Full text of Business Review (Federal Reserve Bank of Philadelphia) : March/April 1982

View original document

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

Federal Reserve Bank of Philadelphia

M ARCH APRIL 1982

Billions $

125 ---------------------------------------------------------------------

Federal Deficits*100- D Faulty Gauge
of Governments Impact
on Financial Markets
75---------------------------------------------------------

at the Right Price

MARCH/APRIL 1 9 8 2

FEDERAL DEFICITS:
A FAULTY GAUGE
OF GOVERNMENTS IMPACT
ON FINANCIAL MARKETS
Brian Horrigan and
Aris Protopapadakis

Federal Reserve Bank of Philadelphia
100 N orth S ixth Street
P hiladelphia, P en n sy lv an ia 19106

The BUSINESS REVIEW is published by
the Departm ent of Research every other
month. It is edited by John J. Mulhern, and
artw ork is directed by Ronald B. Williams.
The REVIEW is available w ithout charge.
Please send subscription orders and
changes of address to the Department of
Research at the above address or telephone
(215) 574-6428. Editorial communications
also should be sent to the Departm ent of
Research or telephone (215) 574-6426.
Requests for additional copies should be sent
to the Departm ent of Public Services.
*

The Federal Reserve Bank of Philadelphia
is part of the Federal Reserve System —-a

HEALTH CARE:
GETTING THE RIGHT AMOUNT
AT THE RIGHT PRICE
Laurence S. Seidman

System which includes twelve regional banks
located around the nation as well as the
Board of Governors in W ashington. The
Federal Reserve System was established by
Congress in 1913 prim arily to manage the
nation’s m onetary affairs. Supporting func
tions include clearing checks, providing coin
and currency to the banking system, acting
as banker for the Federal government, super
vising commercial banks, and enforcing
consumer credit protection laws. In keeping
with the Federal Reserve Act, the System is
an agency of the Congress, independent
administratively of the Executive Branch,
and insulated from partisan political pres
sures. The Federal Reserve is self supporting
and regularly m akes paym ents to the United
States Treasury from its operating surpluses.

FEDERAL RESERVE BANK OF PHILADELPHIA

Federal Deficits:
A Faulty Gauge
of Government’s Impact
on Financial Markets
by Brian Horrigan and Aris Protopapadakis*
In the ongoing debate about the impact of government borrowing on financial markets, the
focus usually centers on the size of Federal budget deficits. In the following article, the authors
argue th a t looking only a t th e d e fic it c a n m ak e for m islead in g conclusions about government’s
influence on the credit markets. They propose a more comprehensive measure which often
behaves differently than the Federal deficit. The views expressed here are those of the authors
and should not be identified as official views of the Federal Reserve Bank of Philadelphia or the
Federal Reserve System.—Donald J. M ullineaux, Senior Vice President and Chief Economist,
Federal Reserve Bank of Philadelphia.
more expenditure cuts for 1982, and these
projections have sparked a lively debate
within the Adm inistration on w hether to
propose sizable tax increases for 1983. Some
members of Congress continue to advocate
rolling back recent tax cuts or increasing
other taxes in order to reduce the deficit.
People are concerned about budget deficits
because they equate them w ith increased
government borrowing from the private sec
tor and increased government competition
with private investors. They fear that when
the U.S. Treasury borrows more, fewer
funds will be available for private invest
ment and interest rates will rise. But does a
bigger budget deficit necessarily mean that

New spapers and magazines frequently
w arn about the dangers of big Federal budget
deficits, claiming that the recent large deficits
have pushed interest rates to record highs.
The continuing debate over tax and expendi
ture cuts illustrates the importance many
people attach to Federal budget deficits.
Projections of large deficits appear to have
prompted the Adm inistration to request
*Brian Horrigan received his Ph.D. from the Univer
sity of California at Los Angeles and joined the Phila
delphia Fed in 1980. He specializes in monetary and
financial economics. Aris Protopapadakis is Research
Officer and Economist at the Philadelphia Fed. He
received his Ph.D. from the University of Chicago.

MARCH/APRIL 19 8 2

BUSINESS REVIEW

can borrow directly from the Treasury via the
Federal Financing Bank.1 The Treasury lends
to these agencies and to the Federal Financing
Bank by borrowing directly from the public.2
This kind of Treasury borrowing also does
not appear in the unified Federal budget.
Thus, even if the unified budget is balanced,
gross borrowing from the public can be large.
The annual increase in total Federal debt
includes all Federal borrowing, w hether the
Treasury is involved in it or not.3 Column 1 in
Figure 1 gives the Federal budget deficits as
reported by the Treasury while column 2 in
Figure 1 gives total Federal borrowing. The
data show that in some years total Federal
borrowing was over $20 billion more than the
Federal budget deficit.
In addition to off-budget borrowing, there
are other government obligations that should
be taken into account in a comprehensive
measure of the debt (see WHAT IS FEDERAL
DEBT? overleaf). Since it is not possible to
measure these obligations accurately, we do
not include them in the calculations that
follow. Adding accurate estimates of these
obligations to the measures of borrowing
developed here could change some of the
conclusions.

the government sector is a bigger drain on
credit markets? We argue that the deficit is not
a reliable indicator of governm ent’s drain on
credit markets. The Federal deficit is an
incomplete measure of government borrowing
because it does not include all government
borrowing. More importantly, all govern
ment borrowing must be adjusted for inflation
before it can be used as a gauge of govern
ment’s competition w ith private borrowers.
An alternative m easure which we call “gov
ernment net borrowing” accounts for all
government borrowing and is adjusted for
inflation to do a better job of gauging govern
ment’s drain on the credit markets.
GOVERNMENT GROSS BORROWING
So far as the credit m arkets are concerned,
what m atters is how much the government
sector borrows from the public. The Federal
budget deficit m easures only part of the
government sector’s borrowing activity.
Other government units and related bodies—
such as off-budget Federal agencies and state
and local governm ents—also compete for
funds in the credit m arkets by issuing their
own debt, and these agencies often lend
funds to the Treasury as well. To obtain the
right total, the borrowing of all government
units has to be added together and w hat they
lend to each other has to be subtracted out.
We label the resulting magnitude “govern
ment gross borrow ing.” Government gross
borrowing m easures the amount of money
the government sector borrows from the
public.
Off-Budget Agencies Borrow, Too, . . .
The Federal government borrows funds that
do not appear in the Federal budget. Federally
owned agencies, such as the Postal Service
and the Tennessee Valley Authority, have
the authority to borrow in the credit markets,
but their activity does not explicitly appear
anywhere in the unified Federal budget.
Also, some Federally sponsored agencies,
such as the Farmers’ Home Administration
and the Rural Electrification Administration,

■^For a detailed analysis of the Federal government’s
off-budget activities, see David Resler and Richard Lang,
“Federal Agency Debt: Another Side of Federal Bor
rowing,” Review, Federal Reserve Bank of St. Louis,
November 1979. Also see John Fialka, “ Growing Giant:
U .S. Lender, Bigger Than Citibank,” The Wall Street
Journal, December 15, 1981; and H. Leonard and E.
Rhyne, “Federal Credit and the ‘Shadow Budget’,” The
Public Interest, Fall 1981.
2For example, as of the end of June 1981, the Student
Loan Marketing Association (SLM A) owned $3.4 billion
of Federally guaranteed student loans. The SLMA pur
chased the loans by issuing debt. The Federal Financing
Bank (FFB) purchased the SLMA debt by issuing its own
debt, and the Treasury in turn purchased the FFB debt.
In effect, the Treasury borrowed money from the public
to lend to students.
^A more precise calculation would involve using the
market value of the new Treasury issues rather than
their par value. However, the differences between par
and market value are small.

FEDERAL RESERVE BANK OF PHILADELPHIA

FIGURE 1

ANNUAL INCREASES
IN TOTAL GOVERNMENT DEBT HELD BY THE PUBLIC
CAN BE QUITE DIFFERENT FROM THE BUDGET DEFICIT*
(1)

(2)

(3)

(4)

Year

Reported
Federal
Budget
Deficits

Increases in
Total
Federal Debt

Increases
in Total
Government
Debt

Increases in
Privately Held Total
Government Debt
(Gross Borrowing]

1981

61.6t

98.Of

119.3t

89.Of

1980

61.2

84.5

108.9

90.3

1979

14.8

54.5

72.7

43.4

1978

29.2

68.5

90.9

60.2

1977

46.4

63.7

80.9

53.2

1976

53.1

77.9

91.0

63.0

1975

69.3

83.6

97.2

83.3

1974

11.5

23.3

38.1

24.7

1973

5.6

20.4

33.3

11.9

1972

16.8

25.1

39.3

26.0

*In billions of dollars. All figures are reported on a calendar year basis.
tPreliminary estimates.

SOURCES:
Federal deficits are from the Economic Report of the President 1982. Deficits are calculated by the NIPA
method, which is based on accrual, unlike the unified budget deficit, which is based on cash flow.
For 1972-76, Federal debt outstanding, Federal debt held by agencies, Federal debt held by state and local
governments, and Federal debt held by the Federal Reserve are taken from the Annual Statistical Digest (19701979). After 1976, these data are taken from the Federal Reserve Bulletin, January 1982.
State and local government data are taken from the Flow of Funds Outstanding, September 1981. State and
local debt outstanding data are from p.39, line 2, while internal holdings of state and local debt and holdings of the
retirement funds are from p. 39, lines 9 and 15 respectively.

BUSINESS REVIEW

MARCH/APRIL 1 9 8 2

WHAT IS FEDERAL DEBT?
In this article, we define the Federal debt as the sum of all the notes, bonds, and bills issued by the
Treasury and other Federally owned agencies. But is this all the Federal debt? Debt is nothing more
than an obligation, and the Federal government has many obligations that do not take the form of
Treasury debt. An important example of obligations not included in the Federal debt is the Federal
program of loan guarantees for private debt. The Federal government guarantees hundreds of
billions of dollars of private loans against default risk, and it also has assumed hundreds of billions of
dollars’ worth of insurance commitments. According to the Treasury (as reported by U. S. News and
World Report, May 4, 1981), Federally guaranteed private loans were $323.6 billion in 1980, and
Federal insurance commitments were $2,217.4 billion.
The majority of the loan guarantees are for mortgages and housing loans ($219.7 billion). It would
be absurd to add private mortgages to the national debt just because the Federal government
guarantees the mortgages. If, by chance, none of the mortgages defaulted, the guarantees would cost
the Treasury nothing. But if all of the mortgages defaulted, the Treasury would be stuck with having
to pay off all of the mortgages. It would also end up owning the housing behind these mortgages. A
sound strategy for the Treasury is not to include loan guarantees in the Federal debt; instead it could
create a sinking fund to cover loan defaults, and make a fixed payment into the sinking fund every
year. The annual payment would have to be large enough to keep the fund liquid and should be
adjusted with the default experience. That way, the cost of these guarantees would appear in the
budget, and Congress and the public would be forced to recognize and deal with the cost of loan
guarantees. The same principle applies to insurance commitments.
Another serious problem with measuring the Federal debt concerns the actuarial deficits of the
retirement and compensation programs of the Federal government. The Federal government
obligates itself to pay retirement benefits to members of the armed forces and the Civil Service. It
cannot morally renege on those obligations. If the government does not fund the retirement programs (as
private pension and life insurance programs do), then the debt of the Federal government increases—
that is, the government has committed itself to pay benefits for which it doesn’t have funds. In 1980,
the actuarial deficit of retirement and compensation programs (military, Civil Service, veterans,
railroad, Foreign Service, Public Health Service) was estimated at $631 billion. These liabilities are
part of the Federal debt and should be included in it. If the government commits itself to funding
these liabilities fully, then it should create an asset position that exactly offsets its total pension
liabilities. We have not included unfunded pension liabilities in the estimates of government net
borrowing only because the estimates of the actuarial deficits are unreliable.
The above principle does not apply to Social Security. Social Security benefits and taxes are set by
Congress and may be changed at any time. The $1,464-billion actuarial deficit of the Social Security
trust funds in 1980 only indicates that Social Security needs reform, not that the Federal debt is
mismeasured. Changes in the law could easily eliminate the entire actuarial deficit of the Social
Security Administration.

available funds. Therefore, from the view
point of the private credit markets, the correct
measure of government borrowing must
include Federal, state, and local government
borrowing, not Federal borrowing alone.
Column 3 in Figure 1 shows the annual
borrowing of the combined Federal, state,
and local governments for the past decade.

. . . As Do State and Local Govern
ments. Even adding in the off-budget Federal
agencies doesn’t give a complete picture of
government borrowing. A large portion of
government financing activity occurs at the
state and local levels. It does not m atter to
private borrowers whether the Federal, state,
or local government competes w ith them for

FEDERAL RESERVE BANK OF PHILADELPHIA

The consolidated government borrowing is
always larger than Federal borrowing alone,
and it is much larger than the Federal deficits.
For instance, though the 1979 Federal deficit
was less than $15 billion, total government
borrowing was almost $73 billion. But not all
of the increases in the Federal, state, and
local debt represent a drain on private credit
markets; some of this debt is purchased by
Federal agencies, by the Federal Reserve
System, and by state and local governments.
Not All Government Debt Is Held by the
Public. A sizable portion of Federal debt is
currently owned by Federal agencies,
primarily the Social Security Administration.
Since Social Security receipts almost always
exceed outlays (they have in 9 of the last 10
years), the Social Security Adm inistration
purchases more Federal debt each year. Debt
issued by the Treasury doesn’t affect the
credit m arkets if it is purchased by a Federal
agency such as the Social Security Administra
tion. Thus, increases in debt holdings of
Federal agencies must be subtracted from the
total increase in Federal debt. Increases in
the Federal Reserve System holdings of
Treasury debt must be subtracted for the
same reason.4
And so m ust holdings of state and local
governments. These governments typically
are prohibited by their constitutions from
running current account deficits. On average,
they run surpluses which they often use to
purchase their own debt and Treasury debt.
To gauge the impact of government borrowing

in the credit m arkets, increases in state and
local government debt holdings must be sub
tracted from the total increase in government
debt as well.
The calculations for 1980 illustrate the
magnitude of the adjustments discussed
above. In 1980, Federal debt increased by
$84.5 billion while the state and local debt
increased $24.4 billion, for a total increase of
$108.9 billion. Of this increase, the Fed
purchased $3.8 billion, Federal agencies
purchased $5.4 billion, and state and local
governments purchased an additional $9.4
billion. Thus, only the remaining $90.3 billion
of government debt was available for pur
chase by the public.
Column 4 in Figure 1 shows the increases
in the consolidated government debt held by
the public—government gross borrowing.
This borrowing is always larger than the
reported Federal budget deficit, but in some
cases it is smaller than the increases in total
Federal debt. Gross borrowing is smaller
than increases in the Federal debt whenever
agencies, the Federal Reserve, and state and
local governments buy back more debt than
they issue.
Gross borrowing is an accurate measure of
the money government borrows from the
public to finance its expenditures. Compared to
this measure, Federal deficits understate the
amount of money government borrows. But
even gross borrowing may be an inadequate
and misleading m easure of the government
sector’s impact on credit markets, because

4The case for subtracting debt held by the Federal
Reserve is less clear cut than that for Federal agencies
and state and local governments. The Federal Reserve
annually purchases a certain amount of Treasury debt,
and in that respect it acts just like a Federal agency. It
purchases this debt, however, by selling new reserves to
the banking system. One could argue that the Federal
Reserve is only converting interest-bearing Treasury
debt to non-interest-bearing Federal Reserve debt, and
that this debt represents as much of a demand on the
credit markets as Treasury debt. Those who believe that
government borrowing can crowd out private invest-

ment assume that consumers consider purchases of
government debt and private corporate debt equivalent.
Consumers do not realize that excess government debt
may mean increased future taxes. There is not much
disagreement, however, that individuals do not view
purchases of bonds (government or private) and money
as being equivalent. Thus the response of the financial
markets to increases in the supply of reserves (and
consequently money) will be different than their
response to increases in the supply of government
bonds, so that reserves and government debt should not
be added together.

MARCH/APRIL 19 8 2

BUSINESS REVIEW

sector without affecting consumption and
investment, because the inflation premium
makes enough funds available to finance the
additional borrowing. Therefore, judging the
impact of government borrowing in the credit
markets without accounting for the effect of
inflation is highly misleading. In fact, two
economies can be identical in real terms, but
if they experience different inflation rates,
the government deficits and the amounts of
new debt the two governments must issue
can behave very differently.
Figure 2 gives an example of two such
hypothetical economies. Transylvania and

gross borrowing greatly depends on the in
flation rate. Gross borrowing seriously over
states government’s impact on credit markets
when prices are rising, because inflation
increases the interest rate government must
pay on its debt while it reduces the real value
of government bonds held by the public.
GOVERNMENT BORROWING
AND CREDIT MARKETS:
WHAT’S THE CONNECTION?
A higher inflation rate automatically results
in larger government gross borrowing,
because interest rates are higher when infla
tion is higher. But does an inflation-induced
rise in government borrowing mean that the
government is competing for more funds in
the credit m arkets? Only when gross bor
rowing rises more rap idly than prices is
government a drain on the credit markets.
Therefore, gross borrowing figures need to
be adjusted for the effect of inflation to get a
good m easure of governm ent’s impact on
credit markets.
As inflation increases, the interest that
government pays on its debt rises.5 The
higher interest compensates bondholders for
the inflation-caused erosion of the real value
of their bonds (see INFLATION AND
INTEREST RATES). If these people are to
restore the purchasing pow er of their bondholdings, they must use the portion of the
interest paym ent that compensates them for
inflation—the inflation prem ium —to pur
chase additional bonds. Therefore, increases
in government debt that keep the real value
of the debt constant don’t add to government’s
claims on the financial resources available
for private investment.
Inflation causes government borrowing
requirem ents to increase. But this increased
demand for funds can be met by the private

INFLATION
AND INTEREST RATES
Interest rates, including those on govern
ment debt, are influenced by inflation because
interest involves payment in the future, and
tomorrow’s dollars may be worth far less in
terms of goods and services than are today’s
dollars. For example, if a $100 loan today is
repaid with $102 a year from now, the nominal
interest rate on that loan is 2 percent. If there
is no inflation, the 2 percent is also the real
interest rate—real because $102 buys 2 per
cent more goods than $100 does. But if there is
inflation, the real interest rate differs from
the nominal interest rate. Inflation causes the
purchasing power of the dollar to depreciate;
future dollars buy fewer goods than current
dollars. Lenders want compensation for any
expected depreciation of their dollars caused
by inflation. If anticipated inflation rises
from zero to 10 percent, for instance, the
nominal interest rate must increase by 10
percentage points (to 12 percent) just to hold
the purchasing power of the principal
constant. Only in this way will the real interest
rate remain at 2 percent; 12 percent more
dollars ($112) buys 2 percent more goods after
the price level rises by 10 percent. The
additional $10 of interest payment (the
inflation premium) doesn’t represent real
income, because it only offsets the lost pur
chasing power of the $100 principal.

5The Federal government alone has accumulated a
large debt ($1 trillion), and a significant part of its budget
goes to interest payments on this debt (almost $98 billion
in fiscal 1981).

FEDERAL RESERVE BANK OF PHILADELPHIA

Ruthenia have the same unchanging real
(inflation-adjusted) consumption and invest
ment, real interest rates, real government
purchases and taxes, and real national debt.
The two economies have different rates of
inflation, though. Transylvania has no in
flation, while Ruthenia m aintains a steady
10-percent rate of inflation. Every year,
Ruthenia’s nom inal consumption and invest
ment, nom inal government purchases and
taxes, and nom inal debt rise by 10 percent,
but in real terms nothing changes. Transyl

vania has a balanced budget, while Ruthenia
has an ever increasing budget deficit and
increasing gross borrowing. Yet this budget
deficit (or gross borrowing) has no impact on
the Ruthenian economy because the real value
of government debt does not change. The
budget deficit (100 billion Ruthenian dollars
in the first year) is exactly equal to the inflation
premium the government pays on its debt,
and it serves to keep the real value of the debt
constant.
The quantity that correctly measures the

FIGURE 2

INFLATION MEANS THAT TWO ECONOMIES
CAN BE IDENTICAL IN REAL TERMS,
BUT HAVE VERY DIFFERENT BUDGET DEFICITS*
(3)

(4)

(5)

(6)
(7)
(8)
Govern- GovernGovernment
ment
ment
Expenditures
Gross
Net
Private
for Goods
Interest
Budget Government Borrow- Borrow- Consumption
Year
& Services
Payments Taxes Deficit
Debt
ing
ing
& Investment
(2)

(1)

TRANSYLVANIA
Inflation 0%, Nominal & Real Interest Rate 2%
1
2
3

600
600
600

20
20
20

620
620
620

0
0
0

1,000
1,000
1,000

0
0
0

2,400
2,400
2,400

RUTHENIA
Inflation 10%, Nominal Interest Rate 12%, Real Interst Rate 2%
(real values in parentheses)
1
2
3

600 (600)
660 (600)
726 (600)

120
132
145.2

620
682
750.2

100
110
121

1,000 (1,000)
1,100 (1,000)
1,210 (1,000)

*In billions of Transylvanian and Ruthenian dollars.

100
110
121

0
0
0

2,400 (2,400)
2,640 (2,400)
2,904 (2,400)

BUSINESS REVIEW

MARCH/APRIL 1982

impact of government borrowing on the credit
markets of both economies is government
net borrowing, shown in column 7, Figure 2.
Government net borrowing is the change in
the real value of the government debt,
expressed in current dollars. While gross
borrowing is very different for the two
countries, net borrowing is the same, reflecting
the fact that the two economies are identical
except for inflation.
But how is Ruthenia’s inflation-induced
government gross borrowing financed with
out causing a drain on the credit m arkets?
The households in Ruthenia provide the
funds by saving the inflation premium
component of the interest payments on
government debt. This is the only saving
strategy that allows them to m aintain both
the real value of their consumption and the
real value of their w ealth in the face of rising
prices. Thus the increase in the dollar savings

of the households is just equal to the dollar
increase in government borrowing, leaving
both real savings and real investment un
changed. A numerical example of a typical
Ruthenian household may serve to illustrate
the case.
Consider a family w ith wage income of
$25,000 and accumulated savings of $20,000,
all invested in one-year government bonds.
Suppose there is no inflation and the interest
rate is 2 percent, resulting in $400 of interest
payments. To simplify the example assume
that this family consumes all its wage and
interest income—it undertakes no new saving.
Over time, its assets (bonds) rem ain at
$20,000 and its consum ption at $25,400
(Figure 3, panel a).
If inflation suddenly increases to 10 per
cent and is expected to stay there, the interest
rate rises to 12 percent (fully reflecting
inflation), and the family’s wages rise at the

FIGURE 3

TO KEEP REAL CONSUMPTION CONSTANT,
HOUSEHOLDS MUST SAVE MORE
WHEN THERE IS INFLATION
(1)
Wage
Year Income

(2)
Interest
Income

(3)

(4)

(5)

Current
Current
Total
Value of
Value of
Income Consumption Saving

(6)

(7)

(8)

25,400
25,400
25,400

20,000
20,000
20,000

20,000
22,000
24,200

20,000
20,000
20,000

Current
Real Value of Value Real Value
Consumption of Assets of Assets

(a) Inflation 0%, Interest Rate 2%

1
2

25,000
25,000
25,000

400
400
400

25,400
25,400
25,400

0
0
0

(b) Inflation 10%, Interest Rate 12%
1
2
3

25,000
27,500
30,250

2,400
2,640
2,904

27,400
30,140
33,154

25,400
27,940
30,734

2,000
2,200
2,420

25,400
25,400
25,400

FEDERAL RESERVE BANK OF PHILADELPHIA

to increases in net borrowing and therefore
would not represent a drain on credit markets.
Net borrowing is the correct gauge of any
potential crowding out of private borrowers
from the credit m arkets.7
The argument so far is made as if inflation
is fully anticipated. But, realistically, inflation
is never fully anticipated, and forecasts of
inflation are often far off the mark. Under
these circumstances, is government net bor
rowing still the correct measure of the govern
ment’s impact on the credit m arkets? As
discussed in detail in the Appendix, govern
ment net borrowing is a correct measure
even when inflation is not fully anticipated.

10-percent inflation rate [Figure 3, panel b).
For the first year, the family’s total income is
higher because of the higher interest rates.
Can this family still consume all its income
and m aintain the purchasing power (real
value) of its assets? Obviously not, because
inflation erodes the purchasing power of its
bonds. If this family consumed all its new
income, by the end of the third year its assets
would be worth only $16,529 in today’s
Ruthenian dollars. Instead, it must save the
inflation premium built into the nominal
interest rate and buy more government bonds
w ith that money. Only this behavior will
allow the family’s real consumption and its
real assets to rem ain the same as before.
Figure 3 (panel b) shows the details of the
family’s new saving strategy. The key point
is that the inflation premium built into in
terest rates is not truly income. Rather, it
compensates investors for the loss of the
purchasing power of their nominal invest
ments (bonds). The family in the example
must save all of the inflation premium com
ponent of the interest payments to keep its
real w ealth constant. In dollar terms (though
not in real terms), this family is saving more
than it used to, making more funds available to
buy government bonds.
The examples about government and house
hold finances show that inflation causes
budget deficits and government gross bor
rowing to increase. But this increase can be
exactly met by an equal increase in the dollar
savings of the households.*6 Thus, though
such inflation-induced deficits may seem
alarmingly large, they are not due necessarily

IS GOVERNMENT A NET BORROWER?
With an inflation-adjusted measure of
government borrowing, it is possible to find
out w hether the government sector might be
crowding out private investment by calcu
lating the net borrowing of governm ent.8
Columns 1 and 2, Figure 4 (overleaf), show
Federal net borrowing and total net bor
rowing, respectively. These figures show
that government net borrowing has been far
smaller than the Federal deficit or gross

simplifies that task, without changing the conclusion.
Another feature of our example is the absence of taxes
on interest income. That omission is readily remedied by
thinking about these rates of interest as after-tax rates.
7See G. V. Jump, “Interest Rates, Inflation Expec
tations, and Spurious Elements in Measured Real Income
and Savings,” American Economic Review, December
1980, and J. Siegel, “Inflation-Induced Distortions in
Government and Private Saving Statistics,” Review of
Economics and Statistics, February 1979 for a similar
analysis. The Economic Report of the President 1982
also adjusts deficits for inflation. See Chapter 4,
Appendix.

6The examples in the text and in the appendix assume
that inflation is neutral—that is, real GNP, the real rate
of interest, and real investment are not affected by infla
tion. Given the current structure of tax laws it is highly
unlikely that inflation is neutral in the U .S. However,
though inflation may cause some real variables to change
at the same time as it increases deficits, we try to focus
on the deficits and theirimpact, leaving out the effects of
inflation on the economy. Assuming neutrality greatly

6 T o compute net borrowing, we use a price index to
deflate the end-of-year gross debt. This procedure gives
an estimate of real debt. The annual change in real debt
gives real net borrowing; multiplying that by the price
index gives net borrowing in current dollars. The price
index is the geometric average of the GNP deflators for
the last quarter of the year and the first quarter of the
following year.

BUSINESS REVIEW

MARCH/APRIL 1982

FIGURE 4

NET BORROWING GENERALLY HAS BEEN SMALL
RELATIVE TO INVESTMENT*
(1)

(2)

(3)

(4)

F e d e ra l

G ov ern m en t

N et P riv ate

N et G o v ern m en t

Net Borrowing

Investment

2 3 .2 t
1 9 .7
-1 2 .9
1 .3
1 3 .8
3 7 .7
5 6 .0
-1 5 .6
-1 8 .5
1 .4

1 9 .Of
1 6 .2
-1 4 .6
3.0
1 7 .0
3 6 .8
5 2 .8
-2 0 .0
-1 8 .2
9.5

1 3 0 .2 t
1 3 2 .6
1 9 3 .5
1 8 6 .6
1 5 4 .5
1 1 9 .0
8 9 .2
1 0 5 .4
1 4 5 .6
1 1 5 .5

Total

1981
80
1979
78
77
76
75
74
73
72

Total

2 6 .5
3 2 .9
2 3 .7
1 7 .4
1 2 .0
1 5 .5
1 8 .2
1 8 .6
1 6 .4
1 6 .3

’ Billions of dollars.
tBased on most recently available estimates.
SOURCE: Survey of Current Business. Net real government investment is the annual change in the net physical
capital stock owned by the government sector as reported in the National Income and Product Accounts. This
capital stock includes all equipment and structures owned by Federal, state, and local government and govern
ment-owned enterprizes. Net private investment, column 3, is calculated by adding the net private investment
shown in the National Income and Product Accounts (Gross Investment minus capital consumption allowances)
to net consumption of durable goods. Net consumption of durables is calculated by applying a 20-percent
depreciation rate to the stock of durables and subtracting that from durables consumption in the N ational Income
and Product Accounts.

borrowing figures would suggest. Often net
borrowing is negative: the public reduced its
real holdings of government debt in those
years W hen net borrowing is negative,
government in effect supplements savings
available for private investm ent.9
The figures show that government net

borrowing was substantial only during the
1975 recession and the ensuing recovery.
There is some government net borrowing
also in 1980, the year of a sharp, but short
lived, downturn. It is not surprising that net
borrowing, especially Federal net borrowing,
rises during recessions; the increase in bor-

9T o the extent that inflation is fully anticipated,
negative net borrowing implies a flow of funds to the
public. If inflation is completely unanticipated, there is
no actual flow of funds. However, the unanticipated

capital loss on government bonds will cause households
to save more out of their income to rebuild their wealth
position. Thus negative government net borrowing in
effect increases the supply of private savings.

FEDERAL RESERVE BANK OF PHILADELPHIA

either private investment or government
investment. It is difficult to see how these
relatively small amounts of net borrowing
could have caused the record high interest
rates experienced recently.
Using the concept of government net bor
rowing can help put the projected budget
deficits in perspective. The Adm inistration’s
most recent forecast is a $97-billion deficit
for calendar 1982. This deficit is by far the
largest ever. Nonetheless, this large deficit
represents only about $46 billion in Federal
net borrowing according to our estim ates.11
By historic standards $46 billion of net bor
rowing is large, but it is much less (47-percent
less in real terms) than Federal net borrowing
was in 1975—another recession year. Such
large net borrow ing—and a budget deficit—
would only be a problem if it persists after the
economy comes out of the recession.

rowing coincides with the recession-induced
decline in tax revenues.10
One w ay to assess the potential impact of
government net borrowing on the credit
m arkets is to compare it to net private in
vestm ent (see column 3, Figure 4). The data
show that net government borrowing was
very small relative to net private investment
in the last decade. Thus the potential drain of
government on the credit m arkets has been
relatively small. For instance, in 1980 net
government borrowing was only 12 percent
of net investment and in 1978 it was less than
2 percent. Only during the 1975 recession
was government borrowing large relative to
private investment, and that was a result
mainly of the recession.
Another way to gauge the significance of
government net borrowing is to compare it to
government net investment. Net government
investment m easures the net addition to the
physical capital stock (items such as buildings,
bridges, highways, and defense installations)
owned by the Federal, state, and local govern
ments. These data are shown in column 4,
Figure 4. Government net borrowing is
considerably smaller than government net
investment, except during periods of recession.
Government has been collecting more taxes
than it needs in order to finance its current
expenditures. All of net borrowing and some
tax revenues go to finance government in
vestment projects—a situation which raises
policy issues (see SHOULD GOVERNMENT
INVESTMENT PROJECTS BE FINANCED
WITH TAXES? overleaf).
The results of our analysis show that the
size of government net borrowing usually
has been small compared to the amount of

CONCLUSION
M any people are concerned that large
Federal deficits cause high interest rates and
crowd out private investment. W hatever the
validity of the crowding-out hypothesis, the
unified Federal budget deficit simply is not
the appropriate m easure of government’s
drain on credit markets. The unified Federal
budget deficit does not include the borrowing
of off-budget Federal agencies and of state
and local governments, nor does it exclude
the debt purchased by government agencies,
by state and local governments, and by the
Federal Reserve System. Most importantly,
the meaning of the Federal deficit is distorted

^ P ro jectio ns of Federal borrowing for 1982 are from
Borrowing and Debt Special Analysis E, released by the
Office of Management and Budget. Since detailed 1982
estimates of Federal Reserve, state, and local holdings
of Federal debt are not available, we assume that these
institutions will behave as they did in 1981. Thus, as a
result of a projected increase in Federal debt of $131.3
billion, public holdings have to rise by $90.8 billion. We
also adopt the consensus forecast that the GNP deflator
will grow by 7.3 percent in 1982.

10If the government were to try to hold down its net
borrowing by reducing its expenditures and raising
taxes during a recession, it would destabilize the
economy unnecessarily, and a deeper recession could
result. The potential impact of net government borrowing
must be evaluated over the business cycle and not year
by year.

BUSINESS REVIEW

MARCH/APRIL 19 8 2

SHOULD GOVERNMENT INVESTMENT PROJECTS
BE FINANCED WITH TAXES?
When a private firm undertakes an investment project, it does not usually suspend dividends and
try to finance the project internally. If the firm’s credit standing is good and the proposed project is
expected to be profitable, it borrows in the market or issues new equity; the new investment
generates new cash flows sufficient to pay the additional dividends and interest.
Investment projects, whether private or public, are undertaken because they are expected to yield
benefits that exceed the cost of building and maintaining them. The difference between private and
public investment projects is that while private projects will be undertaken only when their financial
benefits exceed their cost, this rule need not hold for public investment. For example, a local
government may decide to build a bridge to alleviate traffic congestion. The local government could
finance the bridge from additional tax revenues. But the appropriate financing strategy is to borrow
the initial cost of the project and plan to pay for the real portion of the interest charges, for
maintenance, and for depreciation with future taxes or tolls. The project will eventually be paid for
in either case, but debt finance matches the tax payments the community makes to the benefits it
receives more closely than immediate tax finance.
The reason that the government should not finance investment projects with current taxes lies in
the role taxes play in the economy. While taxes raise revenues for the government, they also affect
the decisions individuals make about labor supply and saving. Evidence suggests that an increase in
income and profits taxes decreases saving and labor supply moderately.* Financing investment
projects from current taxes means that tax rates are higher than they need be, unnecessarily reducing
incentives to produce and save.
The Department of Commerce has estimated the net investment of the Federal, state, and local
governments.t Column 4 in Figure 4 shows that government net investment substantially exceeds
government net borrowing except during the 1975 recession and the 1980 downturn. For the last ten
years government net borrowing has covered only part of new government investment. The sum of
government net borrowing from the private sector from 1972 through 1979 amounts to $49 billion (in
1972 dollars), while the sum of government net investment is $138 billion (in 1972 dollars). Thus a
large part of these investments has been and is continuing to be financed by current taxes. This has
meant higher taxes and higher tax rates than necessary. +The economy could benefit from lower tax
rates that would result from financing government investments through borrowing from the
public.

‘ See Aris Protopapadakis, “Supply-Side Economics: What Chance for Success?” Business Review, Federal
Reserve Bank of Philadelphia, May/June 1981.
tThe Department of Commerce provides estimates only of the physicalcapital stock owned by the government.
These estimates do not include financial assets purchased by the government. This exclusion is particularly
important for our estimates, because our net borrowing includes off-budget agencies. Some of these agencies (for
instance the SLMA) purchase financial assets. However, it is very difficult to estimate the market value of these
assets and we do not include them in our net investment figures.
fW e do not argue here that the taxes collected should always be equal to current expenditures and transfers.
Whether optimal revenue raising involves budget deficits or surpluses is not known, because the information
necessary to decide that issue is very difficult to find. We only argue that paying for capital projects with current
taxes is not an optimal strategy. Furthermore, it is generally agreed that the government should not adjust its
taxes and expenditures every year so as to keep its net borrowing constant every year. Rather, the government
should allow net borrowing to rise and fall over the business cycle.

FEDERAL RESERVE BANK OF PHILADELPHIA

show that governm ent has not been a signi
ficant drain on the credit m arkets. Looking to
the future, it is clear that as long as inflation
persists, government can run substantial
budget deficits w ithout crowding out private
investment. But as inflation and inflationary
expectations fall, budget deficits will fall with
out any expenditure cuts or tax increases.

by inflation. The inflation of the last decade
caused interest rates to rise and therefore
caused budget deficits to balloon. These large
deficits do not represent necessarily a drain
on the credit markets.
Government net borrowing is a better
measure of the government sector’s impact
on credit m arkets. The net borrowing figures

APPENDIX. . .

•

APPENDIX

THE CASE
OF UNANTICIPATED INFLATION
The examples in the main text on the relationship between inflation, interest rates, and govern
ment budget deficits assume that inflation is always fully anticipated. But a 10-percentage-point rise
in the inflation rate raises the nominal interest rate from 2 percent to 12 percent only if the public fully
anticipates the inflation, and then only if inflation is neutral. If increases in inflation are not fully
anticipated, the reported budget deficits will not rise sufficiently to hold the real national debt
constant. At the same time, an unanticipated increase in the price level imposes a windfall loss on
bondholders.*
The wealth loss imposed on holders of government bonds by unanticipated inflation is a wealth
gain for the government. An inflation-induced drop in the real value of government bonds is
equivalent to an increase in the taxes of the bondholders. The real value of the outstanding debt falls,
but the interest rate is not high enough to compensate the bondholders for this loss.
The thesis of our article—that the proper measure of the impact of government borrowing is given
by the change in the real value of total government debt—does not depend on whether or not inflation
is unanticipated. It is easiest to see why by considering again the inflationary economy of our
example, Ruthenia.
If the Ruthenian inflation is anticipated, the additional financing needs of the government equal
the inflation premium of the interest payment—$100 billion. But what if the inflation is not
anticipated at all? As long as the government takes no action, there would be no budget deficit and
the net borrowing would be -$100 billion. This sum is the same as the purchasing power loss suffered
by the bondholders. If the government uses net borrowing as a guide for its fiscal policy and tries to
keep net borrowing constant, it would attempt to return to its original net borrowing, $0 in this
example. It can do so by either increasing transfer payments or cutting taxes and running a $100billion budget deficit. If it cuts taxes by $100 billion, individuals in the economy who suffered capital
losses on their bondholdings will use these unanticipated taxes to restore their portfolio without
changing their consumption or saving plans (taxes are unanticipated because the inflation was
unanticipated.) But since the government, by running a $100-billion deficit, is providing the right
quantity of bonds the public needs for the rebuilding of portfolios, consumption and investment will
remain the same, whether or not the inflation is anticipated.
To the extent that each individual is different, the capital losses on bonds will not be exactly offset
by the tax breaks or by the increases in transfer payments for each individual. Thus, any government
action to offset the impact of unanticipated inflation will alter the distribution of wealth and
probably the value of the real variables in the economy, which may be legitimate cause for concern.
Under these circumstances, government net borrowing may not be the only information necessary to
gauge government’s impact on the credit markets.

*If, for example, bondholders require a 2-percent real return on their investment and they expect a 6-percent
inflation rate, the nominal interest rate would be 8 percent. Should the actual inflation rate turn out to be 10
percent, the bondholders realize a real return on their investment of -2 percent.

FEDERAL RESERVE BANK OF PHILADELPHIA

Health Care:

Getting the Right Amount
at the Right Price
by L au ren ce S. Seidm an*

ing skepticism. A m arket-oriented approach
recommends itself more and more to policy
makers. The next few years promise to
witness a fundam ental clash of opposing
strategies, which may decide the direction of
the health sector for the rest of the century.

During the past decade, the consensus
among health policymakers appeared to be
that significant government regulation and
planning would be required to hold down the
cost of health care and make it more easily
accessible to the less well-off members of
society. Although full-fledged national health
insurance w as not enacted, important steps
were taken in that direction. The issue for
many policym akers was not w hether com
prehensive NHI should be legislated, but
only in w hat form.
Recently, however, government regulation
and planning in the health industry, as in
others, have come to be viewed with increas

HOSPITAL COSTS:
DECADES OF HIGH GROWTH

Since the beginning of the drive for national
health insurance, an overriding consideration
has been cost, and cost remains uppermost
in the minds of many of today’s health-care
reformers.
Hospital cost growth has substantially
outpaced the rise in the consumer price
index (CPI) for three decades. In the period
from 1950 to 1965 (prior to the enactment of
M edicare and M edicaid), cost per patient
day in short-stay hospitals increased 8 per
cent per year, while the CPI increased 2

’ Laurence S. Seidman is Assistant Professor of
Economics at Swarthmore College and is associated
with the Philadelphia Fed’s Department of Research. He
has published widely on health care costs and other
policy issues.

MARCH/APRIL 1982

BUSINESS REVIEW

For the average hospital patient today,
hospital care is free: the bill a patient runs up
is irrelevant to him financially, because the
insurer will pay 100 percent of it. (The
physician’s fee generally is not fully covered,
so the patient will pay a fraction of the MD
bill). Thus patients have no incentive to care
about the cost of their own hospital care. But
such care cannot be free to society, because
it consumes real resources. In the end, all
households collectively pay for hospital care
in large part through their health-insurance
premiums.
The situation is similar to restaurant bill
splitting. Suppose a large group goes out to
dinner, and it is agreed, in advance, that the
bill will be split without regard to each
person’s individual order. W hat happens?
Each person realizes that his own order will
have little effect on his financial burden,
which will depend largely on w hat everyone
else orders. So some probably will inflate
their orders, and the result is an inflated bill.
In effect, the U.S. hospital system is
financed by bill-splitting. Through insurance
premiums and taxes (for M edicare and
Medicaid), the citizens split the bill. It is
therefore little w onder that w hen the MD
decides w hether to admit his patient to the
hospital, which hospital to use, what tests to
run, and how long to extend the stay, he
generally ignores cost. His patient would not
want him to consider cost, because it has no
effect on his own financial burden.
There is a broad consensus that because
patients and physicians lack incentives to
keep costs down, too many people end up
demanding too much care, driving the cost
up too high. The disagreement is over what
to do about it.

percent. From 1965 to 1976, cost per day
increased 12 percent, while the CPI increased
5 percent. While half of the price rise came
from an increase in resources used—increase
in the quantity, quality, and style of se rv ic e half of it came from rising costs for labor and
other inputs for hospital care. This unusually
rapid cost rise has led m any critics to ask
w hether care is being provided as efficiently
as it might be—that is, w hether consum ers’
preferences are being reflected accurately in
the assignment of resources and w hether
those resources are being used in such a way
as to get as much care out of them as
possible.
Efficiency, however, is only one side of
the health sector issue. The other is equity.
Although the typical patient is able to pay for
a hospital stay through insurance, some
inadequately insured households remain
unable to do so, and they continue, each
year, to have great difficulty affording the
medical care they need. When hit with catas
trophic illness, they suffer severe financial
hardship.
M edicare for the elderly and M edicaid for
the poor were enacted in 1965 to reduce the
num ber of such households. Yet important
gaps still rem ain. Universal major-risk or
catastrophic protection does not yet prevail
in the U.S. Thus, both efficiency and equity
issues rem ain on the health policy agenda.
While lack of health insurance is the
source of inequities in providing care, ob
servers of the health scene trace efficiency
difficulties to the paym ent of bills by third
parties through health insurance plans.
HEALTH INSURANCE SPREADS
Over the last three decades, the share of
the national hospital bill paid by third-party
insurers—private and public—has increased
strikingly. In 1950, half of all hospital
revenues from patient care came from in
surers. By 1965, the third-party share ex
ceeded 75 percent. Today, nearly 95 percent
of the national hospital bill is paid by insurers.

THE REGULATORY APPROACH
Until recently, the dom inant view among
participants in the health policy debate was
that government should act to bring health
care within the reach of all by making it free
to everyone. Free-care advocates recognized
18

FEDERAL RESERVE BANK OF PHILADELPHIA

regulatory-planning approach have demon
strated that, despite free care, supply limita
tions can succeed in containing the health
share of GNP. In Britain, for example, where
the government owns and operates the hospi
tal system under the National Health Service,
the health share of GNP is roughly half that in
the U.S. (5 percent instead of 9 percent).
Cost containment, however, is not the same
thing as efficiency. It is just as inefficient to
devote too few resources to a given sector as it
is to devote too many. Efficiency requires
achieving just the right level of care, where
further expansion entails a cost not justified
by the benefit to the users. Moreover, effi
ciency requires that a given total supply be
allocated among users according to the urgency
of each one’s demand.
The shortcoming of the regulatory strategy
is not that it cannot control total cost effective
ly. Evidence from abroad shows that in some
cases it can. Rather, the central weakness is
that there is no effective w ay to register the
true preferences of each consum er. As a
result, it is difficult to determine the proper
growth rate of cost and the allocation of
scarce supply among individual consumers.
In virtually all other sectors, where each
person pays according to his own use, con
sumers self-regulate. We seldom worry about
w hether the growth rate of cost in sector X is
appropriate, because we know that each
consumer must balance cost against benefit.
But in the hospital sector, there is no incentive
for self-regulation. Each MD knows that his
insured patient w ants the best and most,
regardless of cost. MDs transm it these in
flated demands to the system. W ithout con
sumer cost sharing, how can the regulators
and planners assess the true intensity of
patient preferences?
In such supply-constrained health systems,
clear emergencies usually are handled with
the urgency they w arrant. But many medical
conditions that may require hospital service
are not clear-cut emergencies. W ith limited
hospital capacity, it is not obvious that those

that their approach would remove incentives
from patients and physicians to weigh cost.
But they counted on government regulation
and planning to fill the vacuum and encourage
efficient resource use.
Over the past decade, several methods
have been tried or proposed to promote
efficiency. The basic problem is that with
free care there is excess dem and—demand
that it would be w asteful and thus inefficient
to satisfy fully. The first method of eliminat
ing waste is to limit supply; the second, to
limit demand.
How can supply be limited? One method is
to require hospitals to obtain certificates of
need from regulators, in advance, before
expanding facilities or acquiring new tech
nology. Certificate-of-need programs were
begun in various states and have been author
ized under the Health Planning Act of 1974.
A second m ethod is to limit the growth in
revenue permitted of any hospital. The Carter
Adm inistration proposed a hospital cost
containm ent m easure which would have
limited revenue growth to approxim ately 9
percent per year. But this proposal has not
been enacted into law. So far, attempts to
limit supply haven’t been notably successful.
The attem pt to limit demand has taken the
form of utilization review. In 1972, Congress
authorized the creation of Professional
Standards Review Organizations (PSROs),
composed of physicians in each local area,
to review cases that appeared to use hospital
facilities w astefully. It was hoped that
PSROs would pressure MDs to reduce
demand for hospital facilities on behalf of
their patients. PSROs, however, haven’t
made a major dent in hospital cost growth
either.
Thus it isn’t surprising that some critics of
free care and regulation have focused on the
cost issue, even going so far as to argue that
having a free-care system guarantees that
the share of GNP going to health costs must
continue to grow, and grow significantly. It
needn’t, though. Other countries using the

BUSINESS REVIEW

MARCH/APRIL 1 9 8 2

with the most urgent nonemergency demand
will obtain the limited supply. Waiting lists
and queues for elective treatm ent are a
common feature of m any regulated, freecare systems, such as the British National
Health Service.
Thus all in all, the regulatory approach
doesn’t offer an easy w ay to decide how
much care to provide and who should receive
it.

that they could afford the care they needed
even if struck with a catastrophic illness.
The crucial feature of the Federal tax ex
clusion, however, is that it is open ended.
Rather than being limited to an amount
sufficient to buy major-risk protection, it
applies without limit. Thus employees have
been encouraged to seek, and employers to
provide, first-dollar, shallow hospital in
surance—insurance that covers all care from
the first dollar spent, w ith no provision for
patient cost sharing. Such high premium
first-dollar coverage does more than assure
major-risk protection; it m akes hospital care
free for the typical patient.
The crucial first step, then, in the marketoriented strategy is to remove the substantial
tax bias towards shallow insurance. In a
recent session of Congress, several bills
were introduced that would place a cap on
the tax exclusion so that only part of the
insurance benefit would be excluded from
taxable income. An employer contribution
above this cap would be included in the
taxable income of the employee exactly as if
it were cash salary.
But wouldn’t it be politically impossible to
set the cap at a level lower than the benefit
level employees are receiving now? It might
seem so, but the incentives can be arranged
in such a w ay as to induce employees to
choose lower benefit levels. The trick is to
give cash paym ents under the cap the same
tax advantages as benefits.
As proposed in a bill introduced last year
in Congress, employees could be allowed to
keep cash tax free below the cap should they
choose less costly insurance. Suppose, for
example, that the cap were set at $1,200 for a
workplace where previously the compre
hensive insurance premium was $1,200. If
an employee preferred to switch to a less
costly policy with an $800 premium, he
would be perm itted to keep the difference
($400) tax free.
An alternative would be to count the entire
contribution of an employer as taxable

A MARKET-ORIENTED STRATEGY
At the end of the last decade, an alternative
approach began to challenge the regulatory
strategy. Although advocates of this marketoriented approach concede that the m arket
does not function effectively in the hospital
sector, they believe that it can be restored to
reasonable health. M oreover, they contend
that achieving equity and restoring the
m arket m echanism can be made compatible.
Advocates of this strategy believe that the
source of the current m arket failure in the
hospital sector is misguided Federal tax
policy. For several decades, Federal tax
treatm ent has encouraged employers to shift
the compensation package aw ay from cash
towards comprehensive, costly health in
surance.
Consider an employer who has decided to
raise com pensation per employee by $100. If
he pays this increase in cash wage or salary,
the $100 will be taxable income; after payroll
and income taxes, the employee may perhaps
keep $70. But if the employer instead buys
the employee $100 of additional health in
surance, this form of com pensation will not
be regarded as taxable income for the em
ployee. Thus the choice facing the employee
is to take $70 more in cash or $100 more in
insurance. The tax exclusion of insurance
obviously biases employee choice in the
direction of insurance.
Undoubtedly, the intent of Congress in
excluding insurance benefits from taxable
income w as to encourage households to
obtain adequate major-risk protection so

FEDERAL RESERVE BANK OF PHILADELPHIA

the W est Coast, give providers an incentive
to limit cost because they are prepaid a fixed
sum, regardless of service rendered, in con
trast to fee-for-service. HMO supporters
believe they can capture an im portant share
of the health sector if households are not
subsidized w hen they buy high-premium
FFS insurance. Removal of the tax subsidy
would make HMOs, if they are more efficient
and less costly, more attractive to consumers.

income for the employee, but to enact a new
health insurance tax credit for all households.
The key feature of the credit is that it would
be a fixed-dollar amount for a household of a
given size and income regardless of how
much the household spent on insurance. The
household no longer would be subsidized for
purchasing insurance above the amount of
the credit.
It seems likely that if tax reform removed
the bias towards insurance, many employees
would prefer a shift in their compensation
package back tow ard cash. W ithout the tax
subsidy, m any probably would prefer in
surance with a more moderate premium,
even though it would require moderate
patient cost sharing (deductibles and coinsurance). M ajor-risk coverage would con
tinue to be purchased, but high-premium
insurance that makes hospital care free would
be m uch less common.
Physicians soon would recognize that the
average patient no longer is covered fully for
hospital care. Of course the typical patient
w ouldn’t nag his doctor about costs while on
his stretcher in the ambulance. Nevertheless,
the recuperating patient would receive a bill
and would bear a fraction of the cost. A n
ticipating the impact on his patient, the
average MD would try to avoid unnecessary
cost.
The new sensitivity of MDs to cost would
be transm itted to hospital managers. Under
free care, a hospital manager has no incentive
to try to provide a given quality at lower cost,
because MDs and patients regard cost as
irrelevant. W ith cost sharing, MDs would
prefer hospital A to hospital B if A provided
the same grade of care at lower cost. Thus,
insurance with cost sharing ultimately would
transmit pressure for efficiency to producers,
just as it does in other sectors.
Removal of the tax subsidy not only would
promote efficiency under traditional fee-forservice (FFS); it also would remove an obstacle
against health m aintenance organizations
(HMOs). Prepaid HMOs, such as Kaiser on

If the market-oriented strategy works as
advocates envision, it should improve ef
ficiency. And an income-related, last-resort
tax credit would ensure that all households
have income-related major-risk (catastrophic)
coverage. The tax credit would assure equity,
while the repeal of the tax exclusion would
promote efficiency.
Consider, for exam ple, a household with
$20,000 of income. W hether or not it has
private insurance, it might have to bear the
first $2,000 (10 percent of income) of expense
(excluding any expense of an insurer on its
behalf) before being eligible for a tax credit.
It might then be entitled to a credit equal to
80 percent of its additional expense, until it
has borne $3,000, or 15 percent, of its income.
At that point, it would be entitled to a 100percent tax credit on additional expense.
The credit would provide an out-of-pocket
ceiling equal to 15 percent of income. W hat
ever a household’s private insurance coverage,
the tax credit of last resort would assure
major-risk coverage. M oreover, the credit
would be refundable, so that a low-income
household would receive a check from the
IRS if its credit exceeded its tax liability.
Thus the incom e-related refundable credit
could replace Medicaid.
No policy prescription, on health care or
anything else, is likely to scrape by without
criticism, and this m arket approach to health
care is no exception. W hat are the typical
objections?
SOME CRITICISM

Critics of the market-oriented strategy
21

BUSINESS REVIEW

MARCH/APRIL 1 9 8 2

give several reasons for their skepticism.
For one thing, they doubt that removing
the tax bias toward shallow insurance in fact
will reduce the prevalence of such insurance.
They believe households strongly prefer
first-dollar coverage and would pay high
premiums even without the tax advantage.
Advocates reply that one reason why the tax
advantage was originally enacted was the
concern that, without it, many people might
choose cash over even major-risk insurance.
Also, critics doubt that insurance contracts
with cost sharing would affect either physi
cian or hospital behavior even if those con
tracts became widespread. They contend
that MDs, not patients, make the decisions,
and that they will be unaffected by the
financial impact on their patients. But even
today MDs often consider financial impact
in the few areas where it matters to the
patient. MDs sometimes tell patients for
example: “I will put you in the hospital
where your insurance will fully cover you,
even though I could treat you as an out
patient, where it would cost you more.” MDs
know that patients appreciate such concern.
Finally, some critics fear that cost sharing
would impose severe burdens on households
unable to afford it. Those critics surely are
correct that removing the tax bias alone
might well produce hardship and would do
nothing to protect households with little or
no private insurance. But the income-related
last-resort tax credit can address these con
cerns. Whatever the cost sharing under a
private insurance policy, an income-related
credit can ensure that every household
receives enough assistance to afford the care
it needs without hardship.

CONCLUSION

Most other countries have adopted a regu
latory-planning approach to the health sector.
Free hospital care has been accepted as
necessary for equity, and regulation has
been relied upon to promote efficiency. Until
recently, this strategy appeared to be the
dominant one in the U.S. as well.
An alternative, however, has emerged to
challenge the regulatory approach—a
market-oriented strategy. To promote ef
ficiency, this strategy would reform the tax
treatment of private health insurance to
remove the bias toward shallow, first-dollar
coverage. With the tax exclusion curtailed,
many employees would shift over time to
less costly insurance that includes patient
cost sharing. With hospital care no longer
free to most patients, physicians would begin
to weigh cost, and hospital managers would
respond to this new sensitivity on the part of
MDs.
The market-oriented strategy would
promote equity though an income-related
last-resort tax credit that would ensure an
out-of-pocket ceiling related to household
income for all households, whatever their
private coverage. The tax credit is a modern
policy instrument that enables assistance to
be provided according to income while pre
serving confidentiality.
The debate over health care provision will
be resolved one way or another in the next
few years. Its outcome will determine the
direction of the health sector for the rest of
the century, with important implications for
social welfare. It is a debate worth watching
carefully.

FROM
THE PHILADELPHIA FED
This new pamphlet presents
some highlights of financial
planning tools authorized by the
Economic Recovery Tax Act of
1981. For your free copy write or
call the Department of Consumer
Affairs, Federal Reserve Bank of
Philadelphia, P.O. Box 66, Phila
delphia, PA 19105 — (215) 5746116.

100 North Sixth Street
Philadelphia, PA 19106

Full text of Business Review (Federal Reserve Bank of Philadelphia) : March/April 1982

FRASER