having the same reaction that paul did which is i wouldn't want to sit in front of congress and say we're very concerned that wages are rising too quickly. so that's a concern. i think it's a complicated question. there's elements of the price process which are independent of labor markets and i think need to factor into the monetary policy as well. >> it seems that at the same time as we've had a long period of low below target inflation in many countries around the world, we've also seen this decline in
12:49 pm
our star, the so-called natural rate of interest, the interest rate that will prevail when all is calm, which as has been pointed out, we don't know what it is. but there's certainly a widespread consensus that it has fallen a lot. it seems to me -- two questions. one is, are these things related? and secondly, this is kind of challenging. how should monetary policy think about a world where inflation, at least for now, seems to be persistently low and the natural rate of interest seems to have fallen and is also persistently low? >> i have a somewhat facetious answer to the first part, which is it looks like u star and r star if such things exist, if you ask why has u star gone down, the answer is we really don't know. if you ask why has r star gone
12:50 pm
down, we really don't know. so the common factor is both of them are caused by we really don't know. [ laughter ] >> i thought the catch-all answer had something to do with can make a story along those lines, but the fact that wage phillips curve is holding up, it is harder for the demographic model, but fit is an older and more experienced workforce, that should be showing up, and this is the less dramatic part of the story, and i buy the demography on the cycle stagnation story, and i believe that is right. but what i would take from it with the general spoi lesson is that it is quite possible and quite a lot of slack and advanced in the whole, and looks like there is, and at the same time very little monetary space in the world of the whole which should give us a lot of anxiety and we should worry a lot, because if we are in the situation now when nothing
12:51 pm
really bad has happened lately, we are in big trouble, if you know, sooner or later something really bad will happen, and so the two factors together mean that we should be worrying a lot about what to do to give us some more space for policy response which is going to give us some of the other questions about the inflation targeting and fiscal policy. >> can i go back to the r-star and the u-star conundrum, and it is a little bit worse than we don't know. >> great. all right. the weakening stagnation seems to be better for the opposite side, and the risky arc, and so the decrease in r-star is even more of a puzzle. >> great. to loretta. fortunately, you have to make
12:52 pm
decisions without all the information. >> well, you always have to make decision without a formula and i want to push back on the slack out there, and if you are talk to the firms and we had a paneling yesterday at the cleveland fed, they routinely tell us, and this has been going on for several years about the difficulty of finding workers that they can hire. you push back and say, are you raising wages, and they say yes. but now they are changing, because they it is not worth hiring out of the pool out there, because when they do bring them in, they don't stay on staff more than a month. so i say, well, do you know where they are going, and they have no idea. it is not bid away for other wages, so what is the response now? the response now is that they are going to be automating more and that is what you are seeing
12:53 pm
in the firms, so i would say that we have done a lot in terms of running policy much more accommodative than we would have in the past. partly, that is because we have brought down our estimates at u-star, and so, in saying, if you were going to say relative to the old u-star we are running the economy very high and the policy is much more accommodating than it would be on the old traditional taylor rules, so on that sense we have taken on board some of the things that you talked about in terms of we have to rethink when we don't see wages going up or we don't see the inflation going, we have to rethink the policy rule, and i think that we have done that, but at some point, there are unintended consequences of doing a policy like that, and so, you may be ending up affecting long-term employment growth because if you automate more in the firms, you won't have the demand for labor is high. so i am sympathetic at wanting to be at maximum employment, and
12:54 pm
this is certainly the goal. i think that the policies that we have in monetary policy is not going to affect the overall labor market in the way that you want to. i think that the things that you pointed out in terms of the things that we talk about when we go out into the district, right, job programs and getting the people the skillsets that they need for running those robots coming into the plant, and all of those things are going to have much bigger impact on public welfare in terms of getting people to work. there is a whole other set of issues of how to get people to work which is another issue, and transportation is an issue, and there is a lot of structural things that have nothing to do with the monetary policy that can improve labor market outcomes. >> i don't understand the problem. so you run the economy hot, and workers that we pull lots of workers into the market, and wages start to rise and firms start to automate which is increasing the pace of productivity growth and you are worried that we will run out of
12:55 pm
jobs? >> well, i worry that there is a whole segment of the population, and right now, they don't have the skills in demand and what is happening to those people unless you are training them for the modern jobs, you are in trouble. you are shifting the district from manufacturing to other types of jobs. there is on this match between the skills that people have that want to enter the labor market and the skills and demand unless we are doing something about that, it is very hard, right, to improve, you know, outcomes, and so i think that is one of the things that the fed listened in conferences around the district. we have heard time and time again. they are not that interested in hitting the 2% inflation target and they are much more concerned about am i going to provide the food on the table for my family, and so i think that we have to think about what tools we use for which problem we are trying to effect. >> and ben, can i steer back to the question of the challenging combination of the low, neutral rates and low inflation? what is the right response of
12:56 pm
monetary policy makers and does it matter where the starting point is, with our inflation close to 2%. japan, it is their inflation expectations are below 1%, and what is the right answer? >> let me start by saying that the decline in r-star is mostly understandab understandable. i think that demographics and the global savings and investment patterns and all of these things that people have talked about can explain most of what is going on there, and i'm more puzzled about the u-star decline than the r-star decline, and i understand that it is outside of the monetary policy, but it is a constraint on the monetary policy, so i want to do the half empty and half full response to the question. this is a correct point that historically the fed has cut the federal funds rate 5.5 percentage points in the recessions, and now we only have just those points of percentages
12:57 pm
given guidance or quantitative easing, and plus thinking about the policy. altogether, they add about three percentage points of space. so my stance is that as long as the nominal interest rates are 2.5 to 3.0, the fed can do most of what it can do at any point in history in terms of the user policy rules. so say we are low in terms of the united states is wrong. but you can talk about the terms and the fiscal policy and the like, but i think that, you know, asking for 5 and 6 percentage points of space and other tools are effective and if appropriately used can
12:58 pm
compensate for some of the loss in the r-star. >> but the other tools are partly used, and some of it has been used already. >> yes. >> and so some -- >> well, it can be used better and more aggressively and now i am talking about the next recession and where the fed is going to be starting where it is, and if there is a medium-sized recession. >> you think 3% percentage pois of regression, and paul, are you as sanguine as that? >> well, because ben knows more about this than anyone, i have been a skeptic. the fact that we are not worried about the effectiveness of qe after all of these years is a persuasive evidence that the evidence ain't that persuasive. and so, it seems dubious, and let's also say that we shouldn't have too much of the parochial u.s. perspective. i mean the other large advanced
12:59 pm
economy the eurozone, how much room do they have for that? a large part of the world is in dire straits here, and we don't know the size, and yes, 5.50 basis points has been the historic cut, but it is way inadequate in the last recession that we had. and we don't think that we will have soot 2008-style shock, and of course, we didn't think that was coming either, so i think that we have still a fundamental issue that we don't know that we don't have a policy issue, because there is a substantial probability that we could. >> are you in agreement? >> well, i was saying much more difficult situation, and particularly in japan, and that
1:00 pm
is going back to inflation expectations, and being inkored, et cetera, and one of the reasons to be aggressive is that if inflation expectations get anchored at a low level, then you are in serious trouble. and so the japanese situation is much different than the u.s. expectations, and in europe they are low, but not as low as japan, and so there is some probability that they will be moving away from the current situation and achieve a more normal stance, but i don't disagree with paul that europe and japan are much more difficult in terms of the federal reserve number.
1:01 pm
i thought that you were going to throw me off with my issues there. >> well, i think it is a reasonable reasonable proposal and should be discussed as such. but now the announcement today has zero credibility. i don't think that it will. and so that is not the time, but if you will invite me in a few years. >> yes. on that we have research coming out of the cleveland fed linked to this. so, it is, you know, the argument that you are raising the inflation for more space seems salient and hard to argue about it, but are the terms going to be more flexible in the pricing, and there is evidence that steady inflation means that the firms are going to change the prices more often and goes back to the original point that
1:02 pm
monetary policy is going to have to act more aggressively for the impact. so you are not getting as much space as you might think in terms of moving it within the model. >> i think that you have it wrong which is suppose that the prices are advancing very fast, and in principle, things would have less need for the monetary policy. >> unless a downturn and less space, you know, in order to get two percentage point of space, you have to be more aggressive on the policy actions, so it is the transmission mechanism is affected because the prices are changing. >> and again, the recession and according to the models that we have, the recession would be less bad. it is underused i think. we can compound it though, right. >> and so if i could say, too, that truthfully, you are saying it is not the time to talk about it, and i agree that the boot strapping and to be completely
1:03 pm
responsible from the years ago and there is no evidence that it works. if we are thinking of laying the initial groundwork for what happens the next time that things go really wrong, it is maybe a good idea to keep alive the 2% inflation target which has a peculiar history, and based on probably just happenstance and the assumptions that turn out to be entirely wrong, and that it was not going to be a good idea, and that comes the next global slump, and so, we suddenly have a fiscal germany and president bernie standards and the united states, and we have chance to have a fiscal monetary expansion that the idea that stopping at 2% inflation is a good idea and a good idea to have out there. >> all right. now, let me ask you about another thing that people talk
1:04 pm
about, and some people say, yeah, basically, we don't have much inflation, but, and that is terrific to enjoy the fruits of the hot labor market. and so we are sowing the seeds of the next disaster by not taking into account enough the stability risk that we run by running low interest rates for such a long time. is that something to worry about, or is that something that women worry about that are going to have nothing to worry about. >> well, so, what we have learned over the past, you know, several years of that, we do have to take financial stability concerns into account when we are thinking of monetary policy and that is not to say to add it to a third mandate and something to take into account, and you have to recognize that if you are running low interest rates and you may create some financial imbalances that may come back to haunt you in the future, and so i think that you
1:05 pm
have to sort of understand that those can occur and nelly when she was at the board, she developed the whole structure for allowing the fmoc to examine those imbalances and actually monitor them. >> and how much of a worry would you say today? >> well, some issues in terms of not financial debt levels being high, and commercial real estate pricing being high, but, you know, there are moderate levels that we can handle them, and it is something that we can think about partly because in theory, we would like to use the macro prudential rules to worry about the financial imbalances and mandate goals and in reality that is the hard thing to do, and because we don't have so many, and the fed did a table top exercise to have a scenario where the financial committee is
1:06 pm
in the stat, and the working paper is out in the cleveland website that is talking about those that is the tabletop, and it is really clear that when you get into the situation where, you know, financial instability issues come to play, we don't have many tools that we can actually use to address them. >> ben, there was some communication about the fed setting expectations, and michael weber made the case that you have to think about household and firm inflation expectations and just encouraging participants and investors is inadequate and he pointed out which we all know it is hard to communicate this stuff to the ordinary people and even the ones in the top half of the iq distribution. are you pointing at me?
1:07 pm
you are at least in the top ten [ laughter ] do the central banks have to worry so much to adjust the consumer expectations or can you get that way by influencing the markets which are influenced by the fed. >> let me say that i am very happy about the desire of the fed and other central banks to do more outreach and to listen to broader public, and the fed is listing series for example and these are very important institutions and they have to be accountable and transparent and need the public to understand what they are doing, and there are limitations to that obviously, but i think that it is very important to try to communicate broadly. from the technical perspective, i think that you do get a significant way there as long as the markets understand that the goals and maybe some of the pricing executives and the people who are in charge of the pricing and the wage setting, and this is going to give you a pretty long way, and the frbis
1:08 pm
model has model consistent expectations, and everybody knows what the target is and behaves accordingly and the second asset setting where the bond market understands what you are doing, and then it turns out to get 2/3 to 3/4 of the way there in the second scenario, and so, you know, i am sure it is the case that a large fraction of the population does not know what the federal reserve is and much less what it is raising as the targeting, and the fed and the other banks would be well served to increase literacy about these things, but i don't think that even some of the policies will be talking about, i am sure, you know, the price level targeting and some of these policies that rely to some extent of the forward looking behavior and simulations and other analysis suggests that you get a good bit of benefit even if the minority of the
1:09 pm
population understands what you are trying to do. >> and so on demand, monetary policy works basically through the housing and the exchange rate, and so, if you can get to minds of the people who set mortgage rates and exchange rates, and this is ul that matters. >> it is not quite that far. >> and the first respect to inflation, and that is the matters of the expectations of the firms, and so it is criminal that we don't have a good survey. >> criminal. >> yes. i mean, it is absolutely needed, because when we are look gt the household, and the expectations play a role in the determination of the wages, and they have the wage, it is for firms in the hr department and so we should have a survey of the hr department of inflation expectations, and this is an important thing to know, and it is not in the u.s. or the
1:10 pm
new zealand example that i referred to this morning is not convincing. i am sure that the people in the hr department are more careful in the household. >> i wonder if before we turn to the audience if paul bannon and loretta could give opinions on what to do with interest rates and loretta needs more to think about, and this monetary framework is already ruled out olivier's raising the inflation target option, but briefly, what is it that you think that given all of the things that we have talked about today what has happened with inflation, and what we understand and what we don't understand, and what is the right thing to do with the framework to put us in a better place than we are today. do you want to start paul? >> no.
1:11 pm
i have to admit that i am baffled, because it feels that the various things that we, our understanding of the economic system such as it is, all seem to be impossible, out of some combination of the politics and this come strainnstraint, and w persuasive, but off of the table. waste targeting is extremely persuasive, but i would say it is off of the table. so i am not at all sure. i am not sure what the answer is. i am supposed to be saying something helpful here, but sometimes we have talked to developing countries and trying to go through the policy options, and they say that would be impossible, and you take out all of the constraints and there turns out to be no space. >> could be. ben? >> the federal reserve is talk about the makeup policy approach and the idea being that if you fall short of the inflation, the target for a period of time as you have recently, there would
1:12 pm
be some compensation for that. the debate is turned on technicals and if it is credible, but there is a relatively simple variance of qualitatively possible that we get a lot of benefit, and for example if we look at the technical variance, and i have proposed the targeting where it is a zero bound, and you fall short of the targeting, you compensate to some extent, and the version of that which is working well is the level price targeting and one year lookback, so in other words, you only keep the rates at zero until you have established over a year that you can hit the target. and so what you can explain with that qualitatively is saying that when we are going to the zero, we want to establish and like the european central bank and robustly converging to the target. we don't want to start the process of normalizing the rates until we have clearly established that we can meet the
1:13 pm
target over a year. that is a fairly simple explanation, and one that would have most of the benefit, because the bottom market would appreciate that the fed is very slow at beginning to raise the rates from the zero balance situation, and does it involve complex pricing calculations and things of that sort. so that is an example, and the makeup policy approach can be done within the context of the current inflation target, and simply defining what it means to hit the target, and in particular that hitting the target just momentarily is not enough, and we want to establish that we are sustainably hitting the target, and simply doing that would have some of the makeup qualities that these optimal theoretical approaches would have and not involve any of the major change, and not create any political scrutiny or confusion, and so that would be my suggestion. olivia. >> so i think that we are
1:14 pm
flirting in the world with stagnation, and here in europe, we have married it. >> right. >> and so in that world the policy is going to be very constrained, and we see it in japan, and we see it in europe. i support what ben is pushing, but i suspect it is very second order relative to the size of the task. so, that maybe eventually you will predict what i am going to say. what about fiscal policy? it seems to me that fiscal policy in the same context is needed. it is cheap. >> in rates alone? >> yes. the cost of debt is very low. it can be used infinitely better. in the u.s. unfortunately, no doubt that our president and our congress will use it probably not right, but use it.
1:15 pm
in europe, it can be used to convince the governments to do it, and it seems to me that the solution is not to do it with more twists on the money, which i think that we have to and not hope it is going to be the solution. the solution is to have a more expansive fiscal policy. >> you want to weigh in on it? >> just the way you think about the inflation expectations is to be well anchored to allow the monetary policy to counteract negative shocks, and for the fiscal policy, same thing, and like if we have fiscal policy on the sustainable path, then you could use it in the way that we would like to use it. and the same issue just like we have to run the monetary policy to keep the inflation policies an cord and it allows us to use the monetary policy more effectively, it is the same in the fiscal side. so we are in a sort of a dilemma
1:16 pm
here. >> time for the questions, and again, raise your hands. stand up, tell us who you are, and be as brief as possible. there is the one here. okay. and then i want to take three and we will see where we go from there. >> this is for the whole panel. >> tell us who you are. >> grace yu, and do you see a future of much more coordinated monetary and fiscal policy, and what would be monetary policy be like in that region. thank you. >> over here, sharon bernstein. >> sorry, jared bernstein. and so instead of wage growth how about targeting labor share, and there are times that you want the labor units to be causing to rebalance factor shares that are whacked in the downturn as we have seen. >> okay. pass it behind you.
1:17 pm
>> thank you very much, mathias matisse, and i teach economic policy across the street, and my question is for the current and the former central bankers really. i can't think of in a decade where the central bankers have been so attacked by politicians like mark carney or the brexiteers or marco draghi or other politicians. is this something that you fear is going to be worse or something that will go away when we get back to more normal times of the economic policy or something that we need for a more central independent economic policy. >> so start with the economic policy, and how does that look? well, i learned everything about the crewman, and they told me it only works as the inflation goes up, and as long as the inflation target is credible.
1:18 pm
you might as well do the fiscal policy and have the monetary policy to be supportive in some way. >> yes, that is right. >> and so there is an idealized by, call it the branch 4% target, and although that is me too. and so, we have a coordinated fiscal expansion to get you to the point, and which is modified by the monetary and that is not anywhere on the table. so if we talk about the actual coordination that might happen, i worry about who is coordinating with whom, and so i would be perfectly happy with mario draghi setting the policy and not happy with the german government to set the ebc policy. >> olivia. >> yes. i think that the two dimensions of the coordination and within the country fiscal and money, and that can be done, and in a very simple coordination and you
1:19 pm
do something, and you tell them to do something, and so i had just been told thou react it to, and do the effect on the outputt if it is negative and it does not imply more than just that and greenspan is an example of that, and coordination across fiscal policies and across countries, and that is a much more difficult issue in the eurozone, and there again, and either the eurozone growth zone recession, and the coordinated fiscal response will be too weak, because when belgium expands, it is not going to get much of the reward. so there, i think that there is a need for coordination. and what happens or not, it should. it probably will not. that is going to be my answer. >> loretta, and when the question was asked about central bankers being attacked, it is not true, because i had visions of paul volcker storming into the room saying you don't know
1:20 pm
what that is. so, there is an attack of central banks around the world, and so, i think that he was wondering how you cope with that as a central banker which does not seem like an evil person at least from my conversations. >> okay. i'll take that. i feel like i am privileged to workt ta feder at the federal r because it is an honorable institution, and working with people like ben, and making sure that the policy is the best policy to hit the goals. so there are going to be challenges thrown in your face and one of them may be the criticisms that we are getting in this environment, but it is part of the job that you have to stick to the knitting and make sure that you are doing the policy as best you can, and there is no simple policy, and so you can see the discussions that we are having in terms of
1:21 pm
what is the best way to go about feeling, and the right policy. so you to, you know, deal with that. i have concern that we get this question all of the time, and if you have it, you know, you can have these considerations that never enter the room in the policy settings, and the reason that you can ask these questions, and there is people who may not know that, and the considerations are hitting the policies, and so we do have a concern about that. >> and the jared's question about targeting the share, and it is a very different animal. and i was targeting nominal wage growth, not real wage growth. labor share is much more viable, and so i would not touch it. >> i wanted to be reminded of ben bernanke's great lines in the transcripts is when the bank analysis revises the gdp growth
1:22 pm
for a number of years and bernanke says if they can revise the gdp retroactively, can we do the same with the interest rates? we will take this one and this gentleman here and then we have to call it a day. >> bernie elie, banking executive. there is coordination of banking and monitory policy and how realistic is that in the united states today given the substantial deficits that we are running in a rapidly rising ratio of the federal debt to gdp? another way to put it. is fiscal policy response to the downturn off of the table given the way that the deficit is in a full employment economy? >> and now, president master, you mentioned an inflation band, and so how high is the band and the standard deviation of roughly 0.4, and katia showed in
1:23 pm
her charts that the upc is currently 1.7, and currently 1.8, and is monetary policy precise enough to push up the inflation a quarter percentage point? >> okay. why don't you take that one, and then we will let him. >> and as a communication device, we know that it is going to be able to have a band which makes sense. so you can run a scope in the band to in harder times go closer to the zero and get to the band of below, so it is more flexibility. >> and like what, one percentage point? >> well, it is not a bad target. just thinking of how you would want to run it higher and normal times and maybe less high in the more dourn. downturn. >> yes, relative to the coordination, i forgot to
1:24 pm
mention that one place that you have had a fiscal response on the market is japan. in some respects the outcome has been disappointing, but one thing that is not a problem is people losing confidence in the ability of the japanese government to repay the debt. japan is far deeper in debt and has far worse long-term demographics than the united states and so the notion that the united states is anywhere close as olivier has documented this, that we are anywhere close to the limits of fiscal policy is utterly at odds with all of the evidence. >> you agree with that? >> yes. there is no justification for doing anything crazy, but no. >> okay. so this could go on for a long time and i want to really thank all of the presenters and we had a pretty good group of people, and i want to thank my colleagues who helped to organize this, and thank all of you for the good questions. if you could on the way out, if
1:25 pm
there is paper or coffee cups, there is a trashcan at the back of the room, and please join me in thanking everyone. thank you.
1:26 pm
1:27 pm
a look at today's live coverage across the c-span channels. at 4:00 p.m. eastern on c-span former ambassador to the united nations and assistant to president obama samantha power talking public policy. later, congresswoman alexandria ocasio cortez will hold a town hall in queens, new york, and focusing on poverty and affordable housing and access to federal benefits. c-span is going to have live coverage of that starting at 5:45 eastastern, and you can se both events free online or with the c-span app. and tonight, we are featuring african-american history preview as a preview of what is
1:28 pm
available on c-span3. tonight, marking the anniversary of slavery in the u.s., and looking at culture and institutions and how they have survived despite slavery and discrimination. watch that tonight at 8:00 p.m. here on c-span, and enjoy american history this weekend and every weekend on c-span3. saturday at 2:30 p.m. eastern on american history tv, winston lord, the author of "kissinger on kissinger" and nixon with his relationship with the secretary of state henry kissinger. >> he had read the books, and nixon was still interested in this and he knew that in order to do that he needed a able, and thoughtful national security adviser. and at 10:00 -- >> little to do except for work, and life seemed rugged indeed.
1:29 pm
here in such a setting, they prepared to search for oil. >> on "real america" the 1948 film "desert venture" on the origins of the saudi arabian oil industry, and sunday at 6:00 p.m. eastern on american artifacts we will preview the "votes for women" exhibit at the smithsonian gallery. >> and she was well ahead of her time. she started her own business as a banker with her sister. she advocated for free love which means sex outside of marriage. >> and at 6:30 p.m. eastern author sophia rosenfeld discussing her book "democracy and truth." >> no one person, king, priest, national research body or specific cast would get to call all of the shots. >> explore our nation's past on american history tv every weekend on c-span3.


info Stream Only

Uploaded by TV Archive on