CHAPTER

16
The IS-LM Model
W e have now explored three of the major issues with which macro-
economics is concerned: gross national product (GNP), money and
distribution. We questioned the appropriateness of GNP as the desirable
end for economic policy, and emphasized the importance of a just distri-
bution as a desirable end, but said little about policies for attaining these
ends. In this chapter, we examine the policy tools at the macroeconomist’s
disposable that can help us attain an economy with sustainable scale, just
distribution, and efficient allocation.
Of course, to know how policies work, we have to know how the
macro-economy works. One way of doing this might be to build on mi-
croeconomic principles to construct a model in which supply and demand
of all goods and services balances simultaneously. This approach would
extend the basic market equation presented in Chapter 8—MUxn*MPPax
= MUyn*MPPay—into a general equilibrium model encompassing all
goods (x, y, z . . .), all commodities (a, b, c . . .), and all consumers (n, m,
o . . .). Such a model can easily become overwhelming. A thousand si-
multaneous equations with a thousand unknowns is hard to come into
mental contact with. It does show that everything depends on everything
else, which is interesting and usefully humbling, but it is also crippling
from a policy perspective to have to face the implication that in order to
predict anything, you have first to know everything. But a smaller system
of two or three or five especially important aggregate sectors interacting
through two or three or five simultaneous equations that reflect key be-
havior can aid the understanding and give basic policy insights. This is the
kind of model that most macroeconomists have sought. They still look at
the whole economy, but they divide it into fewer but larger aggregate
278 • Macroeconomics

sectors than does the general equilibrium model of microeconomics.1 A

model of this type, first offered in 1937 by Sir John Hicks2 and now called
the IS-LM model, has proven to be a good “two-digit” compromise be-
tween completeness and simplicity. It has become the “workhorse” model
in macroeconomics. Below we will explain this model, and then discuss
its applications to ecological economics.
The model divides the economy into two sectors: the real sector (deal-
ing with national income, savings, investment, rates of productivity of
capital, government spending, taxation, etc.) and the monetary sector
(money supply, interest rates, demand for liquid cash balances). The real
sector reflects the theories and insights of classical economics, and the
monetary sector reflects the insights of John Maynard Keynes, which in
1937 were still quite new. The model seeks to explain how the interde-
pendent behavior of consumers and savers, lenders and borrowers, and
monetary authorities interact to determine the level of national income
and the rate of interest.

Box 16-1 The Quantity of Money Theory of Income

Another way of relating the real and monetary sectors in an aggregate
way is through the “identity of exchange,” MV = PQ, where Q is quantity
of final commodities sold to households, P is average price of exchange,
M is stock of money, and V is velocity of circulation of money (number of
times an average dollar is spent per year on final goods and services).
Since by definition V = PQ/M, the equation of exchange is an identity or
truism. To the extent that V is a constant or slow to change, reflecting
stable payment habits and settlement periods, the identity becomes the
“quantity of money theory of income,” stating that changes in PQ are
proportional to changes in M. If the economy is at full employment, it will
be very hard to increase Q in the short run, and the change in PQ will be
mainly a change in P—i.e., inflation. Historically M and P have often
moved in direct proportion, yielding a quantity of money theory of the
price level.

1Our measures of the two most basic magnitudes of macroeconomics, GNP and money, are

too dialectical and uncertain to be able to support exact calculations implicit in complicated mod-
els. As Oskar Morganstern remarked in his classic On the Accuracy of Economic Observations, “eco-
nomics is a two-digit science.”
2J. Hicks, Mr. Keynes and the “Classics,” Econometrica 5(2) (April 1937).
 Chapter 16 The IS-LM Model • 279

■ IS: The Real Sector

Let’s begin with the real or classical sector. The real sector is in equilib-
rium when the supply of goods by firms is just equal to the demand for
goods by households (the lower half of the circular economy in Figure
2.4). Of course, the demand for goods by households is determined by
their income—the money firms pay households for their factors of pro-
duction (e.g., labor), and the supply of goods is determined by the firms’
employment of those factors of production (the upper half of the circular
economy in Figure 2.4). In equilibrium, income (Y) equals output (GNP).
Remember from the circular flow diagram in Figure 2.5 that the equilib-
rium condition for the continued flow of national income at a given level
is that leakages equal injections. In the simplest case, the leakage is sav-
ings (S) by households, the new injection is investment (I) by firms.
Therefore, the equilibrium condition for the real sector is S = I.
But how do S and I get determined? Let r be the interest rate and Y be
national income (GNP). In equilibrium, income paid to the factors of pro-
duction will just equal the output of goods provided by those factors of
production, and the income will be used to purchase the output. Savers
(i.e., households) will save more if their income Y is higher than if it is
lower. Also savers will save more with a higher interest rate r than with a
lower one. Investors (i.e., firms) will borrow and invest more if the inter-
est rate is lower, and if income is higher. In other words, savings is some
function of the interest rate and national income, says S = S(r, Y). Like-
wise, investment is some different function (representing the behavior of
firms instead of households) of the same two variables, say I = I(r, Y).
In equilibrium,
S=I
or
S(r, Y) = I(r, Y)
The above equation is satisfied for all combinations of r and Y such that S
= I, that is, such that savers and investors are both satisfied.
There are many such combinations of r and Y—we have only one
equation with two unknowns. Plotting all the combinations of r and Y that
result in S = I gives us Hicks’ so-called IS curve, short for I = S (Figure
16.1). To reiterate, this is the combination of r and Y that leads to equi-
librium in the real sector: leakages (savings) equal injections (investment),
and the demand for goods is just equal to the supply.
Why is the IS curve drawn with a negative slope? Businesses will only
borrow money to invest if they can make sufficient returns from the in-
vestment to pay off the loan plus interest and still have money left over for
profit. A businessperson would not borrow money at 6% interest to invest
280 • Macroeconomics

Figure 16.1 • The IS curve: At low (high) levels of income Y, there is a corre-
spondingly low level of savings. At high (low) rates of interest r, there is a low
(high) demand for investment. Therefore, at low (high) levels of Y, savings and
investment will only be in equilibrium when r is high (low). If interest rates are
too high for a given level of income, savings (leakages) will be greater than in-
vestment (injections). Firms producing more goods than people consume re-
duce production and the economy shrinks. Firms with excess capacity borrow
less, so the price of borrowing (the interest rate) falls to clear the market. The
converse is true when investment is greater than savings.

in a project expected to return 5% annually on the investment, but would

borrow at a rate of 4%. As interest rates go down further, more and more
investments become profitable, and therefore more investments are made.
More investment leads to higher Y. Therefore, high interest rates lead to
low rates of investment and low income, while low interest rates lead to
high rates of investment and high income. Savings, in contrast, are prob-
ably determined more by income than by interest rates.3 When income is
low, all money has to be spent simply to meet basic consumption needs,
and none is available to save. As income increases, basic consumption
needs require a smaller percentage of income, and more is left over to
save, so in general higher incomes lead to greater savings.
Combining these two tendencies, we would expect that at high levels of
income when lots of money is being saved, investors will only borrow all
that money to invest if interest rates are low. At low levels of income, sav-
ings are low, and unless interest rates are high, businesses will demand more
money than is being saved. For some readers, a diagrammatic explanation
for the negative slope of the IS curve will be easier to follow (Figure 16.2).

3Savings rates should also increase as interest rates increase, as under these circumstances sav-

ings yield higher returns, and consumption has a higher opportunity cost. But empirical evidence
does not support this. One reason may be that if savers are motivated by attainment of a target fu-
ture amount, a higher interest rate would mean less saving is needed to reach the future target.
 Chapter 16 The IS-LM Model • 281

Box 16-2 A Graphic Derivation of the IS curve

Figure 16.2 • A graphic depiction of the derivation of the IS curve.

Figure 16.2 illustrates one way to derive the IS curve. Quadrant I shows
the basic relationship between interest rates and investment—high in-
terest rates lead to low levels of investment, and low interest rates lead
to high levels of investment. There is a negative correlation between in-
terest rates and investment, as depicted on the graph. Quadrant III
shows the relationship between savings and income. Poor people must
spend all their income to meet their basic needs and cannot afford to
save anything. As income increases, people begin to save, so there is a
positive correlation between income and savings, as depicted on the
graph. We know that in equilibrium (which is what the IS curve depicts),
investment equals savings. Quadrant II contains a 45-degree line that al-
lows us to translate a given rate of investment from quadrant I to an
identical rate of savings in quadrant III. Quadrant IV shows the relation-
ship between income and interest in a real sector equilibrium.
If we start with interest rate r, we can see from quadrant I that this will
282 • Macroeconomics

correspond to level of investment I. Dropping a line down from I in quad-

rant I to the 45-degree line then across to quadrant III lets us determine
the equilibrium level of savings S in quadrant III. We can see from quad-
rant III that this level of savings corresponds to income Y. The point in
quadrant IV where income Y meets interest rate r gives us one point on
the IS curve. If we do the same for interest rate r′, we have two points on
the IS curve. We see that a low level of interest leads to equilibrium only
when income is high, and a high level of interest leads to equilibrium
when income is low. Perhaps the simplest way to remember this rela-
tionship is that at low interest rates, investment will be high, and high
investment leads to high income.

Macroeconomics does not assume that the economy is always in equi-

librium, but it does assume that it is at least moving in that direction. For
example, we know that if r rises, then savers will try to save more, and in-
vestors will be less willing to borrow and invest, leading to a condition in
which planned S > I. In other words, savers want to save more than in-
vestors want to invest at the new higher r. This will have two impacts.
First, leakages will be greater than new injections, causing income to fall.
Second, savers earn interest on their savings because investors are willing
to pay that interest to borrow the money. Interest is the price of money.
When the supply of savings is greater than the demand for savings, the in-
terest rate must fall. The mechanism is the same as for any other good, as
explained in Figure 9.2. At a lower Y savers save less, and at a lower r in-
vestors borrow more, and both r and Y continue to fall until I again equals
S at a lower income (Y) and higher interest rate (r) than before. If the in-
terest rate falls, then investment will become greater than savings, and ad-
justment will occur in the opposite manner. These dynamics are indicated
by the arrows in Figure 16.1.

■ LM: The Monetary Sector

We turn now to the monetary sector and the LM curve, which shows the
levels of income (Y) and interest (r) at which the demand for money bal-
ances (money held by people) equals the supply of money. We must first
ask why individuals want to hold money balances when they could easily
exchange them for real assets. From our earlier discussion the answer is
clear—people hold cash balances to avoid the inconvenience of barter.
Keynes referred to this as the transactions demand for money. He also
spoke of a related liquidity preference, meaning that, other things being
equal, people prefer liquid assets to “frozen” assets because they are so eas-
ily convertible into anything else, therefore fungible. Money is the most liq-
uid of all assets. But of course other things are seldom equal, and the cost
 Chapter 16 The IS-LM Model • 283

of holding wealth in the form of fungible money is to forego the interest

that could be had by lending the money, or the utility from spending it on
a real asset or commodity. Yet if too much of your wealth is tied up in
nonliquid forms, you will have difficulty making necessary transactions in
a timely manner and meeting unexpected contingencies. The higher the
national income, the more need for transactions and consequently the
more money everyone will need (a higher transactions demand for
money), and the higher the interest rate will have to be to induce owners
of those transactions balances to sacrifice liquidity by lending them.
The demand for money balances (DM) thus depends on r and Y, by
means of a relation of liquidity preference (L). Thus:
DM = L(r, Y)
The equilibrium condition is that the demand for money equals the sup-
ply of money (SM):
DM = SM
What determines the supply of money? In earlier times it was the ge-
ology and technology of gold or silver mining (a part of the real sector!),
but today we have not real commodity money, but fiat or token money,
controlled by the government through the private banking sector, as dis-
cussed in Chapter 14. For simplicity, the model usually takes SM as given
by the government, equal to M. Thus:
L(r, Y) = M
is the equilibrium condition for the monetary sector, and the LM curve
consists of all those combinations of r and Y such that the aggregate de-
mand for cash balances is equal to the given money supply (Figure 16.3).

Figure 16.3 • The LM curve.

284 • Macroeconomics

Since we have one equation with two unknowns, we cannot get unique
values of the unknowns, but we can determine all those combinations of
r and Y that satisfy our one equation.

Box 16-3 The Federal Reserve Bank

In the U.S., money is not controlled by the democratically elected gov-
ernment but rather by the Federal Reserve Bank (the Fed), a nonelected
“branch” of government. Decisions concerning monetary policy are de-
cided upon by a seven-member board of governors with lesser influence
by the directors of the 12 regional Federal Reserve Banks. Members are
appointed by the President (with Senate approval) for 14-year staggered
terms, and the chair and vice chair are appointed for 4-year terms. De-
spite the importance of monetary policy in the functioning of our econ-
omy, the system is specifically designed to insulate the Fed from
pressure by democratically elected politicians! The Fed is not expected to
respond to voters. This does not mean that the Fed does not have a con-
stituency to which it feels responsible, as we shall discuss later.

Why is the LM (short for L = M) curve drawn with a positive slope?

Let’s ask ourselves what are the consequences on the interest rate (r) of an
increase in income (Y). A larger Y means a larger volume of transactions
and will cause a greater demand for transactions balances. This will lead
to a higher r to compensate for the loss in liquidity from lending those
balances. Thus, a higher Y will require a higher r for money holders to
again be satisfied (for L to equal M). Hence the positive slope of the LM
curve. This relationship seems to be sufficiently clear that a more detailed
graphic explanation is unnecessary.
When the monetary sector is out of equilibrium, what specific mech-
anisms drive it toward equilibrium? Say the monetary authority increases
the money supply, so there is more money available than people actually
desire to hold at the existing interest rate—that is, M > L. Excess money
is used to buy bonds and other nonliquid interest-bearing assets (which
we will refer to jointly as “bonds” for convenience). More money chasing
the same number of bonds will drive up their price.
There are many types of bonds, but in the simplest case, when someone
buys a bond, they are paying something now to receive a fixed amount
when the bond matures. For example, if I pay $50 today for a $100 bond
that matures in 10 years, my rate of return is about 7.2%. An increase in
the money supply might drive the price of the bond up to $60, which pro-
vides a rate of return of only 5.24%. The higher the price for a bond, the
 Chapter 16 The IS-LM Model • 285

lower the interest rate on that bond. Hence, an increase in the supply of
money increases the demand for bonds and drives down the interest rate.
At lower interest rates, there is less opportunity cost to holding
money, and hence a higher demand for money. Lower interest rates also
stimulate investment, leading to economic growth, which further stim-
ulates the demand for money. The result is a new equilibrium at lower
interest rates and higher income. A decrease in the money supply of
course leads to the opposite result. These forces are illustrated by the ar-
rows in Figure 16.3.

■ Combining IS and LM
Putting the IS and LM curves together lets us determine a unique combi-
nation of r and Y (namely r*, Y*) that satisfies both the S = I condition of
the real sector and the L = M condition of the monetary sector (Figure
16.4). The point of intersection is the only point common to both curves,
the only point that gives equilibrium in both real and monetary sectors.
Basically we now have two simultaneous equations determining two un-
knowns, r and Y.
The IS-LM model is used in a comparative statics4 way to analyze the
effect on r and Y of changes in the underlying determinants—namely,

Figure 16.4 • The IS-LM model.

4Comparative statics is the analysis of what happens to endogenous variables in a model (in

this case, r and Y) as a result of change in exogenous parameters (in this case, propensity to save,
efficiency of capital investment, and liquidity preference). It compares the new equilibrium vari-
ables with the old ones, without explaining the precise dynamic path leading from the old to the
new equilibrium.
286 • Macroeconomics

propensity to save, the efficiency (productivity) of capital investment, and

liquidity preference. Of particular interest to policy makers is the impact
of policy variables on r and Y—namely, government expenditure, taxa-
tion, and the money supply. Each of these changes results in a shift in one
of the curves, and consequently in a move along the other curve to a new
intersection point. What we are really interested in, then, is how the econ-
omy moves toward equilibrium after policies or outside (exogenous)
changes push it away.

■ Exogenous Changes in IS and LM

First let’s look at some exogenous changes, those that are basically inde-
pendent of fiscal and monetary policy, and therefore outside the IS-LM
model. Consider an increase in the marginal propensity to save. Such a
change in savings rate might result from fears of an economic downturn
that would lead to lower wages and greater unemployment. (We might
hope that people might one day simply decide to consume less in order
to protect the environment!) In either case, people decide to save more
and spend less of their extra income. This means that now S > I for all the
combinations of r and Y on the IS curve. We need a new IS curve for
which S = I again. If people save more at every r, this means S > I, or leak-
age greater than injection, so the flow of income will fall to the level at
which S = I again.
Even though people are saving a larger fraction of their income, they
will end up having a smaller income out of which to save, with the result
that S will be the same, only now the product of a higher rate of savings
per dollar times a lower number of dollars in income. The very act of sav-
ing more and spending less will have caused a fall in aggregate income to
the extent that the lower income times the higher fraction saved gives the
same total savings, equal to the unchanged level of investment. So every r
will be paired with a smaller Y in the new IS than with the old one. The
IS curve will have shifted to the left. The new intersection with LM will
occur at a lower r* and lower Y* than before. An increase in the marginal
propensity to save will therefore result in a fall in national income and a
fall in the interest rate.
Of course if the marginal propensity to save increases, then the mar-
ginal propensity to consume must decrease. As people consume less,
businesses will be unable to sell their goods, leading to unplanned ac-
cumulations of inventory. This in turn will lead businesses to reduce
planned investment and production—perhaps laying people off. Unem-
ployment resulting from layoffs further decreases consumption, requir-
ing another round of adjustment, lowering Y still more. The final result,
when S = I again, may well be that S will be lower than the level at which
 Chapter 16 The IS-LM Model • 287

we started. Thus, the effort of everyone to save more in the aggregate

could result in everyone actually saving less—the so-called paradox of
thrift. In such a case, a higher savings rate induced by fear of recession
could itself cause a recession—a self-fulfilling prophecy.
Now suppose there’s an increase in the efficiency of investment (an
increase in the marginal productivity of capital), thanks to a new inven-
tion. For example, many people claim that this is exactly what has hap-
pened in today’s “new economy,” in which information technology is
said to have increased productivity. This would increase I, so that I > S
now along the old IS curve. With I > S, injections are greater than leak-
ages out of the circular flow, so the flow of income will grow until S = I
again. The new IS curve will have a higher Y for each r. The curve shifts
rightward. The new equilibrium occurs with a higher Y* and a higher
r*. An improvement in the marginal efficiency of capital raises both in-
come and the interest rate.
Finally, turning to the LM curve, suppose there was an increase in liq-
uidity preference, so that L > M. Such a change could result from increas-
ing uncertainty over future economic conditions, and a desire by people
to be prepared for the unforeseen with cash on hand. Alternatively, the
deregulation of banking in the United States during the mid-1970s al-
lowed certain checking accounts to pay interest. This reduced the oppor-
tunity cost of holding money, and therefore probably increased the
liquidity preference as well. In either case, for any income and associated
level of needed transactions balances, there is a greater willingness to hold
those balances, to hold more than strictly needed. It takes a higher r to in-
duce holders of money to lend. Consequently each level of Y will be as-
sociated with a higher r on the new LM than on the old one. The new LM
will shift upward. The new equilibrium will occur at a higher r* and lower

Box 16-4 Junk Bonds and Timber Companies

Seemingly abstract things like interest rates on bonds and Wall Street
transactions can affect real economic production and the provision our
environmental services. For example, during the 1980s, hostile
takeovers and the introduction of junk bonds on Wall Street led to defor-
estation on the West Coast. How did this happen? Mergers, when two
companies join together, and acquisitions, when one company pur-
chases another, are a normal part of corporate activity in the U.S. Some-
times, however, one company does not wish to be taken over by another.
For example, mergers and acquisitions (M&A) focused primarily on short-
288 • Macroeconomics

term profits can weaken or destroy the company being acquired, leading
to massive layoffs.
Reasonably enough, managers not eager to be laid off will be opposed
to a merger, and under such circumstances, takeover attempts are “hos-
tile.” One company acquires another through the purchase of a control-
ling share of stocks. As soon as someone starts buying enough stock to
control a company, the stock price rises. A company threatened by a
hostile takeover can attempt to defend itself by repurchasing its own
stock, driving up the price of stock even further. To get enough money
for a hostile takeover, the company attempting the takeover can offer
high-yield, high-risk bonds known as “junk bonds” in Wall Street
jargon.a
The best target for takeover is a company that has lots of assets in a
nonliquid form that can be liquidated after takeover to pay off the junk
bonds, but that cannot be sold quickly to defend against the takeover.
Timber companies have valuable assets in the form of forests that can
be liquidated after takeover, but cannot be sold quickly to buy back
stock and prevent a hostile takeover. This made them popular takeover
targets during the 1980s.
A classic example is Charles Hurwitz’s acquisition of Pacific Lumber in
the mid-1980s. Pacific Lumber fought the takeover, but using a combi-
nation of junk bonds and short-term loans, Hurwitz won out, acquiring
with the company 196,000 acres of forest, including the largest unpro-
tected stands of virgin Redwood in the world. Hurwitz was saddled
with an enormous debt and crushing interest payments. To repay the
debt, Hurwitz liquidated much of the forest stock, including many old-
growth redwood groves. Some illegal cuts were conducted on week-
ends and holidays to avoid state regulators. Wall Street innovations
during the 1980s accelerated the decimation of the nation’s last re-
maining virgin forests, and of the environmental services those forests
once provided.b
aDifferentcompanies (and cities and countries) have different credit ratings based on
their financial soundness. Bonds from financially sound companies are themselves
very sound, bonds from less sound companies are not, and the risk of default is
higher.
bN.Daly, “Ravaging the Redwood: Charles Hurwitz, Michael Milken and the Costs of
Greed,” Multinational Monitor, 16 (9) (1994). On-line: http://multinationalmonitor.org
/hyper/issues/1994/09/mm0994_07.html.

Y*. An increase in liquidity preference raises the interest rate and lowers
national income.
The above analysis of changes in propensity to save, efficiency of in-
vestment, liquidity preference, and money supply is summarized in Fig-
ure 16.5.
 Chapter 16 The IS-LM Model • 289

Figure 16.5 • Shifting of the IS and LM curves. The shift from IS1 to IS2 could be
the result of either a decrease in the marginal propensity to save or an increase
in the marginal efficiency of investment. The shift from LM1 to LM2 is caused by
either a reduction in liquidity preference or an increase in the money supply.
Can you work out the changes in r and Y resulting from an increase or decrease
in each of the four parameters, others remaining constant? What about changes
in two or more parameters at the same time?

■ IS-LM and Monetary and Fiscal Policy

Changes in the propensity to save, the efficiency of investment, and liq-
uidity preferences are not brought about directly by policy interventions;
they are affected by psychology and technology and as a result are diffi-
cult, if not impossible, to predict. However, policy makers do have two
sets of economic levers by which they can influence these variables: mon-
etary policy and fiscal policy.
What does the IS-LM model tell us about different monetary and fiscal
policy levers? The analysis of monetary and fiscal policy in macroeco-
nomics can be worked out by tracing the effects on IS or LM of changes
to the money supply and of government taxing and spending.
Monetary policy basically affects the money supply. When the mone-
tary authority (the Federal Reserve, in the U.S.) increases money supply,
the LM curve shifts downward and M > L, which drives interest rates
down, as explained above. Lower interest rates stimulate the economy,
and income grows. If the money supply is increased by too much, there
can be too much money chasing too few goods, and inflation threatens.
Reducing the money supply drives interest rates up, shrinks the economy,
and can help control inflation.
Fiscal policy is basically government expenditure and taxation. When
the government spends money, industry has to produce more goods and
290 • Macroeconomics

services to meet the increased demand. This drives up income and also in-
creases the demand for investments, driving up interest rates. The IS curve
shifts to the right. Decreasing government spending has the opposite effect.
There are three ways the government can finance expenditures. First,
it can impose taxes. When the government increases taxes, people have
less to spend, decreasing demand, and leading to less investment. The
economy shrinks, and interest rates go down. However, if the government
spends the entire tax increase, the stimulus of increased expenditure out-
weighs the contraction caused by taxes, since some of the tax money now
being spent would have been saved. Second, the government can use debt
financing and borrow money. The government borrows money by selling
bonds. An increase in the supply of bonds drives down the price and
drives up the interest rate. Third, the government can use its right to
seigniorage to simply print and spend money, which increases the money
supply. As we noted above, the increased money supply will further stim-
ulate the economy but will have a countervailing impact on interest rates.
Seigniorage-financed fiscal policy seems the logical choice for stimu-
lating the economy, but it carries the threat of inflation. Governments
could dramatically increase their ability to use seigniorage if they in-
creased reserve ratios to 100%, as suggested by Frederick Soddy so long
ago. The government would then be able to print and spend money when
the price index started to fall, and tax and destroy money if inflation
threatened to become a problem. The government would also be able to
target monetary policy much more effectively, using it to address issues of
scale, distribution, and allocation. X and X.
The impact of fiscal and monetary policy depends on how much ex-
cess capacity exists in the economy. Consider a bowling alley in a small
isolated town where the government is undertaking a large project to
stimulate economic growth. When unemployment is high, wages may be
fairly low, and few people have disposable income to spend on bowling.
As a result, the bowling alley is virtually empty. If the government funds
a large project in town, some people are directly employed by the project,
and they spend much of their money in town, inducing other local busi-
nesses to hire to meet the increased demand. People use their extra in-
come to go bowling, and the bowling alley’s income grows.
Now imagine that the government implements the same project in a
town with very low unemployment. Bowling is popular, and the alley is
full every night. The government needs employees for the project, but in-
creasing demand when supply is low drives up wages, the price of em-
ployees. Disposable income increases, but every new bowler at the alley
simply “crowds out” another bowler, who would have to leave the alley.
The alley might like to expand, but the government is borrowing money
to finance its project, driving up interest rates, making it too expensive for
 Chapter 16 The IS-LM Model • 291

the alley owner to expand. The alley can raise its prices with the increased
demand, but it must also pay higher prices for its labor force, and there-
fore can only break even. When an economy is at full employment, the
bowling alley owners might be much better off with an expansionary
monetary policy that lowered interest rates so they could expand. In con-
trast, if the government lowered interest rates when the alley had consid-
erable excess capacity, expansion would do the owners no good at all.
The failure of lowering interest rates to stimulate economies with low
demand is known as a liquidity trap. In general, the economy is some-
where between the extremes of depression and operation at full capacity
(i.e., most bowling alleys are full sometimes, but very few are always full).
While increased government expenditure leads to some degree of crowd-
ing out and increased interest rates, it also increases income.
Table 16.1 summarizes the impacts of fiscal and monetary policy on in-
terest and income. In each case, the impact is obviously the opposite for
the opposite policy.

Inflation and Disinflation

If we looked only at the IS-LM model, and if our goal was continued eco-
nomic growth, the superior policy option would be clear: keep increasing
the money supply to lower interest rates and stimulate investment, and
use fiscal policy when necessary to stimulate demand. However, when we
first presented the LM model, we saw that the real money supply is equal
to the nominal money supply divided by prices—that is, the real money
supply equals M/P. There is, therefore, another path toward equilibrium
between supply and demand for money in response to an increase in
money supply—price inflation. A larger nominal money supply divided
by higher prices can lead to no change in real money supply. The closer
the system is to full output, the less output is likely to increase in response
to lower interest rates, and the more likely that monetary expansion will
result in inflation. Inflation is an increasing general level of prices (not a
state of high prices).
Why are governments and monetary authorities so worried about in-
flation? Are their concerns justified? How does inflation affect the real
economy? The first point to make is that people appear not to like infla-
tion, which alone is some justification for trying to avoid it. Many econo-
mists argue that inflation is regressive, but empirical support for this
argument is difficult to find.5 Empirical evidence does show, however,
that real wages can fall substantially during prolonged episodes of high

5A. Bulir and Anne-Marie Gulde, Inflation and Income Distribution: Further Evidence on Em-

pirical Links, IMF Working Papers, no. 95/86. Washington, D.C.: International Monetary Fund,
1995.
292 • Macroeconomics

■ Table 16.1
EXPECTED IMPACTS OF BASIC MONETARY AND FISCAL POLICIES ON INTEREST RATE AND INCOME
Policy Interest Rate Income
Monetary expansion (–) (+)
can be accomplished by: When economy is weak (high When economy is weak, no impact
• Reduced reserve requirements unemployment, low investment), on income. Known as the liquidity
• Selling bonds on the open market monetary policy may have little trap.
• Lowering the discount (interest) rate to no impact on interest rates.

Tax increase (–) (–)

Taxes (especially progressive Taxes collect more money when
income taxes) help stabilize income grows and less when it
the economy. shrinks.

Increased government expenditure (+) (+)

Can be spent on market or
nonmarket goods.

• Financed by deficit spending (+) (Probably +)

Impact on interest rate may be Income will increase when economy
small when economy is weak, is weak, but may not increase when
large when economy is operating it is already operating at full
at full capacity. capacity; latter condition is known as
crowding out.

• Financed by taxes (+) (+)

Increase in interest rate is less Growth rate is less than occurs with
than occurs with deficit spending. deficit spending.

• Financed by seignorage 0 +
Likely to cause inflation under crowd-
ing out conditions, with no real
growth in income.

inflation.6 In addition, during episodes of high inflation, it is likely that

the wealthy and educated are better able to take advantage of investments
and contracts that protect their money than the poor. Thus, with contin-
uous high inflation, the poor may well lose ground to the rich.
Hyperinflation, often defined as inflation greater than 50% per
month, can also destabilize the economy. In hyperinflation, money fails
not only as a store of value, but also as a medium of exchange. Impacts of
moderate inflation depend to a large extent on whether it is expected or
unexpected.

6B. Braumann, High Inflation and Real Wages, Western Hemisphere Department Series:

Working Paper WP/01/50, May 1, 2001.

 Chapter 16 The IS-LM Model • 293

If everyone expects a certain rate of inflation, and their expectation

comes to pass, then inflation is incorporated into contracts and causes
very few problems. The only groups one would expect to lose from an ex-
pected inflation are holders of money (which pays no interest) and peo-
ple on fixed incomes. However, with expected inflation, most people will
hold less money, and incomes are likely to be inflation adjusted. Disin-
flation is a decrease in the rate of inflation. Deflation is a decline in the
overall price level. Unexpected inflation, disinflation, and deflation have
entirely different outcomes than expected inflation. The most useful way
to assess the impacts of these unexpected changes is to look at debtors
versus creditors.
Unexpected inflation has entirely different outcomes. The most useful
way to assess the impacts of unexpected inflation and disinflation is to
look at debtors versus creditors. When there is unexpected inflation, any
loans with nominal interest rates (i.e., interest rates that are not pegged to
inflation) will be worth less and less every year. Debtors benefit and cred-
itors suffer. For example, people in the 1960s got 30-year house mort-
gages at around 6%. When inflation in the 1970s climbed over 12%, some
home owners ended up paying back less than they originally borrowed.
In general, unexpected inflation systematically redistributes wealth from
creditors (generally the rich) to debtors (generally the poor). The govern-
ment is a net debtor, and therefore benefits, as do the future generations
that are expected to pay off the government’s debts. However, a country
cannot have unexpected inflation forever—eventually it becomes ex-
pected, or else becomes hyperinflation, with its accompanying problems.
What happens when the government tries to cause disinflation or de-
flation? Obviously, just as unexpected inflation benefits debtors, unex-
pected disinflation must benefit creditors. In 1980, a 30-year mortgage at
14% didn’t look so bad when inflation was 13% annually, and people ex-
pected their incomes to rise by at least that rate. By 1986, however, infla-
tion (and wage increases) had fallen to less than 2%, and creditors were
collecting a 12% annual real return on their loans. Thus, existing debtors
suffer and existing creditors benefit from disinflation.
Other impacts of disinflation depend on whether it is brought about by
fiscal or monetary policy. Theoretically inflation can be reduced by de-
creasing aggregate demand or increasing aggregate supply, but policy usu-
ally acts on demand. Fiscal policy can only decrease aggregate demand
through greater taxation or reduced expenditure, both of which should
lower the real interest rate to the benefit of new debtors. Other distribu-
tional impacts depend on the specific policy used. For example, demand
could be reduced by reducing subsidies for big business or by reducing
transfer payments to the poor.
294 • Macroeconomics

THINK ABOUT IT!

Under President Reagan there was a big emphasis on supply-side eco-
nomics, increasing income by providing incentives for production (i.e.,
supply). Policy measures for achieving this include investment subsi-
dies, reduced capital gains taxes, and reduced taxes for the rich. Can
you explain why these policies would theoretically increase supply and
reduce inflation?

The monetary authority, on the other hand, can only act to reduce de-
mand by reducing the money supply, which increases real interest rates,
to the detriment of debtors. Interest-sensitive sectors of the economy, such
as farming and construction, also lose out. If losers are forced into liqui-
dation or bankruptcy, they may be forced to sell their assets at bargain
prices, and it is the well-to-do who maintain the liquidity necessary to
purchase those assets. Thus, recessions may generate corporate mergers
and increased concentration of the means of production.
The claim made for disinflationary policies is that in the short term the
economy suffers, but in the long term stable money allows for steady
growth and higher real wages. The problem is that short-term suffering
can be severe, especially when monetary policy is used to decrease de-
mand. While the jury is still out on the distributional impacts of moder-
ate inflation, the distributional impacts of unemployment caused by
disinflationary policies, as we will see below, are clear.

Unemployment
In the world of microeconomics, involuntary unemployment should not
exist. Prices are set by supply and demand, and when the demand for
labor is low, the price falls. At a lower wage, fewer people are willing to
work, and supply falls accordingly, returning the system to equilibrium.
Clearly, however, unemployment is a persistent problem in modern
economies. We particularly want to examine two issues: the link between
unemployment and inflation, and the implications of unemployment for
distribution.
Some unemployment is inevitable. People are constantly entering and
leaving the labor market, changing jobs and moving from place to place.
Businesses go bankrupt, or suffer downturns and lay people off. It always
takes time to find a new job. This is known as “frictional” or “natural” un-
employment. According to theory, if policy makers tried to reduce unem-

7This theory was originally introduced by Milton Friedman in his 1967 American Economics

Association presidential address.

8In the 1960s, economists found an inverse empirical relationship between unemployment

and inflation, which was dubbed the Phillips curve. But during the 1970s, a number of economies
experienced increasing unemployment and increasing inflation simultaneously.
 Chapter 16 The IS-LM Model • 295

ployment below this level, the result would be greater demand for a fixed
number of workers. Workers would have more bargaining power and
would demand higher wages, thereby causing inflation.7 Thus, a wide-
spread euphemism for “natural” unemployment is NAIRU, the non-
accelerating inflation rate of unemployment.8 There is considerable dis-
agreement over what NAIRU actually is. James K. Galbraith argues that
economists are really quite practical—their estimates of NAIRU simply re-
flect actual unemployment.9
But the link between low unemployment and inflation is not clear em-
pirically. Why not? We offer two explanations. First, in the era of global-
ization, large corporations are free to move their capital and production to
other countries. Even when unemployment is low, corporations can
counter demands for higher wages by a local workforce with the threat of
moving to a lower-wage country. This explains how the low unemploy-
ment of the 1990s in the U.S. was accompanied by stagnant wages and a
diminished share of national income going to wage earners.10 Second, we
must point out that income from production is divided between wages,
profit, and rent. Increased bargaining power by wage earners need not
lead to “wage-push” inflation—it could instead simply increase the share
of income going to wage earners and decrease the share going to rent or
profit. Does increased bargaining power by owners lead to “profit-push”
inflation?
In summary, then, low unemployment increases the bargaining
power of wage earners, which translates into higher wages (though this
effect is diminished by globalization). Higher wages can cause inflation,
which then erodes the higher wages, or it can change distribution pat-
terns between wages and profit. High unemployment, in contrast, in-
creases the bargaining power of corporations and leads to redistribution
toward the owners of capital. Whatever the validity of the theory behind
NAIRU, it is quite clear that monetary authorities pay close attention to
unemployment as an indicator of inflationary pressures. For example,
when unemployment falls too low, the Fed tends to raise interest rates
to reduce investment, employment, and demand. Distributional impacts
of inflation are uncertain, but unemployment caused by disinflationary
policies has clearly negative impacts on some of the poorest sectors in
society.
Finally, it is worth noting that increasing unemployment can set up a
vicious cycle. As people lose jobs, they lose money to purchase goods and
service. With less demand, businesses respond by reducing supply,

9J. K. Galbraith, Well, Excuuuuse Me! The International Economy, December 1995.
10R.J. Gordon, The Time-Varying NAIRU and Its Implications for Economic Policy, NBER
Working Paper No. W5735, May 1997.
296 • Macroeconomics

perhaps laying off more workers to do so, and further reducing demand.
Many fiscal policies such as welfare payments, unemployment insurance,
and other transfer payments are designed to diminish this impact, adding
stability to the economy. Economic stability is a public good, and an im-
portant policy objective.

The Impact of Policies on Scale, Distribution,

and Allocation
Now that we understand the basic elements of fiscal and monetary policy,
we can turn to their particular applications. How we apply these policies,
of course, depends on what we wish to achieve. Mainstream macroecon-
omists primarily pursue continuous economic growth, with a lesser em-
phasis on distribution. Allocation is left to microeconomic forces.
Ecological economists are primarily concerned with the impact of macro-
economic policies on scale (i.e., growth) but with a different goal than
mainstream economists—to make sure that the costs of additional growth
in material throughput are not greater than the benefits. Ecological econ-
omists assume that eventually the costs will exceed the benefits if they
haven’t already. They also place much more importance on distribution
than mainstream economists. In short, ecological economics strives to cre-
ate an economy in which there is no growth in physical throughput, while
avoiding the suffering caused by recession or depression. The allocation
of resources between market and nonmarket goods and services can play
an extremely important role in this regard.

Macro-allocation
As we have discussed earlier, free markets work very well at allocating re-
sources among market goods but very poorly at allocating nonmarket
goods, typically failing to provide them in satisfactory quantities. Many
policy makers already recognize this point, as can be clearly seen in gov-
ernment budgets, the bulk of which are spent on public goods such as de-
fense, health care, education, road systems, bridges, streetlights, national
parks, and so on.11 In fact, few institutions besides government allocate
resources toward nonmarket goods, and only the government is able to
use policy to reduce demand and hence expenditure for market goods and
shift it toward nonmarket goods.
For simplicity, we refer to the allocation between market and nonmar-
11National defense is generally considered a public good, though arms races, nuclear

weapons, and excessive defense expenditures may do more to undermine national security than
to ensure it. To the extent that disease is communicable and individuals are made uncomfortable
by the suffering of others, health care is also a public good.
 Chapter 16 The IS-LM Model • 297

ket goods as macro-allocation, and allocation among market goods as

Macro-allocation is the alloca-
micro-allocation.
tion of resources among mar-
In the private sector, monetary policy directly affects only the market
ket and nonmarket goods and
economy, by stimulating or discouraging investment in the production services.
and consumption of market goods for profit. Why is this so? Monetary pol-
icy acts primarily through its impact on interest rates and hence on bor-
rowing and lending. The private sector invests little in nonmarket goods,
since such goods generate no profit that can be used to pay back loans.
Therefore, lower interest rates will not affect the production of nonmarket
goods by the private sector. Not only will monetary expansion do nothing
to provide public goods and open access resources, it can actually increase
the degradation of these resources if the production of the market goods
is accompanied by negative externalities affecting the environment. Re-
turn to our example of the bowling alley. If lower interest rates induce it
to expand, it will not expand into a void and may expand into some
ecosystem—a wetland, for example—that currently provides valuable
nonmarket services to the local community. As we discussed earlier, such
negative externalities are an inevitable outcome of market production.
Therefore, if our policy objective is sustainable scale, monetary expan-
sion is very problematic. Even if the economic scale is well within the con-
straints imposed by the ecosystem, monetary expansion acts on only one
type of good, market goods. Market goods do not always offer the highest
marginal contribution to human well-being, however. The microeco-
nomic law of the equimarginal principle of maximization thus applies not
only to the scale of the economic system relative to the ecosystem that sus-
tains it, but also to the division of market and nonmarket goods produced
by an economy. In ecological economics, macro-allocation is every bit as
important as micro-allocation.
Theoretically, federal money in a democratic society will be directed to-
ward the goods and services that provide the greatest marginal utility for
society as a whole. As we have discussed, an important role of government
expenditure is to provide nonmarket goods.
It is important to distinguish between two classes of nonmarket goods,
which have different effects on scale. Manmade nonmarket goods affect
scale to the same degree as market goods. If the government project in the
bowling alley town is a big government building, it may also encroach
upon some valuable ecosystem and destroy the services it provides. In
contrast, what if the government project in the bowling alley town created
a park that provides recreation and restores wetlands that serve as filters
for the town’s water supply and as a buffer against catastrophic floods?
Protecting and restoring the ecosystems that provide nonmarket environ-
mental services can effectively decrease scale, or at least help ensure that
we do not surpass optimal scale. As the world becomes more full, the
298 • Macroeconomics

marginal benefits from protecting and restoring ecosystem funds, and

hence the nonmarket services they provide, will increase relative to those
from market goods and human made public goods. As this happens, and
if politicians better come to understand the benefits and public good na-
ture of ecosystem services, more and more federal money should be allo-
cated toward providing such services.
It is important to recognize, however, that government expenditure on
ecosystem funds can still increase scale. How is this so? We must remem-
ber that once the initial expenditure enters the economy, the multiplier ef-
fect takes hold. Money spent to restore ecosystem funds will in turn be
spent by its recipients on market goods—workers restoring the wetland
may spend their money on bowling, pressuring the bowling alley to ex-
pand. The larger the multiplier, the larger the impact on the market sec-
tor of the economy, and the less control the government has over
composition. As most macroeconomics textbooks explain, tax increases
decrease the multiplier and also reduce income. A smaller multiplier in-
creases the ability of the government to affect macro-allocation, and re-
duced income reduces scale. Taxes can also be used to discourage
undesired behaviors, such as pollution, and transfers can be used to en-
courage desired ones, such as environmental preservation. The full impact
of taxation on scale and macro-allocation depends on how the taxes are
spent, but taxes can certainly play an extremely important role in achiev-
ing an optimal scale—a point we will examine at greater length in Chap-
ter 21.
Another important point must be made here. Under traditional analy-
sis of the IS-LM curve, fiscal policy when the economy is operating at full
capacity results in crowding out (remember the full bowling alley) and
should be avoided. However, in terms of macro-allocation and scale, full
output conditions can increase the effectiveness of fiscal policy. With full
employment, if the government spends money to create a park and restore
wetlands, interest rates and labor costs go up, and it is more difficult for
the bowling alley to expand. (Fortunately, the park offers a recreational al-
ternative to bowling that does not displace ecosystem services.) Govern-
ment expenditure on restoring ecosystems under such conditions will
therefore have an unambiguous impact on reducing scale.
What are the distributional impacts of fiscal and monetary policy? Fis-
cal policy in the form of taxation and government transfers can be easily
and effectively distributed as desired. Government transfers such as wel-
fare, unemployment insurance, Medicare, Medicaid, and Social Security
all play an important role in distribution. Corporate welfare programs
12C. M. Sennott, “The $150 Billion ‘Welfare’ Recipients: U.S. Corporations,” Boston Globe, July

7, 1996.
 Chapter 16 The IS-LM Model • 299

(which outweigh transfer payments to the poor12) affect distribution in

the opposite direction. Public goods are available to all, and their provi-
sion improves distribution. In terms of income, progressive taxation can
help reduce gross inequalities in income distribution, a necessary condi-
tion morally and practically if we are to achieve a sustainable scale. Mon-
etary policy can also play an important but narrower role in distribution.
High interest rates caused by tight monetary policy can lead to unem-
ployment, and they favor creditors over debtors, as discussed above in the
section on inflation and disinflation. Low interest rates have the opposite
effect.
In summary, in terms of ecological economic goals, monetary policy is
a blunt instrument directed only toward the production and consumption
of market goods, with limited flexibility in terms of distribution and
macro-allocation. Expansive monetary policy increases scale. Fiscal policy
has far greater flexibility in terms of scale, distribution, and macro-
allocation.

■ IS-LM in the Real World

While the IS-LM model is very useful, it has important limitations.13 The
model is deceivingly simple and does an inadequate job of conveying the
real-life complexity of monetary and fiscal policies. While the model
shows the general impacts of such policies, it fails to incorporate the is-
sues of uncertainty, time lags, and structural changes, as well as the diffi-
culty of choosing the appropriate policy variables to manipulate.
Economists typically have a poor understanding of what is happening
in the economy at any given moment. Is unemployment too high? Is the
economy growing too fast, threatening inflation? Are we headed for re-
cession? Weathermen may be inaccurate at predicting the future, but at
least they can look outside right now and tell you what the weather is.
Economics is less advanced. Viewing the same data concerning the econ-
omy, economists frequently disagree on how to interpret them. For exam-
ple, in the United States, the Federal Reserve Bank raised interest rates in
May 2000, fearing that the economy was growing too fast, and continued
to voice fears of inflation through November 2000. Less than 2 months
later, the Fed initiated the first of several interest rate cuts designed to
ward off recession. Part of the problem is that the economic system is
13See J. R. Hicks, IS-LM: An Explanation, Journal of Post-Keynesian Economics III (2): 139–154

(Winter 1980–1981). Hicks, the originator of the model, expresses reservations about how well it
fits the real world once expectations and dynamics are recognized. He considers the model useful
for understanding the past, but less so for understanding the future. We agree, but feel that many
of Hicks’ caveats apply to all equilibrium models, and that, rare though it is among scholars, Hicks
was too hard on himself!
300 • Macroeconomics

evolving rapidly in response to technological, environmental, cultural,

and structural changes.
Compounding the difficulty of an inadequate understanding of the
economy are the time lags involved in policy. There are two types of lags:
lags in decision making (the inside lag) and lags between the time the de-
cision is made and the policy takes effect (the outside lag). In fiscal policy,
decisions such as tax cuts and expenditure increases are typically debated
at length. Both legislative and executive branches must agree, and appro-
priate legislation must be passed. The decision lag can therefore be sub-
stantial. Once the decision to increase or decrease expenditure has been
made and carried out, the outside lag may be relatively short, as such poli-
cies have an immediate effect on aggregate demand (though the full effect
of the multiplier will take some time). Tax cuts or increases, on the other
hand, have much slower results and are often not even felt until the next
tax year.
The Fed, in contrast, generally has a much shorter decision-making
lag. The Federal Open Market Committee (FOMC), responsible for Fed
Policy, meets about eight times a year.14 Policy is generally decided at the
meeting, and open market transactions can take place almost immediately.
However, the most relevant impact of these policies is on interest rates and
their effect on investment and consumption decisions. Investment deci-
sions are rarely spur of the moment; they generally have a long gestation
period. Thus, the Fed has a short decision-making lag, and a long lag be-
fore the policy takes effect.
These lags are very important to consider when deciding on a policy. It
is quite possible that by the time a decision is made and the resulting pol-
icy takes effect, the problem the policy was designed to address will have
disappeared, and a policy with the opposite effect may even be required.
Another problem is disagreement over what type of policy should be
pursued, and what the impact will be, especially for monetary policy. The
Fed usually tries to manipulate one of two targets: the money supply or
the interest rate. Not only is there considerable debate over which course
the Fed should pursue, there are serious obstacles to achieving either goal.
For example, as Alan Greenspan, current Chairman of the Fed, admitted
in congressional testimony: “. . . we have a problem trying to define ex-
actly what money is . . . the current definition of money is not sufficient
to give us a good means for controlling the Money Supply. . . .” 15
Psychology can also make it difficult to manipulate interest rates. As we

14The Federal Reserve Act mandates that the FOMC meet at least four times a year, and since

1981 it has met eight times a year.

15Congressional testimony, February 17, 2000.
 Chapter 16 The IS-LM Model • 301

discussed earlier, interest rates are ultimately determined by the bond

markets. Bonds, of course, mature in the future, and the amount someone
is willing to pay for a bond depends on their expectation of future infla-
tion. The Fed might implement an expansionary monetary policy to bring
down interest rates, but if bond marketers believe this expansion will in-
stead induce inflation or force monetary contraction in the near future, it
could paradoxically serve to drive interest rates up.
A final problem with policy in countries with independent monetary
authorities is the difficulty in coordinating between monetary and fiscal
policy. This problem can become acute when the monetary authority and
the government have different policy objectives. The elected government
is mainly concerned with growth and employment, two issues that affect
voters, and hence their elected representatives. In contrast, the Fed is
mainly concerned with “sound” money (i.e., no inflation) and has fre-
quently pursued this policy in the past, even when it has caused signifi-
cant hardship in the form of unemployment and worse.

Box 16-5 Why Is the Fed So Anti-Inflation?

From our discussion of inflation, it would seem that inflation is less
harmful than unemployment induced by anti-inflationary policies. Why
then is the Fed so anti-inflation? In answering this question, it is worth
bearing in mind who the natural constituency of the Fed is. Most mem-
bers of the FOMC are bankers or Wall Street professionals, and the Fed
seems to listen closely to the concerns of these groups. These two
groups form the bulk of the wealthy creditors who benefit from low infla-
tion and disinflationary episodes, and who are unable to increase their
share of national income as readily during inflationary periods.a
aW. Greider, Secrets of the Temple: How the Federal Reserve Runs the Country, New
York: Simon & Schuster, 1987.

Despite its shortcomings, the IS-LM model is a vast improvement over

prior models. It is a two-sector general equilibrium model, the sectors
being the real sector and the monetary sector. Before Hicks’ model, econ-
omists often tried to explain the interest rate as a purely monetary phe-
nomenon (liquidity preference and money supply) or a purely real
phenomenon (savings and investment). There was an investment rate of
interest and a money rate of interest, and confusion about which set of
factors “really” determined the interest rate. Hicks showed that the real
302 • Macroeconomics

and monetary sectors simultaneously interact to determine both the in-

terest rate and national income. But Hicks said nothing about the ecosys-
tem and biological rates of growth. In 1937 the world was still considered
“empty.” Thus, the IS-LM model treats all economic growth as identical—
it does not distinguish between government expenditures on market
goods, manmade public goods, or investments in ecological restoration,
nor does it address distribution.

■ Adapting IS-LM to Ecological Economics

How might the IS-LM model be adapted to ecological economics? Re-
membering our basic vision of the macroeconomy as a subsystem of the
finite and nongrowing ecosystem, the most obvious suggestion would be
to impose an external constraint on the model representing the biophysi-
cal limits of the ecosystem. For example, we could assume a fixed
throughput intensity per dollar of Y (i.e., GNP), so that a given Y in money
terms implied a given physical throughput. Then we could estimate the
maximum ecologically sustainable throughput, convert that into the
equivalent Y, and impose that as an exogenous constraint on the model.
Based on Figure 16.4, it would be represented by a vertical line at the Y
corresponding to maximum sustainable throughput. It would not be a
function of the interest rate at all.16 Let’s call the vertical line EC for “eco-
logical capacity.” It reflects a biophysical equilibrium, not an economic equi-
librium. It is ignored by the actors whose behavior is captured in the IS
and LM curves.17
The most obvious approach is not always the best, but it is usually a
good place to start. Also, this approach closely parallels the macroecono-
mist’s representation of full employment of labor as a perpendicular at the
level of Y corresponding to full employment at an assumed labor intensity
of GNP. Our EC line represents “full employment” of the environment at
an assumed throughput intensity of GNP. Later we will discuss further the
assumption of fixed throughput intensity.
Let’s consider the three possible positions of the biophysical equilib-

16A. Heyes, A Proposal for the Greening of Textbook Macro: “IS-LM-EE,” Ecological Economics

32 (1) (2000); P. Lawn, Toward Sustainable Development: An Ecological Economics Approach, Boca
Raton, FL: Lewis Publishers, 2001. Heyes and Lawn have proposed an EE curve corresponding to
the ecological limits discussed here that would be a function of the interest rate. Several tech-
nologies produce income, some of which require or degrade more natural capital than others. Less
natural capital-intensive technologies require investments and are thus more likely at lower rates
of interests. One problem is that the investments themselves would require natural capital.
17However, it need not be ignored by government, which affects the IS curve. The government

is perfectly capable of investing in environmental services produced by natural capital and other
nonmarket goods. But it is completely ignored by monetary policy (the LM curve), which acts on
the economy through its effect on interest rates and hence market goods.
 Chapter 16 The IS-LM Model • 303

Figure 16.6 • The biophysical equilibrium relative to the economic equilibrium.

rium relative to the economic equilibrium, shown in Figure 16.6. The first
case represents the “empty world” scenario. The biophysical limit is not
binding. The distance Y*C may be thought of as excess carrying capacity.
Most macroeconomists who use the IS-LM model would have this case in
mind, if indeed they thought at all about EC. If the distance Y*C is large
then there is for practical purposes of short run policy no point in con-
ceiving or drawing the EC line.
The second case is the “full world” (or overfull) scenario. The economic
equilibrium has overshot the biophysical equilibrium. The distance CY*,
the overshoot, can be given two interpretations. In neither case, however,
would it represent real income. In the first case we might think of it as a
purely monetary phenomenon, inflation. After hitting EC the real sector
is effectively at the end of its tether. Even though the monetary sector
keeps on churning, it only generates price increases for the same real in-
come C. The other interpretation assumes that the real sector keeps pro-
ducing real output, but by unsustainable drawdown of natural capital.
Thus, CY* would represent capital consumption counted as income. As
natural capital is consumed, the EC line eventually has to shift even far-
ther to the left, increasing the overshoot. Most ecological economists be-
lieve the second case to be a rather accurate description of the present
state of affairs. We seem to be avoiding short-term inflation by long-term
capital drawdown. Most conventional economists do not worry about
long-term capital drawdown and shifting the EC curve farther to the left
because they believe that knowledge is shifting EC to the right and
thereby restoring the empty world situation.
The third case represents a big coincidence under our assumptions.
For the economic equilibrium to coincide with the biophysical equilib-
rium would require either extraordinary good luck, or purposeful coordi-
nation and planning. There is nothing in the model to make it happen,
just as, currently, there seems to be nothing in our institutions or behav-
ior that would make it happen. In Chapters 20–23, we will discuss policy
changes that could theoretically lead to this outcome.
304 • Macroeconomics

Recall that we previously discussed the concept of full employment,

which we might represent by an FE limit for labor similar to the EC limit
for natural capital. Ideally, a FE labor line should coincide with the ISLM
equilibrium point—make IS = LM at full employment. If FE is beyond the
intersection of IS and LM, then policy makers might pursue FE through
growth in Y. But what if FE is beyond EC? The problem is no longer to
pursue FE by growth in Y, but instead through structural change, such as
shifting factor intensity away from fossil fuels and manmade capital (both
of which rapidly draw down natural capital) and toward labor. We have
already explained that when IS = LM beyond EC, we are likely to draw
down natural capital, and it is implicit in the acronym NAIRU (the non-
accelerating inflation rate of unemployment) that going beyond FE results
in inflation. Why the difference? Why doesn’t moving beyond EC also
simply cause inflation? The answer is that natural resources are either free
or cheap to begin with; they are not appropriately priced by the market
mechanism, and excessive use therefore does not affect the price signal.
It remains true, however, that the assumption of constant throughput
intensity of Y is troublesome. We know that throughput intensity of Y
changes with new technology, and with shifts in the mix of goods that
make up Y, even if probably not with factor substitution of capital funds
for throughput flows. Differing assumptions about throughput intensity of
Y can at least be represented by a shift in the EC perpendicular. (Curve
shifting is, as you have no doubt realized by now, not an uncommon de-
vice in economic analysis!) However, in terms of practical policy recom-
mendations, perhaps the best approach would be simply to impose the
ecological constraint as a limit on throughput. For any given technology,
a fixed limit on throughput will also limit Y, but over time, new tech-
nologies and a different mix of goods and services can allow Y to increase
without increasing throughput, thereby threatening the life-support func-
tions of the ecosystem.

BIG IDEAS to remember

■ Macroeconomic model vs. ■ IS-LM analysis of monetary
general equilibrium and fiscal policy
■ Real vs. monetary sectors ■ Comparative statics
■ IS = LM ■ Crowding out
■ MV = PQ ■ Inflation and disinflation
■ Transactions demand for ■ Unemployment
money ■ Macro-allocation
■ Liquidity preference ■ IS-LM adapted to show
■ Relation of bond prices to economic and biophysical
interest rate equilibria