Popular in Social Sciences

Books

City Time: On Being Sentenced to Rikers Island

Independent Scholar David A. Campbell

Culture and Society

The City as a Growth Machine: Toward a Political Economy of Place

New York University Harvey Moltoch

Economics

Gender and Economics

University of California Santa Barbara Shelly Lundberg

American Politics

Keeping the Republic

Boston College Marc Landy

Banks and Policy Organisations

Greening the financial sector

European Investment Bank Matteo Ferrazzi

British Politics

Entrepreneurial thinking: the politics and practice of policy impact

University of Sheffield Matthew Flinders

Culture and Society

Carceral Entrapments: Views from the Prison/Street Interface in India

Indian Institute of Technology Delhi Mahuya Bandyopadhya

Inflation – who cares? Monetary Policy in Times of Low Attention

View more insights in Social Sciences - Economics - Macroeconomics

View more figures from University of Bern

View the page of Oliver Pfauti

Read the Study

The decrease in the volatility and persistence of US inflation was accompanied by a decline in the public’s attention to inflation. But has the public always been so inattentive to inflation? And what do these low levels of attention imply for monetary policy? Oliver Pfauti seeks to answer these questions.

Image courtesy of interviewee. November 30, 2023

Generating Workspace

Export as PDF

This video video discusses the role of inflation expectations in monetary policy, particularly when a central bank is constrained by a zero lower bound on nominal interest rates. It explains that central banks can manage these expectations by committing to keep rates low for extended periods, which could lead to overheating and higher future inflation. This video video mechanism helps to lower real interest rates, encouraging consumption and investment, and can mitigate declines in economic activity.

However, the author points out that public inattention to inflation can hinder the effectiveness of these policies, leading to what is termed an "inflation attention trap." In such scenarios, the economy may remain stagnant at the zero lower bound due to insufficient public responsiveness to the central bank’s forward guidance.

The paper presents two principal contributions: first, it empirically examines how public attention to inflation has evolved over the past 50 years, finding that attention increases during periods of high inflation. Second, it derives the optimal monetary policy under these conditions, emphasizing that accounting for the zero lower bound is crucial. The findings suggest that while low attention may stabilize inflation in normal times, it can become detrimental when the lower bound restricts policy options.

In essence, the paper argues that understanding changes in public attention to inflation is vital for developing effective monetary policies, especially during times of economic uncertainty and fluctuating inflation levels. Recent inflation trends post-pandemic have drawn renewed attention to these dynamics.

Language:

Export as PDF

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:00.265 --> 00:02.845
So, in most monetary policy models,

00:03.925 --> 00:06.885
managing inflation expectations is a very powerful tool

00:07.265 --> 00:08.325
for monetary policy.

00:08.545 --> 00:11.405
And this becomes especially relevant at the zero

00:11.625 --> 00:13.765
or the effective lower bound constraint.

00:14.065 --> 00:16.405
So let's say that a central bank's main instrument,

00:16.595 --> 00:18.805
that short term nominal interest rate

00:19.345 --> 00:21.125
is stuck at the zero lower bound

00:21.145 --> 00:23.005
and it cannot go further down.

00:23.625 --> 00:26.165
Now in theory, what the central bank could do is they could

00:26.165 --> 00:29.925
promise to stay at the lower bound for longer than they have

00:29.925 --> 00:31.645
to, which then basically means

00:31.645 --> 00:33.245
that the central bank promises

00:33.245 --> 00:36.325
that they would let the economy overheat a little bit in the

00:36.325 --> 00:38.725
future and we would expect the future

00:39.485 --> 00:41.285
economic boom and higher inflation.

00:42.185 --> 00:45.085
Now this should then increase people's expectations about

00:45.085 --> 00:46.805
future inflation already today.

00:47.225 --> 00:49.405
And because of these high inflation expectations,

00:49.585 --> 00:50.805
the real interest rates,

00:50.825 --> 00:55.125
so the nominal rate minus inflation expectations decreases

00:55.705 --> 00:58.365
and therefore households do not want to save as much

00:58.385 --> 00:59.885
and they would rather consume more.

01:00.105 --> 01:02.765
And because of the higher inflation expectations,

01:02.775 --> 01:04.645
firms will want to set higher prices.

01:05.585 --> 01:08.365
So now even though the central bank may not be able

01:08.365 --> 01:10.365
to move the nominal rate further down

01:10.365 --> 01:11.485
because of the lower bound,

01:11.795 --> 01:14.125
they can decrease the real interest rate.

01:14.145 --> 01:17.565
And through this we should then see only a relatively mild

01:17.965 --> 01:19.205
decrease in economic activity

01:19.705 --> 01:21.725
and in inflation, even though the economy

01:21.905 --> 01:23.125
is at the lower bound.

01:24.215 --> 01:26.835
In practice this might be a bit different

01:26.835 --> 01:29.875
because the general public in most advanced economies at the

01:29.875 --> 01:34.515
time when the lower bound was binding, was very inattentive

01:34.615 --> 01:36.755
and poorly informed when it comes

01:36.755 --> 01:39.035
to monetary policy and inflation.

01:39.735 --> 01:41.515
And at the same time, we also saw

01:41.515 --> 01:43.635
that many central banks were struggling

01:43.855 --> 01:46.595
to bring inflation up to their inflation targets.

01:47.415 --> 01:49.875
So in this paper what I ask basically is the question

01:49.875 --> 01:52.635
whether these two observations may be related

01:53.375 --> 01:54.795
and it turns out that they are.

01:55.375 --> 01:58.595
So, as I will explain a little bit in more detail later on,

01:59.035 --> 02:01.635
I find that when people do not pay a lot of attention

02:01.655 --> 02:04.715
to inflation, the economy can get stuck in

02:04.715 --> 02:07.035
what I call an inflation attention trap.

02:07.535 --> 02:11.515
So these are long periods in which the economy gets stuck at

02:11.515 --> 02:15.315
the zero lower bound because people are inattentive and

02:15.315 --> 02:16.795
therefore these promises

02:16.795 --> 02:18.355
or forward guidance policies

02:18.355 --> 02:22.035
that the central bank makes are not very effective in

02:22.235 --> 02:23.755
stimulating people's expectations.

02:24.715 --> 02:26.455
So this is why then the economy may get

02:26.455 --> 02:27.735
stuck at the lower bound.

02:28.355 --> 02:30.735
And I show that this actually matters quite a bit

02:30.795 --> 02:32.495
for optimal monetary policy.

02:33.595 --> 02:35.775
And I find that accounting for the lower bound is,

02:35.955 --> 02:38.815
is crucial for these, for the normative implications.

02:39.115 --> 02:40.735
So if you do not account

02:40.735 --> 02:42.535
for the zero lower bound constraint,

02:42.965 --> 02:45.575
then lower attention is actually welfare improving.

02:46.355 --> 02:49.015
But that completely changes once we account

02:49.015 --> 02:51.615
for the lower bank because then the economy may run into

02:51.615 --> 02:52.735
these attention traps.

02:52.735 --> 02:54.335
And to prevent this from happening,

02:55.075 --> 02:57.575
the central bank optimally induces a higher

02:57.575 --> 02:58.775
average inflation rate.

02:58.795 --> 03:01.045
But this is costly from a welfare perspective.

03:01.505 --> 03:04.485
So welfare is actually lower when attention is low.

03:05.145 --> 03:08.005
So there are two main contributions to the literature

03:08.005 --> 03:09.165
that I make in this paper.

03:09.825 --> 03:11.205
The first contribution is

03:11.205 --> 03:14.125
that I empirically quantify attention to inflation

03:14.585 --> 03:17.325
and in particular how attention has changed

03:17.915 --> 03:19.245
over the last 50 years.

03:19.785 --> 03:21.405
And I find that it changed a lot

03:21.785 --> 03:24.285
and that attention is higher in times of high

03:24.545 --> 03:25.725
and volatile inflation.

03:26.785 --> 03:29.085
The second main contribution that I make is

03:29.085 --> 03:32.725
that I derive the Ramsey optimal monetary policy when the

03:32.725 --> 03:36.205
central bank faces an occasionally binding lower bound

03:36.585 --> 03:37.645
on the nominal rate.

03:37.985 --> 03:40.245
And when private agents only pay limited

03:40.595 --> 03:41.845
attention to inflation.

03:42.675 --> 03:46.125
Earlier papers that studied the optimal monetary policy

03:46.155 --> 03:48.845
with limited attention did not take the lower bound

03:48.895 --> 03:50.125
constraint into account.

03:50.585 --> 03:51.925
But I find that accounting

03:51.925 --> 03:55.405
for this lower bound actually completely flips the normative

03:55.485 --> 03:57.165
implications of lower attention.

03:57.905 --> 04:00.165
In a first step, I propose a way

04:00.165 --> 04:02.045
how we can estimate people's attention

04:02.105 --> 04:03.845
to inflation using survey data.

04:04.465 --> 04:05.965
So this is based on a model

04:05.965 --> 04:08.125
where the agents decide optimally

04:08.125 --> 04:09.405
how attentive they want to be.

04:09.905 --> 04:11.125
If they are more attentive,

04:11.125 --> 04:13.165
they make smaller forecast errors.

04:13.665 --> 04:15.365
But being attentive is also costly

04:15.365 --> 04:16.645
because then you have to acquire

04:16.705 --> 04:20.125
and process information that takes time and mental effort.

04:20.625 --> 04:22.925
So what then comes out of this model is

04:22.925 --> 04:25.805
that attention can basically be measured by

04:25.905 --> 04:29.645
how strongly agents update their inflation expectations when

04:29.645 --> 04:31.085
they make a forecast error.

04:31.785 --> 04:33.405
And the model also predicts

04:33.405 --> 04:36.725
that attention should be higher when inflation is volatile

04:37.065 --> 04:38.405
or when it is persistent.

04:39.345 --> 04:42.525
So then to estimate attention, I use several surveys

04:42.555 --> 04:45.125
that ask people about their inflation expectations.

04:45.525 --> 04:47.485
I use data on professional forecasters.

04:47.765 --> 04:49.245
I use data on households.

04:50.025 --> 04:53.365
And in this study I focus on the United States going back

04:53.385 --> 04:55.525
all the way to the late, late 1960s.

04:56.265 --> 04:59.285
And this long time span actually allows me

04:59.305 --> 05:01.365
to test whether people's attention

05:01.425 --> 05:03.685
to inflation has changed over time or not.

05:04.185 --> 05:07.725
And I can examine when is attention high and when is it low.

05:08.545 --> 05:11.285
And then later in order to understand the implications

05:11.305 --> 05:13.685
of the empirical findings for monetary policy,

05:14.365 --> 05:18.525
I developed a new casin model where I introduced this idea

05:18.705 --> 05:20.485
of limited attention to inflation

05:21.065 --> 05:23.845
and I imposed that the center bank is potentially

05:23.845 --> 05:27.565
constrained by a zero lower bound on nominal interest rates.

05:28.345 --> 05:30.845
So what I find in the data is that people's attention

05:30.905 --> 05:35.125
to inflation before the covid 19 pandemic was actually at a

05:35.125 --> 05:38.605
historical low, but attention has not always been that low,

05:38.945 --> 05:41.085
but it changed actually quite a bit over time.

05:41.545 --> 05:44.405
So in the seventies and eighties when inflation was high

05:44.545 --> 05:46.205
and volatile and persistent,

05:46.465 --> 05:48.805
people were much more attentive to inflation.

05:49.585 --> 05:51.965
And this is in line with what the underlying model

05:51.985 --> 05:53.325
of attention that I have in the

05:53.325 --> 05:54.765
background actually predicts.

05:55.665 --> 05:56.965
So then given these findings,

05:57.245 --> 05:59.245
I turn to the optimal policy problem

06:00.065 --> 06:02.525
and I show that there's a trade-off of lower attention

06:02.585 --> 06:03.685
for monetary policy.

06:04.545 --> 06:07.085
On the one hand, when attention is low,

06:07.315 --> 06:10.885
inflation expectations respond less to new information.

06:11.105 --> 06:13.645
So inflation expectations become more stable,

06:13.815 --> 06:16.725
which then also helps to stabilize actual inflation.

06:17.025 --> 06:18.445
And this is welfare improving.

06:19.545 --> 06:21.925
On the other hand, it becomes difficult

06:21.945 --> 06:24.765
to affect people's expectations when they're not attentive.

06:25.105 --> 06:27.485
And this becomes an issue at the zero lower bound

06:27.485 --> 06:30.365
because then monetary policy cannot use their usual

06:30.365 --> 06:32.085
instrument than nominal rate,

06:32.465 --> 06:34.885
but they rather have to steer people's inflation

06:34.885 --> 06:38.445
expectations by making promises about the future path

06:38.825 --> 06:39.885
of the nominal rate.

06:40.425 --> 06:41.925
But when people are not attentive,

06:42.055 --> 06:44.165
these promises are not very effective

06:44.185 --> 06:47.405
and the economy may then run into these so-called inflation

06:47.405 --> 06:50.445
attention traps, where you get stuck at the zero lower bound

06:50.445 --> 06:53.525
for much longer because people are not listening to

06:53.525 --> 06:55.485
what the central bank is telling them, and

06:55.485 --> 06:58.045
therefore the central bank cannot manage their expectations.

06:58.945 --> 07:00.805
Now, in terms of the implications

07:00.865 --> 07:03.845
for optimal monetary policy, what I find is

07:03.845 --> 07:06.005
that at accounting for the lower bound completely

07:06.195 --> 07:07.285
changes the picture.

07:08.025 --> 07:10.925
So if we do not account for the zero lower bound,

07:11.355 --> 07:13.725
then lower attention stabilizes inflation.

07:13.835 --> 07:14.965
This is welfare improving.

07:15.555 --> 07:18.085
However, once you do account for the lower bound,

07:18.475 --> 07:20.525
this now changes because now you have

07:20.525 --> 07:22.805
to avoid these attention traps from happening,

07:23.135 --> 07:24.205
which you can achieve

07:24.305 --> 07:26.605
by inducing a higher average inflation rate.

07:27.425 --> 07:30.085
But be because with a higher inflation rate,

07:30.085 --> 07:32.805
what you have is that the nominal rate on average is also

07:32.805 --> 07:35.725
higher and therefore further away from the lower bound so

07:35.725 --> 07:38.445
that the lower bound becomes less likely to bind.

07:39.475 --> 07:42.645
However, this then means that you have to allow, allow

07:42.645 --> 07:44.085
for this higher inflation rate on average,

07:44.145 --> 07:46.285
and this is costly from a welfare perspective.

07:47.265 --> 07:49.485
So the lower people's attention

07:49.625 --> 07:53.645
to inflation is the larger is this increase in the optimal

07:53.645 --> 07:55.405
inflation rate, and therefore the largest

07:55.465 --> 07:56.565
is the welfare loss.

07:57.705 --> 08:00.205
So if you do account for the lower bound welfare

08:00.805 --> 08:02.445
decreases when attention decreases,

08:02.475 --> 08:04.165
whereas if you do not account

08:04.165 --> 08:07.285
for the lower bound welfare actually increases when

08:07.285 --> 08:08.365
attention decreases.

08:09.035 --> 08:11.485
Overall in this paper, what I show is

08:11.485 --> 08:14.565
that people's attention to inflation changes over time,

08:14.945 --> 08:17.205
and it depends on the economic environment.

08:17.545 --> 08:19.765
And this actually turns out to matter quite a bit

08:19.765 --> 08:20.765
for monetary policy.

08:21.625 --> 08:24.405
So the decrease in people's attention to inflation

08:24.695 --> 08:28.205
after the So-called great inflation period in the seventies

08:28.205 --> 08:30.765
and eighties was probably desirable at first

08:30.865 --> 08:31.925
for monetary policy

08:31.925 --> 08:34.765
because this helped to stabilize inflation expectations

08:35.185 --> 08:37.525
and then ultimately actual inflation as well.

08:38.115 --> 08:40.045
However, the paper highlights

08:40.075 --> 08:43.325
that people's low attention may actually become a challenge

08:43.745 --> 08:45.485
or an issue for monetary policy

08:46.115 --> 08:49.485
once the zero lower bound constraint may be become binding.

08:50.305 --> 08:52.645
Now with the recent surge in inflation

08:52.815 --> 08:54.685
after the pandemic, people

08:55.125 --> 08:56.565
Actually started again to become

08:56.565 --> 08:57.725
more attentive to inflation.

08:57.865 --> 08:59.605
As I show in a more recent paper,

08:59.615 --> 09:02.485
which is titled the Inflation Attention Threshold

09:02.745 --> 09:03.885
and Inflation Surge.

09:04.505 --> 09:07.845
And I show that this increase in attention may AC actually

09:07.845 --> 09:09.485
help us to understand some

09:09.485 --> 09:11.725
of the recent inflation surge we have seen.

09:12.345 --> 09:14.965
So again, this highlights the importance

09:15.035 --> 09:17.405
that people's attention to inflation may change

09:17.945 --> 09:21.565
and policy in particular monetary policy may want to take

09:21.565 --> 09:24.245
that into account when thinking about the appropriate

09:24.605 --> 09:26.685
policies in different economic environments.

00:00 - So, in most monetary policy models, managing inflation expectations is a very powerful tool for monetary policy. And this becomes especially relevant at the zero or the effective lower bound constraint. So let's say that a central bank's main instrument,

00:16 - that short term nominal interest rate is stuck at the zero lower bound and it cannot go further down. Now in theory, what the central bank could do is they could promise to stay at the lower bound for longer than they have to, which then basically means that the central bank promises that they would let the economy overheat a little bit in the future and we would expect the future economic boom and higher inflation.

00:42 - Now this should then increase people's expectations about future inflation already today. And because of these high inflation expectations, the real interest rates, so the nominal rate minus inflation expectations decreases and therefore households do not want to save as much and they would rather consume more. And because of the higher inflation expectations, firms will want to set higher prices. So now even though the central bank may not be able

01:08 - to move the nominal rate further down because of the lower bound, they can decrease the real interest rate. And through this we should then see only a relatively mild decrease in economic activity and in inflation, even though the economy is at the lower bound. In practice this might be a bit different because the general public in most advanced economies at the time when the lower bound was binding, was very inattentive

01:34 - and poorly informed when it comes to monetary policy and inflation. And at the same time, we also saw that many central banks were struggling to bring inflation up to their inflation targets. So in this paper what I ask basically is the question whether these two observations may be related and it turns out that they are. So, as I will explain a little bit in more detail later on, I find that when people do not pay a lot of attention

02:01 - to inflation, the economy can get stuck in what I call an inflation attention trap. So these are long periods in which the economy gets stuck at the zero lower bound because people are inattentive and therefore these promises or forward guidance policies that the central bank makes are not very effective in stimulating people's expectations. So this is why then the economy may get stuck at the lower bound.

02:28 - And I show that this actually matters quite a bit for optimal monetary policy. And I find that accounting for the lower bound is, is crucial for these, for the normative implications. So if you do not account for the zero lower bound constraint, then lower attention is actually welfare improving. But that completely changes once we account for the lower bank because then the economy may run into these attention traps.

02:52 - And to prevent this from happening, the central bank optimally induces a higher average inflation rate. But this is costly from a welfare perspective. So welfare is actually lower when attention is low. So there are two main contributions to the literature that I make in this paper. The first contribution is that I empirically quantify attention to inflation and in particular how attention has changed

03:17 - over the last 50 years. And I find that it changed a lot and that attention is higher in times of high and volatile inflation. The second main contribution that I make is that I derive the Ramsey optimal monetary policy when the central bank faces an occasionally binding lower bound on the nominal rate. And when private agents only pay limited attention to inflation.

03:42 - Earlier papers that studied the optimal monetary policy with limited attention did not take the lower bound constraint into account. But I find that accounting for this lower bound actually completely flips the normative implications of lower attention. In a first step, I propose a way how we can estimate people's attention to inflation using survey data. So this is based on a model

04:05 - where the agents decide optimally how attentive they want to be. If they are more attentive, they make smaller forecast errors. But being attentive is also costly because then you have to acquire and process information that takes time and mental effort. So what then comes out of this model is that attention can basically be measured by how strongly agents update their inflation expectations when

04:29 - they make a forecast error. And the model also predicts that attention should be higher when inflation is volatile or when it is persistent. So then to estimate attention, I use several surveys that ask people about their inflation expectations. I use data on professional forecasters. I use data on households. And in this study I focus on the United States going back all the way to the late, late 1960s.

04:56 - And this long time span actually allows me to test whether people's attention to inflation has changed over time or not. And I can examine when is attention high and when is it low. And then later in order to understand the implications of the empirical findings for monetary policy, I developed a new casin model where I introduced this idea of limited attention to inflation and I imposed that the center bank is potentially constrained by a zero lower bound on nominal interest rates.

05:28 - So what I find in the data is that people's attention to inflation before the covid 19 pandemic was actually at a historical low, but attention has not always been that low, but it changed actually quite a bit over time. So in the seventies and eighties when inflation was high and volatile and persistent, people were much more attentive to inflation. And this is in line with what the underlying model of attention that I have in the background actually predicts.

05:55 - So then given these findings, I turn to the optimal policy problem and I show that there's a trade-off of lower attention for monetary policy. On the one hand, when attention is low, inflation expectations respond less to new information. So inflation expectations become more stable, which then also helps to stabilize actual inflation. And this is welfare improving. On the other hand, it becomes difficult

06:21 - to affect people's expectations when they're not attentive. And this becomes an issue at the zero lower bound because then monetary policy cannot use their usual instrument than nominal rate, but they rather have to steer people's inflation expectations by making promises about the future path of the nominal rate. But when people are not attentive, these promises are not very effective and the economy may then run into these so-called inflation

06:47 - attention traps, where you get stuck at the zero lower bound for much longer because people are not listening to what the central bank is telling them, and therefore the central bank cannot manage their expectations. Now, in terms of the implications for optimal monetary policy, what I find is that at accounting for the lower bound completely changes the picture. So if we do not account for the zero lower bound, then lower attention stabilizes inflation.

07:13 - This is welfare improving. However, once you do account for the lower bound, this now changes because now you have to avoid these attention traps from happening, which you can achieve by inducing a higher average inflation rate. But be because with a higher inflation rate, what you have is that the nominal rate on average is also higher and therefore further away from the lower bound so that the lower bound becomes less likely to bind.

07:39 - However, this then means that you have to allow, allow for this higher inflation rate on average, and this is costly from a welfare perspective. So the lower people's attention to inflation is the larger is this increase in the optimal inflation rate, and therefore the largest is the welfare loss. So if you do account for the lower bound welfare decreases when attention decreases, whereas if you do not account

08:04 - for the lower bound welfare actually increases when attention decreases. Overall in this paper, what I show is that people's attention to inflation changes over time, and it depends on the economic environment. And this actually turns out to matter quite a bit for monetary policy. So the decrease in people's attention to inflation after the So-called great inflation period in the seventies and eighties was probably desirable at first

08:30 - for monetary policy because this helped to stabilize inflation expectations and then ultimately actual inflation as well. However, the paper highlights that people's low attention may actually become a challenge or an issue for monetary policy once the zero lower bound constraint may be become binding. Now with the recent surge in inflation after the pandemic, people Actually started again to become

08:56 - more attentive to inflation. As I show in a more recent paper, which is titled the Inflation Attention Threshold and Inflation Surge. And I show that this increase in attention may AC actually help us to understand some of the recent inflation surge we have seen. So again, this highlights the importance that people's attention to inflation may change and policy in particular monetary policy may want to take

09:21 - that into account when thinking about the appropriate policies in different economic environments.

Faculti AI can make mistakes