International Monetary Fund | April 2023 45
The natural rate of interest—the real interest rate
that neither stimulates nor contracts the economy—is
important for both monetary and fiscal policy; it is a
reference level to gauge the stance of monetary policy
and a key determinant of the sustainability of public
debt. This chapter aims to study the evolution of the
natural rate of interest across several large advanced and
emerging market economies. To mitigate the uncertainty
that typically surrounds estimates of the natural rate, the
chapter relies on complementary approaches to analyze
its drivers and project its future path. Common trends
such as demographic changes and productivity slowdown
have been key factors in the synchronized decline of the
natural rate. And while international spillovers have been
important determinants of the natural rate, offsetting
forces have resulted in only a moderate impact on balance.
Overall, the analysis suggests that once the current
inflationary episode has passed, interest rates are likely to
revert toward pre-pandemic levels in advanced economies.
How close interest rates get to those levels will depend on
whether alternative scenarios involving persistently higher
government debt and deficit or financial fragmentation
materialize. In major emerging market economies, natural
interest rates are expected to gradually converge from
above toward advanced economies’ levels. In some cases,
this may ease the pressure on fiscal authorities over the
long term, but fiscal adjustments will still be needed in
many countries to stabilize or reduce debt-to-GDP ratios.
Introduction
In 1979, the Federal Reserve hiked interest rates
from about 10 percent at the start of the year to
almost 14 percent by the year’s end, which in real
terms—after taking account of inflation—amounted to
a rate of interest of about 5 percent.
1
Even at the time
e authors of this chapter are Philip Barrett (co-lead), Christoffer
Koch, Jean-Marc Natal (co-lead), Diaa Noureldin, and Josef Platzer,
with support from Yaniv Cohen and Cynthia Nyakeri. e authors
thank John Williams for very helpful comments.
1
When comparing interest rates, it is important to take account
of inflation. Savings invested at 5 percent when inflation is 2 percent
will buy the same thing as an investment at 3 percent when infla-
tion is zero.
this was viewed as likely insufficient to tame rapidly
rising inflation.
2
And so it proved to be. Inflation
continued to rise, peaking at nearly 15 percent the
following year, requiring even higher interest rates and
a prolonged recession before the situation was brought
under control.
Nearly three decades later as the world faced the
global financial crisis of 2008, the Federal Reserve—
along with central banks worldwide—slashed interest
rates to as close to zero as they thought possible in
nominal and real terms. is time around, however,
commentators and policymakers raised concerns that
interest rates were not low enough to boost demand
and inflation. Once again, these concerns proved
well-founded, with inflation remaining stubbornly low
for much of the next 10 years.
ese two contrasting examples raise an obvious
question. How can it be that in the same country a
real interest rate of 5 percent is sometimes too low but
at other times a real interest rate of zero is too high?
Most answers rely on the idea that a given real
interest rate does not have the same macroeconomic
effects at all times. Instead, the impact is relative to
some reference level. When real interest rates are below
that level, they are stimulatory, boosting demand and
inflation. And when above it, they are contractionary,
lowering output and inflation. If this reference level
moves over time, then the same real interest rate can
be too high or too low at different times.
Macroeconomists call this reference interest rate
the “real natural rate of interest.”
3
e “natural” part
means that this is the real interest rate that is neither
stimulatory nor contractionary and is consistent with
output at potential and stable inflation. Lowering the
real rate below the natural rate is akin to stepping on
the macroeconomic accelerator; raising it above is like
hitting the brake. e natural rate is usually thought of
as independent of monetary policy and instead driven
2
See Goodfriend and King (2005).
3
In many discussions, the “real” part is dropped; this approach is
followed in the chapter. Some economists use the terms “neutral”
and “natural” interchangeably, and some do not. For clarity, this
chapter uses only “natural.”
THE NATURAL RATE OF INTEREST: DRIVERS AND
IMPLICATIONS FOR POLICY
2
CHAPTER
46 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
by real phenomena such as, for instance, technological
progress, demographics, inequality, or preference shifts
for safe and liquid assets.
4
As the preceding discussion suggests, the natural
rate is important for the conduct of monetary policy.
Policymakers need to know the level of the natural
rate in order to gauge the likely impact of their poli-
cies and so assess the stance of monetary policy. e
natural rate also has a critical influence on fiscal pol-
icy. On average over the long term, monetary policy is
typically neither inflationary nor contractionary. And
so the natural rate is also an anchor for real rates over
long periods of time. Because governments typically
pay back debts over long time spans (both through
long-maturity debt and by rolling over short-term
debt), the natural rate is essential in determining the
overall cost of borrowing and the sustainability of
public debts.
Given the importance of the natural rate for both
monetary and fiscal policy, it is not surprising that
the recent surge in inflation and government debt
worldwide has led to renewed interest in this topic.
Real rates have increased a bit as monetary policy has
become tighter in response to higher inflation. But
the uptick remains modest compared with the late
1970s. Whether central banks have raised rates enough
to return inflation to target depends critically on the
level of the natural rate. Similarly, the natural rate will
determine how much of a burden the present-day high
levels of debt will be for governments (see Chapter 3).
In light of these concerns, the chapter seeks to
answer the following questions:
• How has the natural rate evolved in the past across
different economies?
• What has driven this evolution?
• What is the outlook for these drivers and natural
rates in the near and medium term?
• How will this outlook affect monetary and
fiscal policies?
To shed light on these issues, the chapter first reviews
the main stylized facts that characterize real interest
rate trends at different maturities and across different
countries. It then sets out to measure the natural rate.
To mitigate the unavoidable uncertainty associated with
estimations of the natural rate, the chapter will follow
4
In line with a long tradition in monetary economics, monetary
policy is here assumed to be neutral, meaning that it does not affect real
variables over the long term. Borio, Disyatat, and Rungcharoenkitkul
(2019) present an alternative view and implications for the natural rate.
a two-pronged approach. Beginning with a simple
model (Laubach and Williams 2003)—one that lets the
data speak—it moves to a tighter theoretical structure
that imposes more restrictions on the data but allows a
deeper understanding of the underlying drivers of the
natural rate (Platzer and Peruffo 2022). Comparing
estimates from different models provides independent
validation. In addition, alternative scenarios covering
a range of plausible future developments for the main
underlying drivers of the natural rate are considered
for robustness. ese projections provide a long-term
anchor for monetary policy and a crucial input to
analyze debt sustainability in the largest advanced and
emerging market economies.
e main findings of the chapter are as follows:
• Common trends have played an important role in
driving real interest rates down. The natural rate
has declined over the past four decades in most
advanced economies and some emerging markets.
While idiosyncratic factors can explain cross-country
differences, common trends underlying demographic
transitions and productivity slowdowns are key to
understanding the synchronized decline.
• Global drivers have also been important determinants
but on balance have had a limited impact on net cap-
ital flows and corresponding natural rates in advanced
and emerging market economies. As global capital
markets opened and fast-growing emerging market
economies entered the scene in the 1980s and
1990s, foreign factors increasingly shaped long-term
trends in interest rates. High growth in emerging
markets has tended to drive up interest rates in
advanced economies while producing a glut of
savings in emerging markets. These excess savings—
in their quest for safe and liquid assets—have tended
to flow back to advanced economies, pushing
natural interest rates back down. On balance, these
forces seem to have had broadly offsetting effects on
capital flows and a moderate impact on natural rates
over the past half-century.
• Country-specific natural rates of interest are projected
to converge in the next couple of decades. Based on
conservative assumptions on demographic, fiscal,
and productivity developments, it is anticipated that
natural rates in large emerging market economies
will decline, gradually converging toward the low
and steady levels expected in advanced economies.
• As inflation returns to target, the effective lower
bound on interest rates may become binding again.
Post-pandemic increases in interest rates could be
47International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
protracted until inflation is brought back to target
(Chapter 1). However, long-term forces driving
the natural rate suggest that interest will eventually
converge toward pre-pandemic levels in advanced
economies. How close to those levels will depend on
whether alternative scenarios involving persistently
higher government debt and deficit or financial
fragmentation materialize. Because nominal rates
cannot fall far below zero (the effective lower bound
constraint), this could limit central banks’ ability to
respond to negative demand shocks. Thus, debates
about the appropriate level of target inflation at
the effective lower bound could reemerge. Even the
central banks in some emerging market economies
may eventually need to adopt unconventional policy
tools similar to those used by advanced economies
in recent years.
• Despite increased fiscal space, many countries will
have to consolidate. While low natural rates may
ease pressure on fiscal policy, they do not negate
the need for fiscal responsibility. Important gov-
ernment support during the pandemic has strained
public accounts, requiring some budget consol-
idation to ensure long-term debt sustainability.
Various paths to deficit reduction are open, but
delaying action will only make the required steps
more drastic: Larger public debt tends to crowd
out private investment and erode the appeal of safe
and liquid government debt.
Trends in Real Rates over the Long Term
is section lays out some basic facts about how real
interest rates have evolved over the long term. Because
the natural rate is an anchor for real interest rates,
long-term trends in real interest rates are potentially
informative signals about the natural rate itself.
Figure 2.1, panel 1, starts the inquiry by comparing
five different measures of the ex ante real interest rate
for the United States.
5
Different maturities from 1 year
up to 20 years are considered. Despite differences at
high frequencies—the short-horizon measures are
5
Ex ante measures of the real interest rate use actual measures of
inflation expectations, which are either extracted from financial mar-
kets or based on surveys, to deflate the nominal interest rate. Ex post
real interest rates rely instead on realized inflation. Over long periods
of time, ex ante and ex post real interest rates tend to coincide, but
there can be large discrepancies when surprise inflation is expected to
be temporary, as in the most recent episode. Unfortunately, inflation
expectation measures are not always available for long time series,
emerging markets, or both.
unsurprisingly much more volatile—all these measures
share a common long-term trend. Looking through
cyclical fluctuations and term premiums, real rates
have fallen steadily, by about 5 percentage points over
the last four decades across all maturities. Given that
the natural rate of interest is a long-term attractor for
real rates, this suggests that the natural rate of interest
has also fallen, at least in the United States.
To get a sense of whether these developments have
been mirrored elsewhere, Figure 2.1, panel 2, compares
historical ex post real rates in five advanced economies
over a similar period, in this case using three-month
real rates. e broad pattern is the same, with real
rates declining steadily from highs in the 1980s.
Interestingly, the common international component
seems at first glance to have become more important
over time, with countries’ real rates seeming to con-
verge gradually.
1-year 2-year 5-year
10-year 20-year
United States Japan Germany
United Kingdom France
Figure 2.1. Real Interest Rate Trends
(Percent)
1. (Ex Ante) Real Interest Rates at Different Maturities in
the United States
–4
–2
0
2
4
6
8
10
1982 90 2000 10 20 Feb.
23
2. (Ex Post) Real Short-Term Interest Rates in Selected
Advanced Economies
1982 90 2000 10 20 Nov.
22
–10
–5
0
5
10
15
20
Sources: Federal Reserve Economic Data; and IMF staff calculations.
Note: In panel 1, the real interest rates are computed as the difference between
the US Treasury rate at each horizon and the Cleveland Federal Reserve measure
of inflation expectations over the same horizon. In panel 2, the real interest rates
are the difference between the three-month interbank rates and the average of the
realized inflation measured by the consumer price index in the next three months
for each country. Japan’s three-month interbank rates are spliced with rates for
certificates of deposit from 1979 to 2002. Online Annex 2.1 provides details on
data sources and calculations for the figure.
48 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
Figure 2.2 contrasts developments in advanced and
emerging market economies. A shared trend at the
start of the 2000s decoupled later on as real rates con-
tinued to decline in advanced economies but stabilized
at their 2005 level in emerging markets.
Overall, this first look at the data suggests that
the natural rate has likely declined in the past four
decades or so in advanced economies. is downward
trend seems to be increasingly common across coun-
tries and points to some global drivers. e picture
is different in emerging markets, where natural rates
have remained broadly stable over the past 20 years
on average. Because emerging market and advanced
economies’ current accounts are broadly balanced,
the divergence in long-term rates points to remaining
frictions preventing a stronger convergence between
advanced and emerging market economies (Obstfeld
2021).
6
Yet this analysis leaves many important issues
unaddressed. e data, although suggesting that the
natural rate has declined in many advanced economies,
6
Beyond market frictions, weak institutions and lack of investor
protection in recipient countries may also explain the lack of
convergence. An alternative explanation, which is likely to be
particularly relevant for the United States, is that following the
global financial crisis, emerging market debt was not considered
safe, pushing down the real interest rate for the main provider of
safe and liquid assets.
cannot explain why this decline occurred and fail to
distinguish the impact of secular and cyclical factors.
e following sections tackle these concerns.
Measuring the Natural Rate
is section relies on well-known macroeconomic
empirical models to try to estimate the natural rate
of interest. Because the natural rate is an unobserved,
latent variable, any measurement requires some theory.
e approach here is to use a minimal amount of
theory, drawing on simple macroeconomic relation-
ships between aggregate supply and demand, interest
rates, and inflation. Approaches based on aggregate
relationships are a good starting point for developing
a more informed measure of the natural rate because
they are transparent and straightforward. Subsequent
sections use a richer framework based on more exten-
sive microeconomic theory and so speak more to the
underlying drivers of the natural rate.
Single-Country Estimates of the Natural Rate
e first approach is an application of the widely
used Laubach-Williams model (Holston, Laubach, and
Williams 2017; hereafter HLW). is model assumes
a set of relationships between supply, demand, interest
rates, and prices consistent with perhaps the most
standard macroeconomic view of the world, the New
Keynesian model.
7
In this setting, the natural rate is
driven by a variety of shocks, including trend output
growth. Here, it is defined as the real interest rate that
will return output to potential and inflation to target,
once purely transitory shocks to aggregate supply or
demand have dissipated. e intuition for this is that
central banks tend to think about returning inflation
to target in the medium term, because trying to offset
every temporary shock would lead to undue volatility
in interest rates and output.
8
7
See Online Annex 2.2 for a formal description of the model. All
online annexes are available at www .imf .org/ en/ Publications/ WEO.
8
In this framework, financial shocks affect the natural rate only if
they affect potential output. A persistent increase in precautionary
saving or preference for safe and liquid assets would qualify, whereas
purely transitory variation in risk aversion, for example, would not
(Barsky, Justiniano, and Melosi 2014; Gourinchas, Rey, and Sauzet
2022). is definition of the natural rate is consistent with the one
implicit in the theoretical framework of the next section, because it
emphasizes low-frequency movements of the real interest rate in a
world without nominal friction (where output is at potential).
Advanced economies
Emerging market and developing economies
Figure 2.2. (Ex Post) Real Interest Rates in Advanced and
Emerging Market and Developing Economies
(Percent)
–6
–4
–2
0
2
4
6
2003 06 09 12 15 18 22
Divergence after 2011
Source: IMF staff calculations.
Note: The sample comprises 34 advanced economies and 25 emerging market
and developing economies, aggregated using market-exchange-rate-based GDP
weights. Maturity of the bonds is greater than one year. Nominal interest rates are
deflated using consumer price inflation.
49International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
e model is first estimated from data for one
country at a time. As part of the estimation, the model
attempts to figure out what were the most likely
values for several key unobserved variables, including
potential output and the natural rate of interest, given
the (relatively standard) New Keynesian view of the
macroeconomy. is framework also offers a basic
decomposition of changes in the natural rate into two
components: one due to changes in the long-term
growth trend, and one due to other factors, which
can in principle include domestic and foreign drivers.
One drawback, however, is that the HLW model is
designed to apply principally to advanced economies,
for which data can be reasonably described by the
New Keynesian model over a long enough time period.
e richer structural model in the next section has
more to say about emerging markets.
Figure 2.3 summarizes the results from estimating
the HLW model on a sample of six advanced econo-
mies for which sufficient quarterly data exist. It shows
estimates of the natural rate, as well as the part due to
trend growth, for two five-year periods: one covering
the end of the 1970s, the other for the late 2010s.
ese estimates broadly confirm the intuition pre-
sented so far in this chapter: that the natural rate of
interest has declined across advanced economies in the
past 40 years. Despite some variation in the level of the
rate across countries, the magnitude of the decline has
been broadly similar, at a little over 2 percentage points
in most countries. is is much smaller than the over-
all decline in real interest rates over the same period (of
about 5 percentage points), which likely also reflected
the change in the monetary policy stance, particularly
tight at the beginning of the 1980s as central banks
fought historically high inflation.
However, the uncertainty over the estimates of the
natural rate is very large, with the 90 percent confi-
dence interval for the United States ranging from zero
to about 3 percent in the second half of the 2010s.
Uncertainty is a common feature of all estimates of
the natural rate
9
and arises because the estimated
relationships between interest rates and the output gap,
and the output gap and inflation, are both relatively
weak. As a result, fluctuations in output and inflation
provide little information about the overall level of the
natural rate. Yet at least one part of the natural rate
is well estimated: the trend growth component, for
9
See Arena and others (2020) for a related exercise applied to
European countries.
which confidence intervals are much smaller. is is
because data for output are directly informative about
trend growth.
One interesting feature of these results is that the
decline in the natural rate is so similar across advanced
economies despite such differing trend growth compo-
nents. With the exception of Japan, the natural rate
dropped more than implied by the change in growth
rates over the same period. is suggests that some
forces other than domestic growth may be inducing
Natural rate Growth component
1. Canada
2. France
1975–79 2015–19
1975–79 2015–19
3. Germany
4. Japan
1975–79 2015–19 1975–79 2015–19
5. United Kingdom
6. United States
1975–79 2015–19 1975–79 2015–19
Sources: Holston, Laubach, and Williams (2017); and IMF staff calculations.
Note: The ranges show 90 percent confidence intervals.
Figure 2.3. Kalman Filter Estimates of the Natural Rate of
Interest for Selected Advanced Economies
(Percent)
–4
–2
0
2
4
6
–4
–2
0
2
4
6
–4
–2
0
2
4
6
–4
–2
0
2
4
6
–4
–2
0
2
4
6
–4
–2
0
2
4
6
50 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
common movements in the natural rate. at esti-
mated natural rates are more similar across countries
now than 40 years ago is perhaps consistent with the
idea that capital market integration has progressed,
at least among advanced economies. is possibility
motivates an extended version of this model, which
allows for explicit international spillovers through
either real or financial channels (and is explored in the
section “Multicountry Estimates of the Natural Rate”).
The Natural Rate during the COVID-19 Pandemic
Despite its limitation, the closed economy model is
a useful benchmark for addressing two questions that
have gained attention during the post-pandemic infla-
tionary episode in many advanced economies. at
is: How much did policymakers stimulate during the
pandemic? And how fast did they tighten afterward?
One concern when answering these questions is
that any conclusions may unduly rely on the ben-
efit of hindsight. What now might appear to be
policy mistakes may have been perfectly reasonable
decisions for policymakers without the benefit of
perfect foresight.
To illustrate the challenges, Figure 2.4 shows
different vintages of measures of the real and natural
rates. e gap between the two is a summary measure
of whether monetary policy is tight (when the realized
real rate is higher than the natural rate; the gap is posi-
tive) or loose (when the gap is negative). e measures
differ in the data they use. e full-sample estimate (in
red) uses data up to the third quarter of 2022 and so
approximates the current best guess of what the natural
rate was at each point in time. is helps provide an
assessment of the monetary policy stance with the
benefit of hindsight. In contrast, contemporaneous
estimates (in blue) are computed by repeatedly running
the model, extending the data sample by one quarter
each time. is aims to approximate how the real rate
gap might have been assessed at the time.
Early in the pandemic, the two measures differed,
often considerably and usually with the contempo-
raneous estimate presenting a much tighter view of
monetary policy. is is consistent with the idea that
the shocks seen when the pandemic hit were highly
unusual, with both supply and demand moving far
and fast. Faced with contemporaneous data, this model
viewed supply shocks as having a large permanent
component, generating an exceptionally low natu-
ral rate and thus a tight stance for monetary policy.
Subsequent data helped correct this misperception,
with the sharp change in the natural rate early in the
pandemic progressively revised away. A reasonable
interpretation is that policymakers looked through the
immediate crisis, applied their judgment in a way that
Contemporaneous estimate
Estimates based on data up to 2022:Q3
Realized real interest rate
Figure 2.4. Real Rates and Natural Rates: Contemporaneous
and Current Estimates for Selected Advanced Economies
(Percent)
1. Canada
–10
–8
–6
–4
–2
0
2
4
6
Jan.
2019
Jan.
20
Jan.
21
Oct.
21
2. France
–10
–8
–6
–4
–2
0
2
4
6
Jan.
2019
Jan.
20
Jan.
21
Oct.
21
3. Germany
–10
–8
–6
–4
–2
0
2
4
6
Jan.
2019
Jan.
20
Jan.
21
Oct.
21
4. Japan
–10
–8
–6
–4
–2
0
2
4
6
Jan.
2019
Jan.
20
Jan.
21
Oct.
21
5. United Kingdom
–10
–8
–6
–4
–2
0
2
4
6
Jan.
2019
Jan.
20
Jan.
21
Oct.
21
6. United States
–10
–8
–6
–4
–2
0
2
4
6
Jan.
2019
Jan.
20
Ja
n.
21
Oct.
21
Sources: Holston, Laubach, and Williams (2017); and IMF staff calculations.
Note: The ranges show 90 percent confidence intervals. Parameters are estimated
on pre-COVID data. The diamonds represent contemporaneous estimates at
2021:Q4 and realized real interest rates at 2022:Q4 for each country.
51International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
a model cannot, and so delivered moderately stimula-
tory policy.
10
Later in the pandemic, however, policy became
looser. And although the natural rate did rise a little in
most places, looser policy largely came about through
inflation eroding real policy rates. In contrast to the
early pandemic period, the red and blue dots are
generally very close. is says that subsequent data do
not tell us much that was not known at the time. And
so, while policymakers may have had good reasons not
considered here for conservatism in adjusting rates,
the HLW model suggests that policy was loose for a
long time in some countries (October 2022 Global
Financial Stability Report).
Multicountry Estimates of the Natural Rate
One drawback of the HLW approach is that it
involves a closed-economy model; it can only estimate
the natural rate for one country at a time. is is not
an issue when the goal is only to estimate the level of
the natural rate in a particular country. However, the
approach cannot be used for counterfactual analysis that
would try to assess something like the impact of a decline
in foreign potential growth on the domestic natural rate.
One way to address this is to use an explicitly inter-
national model. Wynne and Zhang (2018) proposed
one such framework that allows for two-way interac-
tions between two independent regions using an empir-
ical approach very similar to HLW. e framework
features an important general equilibrium aspect of the
determination of the natural rate via international spill-
overs. is is in line with international macroeconomic
theory that stipulates that when capital is internation-
ally mobile, the determination of natural rates entails
a global dimension (Clarida, Galí, and Gertler 2002;
Galí and Monacelli 2005; Metzler 1951; Obstfeld
2020). is also implies that if there are spillovers from
one country to another, then it stands to reason that
those effects might spill back over to the originator.
Specifically, the natural rate is now allowed to be
affected not just by domestic growth but also by
foreign growth. e intuition is that if foreign growth
increases, so do foreign rates of return, necessitating
greater compensation for domestic investors and driv-
ing up the domestic natural rate. Of course, changes
in the domestic natural rate affect domestic growth,
10
See Holston, Laubach, and Williams (2020) for a discussion on
how to adapt the HLW model to capture the pandemic.
which then spills back to the foreign natural rate
through a similar channel.
Figure 2.5 presents the results from such a model,
with the United States and the rest of the world as
the two regions.
11
As before, this setting suggests that
the natural rate in the United States has declined
by about 2 percentage points in the past 50 years or
so. In contrast, the estimated natural rate in the rest
of the world has been more stable, at least since the
mid-1970s. Two factors are responsible. First, as might
be expected, domestic growth rather than foreign
growth is more important for each (relatively closed)
region. Second, secular slowdown in many advanced
11
It is important to exercise caution when interpreting the quanti-
tative implications of this analysis. e estimation is not disciplined
by current account data, and so the decomposition may lump var-
ious effects together. Moreover, large confidence bands suggest that
inference is highly imprecise.
Natural rate United States growth
Rest of the world growth
Other factors
Figure 2.5. Measuring the Natural Rate: The Role of
International Spillovers
(Percent)
1. United States
–3
–2
–1
0
1
2
3
4
5
1970
80
90 2000 10 19
2. Rest of the World
–3
–2
–1
0
1
2
3
4
5
1970 80 90 2000 10 19
Sources: Wynne and Zhang (2018); and IMF staff calculations.
Note: “Rest of the world” comprises Australia, Austria, Belgium, Brazil, Canada,
China, Finland, France, Germany, Greece, India, Ireland, Italy, Japan, Korea, The
Netherlands, Norway, Portugal, Russia, South Africa, Spain, Sweden, Switzerland,
and the United Kingdom.
52 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
economies is offset by the rise of high-growth emerg-
ing market and developing economies, such as China,
propping up growth in the rest of the world. ese
elements working together have led to a higher and
more stable natural rate outside the United States.
Nevertheless, international spillovers are significant
and important for determining the level of the natural
rate. e analysis suggests two offsetting channels. e
first operates through overseas growth (in red), which
has helped support the natural rate in the United
States. e other channel is shown by the increasing
and negative impact of “other factors” (in yellow).
at this has had a long-lasting and negative effect
on the natural rate in the United States is consistent
with the idea that increased foreign demand for safe
and liquid US assets has depressed returns (Bernanke
2005; Caballero, Farhi, and Gourinchas 2008, 2016,
2017b; Pescatori and Turunen 2015), especially since
the global financial crisis. Note that the converse effect
in the rest of the world is smaller, which reflects the
relative sizes of the two regions.
Overall, this analysis suggests that foreign devel-
opments likely have had two offsetting effects on
natural rates in the United States. Sustained growth in
emerging markets has driven up the US interest rate
while simultaneously producing a glut of savings that
pulled it down again as foreign investors increasingly
demanded safe and liquid US government debt.
While more general than a closed economy model,
this framework still has an important drawback. It has
little to say about the true drivers of the changes in
the natural rate: What causes growth, either foreign
or domestic? What is behind “other factors”? e next
section tackles some of these questions.
Drivers of the Natural Rate
e aggregate macroeconomic models of the pre-
ceding sections can offer a very simple explanation for
why the natural rate has declined: While other factors
do play a role, growth—both foreign and domestic—
seems to be the most important factor. But this is
not very satisfying. “Growth” is a result of different
macroeconomic forces, not a primary force itself. For
example, while both demographic forces and pro-
ductivity growth could be responsible for the secular
decline in growth, each could have potentially very
different implications for the natural rate. Moreover,
these deeper forces may have offsetting effects not fully
captured by this simple decomposition.
Some Theory
Many possible economic mechanisms have been pro-
posed to explain variations in the natural interest rate.
eir importance can vary at different frequencies, with
“macroeconomic” forces more likely to drive long-term
trends and “financial” forces more likely to be import-
ant in the short to medium term, reflecting risk aver-
sion and leveraging cycles.
12
Of course, this distinction
is somewhat artificial because financial forces may drive
secular shifts in behavior that determines saving rates.
13
Macroeconomic Drivers
• Productivity growth: The simplest macroeconomic
theories dictate that the interest rate is pinned down
by growth in aggregate productivity. The idea is that
the rate of interest paid by a borrower must com-
pensate the lender for giving up on alternative use
of those funds, known as their “opportunity cost.”
Higher productivity growth increases the marginal
product of capital and drives up savers’ opportunity
cost, necessitating a higher interest rate to induce
them to lend (Cesa-Bianchi, Harrison, and Sajedi
2022; Mankiw 2022; Solow 1956).
• Demographics: Changes in fertility and mortality
rates have complex and time-varying effects on the
natural rate. Demographic forces have implications
for the economy’s growth rate, its dependency ratio,
and aggregate desired saving for longer retirement
(Auclert and others 2021; Carvalho, Ferrero, and
Nechio 2016; Gagnon, Johannsen, and López-Salido
2021; see Online Annex 2.3).
• Fiscal policy: Increased government borrowing can
lead to higher interest rates because more saving is
required to meet the increased demand for funds.
However, the extent to which this occurs also
depends on how much private investment is displaced
by the additional public debt (Eggertsson, Mehrotra,
and Robbins 2019; Rachel and Summers 2019).
• Market power and the labor share: The impact of
increased market power on the natural rate is ambigu-
ous. Increased market power typically depresses future
production and investment demand, weighing down
on interest rates. But it also reroutes dividends from
laborers to capital owners, with the impact on the
12
See also Rogoff, Rossi, and Schmelzing (2021) for an analysis of
real rate dynamics over the past 700 years.
13
See Eggertsson, Mehrotra, and Robbins (2019) and Mankiw
(2022) for recent reviews and Online Annex 2.3 for detailed descrip-
tion of the theoretical channels.
53International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
natural rate depending on the distribution of these
dividends across cohorts (Ball and Mankiw 2021;
Caballero, Farhi, and Gourinchas 2017b; Eggertsson,
Mehrotra, and Robbins 2019; Mankiw 2022; Natal
and Stoffels 2019; Platzer and Peruffo 2022).
• Other reasons: These include the effect of govern-
ment taxation on the profile of private consumption
and saving (Eggertsson, Mehrotra, and Robbins
2019; Platzer and Peruffo 2022), rising inequality
increasing the overall supply of savings because rich
people tend to save more than poor people (Mian,
Straub, and Sufi 2021a, 2021b, 2021c), and poten-
tial interactions between different channels.
Financial Drivers
• International capital flows and the scarcity of safe
assets: International spillovers from the integration
of global capital markets may have been powerful
drivers of the natural rate. Two main mechanisms
are at work. On one hand, high-growth emerging
markets provide alternative investment oppor-
tunities, resulting in capital outflows and raising
the natural rate in advanced economies (Clarida,
Galí, and Gertler 2002; Galí and Monacelli 2005;
Obstfeld and Rogoff 1997; Obstfeld 2021). On
the other hand, the supply of safe and liquid assets,
primarily US government bonds, has not kept pace
with fast-rising demand, especially from emerging
markets. Their ensuing scarcity may have driven
up their price and lowered their return (Bárány,
Coeurdacier, and Guibaud 2018; Bernanke 2005;
Caballero, Farhi, and Gourinchas 2008, 2016,
2017a, 2017b, 2021; Del Negro and others 2017;
Krishnamurthy and Vissing-Jorgensen 2012).
• Risk aversion and leverage cycles: The quality attributed
to particularly safe and liquid assets (for example,
government bonds in advanced economies) gives rise
to a convenience yield, which is variable and likely
to increase when global stress leads to deleveraging
(Gourinchas, Rey, and Sauzet 2022). Given the safe
haven property of the US dollar, this is especially the
case for US Treasurys whose value increases in periods
of stress, providing protection to risk-averse interna-
tional investors (Gourinchas, Rey, and Govillot 2017).
A New Theoretical Framework
To compare the quantitative impact of these dif-
ferent forces, this chapter relies on a macroeconomic
model (PP) based on Platzer and Peruffo (2022).
is is an important novelty with respect to earlier
literature because the PP model includes in one unified
framework many of the mechanisms discussed in the
previous section and so can explain how the contri-
butions from each of the corresponding economic
forces change the natural rate. is approach avoids
double-counting and having to infer the impor-
tance of each driver from different models calibrated
separately.
14
PP is a “real” macroeconomic model, in the sense
that it abstracts from nominal and financial frictions
that typically underlie cyclical fluctuations. Similarly,
for tractability, uncertainty is assumed away. While
these are reasonable assumptions for the study of
medium- to long-term trends in the real interest rate,
the model is ill-equipped to analyze the impact of the
financial drivers discussed earlier.
15
Nonetheless, PP
still allows for foreign developments to affect domestic
interest rates through their implication for net interna-
tional capital flows.
PP is calibrated to represent eight major global
economies: the United States, Japan, Germany, the
United Kingdom, France, China, India, and Brazil.
ese are the five largest advanced economies and the
three largest emerging market and developing econ-
omies, which cover some 70 percent of global GDP.
Demographic developments, the age-earning profile,
the share of income going to the richest 10 percent,
productivity trends, the retirement age, average pen-
sion replacement rates, labor share, government debt,
and public expenditure inform the country-specific
calibrations.
Before turning to detailed model simulations,
Figure 2.6 compares the overall decline in the natural
rate implied by the PP and HLW frameworks. e
striking similarity between the results obtained with two
very different approaches is reassuring. is mitigates
the uncertainty surrounding HLW point estimates while
bolstering confidence in the microeconomic structure of
the PP framework.
e first exercise for this model is to understand
why the natural rate has declined in the past several
decades. Figure 2.7 presents the estimated change in
the natural rate and its attribution to the different
fundamental forces for each of the eight countries.
14
Full details of the model are in Platzer and Peruffo (2022).
A description of specific calibration and simulations is in
Online Annex 2.3.
15
See the section “Alternative Scenarios” for quantification of the
impact of variations in the convenience yield.
54 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
While no factor clearly dominates over the past
40 years, a set of common forces has driven the natural
rate, explaining part of the international comovement.
All eight countries in the sample experienced
population aging contributing negatively to the change
in the natural rate. is effect was particularly large
in China, Japan, and Germany. Growth in total factor
productivity (TFP) declined in all advanced economies,
at times explaining far more than the final decline in
the natural rate. Fiscal policy is an important offset in
all economies, particularly Japan and Brazil. In Japan,
public debt increased by more than 200 percent of
GDP, lifting the natural rate by more than the negative
contributions from TFP growth or demographics.
In Brazil, it is mainly the large increase in public
consumption, financed by taxation, that explains the
positive contribution of the fiscal driver, even though
the increase in public debt also plays a role. e
contribution of net international capital flows, which
summarizes the net impact of global forces through
international spillovers (discussed in the context of
Figure 2.5), is significant but smaller and goes in the
expected direction.
16
e largest net negative effect
is found in the United States, potentially reflecting
that stockpiling of safe assets by emerging markets
more than offsets capital outflows drawn to attractive
investment opportunities abroad. In contrast, in Japan,
capital outflows seem to dominate, lifting the country’s
natural rate as excess domestic savings are invested in
faster-growing economies abroad. e picture is more
mixed in the three large emerging markets displayed
16
Note that while gross capital flows have increased over time as
capital accounts have liberalized, both in- and outflows have surged
since the 1970s.
Natural rate Kalman filter Natural rate structural model
Figure 2.6. Natural Rate Estimates: Model Comparison
(Percent)
Sources: Holston, Laubach, and Williams (2017); Platzer and Peruffo (2022); and
IMF staff calculations.
Note: The Kalman filter estimates are based on Holston, Laubach, and Williams
(2017). The estimates from the structural model are based on Platzer and Peruffo
(2022). The values from the structural model for 2015–19 are calibrated to overlap
with the Kalman filter estimates. The ranges show 90 percent confidence
intervals.
1. France 2. Germany
3. Japan
5. United States
4. United Kingdom
1975–79 2015–19
1975–79 2015–19
1975–79 2015–19
1975–79 2015–19
1975–79 2015–19
–4
–2
0
2
4
6
–4
–2
0
2
4
6
–4
–2
0
2
4
6
–4
–2
0
2
4
6
–4
–2
0
2
4
6
Demographics TFP growth
Fiscal Capital flows
Other Total change
Figure 2.7. Drivers of Natural Rate Changes from 1975–79 to
2015–19 for Selected Economies
(Percentage points)
–5
–4
–3
–2
–1
0
1
2
3
Sources: Platzer and Peruffo (2022); and IMF staff calculations.
Note: TFP = total factor productivity.
France Germany Japan United
Kingdom
United
States
Brazil China India
55International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
here (Box 2.3 analyzes the importance of international
spillovers for smaller emerging market and developing
economies).
17
The Outlook for the Natural Rate
So far, this chapter has focused on understanding
what has happened to the natural rate and why. While
interesting, this is perhaps less relevant for policy today
than a slightly different issue: What will happen to real
rates in the future?
The Baseline
e same framework used to understand the
drivers of the natural rate can also be used to convert
assumptions about those underlying drivers into pre-
dictions for the natural rate. e baseline projection
presented in Figure 2.8, panel 1, relies on conserva-
tive assumptions for the main drivers: (1) predicted
demographic trends follow United Nations population
projections, (2) public debt follows World Economic
Outlook (WEO) projections until 2028 (and remains
constant thereafter), and (3) all other drivers are
assumed fixed at their 2015–19 levels. In emerging
markets, (4) TFP growth is assumed to converge to
the advanced economies’ average over the long term,
as would be expected as countries get closer to the
technology frontier.
e simulation suggests that natural interest rates
are likely to stay close to pre-pandemic levels in
advanced economies. Because the demographic tran-
sition is already well underway, the residual negative
impact of further aging is expected to be moderate.
At the same time, higher public debt acts as a coun-
terweight, pushing up the natural rate. In emerging
markets, in contrast, the prognosis is for a significant
decline in natural rates. is is the consequence of
slowing productivity growth and an aging population;
in many emerging market economies, the demo-
graphic transition should accelerate in the decades
ahead. In China, for example, a steady decline in the
17
ere are also country-specific forces that drive idiosyncratic
movements in the natural rate. For example, the rise in inequality
during the past half-century has had a large negative impact on
the natural rate in the United States, even more than demographic
changes. Rising inequality is also relevant in India and Japan. e
change in market power is significant for India, which has experi-
enced a large decline in the labor share over recent decades, implying
a corresponding rise in market power in this chapter’s model.
Online Annex 2.3 provides further explanation.
natural rate by about 1.5 percentage points within
the next 30 years is projected, bringing it to about
zero in 2050.
ese projections assume that some degree of
segmentation remains between the capital markets
of advanced economies and emerging markets (see
Figure 2.2 and the analysis in Obstfeld 2021) and that
the balance of capital inflows and outflows stays as it
was in 2019.
Departures from these assumptions are used to craft
alternative scenarios.
Alternative Scenarios
e outlook for a given scenario is highly uncertain.
Many shocks could cause the natural rate to depart
from the baseline paths. And so these paths should
be thought of as illustrative, with a distribution of
future outcomes around them. Within this uncertain
outlook, some specific alternative scenarios stand out
Brazil China
France Germany
India Japan
United Kingdom United States
AE high debt
EM high debt
AE decline in convenience yield
AE higher labor share
Figure 2.8. Simulated Path for Natural Rate of Interest:
Baseline and Scenarios
1. Baseline
(Percent)
–1
0
1
2
3
4
2022 25 30 35 40 45 50
2. Alternative Scenarios
(Percentage point deviation from baseline)
–0.2
0.0
0.2
0.4
0.6
0.8
2022 25 30 35 40 45 50
Sources: Platzer and Peruffo (2022); and IMF staff estimates.
Note: AEs are France, Germany, Japan, United Kingdom, and United States. EMs
are Brazil, China, and India. The lines in panel 2 represent the difference between
the scenarios and the baseline. The decline in convenience yield is simulated in a
version of the model with a positive convenience yield; see Online Annex 2.3.
AE = advanced economy; EM = emerging market economy.
56 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
as particularly germane to the current post-pandemic
conjuncture. (1) Government debt could drift higher,
(2) enthusiasm for holding safe and liquid public
debt could wane, (3) workers’ bargaining power could
increase, (4) deglobalization forces could intensify,
and (5) the energy transition could have important
implications for global saving, investment, and the
natural rate. ese alternative scenarios are reported
in Figure 2.8, panel 2, and in Boxes 2.1 and 2.2 and
are described briefly here. All in all, deviations are
expected to be relatively modest, spanning a range
of about 120 basis points centered on the baseline
scenario. Of course, more sizable effects could be envi-
sioned should combinations of these scenarios happen
simultaneously.
• Higher government debt: As households struggle
to keep up with rising energy expenses and the
ongoing impact of the pandemic, governments may
opt to provide greater financial assistance. Allowing
public debt to increase by 25 percent of GDP above
the baseline by 2050 would increase demand for
private savings and lift the natural rate; however, the
impact should not exceed 5 to 10 basis points for
most countries.
18
• Erosion of the convenience yield, leading to higher
borrowing costs for government in advanced
economies: If investors were to perceive advanced
economies’ government debt as less safe and less
liquid than in the past (for example, if the US
Congress failed to raise the debt ceiling), then the
premium they pay for holding this particular type
of asset would erode as portfolios are rebalanced;
in this scenario, it is assumed that the premium
would return to pre-2000 average levels.
19
This
decline in the convenience yield over the next three
decades would bring up natural rates in advanced
18
e only channel modeled here is the effect of higher
demand for loanable funds from the public sector lifting the
equilibrium interest rate. Higher public debt could in principle
also erode the convenience yield, with a significant effect on
sustainability. is is considered explicitly in the next section,
“Policy Implications.”
19
By considering yield spreads between safe and liquid govern-
ment bonds and the highest-quality corporate bonds, the chapter
focuses here on the spread that most closely reflects the notion
that the convenience yield measures the unique safety and liquidity
characteristics of a government bond (Del Negro and others
2017). Other possibilities include yield spreads with lower-quality
corporate bonds or the equity risk premium (Caballero, Farhi, and
Gourinchas 2017b).
economies (and lower corporate bond yields) by
about 70 basis points.
20
• Higher labor shares in advanced economies: Markups
have increased in the past several decades, raising
the share of income going to capital owners at the
expense of workers (Akcigit and others 2021). As
workers’ bargaining power continues to improve
following the post-pandemic transformation of the
labor market, a return to labor shares prevailing in
the mid-1970s in advanced economies would raise
the natural rate by 6 to 19 basis points by 2050.
• Energy transition: Transitioning to a cleaner and
more sustainable global economy by 2050, as
laid out in the 2015 Paris Agreement on climate
change, would push global natural rates lower in
the medium term because higher energy prices
bring down the marginal productivity of capital
and investment demand. For reasonable scenarios
based on the October 2020 WEO, the effects are
expected to be relatively modest: By 2050, natural
rates are expected to decline by 50 basis points
along a hump-shaped trajectory. If large investment
in low-emission capital and technology is financed
through budget deficits, natural rates could tempo-
rarily climb by 30 basis points (Box 2.1).
• Deglobalization: With increasing geopolitical ten-
sions, the risk of some form of international trade
fragmentation—higher trade barriers, sanctions,
and the like—is elevated. Lower international
trade would push down global output and desired
investment. The effect on the natural rate would
vary across regions, reflecting the shortening of
global value chains. The risk of trade fragmentation
is compounded by the risk of financial fragmenta-
tion (April 2023 Global Financial Stability Report),
whose effect on real interest rates will depend on
countries’ initial external position: Deficit coun-
tries will find it more difficult to finance their
current accounts, while surplus countries will
repatriate excess savings, bringing down the natural
20
e model does not capture the endogenous response of capital
flows to a change in preferences for government bonds by foreign
investors. However, this effect could be sizable for safe asset providers
such as the United States. To get a sense of the possible magnitude
of the effect, it is useful to look at gross foreign portfolio investments
in the United States, which increased by about 79 percent of GDP
(US Bureau of Economic Analysis) from their average level before
2000. Were these flows to reverse, simulations show that this could
result in an increase in the natural rate of roughly 100 basis points in
the United States by 2050.
57International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
interest rate. Effects are between a 40 basis point
decline and a 20 basis point increase, depending on
the region. For trade fragmentation, the effects are
expected to be smaller (Box 2.2).
Policy Implications
Overall, the simulations previously discussed indi-
cate that natural rates will likely remain at low levels
in advanced economies, while in emerging market
economies, they are expected to converge from above
toward advanced economies’ levels. ese patterns will
have important implications for both monetary and
fiscal policy.
Monetary Policy
Once inflation is brought back to target over the
coming years, which may require a protracted period of
high interest rates (Chapter 1), the implication for mon-
etary policy seems clear: Long-term forces suggest that
natural rates will remain low (in advanced economies) or
decline further (in emerging markets), which may limit
the ability of central banks to ease policy by lowering
nominal interest rates. As a result, monetary institutions
may have to resort to the same strategies they employed
in the decade before the pandemic, such as balance sheet
policy and forward guidance. In addition, if deflation-
ary dynamics take hold, many economies may become
trapped for an extended period in a suboptimal equi-
librium characterized by low growth and underemploy-
ment (Summers 2014). To address these challenges, a
larger stabilization role may have to be assigned to fiscal
policy, and coordination between fiscal and monetary
policy might even be necessary. Reopening the debate
about the appropriate level of inflation targets, weigh-
ing the cost of permanently higher inflation against the
benefit of enhanced monetary policy space, may also
be warranted (Blanchard 2023; Galí 2020; IMF 2010;
Chapter 2 of the April 2020 WEO).
Fiscal Policy
Concerns about debt sustainability have recently
resurfaced due to the sharp increase in government debt
following the onset of the COVID-19 pandemic and
the simultaneous rise in policy rates to combat high
inflation. In this context, the key factor for debt sustain-
ability analysis is the difference between the real rate of
interest (r) and the growth rate of the economy ( g). If
growth is higher than the real interest rate, governments
may be able to sustain higher primary budget deficits
without necessarily compromising debt sustainability.
e PP model used earlier in the chapter con-
sidered the impact of the fiscal policy stance on the
natural rate, given that public debt issuance increases
demand for loanable funds. is section studies the
implications of secular movements in the natural rate
for debt sustainability. e analysis relies on a partial
equilibrium framework based on recent work by Mian,
Straub, and Sufi (2022).
21
is framework takes the
natural interest rate and growth projections from the
PP model as given and assesses debt dynamics under
different scenarios for the eight advanced and emerging
market economies presented in the preceding section.
e framework assumes that savers prefer to hold
government debt due to its liquidity and safety
features or due to regulatory requirements. is
means government debt enjoys a premium in financial
markets relative to comparable assets, known in the
literature as the “convenience yield,” which effectively
translates into a discount extended to the govern-
ment on its borrowing costs (Krishnamurthy and
Vissing-Jorgensen 2012; Wiriadinata and Presbitero
2020). However, as the public sector accumulates more
debt, government securities become less attractive to
savers, and the borrowing costs for the government
increases: e convenience yield gets eroded. Because
the interest rate increases with the debt level in this
framework, there is a limit to the size of the primary
deficit governments can sustainably run in the long
term.
22
e sensitivity of interest rates to debt is
important in this context, and its implications are
discussed at the end of this section.
21
Online Annex 2.4 describes the framework in detail. Further
references can be found in Chapter 2 of the April 2022 WEO and
Caselli and others (2022). A framework in which both channels are
mutually operable would be ideal, but it would add a significant
layer of complexity to an already very detailed framework.
22
Of course, stabilizing the debt ratio is only one criterion for
debt sustainability. Furman and Summers (2020) and Blanchard
(2023) discuss stabilizing the debt service ratio, or debt service
costs as a percent of GDP, as an alternative. Chapter 2 of the
October 2021 Fiscal Monitor discusses the merit and limitations
of this approach. In a long-term steady state in which borrowing
costs are pinned down by the natural rate of interest, stabilizing the
debt-to-GDP ratio would also stabilize the debt service ratio. e
two measures would, however, diverge over the business cycle, espe-
cially if interest rates and growth rates move in opposite directions,
as is often the case in emerging markets.
58 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
e projections from the PP model and the
elasticity of the convenience yield to the level of the
debt-to-GDP ratio are used to identify the long-term
debt-stabilizing primary balance for each level of debt.
Given current primary balances, the amount of fiscal
consolidation needed is computed under the baseline
and two of the scenarios presented earlier (high debt
and 1970s labor share). Table 2.1 shows the amount of
fiscal consolidation needed for the United States and
China, the single largest representative of each country
group in our sample.
23
For the United States, consolidation of about
3.7 percentage points of GDP is needed under the
baseline. In the higher-debt scenario, more consoli-
dation is required, at about 3.9 percentage points of
GDP. Under the higher-labor-share scenario, the dif-
ference between the natural rate and long-term growth
becomes less favorable, so that slightly greater consol-
idation is required relative to the baseline. For China,
the needed consolidation is much greater. A deficit
reduction of about 7.6 percent of GDP is required to
stabilize the debt-to-GDP ratio over the long term.
e large consolidation reflects China’s sizable primary
deficit of about 7.5 percent of GDP in 2022. In all
scenarios, it is assumed that fiscal adjustment can be
undertaken either in the near term or over the medium
term; the smaller the primary deficit in 2022, the
smaller the fiscal cost of waiting.
24
Inference about the fiscal space available to govern-
ments is of course uncertain. One important dimen-
sion of uncertainty relates to the sensitivity of interest
rates to debt. An increase in the sensitivity of interest
rates to debt essentially lowers the debt threshold at
which primary surpluses are required for sustainability
and thus erodes the fiscal space available to govern-
ments. Online Annex 2.4 conducts robustness analysis
around this parameter that highlights the importance
of building safety margins to account for changing
market conditions and investors’ risk perceptions
(Caselli and others 2022).
23
is exercise is repeated for the other six large advanced and
emerging market economies in Online Annex 2.4.
24
As noted earlier, this is a partial equilibrium exercise. Fiscal con-
solidation is bound to be more difficult if the effect of deficit reduc-
tion on real GDP is taken into account. Also, for China, the chapter
uses the definition of public debt in the World Economic Outlook
database, which uses a narrower perimeter of the general government
than IMF staff estimates in China Article IV reports. See the 2022
Article IV report on China for a reconciliation of the two estimates
and a debt sustainability assessment based on the broader perimeter
of the general government.
Conclusion
Following four decades of steady decline, real interest
rates appear to have increased in many countries in the
wake of the pandemic. While this uptick clearly reflects
recent monetary policy tightening, this chapter’s analysis
seeks to understand whether the long-term anchor—the
natural rate—has also shifted. is is of key importance
for the pricing of all assets (housing, bonds, equities)
and for monetary and fiscal policy. All else equal, higher
natural rates typically decrease fiscal space—that is,
higher primary surpluses (smaller deficits) are required
to stabilize debt ratios. But they also free up some mon-
etary policy space. Higher natural rates imply higher
nominal rates over the long term, providing central
banks with more space to react to negative demand
shocks without hitting the effective lower bound.
e chapter suggests that recent increases in real
interest rates are likely to be temporary. When inflation
is brought back under control, advanced economies’
central banks are likely to ease monetary policy and
bring real interest rates back toward pre-pandemic lev-
els. How close to those levels will depend on whether
alternative scenarios involving persistently higher
government debt and deficit or financial fragmentation
materialize. In large emerging markets, conservative
projections of future demographic and productivity
trends suggest a gradual convergence toward advanced
economies’ real interest rates.
25
25
Of course, structural policies that boost potential growth and
diminish inequalities, for example, will tend to lean against these
secular trends.
Table 2.1. Required Fiscal Adjustment under
Different Scenarios
(Changes in primary deficit, percentage points of GDP)
Scenarios
Baseline Higher Debt 1970s Labor Share
Near-Term Adjustment
United States –3.71 –3.94 –3.75
China –7.63 –7.69 –7.63
Additional Consolidation Needed for Medium-Term Adjustment (three years)
United States –0.17 –0.18 –0.17
China –0.47 –0.49 –0.47
Additional Consolidation Needed for Medium-Term Adjustment (five years)
United States –0.29 –0.32 –0.29
China –0.87 –0.93 –0.87
Source: IMF staff calculations.
Note: The required fiscal adjustment is the difference from the long-term debt-stabilization
level, calculated as the difference between the 2022 primary deficit from the World
Economic Outlook database and the model-based estimate of the primary deficit that
stabilizes debt to GDP at the long-term rates given projections for the natural rate of
interest and growth.
59International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
is means that the issues associated with the
“effective lower bound” constraint on interest rates
and “low (interest rates) for long” are likely to
resurface.
26
Unconventional policies through active
management of central bank balance sheets and
forward guidance may become standard stabilization
tools, even in emerging markets. Debates about the
appropriate level of inflation target may also reemerge
26
As discussed at length in Eggertsson and Woodford (2003) and
Adrian (2020).
as countries weigh the social cost of higher inflation
against the constraint of ineffective stabilization due
to the effective lower bound. In addition, perma-
nently lower real interest rates also increase fiscal
space—all else equal—and allow fiscal authorities to
take a more active role in stabilizing the economy,
provided fiscal sustainability is ensured (Chapter 2
of the April 2020 WEO). In this case, it is crucial
to clarify the scope and responsibilities of fiscal and
monetary authorities to avoid long-term damage to
the credibility of central banks.
60 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
Policy responses to a transition to a carbon-neutral
world will induce significant structural transformation
that will affect the natural rate (r *) via a number of
channels. is box highlights the crucial role of two
channels: the design of climate policies and the level of
international participation in their implementation.
A comprehensive and global policy package
intended to achieve net zero emissions by 2050 serves
as a benchmark, as simulated in Chapter 3 of the
October 2020 World Economic Outlook.
1
Carbon
taxes—aimed at achieving net-zero emissions by
2050—are imposed globally, starting at between $6
and $20 a metric ton of CO
2
(depending on the
country) and reaching $40 a ton in 2030 and between
$40 and $150 a ton in 2050. e package is fully
financed by the carbon tax revenues—25 percent
recycled toward social transfers, up to 70 percent for
green public infrastructure investment, and the rest
as subsidies to renewable energy sectors—making the
policy budget-neutral.
2
Maintaining budget neutrality
helps isolate the impact of the green transition on r *
absent debt-financed green investments. Although they
are subject to uncertainty and intended to be largely
illustrative, the results from simulating the policy
package yield several insights into how climate policies
can be expected to affect r *.
Different climate mitigation policies affect r* dif-
ferently. Acting alone, carbon taxes depress overall
investment and hence r * (Figure 2.1.1, panel 1). is
is because the carbon tax increases the overall cost
of energy, a complement in production to physical
capital. As a result of frictions, the associated decline
in carbon-intensive activities exceeds the investment
in renewable sources of energy and low-emission
production methods—especially in countries where
e authors of this box are Augustus Panton and
Christoph Ungerer.
1
Simulations are computed with the G-Cubed model, an
open-economy, multicountry macroclimate model (see Liu and
others 2020; McKibbin and Wilcoxen 1999).
2
e scenario differs from the investigation in Chapter 3 of
the October 2020 World Economic Outlook in two ways. First, it
assumes a budget-neutral design rather than deficit financing. Sec-
ond, given the large uncertainty surrounding the impact of green
public investment on output, the simulations take a conservative
approach and do not assume any direct productivity gains from
green public investment. Of course, any amount of progress in
total factor productivity would tend to lift the natural rate.
Net
Avoided damage
Carbon tax
Green infrastructure
Green subsidies
Transfers
Deficit-financed package
1
Budget-neutral package
2
Advanced economies
Top five emitters
Global
Figure 2.1.1. The Global Natural Rate of
Interest and the Green Transition
(Global average, percentage point deviation from
baseline)
Sources: G-Cubed model, version 164; and IMF staff
calculations.
1
The deficit-financed package is based on Chapter 3 of the
October 2020 World Economic Outlook (WEO) but is agnostic
on total factor productivity effects: front-loaded and deficit-
financed green public investment of 1 percent of GDP in the
first 10 years, 80 percent green subsidies to renewable
sectors, carbon tax revenues recycled to households (1/4),
and public debt reduction (3/4).
2
Budget-neutral package uses carbon tax revenues to
finance green public investment, green subsidies, and
household transfers in the same proportion as in Chapter 3
of the October 2020 WEO, but with a much smaller revenue
envelope.
–1.0
–0.5
0.0
0.5
1.0
2023 28 33 38 43 48 55
1. Budget-Neutral Climate Policy Package
2. Impact of Deficit-Financed versus
Budget-Neutral Policies
3. Budget-Neutral Package with Partial
Participation
2023
28 33 38 43 48
55
2023
28
33 38 43 48
55
–0.6
–0.4
–0.2
0.0
0.2
0.4
–0.6
–0.4
–0.2
0.0
0.2
Box 2.1. The Natural Rate of Interest and the Green Transition
61International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
production is carbon intensive. In contrast, public
investment in green infrastructure and subsidies
to renewable energy positively affect investment,
pushing up r *. It is also worth noting that climate
mitigation helps avoid climate-change-related dam-
ages, boosting productivity growth with respect to a
business-as-usual baseline and raising r *.
e net impact on r* depends on the associated
overall fiscal impulse. Panel 2 of Figure 2.1.1 shows an
alternative policy package that includes a temporary
deficit-financed and front-loaded green investment
push. Unlike the budget-neutral policy package, which
depresses r * along the entire transition path, this simu-
lation suggests that a deficit-financed fiscal stimulus—
because it increases demand for private savings—could
have a positive impact on r *.
e macroeconomic impact of the green transition
depends on the number of participating countries. In
Figure 2.1.1, panel 3, the climate policy package
is simulated under three different configurations,
depending on whether all countries, only the five
biggest emitters (China, European Union, India,
Japan, United States), or only advanced economies
participate. Not surprisingly, partial participation
in the program leads to a significantly more muted
impact on r *.
Overall, the short- to medium-term impact of the
green transition on r * depends on the balance of sev-
eral effects. But over the long term, r * would converge
to its pre-climate-policy steady state as economies
become greener and climate policy applies to a shrink-
ing share of economic activity.
Box 2.1 (continued)
62 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
Geoeconomic fragmentation impacts regional
economies through different channels, in particular,
trade, technology diffusion, cost of external financing,
international factor mobility, risk, and provision of
global public goods (see Aiyar and others 2023). is
box uses the IMF’s Global Integrated Monetary and
Fiscal (GIMF) Model
1
to analyze two scenarios of
trade and financial fragmentation between the “US
bloc” (United States, European Union, other advanced
economies) and the “China bloc” (China, emerging
Southeast Asia, remaining countries group).
2
To understand the impact of trade fragmentation
on the natural rate, it is necessary to grasp its impact
on saving and investment in the US and China
blocs (both deflated by the consumption price index
for comparison). e imposition of nontariff trade
barriers—which are assumed to increase by 50 percent
over 10 years—affects saving in two main ways. First,
trade restrictions tend to increase import prices for all
goods, whether intermediate, investment, or consump-
tion. Second, higher import prices for crucial production
inputs act as a negative productivity shock and reduce
output. us, by increasing the price of consumption
(the price of imported consumption goods increases by
about 5 percent to 25 percent depending on the region)
and reducing output, trade barriers tend to reduce saving
and push up the natural rate. Two opposite forces also
determine how trade restrictions impact investment.
First, higher input prices along the global value chain
lower the profitability of production in all regions,
including the “nonaligned bloc,” and depress the volume
of investment demand (see Figure 2.2.1, panel 2). At the
same time, trade restrictions directly increase the relative
price of investment goods (from their higher import
share compared with consumption goods), increasing the
demand for loanable funds, all else equal.
Overall, higher trade barriers between the US and
China blocs will reduce trade between the two regions.
is reduction is partially offset by larger trade within
blocs and with the nonaligned, but the net effect is a
shortening of the global value chain and less global trade
e authors of this box are Benjamin Carton and Dirk Muir.
1
See Kumhof and others (2010) for a description of
the GIMF Model.
2
See Chapter 4 and Online Annex 4.4 for the modification
to the GIMF Model to introduce explicit value chains and the
calibration for eight regions grouped in three blocs: the US bloc,
the China bloc, and the nonaligned bloc. e GIMF Model is
also calibrated so that intermediate inputs (in value chains) and
capital are complements in production.
(–19 percent) and output (–6 percent). Given the struc-
ture of trade, real investment in the China bloc declines
the most due to reshoring (Figure 2.2.1, panel 1).
e impact on real interest rates is modest and var-
ies across regions (Figure 2.2.1, panel 2). Real interest
rates are expected to fall by about 30 basis points in
the China bloc as investment demand declines more
than saving does. In the United States, the positive
impact of lower saving on the natural rate and the
negative impact as a result of the decline in investment
broadly balance each other out. In the nonaligned
bloc, trade diversion implies that investment demand
declines by less than desired saving, which raises the
real interest rate by about 10 basis points.
Investment goods Intermediate goods
Consumption goods Total
Figure 2.2.1. Regional Impact of Trade
Fragmentation Scenario
1. Real Interest Rate
(Percentage point deviation)
–0.4
–0.2
0.0
0.2
United States bloc China bloc Nonaligned
2. Real Investment
(Percent deviation)
–20
–16
–12
–8
–4
0
4
United States bloc China bloc Nonaligned
Sources: IMF, Global Integrated Monetary and Fiscal (GIMF)
Model; and IMF staff calculations.
Note: The fragmentation scenario is a gradual increase in
nontariff barriers between the US bloc and the China bloc for
all types of traded goods (intermediate, investment, and
consumption) over 10 years. The real interest rate is the
average over 10 years, whereas real investment is after
10 years. See Online Annex 2.5 for the country composition
of the blocs.
Box 2.2. Geoeconomic Fragmentation and the Natural Interest Rate
63International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
Geoeconomic fragmentation also has implications for
capital markets. In recent decades, and especially since
the end of the 1990s, capital market integration has
allowed advanced economies—and in particular the
United States—to benefit from low borrowing costs.
Savings from emerging markets have increasingly
sought the safety and liquidity of US government
bonds. is has helped bring down the natural rate of
interest in the United States while lifting it in surplus
countries in Asia and the Middle East (Bernanke
2005; Caballero, Farhi, and Gourinchas 2008, 2016,
2017a, 2017b, 2021). As this process reverses, the
natural rate is likely to increase in the United States
and other advanced economies while decreasing in
emerging markets. In the extreme example of a full
shutdown of capital markets, regional natural rates
would converge to levels that reflect only domestic
drivers such as demographics and productivity.
Figure 2.2.2 presents the macroeconomic impact of a
financial fragmentation scenario assuming the China bloc
reduces its exposure to the US bloc’s Treasury bonds; it
is modeled by reducing the premium paid by foreigners
on US Treasury bonds. e China bloc disposes of net
foreign assets, which pushes down their domestic interest
rate by 40 basis points. In the US bloc, the interest rate
increases by 20 basis points and the net foreign asset
position improves by 10 percent of GDP. e nonaligned
countries experience slight net capital inflows from the
China bloc, as investors look for returns, reducing their
interest rates by about 10 basis points.
Real rate
Net foreign assets (right scale)
–0.6
–0.5
–0.4
–0.3
–0.2
–0.1
0.0
0.1
0.2
0.3
Real interest rates (percentage points)
–30
–25
–20
–15
–10
–5
0
5
10
15
Net foreign assets (percent of GDP)
Sources: IMF, Global Integrated Monetary and Fiscal (GIMF)
Model; and IMF staff calculations.
Note: The fragmentation scenario is a permanent 100 basis
point premium on one bloc’s assets held by the other bloc’s
economic agents. The real interest rate and the net foreign
asset position are reported after 10 years. See Online Annex
2.5 for the country composition of the blocs.
United States
bloc
China
bloc
Nonaligned
Figure 2.2.2. Regional Impact of Financial
Fragmentation Scenario
(Deviation from baseline)
Box 2.2 (continued)
64 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
Do movements in the natural rate of interest in
advanced economies impact real interest rates in
emerging market and developing economies? And if
so, at what horizon? How strong are such associations,
and what determines their strength?
Many emerging market economies have adopted
inflation targeting—orienting monetary policy
toward domestic stabilization goals. Yet policymakers
in those countries may be unable or unwilling to
closely track global natural rates in the short term.
us, for emerging market and developing econ-
omies, real rates’ short-term dynamics may appear
disconnected from global forces (Figure 2.2 and
Obstfeld 2021). Arslanalp, Lee, and Rawat (2018)
examine real interest rates in the Asia and Pacific
region and find that a country’s capital market open-
ness is a key factor for linking domestic and global
long-term real rates.
is box’s analysis focuses instead on short-term
real rates that pertain directly to the monetary
policy stance and are less likely to be swayed
by fluctuations in risk or term premiums.
1
e
importance of global natural rates to individual
countries’ real interest rate dynamics is measured by
the contribution of the US natural rate to emerging
market economies’ individual forecast error variance
decomposition.
e authors of this box are Christoffer Koch and
Diaa Noureldin.
1
is box uses quarterly short-term deposit rates adjusted for
ex post realized inflation. e data are from the first quarter of
2020 to the fourth quarter of 2022, although coverage is uneven,
particularly toward the end of the sample. e primary data
source is the IMF’s International Financial Statistics database.
For countries with short-period gaps, the data are supplemented
with data from Haver Analytics. e emerging market and
developing economies sample consists of Algeria, Bangladesh,
Bolivia, Brazil, Cambodia, Cameroon, Chile, China, Colombia,
Costa Rica, Côte d’Ivoire, Hungary, India, Indonesia, Jordan,
Malaysia, Mexico, Nigeria, Peru, South Africa, ailand,
Türkiye, and Uganda. To avoid spurious regression, the deciding
selection factor is whether each emerging market and developing
economy rate series is cointegrated with the US rate series. e
Phillips-Perron test is used for stationarity of the residual of the
regression of emerging market and developing economy interest
rates on the US natural rate, allowing for up to four lags. Fore-
cast error variance decomposition is computed based on bivariate
vector autoregression models including the US natural rate and
individual countries’ real interest rates.
Figure 2.3.1 shows that at business cycle horizons
of less than five years, domestic real rates dominate.
At horizons beyond a decade, spillover from the
US natural rate matter just as much. is weighted
aggregation masks substantial variation across coun-
tries at longer horizons. e contribution from the
US natural rates tends to be larger for East Asian and
Latin American countries. In large emerging market
economies, such as China and India, about 30 per-
cent of real rate variation is explained by US natural
rates after a decade. After two decades, spillovers are
somewhat stronger in China than in India. Spill-
over effects to African countries, such as Cameroon,
Côte d’Ivoire, and Uganda, are minor, with less
than a 10 percent contribution from US natural rate
spillovers.
What is the role of capital account openness in
explaining this substantial variation across countries?
To gauge its importance, de facto capital openness—
the sum of foreign assets and liabilities as a percent
of GDP (IIPGDP)—is regressed on the cross-country
variation in magnitude of US natural rate spillovers
Contribution of EMDEs’ own real interest rate
Contribution of US natural rate
0
20
40
60
80
100
0 8 16 24 32 40 48 56 64 72 80
Forecast horizon (quarters)
Source: IMF staff calculations.
Note: Forecast error variance decomposition contributions
for each horizon are weighted by GDP weights adjusted for
purchasing power. EMDEs = emerging market and
developing economies.
Figure 2.3.1. Natural Rate Spillovers at
Different Horizons
(Percent)
Box 2.3. Spillovers to Emerging Market and Developing Economies
65International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
to emerging market and developing economies at
80 quarterly horizons. Figure 2.3.2 shows that the
effect of capital account openness becomes significant
only gradually after about a decade. Quantitatively,
a 1 percentage point increase in the gross interna-
tional investment position as a share of a country’s
GDP raises the importance of the US natural rate
in explaining the share of movements in emerging
market and developing economies’ real interest
rates by half a percentage point after a decade and
by 0.9 percentage point after two. So for a country
like Brazil, with an IIPGDP of about 40 percent,
20 percent of the forecast error variance decomposi-
tion of Brazilian real interest rates is attributable to
US spillovers after a decade, and about 36 percent
after two decades. is implies sizable spillovers but
at fairly low frequency.
Estimated impact
90 percent confidence interval
95 percent confidence interval
Source: IMF staff calculations.
Note: Each point on the solid blue line is the estimate of the
coefficient from a cross-section regression of the forecast
error variance decomposition share of the US real natural
rate on the emerging market and developing economies’
capital openness at the displayed forecast horizon from 1 to
80 quarters.
Figure 2.3.2. Estimated Impact of Capital
Openness on Strength of US Spillovers
(Percent)
–0.5
0.0
0.5
1.0
1.5
2.0
0 8 16 24 32 40 48
Forecast horizon (quarters)
56 64 72 80
Box 2.3 (continued)
66 International Monetary Fund | April 2023
WORLD ECONOMIC OUTLOOK: A ROCKY RECOVERY
References
Adrian, Tobias. 2020. “Low for Long and Risk-Taking.” IMF
Staff Discussion Note 2020/15, International Monetary
Fund, Washington, DC.
Aiyar, Shekhar, Jiaqian Chen, Christian H. Ebeke, Roberto
Garcia-Saltos, Tryggvi Gudmundsson, Anna Ilyina, Alvar
Kangur, and others. 2023. “Geoeconomic Fragmentation and
the Future of Multilateralism.” IMF Staff Discussion Note
23/001, International Monetary Fund, Washington, DC.
Akcigit, Ufuk, Wenjie Chen, Federico J. Diez, Romain A.
Duval, Philip Engler, Jiayue Fan, Chiara Maggi, and others.
2021. “Rising Corporate Market Power: Emerging Policy
Issues.” IMF Staff Discussion Note 21/01, International
Monetary Fund, Washington, DC.
Arena, Marco, Gabriel di Bella, Alfredo Cuevas, Borja Gracia,
Vina Nguyen, and Alex Pienkowski. 2020. “It Is Only
Natural: Europe’s Low Interest Rates (Trajectory and
Drivers).” IMF Working Paper 20/116, International Mone-
tary Fund, Washington, DC.
Arslanalp, Serkan, Jaewoo Lee, and Umang Rawat. 2018.
“Demographics and Interest Rates in Asia.” IMF Work-
ing Paper 18/172, International Monetary Fund,
Washington, DC.
Auclert, Adrien, Hannes Malmberg, Frédéric Martenet, and
Matthew Rognlie. 2021. “Demographics, Wealth, and Global
Imbalances in the Twenty-First Century.” NBER Work-
ing Paper 29161, National Bureau of Economic Research,
Cambridge, MA.
Ball, Laurence M., and N. Gregory Mankiw. 2021. “Market
Power in Neoclassical Growth Models.” NBER Working
Paper 28538, National Bureau of Economic Research,
Cambridge, MA.
Bárány, Zsófia, Nicolas Coeurdacier, and Stéphane Guibaud. 2018.
“Capital Flows in an Aging World.” CEPR Discussion Paper
DP13180, Centre for Economic and Policy Research, London.
Barsky, Robert, Alejandro Justiniano, and Leonardo Melosi.
2014. “e Natural Rate of Interest and Its Usefulness for
Monetary Policy.” American Economic Review 104 (5): 37–43.
Bernanke, Ben S. 2005. “e Global Saving Glut and the US
Current Account Deficit.” Remarks at the Homer Jones
Lecture, St. Louis, MO, April 14.
Blanchard, Olivier. 2023. Fiscal Policy under Low Interest Rates.
Cambridge, MA: MIT Press.
Borio, Claudio, Piti Disyatat, and Phuricai Rungcharoenkitkul.
2019. “What Anchors for the Natural Rate of Interest?”
BIS Working Paper 777, Bank for International Settle-
ments, Basel.
Caballero, Ricardo J., Emmanuel Farhi, and Pierre-Olivier
Gourinchas. 2008. “An Equilibrium Model of ‘Global Imbal-
ances’ and Low Interest Rates.” American Economic Review 98
(1): 358–93.
Caballero, Ricardo J., Emmanuel Farhi, and Pierre-Olivier
Gourinchas. 2016. “Safe Asset Scarcity and Aggregate
Demand.” American Economic Review 106 (5): 513–18.
Caballero, Ricardo J., Emmanuel Farhi, and Pierre-Olivier
Gourinchas. 2017a. “Rents, Technical Change, and Risk Pre-
mia Accounting for Secular Trends in Interest Rates, Returns
on Capital, Earning Yields, and Factor Shares.” American
Economic Review 107 (5): 614–20.
Caballero, Ricardo J., Emmanuel Farhi, and Pierre-Olivier
Gourinchas. 2017b. “e Safe Assets Shortage Conundrum.”
Journal of Economic Perspectives 31 (3): 29–46.
Caballero, Ricardo J., Emmanuel Farhi, and Pierre-Olivier
Gourinchas. 2021. “Global Imbalances and Policy Wars
at the Zero Lower Bound.” Review of Economic Studies 88
(6): 2570–621.
Carvalho, Carlos, Andrea Ferrero, and Fernanda Nechio. 2016.
“Demographics and Real Interest Rates: Inspecting the Mech-
anism.” European Economic Review 88 (C): 208 –26.
Caselli, Francesca, Hamid R. Davoodi, Carlos Goncalves, Gee
Hee Hong, Andresa Lagerborg, Paulo A. Medas, Anh D. M.
Nguyen, and Jiae Yoo. 2022. “e Return to Fiscal Rules.”
IMF Staff Discussion Note 22/002, International Monetary
Fund, Washington, DC.
Cesa-Bianchi, Ambrogio, Richard Harrison, and Rana Sajedi.
2022. “Decomposing the Drivers of Global R*.” Bank of
England Staff Working Paper 990, Bank of England.
Clarida, Richard H., Jordi Galí, and Mark Gertler. 2002. “A
Simple Framework for International Monetary Policy Analy-
sis.” Journal of Monetary Economics 49 (5): 879–904.
Del Negro, Marco, Domenico Giannone, Marc Giannoni, and
Andrea Tambalotti. 2017. “Safety, Liquidity, and the Natural
Rate of Interest.” Staff Report 812, Federal Reserve Bank of
New York, New York.
Eggertsson, Gauti B., Neil R. Mehrotra, and Jacob A. Robbins.
2019. “A Model of Secular Stagnation: eory and Quantita-
tive Evaluation.” American Economic Journal: Macroeconomics
11 (1): 1–48.
Eggertsson, Gauti B., and Michael Woodford. 2003. “e Zero
Bound on Interest Rates and Optimal Monetary Policy.”
Brookings Papers on Economic Activity 34 (1): 139 –235.
Furman, Jason, and Lawrence H. Summers. 2020. “A Recon-
sideration of Fiscal Policy in the Era of Low Interest
Rates.” Kennedy School of Business, Harvard University,
Cambridge, MA.
Gagnon, Etienne, Benjamin K. Johannsen, and David
López-Salido. 2021. “Understanding the New Normal: e
Role of Demographics.” IMF Economic Review 69 (2): 357–90.
Galí, Jordi. 2020. “e Effects of a Money-Financed Fiscal Stim-
ulus.” Journal of Monetary Economics 115 (C): 1–19.
Galí, Jordi, and Tommaso Monacelli. 2005. “Monetary Policy
and Exchange Rate Volatility in a Small Open Economy.”
Review of Economic Studies 72 (3): 707–34.
Goodfriend, Marvin, and Robert G. King. 2005. “e Incred-
ible Volcker Disinflation.” Journal of Monetary Economics 52
(5): 981–1015.
Gourinchas, Pierre-Olivier, Hélène Rey, and Nicolas Govillot.
2017. “Exorbitant Privilege and Exorbitant Duty.” Unpublished.
67International Monetary Fund | April 2023
CHAPTER 2
THE NATURAL RATE OF INTEREST: DRIVERS AND IMPLICATIONS FOR POLICY
Gourinchas, Pierre-Olivier, Hélène Rey, and Maxime Sauzet. 2022.
“Global Real Rates: A Secular Approach.” CEPR Discussion
Paper DP16941, Centre for Economic Policy Research, London.
Holston, Kathryn, omas Laubach, and John C. Williams.
2017. “Measuring the Natural Rate of Interest: International
Trends and Determinants.” Journal of International Economics
108: S59–S75.
Holston, Kathryn, omas Laubach, and John C. Williams. 2020.
“Adapting the Laubach and Williams and Holston, Laubach
and Williams Model to the COVID-19 Pandemic.” Federal
Reserve Bank of New York Note, Federal Reserve Bank of New
York, New York.
International Monetary Fund (IMF). 2010. “Rethinking Macro-
economic Policy.” IMF Staff Position Note 10/03, Interna-
tional Monetary Fund, Washington, DC.
Krishnamurthy, Arvind, and Annette Vissing-Jorgensen. 2012.
“e Aggregate Demand for Treasury Debt.” Journal of Politi-
cal Economy 120 (2): 233–67.
Kumhof, Michael, Dirk V. Muir, Susanna Mursula, and Douglas
Laxton. 2010. “e Global Integrated Monetary and Fiscal
Model (GIMF)—eoretical Structure.” IMF Working Paper
2010/034, International Monetary Fund, Washington, DC.
Laubach, omas, and John C. Williams. 2003. “Measuring the
Natural Rate of Interest.” Review of Economics and Statistics
85 (4): 1063–70.
Liu, Weifeng, Warwick J. McKibbin, Adele Morris, and Peter J.
Wilcoxen. 2020. “Global Economic and Environmental Out-
comes of the Paris Agreement.” Energy Economics 90: 1–17.
Mankiw, N. Gregory. 2022. “Government Debt and Capital
Accumulation in an Era of Low Interest Rates.” NBER Work-
ing Paper 30024, National Bureau of Economic Research,
Cambridge, MA.
McKibbin, Warwick J., and Peter J. Wilcoxen. 1999. “e
eoretical and Empirical Structure of the G-Cubed Model.”
Economic Modelling 16 (1): 123–48.
Metzler, Lloyd A. 1951. “Wealth, Saving, and the Rate of Inter-
est.” Journal of Political Economy 59 (2): 93–116.
Mian, Atif, Ludwig Straub, and Amir Sufi. 2021a. “Indebted
Demand.” Quarterly Journal of Economics 136 (4): 2243–307.
Mian, Atif, Ludwig Straub, and Amir Sufi. 2021b. “What
Explains the Decline in r*? Rising Income Inequality versus
Demographic Shifts.” BFI Working Paper, Becker Friedman
Institute, University of Chicago.
Mian, Atif, Ludwig Straub, and Amir Sufi. 2021c. “e Saving
Glut of the Rich.” Unpublished, Harvard University,
Cambridge, MA.
Mian, Atif, Ludwig Straub, and Amir Sufi. 2022. “A Goldilocks
eory of Fiscal Deficits.” NBER Working Paper 29707,
National Bureau of Economic Research, Cambridge, MA.
Natal, Jean-Marc, and Nicolas Stoffels. 2019. “Globaliza-
tion, Market Power, and the Natural Interest Rate.” IMF
Working Paper 19/095, International Monetary Fund,
Washington, DC.
Obstfeld, Maurice. 2020. “Global Dimensions of U.S.
Monetary Policy.” International Journal of Central Banking
16 (1): 73–132.
Obstfeld, Maurice. 2021. “Two Challenges from Globalization.”
Journal of International Money and Finance 110: 1–9.
Obstfeld, Maurice, and Kenneth Rogoff. 1997. Foundations of
International Macroeconomics. Cambridge, MA: MIT Press.
Pescatori, Andrea, and Jarkko Turunen. 2015. “Lower for Lon-
ger: Neutral Rates in the United States.” IMF Working Paper
15/135, International Monetary Fund, Washington, DC.
Platzer, Josef, and Marcel Peruffo. 2022. “Secular Drivers of the
Natural Rate of Interest in the United States: A Quantitative
Evaluation.” IMF Working Paper 2022/030, International
Monetary Fund, Washington, DC.
Rachel, Łukasz, and Lawrence H. Summers. 2019. “On Secular
Stagnation in the Industrialized World.” Brookings Papers on
Economic Activity (Spring): 1–54.
Rogoff, Kenneth S., Barbara Rossi, and Paul Schmelzing. 2021.
“Long-Run Trends in Long-Maturity Real Rates 1311–2021.”
NBER Working Paper 30475, National Bureau of Economic
Research, Cambridge, MA.
Solow, Robert M., 1956. “A Contribution to the eory
of Economic Growth.” Quarterly Journal of Economics
70 (1): 65–94.
Summers, Lawrence H. 2014. “U.S. Economic Prospects:
Secular Stagnation, Hysteresis, and the Zero Lower Bound.”
Business Economics 49 (2): 65–73.
Wiriadinata, Ursula, and Andrea Presbitero. 2020. “e Risks of
High Public Debt Despite a Low Interest Rate Environment.”
VoxEU, August 5, 2020.
Wynne, Mark A., and Ren Zhang. 2018. “Estimating the
Natural Rate of Interest in an Open Economy.” Empirical
Economics 55: 1291–318.
