The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
R E A L B U SIN ESS C Y C L E T H E O R Y :
W IS D O M O R W H IM S Y ?
M artin Eichenbaum
W orking Paper Series
M acro Econom ic Issues
Research D epartm ent
Federal Reserve B ank of Chicago
July, 1990 (W P-90-13)
REAL BUSINESS CYCLE THEORY: WISDOM OR WHIMSY?
by
Martin Eichenbaum*
Northwestern University, NBER and
The Federal Reserve Bank of Chicago
July 1990
This paper is a slightly modified version of a plenary session talk given at the 1990 Meetings,
Society for Economic Dynamics and Control, Minneapolis, Minnesota.
* 1 am very grateful to Craig Burnside, Lawrence Christiano and Sergio Rebelo for their advice
and help in preparing this paper. Thomas Sargent and Mark Watson provided numerous useful
comments.
REAL BUSINESS CYCLE THEORY: WISDOM OR WHIMSY?
ABSTRACT
This p ap er assesses th e em pirical p lau sib ility of th e view th a t aggregate
p ro d u ctiv ity shocks account for m ost of th e v aria b ility in post W orld W ar II US o u tp u t.
W e argue th a t th e ty p e of evidence forw arded by proponents of th is pro p o sitio n is too
fragile to be believable. F irst, our confidence in th e evidence is fu n d am en tally affected once
we ab an d o n th e fiction th a t we actu ally know th e tru e values of th e stru c tu ra l p aram eters
of sta n d a rd R eal B usiness Cycle (R B C ) models. W h a t th e d a ta are telling us is th a t, while
p ro d u ctiv ity shocks play som e role in generating th e business cycle, th e re is sim ply an
enorm ous am o u n t of u n c e rta in ty ab o u t ju st w hat percent of aggregate flu ctu atio n s they
actu ally do account for. T h e answ er could be 70% as K y d lan d an d P re sc o tt (1989) claim ,
b u t th e d a ta co n tain alm ost no evidence against eith er th e view th a t th e answ er is really
5% or th a t th e answ er is really 200%. Second, we show th a t p o in t estim ates of th e
im p o rtan ce of technology shocks are extrem ely sensitive to sm all p e rtu rb a tio n s in th e
theory. Allowing for labor hoarding behavior in an otherw ise sta n d a rd R B C m odel reduces
th e ab ility of technology shocks to account for aggregate flu ctu atio n s by 50%. T his finding
provides some su p p o rt for th e view th a t m any of th e m ovem ents in th e Solow residual
w hich are labelled as p ro d u ctiv ity shocks m ay be an a rtifa c t of lab o r hoarding ty p e
phenom ena.
M artin E ichenbaum
D ep artm en t of Econom ics
N orth w estern U n iv ersity
E van sto n , II 60208
This paper discusses tw o sim ple questions w hich are of fu n d am en tal im p o rtan ce to
m acroeconom ists. F irst — w hich im pulses have been th e p rim ary sources of fluctuations in
postw ar US aggregate o u tp u t? J u s t how im p o rta n t have aggregate technology shocks been?
Second, how reliable are existing answers are to th e first question?
T he answ ers to these questions obviously m a tte r from th e perspective of o p tim al
public policy. B u t ju st as im p o rta n t, the perceived answers also m a tte r because they
influence th e research agenda of m acroeconom ists. A round 1977 it seem ed ju st as obvious
to th e rep resen tativ e g rad u ate stu d en t as it was to M ilton F ried m an or R obert Lucas th a t
m onetary in sta b ility is a critical d eterm in an t of aggregate o u tp u t flu ctu atio n s. G ran ted
there was su b stan tial disagreem ent about th e n a tu re of th e relatio n sh ip betw een m onetary
and real phenom ena. B ut th e critical point is th a t those years were m arked by enorm ous
am ounts of research aim ed a t u n derstanding th e pro p ag atio n m echanism s b y which
m onetary policy affects aggregate econom ic activ ity . T h a t th is was a critical item for
business cycle research was, by and large, sim ply tak en for g ran ted .
T he situ a tio n has clearly changed. Since K ydland and P re s c o tt’s (1982) ap p aren t
d em o n stratio n th a t p ro d u ctiv ity shocks can account for all o u tp u t v aria b ility in th e post
W ar US, th e need for an ad eq u ate th eo ry of m onetary and fiscal sources of in sta b ility has
come to seem m uch less pressing. Not surprisingly, th e am ount of research devoted to these
topics has declined precipitously.
Does th e evidence in fact provide such overw helm ing su p p o rt in favor of th e basic
claim of existing R eal Business Cycle (R B C ) theories so as to ratio n alize th is fu n d am en tal
shift in our view of th e business cycle? In m y view it does not. T his is because th e evidence
in favor of th e p roposition th a t p ro d u ctiv ity shocks can account for m ost of th e v ariab ility
in post W orld W ar II US o u tp u t is sim ply too fragile to be believable.
(I) Sm all p ertu rb a tio n s to th e th eo ry a lte r th e conclusion in a basic way.
(II) Sm all changes in th e sta tistic a l m ethods used a lte r th e conclusion in a basic
1
way.
(III) Small changes in th e sam ple period alter th e conclusion in a basic way.
(IV )
A nd m ost im p o rtan tly , our confidence in th e conclusion is fund am entally
affected once we ab an d o n th e convenient fiction th a t we actu ally know th e tru e values of
the stru c tu ra l p aram eters of stan d ard R B C models.
Indeed, once we qu an tify th e u n certain ty in m odel predictions arising from
u n certain ty about m odel p aram eter values, calib rated or otherw ise, our view of w hat th e
d a ta is telling us is affected in a first order way. E ven if we do n ot p e rtu rb th e s ta n d a rd
theory and even if we im plem ent existing form ulations of th a t th e o ry on th e sta n d a rd
postw ar sam ple period an d even if we use th e sta tio n a ry inducing tran sfo rm a tio n of th e
d a ta th a t has becom e sta n d a rd in R B C studies — even th e n th e stro n g conclusions w hich
m ark th is lite ra tu re are u n w arran ted . W h at th e d a ta are actu ally tellin g us is th a t, w hile
technology shocks alm ost certainly play some role in generating th e business cycle, th e re is
sim ply an enorm ous am ount of u n certain ty about ju st w h at p ercen t of aggregate
fluctuation s th ey actu ally do account for. T h e answ er could be 70% as K y d lan d and
P resco tt (1989) claim , b u t th e d a ta contain alm ost no evidence ag ain st e ith e r th e view th a t
th e answ er is really 5% or th a t th e answ er is really 200%.
U nder th ese circum stances, th e decision to d rastically de—em phasize th e im p o rtan ce
of tra d itio n a l im pulses like m o n etary and fiscal shocks in business cycle research ought to
be viewed as w him sical, in th e sense th a t Learner (1983) uses th a t term . A n inference is
ju st not believable if it is fragile. A nd a decision based on a fragile inference is w him sical.
In th is p ap er I discuss th e first an d fo u rth of th e aforem entioned co n ten tio n s . 1 T o
do th is it is useful to consider th e q u a n tita tiv e im plications of one w idely used R B C th e o ry —
iFor a discussion of points II and III see Christiano and Eichenbaum (1989) and Burnside, Eichenbaum
and Rebelo (1990).
2
the indivisible labor model associated w ith G ary H ansen (1985) and R ichard Rogerson
(1988). A ccording to this m odel, th e tim e series on th e beginning of period t cap ital stock,
kj., tim e t consum ption, c^., and tim e t hours w orked, n^., correspond to th e solution of a
social planning problem w hich can be decentralized as a P a re to o p tim al com petitive
equilibrium . T he planner ranks stream s of consum ption services, and leisure, T —n^.,
according to th e criterion function:
(1)
e
J L / S * {1»(c.) + «(T—n )}.
ut= o
1
1
Here T denotes th e rep resen tativ e ag en t’s tim e endow m ent, E q is th e tim e 0 conditional
expectations o p erato r, 0 is a positive scalar, and (3 is th e subjective discount ra te ,
0
< P <
1.
T here are at least tw o in terp reta tio n s of th e te rm involving leisure in ( 1 ). F irs t, it
m ay ju st reflect th e assu m p tio n th a t individual u tility functions are lin ear in leisure. T he
second in te rp re ta tio n builds on th e assum ption th a t th ere are in d iv isib ilities in labor
supply, so th a t in d ividuals can eith er work some fixed positive n u m b er of hours or not a t
all. A ssum ing th a t ag en ts’ u tility functions are separable across co n su m p tio n an d leisure,
Rogerson (1988) shows th a t a m ark et stru c tu re in w hich individuals choose th e p ro b ability
of being em ployed ra th e r th a n actu al hours w orked will su p p o rt th e P a re to o ptim al
allocation. U nder th ese circum stances, criterion function ( 1 ) rep resen ts a reduced form
preference ordering w hich can be used to derive th e P a re to o p tim al allocation using a
fictitious social planning problem . T h e p aram eter 0 places no restric tio n s on th e elasticity
of labor supply a t th e m icro level of th e individual agent. A t th e m acro level, th e
p aram eter 0 serves only to pin dow n steady s ta te per cap ita hours w orked. F or lin ear (In or
level) solutions of th e ty p e used in th e lite ra tu re , th e value of 0 has no im p act on m odel
3
m om ents like th e v o la tility of hours w orked or th e v o la tility of o u tp u t . 2
O u tp u t, y^, is produced via th e C obb Douglas production function
(®> 7, = Atk { ~ V h t)“
where
0
< a <
1
,
7
* is th e co n stan t unconditional grow th ra te of technology, and A t is an
aggregate shock to technology w hich has th e tim e series rep resen tatio n
(4)
At = A j* 1 ex p (et ).
Here c, is a serially un co rrelated iid process w ith m ean e and s ta n d a rd erro r a , an d pa is a
t
€
scalar satisfying |p a | <
1
.
T h e aggregate resource co n strain t is given by
(5)
ct + k m
- ( l —£)kt < y t .
T he p aram eter 6, w hich governs th e depreciation rate on cap ital, is a positive scalar
satisfying
0
< 6<
1
.
To discuss th e q u a n tita tiv e im plications of th e theory, it is convenient to denote th e
m odel’s s tru c tu ra l p aram eters by th e v ector 'k. G iven a value for 'P, it is straig h tfo rw a rd to
deduce th e m odel’s im plications for a wide v ariety of m om ents w hich m ig h t be of in terest.
F or exam ple, th e an aly st m ight be in terested in u n d erstan d in g th e m o d el’s q u a n tita tiv e
im plications for an object like th e variance of aggregate o u tp u t. E x istin g R B C studies do
th is by conditioning on a p a rticu la r value for V and th e n com pare th e m odel’s p red ictio n
for th e variance of o u tp u t w ith th e corresponding m om ent in th e d a ta . W h en R B C an aly sts
2For a review of some of the solution procedures which have been used in the RBC literature, see
Christiano (1990).
4
say th a t th e m odel accounts for A% of th e variance of o u tp u t, w hat th ey m ean is th a t their
model yields a value of A given by
(6)
A = ^ mW / ^ d,
Here th e n u m e ra to r denotes th e variance of model o u tp u t, calculated for a specific value of
ty, and th e den o m in ato r denotes th e variance of actu al US o u tp u t. T h e claim th a t
technology shocks account for m ost of the fluctuations in p ostw ar US o u tp u t corresponds to
the claim th a t A is a large num ber, w ith th e cu rren t estim ate being betw een .75 and 1.0,
depending on exactly which RB C model is used (see for exam ple H ansen (1988)).
To ev alu ate this claim , we ab stra c t, for th e m om ent, from issues like sen sitiv ity to
sm all p e rtu rb a tio n s in th e theory. As decision m akers, some obvious things we m ight w ant
to know are:
How m uch info rm atio n is th ere in th e d a ta about A?
Is our calculation of A sensitive to sm all p ertu rb atio n s in tf?
A nd ju st w hat is a sm all p ertu rb a tio n in 'P?
U n fo rtu n ately , th e existing R B C lite ra tu re does not offer m u ch help in answ ering
these questions. B asically th is is because th a t lite ra tu re m akes little use of form al
econom etric m ethods, eith er a t th e stage when model p aram eters values are selected, or at
th e stage w hen th e fully param eterized model is com pared to th e d a ta . In stead a v ariety of
inform al techniques, often referred to as "calibration" are used. Irresp ectiv e of w h at o th e r
virtues or defects calib ratio n techniques m ay possess — one lim ita tio n is clear. B y ignoring
sam pling u n c e rta in ty in th e m om ents w hich underlie th e values of
th a t are u ltim a tely
adopted, calib ratio n exercises do not lead, in any n a tu ra l w ay, to a definition of w h at a
sm all p e rtu rb a tio n in $ is. This precludes th e possibility of qu an tify in g th e sam pling
5
u n certain ty inherent in m odel predictions. C onsequently, calib ratio n exercises do not
provide any inform ation ab o u t how loudly th e d a ta speak on any given question.
In recent work Law rence C hristiano and I discuss one way to circum vent these
problem s, in a way th a t is sim ilar in sp irit to existing analyses of E B C m odels, b u t which
uses form al econom etric tools (see C hristiano and E ichenbaum (1990)). T h e basic idea is to
use a version of H ansen’s (1982) G eneralized M ethod of M om ents procedure in w hich th e
estim atio n criterion is set up so th a t, in effect, th e estim ated p aram eter values succeed in
equating m odel and sam ple first moments of the data. As it tu rn s o u t th ese values are very
sim ilar to th e values em ployed in existing RB C studies. F or exam ple, m ost R B C studies
(see for exam ple P resco tt (1986)) assum e th a t th e q u arte rly dep reciatio n ra te , 6, and th e
share of cap ital in th e aggregate production function, ( 1 —a ), equal .025 an d .36,
respectively. O ur procedure yields p o in t estim ates of .021 and .35, respectively.
T he key difference betw een th e procedures does not lie so m uch in th e point
estim ates of
R ath er th e difference is th a t, by using form al econom etrics, our procedure
allows us to tra n s la te sam pling u n certain ty ab o u t th e m om ents which define o u r e stim ato r
of $ in to sam pling u n c e rta in ty regarding
itself. T his info rm atio n leads to a n a tu ra l
definition of w hat a sm all p e rtu rb a tio n in $ is, w hich in tu rn , m akes it possible to qu antify
u n c e rta in ty about th e m odel’s second m om ent im plications. T h e net resu lt is th a t it is
possible to convey how m u ch confidence we have in statem en ts like:
T h e m odel accounts for A% of th e v ariab ility of o u tp u t.
Before rep o rtin g th e results of im plem enting th is procedure for th e m odel discussed
above I m u st digress for one m om ent and discuss th e w ay in w hich g ro w th is handled. In
practice em pirical m easures of objects like yt display m arked tre n d s, so th a t some
statio n ary inducing tran sfo rm a tio n of th e d a ta m u st be ad opted. A v a rie ty of altern ativ es
are available to th e an aly st. F o r exam ple, according to th e balan ced g ro w th m odel
6
described above, th e d a ta ought to be tren d statio n ary , w ith th e In of real variables,
excluding per cap ita hours worked, growing as a linear function of tim e. So one possibility
would be to detren d th e tim e series em erging from th e m odel as well as th e actu al d a ta
assum ing a liner tim e tre n d and calculate th e m om ents of th e linearly d etren d ed series.
A different procedure involves detrending m odel tim e series and th e d a ta using the
filter discussed in H odrick and P resco tt (1980). A lthough our point estim ates of th e vector
'P were not o b tain ed using transform ed d ata, the second m om ents results were generated
using th is tran sfo rm atio n of model tim e series and US d ata.
I
do this for th ree reasons. F irst, m any authors in th e R B C lite ra tu re report results
based on th e H odrick P resco tt (H P ) filter (see for exam ple K y d lan d an d P resco tt (1982),
H ansen (1985), P resco tt (1986), K ydland and P resco tt (1988) and B ackus, K ehoe and
K ydland (1989)). In order to ev alu ate th e ir claim s, it seems d esirable to m inim ize th e
differences betw een our procedures. Second, th e HP filter is in fact a sta tio n a ry inducing
tran sfo rm atio n for tren d sta tio n a ry processes (see King and R ebelo (1988)). So th e re is
nothing logically wrong w ith using H P tran sfo rm ed d ata. Using it ju s t am o u n ts to th e
assertion th a t you find a p a rtic u la r set of second m om ents in terestin g as diagnostic devices.
A nd th ird , all of th e calculations rep o rted in this paper were also done w ith linearly
d etrended d a ta as well as grow th rates. T h e q u alitativ e results are very sim ilar, w hile th e
q u a n tita tiv e results provide even stronger evidence in favor of th e p o in ts I w ish to m ake.
So presenting results based on th e H P filter seems like an ap p ro p ria te conservative
reporting strateg y .
T he first row of T ab le 1 rep o rts results for th e baseline in d iv isib le labor m odel in
w hich th e only shocks to th e environm ents are stochastic shifts in th e aggregate p roduction
technology. Here, a
denotes th e sta n d a rd erro r of th e linearly d etren d ed Solow residuals,
crn denotes th e value of th e sta n d a rd erro r of th e In of per c a p ita hours worked g enerated
by th e estim ated m odel, w hile a
denotes th e corresponding sta n d a rd e rro r of th e In of per
J
capita output. The statistic An denotes the ratio of th e variance of th e In of per capita
7
hours w orked im plied by th e m odel to th e variance of th e In of actu al per c a p ita hours
worked in th e US. T h e v ariab le A denotes th e ratio of th e variance of th e In of per ca p ita
J
o u tp u t im plied by our m odel to th e ratio of th e variance of th e In of per c a p ita post w ar
real o u tp u t. N um bers in parentheses denote th e sta n d a rd errors of th e corresponding
statistics. All u n c e rta in ty in model statistics reflects only u n c e rta in ty regarding th e values
of th e stru c tu ra l p a ra m e te rs . 3
Tw o key features of these results deserve com m ent. F irst, th e sta n d a rd erro r of
is very large. T his is tru e even th o u g h th e sta n d a rd erro r of our p o in t e stim ate of th e
coefficient on cap ital in th e production function is very sm all (see footnote 3). Second, our
point e stim ate of A
equals 80%. This is consistent w ith claim s th a t technology shocks
explain a large percentage of th e v ariab ility in postw ar US o u tp u t. B u t notice th a t th e
sta n d a rd erro r of A is very large. T h ere is sim ply an enormous am o u n t of u n c e rta in ty
y
regarding w hat percent of th e v ariab ility of o u tp u t th e m odel accounts for. As it tu rn s o u t,
this u n certain ty alm ost com pletely reflects u n certain ty regarding th e law of m o tio n of th e
Solow residual, p
an d a , and h ard ly at all u n certain ty regarding th e values of th e o th er
p aram eters of th e m odel . 4
A different w ay to sum m arize th is u n certain ty is to consider th e g rap h of th e
confidence in terv al of A , d ep icted in figure 1. E ach point on th e g rap h is g en erated by
*
fixing A a t a specific value, A , and th e n te stin g th e hypothesis th a t
2
*
2
= A a
The
vertical axis rep o rts th e p ro b ab ility value of our te st s ta tis tic for th e corresponding v alu e of
A. T o see ju st how little info rm atio n th is m odel and th e d a ta con tain reg ard in g A, consider
th e question: W h a t values of A could we reject a t th e 5% significance level? T h e answ er is:
3The data and econometric methodology underlying these estimates are discussed in Burnside,
Eichenbaum and Rebelo (1990). Our point estimates of a, 9, 6, p&, and <J^ equal .655 (.006), 3.68 (.04),
.021 (.0003), .986 (.002) and .0089 (.02). Numbers in parentheses denote standard errors. We compute <7a
using the formula <Ta = [<7*/(l —/?!)]’**.
4See Burnside, Eichenbaum and Rebelo (1990).
8
N ot m any. Even g ran tin g th a t our algorithm breaks down w hen calculating p robability
values for negative values of A, we ought to be very com fortable believing th a t th e model
explains anywhere betw een 5% and
200
% of th e variance in per cap ita US o u tp u t.
E vidently, th e model and th e d a ta , taken together, are alm ost com pletely uninform ative
about th e role of technology shocks in generating fluctuations in US o u tp u t . 5 Decisions
based solely on th e point estim ate of A are w him sical in th e extrem e. If you th o u ght th a t
V
m onetary policy was th e key im pulse in th e business cycle — th e re is v irtu ally no evidence
here to change your m ind.
B u t w hat ab o u t th e point estim ate itself of A ? J u s t how sensitive is it to sm all
y
p ertu rb a tio n s in th e theory?
O ne in terestin g p ertu rb a tio n is to consider th e effects of lab o r hoarding on th e
analysis.
E x istin g
RBC
studies in te rp re t
all m ovem ents
in
m easured
to ta l factor
p ro d u ctiv ity as being th e resu lt of technology shocks or to a m uch sm aller ex ten t as
reflecting classical m easurem ent error in hours w orked. V arious au th o rs, ranging from
Law rence Sum m ers (1986) to R obert Lucas (1989) have conjectured th a t m any of th e
m ovem ents in th e Solow residual which are labelled as p ro d u ctiv ity shocks are actu ally an
a rtifa c t of labor hoarding ty p e phenom enon . 6 To th e ex ten t th a t th is is tru e , em pirical
work w hich identifies technology shocks w ith th e Solow resid u al will system atically
o v erstate th e ir im p o rtan ce to th e business cycle.
In fact, th e re is a su b stan tial am ount of evidence th a t th e tim e series properties of
5Our method for estimating the model’s structural parameters amounts to using an exactly identified
version of Hansen's (1982) Generalized Method of Moments procedure. Presumably the confidence
interval could be narrowed ny imposing more of the model’s restrictions, say via a maximum likelihood
estimation procedure or an over identified Generalized Method of Moments procedure. Using such
procedures would result in substantially different estimates of
thus making comparisons with the
existing RBC literature very difficult. See Christiano and Bichenbaum (1990) for a discussion of this
point.
6For a more general critique of RBC models, see McCallum (1989).
9
Solow residuals are inconsistent w ith th e notion th a t th ey represent exogenous technology
shocks. F or exam ple H all (1988) has argued th a t if Solow residuals represent exogenous
technology shocks, th e n under perfect com petition, th ey ought to be u n co rrelated w ith
different m easures of fiscal and m o n etary policy. As it tu rn s o u t th is im p licatio n is
counterfactual. E vans (1990) has p o inted th a t th e Solow residual is actu ally highly
correlated w ith different m easures of th e m oney supply. H all (1988) him self presents
evidence they are also correlated w ith th e grow th ra te of m ilita ry expenditures. In
in terp retin g his results as evidence of im perfect com petition, H all argues th a t labor
hoarding alone will not produce significant procyclical behavior in th e Solow residual, given
perfect com petition an d flexible prices.
In ongoing research, C raig B urnside, Sergio Rebelo an d I h ave trie d to assess th e
sen sitiv ity of inference based on Solow residual accounting to th e L u cas/S u m m ers critique.
T he m odel th a t we use incorporates a p articu la r ty p e of lab o r ho ard in g in to a perfect
com petition,
com plete m ark ets
RBC
m odel.
Its
purpose
is
to
d em o n strate,
in
a
q u a n tita tiv e w ay, th e frag ility of existing claim s about th e cyclical role of technology
shocks. O ur basic findings can be sum m arized as follows:
(I)
R B C m odels can, in fact, be q u ite sensitive to th e L ucas/S um m ers critiq u e. Allowing
for labor hoarding in our p articu la r m odel reduces th e ab ility of technology shocks to
account for aggregate o u tp u t flu ctu atio n s by over 50%.
(II)
W e find th a t H all’s (1988) conjecture n o tw ith stan d in g , lab o r ho ard in g w ith perfect
com petition an d com plete m ark ets, is fully capable of acco u n tin g for th e observed
correlation betw een governm ent consum ption an d th e Solow residual.
O ur m odel setu p can be described as follows. Suppose, as in th e sta n d a rd indivisible
lab o r m odel, th a t if an in dividual goes to work th e re is a fixed cost, £, d en o m in ated in
term s of hours of foregone leisure. If a person does go to w ork, he stay s th e re for a fixed
10
num ber of hours, h. T h e tim e t criterion of this person is given by
(7)
Here
ln (c f) + 0 n ( T —£ —et h).
denotes tim e t p riv ately purchased consum ption and e^ denotes th e level of tim e t
effort. T he tim e t criterio n function of a person who does not go to w ork is sim ply given by
(8 )
ln (c f) + 0ln(T).
T he aggregate p ro duction technology is given by
(9)
yt = Atkt
( 7
Nt et h) .
Here N^. denotes th e to ta l n um ber of bodies going to work at tim e t.
Proceeding as in Rogerson (1988) it is easy to show th a t, since ag en ts’ c riteria
functions are separable across consum ption and leisure, th e social p lan n er will eq u ate th e
consum ption of em ployed an d unem ployed individuals. T he P a re to o p tim al com petitive
equilibrium corresponds to th e solution of th e social planning problem
M axim ize
CO
(10)
E0 1
ln (c j) + 0 N ,ln (T -
- et h) + 0 ( l- N t )ln (T )} ,
subject to th e aggregate resource co n strain t
(11)
A j k f - V N ^ h ) 0 = cf + g, + kt+1 - (l-i)k t.
11
Here gt represents tim e t governm ent consum ption, w hich evolves according to
(12)
gt = (7t)gt?iexP(/it)’
is a serially u n co rrelated iid process w ith m ean e and sta n d a rd erro r a ^ while pg
w here
is a scalar satisfying |p g | < l . 7
If we assum e th a t th e social p lanner sees th e tim e t realizatio n of th e technology
shocks and governm ent consum ption before he chooses N^. and et , th e n th is m odel is
observationally equivalent to th e sta n d a rd indivisible labor m odel, m odified to incorporate
governm ent consum ption in to th e aggregate resource co n strain t. T h e second row of T able
1
reports th e results of incorporating g^ alone in to th e analysis . 8 W hile th e effect of th is
p ertu rb a tio n is very im p o rta n t for statistic s like th e correlation betw een hours w orked and
real wages (see C h ristian o and E ichenbaum (1990)), its effect on sta tistic s like a
or
is
m inim al.
How can we p e rtu rb th e m odel so as to cap tu re labor hoarding ty p e behavior? O ne
p articu larly sim ple w ay to do this, w hich does not change th e n o n sto ch astic stead y sta te of
th e m odel, is to ju st change th e inform ation stru c tu re facing agents w hen th e y m ake th e ir
m u st be chosen before, ra th e r th a n after,
work decisions. In p a rtic u la r, suppose th a t
tim e t governm ent consum ption and th e level of technology is know n. T o provide a bound
for th e effects of labor hoarding in this setup, we m a in ta in th e assu m p tio n th a t th e shift
length, h, is co n stan t.
T h e basic id ea underlying th is p ertu rb a tio n of th e base line m odel is th a t it is costly
for firm s to vary th e size of th e ir work force. In th e lim it it is sim ply n o t feasible to change
7See Aiyagari, Christiano and Eichenbaum (1989) for a discussion of the effects of government purchases
in the stochastic one sector growth model.
8Our point estimates of Of, 9, 6, pa, <7^, Pg, and <7g equal .655 (.006), 3.68 (.04), .021 (.0003), .986 (.027),
.0089 (.02), .979 (.021) and 0145 (.001). See Burnside, Eichenbaum and Rebelo (1990) for details.
12
employment in response to every bit of new information regarding the state of demand and
technology. One way to capture this is to consider environments where firms must make
their em ploym ent decisions conditional on th e ir views about th e fu tu re sta te of dem and
and technology, and th en ad ju st, w ithin a period of fixed tim e, to shocks along other
dim ensions. In our m odel this ad ju stm en t occurs by varying labor effort and is costly
because w orkers care ab o u t
effective hours
of work.
C onsequently labor m ust
be
com pensated for w orking harder. W e need not be precise ab o u t th e precise com pensation
schem e because th e o p tim al decentralized allocation
can be found by solving th e
ap p ro p ria te social planning problem for our model economy.
B urnside, E ichenbaum and Rebelo (1990) show th a t, in th is m odel, th e In of the
*
*
Solow residual, St , th e In of th e tru e technology shock, A t an d th e In of effort, e^, are, in
equilibrium , tied to g eth er via th e relationship
£
$
St — A j +
(13)
$
,
Here th e superscript * denotes th e deviation of th e In of a variable from its steady s ta te
*
value. T h e equilibrium law of m otion for e^ is of th e form
(14)
et = Tj k* + t 2 N* + t 3 A* + x4 g*
w here th e x ’s are n o n lin ear functions of th e stru c tu ra l p aram eters of th e m odel.
G iven our estim ates of th e stru c tu ra l p aram eters, b o th
an d
are p ositive . 9
This im plies th a t, o th er things equal, it is o ptim al to work h ard er w hen faced w ith a
positive inn o v atio n in governm ent purchases or technology, i.e. effort will be procyclical
9For this model our point estimates of a, 9, 6,
p&, G ^
Pg, and <Tg equal .655 (.006), 4.57 (.17), .021
(.0003), .981 (.027), .0062 (.02), .979 (.021) and .0145 (.001). See Burnside, Eichenbaum and Rebelo
(1990) for details.
13
C onsequently naive Solow residual accounting sy stem atically overestim ates th e level of
technology in boom s, sy stem atically u n derestim ates th e level of technology in recessions
and system atically o v erestim ates th e variance of th e tru e technology shock.
To u n d erstan d th e dynam ic properties of th e m odel, it is useful to consider the
im pulse response functions of th e system , ev alu ated a t our estim ates of th e m odel’s
s tru c tu ra l p aram eters. Excluding th e p aram eters w hich govern th e law of m o tio n of the
true technology shock, these estim ates are alm ost identical to th o se of th e stan d ard
indivisible labor m odel (see footnotes 6 and 8). Figure 2 presents th e response of th e system
to a 1% in n o v atio n in governm ent consum ption. B y assum ption em ploym ent cannot
im m ediately respond to th is shock. However, effort rises by over 15% in th e first period and
th en reverts to its stead y s ta te level. P anel (a) shows th e im plied m ovem ent in th e Solow
residual. Since effort has gone up in th e first period b u t to ta l hours of w ork h a sn ’t changed,
th e Solow residual increases by ab o u t .25%. T his is tru e even though there has been no
technology shock whatsoever. N aive Solow residual accounting falsely in te rp re ts th e increase
in average p ro d u ctiv ity to a shift in technology ra th e r th a n an exogenous increase in
governm ent consum ption. As panel (d) shows, labor p ro d u ctiv ity rises in th e first period by
.1% in response to th e 1% inn o v atio n in governm ent consum ption. Like th e m echanism s
em bedded in Lucas (1970) or H ansen and Sargent (1988), th is sim ple p e rtu rb a tio n of th e
m odel provides, a t le ast in principle, an a ltern ativ e to technology shocks as th e sole
explanation for th e procylical behavior of average p ro d u ctiv ity .
Figure 3 shows how th e system responds to a 1% in n o v atio n in technology. G iven
agents w illingness to in tertem p o rally s u b stitu te effective leisure over tim e , th e y respond to
th e shock in th e first period by increasing effort by ab o u t .4 of a p ercen t. As a resu lt th e
Solow residual rises by 1.3% in response to th e 1% technology shock. A gain naive Solow
residual accounting exaggerates th e tru e m ag n itu d e of th e technology shock.
How do th e se errors tra n s la te in to inference for A ? F ro m th e th ird row of T ab le 1
J
we see th a t th e value of a declines from .017 to .012. T his tra n sla te s in to a 50% red u ctio n
J
14
in A which falls from .82 to .41. E vidently the point estim ate of A^ is qu ite sensitive to
our p ertu rb a tio n of th e theory. Notice also th a t th e sta n d a rd error of A
is reduced
substantially , at least relative to its value in th e sta n d a rd model. B asically this reflects the
fact th a t our point estim ate of the stan d ard error of th e tru e technology shocks drops from
.053 to .032.
Figure 4 plots th e confidence intervals for A im plied by th e th ree models which I
have discussed. In all cases th e m axim al p value occurs at our p o in t estim ate of A .
Allowing for labor hoarding has tw o m ajor effects. F irst, it shifts th e whole d istrib u tio n to
th e left — this reflects th e fact th a t the point estim ate of A is now ab o u t .4 ra th e r th a n
about .8. Second, th e whole graph becomes m ore centered around th e peak. Now a t th e 5%
significance we can reject values of A th a t are less th a n 20% and those th a t exceed 80% .10
Finally before leaving my discussion of th e labor hoarding m odel — let m e point to
one m ore b it of subsidiary evidence in favor of th a t m odel relativ e to existing E B C m odels.
Suppose th a t we regress th e grow th rate of th e Solow residual on th e grow th ra te of
governm ent consum ption. A ccording to existing RB C m odels, this regression coefficient
ought to equal to zero. In fact it equals .184 and is significantly different from z ero .11
Interesting ly , our labor hoarding model im plies th a t th e p ro b ab ility lim it of th is regression
coefficient is .104 w ith sta n d a rd error of .026.12 T aking sam pling u n c e rta in ty in to account
one cannot reject, a t conventional significance levels, th e view th a t th e m odel fully
succeeds in accounting for th e observed correlation betw een th e Solow residual and
governm ent co n su m p tio n .13 S tan d a rd RB C models obviously c a n n o t.14
10In part this increase in precision reflects the fact that we must impose more of the model’s structure in
order to disentangle changes in work effort from technology shocks.
llrrhe standard error of this regression coefficient is .076.
13This standard error reflects sampling uncertainty on our estimates of the model’s structural parameters.
13Hall (1989) argues that time varying effort is not a plausible explanation for explaining this correlation.
To argue this, he first calculates the growth rate of effective labor input required to explain all of the
observed movements in total factor productivity. From this measure he subtracts the growth rate of
actual hours work to generate a time series on the growth rate in work effort. He argues that the implied
movements in work effort are implausibly large. This calculation does not apply to our analysis because it
15
A fter all is said an d done, w hat is m y answ er to th e question ad v ertised in th e title
to this talk: "R eal Business Cycle Analysis: W isdom or W him sy?". M y answ er is — both.
On th e w him sy side, I have tried to convince you th a t th e su b sta n tiv e claim s in this
lite ra tu re regarding th e cyclical role of technology shocks are exceedingly fragile. Decisions
based on those claim s ought to be viewed as whimsical.
O n th e wisdom side we have learned th a t dynam ic sto ch astic general equilibrium
models can be used to successfully organize our th o u g h ts ab o u t th e business cycle in a
q u a n tita tiv e way. W e have learned th a t technology shocks play some role in th e business
cycle. B u t we have not learned ju st how large th a t role is. Finally, to its g reat credit, work
on q u a n tita tiv e R eal Business Cycle m odels has rem inded us th a t em p irical work whose
sole purpose is to answ er th e question: "Is th e M odel T ru e" is n ot likely to very useful. O f
course th e m odel is not tru e. T h a t m uch should have been obvious before we sta rte d . And
it has been obvious to th eo rists all along. To ta k e an obvious exam ple — nobody objects to
Lucas’ (1972) m odel of th e Phillips curve because old people aren ’t ran d o m ly w hisked aw ay
in th e m iddle of th e night via unobservable helicopters. A form al s ta tis tic a l te st w hich
rejected th e m odel because of th a t fact w ouldn’t be very useful or change an y b o d y’s m ind
about anything.
Convincing stru c tu ra l em pirical work ought to address th e question: Does th e model
succeed quantitatively in accounting for those features of th e d a ta it w as designed to shed
light on. B u t good em pirical work also ought to tell us ju st how lo udly th e d a ta speak in
favor of a given hypothesis. A nd ju st as im p o rta n tly it also ought to help us u n d erstan d —
a t w hat cost d id we succeed? W h a t didn't we explain? W h at steps ap p ea r to be th e m ost 4
presumes that there are 710 shocks to productivity, an assumption which is clearly at variance with our
model.
l4Burnside, Rebelo and I are currently pursuing our labor hoarding model to see whether it is
quantitatively consistent with (a) the fact that average productivity leads the cycle, i.e. average
productivity is positively correlated with future output and hours worked, and (b) the fact that average
productivity tends to fall at the end of expansions (see Gordon (1979)). McCallum (1989) points out that
existing RBC models fail to account for the dynamic correlations between average productivity and
output.
16
prom ising in th e in ev itab le and ongoing in teractio n betw een d a ta and th eo ry ?
I
conclude by try in g to draw one final lesson from th e w ay in w hich th eorists
proceed. T heorists are often to ld to be leery of econom etricians bearing free p aram eters.
T hey already know th a t th ey ought to be leery of q u alitativ e conclusions w hich em erge
only under highly specialized assum ptions. T heir response to th is problem is to engage in
theoretical fragility analyses. Indeed, R obert Lucas’ (1989) p ap er on "T he Effects of
M onetary Shocks W hen P rices A re Set in A dvance" provides an excellent exam ple of th is
ty p e of analysis. In m o tiv atin g his paper Lucas w rites:
Models of monetary economies necessarily depend on the assumed conventions about the way in
which business is conducted in the absence of complete markets, about who does what, when,
and what information he has when he does it. Such conventions are necessarily highly specific,
relative to the enormous variety of trading practices we observe. Do the various rigid price
models have enough in common to have useful empirical or policy implications, or does
everything hinge on the accuracy of assumptions in constructing each specific example?
Lucas concludes th a t th e su b stan tiv e im plications w hich em erge from this class of models
are, in fact, q u ite ro b u st. W h eth er one agrees or not is not germ ane to th is ta lk . W h a t is
germ ane is th e effort to address th e question.
U nfortunately , despite som e im p o rta n t exceptions, n o tab ly E d L earner’s (1984)
work on in te rn a tio n a l tra d e and recent work by H ansen, Sargent and R oberds (1990) on th e
tim e series im plications of m artin g ale m odels of consum ption an d tax es, i t ’s h a rd to th in k
of m any analog exam ples in th e em pirical lite ra tu re . T h e tim e has com e for m ore
em piricists to follow su it. A bsent a g reater willingness to engage in em pirical fragility
analysis, s tru c tu ra l em p irical work will sim ply cease to be relevant. W e m ay continue to
publish, b u t our influence will surely perish.
17
References
B urnside, C raig, R ebelo, Sergio T . an d E ichenbaum , M ., "L abor H oarding and th e Business
Cycle," 1990, m an u scrip t, N orthw estern U niversity.
B lanchard, O livier J ., "A T rad itio n a l In te rp re ta tio n of M acroeconom ic F lu ctu atio n s,"
American Economic Review, 1989, 79, 1146—1164.
B ackus, D avid K ., K ehoe, P a tric k J . an d K ydland, F in n E ., "In te rn a tio n a l T rad e and
Business C ycles," 1989, Federal Reserve B ank of M inneapolis W orking P ap er 425.
C hristiano, L aw rence J . an d M artin E ichenbaum , "C u rren t R eal Business Cycle Theories
and A ggregate L abor M arket F lu ctu atio n s", m an u scrip t, N o rth w estern U niversity,
1990.
C h ristian o , L aw rence J ., Solving th e S tochastic G row th M odel by L inear—Q u ad ratic
A ppro x im atio n and by V alue F u n ctio n A pproxim ation, 1990, 8, 23—26.
E vans, C harles L .,"P ro d u c tiv ity Shocks and R eal Business Cycles," 1990, m an u scrip t,
U niversity of S o u th C arolina.
G ordon, R o b ert J ., "T h e E n d —of—E xpansion P henom enon in S h o rt-R u n P ro d u c tiv ity
B ehavior," Brookings Papers on Economic Activity, 1979 No. 2, pp. 447—61.
H all, R o b ert E ., "T h e R elation betw een P rice and M arginal C ost in U.S. In d u stry ,"
Journal of Political Economy, O ctober 1988, 96, 921-47.
H all, R o b ert E ., "In v arian ce P ro p erties of th e Solow R esidual", 1989, N B E R W orking
P ap er No. 3034.
H ansen, G ary D ., "Indivisible L abor an d th e Business C ycle," Journal of Monetary
Economics, N ovem ber 1985, 16, 309—28.
H ansen, G ary D ., "T ech n ical Progress an d A ggregate F lu ctu atio n s," 1988, m a n u scrip t,
U niversity of C alifornia, Los Angeles.
H ansen, G ary D . a n d S argent, T hom as J ., "S traig h t T im e and O ver—T im e in
E qu ilib riu m ," Journal of Monetary Economics, M arch /M ay 1988, 21, 281—308
H ansen, Lars P ., S arg en t, T hom as J . an d R oberds, W illiam , "Im p licatio n s of E x p ected
P resen t V alue B udget B alance: A pplication to P o stw ar D a ta ," 1990, m a n u scrip t,
U niversity of Chicago.
H ansen, L ars P ., "L arge Sam ple P ro p erties of G eneralized M ethod of M om ents
E stim ato rs," Econometrica, 1982, 50, 1029-54.
H odrick, R o b ert J . a n d P re s c o tt, E d w ard C ., " P o s t- W a r U.S. B usiness Cycles: A n
E m p irical In v estig atio n , m an u scrip t, C arnegie—M ellon U n iv ersity , 1980.
K ing, R o b ert G . a n d R ebelo, Sergio T ., "Low Frequency F ilterin g an d R eal B usiness
C ycles," m a n u scrip t, U n iversity of R ochester, 1988.
18
K ydland, F in n E . an d P re sc o tt, E dw ard C., "T im e to B uild and A ggregate F lu ctu atio n s,"
Econometrica, N ovem ber 1982, 50, 1345—70.
K ydland, F in n E. an d P re sc o tt, E dw ard C., "T h e W ork W eek of C ap ital and Its Cyclical
Im plications," Journal of Monetary Economics, M arch /M ay 1988, 343—60.
K ydland, F in n E. an d P re sc o tt, E dw ard C., "H ours and E m ploym ent V ariatio n in Business
Cycle T heory," A ugust 1989, In s titu te for E m pirical Econom ics, D iscussion P ap er
No. 17.
Learner, E .E ., "L e t’s T ak e th e Con O ut of Econom etrics," American Economic Review,
M arch 1983, 73, 31-13.
Learner, E .E ., "Sources of In tern a tio n a l C o m p arativ e A dvantage: T heory and Evidence,"
1984, M IT P ress, C am bridge MA.
Lucas, R o b ert E. J r., "C ap acity , O vertim e, and E m pirical P ro d u ctio n F u n ctio n s,"
American
Economic Review, Papers and Proceedings, 1970, 6, 1345—1371.
Lucas, R o b ert E. J r., "E x p ectatio n s and th e N eu trality of M oney," Journal of Economic
Theory, 1972, 4, 103-124.
Lucas, R o b ert E. J r ., "T h e Effects of M onetary Shocks W hen P rices A re Set In A dvance,"
N ovem ber 1989, m an u scrip t, U niversity of Chicago.
M cC allum , B e n n e tt T ., "R eal Business Cycle M odels," in M odern B usiness Cycle T heory,
ed. by R obert J. B arro, 1989, 16—50, H arv ard U niversity Press.
P re sc o tt, E d w ard , C ., "T h eo ry A head of Business Cycle M easu rem en t," F ederal R eserve
B ank of M inneapolis Quarterly Review, F all 1986, 10, 9—22.
R ogerson, R ., "Indivisible Labor, L otteries and E quilibrium ," Journal of Monetary
Economics, J a n u a ry 1988, 21, 3—17.
Sum m ers, Law rence J ., "Som e Skeptical O bservations on R eal Business C ycle
T h eo ry ,"F ed eral Reserve B ank of M inneapolis Quarterly Review, F all 1986, 10,
23-27.
19
TABLE 1
SUM M A RY S T A T IS T IC S
a
n
n
A
y
S t an d ard
M o d el
(C o n stan t
G overnm ent)
.053
(.0 4 6 )
.012
(.005)
.017
(.0 0 7 )
.52
(.4 7 )
.78
(.6 4 )
S ta n d a rd
M o d el
(V ariable
G overnm ent)
.053
(.0 4 6 )
.013
(.0 0 5 )
.017
(.0 0 7 )
.64
(.4 7 )
.82
(.6 4 )
L abor
H o ard in g
.032
(.0 3 1 )
.0 12
.51
.41
( . 11)
US
D a ta
.012
(.0 0 2 )
(. 002)
.016
.019
(. 002)
(. 002)
1
(. 21 )
-value
Standard Model (Constant Government)
V a r ia n c e o f Y (M o d el) /
V a r ia n c e o f Y (D a ta )
SHOCK TO TECHNOLOGY
tim e
tim e
SHOCK TO GOVERNMENT
t im e
tim e
-value
V a r ia n c e o f Y (M o d el) /
V a r ia n c e o f Y (D a ta )
Federal Reserve Bank of Chicago
RESEARCH STAFF MEMORANDA, WORKING PAPERS AND STAFF STUDIES
The following lists papers developed in recent years by the Bank’s research staff. Copies of those
materials that are currently available can be obtained by contacting the Public Information Center
(312) 322-5111.
Working Paper Series—A series of research studies on regional economic issues relating to the Sev
enth Federal Reserve District, and on financial and economic topics.
Regional Economic Issues
*WP-82-l
Donna Craig Vandenbrink
“The Effects of Usury Ceilings:
the Economic Evidence,” 1982
David R. Allardice
“Small Issue Industrial Revenue Bond
Financing in the Seventh Federal
Reserve District,” 1982
WP-83-1
William A. Testa
“Natural Gas Policy and the Midwest
Region,” 1983
WP-86-1
Diane F. Siegel
William A. Testa
“Taxation of Public Utilities Sales:
State Practices and the Illinois Experience”
WP-87-1
Alenka S. Giese
William A. Testa
“Measuring Regional High Tech
Activity with Occupational Data”
WP-87-2
Robert H. Schnorbus
Philip R. Israilevich
“Alternative Approaches to Analysis of
Total Factor Productivity at the
Plant Level”
WP-87-3
Alenka S. Giese
William A. Testa
“Industrial R&D An Analysis of the
Chicago Area”
WP-89-1
William A. Testa
“Metro Area Growth from 1976 to 1985:
Theory and Evidence”
WP-89-2
William A. Testa
Natalie A. Davila
“Unemployment Insurance: A State
Economic Development Perspective”
WP-89-3
Alenka S. Giese
“A Window of Opportunity Opens for
Regional Economic Analysis: BEA Release
Gross State Product Data”
WP-89-4
Philip R. Israilevich
William A. Testa
“Determining Manufacturing Output
for States and Regions”
WP-89-5
Alenka S.Geise
“The Opening of Midwest Manufacturing
to Foreign Companies: The Influx of
Foreign Direct Investment”
WP-89-6
Alenka S. Giese
Robert H. Schnorbus
“A New Approach to Regional Capital Stock
Estimation: Measurement and
Performance”
**WP-82-2
*Limited quantity available.
**Out of print.
Working Paper Series
( c o n t'd )
WP-89-7
William A. Testa
“Why has Illinois Manufacturing Fallen
Behind the Region?”
WP-89-8
Alenka S. Giese
William A. Testa
“Regional Specialization and Technology
in Manufacturing”
WP-89-9
Christopher Erceg
Philip R. Israilevich
Robert H. Schnorbus
“Theory and Evidence of Two Competitive
Price Mechanisms for Steel”
WP-89-10
David R. Allardice
William A. Testa
“Regional Energy Costs and Business
Siting Decisions: An Illinois Perspective”
WP-89-21
William A. Testa
“Manufacturing’s Changeover to Services
in the Great Lakes Economy”
WP-90-1
P.R. Israilevich
“Construction of Input-Output Coefficients
with Flexible Functional Forms”
WP-90-4
Douglas D. Evanoff
Philip R. Israilevich
“Regional Regulatory Effects on
Bank Efficiency”
WP-90-5
Geoffrey J.D. Hewings
“Regional Growth and Development Theory:
Summary and Evaluation”
WP-90-6
Michael Kendix
“Institutional Rigidities as Barriers to Regional
Growth: A Midwest Perspective”
Issues in Financial Regulation
WP-89-11
Douglas D. Evanoff
Philip R. Israilevich
Randall C. Merris
“Technical Change, Regulation, and Economies
of Scale for Large Commercial Banks:
An Application of a Modified Version
of Shepard’s Lemma”
WP-89-12
Douglas D. Evanoff
“Reserve Account Management Behavior:
Impact of the Reserve Accounting Scheme
and Carry Forward Provision”
WP-89-14
George G. Kaufman
“Are Some Banks too Large to Fail?
Myth and Reality”
WP-89-16
Ramon P. De Gennaro
James T. Moser
“Variability and Stationarity of Term
Premia”
WP-89-17
Thomas Mondschean
“A Model of Borrowing and Lending
with Fixed and Variable Interest Rates”
WP-89-18
Charles W. Calomiris
“Do "Vulnerable" Economies Need Deposit
Insurance?: Lessons from the U.S.
Agricultural Boom and Bust of the 1920s”
^Limited quantity available.
**Out of print.
3
Working Paper Series
(c o n t'd )
WP-89-23
George G. Kaufman
“The Savings and Loan Rescue of 1989:
Causes and Perspective”
WP-89-24
Elijah Brewer III
“The Impact of Deposit Insurance on S&L
Shareholders’ Risk/Return Trade-offs”
WP-90-12
Herbert L. Baer
Douglas D. Evanoff
“Payments System Risk Issues on A Global
Economy”
Macro Economic Issues
WP-89-13
David A. Aschauer
“Back of the G-7 Pack: Public Investment and
Productivity Growth in the Group of Seven”
WP-89-15
Kenneth N. Kuttner
“Monetary and Non-Monetary Sources
of Inflation: An Error Correction Analysis”
WP-89-19
Ellen R. Rissman
“Trade Policy and Union Wage Dynamics”
WP-89-20
Bruce C. Petersen
William A. Strauss
“Investment Cyclicality in Manufacturing
Industries”
WP-89-22
Prakash Loungani
Richard Rogerson
Yang-Hoon Sonn
“Labor Mobility, Unemployment and
Sectoral Shifts: Evidence from
Micro Data”
WP-90-2
Lawrence J. Christiano
Martin Eichenbaum
“Unit Roots in Real GNP: Do We Know,
and Do We Care?”
WP-90-3
Steven Strongin
Vefa Tarhan
“Money Supply Announcements and the Market’s
Perception of Federal Reserve Policy”
WP-90-7
Prakash Loungani
Mark Rush
“Sectoral Shifts in Interwar Britain”
WP-90-8
Kenneth N. Kuttner
“Money, Output, and Inflation: Testing
The P-Star Restrictions”
WP-90-9
Lawrence J. Christiano
Martin Eichenbaum
“Current Real Business Cycle Theories
and Aggregate Labor Market Fluctuations”
WP-90-10
S. Rao Aiyagari
Lawrence J. Christiano
Martin Eichenbaum
“The Output, Employment, And Interest Rate
Effects of Government Consumption”
WP-90-11
Benjamin M. Friedman
Kenneth N. Kuttner
“Money, Income, Prices and Interest Rates
After The 1980s”
WP-90-13
Martin Eichenbaum
“Real Business Cycle Theory: Wisdom or
Whimsy?”
*Limited quantity available.
**Out of print.
4
Staff Memoranda—A series of research papers in draft form prepared by members of the Research
Department and distributed to the academic community for review and comment. (Series discon
tinued in December, 1988. Later works appear in working paper series).
**SM-81-2
George G. Kaufman
“Impact of Deregulation on the Mortgage
Market,’’ 1981
**SM-81-3
Alan K. Reichert
“An Examination of the Conceptual Issues
Involved in Developing Credit Scoring Models
in the Consumer Lending Field,” 1981
Robert D. Laurent
“A Critique of the Federal Reserve’s New
Operating Procedure,” 1981
George G. Kaufman
“Banking as a Line of Commerce: The Changing
Competitive Environment,” 1981
SM-82-1
Harvey Rosenblum
“Deposit Strategies of Minimizing the Interest
Rate Risk Exposure of S&Ls,” 1982
♦SM-82-2
George Kaufman
Larry Mote
Harvey Rosenblum
“Implications of Deregulation for Product
Lines and Geographical Markets of Financial
Instititions,” 1982
*SM-82-3
George G. Kaufman
“The Fed’s Post-October 1979 Technical
Operating Procedures: Reduced Ability
to Control Money,” 1982
SM-83-1
John J. Di Clemente
“The Meeting of Passion and Intellect:
A History of the term ‘Bank’ in the
Bank Holding Company Act,” 1983
SM-83-2
Robert D. Laurent
“Comparing Alternative Replacements for
Lagged Reserves: Why Settle for a Poor
Third Best?” 1983
**SM-83-3
G. O. Bierwag
George G. Kaufman
“A Proposal for Federal Deposit Insurance
with Risk Sensitive Premiums,” 1983
*SM-83-4
Henry N. Goldstein
Stephen E. Haynes
“A Critical Appraisal of McKinnon’s
World Money Supply Hypothesis,” 1983
SM-83-5
George Kaufman
Larry Mote
Harvey Rosenblum
“The Future of Commercial Banks in the
Financial Services Industry,” 1983
SM-83-6
Vefa Tarhan
“Bank Reserve Adjustment Process and the
Use of Reserve Carryover Provision and
the Implications of the Proposed
Accounting Regime,” 1983
SM-83-7
John J. Di Clemente
“The Inclusion of Thrifts in Bank
Merger Analysis,” 1983
SM-84-1
Harvey Rosenblum
Christine Pavel
“Financial Services in Transition: The
Effects of Nonbank Competitors,” 1984
SM-81-4
**SM-81-5
*Limited quantity available.
**Out of print.
Staff Memoranda
(con t'd)
SM-84-2
George G. Kaufman
“The Securities Activities of Commercial
Banks,” 1984
SM-84-3
George G. Kaufman
Larry Mote
Harvey Rosenblum
“Consequences of Deregulation for
Commercial Banking”
SM-84-4
George G. Kaufman
“The Role of Traditional Mortgage Lenders
in Future Mortgage Lending: Problems
and Prospects”
SM-84-5
Robert D. Laurent
“The Problems of Monetary Control Under
Quasi-Contemporaneous Reserves”
SM-85-1
Harvey Rosenblum
M. Kathleen O’Brien
John J. Di Clemente
“On Banks, Nonbanks, and Overlapping
Markets: A Reassessment of Commercial
Banking as a Line of Commerce”
SM-85-2
Thomas G. Fischer
William H. Gram
George G. Kaufman
Larry R. Mote
“The Securities Activities of Commercial
Banks: A Legal and Economic Analysis”
SM-85-3
George G. Kaufman
“Implications of Large Bank Problems and
Insolvencies for the Banking System and
Economic Policy”
SM-85-4
Elijah Brewer, III
“The Impact of Deregulation on The True
Cost of Savings Deposits: Evidence
From Illinois and Wisconsin Savings &
Loan Association”
SM-85-5
Christine Pavel
Harvey Rosenblum
“Financial Darwinism: Nonbanks—
and Banks—Are Surviving”
SM-85-6
G. D. Koppenhaver
“Variable-Rate Loan Commitments,
Deposit Withdrawal Risk, and
Anticipatory Hedging”
SM-85-7
G. D. Koppenhaver
“A Note on Managing Deposit Flows
With Cash and Futures Market
Decisions”
SM-85-8
G. D. Koppenhaver
“Regulating Financial Intermediary
Use of Futures and Option Contracts:
Policies and Issues”
SM-85-9
Douglas D. Evanoff
“The Impact of Branch Banking
on Service Accessibility”
SM-86-1
George J. Benston
George G. Kaufman
“Risks and Failures in Banking:
Overview, History, and Evaluation”
SM-86-2
David Alan Aschauer
“The Equilibrium Approach to Fiscal
Policy”
*Limited quantity available.
**Out of print.
Staff Memoranda
(con t'd)
SM-86-3
George G. Kaufman
“Banking Risk in Historical Perspective”
SM-86-4
Elijah Brewer III
Cheng Few Lee
“The Impact of Market, Industry, and
Interest Rate Risks on Bank Stock Returns”
SM-87-1
Ellen R. Rissman
“Wage Growth and Sectoral Shifts:
New Evidence on the Stability of
the Phillips Curve”
SM-87-2
Randall C. Merris
“Testing Stock-Adjustment Specifications
and Other Restrictions on Money
Demand Equations”
SM-87-3
George G. Kaufman
“The Truth About Bank Runs”
SM-87-4
Gary D. Koppenhaver
Roger Stover
“On The Relationship Between Standby
Letters of Credit and Bank Capital”
SM-87-5
Gary D. Koppenhaver
Cheng F. Lee
“Alternative Instruments for Hedging
Inflation Risk in the Banking Industry”
SM-87-6
Gary D. Koppenhaver
“The Effects of Regulation on Bank
Participation in the Market”
SM-87-7
Vefa Tarhan
“Bank Stock Valuation: Does
Maturity Gap Matter?”
SM-87-8
David Alan Aschauer
“Finite Horizons, Intertemporal
Substitution and Fiscal Policy”
SM-87-9
Douglas D. Evanoff
Diana L. Fortier
“Reevaluation of the Structure-ConductPerformance Paradigm in Banking”
SM-87-10
David Alan Aschauer
“Net Private Investment and Public Expenditure
in the United States 1953-1984”
SM-88-1
George J. Benston
George G. Kaufman
“Risk and Solvency Regulation of
Depository Institutions: Past Policies
and Current Options”
SM-88-2
David Aschauer
“Public Spending and the Return to Capital”
SM-88-3
David Aschauer
“Is Government Spending Stimulative?”
SM-88-4
George G. Kaufman
Larry R. Mote
“Securities Activities of Commercial Banks:
The Current Economic and Legal Environment'
SM-88-5
Elijah Brewer, III
“A Note on the Relationship Between
Bank Holding Company Risks and Nonbank
Activity”
SM-8 8 -6
G. O. Bierwag
George G. Kaufman
Cynthia M. Latta
“Duration Models: A Taxonomy”
G. O. Bierwag
George G. Kaufman
“Durations of Nondefault-Free Securities”
*Limited quantity available.
**Out of print.
Staff Memoranda
(con t'd)
SM-88-7
David Aschauer
Is Public Expenditure Productive?”
SM-88-8
Elijah Brewer, III
Thomas H. Mondschean
Commercial Bank Capacity to Pay
Interest on Demand Deposits:
Evidence from Large Weekly
Reporting Banks”
SM-88-9
Abhijit V. Banerjee
Kenneth N. Kuttner
Imperfect Information and the
Permanent Income Hypothesis”
SM-88-10
David Aschauer
Does Public Capital Crowd out
Private Capital?”
SM-88-11
Ellen Rissman
Imports, Trade Policy, and
Union Wage Dynamics”
Staff Studies—A series of research studies dealing with various economic policy issues on a national
level.
SS-83-1
**SS-83-2
Harvey Rosenblum
Diane Siegel
‘Competition in Financial Services:
the Impact of Nonbank Entry,” 1983
Gillian Garcia
‘Financial Deregulation: Historical
Perspective and Impact of the Garn-St
Germain Depository Institutions Act
of 1982,” 1983
^Limited quantity available.
**Out of print.
@FedFRASER