Full text of Economic Perspectives (Federal Reserve Bank of Chicago) : First Quarter and Second Quarter 2010, Volume XXXIV, Issues 1 and 2

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

Federal Reserve Bank
of Chicago

Thirteenth
International
Banking
Conference

First Quarter and Second Quarter 2010

Economic__

perspectives

First Quarter
2

Interest rates following financial re-regulation
Jeffrey R. Campbell and Zvi Hercowitz

14

Measuring the equilibrium real interest rate
Alejandro Justiniano and Giorgio E. Primiceri

Second Quarter
28

What is behind the rise in long-term unemployment?
Daniel Aaronson, Bhashkar Mazumder, and Shani Schechter

52

Do labor market activities help predict inflation?
Luojia Hu and Maude Toussaint-Comeau

Economic.

perspectives

President
Charles L. Evans

Senior Vice President and Director of Research
Daniel G. Sullivan
Research Department
Financial Studies
Douglas D. Evanoff, Vice President
Macroeconomic Policy Research
Jonas D. M. Fisher, Vice President
Microeconomic Policy Research
Daniel Aaronson, Vice President
Regional Programs
William A. Testa, Vice President

Economics Editor
Lisa Barrow, Senior Economist
Editors
Helen O'D. Koshy
Han Y. Choi

Graphics and Layout
Rita Molloy
Proofreading
Sheila A. Mangier

Production
Julia Baker

Economic Perspectives is published by the Economic
Research Department of the Federal Reserve Bank of Chicago.
The views expressed are the authors’ and do not necessarily
reflect the views of the Federal Reserve Bank of Chicago
or the Federal Reserve System.
©2010 Federal Reserve Bank of Chicago
Economic Perspectives articles may be reproduced
in whole or in part, provided the articles are not reproduced
or distributed for commercial gain and provided the source
is appropriately credited. Prior written permission must
be obtained for any other reproduction, distribution,
republication, or creation of derivative works of Economic
Perspectives articles. To request permission, please
contact Helen Koshy, senior editor, at 312-322-5830 or email
Helen.Koshy@chi.frb.org.

Economic Perspectives and other Bank
publications are available at www.chicagofed.org.

s chicagofed. org
ISSN 0164-0682

Contents
First and Second Quarters 2010, Volume XXXIV, Issues 1 and 2

First Quarter
2

Interest rates following financial re-regulation
Jeffrey R. Campbell and Zvi Hercowitz

This article uses a calibrated general-equilibrium model of lending from the wealthy to the middle
class to evaluate the effects of tightening household lending standards. The authors simulate a rise
in down payment and amortization rates from their average values in the late 1990s and early 2000s
to levels more typical of the era before the financial deregulation of the early 1980s. Their results
show a drop in loan demand. This substantially lowers interest rates for an extended period.
Counterintuitively, tightening lending standards makes borrowers better off.

14

Measuring the equilibrium real interest rate
Alejandro Justiniano and Giorgio E. Primiceri
The equilibrium real interest rate represents the real rate of return required to keep the economy’s
output equal to potential output. This article discusses how to measure the equilibrium real interest
rate, using an empirical structural model of the economy.

Second Quarter
28

What is behind the rise in long-term unemployment?
Daniel Aaronson, Bhashkar Mazumder, and Shani Schechter
This article analyzes what is behind the recent unprecedented rise in long-term unemployment
and explains what this rise might imply for the economy going forward. In particular, the authors
attribute the sharp increase in unemployment duration in 2009 to especially weak labor demand
and, to a lesser degree, extensions in unemployment insurance benefits.

52

Do labor market activities help predict inflation?
Luojia Hu and Maude Toussaint-Comeau
The authors conduct an empirical analysis of the role of labor market activities in inflation and
conclude that wage growth is not very informative for predicting price inflation. But price inflation
does seem to help predict wage growth.

64

International Banking Conference
Macroprudential Regulatory Policies: The New Road
to Financial Stability?

Interest rates following financial re-regulation
Jeffrey R. Campbell and Zvi Hercowitz

Introduction and summary
Mortgages and other forms of household borrowing
typically require collateral, such as a house or car.
Typical loan contracts require borrowers to take an
initial equity stake in the collateral (the down payment)
and to increase ownership further by repaying the loan’s
principal before the collateral hilly depreciates (amor­
tization). Since the New Deal, government regulation
has substantially influenced these terms of private
contracts. In the 1940s and early 1950s, the Federal
Reserve Board imposed stringent minimum down pay­
ment rates and maximum amortization periods for home
mortgages, auto loans, and loans to purchase other con­
sumer durable goods. The suspension of these regula­
tions in 1953 allowed consumer credit to grow steadily
until the credit crunch of August 1966. The financial
deregulation wave of the late 1970s and early 1980s
triggered innovations in consumer lending that further
decreased households’ ownership stakes in their housing
and other tangible property. Many observers have blamed
precisely this deregulation for the most recent finan­
cial crisis, so it seems very possible that households’
required ownership stakes will be rising as policy­
makers look at their options for improving the regula­
tion of consumer loans and other financial contracts.
In this article, we employ a model of lending from
the wealthy to the middle class to evaluate the effects
of raising the equity requirements of household debt.
We build on our earlier analysis of the Carter-Reagan
financial deregulation in Campbell and Hercowitz (2009).
In that article, we found that lowering equity require­
ments raises the demand for household credit and
thereby increases the interest rate. This resembles the
simultaneous increases in household debt and interest
rates during the mid-1980s, even though we abstract
from rising government deficits, which are the standard
explanation for that period’s high interest rates. In this

2

article, we examine the implications of reversing this
process by increasing down payment rates for new
loans and by forcing all loans to amortize faster. The
model’s results show that this reform reduces loan
demand. The interest rate falls 78 basis points over
three years and then very slowly returns to its level
before the reform. In an alternative version of our
model in which producers cannot absorb the capital
freed by tightening household lending standards, the
interest rate falls 129 basis points over the three years
after the reform. These results are potentially of inter­
est to monetary policymakers because they can guide
an assessment of how financial market reforms im­
pact the “neutral” interest rate required to keep the
economy’s output at its potential in the absence of
business cycles.
In the model, saving households are rentiers living
off of their wealth, so the low interest rate unambigu­
ously harms them. Nevertheless, the low rate has two
beneficial effects for borrowers. First, the lower interest
rate reduces the carrying cost of debt. Second, the
lower interest rate brings down the user cost of capital
and thereby encourages investment. These investments
increase the demand for labor and thereby raise wages.
Overall, the model’s predictions show that borrowers’
welfare gains are equivalent to raising their consump­
tion permanently by 0.9 percent. If we treated the house­
hold credit market in isolation from the rest of the
economy, then this second effect would be absent.
Jeffrey R. Campbell is a senior economist in the Economic
Research Department at the Federal Reserve Bank of
Chicago. Zvi Hercowitz is a professor ofeconomics at the
Eitan Berglas School ofEconomics, Tel Aviv University.
The authors are grateful to R. Andrew Butters and Ross
Doppelt, who both provided superb research assistance.
They also thank the Pinhas Sapir Center at Tel Aviv
University forfinancial support.

1Q/2010, Economic Perspectives

In fact, such a market-by-market analysis would be
misleading; the reform makes borrowers slightly worse
off after shutting down its indirect effect on wages.
If tighter lending standards changed neither the
interest rate nor wages, then they must harm borrowers
by limiting their choices. Following this intuition about
the “direct” effects alone leads to the conclusion that
tighter lending standards primarily harm borrowers. Our
results show that this intuition can easily be overturned
by a complete equilibrium analysis that accounts for
the “indirect” effects of changing prices. Since the re­
form helps some households at the expense of others,
its assessment requires us to weight the households’
utility changes. Even with a specific assumption about
these weights, the result is only a partial assessment,
since we have nothing to say about how tightening house­
hold lending standards changes systemic economic risk.
Our article proceeds as follows. In the next section,
we review the history of interest rates and household
debt markets in the United States, paying particular
attention to households’ ownership stakes in their
tangible property. Then, we present the model and
derive its long-run implications for debt and interest
rates. We show that financial re-regulation has no longrun effect on interest rates, leaves saving households
worse off, and improves borrowers’ welfare. Finally,
we present the complete analysis of the reform.

Household debt and interest rates
in the United States
The rise of mass production techniques early in
the twentieth century created a large volume of stan­
dardized capital goods, such as automobiles, which
could serve as collateral for credit extended to house­
holds. By the 1920s, most durable household goods
could be bought “on credit” directly from their retailers.
The home mortgage market of that decade bears a
remarkable resemblance to that of the 1990s and 2000s.
First mortgages had low loan-to-value ratios, and house­
holds often financed the first mortgage’s required down
payment with second and third mortgages. All of these
mortgages matured in only a few years, and they re­
quired no repayment of principal before maturing.1
The Great Depression, World War II, and the Korean
War dramatically increased government involvement
in consumer credit markets. In the early 1930s, the
federal government purchased large volumes of “under­
water” mortgages. These were loans with principals
exceeding the value of their collateral. It then refinanced
them with 15-year amortized mortgages, which built
in the gradual repayment of the principal over the
15-year amortization period. This amortization directly
served the policy goal of avoiding a wave of mortgage

Federal Reserve Bank of Chicago

defaults arising from a sudden lack of refinancing
options. The 15-year amortized mortgage and its
30-year cousin accounted for most household debt from
the 1930s through the 1980s, even though they required
substantial down payments from borrowers.2 The move
from interest-only short-term loans to long-term am­
ortized debt reduced systemic risk at the cost of keep­
ing potential homeowners with insufficient funds for
a mortgage’s down payment out of the market. With
the onset of World War II, the Federal Reserve Board
tightened loan standards further by issuing Regulations X
and W. These dictated restrictive maximum loan-tovalue ratios and amortization periods for home mort­
gages (Regulation X) and other collateralized consumer
credit (Regulation W).
The Federal Reserve suspended enforcement
of Regulations X and W near the end of the Korean
War in 1953. Figure 1 illustrates the evolution of
credit markets since 1952. The data come from the
Federal Reserve Board’s Flow ofFunds Accounts of
the United States. The dashed line in figure 1 shows
the ratio of all mortgage debt on owner-occupied
housing relative to this housing stock’s value, and the
solid line represents the ratio of all household debt to
all tangible assets of households, which include the
stock of owner-occupied real estate and the stock of
automobiles owned by households. Since these are
both useful measures of household leverage (the use
of debt to finance investment), we refer to them hence­
forth as leverage ratios.
The wartime credit restrictions made these lever­
age ratios very low: They both equal about 0.195 in
the first quarter of 1952. Throughout the 1950s, both
ratios rise dramatically. The overall leverage ratio
(the solid line in figure 1) peaks at 0.38 in the fourth
quarter of 1965. At that time, the Federal Reserve’s
Regulation Q placed a cap on the permissible interest
rate paid on savings accounts. During the credit crunch
of August 1966, market interest rates exceeded this
cap, and the resulting outflow of funds from savings
and loans and other traditional sources of mortgages
reduced the availability of household credit.
The mid-1960s marked a turning point for house­
hold leverage ratios. They declined (not always steadily)
until the enactment of the Gam-St Germain Depository
Institutions Act in the last quarter of 1982. This act and
the Monetary Control Act of 1980 eliminated many
restrictions on mortgage lending. Along with the con­
current growth of mortgage debt securitization, these
changes fueled a second wave of post-war household
leverage growth. In the first quarter of 1983, both ratios
equaled about 0.30. By the first quarter of 1995, they
both equaled 0.41.

3

Throughout the credit expansion of
the late 1990s and the early 2000s, these
ratios rarely exceeded 0.45. Home prices
began to decline in the middle of 2006,
mechanically raising the household lever­
age ratios. This continued until the first
quarter of 2009, when both ratios equaled
about 0.58. The most recently available
data come from the second quarter of
2009, and they show the leverage ratios
declining. Of course, the leverage ratios’
common recent spike emanated from a
loss in the value of previously mortgaged
properties rather than from any deliberate
loosening of mortgage terms. With their
mortgages considered underwater, many
homeowners chose to delay repayment or
default outright. The financial turmoil
that arose from the resulting impairment
of mortgage debt has led most observers
to reassess the need for tighter mortgage
standards. Therefore, we expect these
household leverage ratios to continue
their declines as creditors write off their
bad debts (thus reducing household in­
debtedness) and as lenders raise required
down payments and principal repayment
rates on newly issued loans. Furthermore, the possi­
bility of congressionally mandated changes to finan­
cial market regulation might either directly or indirectly
lead to tighter standards for household credit.
We expect tighter loan standards to reduce demand
for credit, thereby lowering interest rates. To get a sense
of how much lower we could expect them to go, we
plot the yield on three-year constant-maturity zerocoupon U.S. Treasury debt in figure 2. To account
for the effects of anticipated inflation on these interest
rates, we have subtracted from each of them the most
recent four-quarter percentage change in the Personal
Consumption Expenditures Price Index. The yield’s
average over the time period plotted (the fourth quarter
of 1953 through the third quarter of 2009) is 2.6 percent.
The most noticeable feature of the data is the
familiar rise in real interest rates associated with the
Federal Reserve’s policy of targeting the growth rate
of money that began in the fourth quarter of 1979 and
ended in the fourth quarter of 1982. To get a better sense
of the relationship between credit demand and inter­
est rates, we remove this period and that of the recent
financial crisis from the analysis. For the remainder,
we have calculated average interest rates for the peri­
ods defined by turning points of the household lever­
age ratios in figure 1: These are 1953:Q4-1966:Q3,

4

1966:Q4-1979:Q3, 1983:Q1-1995:Q4, and
1996:Ql-2007:Q2. The results are 1.94 percent,
1.33 percent, 4.50 percent, and 2.45 percent. Thus,
it appears that the interest rate rose at the same time
household leverage ratios were growing in the 1980s
and early 1990s, and a decline in interest rates accom­
panied the end of both growth spurts in figure 1. An
explanation of interest rates that focuses only on house­
hold leverage ratios is clearly incomplete. For exam­
ple, contemporaries attributed the high interest rates
of the 1980s to that era’s high government deficits.3
Nevertheless, the association between interest rates
and changes in household leverage seems strong enough
to merit further quantitative exploration. We next present
a theoretical framework for doing so.

A model of household debt and
interest rates
Much of modem macroeconomic theory builds
on the useful fiction that identical infinitely lived
households populate the economy. This will not do
for the question at hand because two identical house­
holds have no incentive to lend to each other. Accord­
ingly, our model of household debt and interest rates
has two representative households, which we call the
borrower and the saver. The borrower is less patient

1Q/2010, Economic Perspectives

of government-set equity requirements. To
consider the effects of anticipated increas­
es in equity requirements on the interest
rate, we now reverse that experiment by
raising the equity requirement parameters.
Next, we present the model of house­
hold debt and interest rates. We begin by
describing the two households’ preferences.
We then lay out the economy’s technolo­
gy for producing goods, and we finish
with a discussion of both households’
consumption and savings choices in a
competitive equilibrium.

than the saver. The difference in patience motivates the
(heads of) househoids to five up to the names we have
assigned them, ff the borrower’s debts were iimited oniy
by her ability to repay them, then she wouid never stop
borrowing more. As time passes, she wouid spend more
and more on interest payments and fess and fess on her
own consumption.4 This is grossiy unrealistic for the
United States as a whoie. Another feature of our modei—
coiiaterai requirements—inhibits the never-ending
expansion ofdebt. In the Iongrun, the saver’s consumptionsavings decisions determine the interest rate. At that rate,
the borrower wouid like to expand her debts. However,
the coiiaterai requirement inhibits her from doing so.
As noted previously, most household debts require
the borrower to hold an equity stake in the good serv­
ing as collateral. The borrower’s down payment is the
equity stake at purchase, and the equity stake grows
as the borrower repays the loan’s principal, hi the model,
two parameters determine the borrower’s equity re­
quirements. Because the history of household debt in
the United States indicates that government regulation
substantially influences equity requirements, we view
the two equity requirement parameters as being set by
policy.5 In Campbell and Hercowitz (2009), we mod­
eled the expansion of leverage following the financial
market deregulation of the early 1980s as a reduction

Federal Reserve Bank of Chicago

Consumer choices
Both the saver and the borrower value
the consumption of two goods. The first
good is nondurable and stands in for items
such as food, energy, and entertainment
services. The second good represents the
use of durable goods such as housing,
furniture, automobiles, and consumer
electronics. Both individuals can adjust
their consumption of these goods once
every calendar quarter.
We denote the quantity of the nondu­
rable good consumed in quarter t by the
borrower with C,. The analogous quantity for the saver
is C,. Similarly, we represent the quantities of the dura­
ble good used by the borrower and saver in quarter t
with S, and S,. Henceforth, we use A and A to repre­
sent borrower- and saver-specific versions of A.
If these households are to make consumption and
savings decisions, then they need to know how to trade
off nondurable and durable consumption in the present
quarter and how to balance consuming more of either
good today versus saving to enable more consumption
in the future. For this, we suppose that they plan how
much of both goods to consume in the present quarter
and in everyfuture quarter. We denote a plan for the
borrower’s nondurable consumption from quarter t
onward with C'= (c,,C,+1,C,+2, ...) = ^C,,C'+1). The

borrower’s analogous plan for durable consumption
is S' = (s„S,+1,S,+2, ...) = (s,,S'+1). We suppose
that for each possible plan, the borrower computes a
utility value C/^C', S'), using the following formula:

C/(c',S') = 01nS, +(l-0)lnC, +PU(C'+1,S'+1).

The parameters 0 and (3 both lie between zero
and one. This says that the utility value of following
a plan equals the fe/icity from the current quarter’s

5

consumption, 01nS, + (l- 0)lnC,, plus the value of
continuing to follow the plan discounted by (3.6
The saver’s utility value of a given plan can be
calculated from his analogous equation:

C/(c',5') = 01n5,+(l-0)lnC,+pt?(c'+1,5'+1).

The value of 0 here equals its value in the borrower’s
utility rule, so both households agree on how to divide
an allocation of income for the current quarter between
nondurable goods and (the services from) durable
goods to make felicity as large as possible. However,
the saver’s discount factor (3 exceeds the borrower’s
discount factor (3. In this sense, the borrower is less
patient than the saver. The borrower would prefer to
trade the saver’s best possible consumption plan for
one of equal cost, but with higher consumption in the
present and lower consumption in the future.

Production of income and accumulation of wealth
Each quarter, the economy inherits three stocks
of capital goods from the previous quarter. The first is
the stock of market capital. We denote the number of
machines in the stock of market capital available in
quarter t with W(. Combining these machines with Nt
hours of work (provided in principle by either house­
hold) yields an output of Yt = K“N'~a, measured in
units of the nondurable consumption good. After pro­
duction, a fraction A, of the machines stop working.
Investments in machines, /(, can replace those lost to
depreciation and (if sufficiently large) expand the
stock of machines available for the next quarter. Thus,
Kl+l=(l-X)Kl+I,.
The remaining two stocks inherited from the pre­
vious quarter are the two households’ stocks of home
capital, that is, durable goods. We assume that the
flow of services from a stock of home capital is pro­
portional to its size, so that we use S, and S, to rep­
resent each of the households’ durable goods stocks
as well as the flows of services forthcoming from
them. Just as with market capital, the home capital
goods depreciate and can be replaced and expanded
with investment. The two stocks’ common deprecia­
tion rate equals 5, and their respective investments
are X, and Xr Therefore,

5,+1=(l-5)5,+W„

6

and

5,+1=(l-5)5,+A,.
All income in the economy can be directed to­
ward one of the following uses: each household’s
nondurable consumption, investment in each house­
hold’s stock of home capital, or investment in the
stock of market capital. It is costless to convert one
unit of income into one unit of any of these goods.
Since the uses of income cannot exceed that avail­
able, we have

C.+C.+X.+X.+I,^ Y„
The households face two other substantial limits
on their accumulation of capital. First, the machines
in the stock of market capital may not be converted
into consumption goods of either kind. This makes
sense for most capital goods because blast furnaces
and airliners are of little use to the typical consumer.
We impose this limit by requiring that It > 0. Second,
neither household may sell durable goods from their
stocks of home capital. That is, Xt>0 and X, > 0.
Obviously, households can and do sell their durable
goods all of the time. However, we find this assump­
tion reasonable when we suppose that the model’s
borrower and saver represent two classes of individu­
als with different tastes. If the saver is rich and con­
sumes mansions while the borrower is middle class
and consumes bungalows, then the restriction means
that we cannot convert mansions into bungalows and
vice versa.
Trade and competition
We have now described how the two households
rank consumption plans and the technology available
for implementing them. We will now present how the
households implement these plans by reviewing a
typical quarter’s trades in the sequence they occur.
We then describe the collateral requirements that re­
strict the households’ debts and finish with a presen­
tation of the conditions required for markets to clear.

The sequence of trades in a quarter
At the beginning of quarter /, the households own
their stocks of durable goods; stocks of market capital,
K, and K,; and financial assets (bonds), B, and B,.
Production takes place at a representative firm.
It rents capital from the households and combines it
with labor to produce income. The cost of renting one
machine in quarter t is Ht, and the cost of one hour of

1Q/2010, Economic Perspectives

work is
Capital and labor employed at each firm
are chosen to maximize its profits. After production
takes place, the representative firm makes its required
rental payments to the owners of capital, returns the
undepreciated capital goods to their owners, and pays
its wage bill. We think of the saver as representing the
wealthiest families in the United States, so we suppose
that he spends all of his time on leisure activities and
offers none to the labor market. The borrower represents
the middle class, so we suppose that she offers A hours
of work to tlie market regardless of the wage she earns
for each one. Thus, the saver’s wage income equals
zero always, while the borrower’s is WN.
The funds available to either household is the sum
of that household’s labor earnings, the rents it receives
for the use of its market capital, and its stock of bonds.
It can put these funds to one of four uses. Three of
these—nondurable consumption, investment in home
capital, and investment in market capital—have already
been covered. The fourth use of funds is the purchase
of new bonds. All bonds pay one unit of the nondura­
ble consumption good in the next quarter, and their
price in the current quarter is \!Rt, where Rt is the
gross rate of interest. With this in place, we can write
the two households’ budget constraints as

Ct + X, +1, + Bl+i /R^N + HJf+B,

and

C,+A,+/,+B,+1/^< ll,K, ■ B„

Collateral requirements
A household can choose any positive value of bonds
(A’z. i or /?,.) that is consistent with its budget constraint.
When either of these bond stocks is negative, we say
that household is indebted. An indebted household
must pledge some or all of its home capital stock as
collateral. We denote the maximum debts that can be
collateralized by the two households’ home capital
stocks with V, and V,. So, we require:

-B,< Vt
and
-B< V,.

Federal Reserve Bank of Chicago

We specify these maximum debts with

Vl+l =(1-^)F; + (1-

and

Here, 1 - it is the maximum loan-to-value ratio allowed
for household debt, and <|> is the rate at which the princi­
pal must be repaid. If 4> = 8, then the borrower must
repay the principal only to the extent that depreciation
erodes the collateral’s value. If instead <[> > 5, then the
borrower must accumulate equity in the collateral as
it ages. We adopt the specification requiring the geo­
metric repayment of principal because it greatly sim­
plifies the ensuing analysis.

Market clearing and equilibrium
The evolution of the model economy can be com­
pletely described by a collection of plans for current
andfuture nondurable consumption, durable consump­
tion, market capital, and collateral values, as well as
the sequences of the wage rate, the rental rate of capital,
and the interest rate. We say that such a collection is
an equilibrium if the households’ consumption plans
maximize their utility values given their incomes; the
representative firm maximizes its profit given the wage
and interest rate; and the demands for bonds, market
capital, and labor always equal their corresponding
supplies. The interested reader can find a more tech­
nical definition of equilibrium in box 1.
The model’s steady state
Next, we examine how the steady-state values of
the model’s key outcomes change with parameters so
that we can gain intuition valuable for interpreting the
model’s dynamics. By definition, a steady state is an
equilibrium in which all of the variables are constant
over time. Therefore, a household’s borrowing con­
straint binds either always or never. It is tedious but not
difficult to show that only the less patient household’s
borrowing constraint binds in the steady state.
For our purposes, the three key variables of inter­
est are the interest rate and the two households’ lever­
age ratios (their stocks of outstanding household
debts divided by the values of their household capital
stocks). To characterize all of these variables, we first
need to consider both households’ optimal consump­
tion and savings choices. Suppose that the saver begins
with a utility-maximizing steady-state consumption
plan with nondurable consumption C and changes it
slightly by decreasing consumption in year t by A > 0,

7

BOX 1

Equilibrium definition

Building upon the notation we used for the two
households’ consumption plans, we denote the path
for any quantity or price .4 with A' = (A, A , ...).
For a collection of plans to form an equilibrium,
they must satisfy the following five conditions.
1)

Given
V,, R', H1, and IV', the bor­
rower’s plans for C', S', X', K', I', B', and V'

■ are consistent with the initial given values
of Kt, S„ Bt, and Ij;
■ obey the rules for accumulating market
capital, home capital, and collateral value;
■ satisfy the borrower’s borrowing and
budget constraints in every quarter; and
■ yield a higher utility value for the borrower
than any other plans that satisfy this condi­
tion’s other requirements.
2)

Given Kt,St,Bt, Vt,R', H', and W', the saver’s
plans for C', S', X', K', I', B', and V'

■ are consistent with the initial given values
of£, S„ /?,,and (<;
■ obey the rules for accumulating market
capital, home capital, and collateral value;

investing the proceeds in bonds, and consuming the
principal and interest in year t + 1. By its construction,
this experiment leaves consumption in all years after
t + 1 unchanged. If A is small, then the utility loss
in year t is A / C and the discounted utility gain in
year t + 1 equals (3RA / C. Here, R is the steady-state
interest rate. Since the original consumption plan max­
imized utility, this slight change cannot increase utility.
The change also cannot lower utility because if it did,
then the analogous experiment that increases consump­
tion in year t by borrowing A and repaying it in year
t + 1 would increase utility. Therefore, we have that:

■ satisfy the saver’s borrowing and
budget constraints in every quarter; and
■ yield a higher utility value for the saver
than any other plans that satisfy this condi­
tion’s other requirements.
3)

For all r > 0, 5(+1 + 5(+1 = 0.

4)

ForallT>0,^(+i = f(+i+^(+1.

5)

For all t > 0, X’ and Nare the capital and labor
choices that maximize the representative firm’s
profits given H and IF.

The first two conditions just require each of
the households to do the best they can (measured
with their utility values) with what they have got.
The third condition states that the net supply of
risk-free bonds in the economy equals zero. Thus,
if one household wishes to borrow, the other must
lend. The fourth condition says that the economy’s
stock of market capital must equal the sum of the
two households’ market capital stocks. And the final
condition requires the rental rate of capital and the
wage rate to induce the profit-maximizing represen­
tative firm to rent the entire available capital stock
and employ all of the available hours of work.

back at the interest rate 1 / (3 in year t + 1 would increase
her utility. However, the collateral requirements pre­
vent her from doing this. Since the borrower exhausts
her borrowing opportunities in the steady state, we
can calculate her leverage ratio as:

(1- 7t)S

B
a-

<t>

’

Thus, increasing either ji or <J> directly reduces the
borrower’s leverage ratio in the long run. Since the
saver purchases bonds, we set his leverage ratio to zero.

Quantitative analysis of increasing
equity requirements
Eliminating common terms from both sides yields
our first important result, R = 1 / [3. That is, the saver’s
discount rate determines the steady-state rate of inter­
est alone. Credit market regulation that changes either
3i and 0 has no long-run effect on the interest rate.
Since the borrower is less patient than the saver,
the experiment of borrowing A in year t and paying it

8

Although the steady-state analysis reveals that
equity requirements have no long-run effect on inter­
est rates, it does not rule out substantial short-run effects
in the wake of a reform. Investigating this possibility
requires a quantitative analysis of the model’s equi­
librium, which we provide here. For this, we assign
values to the model’s parameters that reflect the equity

1Q/2010, Economic Perspectives

TABLE 1
Calibrated parameter values
Equity requirement
High

Low

ji

<1

0.16

0.0315

0.11

a

X

6

0.3

0.025

0.01

P

P

1

1

1.01

1.015

e

0.37

0.0186

Note: See the text for further details.
Source: Campbell and Hercowitz (2009).

requirements of household debt typical of the late 1990s
and early 2000s. After calculating the model’s steady
state with these values, we raise the equity requirement
parameters to values more typical of the period be­
fore the financial deregulation of the early 1980s. We
then calculate the model’s equilibrium paths for all
quantities and prices when households start with the
capital and debt stocks from the initial steady state
(associated with low equity requirements) but face the
new higher equity requirement parameters. In the long
run. the economy’s interest rate, its capital stocks, and
the debt owed by the borrower to the saver converge
to their values in the steady state calculated with the
new parameters. We focus on the model economy’s
transition from the initial steady state to the other
steady state following the parameter change.
Table 1 lists the parameter values we use for this
experiment. All of them are taken from our earlier anal­
ysis of credit market deregulation in Campbell and
Hercowitz (2009). We consider two configurations for
the equity requirement parameters: high and low. In
both cases, ji equals the average of typical down pay­
ments on homes and automobiles weighted by their
shares of durable purchases, and 0 equals the average
repayment rates of home mortgages and automobile
loans weighted by their shares of household debt. The
parameters for the high regime were chosen using ob­
servations of household debt and loan terms from before
the financial liberalization of 1983, while the choice
of the low regime’s parameters used similar observa­
tions from 1995 through 2001. The required down pay­
ment for a home capital good equals 16 percent of its
value in the high regime and 11 percent in the low re­
gime. The model’s remaining parameters are held con­
stant across the two regimes. Campbell and Hercowitz
(2009) provide justification for the specific values chosen.
We note here only that the choice of (3 produces an
annual steady-state interest rate of 4.02 percent.
For our experiment, we start the economy at the
model’s steady state calculated with the parameters

Federal Reserve Bank of Chicago

from the low regime. We suppose that, without
warning, the parameters switch to those of the high
regime. Both of the model’s households expect the
change to be permanent. Given the initial values of
S,,V,, B„ Bt, S,, and Kt, from the steady state associated
with low equity requirements, we calculate the model’s
equilibrium. Figure 3 contains plots of the resulting
equilibrium paths for the model’s key variables. Panels A,
B, C, and D plot the values of both households’ consump­
tion choices, and panels E and F display the evolution
of the productive capital stock and the wage. All of these
have been scaled so that their values in the original
steady state equal 100 percent. Panel G shows the interest
rate in annual percentage points, and panel H shows
the household leverage ratio in percentage points.
In the model, there are two reasons for the borrower
to purchase durable goods: They create a flow of valu­
able services, and they enable the expansion of debt.
The re-regulation of household debt markets reduces
the size of this second incentive, and so the reform
initially makes the borrower wish to reduce her stock
of durable goods. Indeed, the borrower purchases «o
durable goods for six quarters following the re-regulation
(figure 3, panel A). This decline in durable purchases
together with the acceleration of principal repayment
required by the higher value of <j) reduces loan demand,
so both the interest rate and the household leverage ratio
fall as expected. The leverage ratio starts at 38.17 per­
cent, falls rapidly while the borrower purchases no
durable goods, and then declines more gradually toward
its new long-run level of 23.37 percent (figure 3,
panel H). The interest rate falls rapidly from its initial
value of 4.02 percent to its trough three years after
re-regulation, 3.24 percent—a decrease of 78 basis
points (figure 3, panel G). Thereafter, the interest rate
rises very slowly towards its steady-state value. Even
25 years after re-regulation, the interest rate is 36 basis
points below its original value. Apparently, it takes a
long time indeed to reach the long run.

9

FIGURE 3

The model’s equilibrium following financial re-regulation
A. Borrower’s durable goods (St)

B. Borrower’s nondurable consumption (C,)

C. Saver’s durable goods (S,)

D. Saver’s nondurable consumption (C,)

E. Stock of market capital (K,)

F. Wage (ItV,)

G. Annual Interest rate (400 x (R, - 1))

H. Household leverage ratio (-B,/(S, +S,))

Notes: Panels A through F indicate the variable relative to its value in the initial steady state, which has been set to equal 100 percent
(the dashed horizontal line). The values on the vertical axis in each panel are the variable’s minimum and maximum values attained
in the 100 quarters following re-regulation.

10

1Q/2010, Economic Perspectives

A note on welfare
TABLE 2
In this article, we have examined
Consumption-equivalent welfare changes
interest rates in the wake of the deregula­
Borrower
Saver
tion and re-regulation of financial markets.
Appropriate monetary policy requires
Baseline experiment (percent)
0.9
-8.4
understanding and forecasting persistent
Fixed K (percent)
-0.1
-4.8
interest rate changes, so our results can
Note: See the text for further details.
contribute to that discussion. However,
for those who set financial market policy,
the interest rate serves only as a means to
This would be impossible if the interest rate she pays
an end. Policymakers instead concern themselves
on her debts and the wage she receives for her labor
with how adopting a given policy changes the welfare
were held constant. Of course, both of these variables
of borrowers and savers. In the model, we can measure
also change in the short run, and the changes are favor­
welfare with the two households’ utility values after the
able to the borrower: The interest rate falls, and the
policy change. Comparing these with the analogous
wage rises. These two are actually tightly connected.
utility values from the pre-reform steady state provides
The interest rate decline increases the capital employed
the desired welfare assessment.
by the representative firm, which in him raises wages.
Before reporting on the actual welfare changes,
Put differently, the re-regulation induces the saver to
it is worth returning to figure 3. It shows that after
invest more in productive capital and thereby benefit
25 years, the saver consumes much less of both goods
the borrower indirectly with higher wages.7
than he did before the reform (panels C and D). At
To determine whether the “direct” effect of lower
the same time, the borrower consumes more of both
interest rates or the “indirect” effect of higher wages
goods (panels A and B). Although the economy has
contributes more to the borrower’s welfare gain, we
not yet reached its new steady state in that time, these
have mn an experiment with the model in which we
changes also characterize the long run. Therefore, the
hold the stock of market capital fixed at its original
reform unambiguously increases the borrower’s wel­
steady-state level. Put differently, we force the saver
fare while decreasing the saver’s.
to replace depreciated market capital and do not allow
The long-run welfare changes are only tangentially
any further investment. In this experiment, the inter­
interesting for policymakers; they care about the total
est rate falls 129 basis points (to 2.73 percent) over
welfare change that accounts for the short-run transi­
three years; the wage remains constant by construction.
tion from one steady state to another. In the short
In table 2, the row for fixed K reports the consumption
run, the borrower’s consumption of both goods falls
equivalent welfare changes analogous to those from the
(figure 3, panels A and B). The saver’s nondurable
previous experiment. Even though the fall in interest
consumption slowly trends down (panel D). The saver’s
rates is much larger than before, the borrower’s welfare
durable purchases rise to peak at about 10 percent above
gain becomes a loss. The change also cuts the saver’s
their pre-reform level, and then fall to their new steady
welfare loss substantially. Apparently, the indirect effects
state value (panel C).
of financial re-regulation on consumer welfare can
In principle, the borrower’s short-run utility loss
easily dominate its more easily envisioned direct effects.8
could dominate her welfare calculation. This would
Since tightening consumer lending standards helps
be intuitive because re-regulation imposes a constraint
one household at the expense of the other, it is impos­
on her decisions. The actual utility changes reported
sible to unambiguously state that such a policy change
in table 2 show that this is not the case. The utility
helps or hurts “society as a whole.” A policymaker
values themselves have no meaningful units, so all
who cares only for the borrower would prefer tighter
of the table’s entries give the permanent percentage
lending standards, while one who represents the saver’s
change in the consumption of both goods (starting
interests would be against them. A policymaker who
from the original steady state) required to make the
wishes to keep both households’ considerations in mind
household’s utility equal to its post-reform value.
can come to either conclusion depending on the weights
In table 2, the first row reports the results for the
she assigns to the two households’ preferences. We
experiment plotted in figure 3. The borrower’s welfare
have been silent regarding how many “real” households
change equals that from permanently and instantly rais­
the borrower and saver each represent because that
ing her consumption of both durable and nondurable
detail is actually irrelevant for the model’s equilibrium.
goods by 0.9 percent. The borrower is better off, even
As long as no single household thinks that it can
though she faces tighter constraints on her borrowing.

Federal Reserve Bank of Chicago

11

influence the wage or interest rate, nothing changes if
we divide either household into 10, 100, or 1,000
smaller (but identical) households.
Nevertheless, the number of “actual” borrowers
and savers clearly matters for a policymaker’s welfare
calculations. In our favored interpretation of the model,
the saver represents the 5 percent or 10 percent of
households with the highest wealth, and the borrower
represents the remainder. If 5 percent of households
are savers, then tightening lending standards increases
the average utility value of all households. However, the
same tightening decreases average utility if 10 percent
of households are savers. Therefore, we have little con­
crete advice to give a policymaker who wishes to base
her judgment on changes in average utility. That is, we
can identify winners and losers from tightening lend­
ing standards, but assessing whether or not this im­
proves society lies well beyond our capabilities.

Conclusion
Empirically, times of expanding home leverage
have had higher-than-average interest rates. Interest
rates in the United States during the post-Korean
War surge in household leverage were about 60 basis

points higher than their average in the period immedi­
ately after the leverage ratio had peaked. Similarly,
interest rates fell about 200 basis points when the sec­
ond sustained increase in household leverage ratios
ended in 1995 (recall our discussion of figures 1 and 2).
Of course, macroeconomic events other than changes
in credit market regulation substantially influence in­
terest rates. Nevertheless, these results give a range
within which reasonable model predictions for the
interest rate effects of financial re-regulation should
fall. In the baseline version of our model in which
the saver accumulates market capital, the interest rate
falls 78 basis points over three years after financial
re-regulation. Thereafter, the interest rate rises very
slowly back to its original level. If we instead suppose
that the stock of market capital is fixed and cannot be
augmented, the analogous decline is about 130 basis
points. These two specifications embody two extreme
assumptions on the costs of adjusting market capital:
none and infinite. Accordingly, we argue that any
persistent decline in interest rates between 78 basis
points and 130 basis points is a reasonable forecast
in the wake of financial re-regulation.

NOTES
’See Semer et al. (1986) and Olney (1991) for more information
about household credit markets before the Great Depression.

2Green and Wachter (2005) provide a history of the spread of amortized
mortgages in the United States.
3See Friedman (1992) for a discussion of government deficits and
interest rates in the 1980s. Campbell and Hercowitz (2009) argue
that rising demand for credit must have contributed to that decade’s
high interest rates because otherwise household indebtedness would
have declined as government deficits increased interest rates.

4Becker (1980) describes this long-run behavior of household indebt­
edness in detail.

5For an alternative view, see Kiyotaki (1998). He discusses one
environment of limited commitment in which the creditors require
down payments because collateral loses value upon repossession.
6To calculate U(C‘9 S'), choose a large integer r and artificially
set U (C‘+x, S,+i) to zero. Next, use the equation to calculate
U(C‘+'~\S‘+'~1\U(C^~2,S‘+1~2),..., U(C‘,S‘). This calculation
is obviously incorrect because the assumption upon which it is
predicated is false. However, the error will generally be proportional
to (3T, which gets very small as r becomes larger.

12

7It is important to note here that the borrower’s welfare increase does
not reflect a paternalistic assumption built into the model that regu­
lators can make better financial decisions than individual borrowers.
Instead, it reflects the benefits accruing to all borrowers from them
simultaneously reducing their loan demand. In this sense, financial
re-regulation has the same effects as would the formation of a bor­
rowers’ cartel to limit the demand for loans. All of the borrowers
are made better off if they stick to the cartel agreement, but each
one of them would like to deviate and expand her indebtedness so
long as the others conform.
8In this experiment, the saver’s welfare improves when his choices
over market capital are restricted. Just as before with the borrower’s
welfare following financial re-regulation, this welfare improvement
can be interpreted as a cartelization of savers. If all savers commit
to not increasing market capital, they can all avoid paying higher
wages on the transition path. This increases their welfare, even
though it further reduces the interest rate. Of course, each individual
saver would like to expand his purchases of market capital if all
other savers stick to the cartel agreement.

1Q/2010, Economic Perspectives

REFERENCES

Becker, R. A., 1980, “On the long-run steady state in
a simple dynamic model of equilibrium with hetero­
geneous households,” Quarterly Journal ofEconomics,
Vol. 95, No. 2, September, pp. 375-382.
Campbell, J. R., and Z. Hercowitz, 2009, “Welfare
implications of the transition to high household debt,”
Journal ofMonetary Economics, Vol. 56, No. 1,
January, pp. 1-16.
Friedman, B. M., 1992, “Learning from the Reagan
deficits,” American Economic Review, Vol. 82, No. 2,
May, pp. 299-304.

Kiyotaki, N., 1998, “Credit and business cycles,”
Japanese Economic Review, Vol. 49, No. 1, March,
pp. 18-35.
Olney, M. L., 1991, Buy Now, Pay Later: Advertising,
Credit, and Consumer Durables in the 1920s, Chapel
Hill, NC: University of North Carolina Press.
Semer, M. P., J. H. Zimmerman, J. M. Frantz,
and A. Ford, 1986, “Evolution of federal legislative
policy in housing: Housing credits,” in Housing and
the New Financial Markets, R. L. Florida (ed.), New
Brunswick, NJ: Center for Urban Policy Research,
pp. 25-31.

Green, R. K., and S. M. Wachter, 2005, “The American
mortgage in historical and international context,”
Journal ofEconomic Perspectives, Vol. 19, No. 4,
Fall, pp. 93-114.

Federal Reserve Bank of Chicago

13

Measuring the equilibrium real interest rate
Alejandro Justiniano and Giorgio E. Primiceri

Introduction and summary
In conducting monetary policy, policymakers find it
usefiil to monitor the performance of the economy
relative to some benchmark. For instance, the policy
decision whether to raise or lower the short-term nominal
interest rate might be affected by the deviations of cur­
rent inflation from policymakers’ comfort zone, of
output from potential output, and of the real interest
rate (current nominal rate minus expected future in­
flation) from its equilibrium value (the rate that would
be consistent with output at its potential level). Unfor­
tunately, these benchmark concepts are not directly
observed in the data, but can only be defined in the
context of a specific theoretical framework.
Over the past decade, the new Keynesian model
has become the workhorse for the analysis of mone­
tary policy. This model departs from the neoclassical
framework of the 1980s by assuming imperfect com­
petition in goods and labor markets and “sticky” (mean­
ing rigid or inflexible) prices and wages—neoclassical
models assume prices and wages are flexible and ad­
just quickly. These ingredients in the new Keynesian
model alter the transmission of fundamental shocks
perturbing the economy and allow monetary policy
to have temporary real effects.
The equilibrium real interest rate is a crucial
concept in the new Keynesian class of models. This
rate represents the real rate of return required to keep
the economy’s output equal to potential output, which,
in him, is the level of output consistent with flexible
prices and wages and constant markups in goods and
labor markets (Woodford, 2003; and Gall, 2008).1 Mean­
while, the difference between the ex ante real interest
rate—the nominal interest rate minus expected infla­
tion—and the equilibrium real interest rate is defined
as the renZ interest rate gap.
In the new Keynesian model, the real interest rate
(RIR hereafter) gap is central to the determination of

14

output and inflation. Loosely speaking, if this RIR gap
is positive, output will decline relative to potential.
This is because people will be inclined to postpone
spending decisions today to take advantage of higher
returns to savings. All else being equal, a negative out­
put gap will then put downward pressures on prices and
wages because of weaker aggregate demand. Converse­
ly, a negative RIR gap will typically be associated with
a positive output gap, setting in motion inflationary
forces—higher demand leads to higher prices.
The main policy implication of this observation
is that policymakers concerned with maintaining out­
put close to its potential level should set short-term
nominal interest rates—the policy instrument of most
central banks—in order to minimize the RIR gap. In
the absence of a trade-off between stabilizing inflation
and output, this simple policy prescription would also
completely stabilize inflation. In practice, however,
there may well be a trade-off between the two objectives
of output and inflation stabilization.2 Nonetheless, the
equilibrium RIR constitutes a natural benchmark for
the conduct of monetary policy, and the RIR gap can
be viewed as providing some indication of the stance
of monetary policy (Neiss and Nelson, 2003).
While the equilibrium RIR is theoretically appeal­
ing, its use in guiding monetary policy decisions fac­
es at least two major hurdles. First and foremost, the
equilibrium RIR is not directly observable in the data,
limiting its usefulness as a target for monetary policy
in practice.3 Moreover, rather than being constant, the

Alejandro Justiniano is a senior economist in the Economic
Research Department at the Federal Reserve Bank of
Chicago. Giorgio E. Primiceri is an assistant professor in
the Department ofEconomics at Northwestern University.
The authors are grateful to Anna Paulson, Richard Porter,
Spencer Krane, and seminar participants at the Federal
Reserve Bank of Chicago for helpful comments.

1Q/2010, Economic Perspectives

equilibrium RIR fluctuates over time in response to a
variety of shocks to preferences and technology that
perturb the economy.
Second, setting nominal interest rates to track the
equilibrium RIR may not be feasible at times because
of the existence of the zero bound', that is, nominal in­
terest rates cannot be set lower than zero. Indeed, the
equilibrium RIR may fall enough to induce a positive
RIR gap, even with the nominal interest rate at zero.
Output would then decline below potential, engender­
ing deflation. In this way, the gap helps us to gauge the
constraint imposed by the zero bound on monetary
policy. With short-term nominal interest rates now
at historically low levels in the United States and a
number of other industrialized countries, this scenario
is receiving a lot of attention from both the academic
community and policymakers.
Given the importance that the equilibrium RIR
plays for the design of monetary policy in modem
macroeconomic models, our purpose in this article
is to provide an estimate of this unobservable vari­
able. We do so by inferring it from an empirical new
Keynesian model fitted to U.S. quarterly data on a
few key macroeconomic variables from 1962:Q1
through 2008:Q4.4
Specifically, our analysis accomplishes three
objectives. First, we describe the historical evolution
of the equilibrium RIR. We find that this rate has
been negative at times, particularly in the late 1970s
and, most interestingly, during the latest recession.
Second, we estimate the short-term RIR gap as
the difference between the current (as opposed to fu­
ture) ex ante RIR and the equilibrium RIR. This pro­
vides some indication of the stance of monetary policy.
Consistent with the anecdotal view, the estimated short­
term RIR gap suggests that policy was loose during
most of the 1970s. In contrast, policy would seem to
have been tight at the end of our sample. However,
this mostly reflects the zero bound problem—policymakers’ inability to lower short-term nominal interest
rates below zero—and provides a rationale for the
nonconventional policy measures undertaken by the
Federal Reserve during the most recent recession, such
as direct purchases of longer-term securities and the
creation of special facilities and programs (for example,
the Term Asset-Backed Securities Loan Facility, or
TALF) intended to increase access to credit.
Finally, we compare the evolution of the short­
term and long-term RIR gaps, where the latter is de­
fined as the sum of the current and expected future
short-term RIR gaps or, alternatively, the difference
between the ex ante long-term RIR and the equilibrium
long-term RIR. Long-term rates reflect the path of

Federal Reserve Bank of Chicago

current and expected future short-term rates. There­
fore, long-term gaps summarize private expectations
about future macroeconomic outcomes and monetary
policy, providing a more forward-looking measure of
the policy stance. For instance, according to this mea­
sure, policy was not loose in the 2002-06 period, which
preceded the recent economic downturn. This charac­
terization of the policy stance contrasts with what is
suggested by the short-term RIR gap and, in particular,
with the view of several commentators (see, for in­
stance, Taylor, 2007).
Several papers have tackled the estimation of the
equilibrium RIR before, most notably Laubach and
Williams (2003) and Kozicki and Clark (2005). In con­
trast to these earlier studies, our estimate of the equi­
librium RIR is based on a micro-founded model, which
builds on the optimizing behavior of households and
firms seeking to maximize their utility and profits. In
this respect, this article is related to the approach of
Neiss and Nelson (2003), Amisano and Tristani (2008),
and, in particular, Edge, Kiley, and Laforte (2008). How­
ever, in contrast to these earlier studies, we stress the
importance of both current and expected future RIR
gaps for the determination of macroeconomic outcomes.
As with all empirical work based on structural
models, our results may be sensitive to some aspects
of the model specification. To illustrate this point, we
compare our results across two models that differ in
scale, shocks, and transmission mechanisms of these
disturbances.
The article is organized as follows. First, we pro­
vide a brief description of our baseline model economy.
Then, we describe the data and the estimation approach.
Next, we present the main results—that is, we present
our estimates of the equilibrium RIR and RIR gaps. We
also discuss the robustness of these estimates when
inferred from a larger-scale model. We conclude with
a few comments and caveats to our analysis, particu­
larly with regard to the current economic situation.
More specifically, we note how the larger-scale model
also suggests the presence of positive short-term and
long-term RIR gaps for the fourth quarter of 2008. This
provides a further rationale for the Federal Reserve’s
response to the current crisis with nonconventional mea­
sures to ease monetary policy. We do, however, em­
phasize the need to enhance these models’ ability to
capture the interplay between the financial sector and
the real economy, particularly in light of recent events.

The model
In this section, we sketch our baseline new
Keynesian model and analyze two of its key equilib­
rium relations—the aggregate demand and supply

15

equations. The presentation is mostly narrative, with
most of the technical details relegated to the appendix.
Interested readers can refer to Justiniano and Primiceri
(2008) for greater details on the model, or they can
see the comprehensive treatment of new Keynesian
models in Woodford (2003) and Gali (2008), as well
as the excellent primer by Gali and Gertler (2007). For
simplicity, relative to Justiniano and Primiceri (2008),
the model here abstracts from the roles of habit forma­
tion, indexation, and endogenous capital accumulation.
We present the results based on a larger-scale model
with these additional features as a robustness check
in a later section.
There are five types of agents in our model econ­
omy: 1) households, 2) employment agencies, 3) firms
producing intermediate goods, 4) firms producing final
goods, and 5) the monetary authority. We now briefly
describe the behavior of each of them.
Households
We assume that we have a large number of house­
holds seeking to maximize their stream of current and
expected future utility, which depends positively on
their consumption of a single final good and negatively
on the number of hours they work for the production
of intermediate goods. Each household is the sole sup­
plier of a specialized type of labor that it sells to the
employment agencies in exchange for wages. Rather
than taking wages as given—as under the neoclassi­
cal assumption of perfect competition—each house­
hold has some market power and can post its wage.
This, in turn, determines the amount of their special­
ized labor demanded by the employment agencies.
We introduce sticky wages in the labor market
by assuming that at each point in time only a random
fraction of households can change their posted wage.
Hence, when setting its wage, each household takes
into consideration not only current but also future
labor demand and costs of working. For example, if
future labor demand is expected to rise, households
will preemptively post higher wages, since they
might not be able to do so in the near future.
Finally, all households have access to savings
through two types of assets: one-period government
bonds and state-contingent securities, which pay only
if a certain future state is realized. The former are used
to smooth consumption over time. State-contingent secu­
rities serve instead to insure against the idiosyncratic
risk arising from the uncertainty about the length of time
before households will be able to reset their wages.

Employment agencies
Employment agencies mediate the demand
and supply of labor between households and firms

16

producing intermediate goods. Their role is to purchase
all types of specialized labor supplied by households
and bundle them into a single homogenous labor in­
put sold to intermediate goods firms. Employment
agencies operate in a perfectly competitive market,
taking the wage received for the labor bundle as given
and making zero profits.

Intermediate goods producers
A large number of intermediate goods producers
combine technology with labor inputs purchased from
employment agencies to produce differentiated inter­
mediate goods, which are then sold to final goods
producers. Each of the intermediate goods producers
has some market power and can therefore post the
price of its good. This, in turn, determines the amount
of its output demanded by the final goods producers.
We introduce sticky prices in the goods market
by assuming that at each point in time only a random
fraction of firms can change their posted price. Hence,
when setting its price, each firm takes into consideration
not only current but also future demand and marginal
costs, where the latter depend on wages. For example,
if future demand is expected to rise, producers will
preemptively increase prices, since they might not
be able to adjust them in the near future.
Final goods producers
Final goods producers mediate between interme­
diate goods producers and households. They produce
the final good by bundling all intermediate goods into
a single final homogenous commodity purchased by
households. Final goods firms maximize profits as well,
but in contrast to the intermediate goods producers,
they operate under perfect competition, taking the price
for the final good as given and making zero profits.

Monetary authority
The central bank determines monetary policy by
setting the short-term nominal interest rate in response
to price inflation and output growth. This interest rate
rule is a variant of the instrument rule proposed by
Taylor (1993), the Taylor rule, which approximates
the historical behavior of the U.S. federal funds rate.
According to this rule, nominal interest rates rise
more than one-to-one with inflation and fall in re­
sponse to output contractions.

Demand, supply, and the equilibrium RIR
Before presenting our estimation results, we
highlight the main insights of the two crucial equilib­
rium relations in the model. This helps explain the
roles of the equilibrium RIR and RIR gaps in the
determination of output and inflation.

1Q/2010, Economic Perspectives

Aggregate demand
In the model, aggregate spending is determined
by the behavior of the representative household, which
seeks to smooth consumption over time by investing
its savings in one-period government bonds. This
optimizing behavior results in the following (loglinearized) aggregate demand equation, which is
also known as the IS equation:
J)

y, = E,y,+i^r„

where yt and rt are output and the RIR, respectively,
and the hat symbol (*) denotes deviations from the
equilibrium level. Hence, yt denotes the output gap,
and rt stands for the short-term RIR gap. Intuitively,
according to the aggregate demand equation, fluctua­
tions in the short-term RIR gap induce deviations of
the output gap from its expected future value, E,y,+},
where the operator Et denotes households’ expecta­
tion of future values conditional on the information
available today.
Equation 1 can be iterated forward to express the
output gap today only as a function of the current and
expected future short-term RIR gaps. This procedure
yields the expression

2)
7=0

by which the output gap is negatively associated
with the long-term RIR gap. The latter corresponds
to the sum of current and expected future short-term
RIR gaps.5 Notice, therefore, that if the long-run RIR
gap is negative, the output gap will be positive, and
vice versa.
Aggregate supply
In terms of the supply side, intermediate goods
Anns set prices according to the current and expected
fiiture evolution of marginal costs and demand condi­
tions. Profit-maximizing behavior results in the fol­
lowing (log-linearized) aggregate supply or Phillips
curve equation:
3)

7i(=p£,jt(+1 + o( + X ,,

where jt, and st stand for price inflation and real mar­
ginal costs, respectively, and \ (is a markup shock
that represents exogenous variation to the level of mark­
up desired by intermediate goods producers. Finally,
P is a constant very close to one that represents the
temporal discount factor, and k is a positive constant

Federal Reserve Bank of Chicago

that is inversely related to the degree of price sticki­
ness. Intuitively, inflation exceeds its expected future
level either if real marginal costs increase or if inter­
mediate goods firms change their desired markup of
prices over marginal costs for other reasons exogenous
to the model.
To highlight the importance of the RIR gap for
inflation determination, we briefly analyze a special
case of our model obtained by assuming perfectly
flexible wages. Under this assumption, real marginal
costs are proportional to the output gap. Hence, all
else being equal, a positive output gap will cause in­
flation to rise relative to its expected future level. More­
over, if the output gap is projected to remain positive
in the future, expected future inflation will also increase,
further fueling the rise in current inflation. That is, cur­
rent and expected future RIR gaps engender pressures
on prices through their effects on aggregate demand.
This crucial insight also holds in our general model
with wage rigidities, although with sticky wages the
link between the output gap and real marginal costs
is more complex.

RIR gaps and monetary policy
Equations 1 and 3 highlight the importance of RIR
gaps for output and inflation determination. Current
and future expected deviations of ex ante RIRs from
their corresponding equilibrium values affect the out­
put gap, which, in turn, influences the inflation rate.
Since the ex ante RIRs depend on the nominal interest
rates set by the monetary authority, the conduct of
monetary policy is central to the behavior of the RIR
gaps and, hence, output and inflation.
Consider, for instance, a central bank that seeks
to stabilize price inflation and the output gap. Absent
any markup shocks (Xn(), the central bank can achieve
full stabilization of both output and inflation by com­
mitting to set nominal interest rates according to an
appropriate instrument rule that delivers a zero RIR
gap at every point in time.
However, as we mentioned in the introduction,
tracking the equilibrium RIR may not be feasible when
the zero bound on nominal interest rates becomes bind­
ing. Put another way, sometimes the equilibrium RIR
may fall enough that, even with the short-term nominal
interest rate at zero, positive RIR gaps would emerge.
In this case, according to the model, output would
decline relative to potential and inflation would fall.
Even abstracting from the zero bound, in practice
optimal monetary policy is more involved than the
simple prescription of tracking the equilibrium RIR.
This is due to the fact that markup shocks bring about
a trade-off between stabilizing the output gap and in­
flation.6 Nonetheless, despite these considerations,

17

the equilibrium RIR remains an important reference
point for the conduct of monetary policy, assuming
that it can be accurately estimated and forecasted.
This is the task we undertake next.

Model solution and estimation
In this section, we provide a brief overview of
the approach that we adopt to estimate the model’s
parameters and to infer the evolution of the latent
(unobservable) variables. The discussion is somewhat
technical, although we do not aim to provide a com­
prehensive overview of the techniques we used. For
more details on these techniques, interested readers
should refer to An and Schorfheide (2007).
Model solution and state-space representation
The model we described in the preceding section
has a solution of the form

4)

|, = G(0)|,_1 + AT(0)e„

where the state vector collects all variables except
for the shocks. The elements of are expressed in
(log) deviations from the model’s nonstochastic steady
state, which corresponds to the constant values of all
variables that the economy would converge to in the
absence of shocks. The shocks inducing temporary
deviations from the steady state are stacked in the
vector §,. Meanwhile, G (0) and M (0) are matrices
whose elements are functions of the vector of model
structural parameters, denoted by 0. Our goal is to es­
timate these parameters and to uncover the historical
behavior of the unobserved variables in the state vector.
In fact, while some elements of the state vector are
directly observed in the data (for instance, inflation and
output), others are not (such as the equilibrium RIR
and expected inflation). Therefore, in order to esti­
mate the model, equation 4 must be combined with an
additional set of equations specifying which elements
of the state vector are observed in the data.
The general form of this additional set of equa­
tions is

5)

x,=Z(^ + C(0)),

where Z is a matrix mapping the elements of ( into
.r (the vector of observable data) and where C is a
vector of constant terms (which may depend on 0)
representing the steady state of the observable elements
of (. Equations 4 and 5 constitute the transition and
measurement equations of a linear state-space model.

18

Data
We estimate the model, using five series of U.S.
quarterly data: 1) real per capita gross domestic product
(GDP), 2) per capita hours worked, 3) real per capita
wages, 4) quarterly inflation, and 5) the short-term
nominal interest rate. We constmct real GDP by dividing
nominal GDP by the population aged 22-65 and the
GDP deflator.7 For hours, we use a measure of hours
in all sectors of the economy following Francis and
Ramey (2008). This is also our source for the popula­
tion series. Real wages correspond to nominal compensa­
tion of employees from the U.S. Bureau of Economic
Analysis’s national income and product accounts
(NIPAs), divided by hours and the GDP deflator; for
the nominal interest rates, we use the effective federal
funds rate. The sample period spans 1962:Q1 through
2008:Q4.8 We do not de-mean or de-trend any series.
Bayesian inference
The state-space representation of the model allows
us to use a very powerful algorithm known as the Kalman
filter to estimate the parameters 0 and retrieve the most
likely path of the unobservable elements of (,• We
discuss each in turn.
A natural way to estimate the model is to find the
value of the parameters 0 that maximizes the likelihood
function. The likelihood function summarizes all in­
formation about 0 contained in a sample of data and
plays a pivotal role in econometrics and statistics. The
likelihood function of our state-space model can be
evaluated using the Kalman filter.
In practice, however, the likelihood function
associated with most modem macroeconomic models
is typically a complicated nonlinear function of the
model parameters. This makes finding a unique value
that maximizes the likelihood a rather arduous task.
For this reason, most of the recent literature estimating
macro models has turned to Bayesian methods, which
discipline the set of plausible values for 0 through the
use of prior information.
Bayesian inference then seeks to characterize the
distribution of 0 that results from combining the like­
lihood function with the prior information. This is known
as the posterior distribution, from which we can com­
pute the location of a parameter (mean or median)
and a measure of uncertainty. For instance, the uncer­
tainty surrounding 0 can be conveyed by reporting
posterior probability bands that contain the range of
values that parameters are likely to take with, say,
99 percent probability.
Prior beliefs about the elements of 0 may be in­
formed by theory or simply reflect and summarize

1Q/2010, Economic Perspectives

FIGURE 1

Equilibrium real interest rate, 1962-2008

Note: The dashed lines are the 99 percent posterior probability bands.
Sources: Authors’ calculations based on data from Haver Analytics and
the U.S. Bureau of Labor Statistics.

information not contained in the estimation sample.
In practice, this prior information is formulated by
specifying a certain distribution for each element of
the parameter vector, centered at a particular value
(mean) and with an associated measure of uncertainty
(standard deviation).
Once we have estimated the model’s parameters,
we can employ the Kalman filter to sequentially and
systematically update our guess for the unobserved
elements of the state vector. More precisely, at each
point in time, our guess for
based on data avail­
able in the previous quarter, is updated after we ob­
serve the data for the current period. This filtered
(or one-sided) estimate for the state vector forms the
basis for our guess on the value of the state vector
next period, which we also update once we have data
for the next quarter, and so on.
Having followed this procedure for all periods,
we can go back and revise the filtered estimate of
conditional not only on information up to time t but
also on the whole sample of data. We call the state
vector emerging from this procedure the smoothed
(or two-sided) estimate. We analyze these estimates
in the next section.

Equilibrium RIR and RIR gaps
in the estimated model
We do not report the estimated parameters in this
article. They are similar to those of Justiniano and
Primiceri (2008), who use a longer sample. Here, we

Federal Reserve Bank of Chicago

focus on our estimates of the equilibrium
RIR and the RIR gaps.
The equilibrium RIR
Figure 1 plots the smoothed estimate
of the equilibrium RIR (solid blue line).
It is also important to characterize the
uncertainty surrounding the estimated
equilibrium RIR, particularly since this
is cited as a possible concern regarding
its usefulness for monetary policy analy­
sis. Therefore, we also report uncertainty
bands (dashed black lines), which repre­
sent the values this variable is likely to
have taken with 99 percent probability.
We first highlight a few properties of the
smoothed estimate and later discuss
these probability bands.
The first thing to notice is that the
inferred equilibrium RIR has fluctuated
substantially over our sample, with a
standard deviation of 1.94 percent around
a mean of 2.6 percent (annualized).9
A second interesting feature of figure 1 is that the
equilibrium RIR has turned negative in a few instances.
This occurred around 1975 and the end of2008—two
recession dates, as determined by the National Bureau of
Economic Research—and during the 2003-04 period.
These episodes were characterized by a substantial
decline in the federal funds rate in response to weak
economic conditions. However, the 2008 episode is
the only one in our sample for which the uncertainty
bands are completely below zero.
Indeed, the third interesting observation is that
the equilibrium RIR has plummeted in the latest part
of the sample. In particular, during the latest recession,
the equilibrium RIR seems to have recorded by far its
largest decline, with an estimate for 2008:Q4 of roughly
-2.15 percent.
The tightness of the posterior probability bands
deserves some comment. In particular, the precision
with which the equilibrium RIR is estimated perhaps
seems implausible, especially considering that these
bands account for the uncertainty surrounding both
the unobserved states and the model parameters. It is
important to keep in mind, however, that these proba­
bility bands abstract from model uncertainty. That is,
alternative specifications of the model (for example, a
different historical characterization of U.S. monetary
policy or a model with additional propagation mecha­
nisms and/or shocks) might deliver different esti­
mates of the equilibrium RIR. We return to this issue
in the section explaining the larger-scale model.

19

This being said, the cross-sectional
dispersion at different points in time is
larger than perhaps suggested visually by
figure 1. For example, figure 2 plots the
posterior distribution of the equilibrium
RIR for the last point in the sample,
2008:Q4. Values of the equilibrium RIR
are on the horizontal axis, with the verti­
cal line drawn at the median of-2.15
percent, which coincides with the esti­
mate reported in the previous figure.
Notice that this distribution has a range
from roughly -4 percent to -0.5 percent,
with hardly any weight assigned to val­
ues close to zero. Therefore, our modelbased estimates suggest that it is quite
likely that the equilibrium RIR became
negative in 2008. To what extent did this
induce positive RIR gaps? We address
this key issue next.

The short-term RIR gap
The ex ante RIR is given by the dif­
ference between the nominal interest rate
and the inflation rate expected for next
quarter. While the former is directly ob­
servable in our data, the latter is part of
the unobservable state vector and must
be backed out using the Kalman filter.
Figure 3 shows the smoothed esti­
mate of the ex ante RIR (blue line) to­
gether with the equilibrium RIR (black
line). The mean of the ex ante RIR is
2.37 percent (annualized) with a standard
deviation of 2.45 percent. These statistics
are similar to those corresponding to the
equilibrium RIR. The overall contours of
these two series coincide, although they
have differed at times.
In order to highlight the discrepan­
cies between the ex ante RIR and the
equilibrium RIR, figure 4 plots their dif­
ference together with its 99 percent prob­
ability bands. We refer to this difference
as the short-term RIR gap, in order to
distinguish it from the long-term gap that
we analyze next. Note that the short-term
gap has also fluctuated considerably over
time, with an average of-0.33 percent
and a standard deviation of 1.28 percent.
As we noted earlier, the short-term
RIR gap is commonly taken as a measure
of the monetary policy stance. And indeed,

20

1Q/2010, Economic Perspectives

To this end, figure 5 compares the
short-term RIR gap (blue line) with the
Short-term real interest rate gap, 1962-2008
evolution of the long-term one (black
line). Although the two series often move
together—the correlation coefficient is
equal to 0.56—the message about the
stance of monetary policy implied by
the two lines differs during some his­
torical episodes.
The 2002-06 period provides an in­
teresting example. In 2002:Q3 the federal
funds rate stood at 1.75 percent, but it had
declined to 1 percent by 2003 :Q3, and re­
mained there for the next three quarters.
The federal funds rate then rose gradually,
reaching 5.25 percent in 2006:Q3. Some
have argued that monetary policy was
too accommodative during this period
Note: The dashed lines are the 99 percent posterior probability bands.
(for example, Taylor, 2007). Although
Sources: Authors’ calculations based on data from Haver Analytics and
the negative value of the short-term RIR
the U.S. Bureau of Labor Statistics.
gap seems to accord with this claim (blue
line), the positive value of the long-term
RIR gap (black line) does not support the
view that policy was too expansionary. In particular,
at least for some episodes, the evolution of the RIR
gap aligns well with the anecdotal characterization of
it suggests that the private sector expected a decline
of the equilibrium RIR or a monetary tightening.
monetary policy that we see in the literature. For in­
stance, according to our estimates, the equilibrium RIR
The difference between short-term and long-term
exceeded the ex ante real interest rate during most of
gaps toward the end of the sample is also informative.
the 1970s, exactly when U.S. inflation was at histori­
For instance, our estimate of the short-term RIR gap
cally high levels. This is consistent with the view that
in 2008 :Q4 is roughly 1.5 percent. This suggests that,
monetary policy during this period was characterized
according to the model, the federal funds rate of 0.5
by an insufficient response to the rise in inflation (Clarida,
percent was probably above the equilibrium RIR. Fur­
Gali, and Gertler, 2000). Similarly, the significant in­
thermore, it suggests that the zero bound on nominal
crease in the short-term RIR gap in the early 1980s
interest rates would have been binding before addi­
accords well with the conventional view that the disin­
tional interest rate cuts could have closed the short­
term RIR gap. In addition, the estimated long-term
flation in the U.S. economy was engineered by a sub­
stantial policy tightening under then-Federal Reserve
RIR gap exceeds 3 percent. Taken at face value, this
Chairman Paul Volcker.
would suggest that at the end of 2008, positive short­
term gaps were expected to persist and the zero
The long-term RIR gap
bound was expected to bind beyond a single quarter.
While the behavior of the short-term RIR gap pre­
Before we interpret this result as indicative of
sented in figure 4 squares quite well with the convention­
contractionary monetary policy, we must acknowl­
al view, there are a few caveats that call for caution in
edge that these gaps can only reflect the stance of
interpreting this gap as a good proxy for the stance of
conventional monetary policy. By this we mean the
monetary policy. In particular, as we explained earlier,
Federal Reserve’s management of the short-term
it is important to recognize that the whole path of ex­
nominal interest rate. During the current economic
pected future short-term RIR gaps—rather than just its
crisis, the Federal Reserve has also employed a vari­
contemporaneous value—matters for the determination
ety of nonconventional policy measures; and these
of output and inflation in the new Keynesian model (see
measures have been reflected in the changing size and
equation 2, p. 17). From this perspective, we might judge
composition of the Federal Reserve’s balance sheet.
the monetary policy stance better by looking at the long­
Our simple analysis suggests that these measures
term RIR gap, which summarizes the information con­
have been appropriate, insofar as both the short-term
tained in the current and expected future values of the
RIR and long-term RIR exceeded the equilibrium
federal funds rate, inflation, and the equilibrium RIR.
FIGURE 4

Federal Reserve Bank of Chicago

21

RIR. However, these extraordinary mea­
sures are not reflected in our analysis of
the short-term and long-term RIR gaps.

A larger-scale model
The baseline model can be summa­
rized in a few simple equations that, as
discussed, clearly highlight the role of
the equilibrium RIR for the dynamics
of output and inflation. This simplicity,
however, comes at the expense of ab­
stracting from other features that impart
more realism to the model. In particular,
additional shocks can be included and
other mechanisms added (such as
endogenous capital accumulation)
through which disturbances influence
the evolution of the economy. For this
reason, we test the robustness of our
main conclusions by using a larger-scale
model estimated on a richer data set.
This extended model is discussed in
Justiniano and Primiceri (2008) and
is based on the well-known studies of
Christiano, Eichenbaum, and Evans
(2005) and Smets and Wouters (2007).
Relative to our baseline model, the larger-scale
model includes the additional propagation mecha­
nisms provided by endogenous capital accumulation,
investment adjustment costs, a choice of capital utili­
zation, habit formation in consumption, and index­
ation in both prices and wages. These features are
essentially meant to increase the length of time for
which a given shock will affect the evolution of the
economy. There are three additional disturbances per­
turbing the model economy, specifically, shocks to the
marginal efficiency of investment, to the disutility of
labor, and to government spending. Finally, we esti­
mate the model over the same sample, 1962:Q1 through
2008:Q4, but we incorporate additional data on con­
sumption and investment.
Figure 6 reports the smoothed estimates of the
equilibrium RIR and the ex ante RIR, as well as the
short-term and long-term RIR gaps. In each panel,
the black line reproduces the estimates from the base­
line model and the blue line corresponds to estimates
from the extended model.
Panel A highlights the fact that the cyclical pat­
tern of the equilibrium RIR is very similar across models,
although the equilibrium RIR is substantially more
volatile in the larger-scale model.10 One implication
of this finding is that, according to the extended model,
the equilibrium RIR has declined below zero more

22

frequently than what is predicted by our baseline frame­
work. Furthermore, the decline in the current down­
turn, while substantial, is not as dramatic by historical
standards as suggested by the baseline model.
Since the inferred ex ante RIR (panel B) is almost
identical across models, it is not surprising that the
short-term RIR gap (panel C) and long-term RIR gap
(panel D) are more volatile in the larger-scale model as
well. Notice also that the estimates from our baseline
model and larger-scale model co-move more closely in
the case of the long-term gap, for which the two lines
essentially overlap during the latest part of the sample.
Regarding the 2002-06 period, the discrepancy
between the short-term and long-term RIR gaps is far
less evident in the larger-scale model than in our
baseline model. However, both measures in the largerscale model remain positive or very close to zero.
This confirms our earlier observation that policy may
not have been as accommodative during this period
as has been suggested (for example, Taylor, 2007).
Consistent with the baseline model, the larger-scale
framework also predicts large positive short-term and
long-term RIR gaps for the fourth quarter of 2008.
However, the same caveats we raised earlier about in­
terpreting these endpoint estimates as reflecting the
policy stance apply to the larger-scale model as well.

1Q/2010, Economic Perspectives

FIGURE 6

Real interest rate levels and gaps in the baseline model and larger-scale model, 1962-2008
B. Ex ante real interest rate
percent

A. Equilibrium real interest rate
percent

12 h

1965

70

75

’80

’85

'90

’95 2000

’05

D. Long-term real interest rate gap
percent

C. Short-term real interest rate gap
percent

Baseline model
Larger-scale model
Sources: Authors’ calculations based on data from Haver Analytics and the U.S. Bureau of Labor Statistics.

Overall, despite some obvious discrepancies, we
view these results as an important assessment of robust­
ness of our main findings. Furthermore, they suggest—
in line with our earlier hypothesis—that model
uncertainty is likely to be a crucial factor surrounding
the measurement of the unobservable equilibrium RIR
and related components. This source of uncertainty is
sometimes ignored in studies presenting model-based
estimates of the RIR, although our findings suggest
that this should be a major issue for further empirical
work in this area.

Federal Reserve Bank of Chicago

Conclusion
In this article, we study the evolution of the equi­
librium RIR and RIR gaps, using both a prototypical
new Keynesian model and a larger-scale model simi­
lar to those in Christiano, Eichenbaum, and Evans (2005)
and Smets and Wouters (2007). Our estimates point to
a substantial degree of time variation in the equilibri­
um RIR. Moreover, we find that this rate has sometimes
become negative in the post-war period. In particular,
our analysis suggests that the equilibrium RIR fell
sharply below zero toward the end of 2008 (although

23

the magnitude of this decline relative to historical
standards is model dependent), resulting in positive
short-term and long-term expected RIR gaps. This pro­
vides some support for the Federal Reserve’s response
to the current crisis with nonconventional measures
to ease monetary policy.
We conclude by noting that the models we use
here, even the larger-scale one, are to some extent

very stylized and have some shortcomings. One of
these shortcomings is the absence of an explicit theo­
retical framework of the financial sector and financial
frictions. It would be useful to analyze how the intro­
duction of these additional features would affect our
results (as, for instance, in Christiano, Motto, and
Rostagno, 2007). These features seem particularly
relevant for the analysis of current economic events.

NOTES
1 Hence, we could alternatively refer to the equilibrium real interest
rate as the real interest rate at potential. We prefer the former ter­
minology because it is more popular in the literature and policy
discussions, as exemplified by the discussion in Ferguson (2004).
Meanwhile, potential output is proportional, but lower than the
efficient level of output. The efficient level of output is the level
of output under perfect competition and, therefore, with zero mark­
ups. In the goods market, the markup is defined as the amount by
which prices exceed the marginal cost of production. In the labor
market, the markup is defined as the excess of wages over the mar­
ginal cost of supplying labor.

Exogenous variations in desired markups, usually referred to as
markup shocks, introduce such a trade-off (see, for example,
Clarida, Gall, and Gertler, 1999).
Potential output is not directly observable either, and the policy
implications of its measurement have received substantial attention
following the work of Orphanides (2001). See also Justiniano and
Primiceri (2008).

4We also estimate the model’s unknown parameters and subse­
quently extract all unobserved model-based variables, such as
expected inflation next period.

24

5While seemingly daunting to compute, the long-run rates can
be backed out from the Lagrange multiplier of the household’s
budget constraint.
6If wages are rigid, optimal monetary policy must attribute some
weight to wage inflation stabilization as well.

7All data except for hours are from Haver Analytics. We are very
grateful to Shawn Sprague, of the U.S. Bureau of Labor Statistics,
for providing us the series of hours in all sectors of the economy.
8We use the eight years prior to the sample period to initialize the
Kalman filter.
9This result is consistent with the large degree of time variation
reported by Laubach and Williams (2003) and Edge, Kiley, and
Laforte (2008), but stands in contrast to the analysis of Neiss and
Nelson (2003), who argue that the equilibrium real interest rate
exhibits very little volatility.
10The main reason the equilibrium RIR in the larger-scale model
is more volatile is that this model includes habit formation.

1Q/2010, Economic Perspectives

APPENDIX: MODEL EQUATIONS
We present the main equations of the model for each
of the five classes of agents described in the main text.

Households
The expected discounted stream of utility that
each household j maximizes is given by

oo

A1)

IpX

Competitive employment agencies operate in
competitive markets and bundle each household’s
specialized labor Lt (/) into a homogenous labor
input according to

L,+sU)l+v

(p-

log

1+v

5=0

where C denotes consumption, and the second argu­
ment of the utility function represents the marginal dis­
utility of each household’s specific labor, /,(/). that de­
pends on the parameter v, known as the inverse Frisch
elasticity of labor supply. Future utility is discounted
at the rate 3, and bt is a “discount factor” shock af­
fecting both the marginal utility of consumption and
the marginal disutility of labor. The logarithm of bt
is modeled as a Gaussian autoregressive process of
order 1, denoted as AR(1) for short.
At every point in time /, each household’s sources
and uses of income must be equal, as summarized by
the budget constraint

PC + Tt + Bt <

Employment agencies

+ CM(/) + | [, + fF(/)£,(/),

where /’ is lump-sum taxes and transfers, Bt denotes
holdings of government bonds, Rt is the gross nominal
interest rate, Qf(j) is the net cash flow from participating
in state-contingent securities that insure against idio­
syncratic risk, and n, is the per capita profit that house­
holds get from owning the intermediate goods firms.
Following Erceg, Henderson, and Levin (2000),
we permit in every period only a fraction 1 - of house­
holds to reset their wages to minimize the expected dis­
counted stream of labor disutility for the periods in
which the posted wage is anticipated to remain in place,
OO

. A+,0)1
-tp-

5=0

1+v

This is subject to the labor demand function of em­
ployment agencies specified next. Wages for the re­
maining fraction of households are indexed to
steady-state inflation and productivity.

Federal Reserve Bank of Chicago

Homogeneous labor is sold to intermediate goods
firms. Profit maximization and the zero profit condi­
tion imply a specialized labor demand function,
1 +Aw, t
^■w.t

A,

where W (yj is the wage paid by the employment
agencies to the household supplying labor of typey,
and W is the hourly wage paid by intermediate goods
firms for their homogenous labor input. The demand
schedule for laborj is decreasing in the relative wage
and depends on the elasticity of substitution among
varieties of labor given by A (. Notice that this elas­
ticity is time varying, and we assume that log (1+Aw()
is a Gaussian independent and identically distributed
(i.i.d.) process. In the literature this is referred to as
the wage markup shock, and it is analyzed in detail in
Justiniano and Primiceri (2008).

Intermediate goods producers
A monopolistically competitive firm produces
the intermediate good Yff) with the production function

F(z) = +,£,(/)“.
where £((7) denotes the bundled labor input purchased
from employment agencies for the production of good 7,
and At represents a productivity shock. We model
as nonstationary, with its growth rate following a
Gaussian AR(1) process.
As in Calvo (1983), at each point in time a fraction
of firms cannot reoptimize their prices and index them
to steady-state inflation. The remaining fraction 1 - £,
of firms post a new price P, (i) to maximize the ex­
pected discounted stream of profits for the periods in
which the new price is anticipated to remain in place,

25

E,t

^3’A,+I

Wl+sLl+s(i)},

where A+s is the marginal utility of consumption used
to value future income, subject to the goods demand
function specified in the next section.

Final goods producers

substitution Ap (among varieties of intermediate goods.
This elasticity is time varying, and we assume that
log (1+ A. ) is a Gaussian i.i.d. process. This
disturbance is known as the price markup shock.

Monetary authority
The Taylor type rule for the short-term nominal
interest rate, Rt, is given by

Perfectly competitive firms produce the final good
y by bundling all intermediate goods according to

y=

(' WiV—
r+A!__
r.yz

Profit maximization and zero profit condition for the
final goods producers imply the following demand
function for the intermediate good z:

Y,U) =

where Pt corresponds to the aggregate price level. The
demand schedule for intermediate good z is decreasing
in its relative price, and depends on the elasticity of

26

1/4’

R,
R

n/,-s
s=0
*

'K,

with R being the steady state for the gross nominal
interest rate and eRi being a Gaussian i.i.d. monetary
policy shock. The parameters and
capture how
aggressively the monetary authority responds to vari­
ations in inflation and output growth over the current
and previous three quarters. There is a time-varying
inflation target tt* , which evolves exogenously according
to a Gaussian AR(1) process. Finally, notice that short­
term nominal interest rates are adjusted gradually, as
given by p5, referred to as the smoothing coefficient.

1Q/2010, Economic Perspectives

REFERENCES

Amisano, G., and O. Tristani, 2008, “Perceived pro­
ductivity and the natural rate of interest,” European
Central Bank, mimeo.

Gall, J., 2008, Monetary’ Policy, Inflation, and the
Business Cycle: An Introduction to the New Keynesian
Framework, Princeton, NJ: Princeton University Press.

An, S., and F. Schorfheide, 2007, “Bayesian analysis
of DSGE models,” Econometric Reviews, Vol. 26,
Nos. 2-4, March, pp. 113-172.

Gali, J., and M. Gertler, 2007, “Macroeconomic
modeling for monetary policy evaluation,” Journal of
Economic Perspectives, Vol. 21, No. 4, Fall, pp. 25—46.

Calvo, G., 1983, “Staggered prices in a utility­
maximizing framework,” Journal ofMonetary’
Economics, Vol. 12, No. 3, September, pp. 383-398.

Justiniano, A., and G. E. Primiceri, 2008, “Potential
and natural output,” Northwestern University, mimeo.

Christiano, L. J., M. Eichenbaum, and C. L. Evans,
2005, “Nominal rigidities and the dynamic effects of
a shock to monetary policy,” Journal ofPolitical
Economy, No\. 113, No. 1, February, pp. 1-45.
Christiano, L. J., R. Motto, and M. Rostagno, 2007,
“Financial factors in business cycles,” Northwestern
University, mimeo.

Clarida, R., J. Gali, and M. Gertler, 2000, “Mone­
tary policy rules and macroeconomic stability: Evidence
and some theory,” Quarterly Journal ofEconomics,
Vol. 115, No. 1, February, pp. 147-180.

___________ , 1999, “The science of monetary policy:
A new Keynesian perspective,” Journal ofEconomic
Literature, Vol. 37, No. 4, December, pp. 1661-1707.

Edge, R. M., M. T. Kiley, and J.-P. Laforte, 2008,
“Natural rate measures in an estimated DSGE model
of the U.S. economy,” Journal ofEconomic Dynamics
and Control, Vol. 32, No. 8, August, pp. 2512-2535.
Erceg, C. J., D. W. Henderson, and A. T. Levin, 2000,
“Optimal monetary policy with staggered wage and
price contracts,” Journal ofMonetary’ Economics,
Vol. 46, No. 2, October, pp. 281-313.

Ferguson, R., 2004, “Equilibrium real interest rate:
Theory and application,” speech to the University of
Connecticut School of Business Graduate Learning
Center and the SS&C Technologies Financial
Accelerator, Hartford, CT, October 29, available at
www.federalreserve.gov/boarddocs/speeches/2004/
20041029/default.htm.

Kozicki, S., and T. Clark, 2005, “Estimating equilib­
rium real interest rates in real time,” North American
Journal ofEconomics and Finance, Vol. 16, No. 3,
December, pp. 395-413.

Laubach, T., and J. C. Williams, 2003, “Measuring
the natural rate of interest,” Review ofEconomics and
Statistics, Vol. 85, No. 4, November, pp. 1063-1070.

Neiss, K. S., and E. Nelson, 2003, “The real interest
rate gap as an inflation indicator,” Macroeconomic
Dynamics, Vol. 7, No. 2, April, pp. 239-262.
Orphanides, A., 2001, “Monetary policy rules based
on real-time data,” American Economic Review, Vol. 91,
No. 4, September, pp. 964-985.

Smets, F., and R. Wouters, 2007, “Shocks and frictions
in U.S. business cycles: A Bayesian DSGE approach,”
American Economic Review, Vol. 97, No. 3, June,
pp. 586-606.
Taylor, J. B., 2007, “Housing and monetary policy,”
m Housing, Housing Finance, and Monetary’ Policy’,
Jackson Hole Economic Symposium Proceedings,
Federal Reserve Bank of Kansas City, pp. 463-476.

___________ , 1993, “Discretion versus policy rules
in practice,” Carnegie-Rochester Conference Series on
Public Policy’, Vol. 39, No. 1, December, pp. 195-214.
Woodford, M., 2003, Interest and Prices: Founda­
tions of a Theory’ ofMonetary’ Policy’, Princeton, NJ:
Princeton University Press.

Francis, N. R., and V. A. Ramey, 2008, “Measures
of hours per capita and their implications for the
technology-hours debate,” University of California,
San Diego, mimeo.

Federal Reserve Bank of Chicago

27

What is behind the rise in long-term unemployment?
Daniel Aaronson, Bhashkar Mazumder, and Shani Schechter

Introduction and summary
As we entered 2010, the average length of an ongoing
spell of unemployment in the United States was more
than 30 weeks—the longest recorded in the post-World
War II era. Remarkably, more than 4 percent of the labor
force (that is, over 40 percent of those unemployed)
were out of work for more than 26 weeks—we consider
these workers to be long-term unemployed. In contrast,
the last time unemployment reached 10 percent in the
United States, in the early 1980s, the share of the labor
force that was long-term unemployed peaked at 2.6 per­
cent. Although there has been a secular rise in long­
term unemployment over the last few decades, the sharp
increases that occurred during 2009 appear to be out­
side of historical norms. Further, this trend may present
important implications for the aggregate economy
and for macroeconomic policy going forward.
The private cost of losing a job can be sizable.
In the short run, lost income is only partly offset by
unemployment insurance (UI), making it difficult for
some households to manage their financial obligations
during spells of unemployment (Gruber, 1997; and
Chetty, 2008). In the long run, permanent earnings
losses can be large, particularly for those workers
who have invested time and resources in acquiring
knowledge and skills that are specific to their old job
or industry (Jacobson, LaLonde, and Sullivan, 1993;
Neal, 1995; Fallick, 1996; and Couch and Placzek,
2010). Health consequences can be severe (Sullivan
and von Wachter, 2009). Research even suggests that
job loss can lead to negative outcomes among the
children of the unemployed (Oreopoulos, Page, and
Stevens, 2008) and to an increase in crime (Fougere,
Kramarz, and Pouget, 2009).
All of these costs are likely exacerbated as un­
employment spells lengthen. The probability of find­
ing a job declines as the length of unemployment
increases. Although there is some debate as to exactly

28

what this association reflects, it is certainly plausible
that when individuals are out of work longer, their
labor market prospects are diminished through lost
job skills, depleted job networks, or stigma associated
with a long spell of unemployment (Blanchard and
Diamond, 1994).1 For risk-averse households that can­
not insure completely against a fall in consumption as
they deplete their precautionary savings, the welfare
consequences of job loss rise as unemployment dura­
tion increases. Welfare implications are particularly
severe during periods of high unemployment for indi­
viduals with little wealth (Krusell et al., 2008).
In this article, we analyze the factors behind the
recent unprecedented rise in long-term unemployment
and explain what this rise might imply for the economy
going forward. Using individual-level data from the
U.S. Bureau of Labor Statistics’ Current Population
Survey (CPS), we show that all of the substantial rise
in the average duration of unemployment between the
mid-1980s and mid-2000s can be explained by demo­
graphic changes in the labor force, namely, the aging of
the population and the increased labor force attachment
of women (Abraham and Shinier, 2002). But only onehalf of the increase in average duration of unemploy­
ment at the end of 2009 relative to that of the early 1980s
may be due to demographic factors. This suggests
that other factors have come into play more recently.

Daniel Aaronson is a vice president and economic advisor
in the Economic Research Department at the Federal Reserve
Bank of Chicago. Bhashkar Mazumder is a senior economist
in the Economic Research Department and the executive
director of the Chicago Census Research Data Center at
the Federal Reserve Bank of Chicago. Shani Schechter is
an associate economist in the Economic Research Department
at the Federal Reserve Bank of Chicago. The authors thank
Lisa Barrow, Bruce Meyer, and Dan Sullivan for helpful
comments and Constantine Yannelis for excellent research
assistance.

2Q/2010, Economic Perspectives

In particular, we attribute the sharp increase in unem­
ployment duration in 2009 to especially weak labor
demand, as reflected in a low rate of transition out of
unemployment into employment, and a smaller portion
of this increase (perhaps 10 percent to 25 percent) to
extensions in unemployment insurance benefits.2
We show that, in any given month, individuals
with longer unemployment spells are less likely to be
employed the following month. This suggests that the
average ongoing spell of unemployment is likely to
remain longer than usual well into the economic re­
covery and expansion, plausibly keeping the unemploy­
ment rate above levels observed in past recoveries.
For example, we find that if the current distribution
of unemployment duration resembled historical dis­
tributions, the unemployment rate would be roughly
0.4 percentage points lower than it is today. Neverthe­
less, we find no evidence that high levels of long-term
unemployment will have a sizable impact on compen­
sation growth going forward.
We begin by presenting some descriptive facts
about trends and business cycle movements of unem­
ployment duration. We then analyze how much of the
increase in the recent average duration of unemploy­
ment compared with that of the previous severe reces­
sion and its aftermath (in 1982-83) can be explained by
changes in the demographic, industrial, and occupa­
tional composition of the labor force versus changes
in the average duration of unemployment within the
various groups. We next consider how much of the
remaining increase can be attributed to weak labor
demand and extensions of unemployment benefits.
Finally, we examine how high levels of long-term
unemployment may affect the unemployment rate
and compensation growth going forward.

The rise of long-term unemployment
We begin by reviewing some facts about unemploy­
ment spell length. Long-run estimates of unemploy­
ment duration are available back to the late 1940s from
the Current Population Survey, a monthly survey of
60,000 or more households. Respondents are 16 years
and older and are asked to classify themselves as em­
ployed, unemployed, or out of the labor force. Those
unemployed are further asked how long, in weeks, their
unemployment has lasted. As a result, the CPS duration
measures are based on ongoing spells of unemploy­
ment and are not measures of completed spell length.
Figure 1 plots the average (and median) duration
of unemployment from 1948 (and 1967) through the
end of 2009. Over the past half century, the average
length of spells of unemployment have increased, from
11.3 weeks in the 1960s to 11.8 weeks in the 1970s,

Federal Reserve Bank of Chicago

11.9 weeks in the 1980s, 15.0 weeks in the 1990s, and
17.4 weeks in the 2000s.3 Figure 2 plots the share of the
unemployed that are short-term (fewer than five weeks)
versus long-term (more than 26 weeks). There has been a
pronounced shift over time in the composition of the
unemployed by duration, with a particularly sharp
change in 2009. Long-term unemployment accounted
for 10 percent of the unemployed in the 1950s and
1960s; it reached 26 percent in the early 1980s; and it
averaged roughly 20 percent between 2002 and 2007,
but reached 40 percent as of December 2009.4 By the
end of last year, over 4 percent of the labor force was
long-term unemployed.
The average duration of unemployment is counter­
cyclical—that is, it increases when the overall econo­
my is shrinking, as figure 1 makes clear. Therefore,
figure 3, panel A presents a scatter plot of average dura­
tion of unemployment against the unemployment rate
to provide a simple way of comparing durations con­
ditional on the unemployment rate. Each blue or black
box represents a month. The black line represents the re­
lationship between the unemployment rate and average
duration of unemployment over the period 1948-2007.
Because the line is upward sloping, it illustrates that
worse labor market conditions (higher unemployment
rates) are associated with longer unemployment spells.
In particular, through 2007, an extra 1 percentage point
on the unemployment rate was associated with spells
that lasted 1.2 weeks longer on average.
For the most recent period, we use black boxes to
represent months between December 2007 (the start
of the most recent recession) and December 2009 in
figure 3, panel A. Note that all the black boxes lie near
the top of the cloud of blue boxes, highlighting that the
average unemployment spell tends to be much longer
now for any given unemployment rate. As the economy
weakened and the unemployment rate rose, the length
of unemployment spells increased—and at a pace that
was fairly typical for a recession. This is represented by
the black boxes that lie roughly parallel to the black line.
But, starting in June 2009 (the half dozen or so black
boxes on the right side of panel A), unemployment spells
began to lengthen to unprecedented levels. Much of this
spike in average duration of unemployment is driven
by the unmistakable increase in the share of the unem­
ployed out of work for more than 26 weeks, highlighted
by the black boxes in figure 3, panel B. For instance,
the average length of unemployment during the last
six months of 2009 was over seven weeks longer than
that of the first six months of 1983, when unemploy­
ment had peaked at 10.8 percent.
Looking forward, we should expect to see a his­
torically long average duration of unemployment for

29

FIGURE 1

Average and median duration of unemployment, 1948-2009

Note: The shaded areas indicate official periods of recession as identified by the National Bureau of Economic Research; the dashed
vertical line indicates the most recent business cycle peak.
Source: U.S. Bureau of Labor Statistics, Current Population Survey, from Haver Analytics.

30

2Q/2010, Economic Perspectives

FIGURE 3

Unemployment rate, duration of unemployment,
and long-term share of unemployment, 1948-2009
A. Average duration of unemployment versus unemployment rate
average duration of unemployment in weeks

B. Long-term share of unemployment versus unemployment rate
percentage of the unemployed who are long-term unemployed

0

2

4

6
unemployment rate

■ January 1948-November 2007

8

10

12

■ December 2007-December 2009

Notes: In panel A, the black line represents the relationship between the unemployment rate and average duration of unemployment
over the period January 1948-November 2007. In panel B, the black line represents the relationship between the unemployment rate
and the share of the unemployed who are more than 26 weeks unemployed over the period January 1948-November 2007.
Source: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics, Current Population Survey, from Haver Analytics.

some time, since it is typical for average spell length
to rise well past the business cycle trough. This is appar­
ent in figure 4, which plots the cyclical pattern in the
average duration of unemployment versus the unem­
ployment rate for several selected cycles. In both the
mid-1970s and the early 1980s (blue lines), average

Federal Reserve Bank of Chicago

duration stayed persistently high, even as the unem­
ployment rate began to decline.5 As labor demand
picks up early in a recovery, employers might turn to
unemployed workers with shorter spells first, leaving
the unemployment pool increasingly composed of
those with relatively longer spells. Sequential hiring

31

FIGURE 4

Unemployment rate versus average duration of unemployment for selected business cycles

+

December 2007-December 2009

—July 1981-December 1983

—■- December 1973-November 1976
Source: U.S. Bureau of Labor Statistics, Current Population Survey, from Haver Analytics.

patterns like this may be due in part to a selection
effect: Those who are less employable are the ones who
are likely to remain unemployed longer and are less
likely to be rehired. However, the lower reemployment
probability of the long-term unemployed may also be
due to diminished job skills, weakened social networks,
and the assumption by some employers of poor worker
quality that accompany those with longer spells. De­
clines in job separations, which we discuss in more de­
tail later, may also reduce the number of short spells
of unemployment in the early stages of a recovery.

Unemployment duration versus other labor
market measures
It is important to emphasize that the recent spike
in the duration of unemployment not only is quite large
by historical standards but also stands out relative to the
recent deterioration in many other key labor market indi­
cators, including three key measures used to gauge labor
market slack: the unemployment rate, a broader unem­
ployment rate (the U.S. Bureau of Labor Statistics’ U-6
rate),6 and total payroll employment. That observation
can best be seen from a very simple statistical model
that uses gross domestic product (GDP) growth to
generate out-of-sample forecasts of these labor market

32

measures. This exercise when applied to the unem­
ployment rate is the basis for what is often referred
to as “Okun’s law.”7 We follow Aaronson, Brave and
Schechter (2009) and use two samples to estimate
these relationships: 1) all data from the first quarter
of 1978 through the second quarter of 2007 and 2) data
solely from the recessions during that period.
Figure 5 shows the results for four measures of
the labor market—namely, the unemployment rate, the
U-6 rate, total payroll employment, and the average
duration of unemployment. Each panel of figure 5 con­
tains three colored lines. The blue line represents the
actual data, the black line is the forecast based on the
data from our frill sample, and the gray line is the forecast
based on only recession periods in the frill sample. Note
that the recession sample forecasts use the recessionperiod coefficients to forecast through the end of
2009, even though the recession likely ended earlier.
Across all the measures in figure 5, the forecasts
based on the full sample of data consistently under­
predict the deterioration in labor market conditions. For
example, the unemployment rate forecasted (panel A)
at the end of 2009 lies roughly 2 percentage points
below the actual unemployment rate, a finding noted by
many commentators who worry that Okun’s law no

2Q/2010, Economic Perspectives

FIGURE 5

Out-of-sample forecasts of labor market variables, 2007-09
B. U-6 rate

A. Unemployment rate
percent

2007

2008

2009

2007

2008

C.Total payroll employment

D. Average duration of unemployment

millions of jobs

weeks

2007

2009

2008

-------- Actual

--------

Predicted based on full sample

2007

--------

2008

2009

2009

Predicted based on recession sample

Notes: Each panel is based on a simple statistical model that uses gross domestic product growth to generate out-of-sample forecasts
of the labor market measure. All data from the first quarter of 1978 through the second quarter of 2007 are used to estimate the full
sample forecasts, while data from just the recessions in this period are used to estimate the recession sample forecasts. The U-6 rate
is a broader unemployment rate from the U.S. Bureau of Labor Statistics (see note 6 for details).
Sources: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics, Current Population Survey, basic monthly files;
and Haver Analytics.

longer applies and labor markets are not functioning
as in the past. However, if we use the recession sample
(gray line), this simple activity model does a remarkably
good job at forecasting the cumulative rise in the stan­
dard (panel A) and broader (panel B) unemployment
rates and the fall in total payroll employment (panel C).
That is, labor markets have mostly evolved about as
we would expect given the severity of the recession.
But such a conclusion is not warranted for unem­
ployment duration (panel D of figure 5).8 Forecasts
based on both the full and recession samples fail to
predict by up to over a month the dramatic rise in
that series, starting in the fourth quarter of 2008.
The remainder of this article is therefore focused on
explaining the causes of the strikingly unusual increase
in the length of unemployment spells.

Federal Reserve Bank of Chicago

Who are the long-term unemployed and
how have they changed over time?
Figure 3 (p. 31) highlights the spike in average
unemployment duration and long-term unemployment
in 2009. It also illustrates that unemployment duration
was already historically high going into the recent
recession given the unemployment rate at the time.
Relative to the black regression line that predicts dura­
tion based on the contemporaneous unemployment rate
(figure 3, panel A), the black boxes there suggest that
unemployment spells were already about four to five
weeks higher, on average, than those during 1948-2007.
For that reason, at least part of the explanation for
current lengths of unemployment happened years ago.
Accordingly, table 1 examines the background char­
acteristics of the long-term unemployed, in particular

33

gender, age, marital status, race, education, industry,
and occupational background in 2009, in 1983 (when
unemployment rates last reached 10 percent—and for
the sake of comparison in the aftermath of a similarly
severe recession), and in 2005-07 (before the start of
the recent downturn). We also compare the distributions
of these characteristics to their distributions in the en­
tire labor force in the second set of columns.9 In the
third set of columns, we report the ratio of the share
of the long-term unemployed to the share in the labor
force for each group. A number above 1 would imply
that long-term unemployment was unconditionally
more common in that group than would be expected
given their representation in the labor force.
In the early 1980s, long spells of unemployment
tended to be concentrated among factory and machine
workers, who made up 29 percent of the labor force
but 55 percent of the long-term unemployed, or nearly
twice their representation in the work force (final row
of table 1). Consequently, the long-term unemployed
also tended to be heavily male (first column, first row)
and only one in five long spells were from individuals
with at least some college education (first column,
fifteenth and sixteenth rows).
In 2009, factory and machine workers (and con­
struction and manufacturing workers in general), males,
and those with no college education still represented a
larger share of the long-term unemployed than they did
of the labor force (third column versus sixth column).10
However, the long-term unemployed became sectorally
more diverse.11 For example, in 2009, the long-term
unemployed were more likely to come from professional
and business services and finance, insurance, and real
estate relative to 1983, while the share of manufactur­
ing/factory workers went down. Generally, in 2009,
long-term unemployment was more equally weighted
across industry, occupation, education, gender, and
age groups, and was therefore more representative of
the labor force and the population than it had been two
and a half decades ago.
Many important demographic shifts in the labor
force have occurred concurrently with changes in the
average length of unemployment. This has led several
researchers (for example, Abraham and Shimer, 2002;
Valletta, 2005; and Mukoyama and §ahin, 2009) to
suggest a link between work force trends and unemploy­
ment duration. These links can be caused by differences
in the propensity to be rehired in a timely fashion
after job loss for particular demographic groups. For
example, increases in college experience, as well as
the general skills that education provides, might en­
able workers to be more adaptable and thus find job
matches more quickly (of course, more job-specific

34

or industry-specific skills could potentially slow the
process down).
In table 2, we provide a simple breakdown of
changes in average duration of unemployment, using
an approach called a Blinder/Oaxaca decomposition.12
This decomposition enables us to estimate how much
of the rise in unemployment duration is due to com­
positional changes in the pool of unemployed workers
(for example, age, gender, education, and industrial
composition); how much is due to longer spell lengths
within each group (for example, longer spells among
women or construction workers), holding the compo­
sition constant; and how much is due to interactions
between changes in compositional effects and coeffi­
cients. We calculate these changes over two time periods
roughly 20 to 25 years apart. First, we compare 1985-86
to 2005-06, when the economy was in the midst of
expansions. Second, we examine two periods in our
sample where unemployment was 10 percent or higher—
the first six months of 1983 and the last six months of
2009 (that is, 1983:Q1-Q2 and 2009:Q3-Q4).
We find that most changes in the composition of
the work force account for little of the increase in aver­
age duration of unemployment.13 The notable exception
is the age structure of the population. Younger workers
in the midst of a long unemployment spell tend to have
shorter spells of unemployment than older workers in
the same situation (Abraham and Shimer, 2002). There­
fore, as the labor force has become older, average spells
have tended to become longer. In table 2 (first column,
second row), we show that changes in age can account
for 0.7 weeks of the 1.3 increase in weeks from the mid1980s to the mid-2000s, or about 53 percent. Yet, the
changing age composition only accounts for about
25 percent of the rise in duration across the two periods
of high unemployment (second column, second row).
This suggests that as the baby boom generation con­
tinues to transition out of the labor force over the next
decade, we should expect the average duration of un­
employment to slowly fall.
The results of the decomposition also suggest that
rising length of unemployment among women (holding
the share of women in the labor force fixed) can account
for virtually the entire increase in the average duration
from the mid-1980s to the mid-2000s (table 2, first col­
umn, ninth row). This corresponds to the greater labor
force attachment of women in recent decades and con­
firms Abraham and Shimer (2002), whose findings have
a similar pattern. Change in unemployment duration
within industries (first column, twelfth row) can also
account for some of the secular pattern across expansions.
However, both the female and industry effects
can explain a notably smaller share of the total

2Q/2010, Economic Perspectives

Federa l Res erv e Bank of Chicag o

TABLE 1

Descriptive statistics: Long-term unemployed and labor force

Long-term unemployed

Ratio of long-term
unemployed share to
labor force share

Labor force

1983

2005-07

2009

1983

2005-07

2009

1983

2005-07

2009

Gender

Male
Female

67.6
32.4

55.5
44.5

58.6
41.4

56.0
44.0

52.4
47.6

52.2
47.8

1.21
0.73

1.06
0.93

1.12
0.87

Age

16-24
25-54
55-64
65 and over

23.6
65.5
9.8
1.1

21.0
62.1
13.5
3.5

18.3
63.4
14.3
4.0

21.1
64.9
11.1
3.0

14.1
67.7
14.1
4.1

13.1
66.4
15.7
4.8

1.12
1.01
0.88
0.38

1.49
0.92
0.95
0.84

1.40
0.95
0.91
0.85

Marital
status

Not married
Married

50.9
49.1

65.7
34.3

62.4
37.6

39.0
61.0

43.8
56.2

44.1
55.9

1.30
0.81

1.50
0.61

1.42
0.67

Race

White
Black
Hispanic
Other

70.6
20.6
3.1
5.6

56.9
23.0
8.5
11.6

59.8
18.4
6.9
14.9

83.1
8.8
3.1
5.0

73.9
8.6
6.5
11.0

72.9
8.8
6.7
11.6

0.85
2.34
0.99
1.14

0.77
2.68
1.30
1.06

0.82
2.10
1.02
1.29

Education

Less than high school
High school graduate
Some college
College graduate

32.8
46.3
13.4
76

23.2
35.6
24.5
16.7

18.9
37.1
27.4
16.5

29.3
379
17.0
15.9

17.2
30.9
26.6
25.3

15.9
30.3
27.2
26.7

1.12
1.22
0.79
0.48

1.35
1.15
0.92
0.66

1.19
1.23
1.01
0.62

Industry

Agriculture, fishing, forestry, and mining
Utilities and sanitation
Construction
Manufacturing
Wholesale trade
Retail trade
Transportation and warehousing
Finance, insurance, and real estate and leasing/rental
Professional/business services and information
Health, education, and social services
Other services
Government

5.1
0.8
11.4
34.6
3.2
9.7
4.5
2.6
5.2
8.1
11.1
3.0

1.7
0.6
10.2
13.8
2.6
13.8
3.8
4.7
15.8
13.4
16.3
3.0

2.0
0.7
14.1
15.7
2.3
12.2
3.9
5.8
15.3
11.4
14.4
1.8

5.2
1.4
6.7
19.1
4.1
11.8
4.0
6.0
7.5
16.5
12.9
4.8

2.6
1.1
8.1
10.9
3.0
11.7
4.2
7.0
12.2
21.4
13.0
4.8

2.6
1.2
75
10.2
2.6
11.4
4.1
6.7
12.6
22.7
13.5
4.9

0.99
0.62
1.70
1.81
0.78
0.82
1.13
0.44
0.69
0.49
0.86
0.63

0.66
0.55
1.25
1.26
0.88
1.18
0.90
0.67
1.29
0.62
1.26
0.62

0.75
0.58
1.89
1.53
0.90
1.07
0.94
0.87
1.22
0.50
1.07
0.35

4.8
5.3
6.3
10.7
14.6
2.6
55.0

8.1
11.3
11.7
14.1
21.5
1.2
31.6

10.1
10.5
11.5
13.6
18.0
1.0
35.0

10.2
14.8
11.3
15.6
14.4
4.4
29.1

14.7
20.3
11.4
13.6
16.6
0.8
22.6

15.3
21.3
11.0
13.0
17.4
0.8
21.1

0.47
0.35
0.56
0.68
1.02
0.60
1.89

0.55
0.56
1.03
1.04
1.29
1.57
1.40

0.66
0.49
1.04
1.05
1.03
1.21
1.66

Occupation Executive and managerial
Professional and technical
Sales
Administrative support
Services
Farming, forestry, and fishing
Factory/machine workers

Notes: All values are in percent. Some columns may not total because of rounding.
Source: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics, Current Population Survey, basic monthly files.

change in spell length when we compare
the changes across the two periods of high
unemployment in the second column of
table 2. For example, changing coeffi­
cients for women can only account for
about 35 percent of the rise in the average
duration of unemployment across the two
periods of high unemployment (second
column, ninth row). Industry effects al­
most completely disappear (second col­
umn, twelfth row). Just under one-half
(2.8 weeks out of 6.2 weeks) of the in­
crease in duration from the first half of
1983 to the second half of 2009 is ex­
plained by direct shifts in composition
and coefficients (second column, seventh
and thirteenth rows).
Overall, the decomposition suggests
that although demographic factors can
account for much of the secular increase
in unemployment duration, they can only
account for a portion of the especially
sharp rise in durations that has accompa­
nied this most recent recession. This sug­
gests that other factors must be driving
this phenomenon—the topic that we him
to next.

TABLE 2

Decomposition of the secular change in the average duration
of unemployment, 1980s to 2000s

Total change to explain

1983:Q1-Q2
to
2009:Q3-Q4

1.3

6.2

Due to changes in composition
Age
Gender
Race
Education
Industry
Total

0.7
-0.1
0.0
-0.1
0.0
0.5

1.6
0.0
0.1
-0.5
0.0
1.2

Due to changes in coefficients
Age
Gender
Race
Education
Industry
Total

0.0
1.6
-1.0
-0.3
1.3
1.6

-0.3
2.1
0.0
-0.3
0.1
1.6

Interactions between changes
in composition and coefficients

-0.8

3.3

Notes: All values are in weeks. See note 12 for further details. The second column
does not total because of rounding.
Source: Authors’ calculations based on data from the U.S. Bureau of Labor
Statistics, Current Population Survey, basic monthly files.

Labor market transitions, the unemployment
rate, and unemployment duration
In order to better understand the causes of the re­
cent sharp rise in long-term unemployment, it is useful
to develop a framework for studying labor market dy­
namics during the business cycle. In this section, we
begin formulating this framework by showing how
movements between being employed, unemployed, and
out of the labor force (labor market transitions) have
contributed to cyclical patterns in unemployment his­
torically and during the most recent recession. We then
generate a model that uses labor market transitions to
create counterfactual scenarios that would correspond
to alternative views of what may be driving labor mar­
kets. Finally, we use this apparatus to provide some in­
sight into the causes of long-term unemployment. We
also use these results later to analyze the implications
of long-term unemployment for the aggregate economy
going forward.
To measure labor market transitions, we exploit
the fact that the CPS interviews whatever household
unit is living at a particular address for four consecu­
tive months, skips the address for eight months, and
then returns for more interviews over four consecutive
months. This allows us to track many household units

36

1985-86
to
2005-06

over time. We follow previous studies that have used
matching algorithms to identify individuals who are
living at the same address in consecutive months and
build a panel data set containing the labor market status
of individuals at multiple points in time. Specifically,
we consider the nine possible transition probabilities
(transition rates) across the states of employment (£),
unemployment (£/), and out of the labor force (O).

Transition rates and the unemployment rate
Figure 6 plots these nine seasonally adjusted
monthly transition rates (blue lines), along with sixmonth moving averages of each to smooth out some
of the noise in the data (black lines). Two key transi­
tions for explaining past changes in the unemployment
rate are movements from employment to unemploy­
ment (EU) and unemployment to employment (f //:’).14
The EU transition rate measures the fraction of em­
ployed individuals who separate from their employer
and move into unemployment. We will hereafter refer
to this as the “separation rate.”15 The UE rate is some­
times referred to as the hiring rate. Shimer (2007) has
argued that most of the variation in the unemployment
rate is due to fluctuations in the hiring rate rather than
the separation rate, although this conclusion has been dis­
puted by Elsby, Michaels, and Solon (2009), who argue
that both rates have been of significant importance.16

2Q/2010, Economic Perspectives

Movements in the Tit/have been particularly pro­
nounced in this recession: The separation rate has risen
by nearly 70 percent. This disproportionate hike is shown
more clearly in figure 7 (p. 40), where we compare the
proportional change in the EU and UE transition rates
in the current business cycle with the recessionary
periods in 1981-82 and 2001. The EU transition rate
followed its historical pattern during 2008 but then
began rising sharply early in 2009. Relative to the
acceleration in the EU rate, the UE rate appears to
have fallen more gradually, though proportionately
more than in previous recessions.
To assess how important the transitions out of
employment versus transitions out of unemployment
have been in explaining the rise in the unemployment
rate during the most recent downturn, we perform some
simple simulations. We start with the actual levels of
those who are employed, unemployed, and out of the
labor force and the smoothed values of all nine of the
labor market transition rates at the end of 2007. We
then use the actual transition rates starting in January
2008 to simulate the new counts of individuals in each
labor market state for each month going forward. This
is described in greater detail in box 1 (p. 41). With
some basic adjustments, we are able to match the
actual monthly unemployment rates through the end
of 2009 almost perfectly.
We then conduct the following two experiments.
First, we hold all transition rates constant at their
December 2007 values except for the three transitions
that start with the employment state in the initial month
(EE, EU, and £0).17 Those transition rates are allowed
to vary according to what actually transpired in 2008
and 2009. In essence, this exercise, which is plotted
as the dark blue line on figure 8 (p. 42), captures the
effects of transitions out of employment into non­
employment (being either unemployed or out of the
labor force) on the aggregate unemployment rate.18
Analogously, we do a second experiment where only
the transitions from the state of unemployment (UE,
UU, and UO) are allowed to change. This captures
the effects of the fall in the exit rate out of unemploy­
ment into being either employed or out of the labor
force. Those results are shown as the light blue line
in figure 8. The black line is the actual unemployment
rate, and the gray one is the actual unemployment
rate in December 2007.
We find that the changes in the transition rates
out of employment (all else being equal) would only
raise the unemployment rate by 1 percentage point by
the end of 2009. In contrast, changes in the transition
rates out of unemployment would raise the unemploy­
ment rate by 2.2 percentage points. Broadly speaking,

Federal Reserve Bank of Chicago

this suggests that the combined effects of moving out
of unemployment (UE, UU, and UO)—including,
prominently, the transition into a job—explain more
of the actual increase in the unemployment rate over
the past two years than the combined effects of mov­
ing out of employment (EE, EU, and EO).19
Transition rates and unemployment duration
We next turn to using these exercises to explain
unemployment duration. The simulation is similar as
before except that we now explicitly incorporate the
distribution of unemployment duration into the analysis
by using five-week “bins” of unemployment spells
(that is, 0-4 weeks, 5-9 weeks, and so on). We start
with the distribution of unemployment duration at the
end of 2007 and use estimates of the actual transition
rates into and out of unemployment for each bin, along
with estimates for the other transition rates, to update
the distribution of duration each month. We again find
that the simulation does extremely well at replicating
the sharp rise in the average duration of unemployment
during 2009.20
In figure 9 (p. 42), we show that if only the EE,
EU, and /TO followed their actual paths and all the other
transition rates stayed constant at their December 2007
values, the average duration of unemployment would
have only increased slightly, to about 19 weeks by the
end of 2009 (dark blue line). If, however, only UE, UU,
and UO followed their actual paths and the other tran­
sition rates stayed flat, unemployment duration would
have increased to nearly 23 weeks (light blue line). So
it appears that for both the unemployment rate and the
average duration of unemployment, transition rates from
the starting state of unemployment have been the im­
portant driving influences.21

Simulated effects of federal unemployment
insurance benefit extensions
As noted previously, the spike in the average dura­
tion of unemployment starting in mid-2009 is hard to
explain using demographics or the standard association
with deteriorating GDP growth. One plausible explana­
tion is the unprecedented extension of unemployment
insurance benefits. The maximum number of weeks
of eligibility rose from 26 weeks to 39 weeks in July
2008 with the passage and creation of the Emergency
Unemployment Compensation (EUC) federal program.
Since then, extensions have risen at varying rates, de­
pending on the unemployment situation of individual
states.22 Figure 10 (p. 43) plots the weighted national
average of the maximum number of weeks of unem­
ployment benefit receipt allowed (blue line); the
weights for this average are based on the size of the
unemployment pool in each state. As of January 2010,

37

FIGURE 6

Labor market transition rates, 1976-2009
A. Employment to employment
transition rate

D. Employment to unemployment
transition rate

B. Unemployment to employment
transition rate

E. Unemployment to unemployment
transition rate

C. Out of the labor force to employment
transition rate

F. Out of the labor force to unemployment
transition rate

2Q /2 010, Econom ic Per spe ctiv es

_

FIGURE 6 ( continued)

©\
o
o
rq
i
r©\

u.
©
42
©

Federal Reserve Bank of Chicago

unemployed workers in 14 states were allowed the
maximum of 99 weeks of UI benefits and the na­
tional average was 90 weeks. By contrast, in 1983,
the maximum potential duration of UI coverage in
any state had reached 55 weeks.
In order to estimate the possible effect of UI
benefit extensions on unemployment duration, we
use previous studies of the effect of an additional
week of maximum benefits on average duration.
A prominent example in this literature is Katz and
Meyer (1990), who use a rich statistical model and
administrative data from the UI system to estimate
the probability of leaving unemployment during
the early 1980s recessions. They identify the impact
of UI through variation in maximum benefits both
within and across states that shift as a result of eli­
gibility rules and legislative changes. They find that
the average duration of unemployment rises by
0.16 weeks to 0.2 weeks for each additional week
of benefits extended.
Katz and Meyer (1990) face the difficult prob­
lem of disentangling the effects of UI benefit exten­
sions from the effects of poor economic conditions
that typically prompt benefit extensions in the first
place. When the economy is in a recession, longer
spells of unemployment are expected irrespective
of the generosity of the unemployment insurance
program. To get around this problem, Card and
Levine (2000) use an increase in the maximum
number of weeks of benefit eligibility in New Jersey
in 1996; this increase was unrelated to the state of
the economy at the time. In fact, this particular ex­
tension, which was driven by political considerations,
took place in the midst of an expansion and there­
fore might be less susceptible to the bias of reces­
sion-driven extensions. Indeed, they find a smaller
effect than Katz and Meyer (1990) and much of
the rest of the literature; mean duration rises by
about 0.1 weeks for each additional week of bene­
fits. In order to reflect our uncertainty over the true
effect, we use both estimates.
We begin our analysis in June 2008, when maxi­
mum UI eligibility was 26 weeks and unemploy­
ment spells lasted about 17 weeks on average (the
six-month mean from January through June 2008).
We then calculate an estimated effect of the exten­
sion in unemployment benefits for each subsequent
month beginning with July 2008.23 For such a calcu­
lation, two additional inputs are required. First, we
need the share of the unemployed who are actually
receiving benefits because they are the only ones
who would be directly affected by policy changes.
The black line in figure 10 (p. 43) shows that the

39

FIGURE 7

Employment-to-unemployment (EU) and unemployment-to-employment (UE) transitions
over selected business cycles

------

EU (current)

UE (current)

------EU(2001)

UE (2001)

........ EL/(1981)

UE (1981)

Notes: The most recent business cycle peak was in December 2007, according to the National Bureau of Economic Research; July 1981
and March 2001 were the business cycle peak months corresponding with the recessions beginning in 1981 and 2001, respectively.
These three dates correspond with month 0 in this figure.
Source: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics, Current Population Survey, basic monthly files.

share of the unemployed receiving UI surged from
41 percent in July 2008 to 67 percent in December 2009.
Second, we must assume a time period over which to
distribute the full effect of the extension in benefits.
We distribute the effects of the initial 13-week exten­
sion of benefits that took place in July 2008 over a hill
year.24 The additional extensions that increased the
maximum potential duration in the UI system beyond
52 weeks, beginning in December 2008, are spread
over two years because the much larger extensions are
likely to alter behavior over a longer period of time.
Using these inputs, our estimates based on the
Katz and Meyer (1990) elasticity suggest that the exten­
sion of UI benefits during 2008 and 2009 may account
for as much as 3.1 weeks of the 12-week increase in
the average duration of unemployment that took place
over this period. The estimate based on the Card and
Levine (2000) analysis suggests that it could explain
about 1.6 weeks. These assessments, which we con­
sider our range of preferred estimates, suggest that
the effect of unemployment insurance extensions on
the average duration of unemployment is on the order
of 10-25 percent of the total increase since July 2008.

40

Alternatively, if we spread the effect of the extensions
beginning in December 2008 over just one year, this
would raise our estimates of the contributions to be­
tween 3 weeks and 6 weeks, or between 25 percent
and 50 percent. It is also important to note that our
calculations have not considered the potential effect
of the “reach back” provision in the EUC program
that allowed extensions for those who had exhausted
their unemployment benefits as early as May 1, 2007.
It is possible that this provision could further raise
our estimates of the impact of UI benefits.
Effects of unemployment insurance benefit
extensions on the unemployment rate
We can also utilize the transition data to examine
whether movements from unemployment to out of the
labor force (UO) and those in the opposite direction
(OU) yield additional clues about possible changes
in classification between the non-employed that may
have arisen as a result of UI benefit extensions. In
figure 11, panel A (p. 44), we plot all the possible
transitions from unemployment during the current
business cycle (blue lines) and compare them with

2Q/2010, Economic Perspectives

BOX 1

Methodology for simulating paths of the unemployment rate and unemployment durations
In this article, we use transitions across different
labor market states as a tool for simulating the paths
of the unemployment rate and unemployment dura­
tions. This allows us to consider alternative scenarios
for the path of the unemployment rate or durations
either historically or going forward based on changing
the paths of particular transitions. Using this approach,
we can infer the relative importance of particular
economic phenomena that are related to certain tran­
sitions as described in the text. An important caveat
is that this is a mechanical approach that may or may
not correspond well to changes in the actual economy.
For example, conditions that may change a particular
transition rate may also affect other transition rates
in ways that we may not consider.
If we consider time as discrete and denote it as
t and let .r stand for a particular labor market state, that
is, employed, unemployed, or out of the labor force,
(.r = {E,U,O}), then each of the nine possible transitions
are defined as the probability of being in a particular
labor market state in t conditional on having been in
the same or different labor market state in the prior
period. For example, the EU transition is defined as:

EU= Prob (x(= U|vf_i = E).

For each initial state in / - 1, the three possible
transitions must sum to 1. For example, EE + EU +
EO = 1. Each of the nine transitions for each month
is estimated empirically using the matched Current
Population Survey as described in the text. To imple­
ment the simulation we start by inputting the levels of
those who are employed, unemployed, and out of the
labor force for a chosen base period. We then simulate
the next period’s level of E, U, and O, using the assumed
levels for the base period and an assumed path for
the transition probabilities. For example, if we wish
to simulate the actual path of the unemployment rate
through 2008 and 2009, we define December 2007
as our base period and then use the actual estimated
values of the transition probabilities for January 2008
through December 2009.1

those during the 2001 recession and its aftermath
(black lines). The series with the boxes represent the
UE transition rates and are identical to those shown in
figure 7 (p. 40). To this we add UU transition rates
(diamonds) and UO (circles) rates. It appears that the
UU and UO rates in the current recession track the rates
in 2001 reasonably well for the first 16 or so months
of the downturn before beginning to diverge. In con­
trast, the UE rates diverge earlier in the cycle. One

Federal Reserve Bank of Chicago

For example:
EJanOS =EDec07 *EEJan08 +UDec07 xEUJanOS +ODec07 *EOJan08

We use several methods to pose alternative tran­
sition rates, depending on our question of interest. To
address the relative importance of transitions from
employment versus transitions from unemployment,
we start with a baseline path where all of the transi­
tion rates are constant. We then change the paths of
all three transition rates from either employment or
unemployment simultaneously. For example, we sim­
ulate the effects arising only from changes from the
employment state by changing the paths of EE, EU,
and EO simultaneously.
A second approach is used when we wish to hold
the UO and OU transition rates fixed at a particular
rate. In this case, we allow the UE and OE rates to fol­
low their actual paths and then adjust the UU and OO
so that the probabilities from U and from O each sum
to 1. Finally, for the simulation that attempts to repro­
duce the forecast of the unemployment rate according
to the Blue Chip Economic Indicators, we assume
that the EU, EE, UU, and UE rates take five years to
return to their historical average values. We then ad­
just the EO and UO rates so that the three transitions
from E and from U sum to 1.
’Rather than immediately going from the base period to the
first period of the simulation, we first use the transition rates
from the base period and run about ten iterations of the model
so that the values of E, U, and O and the implied unemploy­
ment rate reach a steady state, where they are unchanging. We
then proceed to use the steady-state values for the simulation.
The steady-state values may differ from the actual values in the
base period. For example, the steady-state value of the unemploy­
ment rate in December 2007 is about 80 percent of the actual
value. We therefore scale the subsequent values of the simula­
tion by a factor of 1.25. This discrepancy is likely due in part
to the inability to account for month-to-month compositional
changes that arise from the fact that individuals enter or exit
the working age population. Measurement error and differences
between the complete population and the matched sample may
also play a role. This approach assumes that although we can­
not match the level of unemployment, we can match changes
over time.

possible reason for this pattern is that individuals who
would have normally dropped out of the labor force at
this point in the cycle chose to remain unemployed—
perhaps to continue to collect unemployment benefits.
In figure 11, panel B (p. 44), we focus only on the
rate of UO transitions and add data from the 1981-82
recession (dotted). This panel shows that the UO path
during the current recession resembles the UO path dur­
ing the 1981-82 recession reasonably well, suggesting

41

FIGURE 8

Counterfactual effects of changing labor market transition rates on the unemployment rate, 2008-09
percent

Jan
2008

Mar

May

Jul

Sep

Nov

Jan
2009

Mar

May

Sep

Jul

Nov

------

Actual unemployment rate

------ Unemployment rate if only UE, UU, and UO change

------

Unemployment rate if only EE, EU, and EO change

------ Unemployment rate in December 2007

Note: E indicates employment, U indicates unemployment, and O indicates out of the labor force.
Sources: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics, Current Population Survey, basic monthly files;
and Haver Analytics.

FIGURE 9

Counterfactual effects of changing labor market transition rates on the duration of unemployment, 2008-09

2008

2009
Actual average duration of unemployment

Average duration of unemployment
if only EE, EU, and EO change

Average duration of unemployment
if only UE, UU, and UO change

Baseline duration, 17 weeks

Notes: E indicates employment, U indicates unemployment, and O indicates out of the labor force. The average duration of unemployment
in December 2007 was about 17 weeks, so we use this duration as our baseline.
Source: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics, Current Population Survey, basic monthly files.

42

2Q/2010, Economic Perspectives

FIGURE 10

Maximum potential duration of unemployment insurance (UI) benefits and
the share of unemployed receiving benefits, 2008-09

2008

2009
------

Maximum potential duration of UI benefits, weighted national average (left-hand scale)

------

Share of unemployed receiving UI benefits (right-hand scale)

Notes: The weights for the national average duration are based on the size of the unemployment pool in each state. The share of
unemployed receiving UI benefits is the number of individuals with continuing claims under federal and state unemployment insurance
programs divided by the number of unemployed; this fraction is seasonally adjusted.
Sources: Authors’ calculations based on data from the U.S. Department of Labor, the U.S. Bureau of Labor Statistics, and Haver Analytics.

that the departure from the 2001 pattern may simply
reflect the greater severity of the current recession.
That said, figure 11, panel C suggests that the rate of
OU transitions in the current recession appears to
move substantially higher in percentage terms than
the patterns observed during the previous downturns.
We conduct a simulation motivated by figure 11
to ask how different the unemployment rate would be
had the paths of the UO and OU transitions stayed con­
stant at their values 16 months after the start of the
recession (April 2009). In order to ensure that the proba­
bilities from a particular state add up to 1, we allow
the UE and OE rates to follow their actual paths and
adjust the UU and OO rates so that the probabilities
sum to 1. Figure 12 shows that under this counterfactual scenario the result of this exercise would be to
lower the unemployment rate to 9.3 percent as of
December 2009—about 0.7 percentage points below
the actual unemployment rate that month. Although
this is a relatively crude and mechanical approach, it
nonetheless provides a magnitude for the possible
effect of unemployment insurance benefit extensions
on the unemployment rate.

Federal Reserve Bank of Chicago

Implications of long-term unemployment
for the aggregate economy
In this section, we consider how the increase in
long-term unemployment may affect the economy going
forward. We consider the effects of the unemployment
duration structure on the unemployment rate and then
on compensation growth.
Effects of duration structure on the
unemployment rate
As unemployment spells lengthen, the probability
of finding a job in a given time period declines—an
association that is robust across time and demographic
groups. The pattern is illustrated in figure 13 (p. 46),
which plots the probability of being employed today
for various lengths of unemployment duration in the
previous month (horizontal axis). For example, at
0-4 weeks of unemployment, the average probability
of finding a job in the following month is 34 percent,
but at 25-29 weeks, it is only 19 percent.25 As much
as this phenomenon is due to diminished job skills
and weakened social networks, it could have a real
impact on the labor market recovery while the broader
economic recovery takes hold.

43

FIGURE 11

Labor market transitions from non-employment
A.Transitions from unemployment, current recession versus 2001 recession
index, March 2001 and December 2007 = 1.0

-■-

UE (current)

—♦—

(75(2001)

UU (current)

—®—

UO (current)

(7(7 (2001)

-9-

(70(2001)

B. Transitions from unemployment to out of the labor force ((70)
index, July 1981, March 2001, and December 2007 = 1.0

C.Transitions from out of the labor force to unemployment (0(7)
index, July 1981, March 2001, and December 2007 = 1.0

------

OU (current)

-- -

OU (2001)

.......

0(7(1981)

Notes: E indicates employment, (7 indicates unemployment, and O indicates out of the labor force. The most recent business cycle peak
was in December 2007, according to the National Bureau of Economic Research; July 1981 and March 2001 were the business cycle peak
months corresponding with the recessions beginning in 1981 and 2001, respectively. These three dates correspond with month 0 in this figure.
Source: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics, Current Population Survey, basic monthly files.

44

2Q/2010, Economic Perspectives

FIGURE 12

Counterfactual effects of holding unemployment-to-out-of-the-labor-force (UO)
and out-of-the-labor-force-to-unemployment (OU) transition rates fixed from April 2009 onward
on the unemployment rate, 2008-09

------- Actual unemployment rate
------- Simulated unemployment rate, with UO and OU fixed
Sources: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics, Current Population Survey, basic monthly files; and
Haver Analytics.

In order to investigate this possibility, we use our
transition rate model but substitute aggregate transition
rate probabilities for movement from unemployment
(UE, UU, and UO) with analogous transition rate proba­
bilities for each live-week bin of unemployment dura­
tion. We start by simulating a baseline path that roughly
matches the January 10,2010, forecast of the unemploy­
ment rate through 2011 according to the Blue Chip
Economic Indicators (Blue Chip), a survey ofAmerica’s
top business economists (Aspen Publishers, 2010). We
then pose an alternative path where the only change is
to make the share of the unemployed in each live-week
bin at the beginning of the simulation (January 2010)
match their mean historical values. In figure 14, panel A,
we show that this alternative initial distribution of
duration would immediately lower the unemployment
rate by about 0.4 percentage points relative to the Blue
Chip path. We find, however, that duration quickly re­
verts back to high levels (figure 14, panel B) and that
the unemployment rate path converges to what it would
have been had the model started with the actual dis­
tribution of duration. The main lesson we take from
this exercise is that the unemployment rate is probably
about half a point higher than it would be if unemploy­
ment spell lengths were at more historical levels.

Federal Reserve Bank of Chicago

Effects of duration structure on compensation growth
Lastly, we consider the possible effects of higher
long-term unemployment rates on aggregate wage
growth. It is not obvious a priori what the expected
effects should be. If the long-term unemployed are
readily employable and can fulfill vacancies, then there
is a sense in which they may be more eager to return
to work at the prevailing wage than individuals with
short unemployment durations. In this case, the long­
term unemployed may reduce wage pressures. If, how­
ever, many of the long-term unemployed are more
akin to individuals who have stopped searching for
work and have left the labor force, perhaps because
of a geographical or skills mismatch, then they may
play little role in bidding down wages.
Since this is ultimately an empirical question, we
undertake a simple exercise using Phillips curve style
regressions to address this. We use data on year-overyear growth in real compensation per hour. Figure 15
(p. 48) shows that, as expected, there is a negative rela­
tionship between compensation growth and the unem­
ployment rate. The black boxes signify the values starting
with 2008:Ql, when the recession began. We regress
compensation growth on the unemployment rate for the
post-1975 period and calculate the predicted values.

45

FIGURE 13

Probability of reemployment, by unemployment duration

weeks of unemployment
Source: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics, Current Population Survey, basic monthly files.

We then add the share of the unemployed in each of
the five-week bins of unemployment duration to the
regression and reestimate the model. We plot both sets
of the forecasted values, along with the actual growth
rate of real compensation, in figure 16. We find that
there is little difference in magnitude between the two
forecasts. For much of the past 20 years, the predictions
that incorporate unemployment duration are slightly
lower than those that do not. However, there is little
economically important difference in the most recent
period. Overall, this suggests that at least for predicting
aggregate compensation trends, there is no clear-cut
indication that rising unemployment duration will sig­
nify any more or less slack than the information con­
tained in the unemployment rate. This might be because
rising unemployment duration produces countervailing
forces on wage pressures as hypothesized earlier. How­
ever, the statistical model used to estimate these rela­
tionships is based on historical associations, whereas
the current distribution of unemployment spell length
is unprecedented.

Conclusion
The average length of an ongoing spell of unem­
ployment topped 30 weeks in December 2009, with
more than 40 percent of the unemployed out of work for
over six months. These numbers far exceed anything

46

recorded in the post-World War II era. In this article,
we analyze the factors behind this historically unprece­
dented rise in long-term unemployment and explain
what it might imply for the economy going forward.
We show that roughly half of the rise relative to pre­
vious deep modem recessions was due to demograph­
ic factors in place well before the recession began.
The remaining unexplained increase is due primarily
to especially weak labor demand, reflected in low levels
of hiring. Perhaps 10-25 percent of the increase in
long-term unemployment from mid-2008 to the end
of 2009 is associated with extensions of unemployment
insurance benefits. These estimates for the current
business cycle constitute a notable departure from
historical patterns in transitions between being unem­
ployed and out of the labor force. Some simple coun­
terfactual estimates suggest that had these transitions
followed more typical patterns, the unemployment rate
might be about 0.7 percentage points lower. Finally,
we find that high levels of long-term unemployment
typically persist well into an economic recovery, since
firms tend to hire the long-term unemployed last. Some
simple simulations suggest that a historically long
unemployment duration distribution as currently ex­
perienced in the United States could slow the process
of labor market recovery, but it is not expected to
have much of an impact on compensation growth.

2Q/2010, Economic Perspectives

FIGURE 14

Simulated effects of changing the initial distribution of unemployment duration, 2010-11

A. Effects on the January 2010 Blue Chip forecast of the unemployment rate
percent

Jan
2010

Mar

-------

May

Jul

Blue Chip forecast

Nov

Sep

Jan
2011

Mar

May

Jul

Sep

Nov

------- Alternative initial distribution of unemployment duration

B. Effects on the path of the average duration of unemployment
weeks

------- Baseline, based on Blue Chip forecast

------- Alternative initial distribution of unemployment duration

Note: Blue Chip forecast refers to the forecast of the unemployment rate through 2011 according to the Blue Chip Economic Indicators.
Sources: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics, Current Population Survey, basic monthly files; and
Aspen Publishers (2010).

Federal Reserve Bank of Chicago

47

FIGURE 15

Real compensation growth versus the unemployment rate, 1949-2009

■ 1949:Q1-2007:Q4

■ 2008:Q1-2009:Q4

Note: The black line represents the relationship between the unemployment rate and the percent change of compensation from
a year ago over the period 1949-2009.
Source: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics from Haver Analytics.

FIGURE 16

Real compensation growth, actual versus predicted, 1949-2009

------

Actual

------

Predicted based on the unemployment rate only
Predicted based on the unemployment rate and unemployment duration

Source: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics from Haver Analytics.

48

2Q/2010, Economic Perspectives

NOTES
’Alternatively, the relationship between the length of time out of
work and the diminishment of work prospects could be picking up
unobserved differences in worker quality between those who are
unemployed for short and long spells (Ham and Rea, 1987; Kiefer,
1988; and Machin and Manning, 1999). In this case, longer spells
in and of themselves do not lead to worse outcomes. It is very diffi­
cult to convincingly identify which of these channels dominates
without strong statistical assumptions.

9A11 of the inferences here are the same if the base of comparison
is the full population rather than the labor force.

2Based on transition patterns between being employed, unemployed,
and out of the labor force altogether, we estimate that UI extensions
increased the unemployment rate by roughly 0.7 percentage points
during 2008-09.

12To implement the Blinder/Oaxaca decomposition, a separate re­
gression is run for each time period. The change in average duration
of unemployment over the two periods is then decomposed into
a portion due to changes in the levels of the explanatory variables
(for example, the fraction of females and the fraction that has
completed less than high school), a portion due to changes in the
coefficients on these explanatory variables, and a residual term
that captures the effects of the interactions (that is, simultaneously
changing the levels and coefficients).
Specifically, let unemployment duration D be specific to an
individual i and a time period t. To keep things simple, we use two
time periods—the 1980s, which is indexed as t= 1, and the 2000s,
which is indexed as t = 2. We show the results by comparing ex­
pansions (1985-86 versus 2005-06 in the first column of table 2 on
p. 36) and comparing periods of high unemployment (first half of
1983 versus second half of 2009 in table 2, second column). Duration
is determined by characteristics Xjt (for example, gender and age)
that are also specific to individual i and time period t.
We can write this statistical model as D f = X.h + e.t, where e.t
is an error term. The decomposition is then D - D? = (X} -Xfib2
+ Xfb1 - Z?2) + (X} - X2)(b] - bfi. The first term after the equal sign is
reported in the first set of rows in table 2 (“due to changes in com­
position”). The second term is reported in the second set of rows (“due
to changes in coefficients”), and the third term is the row labeled
“interactions between changes in composition and coefficients.”
Running this decomposition on the share of the unemployed
undergoing long-term spells of unemployment yields similar results.
Those are available upon request.

3Long-term unemployment is a good deal less common in the United
States than in much of the developed world (for example, Machin
and Manning, 1999). As of 2008, the last year for which comparable
data are available, the share of the unemployed out of work more
than six months was two times, and in some cases four times, higher
in Belgium, the Czech Republic, France, Germany, Greece, Hungary,
Italy, Luxembourg, the Netherlands, Portugal, Switzerland, the
United Kingdom, and Japan.
4The most recent numbers from the Current Population Survey
are still well below the prevalence of long-term unemployment
during the Great Depression. Unfortunately, national data on un­
employment duration before World War II are not systematically
available. Definitions of unemployment also varied across surveys
and are different from the modem one. That said, Eichengreen and
Hatton (1988) report that more than a third of males who were
looking for work in 1930 had been unemployed for at least 14 weeks
and 55 percent of ongoing unemployment spells had lasted at least
six months in 1940. Eichengreen and Hatton also reproduce data
from Woytinsky (1942), showing the year-to-year changes in un­
employment duration in Philadelphia during the 1930s. In 1933,
for example, over 80 percent of the unemployed had spells of at
least six months. Chatterjee and Corbae (2007) describe a special
January 1931 census of the unemployed in Boston, New York,
Philadelphia, Chicago, and Los Angeles, which reported that
45 percent, 61 percent, 45 percent, 61 percent, and 33 percent
were jobless for at least 18 weeks, respectively.
5As can also be seen in figures 1 and 2 (p. 30), it took particularly
long for average and median unemployment duration and the share
of the long-term unemployed to return to pre-recession levels fol­
lowing the 1990-91 and 2001 recessions.
6The U-6 rate, available since 1994, includes marginally attached
workers and part-time workers who want and are available for full­
time work but had to settle for a part-time schedule for economic
reasons. The U.S. Bureau of Labor Statistics classifies individuals
as “marginally attached” if they “indicate that they want and are
available for a job and have looked for work sometime in the re­
cent past” but are not currently looking. We derived a simulated
U-6 series from 1978 onward based on similar questions in the
CPS. The simulated series replicates the actual reported series
from 1994 onward.
7Okun’s law simply states a linear negative relationship exists
between economic activity (that is, GDP growth) and the unem­
ployment rate.
8To be clear, there are other series that are hard to forecast within
this simple statistical model. We also underpredict the increase in
those who are part-time workers for economic reasons and the frac­
tion of the population outside of the labor force but not marginally
attached. These results are not reported but available upon request.

Federal Reserve Bank of Chicago

10This is true even when controlling simultaneously for all of the
characteristics listed in table 1 (p. 35) in a regression framework
11 See Aaronson and Sullivan (1998) for similar results on job
displacement and job insecurity.

13Notably, changes in industrial structure have little impact. See, for
example, Rissman (2009), Valletta and Cleary (2008), and Aaronson,
Rissman, and Sullivan (2004) on the role of sectoral reallocation
on labor market conditions during recent recessions.
14Movements between being in and out of the labor force play a much
smaller role in explaining shifts in the unemployment rate, so this
discussion largely abstracts from these transitions for simplicity.
But we return to transitions between being unemployed and out of
the labor force (UO and OU) during the most recent recession later
in the article.
15The term “separations,” however, is often used elsewhere to
represent all transitions out of a particular job, including job-to-job
transitions. The separation and hiring rates reported in the U.S. Bureau
of Labor Statistics’ Job Openings and Labor Turnover Survey (JOLTS)
and Business Employment Dynamics (BED) survey also include out
of the labor force transitions.
16Mazumder (2008), using the U.S. Census Bureau’s Survey ofIncome
and Program Participation (SIPP), also finds that the separation
rate has been of somewhat greater importance in recent recessions
than suggested by Shimer (2007).

17Typically, EE is a continuously employed person. However, it can
also be someone who transitions from one job to another without a
spell of non-employment.

49

18It is important to note that there is an “adding up” constraint because
the three probabilities must sum to 1. Therefore, it is not possible
to vary the paths of all three variables simultaneously.
19It should be noted that this transition model is not additive. Allowing
all of the transitions starting from E and all the transitions starting
from U to follow their actual course (simultaneously) accounts for
about 3.4 percentage points of the actual increase of 5 percentage
points in the unemployment rate, leaving some significant share of
the increase attributable to changes in transitions starting from O.
20We also match the rise in the share of the long-term unemployed
quite well.
21While this result is unlikely to be surprising, it should be noted
that it need not be the case. The length of unemployment can in­
crease, with a lag, from a surge in job separations.

22Some of the variation in federal extensions occurs at the state level
because of state-specific triggers for unemployment insurance ben­
efit extensions that depend on the severity of unemployment at the
state level.
23Specifically, for each month we multiply the difference in the
maximum eligibility of UI benefits over and beyond 26 weeks
by the elasticity of a one-week increase in extensions on average

duration of unemployment. This product is scaled by the fraction
of the unemployed receiving UI benefits in that month. The resulting
estimate represents the full effect of the extension over some period
of time. We then divide this effect by 12 to effectively spread out the
total effect over the next 12 months. Finally, we take a running sum
of the effects over the previous 12 months. Starting in December 2008,
when maximum UI eligibility exceeded one year, we began to spread
the effect over the next two years. See note 24 for more details on
the choice of how long to spread out the impact of the extension.
24The effect of an extension on the average duration of unemploy­
ment is not instantaneous. For example, the elasticity of 0.2 from
Katz and Meyer (1990) is based on simulating their model on indi­
viduals over a two-year period. They found similar results from
simulating the model over one year or three years. If we were to
spread the effect of the initial increase in benefits over two years,
this would lower the estimated contributions of the UI extensions
only up until October 2009, but would have no effect on the total
contributions as of December 2009.

25Note that there is no spike at 26 weeks, the typical maximum
number of weeks of UI eligibility. Although the CPS does not
show a spike, administrative unemployment insurance records
typically do (see, for example, Ham and Rea, 1987; Katz and
Meyer, 1990; and Meyer, 1990, 1995).

REFERENCES
Aaronson, Daniel, Scott Brave, and Shani Schechter,
2009, “How does labor adjustment in this recession
compare with the past?,” Chicago Fed Letter, Federal
Reserve Bank of Chicago, No. 263, June.

Blanchard, Olivier Jean, and Peter Diamond,
1994, “Ranking, unemployment duration, and wages,”
Review ofEconomic Studies, Vol. 61, No. 3, July,
pp. 417-434.

Aaronson, Daniel, and Daniel G. Sullivan, 1998,
“The decline of job security in the 1990s: Displace­
ment, anxiety, and their effect on wage growth,”
Economic Perspectives, Federal Reserve Bank of
Chicago, Vol. 22, No. 1, First Quarter, pp. 17—43.

Card, David, and Phillip B. Levine, 2000, “Extended
benefits and the duration of UI spells: Evidence from
the New Jersey extended benefit program,” Journal
ofPublic Economics, Vol. 78, Nos. 1-2, October,
pp. 107-138.

Aaronson, Daniel, Ellen R. Rissman, and Daniel
G. Sullivan, 2004, “Can sectoral reallocation explain
the jobless recovery?,” Economic Perspectives,
Federal Reserve Bank of Chicago, Vol. 28, No. 2,
Second Quarter, pp. 36-49.

Chatterjee, Satyajit, and Dean Corbae, 2007,
“On the aggregate welfare cost of Great Depression
unemployment,” Journal ofMonetary’ Economics,
Vol. 54, No. 6, September, pp. 1529-1544.

Abraham, Katherine G., and Robert Shimer, 2002,
“Changes in unemployment duration and labor force
attachment,” in The Roaring Nineties: Can Full
Employment Be Sustained?, Alan B. Krueger and
Robert M. Solow (eds.), New York: Russell Sage
Foundation and Century Foundation, pp. 367-420.
Aspen Publishers, 2010, “Quarterly Blue Chip
forecasts,” Blue Chip Economic Indicators, Vol. 35,
No. 1, January 10, p. 5.

50

Chetty, Raj, 2008, “Moral hazard versus liquidity
and optimal unemployment insurance,” Journal of
Political Economy, Vol. 116, No. 2, pp. 173-234.
Couch, Kenneth A., and Dana W. Placzek, 2010,
“Earnings losses of displaced workers revisited,”
American Economic Review, Noi. 100, No. 1,
March, pp. 572-589.

Eichengreen, Barry, and T. J. Hatton (eds.), 1988,
Interwar Unemployment in International Perspective,
Dordrect, Netherlands: Kluwer Academic Publishers.

2Q/2010, Economic Perspectives

Elsby, Michael W. L., Ryan Michaels, and Gary
Solon, 2009, “The ins and outs of cyclical unemploy­
ment,” American Economic Journal: Macroeconomics,
Vol. 1, No. 1, January, pp. 84-110.
Fallick, Bruce C., 1996, “A review of the recent empiri­
cal literature on displaced workers,” Industrial and Labor
Relations Review, Vol. 50, No. 1, October, pp. 5-16.

Fougere, Denis, Francis Kramarz, and Julien Pouget,
2009, “Youth unemployment and crime in France,”
Journal of the European Economic Association,
Vol. 7, No. 5, September, pp. 909-938.
Gruber, Jonathan, 1997, “The consumption smoothing
benefits of unemployment insurance,” American
Economic Review, Vol. 87, No. 1, March, pp. 192-205.

Ham, John C., and Samuel A. Rea, Jr., 1987, “Un­
employment insurance and male unemployment dura­
tion in Canada,” Journal ofLabor Economics, Vol. 5,
No. 3, July, pp. 325-353.
Jacobson, Fouis S., Robert J. FaFonde, and
Daniel G. Sullivan, 1993, “Earnings losses of dis­
placed workers,” American Economic Review,
Vol. 83, No. 4, September, pp. 685-709.

Katz, Fawrence F., and Bruce D. Meyer, 1990,
“The impact of the potential duration of unemployment
benefits on the duration of unemployment,” Journal of
Public Economics, Vol. 41, No. 1, February, pp. 45-72.
Kiefer, Nicholas M., 1988, “Economic duration data
and hazard functions,” Journal ofEconomic Literature,
Vol. 26, No. 2, June, pp. 646-679.
Krusell, Per, Toshihiko Mukoyama, Ayyegiil §ahin,
and Anthony A. Smith, Jr., 2008, “Revisiting the
welfare effects of eliminating business cycles,”
Federal Reserve Bank of New York, working paper,
December, available at www.newyorkfed.org/
research/economists/sahin/kmss.pdf.

Machin, Stephen, and Alan Manning, 1999, “The
causes and consequences of long-term unemployment
in Europe,” in Handbook ofLabor Economics,
Vol. 3C, Orley C. Ashenfelter and David Card (eds.),
Amsterdam: Elsevier, North-Holland, pp. 3085-3139.

Meyer, Bruce D., 1995, “Lessons from the U.S.
unemployment insurance experiments,” Journal of
Economic Literature, Vol. 33, No. 1, March, pp. 91-131.
___________ , 1990, “Unemployment insurance and
unemployment spells,” Econometrica, Vol. 58, No. 4,
July, pp. 757-782.
Mukoyama, Toshihiko, and Ay$egiil §ahin, 2009,
“Why did the average duration of unemployment
become so much longer?,” Journal ofMonetary
Economics, Vol. 56, No. 2, March, pp. 200-209.

Neal, Derek, 1995, “Industry-specific human capital:
Evidence from displaced workers,” Journal ofLabor
Economics, Vol. 13, No. 4, October, pp. 653-677.
Oreopoulos, Philip, Marianne Page, and Ann Huff
Stevens, 2008, “The intergenerational effects of worker
displacement,” Journal ofLabor Economics, Vol. 26,
No. 3, pp. 455-483.

Rissman, Ellen R., 2009, “Employment growth:
Cyclical movements or structural change?,” Economic
Perspectives, Federal Reserve Bank of Chicago,
Vol. 33, No. 4, Fourth Quarter, pp. 40-57.
Shimer, Robert, 2007, “Reassessing the ins and outs
of unemployment,” National Bureau of Economic
Research, working paper, No. 13421, September.

Sullivan, Daniel G., and Till von Wachter, 2009,
“Job displacement and mortality: An analysis using
administrative data,” Quarterly Journal ofEconomics,
Vol. 124, No. 3, August, pp. 1265-1306.

Valletta, Robert G., 2005, “Rising unemployment
duration in the United States: Causes and consequences,”
Federal Reserve Bank of San Francisco, working
paper, May, available atwww.frbsf.org/economics/
economists/rvalletta/RV_duration_5-05_new.pdf.
Valletta, Robert G., and Aisling Cleary, 2008,
“Sectoral reallocation and unemployment,” FRBSF
Economic Letter, Federal Reserve Bank of San
Francisco, No. 2008-32, October 17.
Woytinsky, W. S., 1942, Three Aspects ofLabor
Dynamics, Washington, DC: Social Science Research
Council.

Mazumder, Bhashkar, 2008, “New evidence on
labor market dynamics over the business cycle,”
Economic Perspectives, Federal Reserve Bank of
Chicago, Vol. 31, No. 1, First Quarter, pp. 36-46.

Federal Reserve Bank of Chicago

51

Do labor market activities help predict inflation?
Luojia Hu and Maude Toussaint-Comeau

Introduction and summary
Price stability is an important element in maintaining
a healthy economy. Volatile prices, especially when
unanticipated, can have a negative impact on aggregate
demand, as people are not able to adjust and protect
the real value of their financial wealth.1 Such uncer­
tainty can result in disruptions in business planning
and reductions in capital investment spending, which
could be detrimental for the long-run growth potential
of the economy. In addition, inflation can impact eco­
nomic welfare as wealth and income redistributions
occur among different agents (Doepke and Schneider,
2006; and Franke, Flaschel, and Proano, 2006).
As experiences of some Latin American countries
with hyperinflation have shown, economic growth can
be seriously impaired by very high inflation (Heyman
and Leijonhufvud, 1995; and Rogers and Wang, 1993).
But even at much less severe levels, inflation matters.
The U.S. recessions of 1973-1975, 1980, and 1981-82
were all preceded by elevated levels of inflation
(Gordon, 1993).
Because of the intrinsic role of price stability in
a healthy economy, controlling inflation is a primary
objective of monetary policymakers. Understanding
the nature of business cycles and short-run inflation
dynamics is essential for the appropriate conduct of
monetary policy (Svensson, 1997; and Clarida, Gali,
and Gertler, 2000). In order to control inflation effec­
tively, policymakers need to identify key indicators
that help to predict inflation. Among these factors,
labor market activities and, in particular, wages are
closely watched. Indeed, since Phillips’ (1958) paper
demonstrated that there is an inverse relationship be­
tween the rate of change in money wages and the rate
of unemployment, the relevance of the labor market and,
in particular, the link between wages and prices have
been taken as given, as noted in Fosu and Huq (1988).

52

It is unclear whether wage inflation causes price
inflation or vice versa. If rising demand for goods and
services reduces unemployment (causing it to fall
below some natural rate), inflationary pressures might
develop as firms bid against each other for labor and
as workers feel more confident in pressing for higher
wages. Then higher wages could lead to still higher
prices. (In an extreme case, this might lead to a wageprice spiral, which we saw in the United States during
the 1970s [Perry, 1978]).
However, if rising demand for goods and services
(for example, too much money chasing too few goods)
induces firms to raise their prices, these price increases
and greater profits could entice workers to demand
higher wages. In such an environment, inflation could
lead to wage growth (Friedman, 1956; Cagan, 1972;
and Barth and Bennett, 1975).
If productivity growth drives higher wages, the
firm does not have to pass on higher wages into higher
prices. Increased productivity therefore should curb
inflationary pressures.2
A large body of research has aimed to model the
inflation process empirically. However, as a recent
review indicates, there is no consensus view of the
best explanation for inflation (Rudd and Whelan, 2005).
The literature focusing on how the labor and product
markets interact has also produced mixed results. Much
of the evidence suggests that wage growth, even
adjusted for productivity, is not a causal factor in
determining price inflation. However, inflation does

Luojia Hu is a senior economist and Maude ToussaintComeau is an economist in the Economic Research
Department at the Federal Reserve Bank of Chicago.
The authors thank Alejandro Justiniano and seminar
participants at the Federal Reserve Bank of Chicago for
insightful comments and suggestions. They are also grate­
ful to Kenley Pelzerfor her excellent research assistance.

2Q/2010, Economic Perspectives

help predict wages (Mehra, 1991, 1993, and 2000;
Huh and Trehan, 1995; Emery and Chang, 1996;
Hess, 1999; and Campbell and Rissman, 1994).
In this article, we revisit this question by conduct­
ing an empirical analysis of the role of labor market
activities in inflation, including an examination of the
relationship between productivity-adjusted labor costs
(unit labor costs), unemployment, and price inflation.
We contribute to the body of existing evidence with our
use of updated and more recent data, including data for
the past ten years. After incorporating alternative em­
pirical approaches and elements from previous studies,
we reach a fairly simple conclusion. Wage inflation is
not very informative for predicting price inflation, espe­
cially during the period from 1984 onward, which has
been dubbed by economists as “the Great Moderation.”
However, price inflation does seem to help predict wages.
We find that the unemployment data contain additional
information for both wages and prices, which supports
a Phillips curve type of relationship between them
(Stiglitz, 1997).
In the next section, we provide a brief review of
the theoretical and empirical approaches to modeling
price and wage inflation. Then we present our data
and discuss the econometric model of the wage and
price relationship. Finally, we test for the direction of
causality between wages and prices.

Theoretical background: Modeling inflation
Irrespective of the causes for inflation, the tight
relationship between wages and prices follows the
paradigm of the profit-maximizing firm. In its simplest
form, the firm hires labor until the cost of hiring one
additional worker equals the revenue that she generates.
The cost of an extra labor unit (worker) is taken as the
going wage rate (assuming that workers are homoge­
nous and the firm hires in a spot labor market, where
transactions happen immediately). The firm sells its
product in a spot market. The additional revenue that
the firm gets from hiring one additional worker is equal
to the market price of the product times the extra out­
put that she produces. In such a market, the output price
is determined by the price of the labor inputs and their
productivity. This implies productivity-adjusted nom­
inal wages grow at the same rate as product prices. In
this simplified world, where the firm is a price taker
in labor and product markets, the price inflation-wage
inflation gap is always equal to zero.
If these assumptions are relaxed, some conditions
that arise can weaken the tight link between wages
and prices in the short run. As discussed in Campbell
and Rissman (1994) and Huh and Trehan (1995),

Federal Reserve Bank of Chicago

labor market imperfections and certain frictions, such
as adjustment costs (for example, the cost of changing
employment or the presence of nominal wage rigidities),
can create a wedge between the marginal product of
labor and the wage rate. Such a wedge would weaken
the simple framework’s strong connection between price
and wage inflation. In this case, in the short run there
would be a deviation away from the long-run equilib­
rium between price inflation and productivity-adjusted
wage growth. However, over time, price and wage in­
flation should revert to their equilibrium relationship.
The original Phillips curve model was formulated
as a wage equation relating wage inflation to the un­
employment gap. But the idea that systematic move­
ments in prices and wages may be correlated is linked
to the rationalization of other formulations of the model,
such as the expectation-augmented Phillips curve.
Attributed to versions of work by Robert J. Gordon
and also known as the Gordon triangle model of infla­
tion, the expectation-augmented Phillips curve sug­
gests that prices are set as a markup over productivityadjusted wages and are affected by cyclical demand
dynamics, such as unemployment gaps or output gaps,
and supply shocks, such as oil price shocks. In turn,
wages are a function of expected prices and demand and
supply shocks. Expected prices depend on past prices
(Gordon, 1982, 1985; and Stockton and Glassman,
1987). The Gordon triangle model implies a relation­
ship between wages and prices that runs in both direc­
tions in the long run. If the proposition is correct (and
assuming the markup is constant or slow-moving),
then long-run movements in prices and labor costs are
correlated. In the short run, if prices are slow to respond
to shocks in the labor market (and we allow for short­
term dynamics in such behavior), we should also further
expect that short-run movements in labor costs would
help predict short-run movements in prices. A number
of previous researchers have sought to establish the
direction of causation between wages and prices, using
the framework of the expectation-augmented Phillips
curve and Granger causality tests3 (Mehra, 1991, 1993,
2000; Huh and Trehan, 1995; Gordon, 1988, 1998;
Emery and Chang, 1996; Hess, 1999; Campbell and
Rissman, 1994; and Ghali, 1999).4
As noted by Stock and Watson (2008, p. 1), the
traditional backward-looking Phillips curve “continues
to be the best way to understand policy discussions
about the rates of unemployment and inflation.” Much
of the evidence in the empirical literature based on
the backward-looking Phillips curve suggests that
wages are not a causal factor in determining inflation.
However, price inflation does help predict wages.

53

As in much of the literature, Campbell and Rissman
(1994) and Mehra (2000) find that wages do not help
predict prices. Ghali (1999), a rare exception, finds
that they do. In terms of econometric methodology,
all three papers include an error correction (EC) term
in the Gordon triangle model to accommodate some
nonstationarity in the series and allow for co-integra­
tion (the long-run relation between prices and wages).
Once they establish a relationship, they test for the
direction of the causality via Granger causality tests.
The measures of prices and wages in the three papers
are similar. Prices are measured by the gross domes­
tic product (GDP) price deflator, and productivityadjusted wages are measured by unit labor costs.
In these papers, the authors consider different
sample periods. Campbell and Rissman (1994) use
data that cover 1950:Ql-1993:Q3; Mehra (2000)
uses data for 1952:Q1-1999:Q2 and considers some
subsample periods; and Ghali (1999) uses data cover­
ing 1959:Q1-1989:Q3. The analyses also differ in the
way they transform the data to ensure stationarity.5
As we mentioned previously, to accommodate the
co-integration relation of the time series, all three
papers use error correction models (ECM); however,
Campbell and Rissman assume a known (one-to-one)
error correction, while Mehra and Ghali assume that
the EC is unknown and estimate the co-integration
equation. The papers differ in the ways they capture
short-run dynamics of supply and demand factors.
Campbell and Rissman consider only demand, which
they proxy with an unemployment rate variable. Mehra
also uses only demand, but proxies it with the output
gap and changes in unemployment rates. Ghali uses
both demand (output gap) and supply (relative import
prices). The papers also differ in their assumptions
regarding the exogeneity of the demand and supply
variables. Campbell and Rissman and Mehra assume
that the demand factors are exogenous in the long-run
equilibrium relation, so their variables do not enter
the co-integration equation. But Ghali allows both de­
mand and supply variables to enter the co-integration
equation. Finally, they use different estimation methods.
Campbell and Rissman use ordinary least squares
(OLS). Mehra also uses OLS, but includes a first-step
estimation of the co-integration equation between
prices and wages. Ghali uses full maximum likelihood
estimation (MLE), a technique that allows for multi­
ple co-integration equations among prices, wages,
and the demand and supply variables.
In this article, we incorporate various elements
of these three papers to conduct (in-sample) forecasting
of wage and price inflation within an expectationaugmented Phillips curve framework. To be consistent

54

with the literature that suggests that the time period
matters, we conduct the analysis on both a full sample
(which includes updated data for the past ten years),
1960:Ql-2009:Q2, and a subsample, 1984:Ql-2009:Q2.
We then conduct in-sample causality tests of several
versions of the error correction model: 1) assuming
a known versus an unknown co-integration relation;
2) including both supply shocks and demand dynamics
with alternative measures; and 3) treating supply shocks
and demand as exogenous versus endogenous.
A number of studies have looked into a new ver­
sion of the Phillips curve model, the so-called new
Keynesian Phillips curve, or NKPC (Chadha, Masson,
and Meredith, 1992; and Fuhrer and Moore, 1995);
however, this approach is not within the scope of our
work in this article. This new model emphasizes stag­
gered (spread out over time) nominal wages and assumes
price setting by forward-looking agents. The main
difference between the traditional Phillips curve and
the NKPC is that in the latter, expected future inflation
is the determinant of current inflation, whereas in the
traditional expectation-augmented Phillips curve,
lagged inflation plays a major role. As formalized in
Yun (1996), the Calvo (1983) model of staggered pric­
ing and the Taylor (1980) model of staggered contracts
are the workhorses of the NKPC. For example, Gali
and Gertler (1999) and Mehra (2004) use a specifica­
tion of the NKPC inflation model in which current
inflation is modeled as a function of contemporaneous
demand factors and of both lagged and expected infla­
tion. Sbordone’s (2002) model also emphasizes stag­
gered nominal wage and price setting by forward-looking
agents, but allows for imperfect competition with nom­
inal price rigidity, implying an equilibrium pricing con­
dition whereby current inflation is linked to lagged
inflation and expected future real marginal costs. In
sum, the main differences among these different studies
are both the degree to which forward-looking, as
opposed to backward-looking, elements matter and
the way in which the inertia in prices is introduced
(Calvo prices versus Taylor contracts).

Data
As a starting point, we take a look at the data on
wages and prices and the other demand and supply
economic indicators for our sample period, 196O:Q12009:Q2. We define prices as the GDP deflator con­
sistent with the three papers we discussed earlier—
Campbell and Rissman (1994), Mehra (2000), and
Ghali (1999).6 For wages, we use unit labor costs for
the nonfarm business sector (ULC). ULC is nominal
wages, adjusted for labor productivity (ULC = IT x L/Y,
where W equals nominal wages, L equals hours per

2Q/2010, Economic Perspectives

BOX 1

Definitions of variables

p = log(GDP deflator), where GDP is gross
domestic product
w = log(ULC), where ULC is unit labor costs
for the nonfarm business sector

3ip = Ap, quarter-to-quarter growth rate
of GDP deflator
ji" =

Aw, quarter-to-quarter growth rate of ULC

g = log(real GDP/potential GDP); that is,
the output gap

u = unemployment rate - nonaccelerating
inflation rate of unemployment (NAIRU);
that is, the unemployment gap
imp = log(relative import price deflator inclusive
of oil/GDP deflator)

worker, and Y equals output, implying ULC= WI(YIL)\
Box 1 summarizes the definition of the variables used
in this analysis.
Figure 1 charts the time series of the GDP price
deflator and ULC over the period 1960:Ql-2009:Q2.
This chart clearly shows the correlation between the
two series.
In figure 2, we report the quarter-toquarter change (annualized) in the two
series.7 We note two distinctive periods:
Inflation and wage growth increased in
quite dramatic fashion in the 1970s (this
is the period known for the wage-price
spiral phenomenon). From the mid-1980s
onward, we see a tapering off of inflation
and wage growth.8 Looking more closely
at the co-behavior of the two series, from
the mid-1960s up to 1984, the two series
show quite a lot of co-movement. From
1984 onward, there appears to be much
less co-movement between wage growth
and price inflation. In fact, while wage
growth continues to fluctuate, price infla­
tion remains markedly low and stable.
This figure suggests that the relationship
between the two series may not be stable
over the full sample period and that, as
others analyzing trends in inflation and
wage growth have suggested, these series
may not have a “normal,” or built-in, level
and therefore shocks to them could be
quite persistent (Fuhrer and Moore, 1995;
and Benati, 2008).

Federal Reserve Bank of Chicago

The difference between the quarter-to-quarter in­
flation rate of the GDP deflator (ji") and growth rate
of ULC (jiw) is shown in figure 3. The difference can
be viewed as representing a deviation from the longrun equilibrium (assuming a one-to-one or unit rela­
tionship, EC =np - ji"). This is clearly a simplifying
assumption. We later consider versions of the model
that assume constrained co-integration, where we im­
pose unit coefficients, but we also consider a version
of the model with unconstrained co-integration, where
we estimate the coefficients for the error correction term.
Following the theoretical proposition of the profit­
maximizing firm, such a deviation should revert to its
mean in the long run. Consistent with this, in figure 3
we note that the disequilibrium term has been fluctuating
around a mean of zero (that is, it has not gone up or
down over time in a discernible trend). There is clearly
a long-term relation between the two series, but it is
unclear whether there is a causal relationship or, if
there is, which one causes the other.
Figures 4 and 5 report the measures of excess
demand or slack in the economy—that is, the unem­
ployment gap and output gap. As noted in box 1, the
unemployment gap is the difference between the civil­
ian unemployment rate and the nonaccelerating infla­
tion rate of unemployment (NAIRU). The NAIRU is
provided by the Congressional Budget Office (CBO),

55

and it is an equilibrium rate that does not
tend to increase or decrease the inflation
rate. The output gap is the logarithm of
the ratio of real GDP to potential real GDP.
Potential real GDP is also estimated by
the CBO. As can be expected, we note in
these figures that unemployment increased
and the output gap decreased in periods
of economic slowdown (for example, in
the 1970s, 1980s, and early and late 2000s).
Finally, figure 6 shows the time
series of the relative prices of imports,
a measure of supply shocks. The role
of import prices is fairly obvious. The
aggregate supply curve should shift
when input prices change, and input
prices are affected by the prices of im­
ports. The figure shows that prices of
imports changed very little in the 1960s
and early 1970s. They increased substan­
tially in 1974 and again in 1979-80.
Since 1981, relative import prices have
changed very little. We would therefore
expect that this variable should be rela­
tively less important for explaining infla­
tion in the past three decades.

Empirical estimation
To make clear the hypotheses that
we will be testing, it is useful to describe
in more specific terms the expectationaugmented Phillips curve model. The
basic relationships are represented by
the following system of equations:
1)

2)

< =k0+kl ne,’p +k2DD, +k3SSw„

3)
j
where rcf is the first difference of the log
of the price level; n,” is the first difference
of the log of the nominal rate of ULC; DDt
is a vector of demand pressure variables,
which include g (the output gap) and/or
u (the unemployment gap) as defined
previously.9 The term ne,-p is the expected
inflation level, SS represents supply
shocks affecting the price equation, and

56

2Q/2010, Economic Perspectives

5'5'w( represents supply shocks affecting
the wage equation. Such supply shocks
are proxied by imp (the relative import
prices inclusive of oil) and two period
dummies indicating President Nixon’s
price and wage control periods. (The
first period is 1971:Q3 -1972:Q4, and
the second period is 1973:Q1-1974:Q4.)
As can be seen, equation 1 reflects
the idea that prices are a markup over
productivity-adjusted wages and are af­
fected by cyclical demand and relative
supply shocks. Equation 2 shows that
wages are affected by demand and supply
and expected price level. Equation 3 shows
that expected inflation is a function of
past prices. Further, to accommodate the
statistical features of the time series, we
include an error correction term in the
Gordon triangle model (equations 1-3).
We also keep the demand and supply
variables to affect the short-run dynamics
of prices and wages. This is represented
as follows:

= «*(<! - <_j)+2L y!a</
/=1

+£x;a<;+£p;w,
/=1

/=1

+2L *X-Z + £!/=1

A< = a2 (<, -

+ £ y?A<_,
/=1

+ £xX-(+XpM
/=1

l=\

+2L ^z’^-Z + A
Z=1

The error correction term
(EC —
j) allows for a long-run
equilibrium relationship between price
and wage inflation. The parameter a there­
fore reflects long-run dynamics, and y and
A, capture short-run dynamics. The term
sj is the residual from the price equation,
while s2 is the residual from the wage re­
gression. L is the maximum number of lags
on the various variables needed to make
the random disturbances serially uncorre­
lated. Again, as previously noted, DD and
SS are vectors of variables representing

Federal Reserve Bank of Chicago

57

demand and supply shocks affecting price
and wage inflation, as in some previous
studies (for example, Mehra, 2004; and
Hess and Schweitzer, 2000).
We have the following hypotheses
concerning the joint short-run and long-run
equilibrium relationships in wages and prices:
Hypothesis 1 is that wages do not predict
prices (//„ :a‘ = 0,Z‘ = 0,..., Zj. = 0).
Hypothesis 2 is that prices do not predict
wages (Wo: a2 = 0,y2 = 0,..., y2 = 0).
Recalling that the parameter a re­
flects long-run dynamics, while y and A,
capture short-run dynamics, we test for
the hypotheses and determine the sources
of the short-run and long-run co-move­
ments between wages and prices, using
Granger causality tests. Our test for
Granger causality involves examining
whether lagged values of one series (that
is, wages) have significant explanatory
power for another variable (that is, prices).
In this exercise, both variables may
Granger-cause one another.10 Both series
in question may also be co-integrated.
Recall that by incorporating an error correction term
in the Granger causality tests, we allow the series in
levels to catch up with or equal one another. The sig­
nificance of the error correction term in the Granger
causality test would signal the fact that the series in
question are driven to return to a long-run equilibrium
relationship that is causal.

Granger causality test results
Before conducting the Granger causality tests,
we examined the stationarity of the series. The results
of the augmented Dickey-Fuller (ADF) unit root tests
for price inflation and wage inflation confirmed that
we cannot reject the null hypothesis of a unit root at
the 1 percent level—that is, the growth rates of prices
and wages are both integrated of order one, 1(1). Also,
for the full sample period (1960:Ql-2009:Q2), the
relative import prices (imp), unemployment gap (;/),
and output gap (g) are also all 1(1).
Table 1 presents the results of our tests for Granger
causality between wages and prices for the full sample
period, 1960:Ql-2009:Q2, and for a subsample period,
1984:Ql-2009:Q2. In this bivariate model, we assume
that DD = 0 and SS = 0. The regression includes lagged
prices and lagged unit labor cost growth. The number
of lags for each variable is set to four (L = 4). Panel A
of table 1 reports the evidence on whether the column
variables Granger-cause price inflation, while panel B

58

shows the evidence for whether the column variables
Granger-cause wage growth. The error correction col­
umn refers to the long-run effect; the wages column
(in panel A) and prices column (in panel B) refer to
the short-run effect; and the joint hypothesis column
refers to the long- and short-run effects. Each column
reports the p value, the level of statistical significance
with which one can reject the null hypothesis. A high
p value should be taken as evidence that the column
variable does not Granger-cause price or wage inflation.
Referring back to the ECM, to be co-integrated,
at least one of the a in the two equations should not
be equal to zero. Looking at panel A of table 1 for the
full sample period, 1960:Ql-2009:Q2, the highp val­
ue in the error correction column means that a1 = 0.
Therefore, we can say that prices do not catch up with
wages in the long run. But rather wages adjust to catch
up with prices (a2 # 0), per the lowp value for error cor­
rection in panel B. The highp value for hypothesis 1
of the joint test of error correction and wages (panel A,
third column) suggests that wages don’t help predict
prices in either the short run or the long mn (at a 5 per­
cent significance level).
To summarize the results in table 1, wages do not
cause price inflation in our Granger causality tests. How­
ever, prices do cause wage inflation. Wages, but not
prices, adjust to maintain the long-run equilibrium re­
lationship. This is true for both the full sample (1960 :Q 12009:Q2) and the subsample (1984;Ql-2009;Q2).

2Q/2010, Economic Perspectives

TABLE 1

Granger causality test: Bivariate model
A. Are prices caused by

B. Are wages caused by

Error
correction

Wages

Hypothesis 1:
Joint test of
error correction
and wages

1960: Q1-2009: Q2

0.32

0.34

0.06

0.00

0.06

0.00

1984:Q1-2009:Q2

0.82

0.29

0.27

0.00

0.38

0.00

Period

Error
correction

Prices

Hypothesis 2:
Joint test of
error correction
and prices

Notes: The number of lags for each variable is set to four. Each column reports the p values, indicating the level of statistical significance for the test
that the column variable does not Granger-cause either price inflation or wage inflation. See the text for details on hypothesis 1 and hypothesis 2.
Sources: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics from Haver Analytics.

Besides the price-wage inflation gap, our model
stipulated that there are other short-run demand and
supply determinants of price and wage inflation. To
allow for these cyclical (that is, excess) demand factors
to additionally affect wages and prices in the short
run, we add the unemployment gap (and, alternatively,
the output gap, which we do not report in the table) to
our regressions. We also add supply variables, as proxied
by the relative import prices and dummy variables for
the Nixon price and wage control periods. We run the
regressions with these demand and supply control
variables in differences, as we found that they were I( 1).
Both demand and supply control variables include their
lags, which were set to four. And again, we include the
error correction term.
Table 2 reports the p values from the Granger
causality tests for this augmented model. The results
in both panels A and B of this table suggest that wages
do not predict prices; however, prices do predict wages.
In the long run. wages adjust to the error correction,
while prices do not. In other words, price and wage
inflation move together in the long run because wages
adjust to close the gap, and not because price inflation
responds to wage growth.
As for the additional regressors, for the full sam­
ple, the unemployment gap has additional predictive
power for both price and wage inflation, while the
relative import prices only help predict price inflation.
For the subsample, 1984:Ql-2009:Q2, the unemploy­
ment gap only helps predict price inflation, while the
relative import prices do not help predict either. Using
alternative measures of excess demand (for example,
changes in the unemployment rate or the output gap)
yields qualitatively similar results, which we do not
report here.
We find the result of the informational content of
the unemployment gap for both wage and price infla­
tion interesting; it suggests that such cyclical variables

Federal Reserve Bank of Chicago

play an important short-term role in determining infla­
tion (Campbell and Rissman, 1994). In the tradition
of a Phillips curve type of relationship, price inflation
thus appears to be still very much a labor market phe­
nomenon (Stiglitz, 1997).
The two models that we have discussed thus far
constrain the co-integration relationship between price
inflation and wage inflation to be one to one, which can
be justified by theory under the assumption of perfect
competition and a Cobb-Douglas production function
(for example, as in Campbell and Rissman, 1994).
However, this might be too restrictive an assumption.
We relax this restriction and consider a general­
ized model, allowing for an unconstrained co-integra­
tion relationship between prices and wages (that is, in
the unconstrained case, we estimate the coefficients for
the error correction terms). Moreover, we also allow
the supply and demand variables to enter the long-run
equilibrium relation. In other words, the supply and
demand variables are now treated as endogenous and
could enter the error correction. (For simplicity, we do
not reproduce the new augmented generalized ECM,
but note that this means our model now gets augmented
by two more equations with the demand and supply
variables on the left-hand side). First, looking at the
p value results for the joint short-run and long-run
hypothesis between wages and prices based on the
unconstrained model in table 3, panels A and B, we
note that similar to the results in table 2, wages do not
help predict prices, but prices do help predict wages.
Recall that in this new unconstrained model, the
coefficients for all the variables are being estimated.
This ECM was estimated by the maximum likelihood
estimation technique. For the full sample, the model
was found to be co-integrated with rank 2 (that is, it
has two unique co-integration relationships). The two
estimated co-integration relationships, with the stan­
dard errors in parentheses, are as follows:

59

TABLE 2
Granger causality test: Multivariate model

A. Are prices caused by

Error
correction

Wages

Unemployment
gap

Relative
import
prices

1960:Q1-2009:Q2

0.26

0.63

0.00

0.00

1984:Q1-2009:Q2

0.87

0.16

0.00

0.10

Period

B. Are wages caused by

Hypothesis 1:
Joint test of
error correction
and wages

Relative
import
prices

Hypothesis 2:
Joint test of
error correction
and prices

Error
correction

Prices

Unemployment
gap

0.29

0.00

0.00

0.00

0.09

0.00

0.17

0.00

0.21

0.05

0.75

0.00

Notes: The number of lags for each variable is set to four. Each column reports the p values, indicating the level of statistical significance for the test that the column variable does not Granger-cause either price
inflation or wage inflation. See the text for details on hypothesis 1 and hypothesis 2.
Sources: Authors’calculations based on data from the U.S. Bureau of Labor Statistics and Congressional Budget Office from Haver Analytics.

TABLE 3
Granger causality test: Multivariate model with unknown co-integration parameters

B. Are wages caused by

A. Are prices caused by

Period

Error
correction

Wages

Unemployment
gap

Relative
import
prices

Hypothesis 1:
Joint test of
error correction
and wages

Error
correction

Prices

Unemployment
gap

Relative
import
prices

Hypothesis 2:
Joint test of
error correction
and prices

2Q /2 010, Econom ic Per spe ctiv es

1960:Q1-2009:Q2

0.25

0.40

0.00

0.00

0.06

0.00

0.00

0.09

0.01

0.00

1984:Q1-2009:Q2

0.62

0.12

0.00

0.04

0.15

0.00

0.22

0.53

0.21

0.00

Notes: The model was estimated by using the maximum likelihood estimation technique. The number of lags for each variable, chosen by the Akaike Information Criterion (AIC), is set to four. Each column reports
the p values, indicating the level of statistical significance for the test that the column variable does not Granger-cause either price inflation or wage inflation. See the text for details on hypothesis 1 and hypothesis 2.
Sources: Authors’ calculations based on data from the U.S. Bureau of Labor Statistics and Congressional Budget Office from Haver Analytics.

np — 2.84- 1.40z/ + 8.27 imp,
(0.35)

rc" =1.89- 1.98z/ +11.76zwp.
(0.42)

Conclusion

(2.35)

(2.75)

As can be seen, the unemployment gap and rela­
tive import prices variables enter both co-integration
equations significantly. However, we find that the ad­
justment parameters on the error correction terms in
the equations of the unemployment gap and relative
import prices were statistically insignificant. (For
simplicity, we do not report the unemployment gap
and relative import prices equations here.) This sug­
gests that these two variables do not adjust (as wages
and prices do) to maintain the long-run equilibrium
relations. In fact, the likelihood ratio test for the null
hypothesis that the adjustment parameters in the un­
employment gap and relative import prices equations
are jointly zero has ap value of 0.68.
For the subsample, 1984:Ql-2009:Q2, the unem­
ployment gap is 1(2) instead of I( 1). After replacing the
unemployment gap by its first difference, the model is
estimated to have one co-integration relation. In this case,
the unemployment gap and the relative import prices
do not even enter the co-integration equation signifi­
cantly. The unemployment gap and the relative import
prices appear to be exogenous in the long-run equilib­
rium, especially in the subsample period.

Much research has been devoted to not only iden­
tifying the causes of inflation but also gauging which
economic indicators could best measure and predict
inflation. Using more recent and updated data, we an­
alyzed labor market indicators, namely, productivityadjusted wages and unemployment (as well as supply
shock and demand factors), to determine the extent to
which they contain information to help predict inflation.
Similar to previous research, we have found that
wage growth does not cause price inflation in the
Granger causality sense. We found this to be particu­
larly true for the period from 1984 onward (referred
to as the Great Moderation by economists). By con­
trast, price inflation does cause wage growth in the
Granger causality sense. Moreover, unemployment
has additional predictive power for inflation for the
full sample (1960:Ql-2009:Q2), as well as our sub­
sample (1984:Ql-2009:Q2). The unemployment gap
is therefore a useful indicator for inflation.
As the data indicate, in recent years wage growth
has been particularly slow. Given this, some analysts
think that we do not have to be overly concerned
about future inflation. Our findings in this article,
however, do not support the claim that slow wage
growth is a harbinger of low inflation.

NOTES
'For further discussion of the effects of inflation, see, for example,
Dossche (2009).

2Federal Reserve Chairman Ben S. Bernanke (2008) noted in a
speech that we are unlikely to see the 1970s type of wage-price
spiral in today’s economy. Crucial productivity gains that help
blunt inflationary forces were among the several factors cited.
Also, inflation expectations, although somewhat on the rise, are
much lower than they were in the mid-1970s.
3Granger causality is a statistical methodology for demonstrating
whether a variable contains information about subsequent move­
ments in another variable.
4Stock and Watson (2008) provide a survey of the literature of the
past 15 years, which looks at out-of-sample forecast evaluations
based on Phillips curves as well as other inflation forecasting models.

5Mehra (2000) and Ghali (1999) treat prices and wages as integrated
of order one, 1(1), while Campbell and Rissman (1994) treat the
growth rates of prices and wages as 1(1).
6We also conducted the analysis using the U.S. Bureau of Economic
Analysis’s Personal Consumption Expenditures Price Index as the
price measure. Generally, the results were similar.

Federal Reserve Bank of Chicago

7The results in the subsequent analysis are largely robust to an
alternative measure of inflation using a four-quarter change in
price.
8As mentioned earlier, this period has been dubbed by economists
as the Great Moderation, when macroeconomic indicators were
remarkably stable (see, for example, Bernanke, 2004; Kim and
Nelson, 1999; McConnell and Perez-Quiros, 2000).
9Several explanations have been offered in the literature to motivate
unemployment in a wage and price equation. Beside the Phillips
(1958) underlying model of change in wages as a function of the
unemployment rate, the literature of efficiency wages provides
some motivation (for example, Shapiro and Stiglitz, 1984). Huh
and Trehan (1995) provide a summary of the logic of the efficiency
wage approach in explaining the inclusion of unemployment in a
wage and price equation. Also, see Ghali (1999) and Gordon (1988).
10More specifically, the Granger causality test is a two-step regres­
sion procedure used to examine the direction of causality between
two series. For example, to determine whether there is causality
running from p to w, w is first estimated as a function of past values
of w (this is called the restricted equation). Then w is estimated as
a function of past values of w and past values ofp (this is called
the unrestricted regression). There is causality in the Granger sense
from p to w if the inclusion of the past values ofp significantly
improves the estimation of w (that is, by an F test).

61

REFERENCES

Barth, J. R., and J. T. Bennett, 1975, “Cost-push
versus demand-pull inflation: Some empirical evi­
dence,” Journal ofMoney’, Credit, and Banking,
Vol. 7, No. 3, August, pp. 391-397.
Benati, L., 2008, “Investigating inflation persistence
across monetary regimes,” Quarterly Journal of
Economics,No\. 123, No. 3, August, pp. 1005-1060.
Bernanke, B. S., 2008, “Remarks on Class Day 2008,”
speech at Harvard University, Cambridge, MA, June 4,
available at www.federalreserve.gov/newsevents/
speech/bemanke20080604a.htm.

___________ , 2004, “The Great Moderation,” remarks
at the meetings of the Eastern Economic Association,
Washington, DC, February 20, available at www.
federalreserve.gov/BOARDDOCS/SPEECHES/
2004/20040220/default.htm.
Cagan, P., 1972, “Monetary policy,” in Ecozzozzzz'c
Policy and Inflation in the Sixties, W. Fellner (compiler),
Washington, DC: American Enterprise Institute for
Public Policy Research, pp. 89-154.
Calvo, G. A., 1983, “Staggered prices in a utilitymaximizing framework,” Journal ofMonetary’
Economics, Vol. 12, No. 3, September, pp. 383-398.

Campbell, J. R., and E. R. Rissman, 1994, “Long-run
labor market dynamics and short-run inflation,”
Economic Perspectives, Federal Reserve Bank of
Chicago, Vol. 18, No. 2, March/April, pp. 15-27.

Chadha, B., P. R. Masson, and G. Meredith, 1992,
“Models of inflation and the costs of disinflation,”
IMF StaffPapers, Vol. 39, No. 2, June, pp. 395-431.
Clarida, R., J. Gali, and M. Gertler, 2000,
“Monetary policy rules and macroeconomic stability:
Evidence and some theory,” Quarterly Journal of
Economics, No\. 115, No. 1, February, pp. 147-180.
Doepke, M., and M. Schneider, 2006, “Inflation
and the redistribution of nominal wealth,” Journal
ofPolitical Economy, Vol. 114, No. 6, December,
pp. 1069-1097.

Dossche, M., 2009, “Understanding inflation dynamics:
Where do we stand?,” National Bank of Belgium,
working paper, No. 165, June.

62

Emery, K. M., and C.-P. Chang, 1996, “Do wages
help predict inflation?,” Economic Review, Federal
Reserve Bank of Dallas, First Quarter, pp. 2-9.

Fosu, A. K., and M. S. Huq, 1988, “Price inflation
and wage inflation: A cause-effect relationship?,”
Economics Letters, Vol. 27, No. 1, pp. 35-40.
Franke R., P. Flaschel, and C. R. Proano, 2006,
“Wage-price dynamics and income distribution in a
semi-structural Keynes-Goodwin model,” Structural
Change and Economic Dynamics, Vol. 17, No. 4,
December, pp. 452-465.
Friedman, M., 1956, “The quantity theory of money—
A restatement,” in Studies in the Quantity Theory’ of
Money, M. Friedman (ed.), Chicago: University of
Chicago Press, pp. 3-24.
Fuhrer, J., and G. Moore, 1995, “Inflation persistence,”
Quarterly Journal of Economics, Vol. 110, No. 1,
February, pp. 127-159.

Gall J., and M. Gertler, 1999, “Inflation dynamics:
A structural econometric analysis,” Journal ofMonetary’
Economics, Vol. 44, No. 2, October, pp. 195-222.
Ghali, K. H., 1999, “Wage growth and the inflation
process: A multivariate co-integration analysis,”
Journal ofMoney, Credit, and Banking, Vol. 31,
No. 3, part 1, August, pp. 417-431.
Gordon, R. J., 1998, “Foundations of the Goldilocks
economy: Supply shocks and the time-varying
NAIRU,” Brookings Papers on Economic Activity,
Vol. 1998, No. 2, pp. 297-333.

___________ , 1993, Macroeconomics, 6th ed., New
York: HarperCollins.
__________ , 1988, “The role of wages in the infla­
tion process,” American Economic Review, Vol. 78,
No. 2, May, pp. 276-283.
___________ , 1985, “Understanding inflation in the
1980s,” Brookings Papers on Economic Activity,
Vol. 1985, No. 1, pp. 263-299.
___________ , 1982, “Price inertia and policy ineffec­
tiveness in the United States, 1890-1980,” Journal
ofPolitical Economy, Vol. 90, No. 6, December,
pp. 1087-1117.

2Q/2010, Economic Perspectives

Hess, G. D., 1999, “Does wage inflation cause price
inflation?,” University of Rochester, William E. Simon
Graduate School of Business, Bradley Policy Research
Center, Shadow Open Market Committee, policy state­
ment and position paper, No. 99-2, September, avail­
able at https://urresearch.rochester.edu/institutional
PublicationPublicView. action?institutionalltemld=53 9.

Hess, G. D., and M. E. Schweitzer, 2000, “Does wage
inflation cause price inflation?,” Federal Reserve Bank
of Cleveland, policy discussion paper, No. 10, April.

Heyman, D., and A. Leijonhufvud, 1995, High
Inflation, New York: Oxford University Press.
Huh, C. G., and B. Trehan, 1995, “Modeling the
time-series behavior of the aggregate wage rate,”
Economic Review, Federal Reserve Bank of San
Francisco, No. 1, pp. 3-13.

Kim, C.-J., and C. R. Nelson, 1999, “Has the U.S.
economy become more stable? A Bayesian approach
based on a Markov-switching model of the business
cycle,” Review ofEconomics and Statistics, Vol. 81,
No. 4, November, pp. 608-616.
McConnell, M., and G. Perez-Quiros, 2000, “Out­
put fluctuations in the United States: What has changed
since the early 1980s?,” American Economic Review,
Vol. 90, No. 5, pp. 1464-1476.

Mehra, Y. P., 2004, “The output gap, expected future
inflation, and inflation dynamics: Another look,”
Topics in Macroeconomics, Vol. 4. No. 1, article 17,
pp. 1-17.
___________ , 2000, “Wage-price dynamics: Are they
consistent with cost push?,” Economic Quarterly,
Federal Reserve Bank of Richmond, Vol. 86, No. 3,
Summer, pp. 27—43.
___________ , 1993, “Unit labor costs and the price
level,” Economic Quarterly, Federal Reserve Bank
of Richmond, Vol. 79, No. 4, Fall, pp. 35-51.
__________ , 1991, “Wage growth and the inflationary
process: An empirical note,” American Economic
Review, September, Vol. 81, No. 4, September,
pp. 931-937.

Phillips, A. W., 1958, “The relationship between un­
employment and the rate of change of money wages
in the United Kingdom, 1861-1957,” Economica,
Vol. 25, No. 100, November, pp. 283-299.
Rogers, J. H., and P. Wang, 1993, “High inflation:
Causes and consequences,” Economic Review, Federal
Reserve Bank of Dallas, Fourth Quarter, pp. 37-51.
Rudd, J., and K. Whelan, 2005, “New tests of the
new Keynesian Phillips curve,” Journal ofMonetary’
Economics, Vol. 52, No. 6, September, pp. 1167-1181.

Sbordone, A. M., 2002, “Prices and unit labor costs:
A new test of price stickiness,” Journal ofMonetary’
Economics, Vol. 49, No. 2, March, pp. 265-292.
Shapiro, C., and J. E. Stiglitz, 1984, “Equilibrium
unemployment as a worker discipline device,” Hzner/cnn
Economic Review, Vol. 74, No. 3, June, pp. 433M44.
Stiglitz, J. E., 1997, “Reflections on the natural rate
hypothesis,” Journal ofEconomic Perspectives,
Vol. 11, No. 1, Winter, pp. 3-10.
Stock, J. H., and M. W. Watson, 2008, “Phillips
curve inflation forecasts,” National Bureau of Economic
Research, working paper, No. 14322, September,
available at www.nber.org/papers/wl4322.pdf.

Stockton, D. J., and J. E. Glassman, 1987, “An
evaluation of the forecast performance of alternative
models of inflation,” Review ofEconomics and Statistics,
Vol. 69, No. 1, February, pp. 108-117.

Svensson, L. E. O., 1997, “Inflation forecast target­
ing: Implementing and monitoring inflation targets,”
European Economic Review, Vol. 41, No. 6, June,
pp. 1111-1146.
Taylor, J. B., 1980, “Aggregate dynamics and staggered
contracts,” Journal ofPolitical Economy, Vol. 88,
No. 1, February, pp. 1-23.

Yun, T., 1996, “Nominal price rigidity, money supply
endogeneity, and business cycles,” Journal ofMonetary
Economics, Vol. 37, No. 2, April, pp. 345-370.

Perry, G. L., 1978, “Slowing the wage-price spiral:
The macroeconomic view,” in Curing Chronic Inflation,
A. M. Okun and G. L. Perry (eds.), Washington, DC:
Brookings Institution, pp. 259-291.

Federal Reserve Bank of Chicago

63

Thirteenth Annual International Banking Conference

Macroprudential Regulatory Policies:
The New Road to Financial Stability?
September 23-24, 2010
Federal Reserve Bank of Chicago
In conjunction with the International Monetary Fund, the Federal Reserve Bank of
Chicago will hold its thirteenth annual International Banking Conference on
September 23-24, 2010. The purpose of these conferences is to address current
issues affecting international financial markets. This year, we examine the role of
macroprudential regulation in the financial sector. Shocked by the experience of the
last few years, many argue that the more traditional microprudential regulatory tools
are inadequate to create a safe and stable financial system. The microprudential
paradigm relies on the presumption that the financial system as a whole can be
made safe by ensuring individual financial institutions are made safe. This ignores
interconnections and externalities, whereby the actions of one financial institution
or events in financial markets can lead to spillover effects that adversely affect gen­
eral market conditions, other financial institutions, and ultimately the economy as a
whole. Instead, it is argued, there is a need for both microprudential approaches to
regulate individual institutions and macroprudential approaches to manage the overall
financial system risks. However, a number of important questions must be answered.
What are the theoretical motivations for such regulation? How would it interact with
other regulatory and macroeconomic policies, especially monetary policy? What
would be the specific macroprudential tools? Who should have control over the
macroprudential tools? How should a macroprudential regulator be structured?
Where should it be housed? How can macroprudential policies be structured across
national borders? What role, if any, can market discipline play in supporting macro­
prudential objectives? These and related issues will be addressed at the two-day
conference.

As always, the conference will focus on the implications for public policy. It will feature
keynote presentations by Paul Volcker, Chairman of the U.S. President’s Economic
Recovery Advisory Board and former Chairman of the Federal Reserve System; and
Jaime Caruana, General Manager of the Bank for International Settlements. The
makeup of the conference is truly international. The audience consists of represen­
tatives from central banks, regulatory and supervisory agencies, financial institutions,
trade associations, and academic institutions from around the globe. Last year,
attendees came from some 30 countries.
Save the date and plan on attending the conference in September. Additional infor­
mation, including the full agenda and conference and hotel registration details, will
be posted soon at:

www.chicagofed.org/lnternationalBankingConference
Location:

Contact:

Federal Reserve Bank of Chicago
230 South LaSalle Street
Chicago, IL 60604-1413

Ms. Blanca Sepulveda
(312) 322-8340
Blanca.Sepulveda@chi.frb.org