View original document

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


A review from the
Federal Reserve Bank
III of Chicago

Liquidity effects, the monetary
transmission mechanism,
and monetary policy
Can the states solve the
health care crisis?
Call for conference papers

Call for
1993 Conference on
Bank Structure

C ontents
Liqu id ity effects, the m onetary
tran sm issio n m ech an ism ,
and m onetary p o lic y ........................................................................................... 2
Law rence J. C hristiano and
M artin Eichenbaum

Policymakers need economic models to analyze
the consequences of alternative monetary policies.
To be useful, models must at least reproduce
the consequences of simple policy actions, for
example, that interest rates drop in response to
positive money supply shocks. However, existing
general equilibrium business cycle models do not pass
this test. The authors describe a new model which
does have the property that positive money shocks
result in lower interest rates. They discuss some of
the policy implications of this model.

C an the states solve the
health care c r is is ? .............................................................................................. 15
Richard H. M atto o n

Can health care costs be controlled while insuring
universal coverage? While the policy solutions at
the federal level are elusive, state health care
experiments may be showing promise.

C a ll for co nference papers

Karl A. Scheld,

S e n io r V ice P r e s id e n t a n d

D ir e c to r o f R e s e a r c h

Editorial direction

Carolyn McMullen, e d ito r , David R. Allardice, r e g io n a l
s tu d ie s, Herbert Baer, f in a n c ia l s tr u c tu r e a n d re g u la tio n ,
Steven Strongin, m o n e ta r y p o lic y ,
Anne Weaver, a d m in is tr a tio n

Nancy Ahlstrom, ty p e s e ttin g c o o r d in a to r ,
Rita Molloy, Yvonne Peeples, ty p e s e tte r s ,
Kathleen Solotroff, g r a p h ic s c o o r d in a to r
Roger Thryselius, Thomas O’Connell,
Lynn Busby-Ward, John Dixon, g r a p h ic s
Kathryn Moran, a s s is ta n t e d ito r


Novem ber/Decem ber 1992 Volum e XVI, Issue 6
P E R S P E C T I V E S is published by
the Research Department o f the Federal Reserve
Bank of Chicago. The views expressed are the
authors’ and do not necessarily reflect the views
of the management of the Federal Reserve Bank.
Single-copy subscriptions are available free of
charge. Please send requests for single- and
multiple-copy subscriptions, back issues, and
address changes to Public Information Center,
Federal Reserve Bank o f Chicago, P.O. Box 834,
Chicago, Illinois 60690-0834, or telephone
Articles may be reprinted provided source is
credited and The Public Information Center is
provided with a copy o f the published material.

ISSN 0164-0682

Liquidity effects, the m onetary
transm ission m echanism , and
m onetary policy

Law rence J. C hristiano and
M artin Eichenbaum

Conventional wisdom holds
that an expansionary monetary
policy shock generates a per­
sistent decline in short term
interest rates and a persistent
increase in the level of employment and output.
Using different styles of analysis, Bemanke and
Blinder (1992), Christiano and Eichenbaum
(1992a,c), Eichenbaum (1992), Gali (1992),
King and Watson (1992), Sims (1992), and
Strongin (1992) provide strong empirical sup­
port in favor of the conventional view. These
findings pose an important challenge to macro­
economists. This is because existing quantita­
tive, general equilibrium business cycle models
which allow for capital accumulation are incon­
sistent with the conventional view. For exam­
ple, King (1991) and King and Watson (1992)
discuss the difficulty of generating a negative
interest rate response to positive money supply
shocks in Keynesian type models with sticky
wages and/or sticky prices. This is also the
case for real business cycle models in which
money is introduced simply by imposing cash
in advance constraints on agents (as in Lucas
[1984], Greenwood and Huffman [1987],
Cooley and Hansen [1989], or Christiano
[1991] ), or by incorporating a transactions
demand for money into the analysis (as in Kydland [1989], den Haan [1991], or Marshall
[1992] ). A generic implication of these mone­
tized real business cycle models is that, if mon­
ey growth displays positive persistence, then
unanticipated increases in the supply of money
drive interest rates up, not down. This is be­
cause, in these models, money shocks affect


interest rates exclusively through an anticipated
inflation effect. So, to drive interest rates
down, a positive shock to the supply of money
would have to signal less inflation in the future.
But to obtain this result one must make grossly
counterfactual assumptions regarding the law of
motion for the money supply. Specifically, one
would have to assume that the growth rate of
money displayed substantial negative serial
In our opinion, any convincing explanation
of the empirical facts will involve business
cycle models in which money supply shocks
generate significant, persistent liquidity effects.
Recently, a number of researchers have made
progress in constructing such models. For
convenience we refer to these models as liquid­
ity effect models. Specifically, Lucas (1990),
Christiano (1991), Christiano and Eichenbaum
(1992a), Fuerst (1992a), Grilli and Roubini
(1992), and Schlagenhauf and Wrase (1992)
have constructed general equilibrium models
in which purely transitory liquidity effects
arise. In these models, the liquidity effect can
dominate the initial expected inflation effect
associated with a change in the growth rate of
money. Under these circumstances, the con­
temporaneous effect of an unanticipated in­
crease in the money supply is a fall in the nomiLawrence J. Christiano and M artin Eichenbaum
are professors of economics at Northwestern
University and senior consultants at the Federal
Reserve Bank of Chicago. This article is an exten
sion of their paper, "Liquidity effects and the
m onetary transm ission mechanism," A m e r ic a n
E c o n o m ic R e v ie w , Papers and Proceedings, May
1992, pp. 346-353.


nal interest rate along with an increase in em­
ployment and output.
In this article, we seek to accomplish three
objectives. First, we discuss the basic mecha­
nisms at work in these liquidity effect models.
Second, we investigate one way of generating
persistent (as opposed to purely transitory)
liquidity effects. We argue that once a simpli­
fied version of the model in Christiano and
Eichenbaum (1992a) is modified to allow for
small costs of adjusting sectoral flows of funds,
positive money supply shocks generate long
lasting, significant liquidity effects as well as
persistent increases in aggregate economic
activity. Finally, we discuss some of the policy
implications of this class of models.
The model we analyze builds on a tradition
of theoretical papers which begins with the
premise that the key to understanding the ef­
fects of money supply shocks lies in the differ­
ential impacts that such shocks have on differ­
ent agents in the economy (Grossman and
Weiss [1983], Rotemberg [1984], Woodford
[1987], and Baxter, Fisher, King, and Rouwenhorst [1990]). Following Lucas (1990) and
Fuerst (1992a), we focus on firms and financial
intermediaries as the key subset of agents who
absorb disproportionally large shares of money
supply shocks. To generate this result, we
assume that households make their nominal
consumption-savings decision before observing
the current period realization of monetary poli­
cy. This allows us to capture in a simple way
the notion that firms and financial intermediar­
ies respond relatively more quickly than house­
holds do to movements in asset prices induced
by open market operations.
Consistent with the fact that actual open
market operations involve the financial sector
of the economy, we suppose that cash injec­
tions go to financial intermediaries. These
intermediaries are assumed to be in constant
contact with goods producing firms who need
working capital, that is, cash, to fund their
ongoing operations. As long as the nominal
interest rate is positive, financial intermediaries
will lend out all of the cash at their disposal.
When a positive money supply shock occurs,
households are out of the picture, at least in the
short run. This means that firms must absorb a
disproportionally large share of unanticipated
cash injections. To induce firms to do so vol­
untarily, the interest rate must fall. The down­



ward pressure on interest rates continues as long
as an unusually large percentage of the econo­
my’s cash flows through the financial sector.
The same frictions in agents’ environments
that give rise to a liquidity effect also imply that
constant growth rate rules for the money supply,
of the type advocated by Friedman (1968), will
not be optimal. This is because, in our model, it
is less costly for the monetary authority to direct
cash to financial intermediaries (and ultimately
to firms) via open market operations than it is
for private agents to do so via adjustments in
their nominal consumption-savings decisions.
So it can be welfare improving for the monetary
authority to accommodate various shocks which
impact on agents’ environments. To make this
point concrete, we analyze the response of our
model economy to technology shocks which
affect the marginal productivity of labor and
capital. A key result is that, absent monetary
accommodation, contemporaneous aggregate
employment does not increase in response to a
positive technology shock. The problem is that
the extra working capital necessary to fund an
increase in employment is simply not forthcom­
ing sufficiently quickly from the household
sector. Without a change in the money supply,
interest rates rise dramatically and valuable
social opportunities are wasted. As an alterna­
tive to inaction, the monetary authority could
pursue a version of the Real Bills Doctrine in
which the money supply is increased in re­
sponse to unanticipated improvements in real
production opportunities. In our model, when
such a policy is pursued, contemporaneous
employment and output do increase in response
to favorable technology shocks. Transitory
opportunities do not go unexploited.
The previous finding is suggestive along on
a number of dimensions. First, the perspective
on monetary policy provided by our version of
the Real Bills Doctrine captures the spirit of the
Federal Reserve Act which directs the central
bank to “... furnish an elastic currency, to af­
ford means of rediscounting commercial pa­
per...” (Board of Governors [1988]). Second, in
our example, accommodative monetary policy
has the effect of smoothing nominal interest
rates. Thus, it provides a possible rationale for
interest rate smoothing rules of the sort alleged­
ly pursued by the Federal Reserve in much of
the post war era (Goodfriend, [1991]). Third,
with costs to adjusting sectoral flows of funds,


even fully anticipated changes in the money
supply (say, a period in advance) generate
liquidity effects. Because of this, more fully
developed versions of the model could perhaps
rationalize seasonal smoothing of interest rates
by the Federal Reserve of the sort documented
in Mankiw and Miron (1991).
A sim p le m o d el w ith liq u id ity e ffe c ts

We begin by considering a simplified ver­
sion of the model in Christiano and Eichenbaum (1992a). In this model, optimizing
households, financial intermediaries, and firms
interact in perfectly competitive markets. For
now we suppose that the only source of uncer­
tainty in agents’ environments pertains to the
realization of monetary policy. Later, when we
discuss the policy implications of the model,
we also allow for shocks to the aggregate pro­
duction technology.
At the beginning of each period, the repre­
sentative household possesses the economy’s
entire beginning of period money stock, Mr
The household allocates Qt dollars to time t
purchases of the consumption good, C , and
lends the rest, A/ - Qt, to financial intermediar­
ies. In addition, the household must decide on
how much time, L , to work for firms. The
household ranks alternative streams of con­
sumption and leisure according to the expected
value of the criterion:

(1) I



where £/(C, L ) is given by,
(2) U(Ct, L ) = ( l - y )ln (C) + y In ( T - L ) .
The parameters |3 and y are scalars between
zero and one and T is the household’s endow­
ment of time.
In period t, the household faces a cash in
advance constraint on nominal consumption
(3) PC t <Q + WL.
v 7
I t
Here P and Vf denote the period t dollar price
of goods and labor, respectively. According to
(3), consumption purchases must be fully fi­
nanced with cash that comes from two sources:


Qt and wage earnings, WtL . In addition, the
household must obey its budget constraint,
(4) M

= R( M - Q ) + D + F +
(Q + W L - P C ).

The variable R denotes the gross interest rate in
period t while F and D denote period t divi­
dends received from firms and financial inter­
mediaries, respectively.
The household maximizes (1) subject to
(2), (3), and (4) by choice of contingency plans
for L , C , and Qr Throughout we assume that
the contingency plans for L and C are func­
tions of all model variables dated t and earlier.
In the basic liquidity model, the household’s
contingency plan for Qt is not allowed to be a
function of the period t realization of monetary
policy, that is, the household decides how much
money to send to the financial sector before
seeing the realization of time t monetary policy.
This assumption is intended to capture, in an
analytically convenient way, institutional and
other factors which constrain households’
choices of Qt, at least in the short run. Institu­
tional considerations include the fact that a
nontrivial fraction of A/ is held by firms and
financial intermediaries in the form of retained
earnings or pension funds and cannot be readily
allocated by households to change Q . In addi­
tion, a variety of fixed costs associated with
portfolio decisions, such as those stressed by
Akerlof (1979), render it suboptimal for house­
holds to continually readjust their nominal
consumption-savings plans. To illustrate the
impact of the assumed rigidity in Q , we also
analyze a model which abstracts from that
rigidity. Specifically, we investigate the basic
cash in advance model, where Qt is allowed to
be a function of period t monetary policy.
To simplify the analysis, we suppose that
the money supply changes via lump sum cash
injections to perfectly competitive financial
intermediaries. This means that the representa­
tive financial intermediary has two sources of
funds: cash received from the household sector,
Mt - Qr and lump sum injections of cash by the
monetary authority, X . These funds are lent
over the period in perfectly competitive mar­
kets to firms at the gross interest rate, /?. The
financial intermediary’s net cash position at the
end of the period is distributed, in the form of
dividends, to the financial intermediary’s own­


er, the household, after the consumption good
market has closed.
New goods are produced by perfectly com­
petitive firms via the production function,
(5) f ( K , L ) = AK« (z(L ) - + ( 1-6)*,,
where 0 < a < 1, 0 < 5 < 1 and A is a positive
scalar. Here K is the beginning of period t
stock of capital, 8 is the rate of depreciation on
capital, and f(Kr L ) denotes new period t out­
put plus the undepreciated part of capital. The
variable zt denotes the time t state of technolo­
gy. For now we suppose that z grows at the
constant geometric rate p > 0. Firms must
borrow working capital from financial interme­
diaries to cover their payments to labor. Loans
must be repaid to the financial intermediaries at
the end of period t. Consequently, the total
period t cost associated with hiring labor equals
Rt Wt Lt .
Firms own the stock of capital, which
evolves according to
(6) Kt+ = (1 - 5)X + / ,
where / denotes period t gross investment.
Unlike labor, capital is assumed to be a credit
good, so that firms need not borrow funds from
the financial intermediary to finance investment
activities.1 At the end of the period, after the
consumption good market closes, the firm’s net
cash holdings are distributed to its owner, the
household. The perfectly competitive firm
maximizes the expected value of its dividends
by contingency plans which specify L as a
function of model variables dated period t and
earlier, and / as a function of model variables
dated t— and earlier. This timing specification
captures the idea that employment decisions
can be revised quickly, while investment deci­
sions cannot be revised as frequently as the rate
at which open market operations are carried
out. See Christiano (1991) for the role of the
timing assumption regarding investment in
these types of models.
G e n era tin g a liq u id ity e ffe c t

The key feature of the basic liquidity mod­
el which allows it to generate a substantial
liquidity effect is the assumed rigidity in Q . It
is this assumption which prevents an increase in
the money supply from being distributed pro­


portionally among all agents. To see this, con­
sider the basic cash in advance model. To
keeps things simple, suppose that the growth
rate of money, jc = X/ Af, is an identically and
independently distributed random variable.
Under these circumstances, a positive money
supply shock is neutral: it simply results in a
proportional jump in current and future prices
and wages, leaving all other variables unaffect­
ed. The key to this result is that the nominal
expenditures of all agents respond to the money
shock in equal proportion. Among other things,
this requires that the percentage of the money
stock available to financial intermediaries,
(A/ -Q , + X )/(M; + X ), be invariant to X . But
this requires that Qt be a positive function of X .
In the frictionless world of the basic cash in
advance model this is just what happens.
Knowing that the monetary authority has in­
creased the amount of cash available to the
financial sector, the representative household
reacts by sending less cash to that sector and
more to the consumption sector.
Now if Q does not respond to X , then a
positive money shock increases the fraction of
the money supply in the hands of financial
intermediaries. As long as Rt exceeds one,
financial intermediaries lend all of the cash at
their disposal to firms. But this requires that
firms absorb a disproportionately large share of
new cash injections. For firms to do so volun­
tarily, interest rates must fall. Of course, if the
growth rate of money displays positive persis­
tence, then the expected inflation effects of a
change in the growth rate of money exert coun­
tervailing pressure on interest rates. Under
these circumstances, whether interest rates fall
or rise depends on whether the liquidity effect
or the expected inflation effect is stronger.
Suppose for the moment that the liquidity
effect dominates, so that Rt falls in response to
a positive money shock. To understand the
resulting impact on aggregate employment and
output, it is useful to think in terms of the de­
mand and supply curves for labor. A necessary
condition for the solution of the firm’s optimi­
zation problem is that the marginal cost of an
extra unit of labor equals the marginal product
of that labor. Since the firm must borrow
working capital at the gross interest rate /?. this
requires that RWJPi be equal to the marginal
product of labor. By assumption, the marginal
product of labor is a decreasing function of the


amount of labor employed. So, holding the
interest rate fixed, the demand for labor is a
decreasing function of the real wage, W7P.
This is why the demand curve for labor in Figure
1, labeled DD(P ,/f,^ ), has a negative slope.
But for a given level of W7P , the demand for
labor is a decreasing function of P . So, in Fig­
ure 1, the demand curve labelled DD(R',K ,zt),
R ' < Rt, lies farther from the origin than the
demand curve labeled DD(Rt,
K,,zt). Finally, since the marginal product of
labor is an increasing function of the stock of
capital, Kt, and the level of technology, zt, an
increase in Kt or zt also shifts the demand curve
for labor away from the origin.
A necessary condition for the solution to
the representative household’s optimization
problem is that the marginal utility of leisure
equal the marginal benefit of working: VT/P;
times the marginal utility of consumption. Con­
ditional on a fixed value of consumption, this
condition generates a static upward sloping labor
supply curve that does not directly involve /? .
In Figure 1 this labor supply curve is labelled
55(C). The equilibrium level of employment,
L \ and the real wage, (VT/P; )*, corresponding t
o an interest rate of P , is depicted by the inter­
section of the curves DD(R ,K ,zt) and 55(C).
Notice that if the monetary authority is able to
drive down the interest rate, it can shift the labor
demand curve to the right without inducing a
directly offsetting shift in the labor supply curve.
If the general equilibrium effects on consump­
tion are small, this logic suggests that unantici­


pated expansionary monetary policy disturbanc­
es which drive interest rates down generate
increases in aggregate hours worked and output
as well as the real wage rate.
Q u a n tita tiv e p ro p ertie s o f th e basic
liq u id ity m odel

To investigate the quantitative properties
of the basic liquidity model we calculated the
dynamic response of the system to a shock in
the growth rate of money. For now, we sup­
pose, as in Christiano and Eichenbaum (1992a)
that the growth rate of money, x , evolves ac­
cording to:
(7) Jt, = (l - P > + P ,

Here £wis an independent and identically dis­
tributed random variable with standard devia­
tion o , 0 < p <1, and x denotes the unconditional mean of xt. According to this specifica­
tion, the growth rate of money displays positive
serial persistence. The larger p is, the more
serial persistence there is in x .
Figure 2 displays the response of the basic
cash in advance and liquidity models to a one
standard deviation increase in jc due to a posi­
tive shock in e which occurs in period 5. All
calculations are based on values for the parame­
ters of the model equal to those used in Chris­
tiano and Eichenbaum (1992a).2 Consider first
the response of the system in the basic cash in
advance model. In the impact period of the
shock, the interest rate, R , and investment, I ,
rise, while consumption, C, falls. Consump­
tion and investment respond in
different ways because the rise in
R1acts like a tax on the cash
good, C , and a subsidy on the
credit good, / . Notice also that
time worked, L , falls. This effeet can be viewed as reflecting a
leftward shift in the labor demand
curve and a rightward shift in the
labor supply curve. The former
is induced by the rise in /? , the
latter by the fall in C . Both
shifts contribute to a fall in the
real wage, Wt /P . That L falls
reflects that the shift in the labor
demand curve dominates the shift
in the labor supply curve. With
L down and diminishing margin­
al labor productivity, the margin-



Dynamic response to money growth shock

Nominal interest rates, ft,

1.030 p-

Consumption, C ,*



Basic cash in advance model
Basic liquidity model
Adjustment cost liquidity model


"Units after removing exponential trend, exp (.004 t).
NOTE: Ticks indicate the beginning of quarters.

al cost of hiring labor, Rt Wt/Pt, must rise. Fi­
nally, since L has fallen and the stock of capital
is unchanged, current output must also fall.
With output down, and the stock of money up,
prices rise by more than the percentage change
in the money supply.
I Since 0 < pt < 1, monetary growth contin­
ues to be high relative to its steady state level.



But with the growth rate of money declining,
the inflation rate also declines toward its steady
state value. Consequently, R: declines to its
steady state value from above. With R: declin­
ing, consumption slowly rises to its steady state
value while investment declines to its steady
state level. Since a high value of /? depresses
labor demand, as long as R is high, hours


worked and the real wage stay low relative to
their steady state values and the marginal cost
of hiring labor stays high relative to its steady
state value.
All in all, it seems hard to imagine a sce­
nario more at variance with the conventional
view regarding the effect of an unanticipated
positive shock to the money supply.
In sharp contrast to the basic cash in ad­
vance model, the basic liquidity model implies
that the contemporaneous value of /? falls,
while the corresponding values of C and L rise
in response to a positive money supply shock.
The rise in L can be thought of as occurring
because the fall in /? induces a rightward shift
in the labor demand curve, while the rise in
consumption induces a leftward shift in the
labor supply curve. Both shifts contribute to a
rise in the real wage rate, W7Pf. That L rises
reflects the fact that the shift in the labor de­
mand curve dominates the shift in the labor
supply curve. With L up, and diminishing
marginal labor productivity, the marginal cost
of hiring labor, /? Wt /P , falls. Since the stock
of capital is unchanged, the rise in L implies
that output increases, which mutes the contem­
poraneous rise in the price level. Consequently,
the initial rise in the inflation rate is less than
the initial percentage increase in the money
supply. The intuition regarding the dynamic
response of the system thereafter is similar to
the basic cash in advance model.
We conclude that, at least at a qualitative
level, the basic liquidity model seems quite
promising in terms of its ability to account for
the basic facts which motivate the conventional
view of the effects of money supply shocks.
Still, the model clearly fails on one key dimen­
sion: it cannot generate persistent liquidity
effects. Because households face zero costs of
adjusting sectoral flows of funds over different
periods of time, all flows are instantly adjusted
in the period after a monetary disturbance.
This pattern of adjustment is reflected in Figure
4a (see below), which depicts the dynamic
response of QJQt , to an unanticipated shock in
the money supply. By assumption, the house­
hold cannot adjust the amount of funds it sends
to the consumption sector in the impact period
of the shock (period 5). Therefore, QJQ, ,
equals its steady state value, (1 + ) , when
t = 5. In the next period the household sharply
increases the amount of funds sent to the con­


sumption sector so that QJQlX exceeds (1 + jc).
Thereafter, QJQt , quickly returns to its steady
state value. To a first approximation, the only
period in which firms must absorb a dispropor­
tionate amount of the money stock is period 5.
And this is the only period in which there is a
quantitatively significant liquidity effect.
G e n era tin g a p e rs is te n t liq u id ity e ffe c t

One way to induce persistent liquidity
effects is to modify the environment so that the
financial sector remains more liquid than the
consumption sector for several periods after a
money supply shock. This can be done by
assuming that adjusting is costly. If, because
of these adjustment costs, households increase
Qt by a relatively small amount in the period
after the money shock, then in that period too,
financial intermediaries and firms have to ab­
sorb a disproportionately large share of the
economy’s funds. As long as this is true, li­
quidity effects persist. We show that substan­
tial persistence effects can be generated with
only very small adjustment costs.
Explicitly modeling the reasons why ad­
justing the growth rate of Qt is costly is beyond
the scope of this article. Here we simply adopt
a convenient functional form to investigate the
potential of this mechanism for generating
persistent liquidity effects. Let // denote the
amount of time agents spend on reorganizing
flows of funds. We assume that H is given by
(8) H = d T[zxp[c{QJQ^ - (1 + x))] +
exp[- c(QJQ[ X- (1 + x))] - 2}.
Figure 3 displays this function for c - 150
and d = 0.00005. Notice that // is a symmetric
function about QJQt , = (1 + jc). We refer to a
modified version of the basic liquidity model in
which leisure is defined by ( T - L - Hr) as the
adjustment cost liquidity model. The steady
states of the two models coincide because both
the level and the derivative of H with respect to
QJQt , are zero in steady state.
Figure 2 displays the dynamic response of
this model to a one standard deviation shock in
the growth rate of money. In the impact period
of the shock, the system’s response is identical
to that of the basic liquidity model. But now,
since financial intermediaries remain flush
with cash, the liquidity effect persists. The
financial sector remains relatively liquid be-


cause households persist in sending it a rela­
tively large amount of funds.
The dynamic response of Rt to a money
shock is determined by the relative strength of
the expected inflation and liquidity effects. In
the impact period of the shock, the expected
inflation effect plays no role in determining the
interest rate response.3 This is because, in the
immediate aftermath of the money shock, the
only participants in financial markets are
firms and financial intermediaries. Neither of
these agents cares about inflation when making
their money market decisions. As long as the
interest rate is positive, financial intermediaries
lend all of their funds without regard to expect­
ed inflation. In deciding how much to borrow,
firms simply equate the marginal cost of hiring
labor, R' W , with the value of the marginal
product of labor, P MPLt, where MPLj denotes
the time t marginal product of labor. This
first order condition does not involve the infla­
tion rate.
Anticipated inflation effects do play a role
in the periods after a money shock. This is
because, in these periods, households partici­
pate in financial markets and they do care about
expected inflation when making their money
market decisions. This is why the drop in Rt is
smaller in the period after the shock than it is in
the impact period. Given the assumed law of
motion for jc , the anticipated inflation effect
rapidly dissipates after period 6. But, the li­



quidity effect persists until Q has reached its
new, higher steady state growth path. This
explains the kink in the impulse response func­
tion. As households slowly adjust the growth
rate of Q , the percentage of the money stock
going to the financial sector is reduced and the
interest rate slowly climbs back to its steady
state value. The movements in the other vari­
ables of the system mirror the movements in /?
in the way suggested by our discussion of the
basic liquidity model.
The preceding results establish that, once
costs of adjusting Qt are introduced into the
analysis, money supply shocks lead to persis­
tent liquidity effects. A key remaining question
is how to assess the magnitude of the adjust­
ment costs used in the previous example. To
do this, we consider two measures. The first
measure is the actual amount of time spent by
the household adjusting Qt after a shock to the
money supply. From Figure 4b we see that the
maximal value of 7/ occurs in period 6, at
.0076 of one hour, that is, 27 seconds. So,
according to this metric, the adjustment costs
are very small.
Our second measure is based on the fol­
lowing experiment. Suppose that the represen­
tative household responded to a shock in the
supply of money as if there were no adjustment
costs, that is, as if the parameters c and d were
equal to zero. The resulting sequence of values
for Qt IQt 1would then be the same as those
emerging from the basic liquidity model. We
can measure the time spent on implementing
these changes using (8) for c = 150 and d =
0.00005. The excess of this measure of 7/ over
its value in the adjustment cost liquidity model
represents the time the household avoids wast­
ing by smoothing its adjustment to a monetary
shock. According to Figure 4c, 7/ achieves its
maximal value of one hour in the period after
the shock. Thereafter, / / is approximately
zero. So, all of the persistence in the adjust­
ment cost liquidity model is induced by the
household’s effort to avoid wasting two min­
utes a day during the quarter after the shock.
Evidently, regardless of which metric we use,
the adjustment costs in our example seem quite
We conclude that, once small adjustment
costs are introduced into the analysis, our mod­
el can generate persistent declines in the inter­
est rate following a money supply shock.



Time spent adjusting flows of funds in response to money shock




Basic liquidity model


Adjustment cost
liquidity model


J _____ L

J _____ I_____ I_____ L


S om e p o licy im p lic a tio n s

In this section we discuss some of the
welfare implications of our model. Given the
early stage of the research program, it is prema­
ture to take detailed policy prescriptions emerg­
ing from the model literally. Still, the policy
implications of the model are interesting for at
least two reasons. First, they make very clear
the nature of the frictions built into the model.
Second, they are suggestive of the general
policy principles which might emerge from
future research.
Unlike the model of the previous section,
actual economies are buffeted by a variety of
shocks which affect agents’ production oppor­
tunities and their demand for money. Holding
the growth rate of money fixed in the face of
these types of shocks (say, by adopting the k
percent money growth rule advocated by Fried­
man [ 1968]) will not be optimal. The simplest
way to show this is to modify the adjustment
cost liquidity model and allow for shocks to
technology. Until now, we assumed that the
level of technology, zt, grows at the constant
growth rate p. Suppose instead that the law of
motion for z is given by:


(9) zt = exp (pi + 0 ),
where 0 is a stationary shock to technology
which evolves according to
(10) 0, = p. 0,_, + e9

Here, 0 < pH< 1 and e0 is an independently and
identically distributed shock to 0( with standard
deviation c.0. This specification for the shock to
technology is standard in the real business cycle
literature (see Hansen [1985]).
We assume that 0 is revealed after agents
choose Qr This assumption captures the notion
that, due to a variety of unmodeled costs, house­
holds do not immediately direct cash to the fi­
nancial sector when unexpected productive op­
portunities arise in the firm sector. As before,
we assume that x is realized after agents choose
Qt, and that x evolves according to (7). Finally,
we assume (as before) that firms choose / before
observing X but after observing 0,. The timing
assumptions on / can be interpreted as reflecting
the notion that, in reality, firms have advance
information about changes in their own technol­
ogy, but not about open market operations.


It is straightforward to prove that employ­
ment does not respond contemporaneously to a
technology shock. A key factor underlying this
result is that the wage bill, V L , is independent
of the realization of 0t . This latter result refleets the fact that the quantity of dollars sup­
plied by financial intermediaries, M - Q + X t,
is by assumption independent of 0;. This does



not establish the result, however, since it leaves
open the possibility that the wage rate could
fall, thus permitting an increase in employment.
As it turns out, our specification of preferences,
(2), and the cash in advance constraint, (3), do
rule out this possibility.4
Figure 5 displays the dynamic response of
the adjustment cost model to a shock in tech-


nology that occurs in time period
5.5 In the impact period of the
shock, hours worked do not change,
while consumption, investment, the
real wage, the marginal cost of
hiring labor, /? W/ P , and the nom­
inal interest rate, R , all rise. To
understand the interest rate re­
sponse, it is useful to take as given
the result that, in equilibrium, con­
temporaneous employment does not
change. With this in mind, consid­
er Figure 6 which depicts the de­
mand for and the supply of labor.
The initial equilibrium is given by
the intersection of the demand and
supply curves, DD(R ,K ,zi) and
55(C), at L* and ( WJP )*. A posi­
tive shock to 0^ increases zt and
raises the marginal product of labor.
So, holding R: fixed, the demand curve for
labor shifts rightwards to the curve labeled
DD(R ,K ,z' ), where z ' > zr Abstracting from
shifts in labor supply and given that equilibrium
employment cannot increase, the interest rate
must increase by enough to shift the labor de­
mand curve back to where it was before the
technology shock. In practice, a positive tech­
nology shock leads to a rise in consumption
which causes the labor supply schedule to shift
left, thus mitigating the rise in the interest rate.6
In Figure 6 the new supply curve of labor is
given by 55(C') , where C ' > C .
In the period after the shock, Qt begins to
fall. The increased flow of funds from house­
holds to financial intermediaries permits an
increase in hours worked. With L rising and
diminishing marginal labor productivity, the
marginal cost of hiring labor, /? W /P , slowly
declines, while the real wage, W /P , slowly
rises as the system reverts to its (unchanged)
steady state. As a reference point, Figure 5 also
displays the dynamic response of the economy
to a technology shock in the basic cash in ad­
vance model. Notice that the employment
response in the cash in advance model is uni­
formly larger than the corresponding response
in the adjustment cost liquidity model. Evi­
dently, in the adjustment cost liquidity model,
the economy does not take full advantage of the
improved production opportunities.
This suggests that the representative house­
hold’s welfare (1) could be enhanced if the
monetary authority were to increase the money



supply in response to a positive technology
shock. To investigate this, we modified the law
of motion for jc to allow monetary policy to
respond to shocks in technology. Specifically,
we assume that xt evolves according to

jc =

(1 - “ or 7 + px jc - \, + ext + veQ.
p )x r t

When v > 0, the monetary authority accommo­
dates a positive technology shock by increasing
the money supply.
Figure 5 displays the response of the model
to a one standard deviation shock in technology
when v = 1.5. Notice that with accommodative
monetary policy, hours worked increase imme­
diately in the wake of a positive technology
shock. Notice also that the rise in the interest
rate induced by the technology shock is muted
compared to the situation in which v = 0. In
this sense, accommodative monetary policy
serves to smooth the interest rate. With the
monetary authority increasing the supply of
cash to financial intermediaries after a technol­
ogy shock, there is simply less pressure on the
interest rate.7
C onclusion

In this article, we have investigated a class
of models which is capable of accounting for
the conventional view that positive shocks to
the money supply generate persistent decreases
in short term interest rates as well as persistent
increases in hours worked and output. The
models are clearly at an early stage of develop-


ment. Still, they serve to highlight a key fric­
tion in the actual economy which we believe is
central to understanding the ability of the mon­
etary authority to affect aggregate economic
variables via open market operations. We

believe that this class of models will serve as an
important building block for future research
into the interaction of monetary policy and
aggregate economic activity.

‘We make capital a credit good in order to minimize the
impact of inflation on average employment in the model.
For a further discussion of this point, see Christiano (1991)
and Stockman (1981).
2Specifically, the parameters p, p, a, y, 8, x, a (, and pt
were set equal to (1.03)“25, 0.004, 0.36, 0.797, 0.012, 6.012,
0.014, and 0.30, respectively. Also, A = T< l), and
T = 1,369. See also Christiano and Eichenbaum (1992b).
3In particular, Christiano and Eichenbaum (1992a), foot­
note 14, show that the interest rate response to a money
shock is independent of the value of p<
4Equation (3), together with the loan market clearing
condition, (i) V L r = 4/ - Q t + X r imply (ii), P C = Mf + X .
The unitary elasticity of substitution assumption between
consumption and leisure implicit in our specification of
preferences, (2), implies that the value of consumption is
proportional to the value of leisure, that is, (iii) P C = IF (7
- L t - H ) ( 1 - y)/y. Combining (i), (ii), and (iii), we obtain

M r + X = V ( T - H ) ( 1-y)/y - (Af - 0, + X )(l-y)/y. This
equation determines W t as a function of 0 f, M , and X .
Since the latter are independent of 0f it follows that W is
also independent of 0; . From (i), it is clear that if V f is
independent of 0; , then L must also be independent of 0f.
This establishes our result.

5For p0 and oe0 we use the point estimates obtained by
Burnside, Eichenbaum, and Rebelo (1993): p_ = .9857 and
0 4

^ =- . .

The extent to which the labor supply curve shifts to the left
is minimized by our assumption that investment decisions
are made after the realization of 0(. Because / responds
positively to 0f, the resource constraint limits the extent to
which consumption can increase after a positive technology
7See Fuerst (1992a,b) for a related analysis of optimal
monetary policy.

Akerlof, George A., “Irving Fischer on his head:
the consequences of constant threshold-target
monitoring of monetary holdings,” Quarterly
Journal of Economics, Vol. 93, May 1979, pp.
Baxter, Marianne, Stephen Fisher, Robert S.
King, and K. Geert Rouwenhorst, “Monetary
transmission in an economy with capital,” manu­
script, University of Rochester, 1990.
Bernanke, Ben S., and Alan S. Blinder, “The
federal funds rate and the channels of monetary
transmission,” American Economic Review, Vol.
82, September 1992, pp. 901-921.

Christiano, Lawrence JL “Modeling the liquidi­
ty effect of a money shock,” Federal Reserve
Bank o f M inneapolis, Q u a r t e r l y R e v i e w , V ol. 15,

Winter 1991, pp. 1-34.
Christiano, Lawrence J., and Martin Eichen­
baum, “Liquidity effects, monetary policy and
the business cycle,” NBER Working Paper No.
4129, 1992a.
___________________ , “Current real business
cycle models and aggregate labor market fluctua­
tions,” American Economic Review, Vol. 82,
June 1992b, pp. 430-450.

Board of Governors of the Federal Reserve
System, Federal Reserve Act and Other Statuto­
ry Provisions Affecting the Federal Reserve
System, Washington, D.C., August 1988.

___________________ , “Identification and the
liquidity effects of a monetary shock,” in Politi­
cal Economy, Growth and Business Cycles, A.
Cukierman, L.Z. Hercowitz, and L. Leiderman,
(eds.), MIT Press, forthcoming, 1992c.

Burnside, Craig, Martin Eichenbaum, and
Sergio Rebelo, “Labor hoarding and the busi­
ness cycle,” Journal of Political Economy, forth­
coming, April 1993.

Cooley, T.F., and G. Hansen, “The inflation tax
in a real business cycle model,” American
Economic Review, Vol. 79, September 1989,
pp. 733-748.



den Haan, Wouter J., “The term structure of
interest rates in real and monetary economies,”
manuscript, University of California at San Diego,

Kydland, Finn E., “The role of money in a busi­
ness cycle model,” Discussion Paper 23, Institute
for Empirical Macroeconomics, Federal Reserve
Bank of Minneapolis, 1989.

Eichenbaum, Martin, “Comment on -Interpret­
ing the macroeconomic time series facts: the ef­
fects of monetary policy- by Christopher Sims,”
European Economic Review, Vol. 36, June 1992,
pp. 1001-1012.

Lucas, Robert E., Jr., “Money in a theory of
finance,” in Carnegie Rochester Conference on
Public Policy, Karl Brunner and Allan H. Meltzer,
(eds.), Vol. 21, North-Holland, Amsterdam, Au­
tumn 1984, pp. 9-46.

Friedman, Milton, “The role of monetary poli­
cy,” American Economic Review, Vol. 58, March
1968, pp. 1-17.

____________________ , “Liquidity and interest
rates,” Journal of Economic Theory, Vol. 50,
April 1990, pp. 237-264.

Fuerst, Timothy S., “Liquidity effects, loanable
funds, and real activity,” Journal of Monetary
Economics, Vol. 29, February 1992a, pp. 3-24.

Mankiw, N. Gregory, and Jeffrey A. Miron,
“Interest rates and the conduct of monetary poli­
cy,” in Carnegie Rochester Conference on Public
Policy, Allan H. Meltzer and Charles I. Plosser,
(eds.), Vol. 34, North-Holland, Amsterdam,
Spring 1991, pp. 41-70.

_____________________ , “Optimal monetary
policy in a cash in advance economy,” manu­
script, Northwestern University, 1992b.
Gali, Jordi, “How well does the IS-LM model fit
postwar U.S. data?,” Quarterly Journal of Eco­
nomics, Vol. CVII, May 1992, pp. 709-738.
Goodfriend, Marvin, “Interest rates and the
conduct of monetary policy,” in Carnegie Roches­
ter Conference on Public Policy, Allan H. Meltzer
and Charles I. Plosser, (eds.), Vol. 34, NorthHolland, Amsterdam, Spring 1991, pp. 7-30.
Greenwood, Jeremy, and Gregory W. Huff­
man, “A dynamic equilibrium model of inflation
and unemployment,” Journal of Monetary Eco­
nomics, Vol. 19, March 1987, pp. 203-228.
Grilli, Vittorio, and Nouriel Roubini, “Liquidity
and exchange rates,” Journal of International
Economics, Vol. 32, May 1992, pp. 339-352.
Grossman, Sanford, and Laurence Weiss, “A
transaction based model of the monetary mecha­
nism,” American Economic Review, Vol. 73,
December 1983, pp. 871-880.
Hansen, Gary D., “Indivisible labor and the
business cycle,” Journal of Monetary Economics,
Vol. 16, November 1985, pp. 309-327.
King, Robert S., “Money and business cycles,”
manuscript, University of Rochester, 1991.
King, Robert S., and Mark Watson, “Compar­
ing the fit of alternative dynamic models,” manu­
script, Northwestern University, 1992.


Marshall, D., “Inflation and asset returns in a
monetary economy with transaction costs,” Jour­
nal of Finance, forthcoming, 1992.
Rotemberg, Julio, “A monetary equilibrium mod­
el with transaction costs,” Journal of Political
Economy, Vol. 92, February 1984, pp. 40-58.
Schlagenhauf, Don E., and Jeffrey M. Wrase,
“A monetary, open economy model with capital
mobility,” Discussion Paper 67, Institute for Em­
pirical Macroeconomics, 1992.
Sims, Christopher A., “Interpreting the macroeconomic time series facts: the effects of monetary
policy,” European Economic Review, Vol. 36,
June 1992, pp. 975-1000.
Stockman, Alan, “Anticipated inflation and the
capital stock in a cash in advance economy,” Jour­
nal of Monetary Economics, Vol. 8, November
1981, pp. 387-393.
Strongin, Steven, “The identification of monetary
policy disturbances: explaining the liquidity puz­
zle,” manuscript, Federal Reserve Bank of Chica­
go, 1992.
Woodford, Michael, “Credit policy and the price
level in a cash in advance economy,” in New Ap­
proaches to Monetary Economics, William Barnett
and Kenneth J. Singleton, (eds.), Proceedings of
the Second International Symposium in Economic
Theory and Econometrics, Cambridge University
Press, 1987, pp. 52-56.

Can the states solve the
health care c ris is ?

Richard H. M atto o n

The United States spends a
larger share of gross domestic
product (GDP) on health care
than any major industrialized
nation.1 While U.S. health
expenditures as a share of GDP are estimated to
have topped 13 percent in 1991, Germany and
Japan spend 8.2 and 6.8 percent, respectively.
Even socialist Sweden, known for high expen­
diture levels on public health, spent only 9
percent. To make matters worse, this gap has
been widening. In 1980, total health care costs
as a share of GDP was 9.3 percent for the U.S.
compared to 7.1 percent, on average, for the 24
OECD nations. By 1990, U.S. expenditures
had risen to 12.4 percent of GDP while the
share for the OECD group had increased to
only 7.6 percent.2
Increased health care costs affect all sec­
tors of the U.S. economy. According to a
Washington State study, health costs consume
25 percent of the average private firm’s profits
and translates into a 3 to 5 percent surcharge on
the price of U.S. products when sold abroad.3
Health costs currently account for 15 percent of
the federal budget (up from 10 percent in 1980)
and, if unchecked, are expected to consume 28
percent by 2002, according to the Congression­
al Budget Office. Federal spending for the
Medicaid program alone is on a pace to eclipse
50 percent of all federal benefits targeted for
the poor by 1993.4 In the case of state and local
governments, rising health care costs (particu­
larly those associated with Medicaid payments)
are frequently seen as the primary culprit in
budget deficits. Given that state and local



governments cannot run explicit budget deficits
like the federal government, these rising costs
are forcing reductions in other budget areas.
Most observers agree that the U.S. cannot con­
tinue funding such robust growth in health
At the same time that costs are high and
growing, there is a substantial support for
broadening health care coverage to all citizens.
In particular, those citizens and workers who
are above but close to the poverty level often
lack adequate health care benefits. Several
options have emerged in an attempt to meet the
twin goals of cost containment and universal
access. These range from trying to inject more
market incentives into health care provision and
consumption to adopting government based
national health care insurance. Impatient for
federal action and weary of the failure of pri­
vate markets, many states are trying to craft
their own health plans.
This article discusses why health costs
have been rising at such a rapid rate in the U.S.,
and examines state initiatives aimed at address­
ing this issue.
H o w fa s t are m e d ica l costs rising?

In 1970, personal consumption expendi­
tures for medical care totaled $55 billion. By
1988 the figure had grown to $443 billion, a
nominal increase of 705 percent.5 Table 1
shows the rate of increase by type of medical
Richard H. M attoon is a regional econom ist at the
Federal Reserve Bank of Chicago. The author is
indebted to W illiam Testa and Carolyn M cMullen
for their helpful comments.



Personal consumption expenditures for
medical care by type of services
(billions of dollars)















Drugs, appliances








Net cost of health




All other medical




SOURCE: U.S. Department of Commerce, Bureau of
Economic Analysis, Survey o f Current Business,
various issues.

service. As is shown, the rate of expenditure
growth shows considerable variation depending
on the service in question.
Furthermore, as the comparison with the
Consumer Price Index (CPI) in Figure l shows,
prices for medical care commodities and servic­
es have grown much faster than the CPI for all
items in recent years. Growth rates in the two
major components of medical prices exceeded
the general inflation rate. Other statistics illus­
trate similar gains in expenditures. Figure 2
illustrates national health expenditures as a
share of GNP and personal consumption for


medical care as a share of disposable income
from 1970 to 1988. Regardless of how it is
measured, the trend is clearly for health care
expenditures to consume a growing share of the
Even more troubling is the possibility that
higher expenditure levels on health care are
not translating into significantly better health
care. Measures of U.S. public health remain
poor when compared to other developed na­
tions. For example, the U.S. ranks eighth in the
world in life expectancy, l lth in maternal mor­
tality, 18th in child mortality, and 22nd in in­
fant mortality.6
W hy are h ea lth care costs
g ro w in g so q u ickly?

The market for health care is unique in that
asymmetric information between buyer and
provider and restricted competition among
suppliers and third party payments are the rule,
not the exception. Incentives embodied in
America’s system of health care are complex
and rarely emphasize cost containment. The
Health Care Financing Administration (HCFA)
identifies three factors as affecting the growth
of personal health care expenditures:7 increases
in the prices charged for services; increases in
the population receiving medical treatment; and
increases in the intensity with which medical
services are used. In 1988, the HCFA estimat­
ed that price increases accounted for 67 percent


Health expenditures compared to GNP
and disposable income

SOURCE: U.S. Department of Health and Human
Services, HCEA; U.S. Department of Commerce, BEA;
S u rv e y o f C u rre n t B u s in es s , various years.


are shared by all policyholders,
muting the magnitude of the cost
Factors affecting growth of personal
increase to any individual.
health care expenditures
Further evidence of the po­
(percent contribution)
tentially distorting effect of insur­
1982 1983 1984 1985 1986 1987 1988
ance can be seen in the percent­
age increases in the cost of medi­
cal services. Services traditional­
ly covered by insurance, such as
hospital stays, have risen much
SOURCE: U.S. Department of Health and Human Services, Health Care
faster than health services like
Financing Administration.
eye exams which tend not to be
included in coverage. From 1970
of growth in personal health care expenditures,
to 1988, hospital service expenditures increased
followed by increases in intensity of service
by more than 800 percent while less frequently
utilization (23 percent) and population changes
or not fully covered items such as drugs and
(10 percent). However, future demographic
eyeglasses showed a gain of less than 350 per­
changes, namely the aging of baby boomers, will
cent. This evidence suggests that some of the
have a profound effect on medical spending,
observed price increases in specific health care
given that people 65 years or older spend four
procedures are related to the distorting effects
times as much on health care as younger people.
which insurance coverage has had on the de­
The pattern for these three factors is shown in
mand for medical services.
Table 2. In order to understand why prices for
Other factors influencing demand and
medical commodities and services have risen so
therefore the price of medical services include
quickly, we must look at the factors affecting
demographic and lifestyle factors and the envi­
supply of and demand for health care.8
ronment. Clearly, as life expectancy grows and
the population ages, demand for medical ser­
D e m a n d fa c to rs
vices increases. The elderly consume a signifi­
One of the primary factors in the spiraling
cant portion of health care as is illustrated by
costs of health care is the low price elasticity of
the fact that a large percentage of health care
demand for medical services. According to
expenditures are spent on the very elderly,
some estimates, the price elasticity of demand
particularly during their last years of life. For
for medical services may be as low as O.2.9
example, the U.S. spends 1 percent of GNP on
This means that the consumption of medicine
health care for elderly people in the last year of
tends to be relatively unaffected by price in­
their lives (Fuchs [1984]). Lifestyle changes
creases so that total expenditures increase over
also play a role. Society’s increasing incidence
time as prices rise.
of alcohol and drug abuse and other abusive
Health insurance often distorts the costs of
behaviors reduce the stock of health while
consuming medical care. An individual covered
increasing health care expenditures. The quali­
by an insurance plan often bears no additional
ty of the environment also plays a role. Prob­
cost for consuming additional units of medical
lems with air and water quality provide envi­
care once they have exceeded the deductible on
ronmental hazards which potentially lead to a
their policy. Consequently, at a given threshold,
greater demand for medical care.
medical care consumption becomes costless
S u p p ly fa c to r s
(other than for the individual’s time) and pre­
sumably the absence of immediate cost encour­
Productivity gains in the provision of med­
ages greater consumption, even though the costs
ical services have been slow for a variety of
are ultimately passed along to consumers in the
reasons. First, insurance and public health
form of higher premiums and/or taxes. As a
reimbursement programs have traditionally
result, consumers are relatively insensitive to
paid medical providers on a cost of service plus
price increases in medical services since the
a small profit basis. Since providers are always
penalty of higher cost medical consumption only
assured of covering their costs there is little
shows up in higher insurance premiums, which
incentive to improve productivity and lower





costs. Similarly, the use of “best practice”
techniques, in which expensive procedures with
sometimes marginal benefits to the consumer of
the service are used, also encourages potentially
wasteful uses of resources.
There is a limited supply of physicians
because, despite a large applicant pool, the
number of medical school seats is limited. This
limited supply of doctors helps to keep the costs
of their services high. Furthermore, increased
specialization among doctors has actually in­
creased the variety of potential services and has
helped create a demand for those services.
The relative abundance of medical specialists
has encouraged patients to seek specialists for
routine medical procedures which could be
treated by general practice physicians who
presumably charge lower fees. Part of the
Canadian system of health care cost control is
to restrict the number and availability of medi­
cal specialists.
Asymmetric information is another impor­
tant factor affecting the cost of health care.
The health care market is one of the few areas
where most consumers are generally unin­
formed about purchasing decisions. Because
medical information is specialized, the consum­
er often has no knowledge as to whether the
treatment prescribed for a given illness is nec­
essary. Without a third party opinion, there is
little reason for restraint in prescribing medical
treatment. Furthermore, many consumers pur­
chase medical care infrequently and conse­
quently do not know whether the price for a
particular procedure is in fact a good price.
Since price advertising is not common in the
medical profession it is very difficult for con­
sumers to develop even a casual understanding
of the costs of the system.
The emphasis on medical technology
also contributes to high costs. The U.S. is a
leader in the development and use of high
technology medical treatments. Much of this
may be due in part to the historical tendency for
insurance coverage to pay for any treatment
without regard to cost. Furthermore, in com­
parison to the Canadian system, technology is
used much more broadly. In Canada, high
technology devices such as CAT scanners
tend to be available only at specific hospitals,
while in the U.S., they are available almost


H ealth care costs in th e
S even th D is tric t

Costs and demand for medical care are not
uniform across the U.S. This lack of uniformity
has led some analysts to favor state based solu­
tions to health care cost and availability con­
cerns, fearing that a federal solution will fail to
recognize local variations in these problems.
This section examines the supply and demand
factors as well as the cost for health care in the
states comprising the Seventh District (Illinois,
Indiana, Iowa, Michigan, and Wisconsin) rela­
tive to the rest of the nation.
One indicator of the District’s cost of health
care can be found in average surgery and hospi­
tal charges. For inpatient surgery, the District’s
average 1989 cost per surgery was $936 versus a
U.S. average of $980. However the average
surgery charge within the District ranged from
$1,099 in Illinois to a low of $817 in Wisconsin
(see Figure 3). The Figure also shows a similar
pattern for outpatient surgeries. Similarly, the
average daily cost per admission for hospitals
was below the national average of $586 in Indi­
ana ($571), Iowa ($431), and Wisconsin ($483).
The costs were above the U.S. average in Illinois
($632) and Michigan ($643).1
District states were found to lag the nation
in utilization containment for health care servic­
es. For example, the national average for surgi­
cal procedures is 74 per 1,000 of population. All
of the District states, except Iowa with 61 per


1,000, had surgical rates which were signifi­
cantly higher (see Figure 4). Similarly, District
statistics for the average length of stay for a
hospital visit are slightly above the U.S. aver­
age. The national average hospital stay is 7.2
days. District states range from 6.6 days for
Indiana and 7.4 for Michigan, Wisconsin, and
Illinois, to 8.2 for Iowa." This higher utiliza­
tion may reflect the fact that health insurance
coverage in District states is generally broader
than the U.S. as a whole. For the U.S., 17.4
percent of the nonelderly population have no
health insurance, while the percentage without
coverage in the District is much lower. Wis­
consin leads the way with only 9.8 percent of
the population without health insurance, fol­
lowed by Michigan (11.9), Iowa (12.7), Illinois
(14.5), and Indiana (16.9).1 Broader coverage
may encourage more active use of medical
In the area of public expenditures for
health and hospitals, District annual per capita
public expenditures were below the U.S. aver­
age. As Figure 5 shows, only Michigan’s pub­
lic spending on health and hospitals is above
the U.S. average. Even in the difficult area of
Medicaid payments per capita, three of the five
District states spent less than the national aver­
age. The U.S. average for Medicaid payments
was $ 185 as compared to $ 151 in Illinois, $ 167
in Indiana, and $149 in Iowa. Only Michigan
and Wisconsin were above at $198 and $209



Other factors which can influence the sup­
ply of and demand for medical services include
the availability of health services (the health
care infrastructure) and health characteristics of
District citizens. With respect to health care
availability, the District is better than average
in access to hospital facilities. While the U.S.
average is 4.1 hospital beds per 1,000 of popu­
lation, District state averages are 4.5 for Illi­
nois, 4.2 for Indiana, 5.2 for Iowa, 4.0 for
Michigan, and 4.5 for Wisconsin.1 In the num­
ber of physicians relative to the population as a
whole the District fares somewhat worse. The
national average for physicians per 100,000
population is 225. All of the District states are
below this figure with Iowa showing the largest
relative deficit at 154 physicians per 100,000
population (see Figure 6). However, it should
be noted that a relatively large supply of physi­
cians does not appear to reduce medical costs.
States such as New York and California have
among the highest prices and expenditure levels
for medical services despite having more doc­
tors per 100,000 population. The concentration
of higher priced medical specialists coupled
with a demand for their services may explain
why the presence of more physicians has not
created price competition which would act to
hold down medical costs in these two states.
In terms of health characteristics of District
citizens two statistics are helpful. Demographically the District has approximately the same


State and local per capita spending
on health and hospitals (1988)

SOURCE: State Policy Research Inc.,
Table 1-20.

S ta te s in P rofile,



Number of physicians per 100,000
population (1986)

SOURCE: State Policy Research Inc., States in
Table 1-14.

P rofile,

percentage of citizens that are 65 or older as the
rest of the nation (roughly 12.5 percent). This
is important since the elderly consume a signifi­
cantly larger share of health care expenditures
than the rest of the population. Since the Dis­
trict’s share of this segment of the population is
roughly similar to the nation, health care costs
attributed to the elderly should rise at the same
rate as the national average. Table 3 shows the
mortality rates from the four leading causes of

death. As expected, the District averages are
roughly comparable to all regions except the
West, where death rates from heart disease, can­
cer, and stroke are all significantly lower.
One area where District states may be able
to make headway in controlling health care costs
is managed care. District figures on participation
in HMOs (health maintenance organizations) and
PPOs (preferred provider organizations) indicate
that District states, like much of the U.S., could
increase participation in these organizations to
control health care costs. Evidence suggests that
these types of health organizations have a greater
incentive for internal cost control which reduces
total medical costs since fees for procedures are
often fixed. Nationally, 13.2 percent of the pop­
ulation is enrolled in HMOs. In the District,
Wisconsin leads the way with nearly 22 percent
of its population in HMOs, followed by Michi­
gan (15.3 percent), Illinois (13 percent), Indiana
(7.8 percent), and Iowa (7.2 percent).1
E xp lain in g th e v a ria tio n in reg io nal
m ed ical costs

Some of the factors which explain the
variation in regional health care costs are obvi­
ous. Areas with a higher cost of living and
higher labor costs tend to have higher medical
costs. Similarly, metropolitan areas with techno­
logically advanced hospitals, concentrations of
medical specialists, and the ability to perform
advanced medical work also tend to have higher
medical costs.
However other forces are at
work. One recent study examining
regional cost differences in Medic­
Death rates—four leading causes
aid was conducted by Jane Sned­
(1987 per 100,000 residents)
don Little of the Federal Reserve
Bank of Boston.1 While Little’s
findings concern variations in state
Medicaid costs, they provide some
interesting insights into potential
2 08.3
explanations for variations in gen­
eral health costs. In her paper, the
rate of reimbursement provided by
Regional averages
Medicaid for nursing home care is
N orth e a st
found to be the most significant
S e ve n th D istrict
factor in explaining regional Med­
N orth C e n tra l*
icaid price variations. States with
high nursing home reimbursement
W e st
2 18.5
rates tended to have high Medicaid
‘ Minus Seventh District states.
costs and vice versa. However,
SOURCE: National Center for Health Statistics, Monthly Vital Statistics
while states with lower reimburse­
Report, September 26, 1989.
ment rates had lower Medicaid



costs, these savings did not appear to spill over
into personal health care costs for nonMedicaid
recipients. Practices such as cost shifting from
Medicaid to other payors (such as private insur­
ers willing to pay higher charges for medical
services) appear to be common, making personal
health care expenditures vary less from state to
state. The tendency to cost shift makes it very
difficult to know what medical services are being
provided in state to state comparisons. Each
state’s medical dollar may be buying different
levels of service and this may explain some of
the regional variation in medical costs.
Cost shifting also occurs in the health care
system as a whole when negotiated discounts are
used to control private health care costs. What
happens is that either an individual company or
group of companies negotiate specific discounts
with a particular provider. For example, a differ­
ent fee structure for services at a particular hospi­
tal might be the focus of such a discount. While
these discounts provide lower costs to the recipi­
ents of the discounts, they may lead the health
care provider to charge even higher fees to com­
panies and individuals not included in the negoti­
ated discount plan. Furthermore, the discount
can also lead to reduced access to health care for
individuals covered by the discount because the
provider has an incentive to serve full paying
patients first, rather than those with the discount.
Negotiated discounts can limit health care costs
for a segment of the population but they may not
reduce the cost of health care for society at large
unless the discount is available to everyone.
Proposals fo r re fo rm in g th e
h ea lth care system

Factors such as regional variations in the
utilization, cost, and availability of health care
have spurred a wide range of state health care
reform measures in the U.S. Virtually all of
these proposals are designed to address two is­
sues: cost containment and access to health care.
The interest in cost containment is obvious given
the escalation in health care prices. Government,
business, and individual consumers all agree that
health care costs cannot continue to rise at cur­
rent rates. There are an estimated 37 million
Americans without health insurance1 who have
less access to medical care and often are forced
to receive medical care in emergency situations,
increasing the cost of treatment. This treatment
is often uncompensated, which induces cost
shifting to privately insured patients. Conse­



quently, cost containment and access to health
care are the twin goals of most proposals.
Most reform proposals are a variation on
three general frameworks: market based re­
forms, “play or pay” proposals, or national
health insurance. All three types of proposals
include the private provision of at least some
medical services and do not favor the adoption
of an all inclusive public health care system,
such as the system in the United Kingdom,
where medical services are provided through
publicly supported facilities staffed with public
Market based reforms

Most of the so called market based reform
proposals assume that a lack of market disci­
pline explains much of the recent rise in medi­
cal costs. Since insurance often makes the
consumption of medical care relatively costless
to consumers, they have no real incentive to
seek lower prices or reduce their consumption
of medical care. If consumers were forced to
bear a greater cost for consuming medical care,
according to this view, they would consume it
more wisely and would have a greater incentive
to limit their use of medical care by adopting
healthier lifestyles. Similarly, if health care
providers found that higher prices for medical
services reduced patient demand, it is assumed
that they would have a greater incentive to
provide more cost efficient care. For example,
the Heritage Foundation has proposed making
health care a taxable benefit.1 The argument is
that because health insurance is tax exempt,
individuals purchase more health insurance (in
terms of taking on broader coverage) than they
would if they actually had to bear the full cost
of the purchase, if only from a tax perspective.
If the full cost of employer sponsored plans was
taxable, people might be more willing to opt for
lower cost managed care and HMO options or
accept coverage which more closely reflects
their lifestyle.
Generally speaking, these proposals are
also geared toward maintaining a system of
private insurance as the most efficient method
for providing health coverage for everyone. In
order to extend private health insurance to the
poor, market based proposals usually contain
voucher and tax credit options. For example,
the Bush Administration’s market based reform
proposal provides for a $3,750 voucher for a
family of four to purchase health insurance.


Opponents point out that the cost of health
insurance for a family of four will probably
exceed the voucher payment. Supporters of
these proposals believe that managed care pro­
grams now being pursued by insurers show
great promise for controlling health care costs.
Since insurers have a profit motive, they are
best positioned to monitor the appropriateness
of medical expenditures and procedures.1 Crit­
ics of market based proposals worry that medi­
cal rationing will result. While making con­
sumers bear more of the cost of consuming
medical care would probably reduce demand
for medical services, it may also lead to under­
consumption of appropriate medical care. To
avoid copayments or higher insurance premi­
ums, individuals may try to avoid consuming
medical care even when beneficial. For exam­
ple, they might avoid diagnostic screenings
where early detection of disease might prevent
more expensive treatment later.
“Play or pay” proposals

These proposals try to expand health care
coverage by requiring that employers provide a
minimal health insurance package for their
employees (play) or pay a payroll tax for a new
health care program designed to cover the unin­
sured with publicly provided health plans.
These proposals usually have guidelines that
would initially apply play or pay standards to
businesses of a certain minimum size (10 em­
ployees) and then only to employees working
more than 17.5 hours per week. Furthermore,
to reduce potential opposition, most play or pay
options include some provision for cost con­
tainment, usually through a form of public rate
setting for medical services.
Play or pay is partially designed to address
the plight of workers in industries where health
insurance coverage tends to be slight. For
example it is estimated that almost 50 percent
of retail and nearly 75 percent of hotel and
restaurant workers are not covered by health
insurance. However, critics of play or pay
point out that the additional costs of pay or play
may encourage businesses to lay off marginal
workers or at least limit their demand for new
workers. The target group of employees might
find themselves not only uninsured but unem­
ployed. Also, it is unclear how employers who
currently provide health care would react to the
play or pay option. Given that the payroll tax


may be cheaper than paying for health insur­
ance premiums, it may be that some employers
would actually choose to repeal coverage and
pay the tax, thereby straining the financial
solvency of the system.2
National health insurance

National health insurance proposals are
usually based on the Canadian health insurance
system in which the government becomes the
primary payer for all medical services. The
provision of medical services remains a private
industry but the government pays for virtually
all medical treatment for all Canadian citizens.
To pay for the system, Canada levies taxes in
lieu of private health insurance premiums.
While secondary insurance plans are available
to cover certain special costs (such as private
hospital rooms, dental services, and eyeglass­
es), almost all other costs from check-ups to
surgery are covered by national health insur­
ance. The government also becomes responsi­
ble for setting rates for medical procedures and
is therefore able to use its single-payer clout to
try to control medical costs. The system also
reduces costs by cutting the administrative
expense associated with dealing with large
numbers of private insurers. (The U.S. has an
estimated 1,500 private insurance companies).
The Canadian system is also designed to
reflect differing regional preferences for medi­
cal care. The central government pays a mini­
mum of 40 percent of each provinces’ health
tab but the remainder is allocated by each prov­
ince. In designing a health care program, each
province must address five national objectives.
These are: 1) the plan must cover all citizens;
2) all medical and basic hospital services must
be included; 3) no user fees or other special
billing fees can be used to limit access; 4) the
plan must be transferable with no change in
coverage if a plan recipient changes jobs; and
5) it must be publicly administered.2
The biggest objection to national health
insurance plans concerns de facto access to
health care. The government sets the reim­
bursement rates for specified procedures only
and establishes the capital spending for hospi­
tals. As a result, highly sophisticated equip­
ment tends to be found in only a handful of
university hospitals. For example, while the
U.S. has nearly 2,000 hospitals with magnetic
resonance imaging machines (MRIs), Canada


has only 15.2 Thus, Canadian hospitals often
have to schedule procedures to reflect the avail­
ability of equipment. This in turn causes long
waits for routine but necessary surgery. The
U.S. General Accounting Office (GAO) esti­
mates that in Canada, six month waits for CAT
scans and one year waits for orthopedic work
such as hip replacements are common. In con­
trast, privately insured Americans are largely
accustomed to a system of medical care on
demand. The benefit of the Canadian system is
evident in lower fees and average per capita
health costs. However, there is some dispute
whether the lower per capita health costs are
the product of the health insurance system or
simply reflect the fact that proportionately
fewer Canadians are either poor or old than is
the case in the U.S. These are the most expen­
sive patients to treat.2
U .S . s ta te based plans

States have grown impatient waiting for
the federal government to devise a national
health care plan. Pressured by constituents and
the growing contribution of health care costs to
state budget problems, more than two dozen
states have passed some form of health care
reform. Some of these programs are sweeping
in scope while others intend to address more
limited areas such as reducing the cost of pub­
licly provided health programs or increasing
health care coverage. The following examples
illustrate the variety of the proposals being
developed at the state level.

Iowa has shown increasing interest in play
or pay reforms to address health care cost and
availability issues. Massachusetts passed a
similar measure in 1988 but the implementation
of the plan has been delayed until 1995 by a
state budget crisis and a severe economic
downturn.2 The Iowa proposal has been de­
signed by a consortium of hospitals, businesses,
labor unions, and insurers. As such it has
broader based support than most reform efforts.
The plan would require all Iowa residents to
have health insurance and would extend cover­
age in two ways. Employers would either have
to provide health insurance for their employees
or pay a 5 percent payroll tax. The payroll tax
would be used to provide health insurance
subsidies on a sliding scale for those uninsured
that have incomes up to 250 percent of the
federal poverty level.



Under the Iowa proposal, cost containment
is achieved in a variety of ways. First, empha­
sis would be placed on using managed care
methods designed to limit the choice of medical
providers for consumers. In addition, while
insurance would still be provided by private
insurance companies, reimbursement rates from
insurers to medical providers would be stan­
dardized. Furthermore, growth in the cost of
health insurance would be limited by establish­
ing a ceiling on the percentage of health insur­
ance premiums that could be claimed as profit
and overhead by the insurer, presumably en­
couraging insurers to reduce overhead in order
to increase profits.
Opposition to Iowa’s play or pay scheme
has been voiced by several parties. The first
source of friction is small business. For those
small businesses which traditionally have not
provided health insurance to their workers, the
play or pay program immediately increases
costs. These Iowa businesses usually claim
they will be less competitive if they face an
additional cost of doing business which small
businesses in neighboring states will not face.
For large companies with comprehensive medi­
cal plans, there is the fear that these businesses
will drop medical coverage (estimated to cost
13 percent of payroll costs) in favor of the
lower priced 5 percent payroll tax. Such an
incentive would lower health care expense for
firms with medical plans but it would also have
the unintended effect of expanding the share of
the population needing to receive insurance
through the state’s health insurance pool which
would be created through the new payroll tax.
In response to this, some have suggested plac­
ing up to a 10 percent payroll tax on larger
companies in an effort to discourage such de­
fections.2 However, these fears may be un­
founded or at least limited in any case. Em­
ployees would certainly object to any unilateral
withdrawal of health benefits, and multistate
companies would be hard pressed to offer med­
ical benefits which differed so drastically from
one location to another.

Oregon has come up with an innovative
but controversial proposal for insuring health
care coverage for nearly all state residents
while containing costs. The future of the plan
is in limbo since the state did not receive the
waiver from the U.S. Department of Health and


Human Services (HHS) which was needed to
implement the plan, but it is expected that the
state will resubmit the proposal once the HHS
objections can be addressed. One aspect of the
plan which is drawing particular attention in­
volves guaranteeing health care for all state
residents below the poverty level through a
system of public rationing of medical care. The
potential use of rationing to control expenses
has been evolving in the state since 1989. Un­
der previous law, Oregon has set out plans to
insure coverage for all citizens through either
private insurance or Medicaid. To make this
affordable, the state proposed limiting the range
of services provided. The idea was to establish
a minimum health plan of specific covered
services which would be extended through
Medicaid to all citizens below the federal pov­
erty level. (Previously, Medicaid in Oregon
only covered citizens with incomes lower than
50 percent of the poverty level.) As a second
step, by 1995, most employers will have to
provide a health benefits package which pro­
vides the same level of coverage as the state’s
Medicaid package provides.
The Oregon proposal attempts to contain
costs by limiting the types of treatments avail­
able to patients. Oregon has circumscribed the
allowable range of health care services by cre­
ating a ranking of possible medical treatments.
The Oregon Health Services Commission com­
pressed 10,000 diagnoses and treatments into
709 “condition-treatment” pairs. For example,
one such pair would be: condition—appendici­
tis; treatment—appendectomy. Next, the com­
mission developed 17 social importance catego­
ries for the 709 condition-treatment pairs.
These categories ranged from those deemed
“essential” to those deemed “valuable to certain
individuals.” An example of an essential treat­
ment would be a procedure which prevents
death and leads to full recovery of the patient.
An example of a treatment which would be
valuable to certain individuals would be a treat­
ment which provides minimal improvement in
the patient’s quality of life or is geared to help­
ing recovery from a self-limiting condition, for
example, plastic surgery for minor scarring.
This procedure created a rank ordering for the
condition-treatment pairs based on the impact
of the treatment on the patient’s quality of life
and the clinical effectiveness of the treatment.
Once the list was established it was submitted
to the legislature and it was the legislature’s


task to determine how much of the list would
receive state funding. The legislature decided to
appropriate enough funds to cover services num­
bered 1 through 587 on the list. Services below
587 would not be covered in the Medicaid pack­
age. Health care providers would be legally
entitled to deny treatment to patients seeking
treatments not covered by the plan. The tradeoff
in the Oregon system is clear. Health care is
extended to a broader audience but at the cost of
limiting the available services.
Critics of the rationing system have ques­
tioned the ethics of such an approach. Is it ethi­
cally correct to ever withhold treatment even if
the benefits of the treatment are marginal? One
of the objections raised by the U.S Department
of Health and Human Services in rejecting the
state’s request for a waiver to implement the
program was the possibility that treatment for
people with disabilities could be denied under
the Oregon ranking system. Since the treatment
might be necessary but may have little effect in
improving the condition of a disabled individual,
it was likely that the Oregon plan would not
cover these medical procedures. This was seen
as potentially discriminatory. By ranking poten­
tial treatments on both a societal and individual
benefit basis, government is in fact determining
the value of a treatment for an individual. Pro­
ponents of the plan counter that rationing of
medical care already exists when individuals are
denied health care coverage. This plan insures
that everyone has at least a minimal level of
medical care guaranteed. More broadly, the
nature of public budgeting is always to ration
dollars between competing goals. Health care
competes with national defense for funding.
Given this, why should rationing amongst health
care benefits be any different?
A less philosophical argument questions
whether the cost containment goals of this ap­
proach will, in fact, work. Given that a diagno­
sis is more of an art than a science, some ques­
tion whether health care providers can circum­
vent limits on coverage and tailor some diag­
noses to fall under covered treatment classes.
Also because the effect of treatment and the
severity of the same illness varies from patient to
patient, ranking systems may be inherently arbi­
trary and doomed to fail. A treatment which
might provide a complete return to health for one
patient may be ineffective for another. As such
the rankings may be arbitrary and ineffective.2



Vermont has recently passed a law requir­
ing that universal health care coverage be made
available to all residents by October 1994. This
legislation has prompted the state to consider
two primary options for providing universal
care. The first would be to adopt a Canadian
style, single-payer health insurance program in
which the state would set medical prices and be
responsible for paying all medical bills. Alter­
natively, the state is also considering a multi­
ple-payer plan in which the state would set all
health care prices, but private insurance compa­
nies would still be used to provide health insur­
ance coverage.
As envisioned, the Canadian style program
would provide comprehensive care to all resi­
dents without any deductibles or copayments by
patients. The system would be financed from
revenues from the state’s income, sales, and
payroll taxes. Proponents of the system believe
that costs will be contained not only through
price setting, but also through a reduction in
administrative cost by eliminating multiple
payers. As of yet it is not certain whether the
single-payer will be a public or private agency.
The multiple-payer approach is based
loosely on the experience of West Germany’s
health plan which was considered successful at
containing health care costs in the 1980s. (A
multiple-payer approach has since been adopted
throughout unified Germany.) The system
consists of a number of nonprofit organizations
(called sickness funds in Germany) which col­
lect insurance premiums from employers and
employees. The premiums are then paid to
regional doctor’s associations which uses the
money to provide health care. The state sets
prices and provides insurance to the unem­
ployed. It is assumed that in Vermont, existing
insurance companies would serve the role of
collecting premiums.
While it is unclear which approach will be
adopted, Vermont has already passed several
health care reforms. The state has established a
three member Health Care Authority with over­
sight authority including the power to negotiate
insurance rates for state residents. Other re­
forms include: establishing a standard set of
benefits; insuring that benefits are portable
from one job to another; subsidizing primary
care coverage for all pregnant women and
teenagers with family incomes up to 225 per­
cent of the federal poverty level; and central­



ized budgeting with nonbinding expenditure
targets for hospitals, clinics, and physicians.2

A more limited approach aimed just at cost
containment is being tried by Virginia. The
state has chosen to revive its certificate of need
program. This program, which was started in
1974 by the federal government, was designed
to create state oversight for hospital expansions.
The belief was that through aggressive growth
plans, hospitals were creating an oversupply of
hospital beds and services and in doing so were
driving costs up through duplication of services.
Unrestricted capital construction and high tech­
nology acquisition encouraged utilization of
expensive hospital facilities for procedures
which could be performed in doctors’ offices.
The certificate of need program required hospi­
tals to prove to a state board that they had a
need for expansion or new equipment.
The program was largely discontinued by
states when federal funding was eliminated in
1986. However, Virginia has returned to the
certification of need concept in an effort to
control costs. Part of the state’s interest in
returning to this form of regulation stems from
the fact that from 1989 to 1991 hospitals spent
$130 million on expansion and new technology
in Virginia. The state estimates that 50 percent
of this would have been denied if the certifica­
tion of need program had been in effect. Fur­
thermore, the certification program uncovers
one of the problems with incentives in the
health care market. Unregulated competition in
health care encourages hospitals to provide the
best technology for patients rather than the best
price per unit of care.
Critics of the certification program point
out that its success has been mixed. One study
by the University of Alabama at Birmingham
found that the program may have encouraged
hospital expansion as hospitals accelerated
growth plans in anticipation of tougher regula­
tion which would reduce future expansion.
Others have criticized the program for reducing
access to health care by limiting the facilities
of hospitals.2

Another more limited reform approach
aimed at controlling Medicaid costs is being
tried in a number of states, but most notably in
Kentucky. These programs are generally de­
signed to introduce managed care reforms into


the Medicaid system. In Kentucky’s case, the
thrust of the reform is to use family physicians
as the primary care giver for Medicaid cases.
In the past, Medicaid patients often had no
family doctor. When treatment was needed the
recipient would go to the hospital or seek out a
specialist for their specific ailment. This meth­
od of treatment was expensive and resulted in
uncoordinated care for patients. By using a
family doctor, additional treatment can be coor­
dinated and costs reduced by using specialists
and hospitals only when they are necessary.
Kentucky estimates that this program will save
the state $120 million this year while covering
288,500 participants.
Critics of the program believe that the set
reimbursement schedules for family physicians
treating Medicaid patients will encourage doc­
tors to reduce the quality of their services.
Also, critics point out that such an approach
may not work well in states with little experi­
ence with managed care programs, particularly
those with few HMOs. For example, Wiscon­
sin estimates that it saves $10 million per year
with its managed care system for 122,000
AFDC recipients; however, it has been unable
to expand the program because of the lack of
HMOs to function as service providers.2

C onclusion

Due to the lack of national consensus on a
health care policy, states are again proving to be
the laboratories of public policy. Through the
state’s varied approaches, a better understanding
of the most promising programs for expanding
coverage and moderating costs may emerge.
Clearly, any solutions will have to address those
aspects of the health care market that have gone
awry. The practice of providing the best possi­
ble care without regard to cost will likely give
way to a more cost effective approach to health
care provision. It is also likely that individual
consumers of health care services will have to
bear a greater share of the cost of consuming
health care. Greater incentives for maintaining
personal health are likely to be adopted, as well
as the use of disincentives for destructive and
self-inflicted health problems related to behav­
iors such as smoking and drinking, which cur­
rently bear no insurance penalty.
Most of all, solutions will have to address
both the cost and access to health care. Based
on the diversity of the proposals under consider­
ation, the state experiments in health care are on
the right track and may well point to an effective
health care policy for the nation as a whole.

'See Health Insurance Association of America, (1990), p. 55.


2See Koretz (1992), p. 22.

l6Little (1992), pp. 43-66.

3See Gardner (1992), p. 28.

l7Rowley (1992), p. 15.


Th e W a ll S tr e e t J o u r n a l,

May 11, 1992, p. 1.

p. 31.

l8Graham (1992), p. 23.

5HIAA, p. 61.

1 Ibid, pp. 22-26.

6See Harden (1992), p. 61.

20Ibid, pp. 22-26.

7Ibid, p. 48.

2lIbid, pp. 14-16, 18, 22.

T or more on supply and demand factors affecting medical
costs, see Fuchs (1972).

22Lays (1992), p. 62.
n N e w Y ork T im es C o m p a n y

(1992), p. A-14.

24Biemesderfer (1992), p. 58.
l0American Hospital Association (1989).
:5Worthington (1992b), pp. 1-2.
"HIAA, p. 64.
26Wiener (1992), pp. 26-30.
l2State Policy Research, Inc. (1990), Table 1-24.
27Worthington (1992a), pp. 15-16.
l3Ibid, Table 1-19.
28Wagar (1992), pp. 20-22.
l4Ibid, Table 1-15.


29Knapp (1982), pp. 16-19.



American Hospital Association, Hospital Sta­

tistics, Chicago, IL, 1989.
Biemesderfer, Susan C., “Running for cover­

age,” State Legislatures, National Conference of
State Legislatures, July 1992, p. 54-59.
Committee for Economic Development, Re­
forming health care: a market prescription, New
York and Washington, D.C., 1987.
Fuchs, Victor, “The basic forces influencing

costs of medical care,” in Essays in the Econom­
ics of Health and Medical Care, Victor Fuchs,
ed., National Bureau of Economic Research,
Columbia University Press, New York, 1972,
pp. 39-50.
___________ , “Though much is taken: reflec­
tions on aging, health, and medical care,” Milbank Memorial Fund Quarterly/Health and
Society Vol. 62, No. 2, 1984, pp. 143-166.

aging population,” Business Week, June 8,
1992, p. 22.
Lays, Julie, “The Canadian system,” State
Legislatures, National Conference of State
Legislatures, July 1992, p. 62.
Little, Jane Sneddon, “Lessons from varia­

tions in state Medicaid expenditures,” New
England Economic Review, Federal Reserve
Bank of Boston, January/February 1992,
pp. 43-66.
New York Times Company, The, “Canada’s

no medical model,” New York Times, May 26,
1992, p. A-14.
Rowley, Storer H., “Prescriptions from
Canada—would universal health care work in
this country?,” Chicago Tribune Magazine,
May 31, 1992.
State Policy Research, Inc., States in Profde,

Gardner, Booth, “Health care reform: the state

Alexandria, Virginia, 1990.

perspective,” Intergovernmental Perspective,
Advisory Commission on Intergovernmental
Relations, Vol. 18, No.2, Spring 1992, pp. 28-30.

Wagar, Linda, “Rebirth of a good idea,” State

Government News, Vol. 35, No. 6, The Council
of State Governments, June 1992, pp. 20-22.

Graham, D. William, “Long-term health care

reform: three approaches,” Intergovernmental
Perspective, Advisory Commission on Intergov­
ernmental Relations, Vol. 18, No.2, Spring 1992,
pp. 22-24.
Harden, Shelda, “Feds take a stab at health care

reform,” State Legislatures, National Confer­
ence of State Legislatures, July 1992, pp. 61-63.

Wessel, David, “The coming federal battle:

health costs vs. poor,” The Wall Street Journal,
May 11, 1992, Sec. l,p . 1.
Wiener, Joshua M., “Oregon’s plan for health
care rationing,” The Bookings Review, Winter
1992, pp. 26-31.
Worthington, Rogers, “Vermont may take

Health Insurance Association of America,

Source Book of Health insurance Data, Wash­
ington, D.C., 1990.
Knapp, Elaine, “Medicaid diet plans,” State

Government News, The Council of State Gov­
ernments, Vol. 35, No. 6, June 1982, pp. 16-19.

health care lead,” Chicago Tribune, July 12,
1992a, Section 1, pp. 15-16.
___________ , “Iowa warms to ‘play-or-pay’
health plan,” Chicago Tribune, May 26, 1992b,
Section 1, pp. 1-2.

Koretz, Gene, “America’s medical tab is grow­

ing faster than ever ... and that’s bad news for an




C a ll fo r Papers

A n Appraisal
The 2 9 th A n n u a l Conference on B a n k Structure a n d Competition, M a y 5 - 7 , 1 9 9 3
The Federal Reserve Bank of Chicago announces its
29th annual Conference on Bank Structure and Com­
petition, which will be held at the Westin Hotel in
Chicago, Illinois, May 5-7, 1993, and invites the sub­
mission of papers for the program.

I the impact of capital requirements on bank
I recent developments under the Community
Reinvestment Act and the Home Mortgage
Disclosure Act

The conference, which is attended each year by over
400 academics, regulators, and financial institution ex­
ecutives, has served as a major forum for the exchange
of ideas on public policy toward the financial services
industry since its inception in 1963. In recent years,
the program has featured some of the most prominent
members of the financial and research communities.

We also encourage the submission of papers on other
topics related to the structure and regulation of the
financial services industry.

The central theme of the 1993 conference will be an
appraisal of the Federal Deposit Insurance Corporation
Improvement Act of 1991, including the following
specific provisions:
capital adequacy
prompt corrective action
interbank exposure
I consumer protection
Other topics on which papers are specifically invited:
the consolidation movement in banking,
with emphasis on recent “megamergers”
the management of risks associated with derivative
financial instruments
the RTC’s asset disposal activities and the current
status of the savings and loan cleanup

The first day of the conference will feature technical
papers of primary interest to an academic audience,
while the final two days are designed to appeal to a
broader audience. Invitations to the conference will
be mailed in mid-March.
If you would like to present a paper at the conference,
please submit two copies of the completed paper or an
abstract with your name, address, and phone number,
and those of any coauthors, by December 31, 1992.
If you are not currently on our mailing list or have
changed your address and would like to receive an
invitation to the conference, please contact us.
Address all correspondence to:
Conference on B ank Structure a n d Competition,
Research Department,
Federal Reserve B ank o f Chicago,
2 3 0 South LaSalle Street,
Chicago, Illinois 6 0 6 0 4 -1 4 1 3

For additional information, call Herbert Baer 3 1 2 -3 2 2 -6 1 9 9 or Douglas E vanoff3 1 2 -3 2 2 -3 8 1 4 .

P u b l i c I n f o r m a tio n C e n t e r

Federal Reserve Bank of Chicago
P.O. Box 834
Chicago, Illinois 60690-0834


D o N o t F orw ard
Address C o rrec tio n Requested
R eturn P o stag e G uaranteed
M a ilin g lab el c o rre c tio n s o r d eletion s:
Make changes or mark Delete From M ailing List

on the label and fax the label to: (312) 322-2341,
or mail to: ATTN: Mail Services,
Federal Reserve Bank of Chicago,
P.O. Box 834, Chicago, Illinois 60690-0834