Full text of Economic Review (Federal Reserve Bank of San Francisco) : Summer 1989, Number 3

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

I'.,conomic
Review
Federal Reserve-Bank
of San Francisco
Summer 1989

Gary C. Zimmerman

Adrian W. Throop

Adrian W. Throop

Number 3

The Growing Presence of
Japanese Banks in California
Reagan Fiscal Policy and the Dollar
Fiscal Policy, the Dollar, and International Trade:
A Synthesis of Two Views

| Table of

^

The Growing Presence of Japanese Banks In California „ 00„*000„«0„ .. 00„„„0«,Qo 3
Gary C. Zimmerman

Reagan Fiscal Policy and the Dollar . . . „„ . . . . . . „ . „ . . . . . . . . . „ . . . » » . . . * . . . . . . 18
Adrian W. Ttiroop

Fiscal Policy? the Dollar? and International Trade:
A Synthesis of Two Views . . . . . . • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . •. . 27
Adrian W. Throop

Federal Reserve Bank o f San Francisco

1

Opinions expressed in the Economic Review do not neces­
sarily reflect the views of the management of the Federal
Reserve Bank of San Francisco, or of the Board of Governors
of the Federal Reserve System.
The Federal Reserve Bank of San Francisco’s Economic Review is
published quarterly by the Bank’s Research Department under the
supervision of Jack H. Beebe, Senior Vice President and Director of
Research. The publication is edited by Barbara A. Bennett. Design,
production, and distribution are handled by the Public Information
Department, with the assistance of Karen Rusk and William Rosenthal.
For free copies of this and other Federal Reserve publicatons, write or
phone the Public Information Department, Federal Reserve Bank of San
Francisco, P.O. Box 7702, San Francisco, California 94120. Phone
(415) 974-2163.

2

Economic Review / Summer 1989

The Growing Presence
of Japanese Banks in California

Gary C. Zimmerman
Economist, Federal Reserve Bank of San Francisco.
Outstanding research assistance was provided by Stacey
Dogan and Alice Jacobson. Editorial committee members were Bharat Trehan, Elizabeth Laderrnan, and Fred
Furlong.

Japanese-owned banking institutions have dramatically
increased their share of the California banking market
since the early 1980s. However, the author finds that,
overall, the increase inforeign bank ownership in the State
has been slight. Japanese bank expansion appears to have
arisen primarily from the rapid increase in trade with
Japan and Japanese direct foreign investment in California. While Japanese banks emphasize commercial lending
and rely more heavily on purchased funding , they have a
substantial, and growing, retail presence in the State.

Federal Reserve Bank of San Francisco

Japanese banks have made significant inroads into the
California banking market since the early 1980s. Over the
period from 1982 through 1988, assets of Japanese-owned
banking institutions in California increased nearly threefold. At the end of 1988, Japanese-owned banks held onequarter of the banking assets and over 30 percent of all
business loans in the State. This expansion has come about
through acquisitions and new entry, as well as through
rapid growth of existing institutions.
California is not the only market where Japanese banks
have made inroads. World-wide, Japanese-owned banking
institutions have grown over the last decade and now
dominate the list of the world's largest banks. Their
overseas branches alone nearly doubled their assets from
1982 to 1987, reaching almost $1.1 trillion by year-end
1987. 1
In California, in particular, this growth has received
considerable attention, raising concerns about its causes
and the possibility that with continued growth, Japanese
banks could come to dominate the California market. This
paper first examines the strong growth and increasing
market shares of two groups of Japanese-owned banking
institutions: commercial banks chartered in California and
the agencies and branches of Japanese banks that operate
in California. (A description of the banking powers and
services of these different types of banking institutions is
provided in the BOX.) Section II then explores possible explanations for the rapid growth of Japanese-owned banking institutions in California, and Section III examines
empirical evidence on growth patterns of banking institutions to assess the potential causes of rapid growth. In
particular, empirical comparisons of the composition of
bank portfolios and funding attempt to determine whether
balance sheets of Japanese banks reflect the factors that
may be causing rapid growth. Section IV provides some
observations and conclusions about the Japanese banking
presence in California. Specifically, growth is likely to
continue, but at a slower pace, with more emphasis placed
on the "retail banking" area.

3

I. Rapid Growth
The growth of Japanese banks, agencies, and branches
in the mid-1980s in the U.S. has been rapid. In California,
the traditional measures of market presence reflect that
growth; assets, loans, and deposits all show dramatic
increases. Between1982and 1988,the assets of Japaneseowned banking institutions in California nearly tripled,
growing from $34.6 billion to $93.4 billion.? As shown
in Chart lA, Japanese-owned institutions' market share,
which was 10.7 percent of the total assets of all banks,

agencies, and branches in the State in 1982, rose to 25.2
percent by theendof 1988.
USiIlg a narrower definition of the banking market,
that is, commercial banks only (excluding agencies and
branches), there still is a strong increase in the market
share of Japanese-owned institutions. As Chart 1B shows,
assets of Japanese-owned banks rose from 3.7 percent to
11.9 percent of the "bank only" market from 1982 to
1988. 3 The Japanese-owned commercial banks more than

Chart 1A
Share of Total Bank, Agency and
Branch Assets for California

Percent

80
Domestic

70
60
50
40

Total Foreign

30
Other Foreign

20
10

Japanese

0

1983

1984

1985

1986

1987

1988

Chart 1B
Share of Bank Only Assets for California
Percent

90
Domestic

80+---------~

70

60
50
40
30
Total Foreign

20+---------_

-

Other Foreign
Japanese

10 ~~=~---~~==:::=
O+----,----.,...----....---r------r------,
1983

4

1984

1985

1986

1987

1988

Economic Review / Summer 1989

tripled their assets during the period, reaching $35.3
billion by year-end 1988. Sixty percent of this increase
resulted from acquisitions of three large foreign-owned
banks by Japanese-owned banks." However, Japaneseowned banks in general also grew at a faster pace than did
other California banks.
Of the ten Japanese-owned commercial banks operating
in the State at year-end 1988, eight were subsidiaries of
Japanese banks. Twosmall banks are Japanese-owned, but
are not owned by banks headquartered in Japan. The
subsidiary banks account for virtually all of the assets of
the Japanese-owned banks in the State. Four of these banks
ranked among the top ten banks in the State at year-end
1988: Union Bank (fifth largest at $15.0 billion in assets),
Bank of California (sixth largest, $6.9 billion), Sanwa
Bank California (seventh largest, $6.4 billion), and Sumitomo Bank of California (ninth largest, $3.7 billion).
Earlier in the year Tokai Bank of California ranked tenth. 5
The twenty-two foreign agencies and six foreign
branches of Japanese-owned banks more than doubled in
size over the 1982 to 1988 period. Their combined assets

Federal Reserve Bank of San Francisco

reached $58.1 billion by year-end 1988. The agencies
($44.7 billion) and branches ($13.4 billion) provide an
array of commercial, money market, and trade-related
banking services, either directly or through international
banking facilities that serve only foreign residents. 6

Shifts in Market Shares
Although the increase in Japanese ownership of banks in
the U.S. has raised concerns about "a wave" of Japanese
financial services acquisitions," expansion has not occurred at the expense of domestic banks' market share in
California. Over the six years ending in 1988, the market
share of all foreign banking institutions increased only
modestly, from 30.8 percent to 31.7 percent. 8 As can be
seen in Chart lA, rather than gaining market share from
domestic banks, Japanese banks essentially have replaced
other foreign banks as the dominant foreign banking power
in California.
In fact, using the narrow market definition, commercial
banks only, foreign market share actually declined from
20.2 percent in 1982 to 14.9 percent in 1988. Most of this

5

decline resulted from the purchase of the British-owned
Crocker Bank by Wells Fargo Bank in 1986.9 This explains
the sharp decline in "other-foreign bank" market share in

Chart lB. Thus, we have seen a shift, rather than an
increase, in foreign ownership.

II. Reasons for Expansion
This section examines a number of factors that may have
played a role in the growth of Japanese banking assets in
California. For example, the growth in tradewith Japan
and the growth of Japan's current account surpluses are
two factors that may have stimulated the growth of Japanese banking institutions in California. Liberalization of
domestic capital markets in Japan is another factor that may
have stimulated investment in the State. Finally, California's location, making it an important point of entry, as
well as the State's perceived attractiveness as a growing
banking market, are factors that may have helped to set the
stage for the upsurge in investment in Japanese banking in
California. These considerations are discussed below.

Banks Follow Trade
One reason for the growth in Japanese banking assets in
California may be the growth of US. trade with Japan.
Because of California's location on the Asia-Pacific Rim,
the State has become an important locus of trade-related
activity. With the increasing presence of Japanese-owned
multinational firms engaging in trade-related activities in
the State, there has been a commensurate increase in the
need for trade-related financing, exchange, clearing, and
other credit and banking services. Japanese-owned banking institutions may have a comparative advantage over the
US. and other competitors in providing these services. If
this is the case, it is logical that Japanese banking assets
would increase in the current environment of growing
trade with Japan.
A study of US. and Japanese banking by Henry Terrell (1979) provides a theoretical and empirical basis for
suggesting that in the context of rapid growth in trade,
Japanese banks do indeed enjoy comparative advantages in
providing banking services in California. 10 In this analysis, Terrell adapts Caves' (1974) model of foreign investment to international banking. He notes that" ... foreign
investment is often associated with product differentiation, which may include possession of intangible assets
such as a firm's knowledge about how to produce and
distribute its product." 11 Terrell applies this concept to
multinational banking, hypothesizing that in foreign markets, foreign banks can differentiate their products from
those of domestic competitors by specializing in services
to multinational firms from their home country. By specializing in such services, Terrell argues, Japanese banks'

6

subsidiaries, agencies, and branches may enjoy a significant competitive advantage over their domestic competitors on account of their pre-existing business relationships
with Japanese firms, as well as their superior knowledge of
Japan's markets, customs, and operations.
Thus, Terrell suggests that "rapid growth of foreign
branch and subsidiary activities by both Japanese [banks in
America] and American banks [in Japan] indicates a
customer preference for obtaining banking services [in a
foreign country]-such as access to credit, deposit, and
payment facilities-from the office of a bank with which
they [already] are familiar [in their home country]. . ." He
also asserts that "Customers much prefer this approach to
the alternatives of either dealing with a local institution
or, more expensively, dealing with a far-distant banking
facility." 12
These assertions provide a testable hypothesis; namely,
that foreign banking activity is related to growth in trade
with a given foreign country, as well as strong local
economic conditions in the host country. Indeed, Terrell finds such a positive correlation in his empirical
analysis. 13
The strong relationship Terrell found in the 1970s between international trade and foreign banking activity
appears to have been borne out in the 1980s. The surge in
trade between Japan and California appears to have increased the demand for Japanese banking services in
California and the US. 14 Chart 2A shows California's
combined volume of imports from and exports to Japan
from 1980 to 1987. The volume of trade increased by
nearly 150 percent during this period.t> Similarly, the
assets of Japanese-owned banking institutions experienced
a surge in growth over this period, as shown in Chart 2B.
Moreover, just as Japan's trade with California came to
take up a larger share of California's total trade volume (35
percent in 1987, versus 30 percent in 1982), Japanese
banking assets in California also accounted for a larger
share of foreign banking assets in California by the end of
1987. (See Charts lA and lB.)
The British experience provides further support for the
view that increased trade helped to stimulate the growth of
Japanese banking assets in California. While commerce
with Japan has soared, British trade with California has
grown only slightly since 1982. (See Chart 2A.) By 1987,
California's $3.1 billion total trade with Britain was

Economic Review / Summer 1989

dwarfed by the $41.1billion trade with Japan. According to
the "banks follow trade" hypothesis, the lackluster growth
in British trade with California should have been associated with lackluster growth in British banking assets
during this period. In fact, as discussed earlier, the British
banks left the State. Thus, even though they held about 15
percent of California's commercial bank assets in 1982,
without growing trade flows to sustain them, a number of
years of subpar performances by their California banks
induced most of the British banks to sell their California
subsidiaries and invest elsewhere.

$ BIllions

DirectForeign Investment
Another factor that may have helped to stimulate the
rapid growth in Japanese banking assets in recent years is
the growth of Japan's external surpluses over most of this
period. Japan's large trade surpluses have made funds
available for direct foreign investment. Coupled with the
1980 amendments to the Foreign Exchange and Foreign
TradeControl Law, which deregulated capital flows into
andout of Japan, these trade surpluses have stimulated a
dramatic increase in Japanese investment in the U.S. in
general arid California in particular.

Chart 2A
Total Volume of Imports and Exports
For the State of California

50

40
30
Japan

20

10

1981

$ Billions

1982

1983

1984

1985

1986

1987

Chart 2B
Japanese Banking Assets in California

100

80
60
Total

40
20

o L -__-===r====::::;===;:==;::---,
1983

Federal Reserve Bank of San Francisco

1984

1985

1986

1987

1988

7

Over 700 Japanese firms have established manufacturing operations in the U. S.; about one-quarter of those are
located in California." The bulk of the Japanese investment through 1986 was in manufacturing, wholesale trade,
and real estate. Finance accounted for under seven percent,
according to US. Commerce Department data.
According to Commerce Department data, by 1986,
Japan accounted for 14.2 percent of total direct foreign
investment in California, up from 9.1 percent in 1982. In
1986, Japan was second only to Canada in terms of direct
foreign investment in California, with $5. 3 ..billion to
Canada's $6.0 billion. However, for Japan, there was a
129.1 percent increase between 1982 and 1986 in the book
value of plant, property, and equipment invested in California. Over the same period, Canada's increase was only
9.2 percent. The Japanese also surpassed Britain, which
slipped to third, despite a 63.6 percent increase in direct
foreign investment over the period.
This increase in direct investment probably has affected
the growth of Japanese banks in California both directly
and indirectly. The direct effects can be seen in the steppedup pace of Japanese acquisitions of existing banks in
California. Another potential channel for increased direct
investment in California may have been increased funding
from the parent banks in Japan. Such funds could have
enabled Japanese banks in California to grow more rapidly
than their competitors.
Increased direct investment also may have provided an
indirect stimulus to Japanese banks in California through
increased demand for banking services on the part of
Japanese-owned commercial firms that have set up operations in California. As in the case of the "banks follow
trade" argument, banks also may follow investment since
Japanese investors probably prefer banking services provided by Japanese-owned banks with which they already
have established business relationships in Japan.

Other Factors
The strength of the California economy is another factor
that may have influenced Japanese bank expansion. In its
own right, California is a large market. Its GDP ranks with
the top eight countries in the world. Its large size, good
location for international trade, and diversified base of
production offer banks an attractive market. In recent
years, the State's economy has outperformed the U.S.
economy, with more rapid growth and greater resistance to
downturns, potentially making California more attractive
than other US. locations to foreign banks.
Thus, if California's attractiveness as a market in its own
right were a factor stimulating the growth of Japanese

8

banking in the State, one would expect to see Japaneseowned banks and agencies and branches growing more
rapidly in California than in the US. generally or in the
New York and Chicago banking markets. Despite the
obvious Japanese interest in California, a comparison of
growth rates for Japanese-owned banking institutions from
1982 to 1988 indicates that Japanese-owned institutions in
the US. generally, and in New York and Illinois particularly, grew at slightly faster rates than did Japanese-owned
banks in California. 17
Likewise, if California were particularly attractive as a
banking market in its own right, one might expect foreign
banks in general to grow rapidly in California. However,
this pattern is not observed. As discussed above, Japaneseowned banking institutions have grown much more rapidly
than other foreign institutions. Charts lA and lB illustrate
the impact that differential growth rates have had on
market share. While the Japanese-owned banks have been
growing rapidly, other foreign-owned banks have been
leaving the market. In several cases, the Japanese have
acquired foreign-owned banks. The pattern also extends to
the agencies and branches, where the Japanese institutions
have grown much more rapidly than the non-Japanese
institutions. Given these observations, the attractiveness of
the California market relative to other US. markets does
not appear to have been a major factor in the rapid growth
of Japanese-owned banks.
Diversification of geographic risks and access to U.S.
markets also may have been a stimulus for the growth of
Japanese banks in California. Specifically, entry into the
California market would provide access to the US. money
markets and diversify an international institution's funding
base by allowing it to raise funds in the large US. financial
markets. Similarly, operations in California provide access
to major corporate borrowers in the US. and an opportunity to diversify lending risks across countries. However, this
may not be a particularly strong argument, since diversification may be attained without access to California, or
even the US., for that matter.
One final explanation that has been offered for the rapid
growth of Japanese banks has to do with potential competitive advantages. Some argue, for example, that Japaneseowned banks have access to funding and/or capital from
their parent banks, which, because of their size and strong
credit ratings, can borrow at lower costs than can the
subsidiary (or than domestic banks)." This advantage,
moreover, may be reinforced by the availability of funds
associated with Japan's large external surpluses. Thus,
borrowing from an overseas parent could provide a fundingadvantage for Japanese-owned subsidiary banks, agencies, and branches. However, the parent must be willing to

Economic Review / Summer 1989

forsake alternative open market returns on the funds in
favor of subsidizing the operations of its California af­
filiates.
In any event, Japanese banks are unlikely to have any
significant funding advantages over their domestic rivals in
domestic funding markets. It seems clear that within
California, Japanese banks price their retail deposits com­
petitively to survive in the retail market. And in the
wholesale markets, Japanese banks offer competitive rates
on large CDs, federal funds, and eurodollars.
In sum, there are a number of factors that potentially

have stimulated the growth of Japanese banks in Califor­
nia. Based on the evidence considered in this section, it
appears that the growth in U.S. trade with Japan, the
increase in Japanese direct investment in the U.S., and
possibly, competitive advantages associated with using
Japanese parents as a funding source have been the major
reasons for the growth of Japanese banks in California and
in the U.S. In the next section, I compare the portfolios of
Japanese-owned banks with those of domestically-owned
banks in California to determine whether investment
and/or funding patterns reflect these influences.

III. A Comparison of Portfolios and Cost Structures
As discussed above, the dramatic increase in trade with
Japan and the growth in Japan’s trade surpluses likely have
been key reasons for the increase in the Japanese banking
presence in California. Trade financing is an area where
Japanese banks may have a competitive edge in light of this
growth in trade and trade surpluses. Therefore, one might
expect to find that Japanese banks have a larger market
share and extend more of this type of credit as a proportion
of assets than do domestic banks. Since trade financing
should appear on commercial banks’ balance sheets as
business lending, commercial and industrial loans, com­
mercial letters of credit, and standby letters of credit
(SLCs) provide useful measures of market presence.19
Data on outstanding balances of commercial and indus­
trial loans and letters of credit show that Japanese-owned
banking institutions in California have, in fact, gained a
large share of these major commercial lending markets. In
business lending, Japanese-owned banking institutions
held 30.1 percent of the California market at year-end
1988. Their share of commercial letters of credit was even
higher, nearly 32 percent. They controlled a 44 percent
share of the rapidly growing SLC market, which is dom­
inated by foreign banks with strong credit ratings.20
Clearly, then, Japanese banks in California have a strong
position in the markets for commercial lending and letters
of credit. This is consistent with the view that the growth in
Japanese-owned banks is associated with their trade orien­
tation.

domestically-owned banks in California. Specifically,
Japanese-owned banks probably extend proportionately
more business-oriented credit, including commercial real
estate loans, and proportionately less consumer-oriented
credit, including residential mortgage loans, than do
domestically-owned banks. Likewise, if direct investment
is a factor in Japanese-owned banks’ growth, there could
be differences in the composition of Japanese-owned
and domestically-owned banks’ liabilities. Specifically,
Japanese-owned banks may rely proportionately more on
borrowings from their parents and proportionately less
on retail-oriented deposits than do domestically-owned
banks.

Chart 3
Business Loans as
Percentage of Total Assets
Japanese Banks vs.
U. S. Owned Banks (California)

Aggregate Portfolio Measures
Other indications of Japanese banks’ trade orientation
potentially may arise in comparisons of Japanese- and
domestically-owned banks’ portfolios; that is, if the growth
in Japanese-owned banks largely is the result of increased
trade, one might expect to see differences between the
composition of the portfolios of Japanese-owned and

Federal Reserve Bank o f San F rancisco

9

Comparisons of ratios of various categories of loans to
total assets and ratios of various types of liabilities to total
assets are made. Percentages for aggregate data on all
Japanese-owned banks are compared with percentages for
aggregate data on all domestic banks. These percentages
are tracked over time, and presented in Charts 3 through 7.
The aggregate data provide a measure of portfolio dif­
ferences between the bank groups, as well as a measure of
portfolio trends over time for each group. (However, these
are relative portfolio measures, scaled by assets; they
obscure differences in the actual growth rates of the asset
and deposit categories across bank groups. For example,
even though small time and savings deposits at Japaneseowned banks have grown more rapidly than at domestic
banks, Chart 7 shows that the even more rapid expansion of
Japanese bank assets has reduced the relative reliance on
small time and savings deposits as a funding source for the
Japanese banks.)
Aggregate data indicate that Japanese banks do have a
much higher proportion of commercial loans than domes­
tic banks have. While Japanese-owned banks held nearly
30 percent of their assets in business loans in 1988,
domestic banks held only 19.5 percent. Chart 3 indicates
that the difference between the two groups grew over the
period when trade with Japan was soaring. It also is
interesting to note that there are no Japanese-owned sav­
ings and loan associations or savings banks in California at
present. While this may be due in part to regulatory

Chart 4
Real E s t a t e L o a n s as
P e r c e n t a g e of T o t a l A s s e t s
J a p a n e s e B a n k s vs.
U . S . O w n e d B a n k s ( Calif orni a)

10

Chart 5
F u n d i n g as
P e r c e n t a g e off T o t a l A s s e t s
J a p a n e s e B a n k s vs.
U. S. O w n e d B a n k s ( C a l i f o r n i a )

restrictions, it also provides support for the argument that
Japanese banks follow trade and thus are trade/businessoriented. Thrift institutions in California traditionally have
not engaged in business finance and instead concentrate
primarily on household finance, an area in which Japanese
banks do not have a comparative advantage, according to
the “ banks follow trade” argument.
In the aggregate, we also find that Japanese-owned
banks have a higher proportion of commercial real estate
loans than domestic banks have. At year-end 1988, Jap­
anese banks had 10.4 percent of their assets in commercial
real estate loans, versus only 6.7 percent for domestic
banks.
It is also interesting to note that while Japanese banks
have a higher proportion of business and commercial real
estate lending, there appear to be no significant differences
in their relative proportions of total real estate lending (as
shown in Chart 4), single family real estate lending, or
consumer loans. Rather, as shown in Table 1, they have a
higher proportion of loans to assets, which is offset by a
smaller proportion of federal funds and repurchase agree­
ment lending, securities holdings, and other assets, which
mostly includes cash and balances due from other banks.
Deposit and liability data also tend to support the view
that Japanese banks are more wholesale-business oriented
than are their domestic counterparts. As can be seen in
Chart 5, in the aggregate, Japanese-owned banks primarily
are funded by domestic deposits, like their domestic bank
competitors. However, from Chart 6, it is clear that

E conom ic R eview / Sum m er 1989

Japanese-owned banks in California rely more heavily on
jumbo CDs ($100,000 and over) than do domestic banks.
Moreover, Chart 6 shows that other borrowings, which
mayinclude federal funds, repurchase agreements, and
eurodollar borrowing, as well as borrowing from parent
banks, have been a more important source of funds for
Japanese-owned banks than for their domestic counterparts. Nonetheless, at present, as a percent of assets,
funding from parent banking organizations is not a particularly large funding source for Japanese subsidiary banks
in California. According to recent bank holding company
data, only about five percent of assets were funded this
way, and only one of the eight subsidiaries had a significantly greater amount of borrowing from its parent. 21
One reason is that most of the subsidiary banks view
such borrowing as a backup, rather than a primary,
source of funds. Finally, Chart 7 indicates that retail time

Federal Reserve Bank of San Francisco

and savings deposits account for just over one-third of
Japanese-owned bank funding, versus closer to fifty percent for domestic banks.
Thus, there are significant funding differences between
Japanese-owned banks and domestic banks. Japaneseowned banks rely more on wholesale deposits and borrowings and lesson retail deposits, consistent with the view
that they are more trade- and business-oriented.
Taken together, then, the aggregate data on lending and
funding patterns roughly are consistent with the trade
patterns,lending support to the view that increased trade
and increased direct investment have been important stimuli to the growth of'.Japanese-owned banks in California.

Comparisons of Individual Bank Portfolios
However, these results could arise from individual
banks' •portfolio decisions, and could be unrelated to the

11

individual banks' ownership status. Thus, it is important to
determine whether the difference is significant at the
individual bank level after controlling for other factors that
might influence portfolio composition. Analysis of the
individual bank data also may be necessary because the
sample of Japanese-owned banks is small, and the larger
institutions might skew the aggregate numbers.
Therefore, cross-sectional OLSQ regressions on individual bank data are employed to identify statistically
significant portfolio differences between Japanese-owned
banks and other banks. Financial statement data are available from the December 1987 Call Reports for all commercial banks in the state. 22 A description of the regression
model is presented in the Appendix, and the results of the
regressions are presented in Tables I and 2.

12

At the individual bank level, the regression results are
consistent with the aggregate portfolio data. Controlling
for size and branch differences, Japanese banks have 10.2
percent more business loans on average than domestic
banks have. Moreover, that difference is statistically significant at the one-percent level, as shown in Table 1.
Similarly, the cross-sectional regressions indicate that
commercial real estate lending at Japanese-owned banks
accounts for nearly seven percent more of the asset portfolio than at domestic banks, and that the difference is
statistically significant at the one-percent level. As shown
in Table I, these findings are consistent with the view that
increases in trade and Japanese direct foreign investment
in California played a part in the expansion of Japanese
bank activity in the State.

Economic Review / Summer 1989

Finally, at the individual bank level, differences in
funding patterns are statistically significant, and in some
cases these differences are also quite sizeable. For exam­
ple, as shown in Table 2, on average, Japanese-owned
banks were 12.2 percent more dependent on large CDs and
7.4 percent more dependent on other borrowing for fund­
ing than were domestic banks.

Chart 6
P u r c h a s e d F u n ds as
P e r c e n t a g e of T o t a l A s s e t s
J a p a n e s e B a n k s vs.
U. S. O w n e d B a n k s (California)

Cost Structures
This greater reliance on borrowed funds potentially
confers a cost advantage to Japanese-owned banks to
the extent that they can borrow more cheaply than can
domestically-owned banks, either because their parents in
Japan are subsidizing them or because their affiliation with
their highly-rated parents enables them to borrow directly
at a lower cost. In the aggregate, the small proportion of
parent funding probably dilutes any potential advantage
significantly. Still, these portfolio measures do not rule out
the possibility that a “ small” advantage might exist.
We cannot rule out the possibility of a funding advan­
tage from the portfolio data. However, if parent banks were
willing to “ subsidize” their California subsidiary banks,

Table 3
Cost of Funds Estim ates for 1988
Dependent Variable:
Independent Variable

Intercept
Assets (millions of $)
Japanese-owned
Other foreign-owned
1 Year Growth Rate
Number of Branches
Equity Capital (mil. $)
Chart 7

% Assets

C o n s u m e r T i m e & Savings a©
P e r c e n t a g e of T o t a l A s s e t s
J a p a n e s e B a n k s vs.
U. S. O w n e d B a n k s (California)

Number of Observations
Adjusted R-squared
F Value
Probability>F
Dependent Variable:

Cost of Interest-Bearing Funds
Parameter
Estimate

Standard Error

0.06170***
-0.00487
-0.00314
0.00548*

0.00076
0.00310
0.00453
0.00293

0 .0 0 0 0 2

0 .0 0 0 0 2

0.00006
0.06743*

0.03563

0 .0 0 0 1 0

360
0.0113
1.687
0.1231
Cost of Other Borrowing

Independent Variable

Parameter
Estimate

Standard Error

Intercept
Assets (millions of $)
Japanese-owned
Other foreign-owned
1 Year Growth Rate
Number of Branches
Equity Capital (mil. $)

0.08011***
-0.01417
0.01225
-0.00013
- 0 .0 0 0 1 0
0.00035
0.17513

0.00996
0.01474
0.03229
0.02250
0.00056
0.00049
0.17103

Number of Observations
Adjusted R-squared
F Value
Probability>F

69
-0.0593
0.356
0.9035

*** Significant at the 1 percent level.
* Significant at the 10 percent level.

Federal Reserve B ank o f San F rancisco

13

we would expect Japanese-owned banks to have a lower
cost of funds than their competitors.
Thus, it is useful to compare Japanese-owned and
domestically-owned banks' average cost of funds for two
categories of liabilities. The overall average costs of
interest-bearing funds and of other borrowings were estimated using cross-sectional OLSQ regressions for 1988to
control for other factors that might influence banks' cost of
funds, such as size, number of branches, equity capital,
and growth over the past year. The results, presented in
Table 3, show no significant difference between Japaneseowned banks and domestic banks in either the overall cost

of interest-bearing funds or the cost of other borrowings
(wbich includes the subsidiary banks' borrowing from
parents). These results tend to refute the argument that
Japanese banks have a funding advantage over domestic
banks'as a result of their ability to rely on low cost funding
from-their parents.
Related to potential funding advantages, it could be
argued that Japanese banks have advantages arising from
differential capital standards between Japan and the U.S.
However, recent agreements on international capital standards for banks should eliminate or reduce any potential
differences across countries. 23

IV. How Will Japanese-Owned Banks Fare in the Future?
If the Japanese banks have an obvious competitive
advantage in California, it is in the area of providing traderelated and commercial banking services. In this market
they can provide services that are tailored to the needs of
Japanese multinational business firms in California. Japanese banks' knowledge of Japan's markets, customs, and
operations likely is superior to that of their other California
competitors.
Using this advantage, the assets of Japanese-owned
banks, agencies, and branches have grown from under 11
percent of the California total in 1982 to over one-quarter
of the state's banking assets and over thirty percent of
commercial loans at year-end 1988.
Japanese-owned banks especially have grown rapidly
during this period. Acquisitions of other foreign banks
accounted for a large part of that growth. Prospects for
future acquisitions may be reduced somewhat by the limited number of "independent" medium-sized banks in the
State as well as the opening of banking in California
to institutions headquartered anywhere in the country.
Japanese-owned commercial banks thus may find more
competition for acquisitions after 1991. Thus, they also
may find it difficult to continue increasing their market
share at the rapid pace of the last six years.
A slowdown in trade with Japan or direct Japanese
investment also could reduce the demand for the services
of Japanese-owned banking institutions and limit their
growth. And they will always be limited by the reality that
trade-related lending is only a small portion of the State's
total banking business.
Over time, a move in the retail direction could minimize
these limitations. It also would tend to minimize portfolio
differences with domestic banks, if other factors remained

14

constant. However, unlike financing trade and the activities of foreign firms, there is no compelling evidence
that Japanese commercial banks have an advantage in
providing retail banking services in California.
Nonetheless, Japanese banks in California, as distinct
from the agencies and branches, have demonstrated an
interest in the "traditional retail banking field." A recent
American Banker review of Japanese banking in the State
suggests that the Japanese banks in California have broadened their customer bases over time, expanding their
provision of banking services to include a wider segment
of the market than just trade or the Japanese community. 24
Moreover, the article suggests that many Japanese banks
are planning to extend their retail and middle market
commercial lending business.
Japanese-owned commercial banks already have a significant investment in the "bricks and mortar" necessary
for a retail banking presence in the State. Over the last five
years, the number of branches has grown, although not as
fast as assets. Japanese-owned banks had 424 branches in
December 1988, up from 247 in 1982. Their expanded
branch network gives them a "delivery system" for retail
deposit services as well as for real estate lending, consumer finance, middle market commercial lending, and
even agricultural lending.
This progression from trade finance to provision of
banking services for the domestic market per se is not
unusual. Terrell noted such a trend in his 1979 article,
finding that it is consistent with Caves' theoretical framework concerning foreign investment.
Thus, to remain competitive in the State in the future,
Japanese banks increasingly will need to compete directly
with the large domestic banks. And their future, like that of

Economic Review / Summer 1989

their competitors, will be greatly influenced by the State's
economic conditions and international economic developments.
California, and the U.S. generally, are obviously important markets to Japanese banks. It should be noted that the
opportunity to increase their size and market share has
come about as the result of acquisitions of other foreignowned banks disillusioned with their retail banking experi-

ence in the Golden State. Despite the "warning" provided
by the departing British, the Japanese banks have made a
sizable long-term commitment to the California banking
market, and they are likely to be an important factor in its
future. Moreover, to the extent that trade and direct investment remain at high levels, the Japanese banks will have a
base for operations that the British apparently lacked.

Appendix
The dependent variables in Tables 1 and 2 are the
individual asset or liability categories as a percentage of
assets. The regressions are estimated across banks, controlling for ownership by using dummy variables for Japanese ownership and for other foreign ownership. Two
other control variables and an error term also are included
in the model. The model appears robust, in that similar
results are generated by adding a control variable for bank
growth over the prior year and by dropping the size and
branch control variables altogether.
Asset size is included as a control variable because
banks in the State range from very large multinationals to
very small local banks, and size may affect their asset and
liability composition. For example, only large banks are
likely to be able to borrow directly in the money markets or

Federal Reserve Bank of San Francisco

to be involved in major corporate financing. Furthermore,
differences in the size mix also may be important across
ownership groups, especially since the Japanese subsidiary banks are relatively large compared to most domestic
banks.
The number of branches for each bank is also included
as a control variable because of possible portfolio differences arising from branch structure. California is a
statewide branching state, and the State's 442 banks operate around 5,000 offices. These branch networks generate
retail deposits and retail lending opportunities and considerable overhead expenses. Different branching patterns
between the ownership groups could account for some of
the variation between groups, especially in terms of access
to retail deposits and retail lending.

15

NOTES
1. Principal Accounts of Overseas Branches of Japanese
Banks, at year-end, 1981 through 1987. Japan Financial
Statistics, Federation of Bankers Associations of Japan,
1989, page 26.
2. Includes total foreign and domestic assets of Japanese-owned banks, subsidiary banks, agencies, and
branches (including Edge Act Corporations). Year-end
1988 data from the December 31,1988, Call Report.
3. Year-end 1988 Call Report data for California banks.
4. Japanese banks have made a significant financial
commitment to penetrating the California market. Together, Bank of Tokyo, Mitsubishi, and Sanwa have invested upwards of $1.3 billion in California acquisitions.
California First (owned by Bank of Tokyo) paid $750 million
to acquire Union Bank in October 1988. Sanwa Bank
of California paid $263 million to acquire Lloyds Bank
California in 1986. Mitsubishi paid $242.5 million to acquire BanCal Tri-State Corp. in 1984. American Banker,
6-20-84, 1-20-87, and 7-20-88.
5. As of 12-31-88, Mitsui Manufacturers Bank ranked
twelfth in the State with $1.6 billion in assets and Tokai
Bank of California, with $939 million in assets, ranked
twentieth. Tokai had ranked tenth through much of 1988.
Two other banks are subsidiaries of Japanese banks, DaiIchi Kangyo Bank, $353 million in assets at year-end 1988,
and Kyowa Bank of California, $114 million in assets.
American Pacific State Bank, $165 million in assets, and
Los Angeles National Bank, $60 million in assets, are also
Japanese-owned, but are not owned by banks.
6. International Banking Facility (IBF) assets, which are
limited to transactions with foreign residents, account for
nearly half of the total foreign agency and branch assets
nationally. See Federal Reserve Bulletin, Assets and Liabilities of U.S. Branches and Agencies of Foreign Banks,
June 30, 1988, page A78.
7. For example, see Richard W. Wright and Gunter A.
Pauli's The Second Wave: Japan's Global Assault on
Financial services, St. Martin's Press, New York, 1987,
page 1.
8. See footnote 2 for the market definition.
9. The market share of the non-Japanese foreign-owned
banking institutions fell from 20.1 percent in 1982 to 6.5
percent at year-end 1988, primarily as a result of the
British retreat from California. British banks controlled
about 15 percent of the assets of banking institutions in the
State at the beginning of the period.
10. Terrell's (1979) study analyzes international banking
on a bilateral basis. He examines the growth of foreign
branch and subsidiary activities of both Japanese and
American banks in the context of foreign investment. Data
for these countries lend themselves to comparisons of
banking and trade in local markets because most of the
banking activities of these two countries are primarily
either for the local market or are trade related.
11. Terrell's interpretation of Caves' basic assumption.

16

12. See Terrell, (1979), page 18.
13. SeeTerrell, (1979), page 26.
14. This process, in fact, closely parallels the expansion
of U.S-: banks overseas a generation earlier, when U.S.
multinational firms were experiencing rapid overseas
growth. See Terrell (1979).
15. Source: Commerce Department. Foreign Trade StatisticsReport: Waterborne Trade, year-end totals. Imports:
Report SM305, Exports: Report SM705.
16. From "Japan's Expanding U.S. Manufacturing Presence: .1987 Update," JEI Report, Japan Economic Institute, December 16, 1988, page 3.
17.• ·Totalassets of Japanese-owned banks, agencies and
branches in the U.S. increased by 216.0 percent over the
period from 1982 to 1988. The increase in New York was
226.3 percent. In California, the increase was only 170.1
percent.
18. Testimony presented at a Special Joint Hearing on
Foreign Investment in California: Banking and Real Estate,
California Legislature, January 25, 1989.
19. SLCs are essentially financial guarantees sold by
large, creditworthy banks. Commercial and industrial
loans provide a potentially cleaner measure of market
share than assets because interbank transactions may
bias assets upwards and inflate market shares.
20. For example, of the ten largest SLC-issuing Japanese-owned agencies in California in May 1988, eight
held Moody's best rating, Aaa, for long-term and senior
debt. The two remaining Japanese-owned agencies had
third highest ratings of Aa2. In contrast, of the top four
domestic banks in the State in May 1988, none had the
highest rating. One had the next highest, Aa1, two had the
third highest, Aa2, and the fourth had a rating of Baa2.
Moody's Credit Opinions, Moody's Investors Service,
New York, May 1988.
21. Source: Federal Reserve Bank of San Francisco, BHC
Reports, December 31,1987. Y-8 Report.
22. 1987 data were used because the late-1988 acquisition of Union Bank (British-owned) by Japanese-owned
California First Bank may have distorted the "Japanese"
bank portfolio. Union Bank's assets were equal to nearly
one quarter of all Japanese bank assets at the time the
merger took place, and to the extent that Union's portfolio
was determined by the former British owners, portfolio
measures would tend to reflect British ownership rather
than Japanese ownership.
23. Also, all banks chartered in the U.S., including foreign
subsidiary banks, are subject to U.S. capital standards.
24. AmericanBanker, "Japanese Banks Tackle California
Middle Market," September 10,1987, and "Bank of Tokyo
Targets California's Middle Market," February 19, 1988.

Economic Review / Summer 1989

REFERENCES
Boyle, Patrick. "Many Japanese Banks With Offices in the
U.S. Get a Boost From Moody's," Los Angeles Times,
January16, 1984.
Bronte, >Stephen. "The Flap over Moody's Downgradings," Euromoney, June 1982.
Caves, Richard. "International Trade, International Investment, and Imperfect Markets," Special Papers in International Economics, Princeton University, Number
10, November 1974.
California Legislature. Senate Committee on Banking and
Commerce and Senate Select Committee on the Pacific .Rim. Foreign Investment in California: Agriculture, Banking and R.eal Estate, Special Joint Hearing.
sacramento, California, January 25, 1989.
Commerce Department. Foreign Trade Statistics Report:

Waterborne Trade, 1980-1987.
Dohner, Robert S.,and Henry S.Terrell. "The Determinants of the Growth of Multinational Banking Organizations: 1972-86," International Finance Discussion
Papers 326. Board of Governors of the Federal Reserve System, June 1988.
Japan Economic Institute. "Japan's Expanding U.S. Manufacturing Presence: 1987 Update," December 16,
1988.
Japan Economic Institute. "Japanese Banks in the United
States," January 22, 1988.
Federation of Bankers Associations of Japan. Japan Fi-

nancialStatistics, 1989.
Hanley, Thomas H., James M. Rosenberg, Carla A.
D'Arista, and Neil Mitchell. "The Japanese Banks:
Positioning for Competitive Advantage," Stock Research, Salomon Brothers, Inc., November 1986.
Heins, John. "Sanwa, They Make Great Stereos," Forbes,
May 30,1988.
Howe, Kenneth. "Japanese Inroads in California Banking," San Francisco Chronicle, February 18,1988.

Federal Reserve Bank of San Francisco

Luke, Robert. "Japanese Banks Tackle California Middle
Market," American Banker, September 10, 1987.
Mendelsohn, M.S. "International i.Loansof Japanese
Banks Surge on Strength of Payments Surplus,"
American Banker, March 4,1986.
Miller, Richard. "Feasting on California," United States
Banker, May 1988.
Mills, Rodney H. "US Banks are Losing their Share of the
Market," Euromoney, February 1980.
Moody's Credit Opinions. Moody's Investors Service, May
1988.
Naff, Clay. "Bank of Tokyo Targets California's Middle
Market," American Banker, February 19,1988.
Nolle, Daniel E., and Charles Pigott. "The Changing Com,~
modity Composition of U.S. Imports from Japan,
Quarterly Review, Federal Reserve Bank of New York,
Spring 1986.
The Economist. "Japanese Banking Booms Offshore,"
November 26, 1988.
Terrell, Henry. "U.S. Banks in Japan and Japanese Banks
in the United States: An Empirical Comparison," Economic Review, Federal Reserve Bank of San Francisco, Summer 1979.
Sakamoto, Tomohiko. "The Japan-U.S. Bilateral Trade,"
Economic Review, Federal Reserve Bank of San Francisco, Spring 1988.
Schaefer, James. "Operating Results of Foreign Based
International Banks in California," mimeo, James T.
Schaefer, CPA, Claremont, CA, June 30, 1988.
Wright, Richard w., and Gunter A. Pauli. The Second

Wave: Japan's Global Assaulton Financial Services,
New York: St. Martin's Press, 1987.
Zimmerman, Gary. "The Growing Presence of Japanese
Banks in California," Weekly Letter, Federal Reserve
Bank of San Francisco, October 28, 1988.

17

Reagan Fiscal Policy and the Dollar

Adrian W. Throop
Research Officer, Federal Reserve Bank of San Francisco.
Research assistance by Eric McCluskey, Jerry Metaxas,
and Panos Bazos is gratefully acknowledged. Editorial
committee members were Frederick Furlong, Jack Beebe,
and Hang-Sheng Cheng.

Whether fiscal policy, monetary policy, or other factors
contributed more to the large swings in the value of the
dollar and the u.s. trade balance in the1980s is the subject
of ongoing debate. Using a simulation from the macroeconomic model developed at the Federal Reserve Bank of
San Francisco, 1 find that the fiscal expansion under the
Reagan Administration was the most important reasonfor
the dollar's appreciation from 1980 to 1985, but that
monetary conditions at home and abroad were primarily
responsible for the dollar's subsequent depreciation
through 1987.

18

During the 1980s, the real exchange value of the dollar
gyrated markedly. From 1980 through early 1985, it appreciated 55 percent. Then through 1987, it depreciated just as
sharply, before enjoying a mild rebound in 1988. Coupled
with these swings in the value ofthe dollar, the U.S. budget
balance worsened very significantly, U.S. interest rates
rose and fell dramatically, and the U.S. external balance
moved sharply into deficit. The reasons for the dollar's rise
and fall have been the subject of considerable debate. This
debate focuses on whether the Reagan fiscal expansion or
the changes in monetary policy-or still other factorsduring the 1980s are more to blame for the movements in
the dollar and interest rates and the resulting decline in the
U.S. trade balance.
Econometric modeling is necessary to sort out the
relative contributions of fiscal and monetary policies. This
article reports the results of such an exercise using a
macroeconometric model developed at the Federal Reserve
Bank of San Francisco.' This model is used to simulate the
economy's path as if there had been no change in fiscal
policy after 1980. The differences between the actual and
simulated values of the dollar, interest rates, and the trade
balance measure the effects of the Reagan fiscal policy on
these variables.
The article is organized as follows. Section I briefly
describes the debate concerning the causes of the movements in the dollar. In Section II, the various dimensions of
the Reagan fiscal policy are quantified and the meaning of
an unchanged fiscal policy is discussed. Then in Section
III, the simulation results from the FRBSF model are
presented. Finally, Section IV provides a summary and
some conclusions.

Economic Review / Summer 1989

I. The Debate
The Reagan years have provided almost a laboratory
experiment on the effects of fiscal expansion in an open
economy. The Reagan Administration's fiscal program
comprised planned reductions in both taxes and spending
to stimulate saving, investment, and work effort in an
economy suffering from low growth and high inflation.
The most prominent feature of this program was the
Economic Recovery and Tax Act of 1981, which cut tax
rates for both businesses and households. Personal tax cuts
were designed to stimulate personal saving by reducing
marginal tax rates. The tax changes for business were
aimed at directing an increasing share of the anticipated
expansion in personal and business saving into investment
in plant and equipment. However, since 1980, the private
saving rate has either fallen or at best remained unchanged,
depending on the measure of saving used.
Total federal receipts as a proportion of GNP dropped
through 1984 and then rose to the same level as in 1980, but
only because of rising Social Security taxes. Most importantly, a buildup in defense spending and continued growth
in spending on entitlement programs, such as Social
Security and Medicare, combined to overwhelm cuts that
were made in other nondefense spending. As a result,
the deficit in the federal government's budget rose from
approximate balance in 1980 to four percent of highemployment GNP in 1986, before declining to 2.4 percent
by 1988.2 Moreover, growing interest payments have
added to this deficit, so that it now absorbs nearly half of
net private saving in the US. economy. 3

If the U.S. economy had been closed to international
flows of capital, the growth in federal borrowing would
have tended to "crowd out" domestic private capital
formation. Because the U.S. economy is highly open,
however, there was a quite different tendency. U.S. interest
rates rose, large capital inflows were attracted from abroad
to finance the federal deficit, the real value of the dollar
rose, and the US. trade balance moved significantly into
deficit.
Whether the fiscal expansion is solely, or even primarily,
responsible for the rise in interest rates and the value of the
dollar, as well as the decline in the trade balance, has been
the subject of lengthy debate, however. At about the same
time that the US. budget deficit was increasing, the
Federal Reserve was attempting to reduce inflation through
tighter monetary policy. Some argue that it was the tightening of monetary policy that pushed up interest rates and
primarily was responsible for the dollar's rise in the early to
mid-1980s. A similar debate arises regarding the causes of
the dollar's decline after 1984. Some argue that the US.
budget deficit as a percent of high employment GNP began
to tum down after 1986, tending to put downward pressure
on US. interest rates and the dollar in this period. But at
the same time, others suggest that the declines in interest
rates and the dollar were primarily the result of the easing
in monetary policy that took place during this period as the
Federal Reserve's disinflationary goals were achieved.

II. Dimensions of Reagan Fiscal Policy
To shed some light on this debate, the FRBSF macroeconometric model is used to simulate the economy's path
as if there had been no change in fiscal policy after 1980,
and the results are compared to the actual path of the
economy. To perform such an exercise, it is useful to
evaluate the components of the Reagan fiscal expansion to
determine what it did and did not change in the economy.
For our purposes, fiscal policy is defined in terms of its
macroeconomic effects, as opposed to specific legislative
changes. Thus, fiscal policy may be altered even when
there are no legislated changes. One such example is the
increase in both marginal and average tax rates that is
generated by inflation when taxes are not indexed to
changes in the price level. Conversely, legislative changes
may be required just to keep the effects of fiscal policy
from changing when, for example, taxes have to be cut in
order to keep revenues from rising as a fraction of GNP.

Federal Reserve Bankof San Francisco

From a macroeconomic point of view, there are two
dimensions to the measurement of an unchanged federal
fiscal policy. First, there should be no change in federal
marginal tax rates that would alter economic incentives.
Forexample, in the FRBSF model the average marginal tax
rate for households significantly affects their after-tax
mortgage rate and, therefore, influences expenditures on
housing. Similarly, business taxes influence the cost of
capital for nonresidential investment and rental housing.
An unchanged fiscal policy is defined as one that does not
alter the marginal tax rates that affect these expenditures,
and therefore does not cause the IS curve to shift.
Second, an unchanged fiscal policy requires federal
outlays and receipts not to change as a fraction of GNP at
high employment. Unchanged receipts would prevent disposable income, and hence consumption, from changing
on account of fiscal policy. With unchanged government

19

receipts and outlays, as well as unchanged marginal tax
rates, there should be no shift in the IS curve and, thus, no
change in aggregate demand due to changes in fiscal
policy.4

Marginal Tax Rates
As shown in Table 1, the Economic Recovery and Tax
Act of 1981 and the Tax Reform Act of 1986 reduced the
average marginal federal tax rate on individual income
from 30 percent in 1980 to 23 percent by 1988. In the
counterfactual simulation that removes the effects of Reagan fiscal policy, the average federal marginal tax rate for
households is held constant at 30 percent from 1980
through 1988, instead of being allowed to fall. As a
result, after-tax interest rates for households are reduced
relative to their historic values. This has the effect of
raising expenditures on consumer durables such as owneroccupied housing relative to the actual expenditures during
this period.
The Tax Act of 1981 also reduced effective tax rates on
business investment by shortening depreciable "tax lives"
and increasing the investment tax credit for purchases of
equipment. The Tax Equity and Fiscal Responsibility Act
of 1982 took back part, but by no means all, of these tax
cuts for business as part of a package to reduce the size of
the federal budget deficit. Then in 1986, the Tax Reform

d

d

=

inal return to equity;
long-term inflation rate;
.al rate at which the capital good
depreciates.

evident from this formula, the rental cost of
fixed capital is equal to some fraction of the price
apital good. This fraction is determined by the real
debt and equity capital and, central to our inquiry, a
ative factor that depends upon the corporate tax
rate, the present value of depreciation and the investment
tax
it. (The investment tax credit was positive for
ent u
86 and always has been zero for commercial an
.al structures and rental housing.) This
factor equals one plus the effective tax rate
f equit -financed investment. The Reagan
ost of business investment
iation (z) and the
lue 0
thereby
cing this multiplica-

20

Economic Review / Summer 1989

Act reduced the corporate income tax rate from 46 percent
to 34 percent, but at the same time eliminated the investment tax credit for equipment and lengthened the tax lives
for residential and nonresidential structures. The net result
of these changes was that by 1988, the effective tax rate on
investment in equipment was about the same as in 1980,
but effective tax rates on rental housing and nonresidential
structures were lower. (See Table I and the accompanying
Box.) In the counterfactual simulation, these effective tax
rates are held at their 1980 values, on balance reducing the
incentive for business investment, and thus its contribution
to aggregate demand, compared with the actual path of
business investment.

Government Spending and Disposable Income
Observed movements in federal outlays and receipts are
partly due to changes in spending that are automatically
triggered by changes in the level of economic activity. The
federal budget measured on a high employment basis
removes these cyclical variations in outlays and receipts
caused by the economy's deviations from its path of high
employment. In contrast, changes in high-employment
outlays and receipts that deviate from a constant proportion
of high-employment GNP constitute a real change in the
direction of fiscal policy> As shown in Table 2, the federal
high-employment budget deficit rose from 0.3 percent of
high-employment GNP in 1980 to 4.0 percent in 1986, and
then dropped back to 2.4 percent of GNP by 1988.
Looking at the components of that deficit, the most
permanent contributor was an increasing ratio of federal
transfer payments to GNP. In contrast, purchases of goods
and services as a proportion of high-employment GNP
rose a little more than one percentage point through 1985,
but returned almost to their 1980 level by 1988. On the
revenue side, there was little net change in the ratio of
total federal tax receipts to GNP between 1980 and 1988.
Although the ratio of total income tax receipts to GNP
declined by two percentage points, a rise in Social Security
taxes offset this decline. In the FRBSF model, the impact
of policy-induced changes in total federal receipts and
transfer payments on household disposable income, and
hence consumption, is captured by the ratio of cyclicallyadjusted federal taxes less transfer payments to highemployment GNP. 6 As shown in Table 2, this ratio
declined from 9.7 percent in 1980 to around 7.0 percent in
1985, with no discernible trend since then.
Also, part of the Reagan fiscal package was a reduction
in the amount of grants-in-aid to state and local governments as shown in Table 2. These governments were able
to absorb the grant reductions and maintain approximately

Federal Reserve Bank of San Francisco

the same level of services by raising taxes toward the end of
the 1981-82 recession (see Weicher [1987]). The change in
the federal high-employment budget captures the overall
impact, with the reduction in grants-in-aid acting as a
proxy for the fiscal restraint achieved through higher state
and local taxes. But the ratio of cyclically-adjusted federal
taxes less transfer payments to high-employment GNP
overstates the total reduction in net taxes and transfers due
to the Reagan program. Consequently, in the simulation,
the reduction in the level of federal taxes less transfer
payments is adjusted for the increase in state and local
taxes.
In the counterfactual simulation of no change in fiscal
policy, the ratios of the high employment values of the
federal budget deficit, federal purchases of goods and
services, and federal taxes net of transfer payments to
high-employment GNP are all maintained at their 1980
levels, after taking into account the effects of reduced
grants-in-aid and higher state and local taxes. This has the
effect of reducing the contribution to aggregate demand of
both government purchases of goods and services and
personal consumption expenditures from what they were
historically, but the impact of these factors is slightly offset
by the effect of the Reagan Administration's policy with
respect to grants-in-aid to state and local governments.

21

III. Simulated Effects of Reagan Fiscal Policy
A key feature of the FRBSF macroeconometric model
that is used for the simulation presented below is that it
treats the real value of the dollar expected in the long run as
an endogenous variable determined by expectations of
future fiscal policy. As a result, current fiscal policy
influences the dollar through two channels. The first channel is the current level of interest rates, or more specifically, the differential between real interest rates at home
and abroad. Assuming, as quite a few macroeconometric
models do, that 1) capital is perfectly mobile internationally in the sense that there are no significant transaction
costs, capital controls, or other impediments to the flow of
capital between countries, 2) domestic and foreign financial assets are perfect substitutes, and 3) trade flows are
slow to adjust, then exchange rates are determined in the
short run by equilibrium in the market for financial assets,
rather than by equilibrium between current international
flows of goods and capital. This implies that a rise in longterm interest rates at home relative to those abroad, which
is produced by an expansionary fiscal policy at home, will
cause the value of the home currency to appreciate until the
difference between its current level and its expected longrun level (that is, its expected depreciation) is equal to the
interest rate differential.
In addition to this short-run asset equilibrium view
embedded in the FRBSF model, the model also incorporates a rational expectations view of the determination of
the expected long-run level of the exchange rate. This
second channel arises because changes in current fiscal
policy alter expectations of future fiscal policy which, in
turn, alter the expected long-run real exchange rate. As
analyzed in detail in Throop (l989b), if market participants
view US. and foreign assets as close substitutes, then an
expectation of future US. budget deficits is likely to
increase the expected real value of the dollar. But if they
believe that US. and foreign securities are relatively
imperfect substitutes, then a depreciation in the expected
real value ofthe dollar is more likely. Thus, the magnitude
of the effect operating through this additional channel
depends on 1) the degree to which changes in current fiscal
policy alter expectations of future fiscal policy, and 2) the
size and direction of the effect of future fiscal policy on the
expected real value of the dollar.
In the exchange rate equation in the FRBSF model, the
budget deficit influences the exchange rate through both
channels, but simulations indicate that the second channel is more important than the first one. (See Throop
[1989b].) With regard to the first channel, a sustained onepercentage point change in the real short-term interest rate

22

differential is estimated to produce a 10 percent change in
the real trade-weighted value of the dollar in the same
direction. The magnitude of this effect is consistent with
an average horizon for investors in the foreign exchange
market of ten years.
With regard to the second channel of influence, the
exchange rate equation in the FRBSF model indicates that
market participants view US. and foreign assets as close
substitutes and changes in structural budget deficits as
being relatively permanent. Thus, a one-percentage point
reduction in the current US. structural budget surplus as a
percent of high-employment GNP produces a six percent
appreciation in the expected real trade-weighted value of
the dollar, while a similar reduction in the weighted
average of foreign budget surpluses depreciates the dollar
by eight percent. 7 These expectational effects are relatively large. In the FRBSF model, it takes about a nine
percent appreciation in the real value of the dollar to
produce an effect on the trade balance that fully offsets the
effect on aggregate demand from a one-percentage point
reduction in the US. budget surplus. The estimated six
percent appreciation generated by the expectational effects
of a budget surplus is fully two-thirds of this.
In modeling the exchange rate and the trade balance, one
needs to take into account the reaction of foreign central
banks to changes in US. interest rates. Floating exchange
rates have diminished the short-run monetary linkages
among national real interest rates. Nonetheless, foreign
central banks continue to pursue macroeconomic stabilization and so continue to respond to changes in US. interest
rates, though to a lesser degree than before. For example,
foreign central banks tend to allow their countries' interest
rates to rise in response to a rise in US. rates to prevent
capital outflows and a depreciation of their currencies that
would result in an increase in aggregate demand and
higher output and inflation. However, matching the rise in
US. interest rates exactly would have a deflationary impact on foreign economies. As a result, foreign central
banks have tended to match some, but not all, of the
changes in US. real interest rates so as to stabilize
aggregate demand. The reaction function in the FRBSF
model indicates that foreign central banks tend to match
about 55 percent of the change in US. real short-term
interest rates on average.
A final important factor in the simulated effects of
Reagan fiscal policy is the impact of the real exchange rate
on the real stock of money through its influence on the price
level in the FRBSF model. In the FRBSF model, a 10
percent appreciation in the real trade-weighted value of the

Economic Review / Summer 1989

dollar reduces prices by 0.8 percent over a period of about
two years through competitive effects on domestic prices
of exports and import substitutes. Because of this, a fiscal
expansion in the FRBSF model causes the real value of the
dollar to appreciate, prices to drop, and the real stock of
money to expand-relieving some of the pressure on
interest rates and allowing real GNP to expand more than it
otherwise would.
The simulated effects of the Reagan Administration's
fiscal policy are shown in Table 3. Column A gives the
actual values of the dollar, net exports, real GNP, and other
variables of interest during the period of the Reagan
Administration's fiscal expansion. The results from a simulation that assumes no fiscal expansion and uses the actual
path for M2 after 1980, are given in column B. The effect

Federal Reserve Bank of San Francisco

of the Reagan fiscal policy on any variable is then equal to
column A less column B.

Effects on the Dollar
This simulation sheds light on the relative contributions
of U.S. fiscal policy and other factors to the sharp 55
percent appreciation in the real trade-weighted value of the
dollar between 1980 and 1985. As shown in the Table and
Chart 1, the real trade-weighted value of the dollar would
have appreciated 30 percent from 1980 to 1985, even if
fiscal policy had not been expansionary.
The Reagan fiscal program caused the dollar to appreciate. an additional 25 percent, by far the largest single
contributor to the dollar's rise. Chart 2 shows the relative

23

contributions of the underlying determinants of the real
trade-weighted value of the dollar with and without the
Reagan fiscal stimulus. Of particular interest is that the
real interest rate differential would have been little different from its historical path even if the Reagan fiscal
expansion had not occurred. Given the historical path of
M2, the Reagan fiscal package is estimated temporarily to
have increased the U.S. real bond rate only 21basis points,
and the differential between the U.S. and foreign real bond
rate only 16 basis points. Taking out these movements in
interest rates associated with the Reagan fiscal package,
the dollar's appreciation would have been only 1.7 percent
less than it actually was," And even if the simulation had
not assumed the historical path for M2 and instead had
assumed a path for monetary policy that kept real GNP on
its historical path, the Reagan fiscal package would be
estimated to have increased the real bond rate differential
only marginally. Thus, the lion's share of the 25 percent
appreciation in the dollar caused by the Reagan fiscal
program was the result of the expectational effects of the
budget deficit, not the rise in the interest rate differential.
Of the 30 percent appreciation in the real value of the
dollar that would have occurred in the absence of any
change in fiscal policy, about equal contributions can be
assigned to a tightening in domestic monetary conditions,
fiscal tightening in the U.S.' major trading partners, and
unexplained speculative factors that appear to have been
present in 1985. Primarily as a result of the Federal
Reserve's attempt to reduce inflation, the U.S. real bond

rate rose between,1980 and 1984. And although foreign
countries also were pursuing policies that raised interest
rates (to counter the inflationary effects of a strong dollar
on their economies), foreign real rates did not rise by as
much as those in the U.S. These monetary factors-rather
than fiscal factors-accounted for most of the rise in the
real interest rate differential.
After 1985, the effects of the Reagan fiscal program on
changes in the value of the dollar were relatively small. In
the absence of the Reagan fiscal expansion, the dollar
would have declined by about as much as it actually did,
but starting from a lower level. In this period, a decline in
the real interest rate differential from about four percentage points in 1985to less than one percentage point in1988
accounts for close to 80 percent of the dollar's depreciation. The decline in this differential primarily was the
result of two developments. During this period, the U.S.
real bond rate declined as the Federal Reserve's disinflationary goals were achieved and monetary policy eased. At
the same time, foreign real bond rates continued to rise as
foreign central banks tightened policy in response to the
inflationary effects of the strong dollar on their economies.
The strong dollar tended to create inflation abroad both
directly through higher prices of tradable goods and indirectly through the boost to aggregate demand from increased exports to the U.S.
Thus, fiscal conditions at home and abroad were relatively more important than any other single factor in
raising the real value of the dollar through 1985. But U.S.

Chart 1
Natural
Logarithm

Real Trade-Weighted U. S. Dollar

4.1
Reagan
Fiscal Polley

4.0
3.9
3.8
3.7
3.6

Unchanged
Fiscal Polley

3.5
3.4
3.3
1980

24

1982

1984

1986

1988

Economic Review I Summer 1989

fiscal policy contributed relatively little to the decline in
the real value of the dollar from 1985 to 1988.

Reagan fiscal policy is estimated to have raised real
GNP by $75.0 billion and reduced the civilian unemployment.rate nearly one percentage point by 1986. However,
becauseof the gradual crowding out of net exportsand, to
a-lesser.exrenr, interest-sensitive expenditures, as well as
the modest tightening in fiscal policy after 1986, by 1988,
the.overallgain to real GNP totalled only $37.5 billion,
and.the civilian unemployment rate was only about onehalfpercentage point less than it wouldhave been without
the Reagan program.
Finally, because the Reagan fiscal package caused the
dollarto appreciate dramatically, it also had a dampening
effect on the price level. On balance, there was a 1.6
percent reduction in the price level by 1988, as the effects
of the dollar's appreciation overwhelmed those from reduced slack in the economy. The dollar's dampeningeffect
on prices also increased the real stock of M2, and thereby
helped to keep interest rates lower than they otherwise
would have been.

Effects on the Trade Balance

Working through its effect on the dollar and incomes,
the Reagan fiscal program accounted for $85.3 billion, or
about 85 percent, of the deficit in real net exports of $99
billion by 1988. In the absence of any change in fiscal
policy, real net exports wouldhave droppedapproximately
$100 billion between 1980and 1986, due to the economy's
rebound from the 1980 and 1982 recessions and the 30
percent appreciation of the dollar. That would have occurred between1980and 1985 evenif fiscal policyhad not
been expansionary. However, the subsequent depreciation
would have tended to restore net exports. In fact, by 1988,
the deficit in net exports would have been only $13.5
billion, or near its average level in the decade of the 1970s
if the fiscal expansion had not occurred.

IV. Summary and Conclusions
The Reagan Administration's fiscal policies have provided close to a laboratory experiment on the effects of
fiscal expansion in an open economy. Simulation of the
FRBSF macroeconometric model shows that significant
effects from this expansionary fiscal policybeganto be felt
on the dollar and the U.S. trade balance after 1982. The
Reagan fiscal packageaccountedfor nearlyone-halfof the

55 percent appreciation in the real trade-weighted valueof
the dollar between 1980 and 1985, and about 70 percentof
the peak value of the deficit in real net exports in 1986,
assuming an unchanged path for M2. The dollar would
have appreciatedabout 30 percent between 1980 and 1985
even in the absence of the Reagan Administration's fiscal
package, however, becausetighteningmonetary conditions

Chart 2
Contributions of Various Factors
to Real Trade-Weighted U.S. Dollar
Contributions to
Natural Logarithm

0.6
Reagan
Fiscal Polley

0.5
Real Interest
Rate Differential

0.4

•

Unchanged
Fiscal Polley

0.3
0.2
0.1

Expected Foreign
Budget Surplus

0.0
~ Expected U.S.

-0.1

~

Budget Surplus

-0.2
1980

Federal Reserve Bank of San Francisco

1982

1984

1986

1988

25

in the US. raised the differential between US. and foreign
real long-term interest rates. Also, even if the Federal
Reserve had reduced interest rates to the extent necessary
to stabilize real GNP in the absence of a fiscal expansion,
the path of the dollar would not have been much lower.
When capital is highly substitutable internationally, and
investors adjust their expectations of the long-run equilibrium value of the dollar relatively quickly, an expansive
fiscal policy has relatively little impact on real interest
rates. Because of this, the increase in the differential
between US. and foreign real interest rates between 1980

and 1985primarily was due to a growing disparity between
monetary conditions at home and abroad, rather than to the
expansive Reagan fiscal policy. The contribution of this
difference in monetary conditions to the strong dollar was
temporary, however, and had disappeared by 1988. In
contrast, although the US. budget deficit on a highemployment basis declined somewhat after 1986, US.
fiscal policy continued to expand income and boost the real
value of the dollar. As a consequence, in 1988, the Reagan
fiscal policy still was contributing $85 billion (in 1982
dollars) to the trade deficit.

NOTES
1. The complete FRBSF macroeconometric model is fully
described in Throop (1989a).
2. The measure of the deficit used here counts the erosion in the real value of the federal debt due to inflation as
a tax revenue.
3. For more detail on these points, see Modigliani (1988).
4. Some might argue that these dual criteria for an unchanged fiscal policy are mutually inconsistent. For example, if marginal tax rates are higher than average rates, as
in fact they generally are, normal growth in the economy
with fixed marginal rates would tend to raise tax receipts
as a proportion of GNP. However, an unchanged fiscal
policy that meets both criteria could be maintained by
reducing average tax rates without changing marginal
rates. In the case of personal income taxes, this could be
done by increasing the standard deduction. The extent of
progressivity in the tax structure is much less for corporations, but here too, the average tax rate could be reduced
without changing the marginal tax rate on the cost of new
investment.

5. For this purpose, I use the U.S. Department of Commerce's "mid-expansion" measure of high employment.
See de Leeuw, et at. (1980), de Leeuw and Holloway
(1982), and Holloway, Reeb, and Dunson (1986).
6. Taxes on corporate profits are given a weight of only
0.5 because corporations on average retain approximately half of their earnings. Thus, only half of any given
change in taxes on corporate profits will affect disposable
income.
7. The budgetary data used in the exchange rate equation on an inflation-adjusted basis are combined federal,
state, and local balances compiled by the OECD. Sources
of these data are Price and Muller (1984) and recent
issues of the DECO Economic Outlook.
8. This contrasts with the conventional view of the effects
of fiscal policy in which a budget deficit's only channel of
influence on the exchange rate is through the real longterm interest rate differential. When expectational effects
are taken into account, as in the FRBSF model, the impact
of a fiscal expansion on the interest differential is very
much reduced. See Throop (1989b).

REFERENCES
de Leeuw, Frank and Thomas Holloway. "The High Employment Budget: Revised Estimates and Automatic
Inflation Effects," Survey of Current Business, April
1982.
de Leeuw, Frank, Thomas M. Holloway, Darwin G. Johnson, David S. McClair, and Charles A. Waite. "The
High Employment Budget; New Estimates: 1955-80,"
Survey of Current Business, November 1980.
Holloway, Thomas, Jane S. Reeb, and Joy D. Dunson.
"Cyclical Adjustment of the Federal Budget and Federal Debt: Undated Detailed Methodology and Estimates," Bureau of Economic Analysis StaffPaper 45,
November 1-986.
Modigliani, Franco, "Reagan's Economic Policies: A Critique," Oxford Economic Papers, September 1988.

26

Price, Robert W R. and Patrice Muller. "Structural Budget Indicators and the Interpretation of Fiscal Policy
Stance in OECD Economies," DECO Economic Studies, Autumn 1984.
Throop, Adrian W "A Macroeconometric Model of the
U.S. Economy," Working Papersin Applied Economic
Theory, No. 89-01, Federal Reserve Bank of San
Francisco, (Revised) March 1989a,
_ _ _ _ . "Fiscal Policy, The Dollar, and International
Trade: A Synthesis ofTwo Views," Economic Review,
Federal Reserve Bank of San Francisco, Summer
1989b.
Weicher, John C, "The Domestic Budget After GrammRudman-and after Reagan," in Phillip Cagan, ed.
Deficits, Taxes, and Economic Adjustments. Washington, D,C,: American Enterprise Institute, 1987,

Economic Review / Snmmer 1989

Fiscal Policy, the Dollar, and International Trade:
A Synthesis of Two Views

Adrian W. Throop
Research Officer, Federal Reserve Bank of San Francisco. Research assistance by Eric McCluskey, Jerry
Metaxas, and Panos Bazos is gratefully acknowledged.
Editorial committee members were Frederick Furlong,
Jack Beebe, and Hang-Sheng Cheng.

In most macroeconometric models, a larger budget
deficit leads to an appreciation ofa nation's currency and
a rise in its trade deficit only to the extentthat it drives up
thedifferential between interestratesat homeandabroad.
In contrast, Mundell'spioneering worksuggeststhata rise
in the budget deficit does these things withoutany change
in the interest rate differential because market expectations adjust instantaneously to the effects of the larger
budgetdeficit. The FRBSF macroeconometric modelsynthesizes these twoapproaches by makingthe expected real
value ofthedollar afunction ofcurrentfiscal policies both
athomeand abroad. The resultis thatbudgetdeficits have
a significantly largerinfluence on theexchange rate, and a
smaller impact on interest rates, than in most macro
models. Consequently, expansionary fiscal policy tendsto
crowd out net exports more than interest-sensitive expenditures.

Federal Reserve Bank of San Francisco

There are two distinct views in the literature on the
causal nexus between fiscal policy and the real value of the
dollar. The view that has gained ascendancy in recent years
is that expansionary US. fiscal policy appreciates the real
value of the dollar only to the extent it puts upward pressure
on US. real interest rates and increases the differential
between US. and foreign real interest rates. According to
this view, the long-run equilibrium value of the dollar does
not change, and the real value of the dollar is bid up in the
short-run to the point where the real interest rate differential compensates for the dollar's expected depreciation in
the future. Because it causes interest rates and the dollar to
rise, expansionary fiscal policy crowds out both interestsensitive expenditures and net exports. Most multicountry
econometric models incorporate this view. However, recent simulations of these models suggest that this view can
explain only about one-half of the dollar's rise after 1980.1
An alternative view consistent with the pioneering work
of Robert Mundell (1963) stresses that because of an
adjustment in expectations, capital inflows can be attracted
to finance a U.S. budget deficit even with no increase in the
real interest rate differential. Assuming the market regards
the US. budget deficit as lasting more than temporarily,
the market's longer-run expectation of the real value of the
dollar will rise. This change in expectations, in turn,
produces an appreciation in the real value of the dollar. The
higher dollar, then, creates a "twin deficit" in the trade
balance, which allows actual capital inflows to take place
without there being any increase in the differential between U.S. and foreign real interest rates. Except to the
extent that the fiscal expansion also raises the world level of
interest rates, only net exports would be crowded out
according to this view.2
A synthesis of these two views is possible. By embedding a rational expectations model of the dollar's
longer-run equilibrium into the short-run dynamics of asset
equilibrium, these two distinct linkages among fiscal
policy, the dollar, and trade imbalances can be captured.
Such a synthesis is modeled empirically in the international sector of the FRBSF macroeconometric model,
which is used for forecasting and policy simulations at the

27

Federal Reserve Bank of San Francisco." This paper describes the international sector of the FRBSF model and
assesses the relative quantitative importance of each of
these two linkages.
The paper is organized as follows. Section I reviews the
conventional approach to modeling the effects of fiscal
policy on the dollar and the trade balance. Section II
discusses the unique features of the international sector of

the FRBSF macroeconometric model and the determinants
ofthe real value of the U.S. dollar during the 1980s in this
model. Section III contrasts the simulated effects of a
sustained shift in fiscal policy on the dollar and the trade
balance obtained from the FRBSF model with those obtained from a conventional model. Finally, Section IV
provides a summary and conclusions.

I. Conventional Model of Fiscal Policy in an Open Economy
Most macroeconometric models, including the FRBSF
model, are disaggregated, dynamic versions of the basic
IS-LM model on the demand side, with gradual wage and
price adjustments on the supply side. In addition, most
models assume the degree of international capital mobility
is relatively high, so that interest rates have a direct and
significant effect on the exchange rate. The most direct
approach assumes perfect capital mobility and perfect
asset substitutability between domestic and foreign bonds,
so that expected yields-including the portion due to
expected changes in exchange rates-are equalized at any
moment in time. The available evidence suggests that this
is a reasonably good approximation to reality for major
industrialized countries." Although not all macroeconometric models assume perfect asset substitutability, the
interest rate differential and expected rate of appreciation
or depreciation in the exchange rate are among the important explanatory variables determining the level of the
exchange rate in most of them. 5
Perfect capital mobility and the trade account's slow
adjustment to changes in exchange rates, in turn, imply
that exchange rates are determined in the short run by
equilibrium in the market for financial assets, rather than
by equilibrium between current international flows of
goods and capital." Assuming securities at home and
abroad are perfect substitutes for one another, the asset
theory of the exchange rate requires that the difference
between the nominal returns on securities of a given
maturity at home and abroad is equal to the expected
percentage change in the nominal exchange rate over that
period. This is called the "open interest parity condition."
If, for example, the rate differential exceeded the expected
depreciation in the exchange rate, market arbitrage would
bid the value of the exchange rate up until its expected
depreciation over the relevant time horizon equaled the rate
differential. It is easily shown that this arbitrage condition
also holds in real, or price-adjusted terms." Thus,

InEXCH - InEXCHe = n[(i-pe) - (i*-pe*)]

28

(1)

where: EXCH = current real value of the dollar, defined
as units of foreign currency per unit of
domestic currency deflated by the ratio
of foreign prices to domestic prices
EXCHe = real value of the dollar expected n years
in future
= U.S. nominal interest rate on security
maturing in n years
i*
= foreign interest rate on security maturing in n years
pe
= expected U.S. (annualized) inflation
rate over n years
pe*
= expected foreign (annualized) inflation rate over n years
The difference between the current real exchange rate
and its expected future value-that is, the expected change
in the real value of the currency-is thus proportional to
the real interest rate differential. A central aspect of this
theory is the importance of term structure effects. For
example, assuming n is 10, a one-percentage point rise in
the one-year U.S. real interest rate relative to the foreign
one-year rate would cause the 1O-year rate in the U.S. to
increase by 0.1 percentage points (assuming future one
year interest rates are not expected to change as well), and
the real value of the dollar to go up only one percentage
point. In contrast, a one percentage-point increase in the
U.S. one-year rate that is expected to last for ten years
would cause the 1O-year U.S. bond rate to rise by one
percentage point and the real exchange value of the dollar
to rise by 10 percentage points. So the same movement in
the one year rate generates different movements in the
exchange rate depending on the change in the 10 year rate.
Thus, if movements in long- and short-term interest rates
are not perfectly correlated, the movement in the long-term
real interest rate differential controls movements in the real
exchange rate. In the short run, the expected real long-run
equilibrium value of the dollar does not change, and the
current exchange rate moves in proportion to the long-term

Economic Review I Summer 1989

interest rate differential. In the longer run, however, the
expected real exchange rate tends to be consistent with the
long-run equilibrium value of the exchange rate, so that
with perfect asset substitutability, real interest rates at
home and abroad tend to be equalized.
The implications of using the open interest parity condition to determine the exchange rate in the short run can be
illustrated in the basic IS-LM framework. For this shortrun analysis, fixed wage rates are assumed, but changes in
the real value of the dollar are allowed to have an impact on
the price level. For the moment, the foreign real interest
rate is assumed to be fixed. The conventional approach
assumes that the expected real value of the dollar in the
long run is a constant, determined by the condition of
purchasing power parity. This implies that the current real
value of the dollar can be expressed as a simple function of
the U.S. long-term real interest rate. The result is that the
absolute value of the slope of the IS schedule (locus of
equilibrium in the goods market) is less than it would be in
a closed economy on account of the indirect response of net
exports to the real interest rate occurring through the real
exchange rate.
A further aspect of the open interest parity condition is
the effect on the slope of the LM schedule (locus of
equilibrium in the money market). An appreciation in the
real value of the dollar reduces prices both directly through
the lower relative price level of foreign goods and indirectly
through the competitive pressures placed on domestic
producers of tradable goods. Therefore, as the real interest
rate and real value of the dollar rise, the price levelfalls and
the real stock of money rises. This effect tends to reinforce
the reduction in the quantity of money demanded at the
higher rate of interest, resulting in a less steeply sloped LM
schedule than in a closed economy.
The effect of a fiscal expansion in this conventional
framework is illustrated in Figure 1. Assume that U.S. and
foreign real interest rates are initially at r 1, with the
equilibrium U.S. real GNP at y.. A fiscal expansion, due
to either an increase in government spending or a cut in
taxes, would shift the IS schedule from IS j to IS2 , raising
the U.S. real interest rate to r2 and U.S. real GNP to h.
Real GNP would rise through an increase in the velocity of
money produced by higher interest rates. However, the rise
in interest rates would offset a portion of the initial effects
of fiscal expansion by contracting domestic investment,
and possibly also consumption, and also by contracting net
exports through the associated appreciation in the real
value of the dollar.
The conventional exchange rate analysis, based on
interest differentials alone, implicitly assumes that any

Federal Reserve Bank of San Francisco

Figure 1
Effects of Fiscal Expansion
Real
Interest
Rate

IS2
IS3
IS1

change in fiscal policy is not expected to last, so that the
expected long-run real value of the dollar is not affected.
Thus, fiscal policy affects the real value of the dollar only
through its influence on the current differential between
U.S. and foreign real interest rates. A U.S. fiscal expansion opens up a positive interest rate differential which
appreciates the exchange value of the dollar so as to equate
expected yields. Moreover, fiscal crowding out in these
models always falls partly on interest-sensitive domestic
expenditures since a positive interest rate differential
would not be sustained if the dollar were to rise by enough
to place all the crowding out on net exports.
One problem with this approach is that changes in fiscal
policy generally are fairly long lasting. As a consequence,
expectations that the exchange rate will return to its
original level will continually be disappointed. Specifically, the actual real exchange value of the dollar will
exceed the expected value as long as the fiscal expansion
lasts. That is, as long as the fiscal expansion lasts, the
expected depreciation will not occur. Thus, it seems logical
that the market eventually would begin to revise its expectation of the long-run exchange rate upward. In doing so,
the current real value also would rise and the positive real
interest differential in favor of the U.S. would fall until the
longer-run equilibrium of no differential between U.S. and
foreign real interest rates eventually would be reached.
This process is depicted in Figure 1. The rise in the
expected real value of the dollar has the effect of shifting
the IS schedule downward (through a reduction in net
exports at any given interest rate), and the LM schedule
downward also (through the increase in the real stock of

29

money caused by an appreciating dollar). As long as the
U. S. real interest rate exceeds the foreign real interest rate,
the current exchange rate will continue to exceed the
expected exchange rate, and the IS and LM schedules will

continue to shift down through an adaptive adjustment of
expectations until a full equilibrium at r j , and Y3 is reached
at the intersection of LM 3 and IS 3 . 8

II. International Sector of the FRBSF Macroeconometric Model
A key feature of the FRBSF macroeconometric model is
that it treats the expected real value of the dollar in the long
run (EXCHe) as an endogenous variable determined by
expectations of future fiscal policy. As a result, current
fiscal policy influences the dollar through another channel
besides the current level of interest rates. By altering
expectations of future fiscal policy, it also influences the
real value of the dollar through its effect on the expected
real value of the dollar. The magnitude of the effect
operating through this additional channel depends upon: 1)
the size of the effect of changes in current fiscal policy on
expectations of future fiscal policy, and 2) the size and
direction of the effect of expected future fiscal policy on the
expected real value of the dollar.
The effect of future fiscal policy on the real value of the
dollar can be modelled in a two-country (the U.S. and the
rest of the world), long run, or full employment, equilibrium framework. Dornbusch (1983) and Dornbusch and
Blanchard (1984) have suggested a useful diagrammatic
approach, shown in Figure 2. The locus offull employment
equilibrium in the United States is given by G~s. This
schedule slopes downward because at full employment an
increase in the real value of the dollar reduces net exports,
and so must be offset by the higher U.S. spending brought
about by a lower U.S. real interest rate. Similarly, the locus
of full employment equilibrium for the rest of the world
slopes upward. A rise in the dollar expands net exports
abroad and so must be offset by a higher real interest rate
abroad to produce an offsetting change in aggregate demand. Assuming perfect capital mobility and perfect asset
substitutability, real interest rates will equalize in the long
run, and full employment equilibrium will occur at the
intersection of these two schedules, at point a. At this
intersection, real interest rates in the two countries are
equal, and the real exchange rate produces trade balances
that are consistent with full employment. Equilibrium
capital flows, in turn, are mirror images of the trade
balances.
A U.S. fiscal expansion increases domestic demand for
U.S. goods and services. This shifts the U.S. schedule to
the right from G~s to Gys because, for any given real
exchange rate, higher U.S. real interest rates are needed to
offset the rise in domestic spending and restore equilibrium. Since some of the increase in U.S. domestic

30

demand is spent on imports, the U.S. fiscal expansion
also shifts up the locus of full employment equilibrium
for the rest of the world, Grow, through the increased
demand for rest-of-the-world net exports. But most of the
rise in world aggregate demand falls on U.S. output, so
GUs shifts up by more than Grow. The larger increase in
demand in the U.S. appreciates the real value of the dollar,
which in turn diverts private demand away from U.S.produced goods towards foreign-produced goods. A general equilibrium is restored at point b, where the higher
level of world interest rates dampens the excess world
aggregate demand created by the U.S. fiscal stimulus, and
the dollar appreciation dampens the relative excess demand for U.S.-produced goods."
However, it is possible for the dollar to depreciate if
investors come to think that at some point, foreigners will
demand a higher return on U.S. assets to absorb an
increasingly large share of U.S. debt in foreign portfolios.
Thus, in explaining current movements in the dollar, a
fundamental issue is whether the market believes U.S. and
foreign assets are, and will continue to be, close to perfect

Figure 2
Effects of U.S. Fiscal Expansion
When U.S. and Foreign Assets
Are Perfect Substitutes
U.S. Real
Interest Rate

Rest-of-World
Real Interest Rate

us

row

R

r

1
row

R

1
us

0

r

0

us
G

0

......

•

Real Value of Dollar (EXCH)
Depreciation
Appreciation

..

Economic Review / Summer 1989

substitutes. If markets come to expect imperfect substitution between U.S. and foreign assets, a risk premium for
holding U.S. assets (that is, a real yield differential) would
have to be included in equation (1) and also in Figure 2.
Figure 3 shows the long-run effect of a fiscal expansion
in this case. Assume there is no risk premium initially
because portfolios are balanced, with the initial equilibrium at point a in Figure 3. The U.S. fiscal expansion
shifts the GUs and Grow schedules upward as in Figure 2.
But as the risk premium, or yield differential, grows with
the accumulation of U.S. debt by foreigners, it drives a
wedge (equal to cd) between real interest rates in the U.S.
and abroad. The new equilibrium is no longer at point b,
but rather at a lower value for the dollar. Indeed, a stable
long-run equilibrium in this case requires an increase in the
risk premium by enough to depreciate the real value of the
dollar. If the risk premium grew by only enough to leave
the real exchange rate unchanged, this still would not
be a stable equilibrium because capital inflows would be
needed to service the interest on the accumulation of U.S.
external debt, resulting in a further increase in the risk
premium. Thus, servicing the accumulated debt without
capital inflows (as required for a stable equilibrium in this
case) requires that the risk premium rise enough to cause
the real value of the dollar to depreciate, thereby generating
an increase in U.S. net exports to balance the current
account. This is illustrated in Figure 3. 10

Figure 3
Effects of U.S. Fiscal Expansion
When U.S. and Foreign Assets
Are Imperfect Substitutes
Rest-of-World
Real Interest Rate

U.S. Real
Interest Rate

r

us
1

row

r

us

0

r

row
r
1

...

Real Value of Dollar (EXCH)
... Depreciation
Appreciation

Federal Reserve Bank of San Francisco

..

0

Exchange Rate Equation in the FRBSF Model
In summary, then, movements in the exchange rate
depend, in part, on the market's expectation of the real
value of the dollar in long-run equilibrium, or the long-run
"anchor" for the dollar in equation (1). The market's
expectation of the dollar's long-run anchor, in tum, is
shaped by expectations of the impact of future domestic
and foreign fiscal policies." To the extent investors alter
their expectations of future fiscal policy in response to
current changes in fiscal policy, the dollar's anchor will be
affected and the current value of the exchange rate will
change. Under the assumption of perfect asset substitutability, or at least a constant risk premium, expectations of a
sustained U.S. fiscal expansion will cause the long-run
anchor for the dollar to rise and the dollar to appreciate.
However, if investors expect that within their investment
horizon the fiscal expansion will significantly increase the
risk premium between foreign and domestic assets, the
long-run anchor for the dollar could fall and the current
value of the dollar could tend to depreciate.
The exchange rate equation in the FRBSF macroeconometric model enables us empirically to examine these
important expectational effects. High employment, or
structural, budget balances as a percent of high-employment GNP are used as an approximate measure of the overall impact of fiscal policy. Structural budget balances are
preferable to actual (non-cyclically-adjusted) balances
because they isolate better the goods market pressures
associated with fiscal policy shifts .12
How expectations of these budget balances are formed is
an open question. The conventional approach assumes that
future budget balances are independent of current budget
balances, so that the expected real value of the dollar is a
constant determined by a condition of purchasing power
parity, possibly modified by a time trend.'> In contrast,
the approach taken here allows for the possibility that
a rational expectation of budget balances at home and
abroad over the relevant investment horizon should depend, at least in part, on current budget balances. Specifically, the effects of anticipated budget surpluses or deficits
are modeled as a function of a four-quarter moving average
of current budget balances. 14
The logarithm of the expected real value of the dollar in
the long run is thus assumed to vary with the current U.S.
budget balance (B), and a weighted average of current
foreign budget balances (B*).15 The signs of the coefficients on the budget balances depend upon the length of
the market's investment horizon and whether the market
regards U.S. and foreign assets as perfect or imperfect

31

substitutes within that investment horizon. The magnitudes of the coefficients on the budget balances depend, in
part, upon the size of the response of expected budget
balances to changes in current budget balances.

(2)
Substituting, equation (2) into equation (1) yields the
exchange rate equation to be estimated as:

InEXCH = ao + n[(i - pe) - (i* - pe*)J
+ alB + a2B*

foreign interest rates and foreign inflation are measured on
a trade-weighted basis.
The Board of Governors' index of the trade-weighted
value of the dollar is used, and the foreign interest rates and
budget balances are for the ten countries in this index.!"
The resulting equation for the real value of the dollar,
estimated over the entire floating rate period, is:
18

18

+ i"!:oaJis-i.;) + i"!:obJP-P*) (4)
.0574B + .0773B*+1.02e_ 1 - .373e~2

InEXCH = 3.44
-

(3)
18

where:

As pointed out earlier, the differential on long-term real
interest rates has a more stable impact on movements in the
real exchange rate. This real bond rate differential can be
decomposed into the nominal bond rate differential and an
expected inflation differential. The nominal bond rate
differential is modeled as a distributed lag on current and
past differentials in nominal short-term interest rates,
following the standard expectations model of the term
structure of interest rates. The expected inflation differential is similarly modeled as a distributed lag on current
and past values of the differential in quarterly inflation
rates, but with separately estimated weights to allow for
the possibility that the process of expectation formation
may differ for nominal interest rates and inflation. The sum
of the weights on the inflation differential is constrained to
be the same as the sum of the weights on the nominal
interest rate differential, but with an opposite sign. Both

I a
i=O i

18

I b,

= .104

i=O

= - .104

Individual coefficients, t statistics, summary statistics,
and exact estimation periods for this and other estimated
equations are shown in the Appendix. The real interest rate
differential is found to have a highly significant influence
upon the real exchange rate, in accordance with the conventional view of exchange markets. A sustained onepercentage point change in the real short-term interest rate
differential is estimated to produce a 10 percent change in
the real trade-weighted value of the dollar in the same
direction. The magnitude of this effect is consistent with
an average horizon for investors in the foreign exchange
market of 10 years.'?
Signs of the estimated coefficients on U.S. and foreign
budget deficits indicate that market participants view U. S.
and foreign assets as close substitutes. Their magnitudes
suggest that they view changes in structural budget deficits

Chart 1
Natural
Logarithm

Real Trade-Weighted U. S. Dollar

4.1

4.0
3.9
3.8

3.7
3.6

3.5
3.4
1973

32

1975

1977

1979

1981

1983

1985

1987

Economic Review I Summer 1989

as being relatively permanent. Thus, a one-percentage
point reduction in the current U.S. structural budget sur­
plus as a percent of high-employment GNP is estimated to
produce a six percent appreciation in the expected real
trade-weighted value of the dollar, while a similar reduc­
tion in the weighted average of foreign budget surpluses
depreciates the dollar by nearly eight percent. The dif­
ference in these coefficients is not surprising, given that the
combined GNP of the foreign countries exceeds that of the
United States. Moreover, these expectational effects are
relatively large. In the FRBSF model, it would take about a
nine percent appreciation in the real value of the dollar to
reduce the trade balance sufficiently to fully offset the
effect on aggregate demand from a one-percentage point
reduction in the U.S. budget surplus, or in other words, to
result in a full crowding out through the trade balance. The
six percent appreciation generated by the expectational
effects of a U.S. budget surplus is fully two-thirds of this.
A plot of actual and predicted values from this equation
for the whole period of floating exchange rates since 1973
(excluding serial correlation terms in predicted values) is
shown in Chart 1. The overall fit is quite good. Although
the variables in the equation do not explain the strength of
the dollar in 1985 very well, this was a period when short­
term speculative factors appear to have been particularly
important. It is of particular interest that the equation
tracks the major movements in the dollar quite well even
though it ignores the potential effects of central bank

interventions in the exchange market, except insofar as the
latter influence interest rates. This result is consistent with
the exchange rate model’s basic premise of highly substitu­
table private capital.
Chart 2 decomposes the predicted real value of the dollar
into its various components. From 1973 to 1980, the real
value of the dollar dropped by 25 percent. The effects on
expectations of the shift in the U.S. government budget
toward surplus and the shifts in foreign government bud­
gets toward deficits account for practically all of this
depreciation.18 The differential between U.S. and foreign
real long-term interest rates rose until 1975, tending to
push the real value of the dollar up during this period, but
by 1980, the differential had returned to its 1973 level,
reinforcing the tendency for the dollar to fall.
Between 1980 and 1985, the real value of the dollar
appreciated sharply by 55 percent. The effect of the U.S.
budget deficit on market expectations was the largest
contributor to this appreciation. The change in expecta­
tions arising from the growing budget deficit accounts for
about 40 percent of the total increase in the real value of the
dollar in this period. Foreign budgets generally were
moving from deficit into surplus, with the associated
expectational effects contributing about 20 percent of the
increase in the dollar’s value. Finally, a rising real interest
differential accounts for about 20 percent of the dollar’s
appreciation, with the remaining appreciation apparently
due to speculative factors.

Chart 2
C o n t r i b u t i o n s of Va r io us F a c t o r s
to Real T r a d e - W e i g h t e d U. S. Dollar
C o n trib u tio n s to
N a tu ra l L o g a rith m

Federal Reserve Bank o f San Francisco

33

During the period from 1985 to 1988 when the dollar
depreciated sharply, the combined effect of changes in the
U.S. and foreign budget deficits on the long-run expectation of the dollar contributed only 10 percent of the total
depreciation in the real value of the dollar. In contrast to the
preceding period, the declining real long-term interest rate
differential accounted for nearly 75 percent of the total
decline in the real value of the dollar. The declining real
interest rate differential, in turn, was primarily due to both
the decline in the US. real bond rate that followed monetary disinflation in the United States and a rising tradeweighted foreign real bond rate. Foreign central banks
raised their interest rates in response to the effects of the
persistently strong dollar. These movements in real interest
rates were conducive to a better domestic macroeconomic
equilibrium and were consistent with the Plaza Agreement
of September 1985, in which the Group of Five agreed to
cooperate in reducing the value of the dollar.
Other Aspects of the International Sector
of the FRBSF Model
The remainder of this section briefly discusses the
response of the trade balance and inflation to changes in
the exchange rate, as modelled in the FRBSF macroeconometric model. These equations are similar to those in most
large-scale econometric models, although this model is
relatively small in size.'? Since the model is fully documented elsewhere (Throop [1989]), only the most pertinent
aspects of the international sector are described here.

Reactions of Foreign Central Banks
In modelling the exchange rate and the trade balance,
one needs to take into account the reactions of foreign
central banks to changes in U.S. interest rates. Floating
exchange rates have diminished the short-run monetary
linkages among national real interest rates. Nonetheless,
foreign central banks continue to pursue macroeconomic
stabilization and so, continue to respond to changes in
U.S. interest rates, though to a lesser extent than before.
For example, foreign central banks tend to allow foreign
interest rates to rise in response to a rise in US. rates, to
prevent capital outflows and a depreciation of their currencies that would result in an increase in aggregate demand
and higher output and inflation. However, matching the
rise in U.S. interest rates exactly would have a deflationary
impact on foreign economies. As a result, foreign central
banks have tended to match some, but not all, of the
changes in US. real interest rates in an effort to stabilize
aggregate demand.
The estimated response function of the trade-weighted

34

foreign real short-term interest rate to changes in the U.S.
real short-term interest rate in the FRBSF model is:
Il(i:-p~*) = .2351l((,-p~)

+ .1601l(is -

+

p~)

.143Il(i,-P~)_1

(5)

2

+

+.0091l(is-P~)-3

.222e_ 1

The short-term inflationary expectations that enter into
short-term real interest rates at home and abroad. are
modeled by a four-quarter moving average of the inflation
rate. Summing the coefficients on the lagged real interest
rates suggests that foreign central banks have matched
about 55 percent of the change in U.S. real short-term
interest rates after three quarters on average.

Exports and Imports
Real exports (GEX82) are modeled as a function ofreal
GNP in the U.S.' ten major industrial trading partners
(ROWGNP82) and the real trade-weighted value of the
dollar (EXCH). The equation for exports is:

2
InGEX82 = - .811 + 1=0
.2- a_ i InROWGNP.H
9

+ 1=2
.2- b

InEXCH

i

2

where:

2- a.i, =
i=O

i

(6)

+ .774e_ 1

9

2- b i.,
i=2

1.75

= - .523

Real nonpetroleum imports (NPM82) are related in a
similar fashion to U.S. GNP and the real trade-weighted
value of the dollar.F" Real imports of petroleum (PM82)
historically have been subject to a number of special
factors, including for a time, a complex system of controls
on U.S. production. But after 1974, the ratio of petroleum
imports to GNP has been significantly and negatively
related to the real price of oil (POlL), as theory would
suggest under stable domestic supply conditions. Oil imports have not been significantly related to the real exchange rate in the expected direction, however, partly
because imports are priced in dollars, and so are not
immediately affected by changes in the value of the dollar.
Moreover, in the longer run, the response of foreign oil
suppliers to the movement in the real value of the dollar has
been quite erratic. The two import functions are:

InNPM82

= -20.1 +

2
i2-

0

a - i InGNP82

i

(7)

9

+ 1=0
.2- b_ i InEXCH - i + .797e_ 1
2

where:

i~O a:.,

9

=

3.01

2- b _ i
i=O

=

+ .384

Economic Review / Summer 1989

In(PM82/GNP82)

(8)

=

- .291

+

.897 In(PM82/GNP82)

- .137InPOIL - .251e

1

I

The overall fit of the export and import equations is quite
good, as shown in plots of actual and predicted values in
Charts 3 and 4 (excluding serial correlation terms from the
predicted values). The absolute value of the price elasticity
of demand for exports exceeds that for nonpetroleum
imports, consistent with other recent work.>' The lags on
the real exchange rate are much longer than on GNP in the
case of both exports and imports. Also, the elasticity of
U.S. nonpetroleum imports with respect to U.S. GNP is
3.01, substantially exceeding the 1.75 elasticity of U.S.
exports with respect to foreign GNP. This difference may
be due to pure income effects; or it may be capturing the
effect of different rates of productivity growth in tradable
goods at home and abroad. 22

rate of inflation is measured by a distributed lag on past
inflation. Changes in the real value of the dollar influence
prices both directly through prices of imports, and indirectly through competitive effects on domestic prices of
exports and import substitutes. The Phillips curve equation captures these relationships by including a distributed
lag on current and past changes in the real trade-weighted
value of the dollar. A second type of "supply shock" to the
price level comes from the real price of oil. Changes in the
real price of oil alter the mark-up of prices over unit labor
costs by changing the price of an important non-labor
input. A distributed lag on the percentage change in the
real price of oil captures this effect. The estimated inflation
equation is:
•

4

.

where:

In the FRBSF macroeconometric model, movements in
the real value of the dollar have a significant impact on the
price level. The inflation equation in the model may be
characterized as an expectations-augmented Phillips curve
that includes the effects of "supply shocks" from changes
in the real price of oil and the real value of the dollar. The
civilian unemployment rate (LHUR), adjusted for changes
in the natural rate of unemployment due to demographics
(U*), is used to measure excess demand, and the expected

.

U*)+.2: ai GDF - i (9)
1=2
6

.

+2:
hi POIL_ 1 +2:
ciEXCH + .388e_ 1
1
0
1=0
11

Inflation and the Dollar

11

GDF = .0847 - .600(LHUR

2: a
i=2

4

i

= 1.0

6

2: hi = .0389 i=O
2: c i =
i=O

- .0794

The sum of the estimated coefficients on past inflation is
not significantly different from one, and so it is constrained
to that value. The lag on past inflation extends about three
years. These results imply a vertical long-run Phillips
curve in which, absent supply shocks, the rate of inflation
at full employment is equal to the rate of inflation inherited
from the past. Equivalently, they reflect an accelerationist
view that excess demand, or an unemployment rate below

Chart 3
Real Exports

Billions
1982 Dollars

550
500
450
400
350
300
250
200
1975

Federal Reserve Bank of San Francisco

1977

1979

1981

1983

1985

1987

35

Chart 4
Real Imports

Billions
1982 Dollars
650
600

Predicted

550

"

500
450
400

Actual

350
300
250
200

1975

1977

1979

1981

full employment, leads to a continuing acceleration in the
inflation rate.23
A 10 percent change in the real price of oil is estimated
to change the U.S. price level by 0.4 percent in the same
direction over five quarters; and a 10percent change in the
real trade-weighted value of the dollar moves the price
level by 0.8 percent in the opposite direction over seven

1983

1985

1987

quarters. Because of the role of the dollar in the inflation
equation, a fiscal expansion in the FRBSF model causes
the real value of the dollar to appreciate, prices to drop, and
the real stock of money to expand-relieving some of the
pressure on interest rates and allowing real GNP to expand
by more than it otherwise would.

III. Effects of a Fiscal Expansion
This section compares the estimated responses of interest rates, the dollar, and the trade balance to a fiscal
expansion obtained from the FRBSF model with those
obtained from the more conventional framework in which
the market's expectation of the future real value of the
dollar is unaffected by current fiscal policy. To represent
the conventional framework, the coefficients on the budget
balances in the FRBSF model's exchange rate equation are
set equal to zero. I assume for a baseline the actual path of
the economy from 1981 through 1988. Monetary policy is
defined in terms of the actual path of nominal M2, which I
initially assume is unaffected by the fiscal expansion. A
simple fiscal change is examined, namely a permanent
increase in government spending equal to one percent of
high-employment GNP.
For simulating this fiscal expansion with the complete
FRBSF model, only two exogenous variables are changed
from their historical paths. These are the value of government spending itself and the ratio ofthe cyclically-adjusted
budget balance to high-employment GNP that appears in

36

the exchange rate equation. The results of this simulation
are compared with those from the conventional framework
which, effectively, changes only the first of these variables
from its historical path.
Table 1 shows the results of these simulations as deviations from the historical baseline path. In all of the
simulations, real government spending is increased by $32
billion in the first quarter, with this increment growing to
$38.8 billion by the 32nd quarter. Simulation A shows the
results from the conventional framework. After two quarters, real GNP rises to a maximum of $44.8 billion, but
then turns down as the lagged effects of higher interest
rates and an inventory adjustment produce a cyclical
downturn. The U.S. real bond rate rises steadily because of
persistent pressure from higher government spending on
short-term interest rates. After 32 quarters, the U.S. real
bond rate has risen 151 basis points, and the differential
between U.S. and foreign real bond rates has increased by
64 basis points. As a consequence, the real trade-weighted
value of the dollar has appreciated 6.9 percent above its

Economic Review / Summer 1989

baseline path. This produces a $16.3 billion decline in
real net exports. At this point, interest rates have nearly
peaked, and the impact of the fiscal expansion on the dollar
and the trade balance is near its maximum. Even so, only
40 percent of the crowding out occasioned by increased
government spending falls on net exports.
The response of the real exchange rate and net exports to
a fiscal expansion in the complete FRBSF model, which
includes the effects of changes in the expected long run real
value of the dollar, is quite different. As shown in Column

Federal Reserve Bank of San Francisco

B of Table 1, after four quarters, the size of the increase in
the real bond rate, and the differential between it and the
foreign real bond rate, is about the same as in Column A.
But the dollar appreciates significantly more in Simulation
B-7 .3 percent, versus 1.2 percent-because of the expectations effect. This stronger appreciation, in tum, produces larger reductions in net exports and actual declines
in domestic prices in subsequent quarters. The larger
increase in the real stock of money from lower prices
combines with the larger decrease in net exports to pull the

37

u. S. real bond rate back to the baseline path by 28
quarters. The real value of the dollar remains high, supported by expectational effects, even though the U.S. real
bond rate and the differential between it and the foreign
real bond rate return to their baseline levels.
Between four and 32 quarters in Simulation B, the real
trade-weighted value of the dollar fluctuates between 8
percent and 5Y2 percent above its baseline path. As a result
of the higher dollar, net exports fall rapidly, reaching close
to their maximum amount of decline after only 16 quarters.
Thus, in Simulation B, real net exports decline by $27.7
billion after 12 quarters and by $31.0 billion after 32
quarters, compared with $14.3 billion and $16.3 billion,
respectively, in Simulation A. By incorporating the expectational effect in Simulation B, not only does the dollar
appreciate much faster, but also after the first year there is
much less pressure on real interest rates. This outcome
puts a significantly higher proportion of the crowding out
from a fiscal expansion on net exports. In the period from
12 to 32 quarters after the fiscal expansion, 80 percent of
the crowding out associated with the increment to government spending falls on net exports in the FRBSF model,
compared with only about 40 percent in the conventional
framework.
After 32 quarters, real GNP in Simulation B is still
$10.5 billion higher than the baseline path, due to both the
interest elasticity of money demand and the increase in real
M2 produced by the dollar's appreciation. In the long run,
however, the economy will tend to return to full employ-

ment as domestic prices adjust, so that fiscal policy will
affect only the composition of output. Simulation C approximates this longer-run solution in the context of the
complete FRBSF model by raising the path of nominal
interest rates about 11 basis points above that in Simulation
B, so that after 32 quarters real GNP returns to its baseline
path.
In this long-run solution, there is an extra 31 basis point
increase in the U.S. real bond rate and an extra 1.7
percentage point appreciation in the real value of the dollar
compared with Simulation B. But the incidence of crowding out does not change significantly. About 80 percent
the crowding out from higher government spending continues tofall on net exports, with the remainder falling on
interest-sensitive consumption and investment. I estimate
that the real value of the dollar would have to appreciate by
nine percent to make crowding out fall entirely on net
exports. It actually appreciates by 7Y2 percent in this
longer-run simulation. Roughly six percentage points of
the appreciation are due to the expectational effect of
the fiscal expansion, while the remaining 1Y2 percentage
points are caused by the rise in the real bond rate differential. Thus, in the longer run, expectational effects continue to be more important than interest rate effects in
appreciating the dollar, and the stronger dollar continues to
be more important than interest rates in determining which
components of aggregate demand will bear the brunt of
crowding out.

IV. Summary and Conclusions
This paper synthesizes two major strands in the literature on open-economy macroeconomics that deal with the
linkages among fiscal policy, the dollar, and international
trade. Assuming perfect capital mobility and perfect asset
substitutability, as well as an instantaneous adjustment of
expectations, Mundell (1963) showed that a fiscal expansion can attract net capital inflows without any increase in
the differential between domestic and foreign real interest
rates. Under the same assumptions with regard to capital
mobility and asset substitutability, the conventional shortrun dynamic analysis of asset equilibrium, expounded by
Dornbusch (1976), Frankel (1979) and others, implies that
a fiscal expansion will attract net capital inflows only
insofar as it increases the differential between domestic
and foreign real interest rates. My analysis suggests that
both the interest rate differential and expectations matter.
The international sector of the FRBSF macroeconometric
model provides a synthesis by embedding a rational expectations model of the dollar's longer-run equilibrium into the
short-run dynamics of asset equilibrium. This is done by

38

including expected fiscal balances for the U.S. and other
countries along with the interest rate differential in the
exchange rate equation.
My econometric estimates suggest that expected budget
balances are significant determinants of long-run expectations of the exchange rate. These estimates also indicate
that market participants believe that an expansionary fiscal
policy will appreciate, rather than depreciate, the real value
of the dollar in the long run, suggesting that they do not
expect risk premia to be significantly affected by the
change in U.S. fiscal policy. Thus, the economy's adjustment to a fiscal expansion is similar to that originally
proposed by Mundell. Interest rates rise by less, and the
value of the dollar rises faster and farther than in conventional macroeconometric models, where the real value of
the dollar is determined solely by the differential between
U.S. and foreign real interest rates. As a result, a fiscal
expansion rapidly crowds out a relatively large amount of
net exports.

Economic Review / Summer 1989

APPENDIX
Selected FRBSF Econometric Model Equations
A. REAL EXCHANGE RATE
18

InEXCH = 3.44 + .~ aJis-i;)_;
(6.66) 1-0
- .0574B
(-3.09)

+

18
.~ bJp -p*)

!l0; -

-r

(7.50)

i -

.373e_ 2
(-2.75)

LAG

3i

bi

0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

.00849
.00829
.00806
.00781
.00754
.00724
.00692
.00658
.00621
.00582
.00541
.00497
.00450
.00402
.00351
.00298
.00242
.00184
.00123

- .00279
.00756
- .00716
- .00780
.00781
- .00775
.00761
.00741
- .00713
- .00678
- .00636
- .00587
- .00531
- .00467
- .00396
- .00318
- .00233
- .00141
-.00041

SUM

.104
(5.74)

- .104
(- 5.87)

R2

=

R.E.
D.W.

=
=

i

1-0

+ .0773B* + 1.02e
(2.06)

B. FOREIGN REAL SHORT-TERM
INTEREST RATE

+

P~*)

= .235 !l Os
(3.32)

.160 AOs - p~) -2
(2.23)

R2
S.E.
D.W.

+ .222e ~l
(1.69)

C.EXPORTS
2

InGEX82 = - .811 + ;koa_; InROWGNP82
(-1.28)

+

9

;k 2b

i

InEXCH

-r

i

+ -

.774e
(8.28)

I

LAG

o

0.591
1.071
0.084

1
2
3
4

5
6
8
9

= real trade-weighted value of the dollar

R2

U.S. short-term interest rate
foreign trade-weighted short-term interest rate
U.S. inflation rate
trade-weighted foreign inflation rate
= 4 quarter moving average of
budget
balance
= 4 quarter moving average of weighted foreign
budget balance

S.E.
D.W.

Federal Reserve Bank of San Francisco

P~L3

(0.13)

1.90

=

Sample Period: 1973.Q2 -1988.Q4

B*

+ .009 !lOs

7

us.

1

Sample Period: 1973:Q2-1987:Q4

.943
.0354
2.00

=
=
=
=

p~)

(1.99)

= .263
= .754

SUM

EXCH
i
i*
p
p*
B

+ .1430s -

p~)

=
=
=

1.75
(13.9)

- .126
- .107
.089
-.072
-.055
-.039
-.024
-.010
- .523
(-5.75)

.988
.0213
1.76

Sample Period: 1972:Q4 - 1987:Q4

= exports in billions of 1982 dollars
GEX82
ROWGNP82 = GNP in 1982 dollars of 10 major industrial trading partners
EXCH
= real trade-weighted value of dollar

39

D.NONPETROLEUMIMPORTS

F.INFLATIONa

2

lnNPM82 = - 20.1 + ~ a
(_ 15.0) i=O

+

i

lnGNP82

9
.~ b i., lnEXCH - i

1=0

GDF = .0847 - .600 (LHUR - U*)
(0.41) (-3.86)

+ - .797e_ 1

+

(8.85)

LAG

o
1

2

1.67
1.16
0.17

4

5
6
7
8
9

R2
S.E.

D. W.

3.01
(16.2)

1
2
3

.051
.047
.044
.040
.037
.033
.029
.026
.022

8

.061
.155
.148
.138
.126
.112
.095

9

.077

10
11

.056
.032

SUM

1.00

4

5
6
7

.384
(2.61)

= .994
= .0266
= 1.81

R2
S.E.
D.W.

= nonpetroleum imports in 1982 dollars
= GNP in 1982 dollars
= real trade-weighted value of dollar

In(PM82/GNP82)= - .291 + .897ln(PM82/GNP82)

-.137lnPOlL -.25Ie_ 1
( - 2.06)
( - 1. 72)

= .843
= .113
= 2.16

Sample Period: 1975. Q1 - 1987. Q4

40

6
.~ c i EXCH

1=0

+

.388e

~.m)

I

.00976
.00733
.00634
.00679
.00867

- .0057
.0097
.0124
.0138
.0140
.0129
.0106

.0389
(2.60)

-.0794
(- 2.69)

= .809
= 1.26
= 2.00

GDF

= annualized percent change in GNP fixed-

LHUR
U*

= civilian unemployment rate
= measure of variation in the civilian unem-

POlL

= annualized percent change in real price of

EXCH

crude oil
= annualized percent change in real tradeweighted value of dollar

weighted price index
I

(- 1.61) (17.2)

PM82
GNP82
POIL

+

Sample Period: 1958:Q2 - 1987:Q4

E. PETROLEUM IMPORTS

R2
S.E.
D. W.

i=O

o

Sample Period: 1972.Q4 - 1987:Q4
NPM
GNP82
EXCH

i

II
i~2ai GDF_ i

LAG

3

SUM

4
~ hi POlL

+

ployment rate due to demographics

a

The personal consumption deflator is used to deflate the
nominal stock of M2 in the FRBSF model. However, its
rate of change is a function of the rate of inflation in the
GNP fixed-weighted price index.

= petroleum imports in 1982 dollars
= GNP in 1982 dollars
= real price of crude petroleum

Economic Review / Summer 1989

NOTES
1. For an overview of these simulation results, see Helkie
and Hooper (1988) and Bryant and Holtham (1988).
2. Mundell (1963) assumed static expectations with regardto the exchange rate in the sense that the exchange
rate inthe future is expected to be the same as today's
exchange rate. As discussed below, however, a rational
adjustment of the market's long run expectat!o~ of the
dollar to changes in the current budget deficit has a
sirnilareffect on the incidence of the fiscal change. The
most forceful recent proponents of this view have.been
Dornbusch (1983) and Blanchard and Dornbus~h (1~84).
Foran earlier comparison of these two alternative Views,
see Hutchison and Throop (1985). For recent surveys that
put Mundell's contribution into historical per~pe~tiveand
further discuss some of the issues covered In this paper,
see Frankel and Razin (1987) and Marston (1985).
3. The FRBSF macroeconometric model is fully described.in Throop (1989).
4. In technical terms, previous research indicates that
risk premia on internationally-traded assets are small,
vary with time, and are difficult to associate systematically
with structural variables. See Danker, et. a/. (1984), Frankel (1982), and Hutchison and Throop (1985).
Although Mundell (1963) implicitly took perfect mobility
to require perfect substitutability, current writers generally
take perfect capital mobility to mean only an absence of
substantial transaction costs, capital controls, or other
impediments to the flow of capital between countries. This
definition of perfect capital mobility implies that t.~e exchange rate would adjust instantaneously to equilibrate
the international demand for stocks of national assets, as
opposed to the more traditional view of adjusting to equilibrate the international demand for flows of goods and
capital. But it leaves open the ques!ion whether dOf!1estic
and foreign assets are perfect or Imperfect substitutes.
See Dornbusch and Krugman (1976) and Frankel (1983).
5. For a survey of the most important mUlticou~try econometric models, see Bryant, et. a/. (1988), especially Chapters 3 and 5. Additional detail on these models may be
found in Part I of the Supplemental Volume.
6. The asset theory of exchange markets was pioneered
by Dornbusch (1976a) and Frankel (1979). See also
Hooper and Morton (1982) and Hutchison (1982) for applications of the asset view. A useful general survey of
modern exchange rate theory is Shafer and Loopesko
(1983).
7. The open interest parity condition in nominal terms is:

Ins - Ins e

= n(i

- i*)

where s is the nominal value of the dollar, defined as units
of foreign currency per unit of domestic currency, and se is
the expected value of the nominal exchange rate. By
definition

s

=

EXCH

p*

P

Federal Reserve Bank of San Francisco

and

se

= EXCHe

p* (1 + pe*)n
p(1 + peyn

where EXCHis the real exchange rate and p and p* are
the U.S. and foreign price levels, respectively. Taking
logarithms and substituting into the arbitrage equation in
nominal terms gives equation (1) in the text.
In theory, the relevant interest rate differential sh?uldbe
after taxes. Although marginal tax rates on real Interest
income differ among industrial countries, no estimates of
these rates are available. For a survey of and some
background papers on what is known about how interest
income and foreign exchange gains and losses are taxed
in various countries, see Tanzi (1984).
8. The ultimate equilibrium at LM3 and IS3 is similar to that
in Mundell's (1963) classic analysis, in which the dollar
appreciates without any increase in the equil.ibrium real
interest rate differential. Mundell assumed static expectations with respect to the exchange rate (meaning that the
exchange rate expected in the ~uture is the s~r:ne. as
today's exchange rate), allowing this short-run equilibrium
to be reached immediately. Also, he ignored the effect of
the currency appreciation on the LM schedule, so !hat the
IS schedule shifted all the way from IS2 to 15 1, leading to a
full crowding out of net exports by the fiscal expansion.
9. Because real interest rates equalize in the long run, the
dollar appreciates without any increase in the equilibriu~
real interest rate differential, just as in Mundell's claSSIC
analysis of a small country with fixed prices. T~e difference in this two-country, full-employment case IS that
because the world interest rate rises, there is some crowding out of U.S. domestic investment, and possibly consumption, in addition to net exports. Howev~r, the smaller
is the country with the fiscal expansion relative to the rest
of theworld, the greater is the crowding out of net exports.
Crowding out from fiscal expansion in a country small
enough to have no significant impac~ on w?rld Intere~t
rates would fall entirely on net exports, Just as In Mundell s
small country case with fixed prices.
A further point is that the crowding out of world-wide
capital formation by a U.S. fiscal expan.s.io~ gr~dually
shifts up the locus of full-employment eqUlllbnu~. In both
countries as capital becomes scarcer, thus raiSing real
interest rates at full employment in both countries. Since
this shifts up the schedules of both countries, there is no
necessary impact on the real value of the dollar. Howev~r,
if consumption spending is a function of net w~alth, as IS
commonly believed, the increase in the r~latlve .wealth
pcsition of the foreign country would shift up It? fullemployment equilibrium relatively more, th~s tending to
depreciate the dollar. To the extent that I.nvestors. e~­
pected the fiscal expansion to have such an Impact Within
their investment horizon, the current value of the dollar
could be affected. Whether this is in fact the case is an
empirical matter.

41

10. Branson (forthcoming) and Sachs (1985) have constructed formal models in which the risk premium in an
open-interest parity condition varies over time in proportion to relative debt positions. Krugman (1985, 1988) correctly points out, however, that risk premiums should also
enter into the expected long-run value of the dollar, consistent with Figure 3. But he suggests in addition that the
market has not correctly assessed the limit to absorption
of dollar-denominated assets by the rest of the world. The
implication that expectations in the foreign exchange
market are irrational, is hard to accept.
Rather, it is more realistic to assume there is a relatively
large potential world demand for dollar assets. The dollar
is universally accepted as a means of international payment and serves as an international store of value to
an extent unmatched by any other asset. Moreover, the
breadth, depth, and resilience of U.S. financial markets
provide a degree of liquidity not available in other assets.
As a result, only a small increase in the U.S. real interest
rate relative to the foreign real interest rate likely would be
required to ensure continued external financing of the
U.S. payments deficit. As a consequence, over relatively
long time horizons, the expectation of a relatively permanent U.S. budget deficit is more likely to lead to an
increase in the expected long-run equilibrium in the real
value of the dollar than a decrease, consistent with the
empirical results discussed below. For a further defense
of this view, see Cheng (1988).
11. The analysis shown in Figures 2 and 3 makes it clear
that the expected real value of the dollar also should
depend on the expected rate of private saving at home
and abroad. Although the U,S, private saving rate declined significantly in the 1980s, prior to that it had been
stable over a long period of time, (See Denison [1958] and
David and Scad ding [1974],) The question whether expectations of long-run private saving rates have changed
significantly is beyond the scope of this article,
12, They are only approximate because the marginal
effects of government spending, transfer payments, and
various taxes on aggregate demand are not exactly the
same, The budgetary data are combined federal, state,
and local balances compiled by the OECD, Sources of
these data are Price and Muller (1984) and recent issues
of the OEeD Economic Outlook,
Both inflation-adjusted and unadjusted structural budget balances were tried, Unadjusted structural budget
balances count the inflation premium in interest paid on
government debt as an outlay, but do not count the
corresponding erosion in the real value of this debt due to
inflation as a receipt. The inflation-adjusted structural
budget balance corrects this by including the erosion in
the value of the debt as tax revenue, These two measures
performed equally well in the exchange rate equation,
But evidence from the consumption function in the
FRBSF econometric model, as well as empirical work by
Eisner and Peiper (1984) and Price and Muller (1984) that
shows real growth in the United States and Europe to be
more closely related to movements in inflation-adjusted

budget balances than to unadjusted ones, supports using
the adjusted measure, The inflation-adjusted measure
is consistent with households' behaving rationally and
therefore saving (and reinvesting) inflation premiums in
the interest on government debt.
Because of this behavior, the private saving rate as
conventionally measured should tend to rise and fall with
theinflation rate. This response of the private saving rate
to inflation is particularly evident in European countries
that-have experienced sharp changes in inflation, However, it is obscured in the U.S, data by simultaneous
movements in the ratio of real wealth to income, which also
influence the saving rate in a life-cycle model of consumption, For further discussion of the inflation-adjusted measure, seeJump (1980), Siegel (1979), and Tanzi, Blejer, and
Teijero (1987),
13, See, for example, Bryant, et. al. (1988), Bryant and
Holtham (1988), and Helkie and Hooper (1988),
14, The alternative of budget balances over four quarters
ahead did not perform as well. Neither did distributed lags
on current and past budget balances,
15, Trade-weights clearly are appropriate for combining
the rest of the world's real interest rates since that is the
way the exchange rate is constructed, However, in the
case of the structural budgets, the relative size of the
country also is important. The larger the country, the
smaller trade generally will be as a proportion of GNP, and
the flatter will be its full employment locus in Figures 2
and 3, Therefore, the impact of a one-percentage point
change in the country's structural budget on its real
bilateral exchange rate with the U,S, would be greater the
larger is the size of that country's economy, Thus, the
weight for the foreign budget balances that I used is the
trade-weight times the relative GNP-weight.
Since the relative effects of domestic and foreign budget balances on the real exchange rate depend upon the
relative size ofthe U.S. and the rest of the world, there is no
reason that the coefficients on the two budget balances
should be constrained to be of equal absolute value, as is
the case with U.S. and foreign interest rates.
16. Multilateral trade weights are used. See Board of
Governors of the Federal Reserve System (1978). The
nominal index is deflated by the ratio of trade-weighted
foreign consumer prices to the U.S. GNP fixed-weight
price index.
17. Hooper (1985,1987) estimates a six percent change
in the real exchange rate for a one-percentage point
change in the real interest rate differential. He uses interest rates on securities with maturities that are usually 10
years, but sometimes less.
18. Although the standard measure of the U.S. federal
fiscal deficit as a percent of high-employment GNP rose
and the state and local government balance was about
unchanged, there was a larger increase in the inflation "tax" on government debt. Hence, the U.S. inflationadjusted structural budget balance rose, See footnote 12.
19, The FRBSF model has only 28 behavioral equations,

Economic Review / Summer 1989

compared with 124 in the Federal Reserve-MIT-Penn
model (Brayton and Mauskopf, 1987), for example.
20. Weighted averages of domestic spending and domestic output also were tried as scale variables, on the
theory that imports depend upon spending as well as
production, but they gave inferior results compared with
real GNP.
21. See Feldman (1982) and Warner and Kreinin (1983).
22. The measured difference in income elasticities would
imply a need for the real value of the dollar to decline
secularly unless there is an offsetting difference in growth
rates of income at home and abroad. A classic study on
income elasticities in world trade, originally pointing out

the need for a secular decline in the real value of the dollar,
is Houthakker and Magee (1969). Subsequent literature
on income elasticities is surveyed in Goldstein and Kahn
(1985). A recent discussion of the effect of income elasticities and productivity growth on the trend in the real
value ot.the dollar is provided in Krygman and Baldwin
(1987). A negative time trend to account for the possible
effectofthe difference in elasticities initially was included
in the equation for the exchange rate (equation (4)), but it
prpved to be statistically insignificant.
23. TheJul1 employment rate of unemployment, at which
inflation tends neither to accelerate nor decelerate, is
estimated at 5% percent in the U.S. economy at present.

REFERENCES
Blanchard, Oliver and Rudiger Dornbusch. "U.S. Deficits,
the Dollar, and Europe," Banca Nazionale del Lavoro,
March 1984.
Board of Governors of the Federal Reserve System.
"Index of the Weighted-Average Exchange Value of
the U.S. Dollar: Revision," Federal Reserve Bulletin,
August 1978.
Branson, William H. and Grazia Marchese. "Internationallmbalances in Japan, Germany, and the United
States," forthcoming, International Payments Imbalances, Federal Reserve Bank of Boston Conference
Series.
Brayton, Flint and Eileen Mauskopf. The MPS Modelof the
U.S. Economy. Board of Governors of the Federal
Reserve System, 1985.
Bryant, Ralph C., Dale W. Henderson, Gerald Holtham,
Peter Hooper, Steven Symansky, eds. Empirical Macroeconomics for Interdependent Economies. Washington, D.C.: The Brookings Institution, 1988.
Bryant, Ralph C. and Gerald Holtham. "The U.S. External
Deficit: Diagnosis, Prognosis, and Cure," in Ralph
Bryant, Gerald Holtham, and Peter Hooper, eds.,

External Deficits and the Dollar: The Pit and the
Pendulum. Washington, D.C.: Brookings Institution,
1988.
Cheng, Hang-Sheng, "Must the Dollar Fall?" Weekly Letter, Federal Reserve Bank of San Francisco, October
7,1988.
Danker, Deborah 1., Richard A. Haas, Dale W. Henderson,
Steven A. Symansky, and Ralph W. Tryon. "Small
Empirical Models of Exchange Market Intervention:
Applications to Germany, Japan, and Canada" Staff
Economic Studies 135. Washington: Board of Governors of the Federal Reserve System, 1984.
David, Paul A. and John L. Scadding. "Private Savings:
Ultrarationality, Aggregation, and Denison's Law,"
Journal of Political Economy, MarchiApril 1974.
Denison, Edward F. "A Note on Private Saving," Review of
Economics and Statistics, August 1958.

FederalReserve Bank of San. Francisco

Dornbusch, Rudiger. "Expectations and Exchange Rate
Dynamics," Journal of Political Economy, December
1976a.
_ _ _ _ . "Comments" on Jeffrey R. Shafer and Bonnie
E. Loopesko, "Floating Exchange Rates After Ten
Years," Brookings Papers on Economic Activity, No.1,
1983.
_ _ _ _ and Paul R. Krugman. "Flexible Exchange
Rates in the Short Run," Brookings Papers on Economic Activity, No.3, 1976b.
Eisner, Robert. HowRealIs theFederalDeficit? New York:
The Free Press, 1986.
_ _ _ _ and Paul J. Peiper. "A New View of the Federal
Debt and Budget Deficits," American Economic Review, March 1984.
Feldman, Robert A. "Dollar Appreciation, Foreign Trade,
and the U.S. Economy," Quarterly Review, Federal
Reserve Bank of New York, Summer 1982.
Fleming, Marcus. "Domestic Financial Policies Under
Fixed and Floating Exchange Rates," I.M.F. Staff
Papers, November 1962.
Frankel, Jeffrey. "On the Mark: A Theory of Floating
Exchange Rates Based on Real Interest Differentials,"
American Economic Review, September 1979.
_ _ _ _ . "A Test of Perfect Substitutability in the Foreign Exchange Market," Southern Economic Journal,
October 1982.
_ _ _ _ . "Monetary and Portfolio-Balance Models of
Exchange Rate Determination," in Jagdeep S. Bhandari and Bluford H. Putnam, eds., Economic Interdependence and Flexible Exchange Rates, Cambridge:
The MIT Press, 1983.
Frenkel, Jacob and Assaf Razin. "The Mundell-Fleming
Model A Quarter Century Later: A Unified Exposition,"
IMF Staff Papers, December 1987.
Goldstein, Morris and Mohsin S. Kahn. "Income and Price
Effects in Foreign Trade," in Ronald W. Jones and
Peter B. Kenen, eds., Handbook of International
Economics, Vol. 2. Amsterdam: North-Holland, 1985.

43

Helkie, William L. and Peter Hooper. "An Empirical Analysis of the External Deficit, 1980-86," in Ralph Bryant,
Gerald Holtham, and Peter Hooper eds., External
Deficits and the Dollar: The Pit and the Pendulum.
Washington, D.C.: Brookings Institution, 1988.
Hooper, Peter. "International Repercussions of the U.S.
Budget Deficit," Aussenwirtschaft, January 1985.
_ _ _ _ and Catherine L. Mann. "The U.S. External
Deficit: Its Causes and Persistence," Conference
on U.S. Trade Deficit-Causes, Consequences and
Cures, Federal Reserve Bank of St. Louis, October
1987.
Hooper, Peter and John Morton. "Fluctuation in the Dolla~:
A Model of Nominal and Real Exchange Rate Deterrnination," Journal of International Money and Finance,
1982, 1.
Houthakker, H.S. and Stephen P. Magee. "Income and
Price Elasticities in World Trade," Review of Economics and Statistics, May 1969.
Hutchison, Michael M. "Intervention, Deficit Finance and
Real Exchange Rates: The Case of Japan," Economic
Review, Federal Reserve Bank of San Francisco,
Winter 1984.
_ _ _ _ and Adrian W. Throop. "U.S. Budget Deficits
and the Real Value of the Dollar," Economic Review,
Federal Reserve Bank of San Francisco, Fall 1985.
Jump, Gregory V. "Interest Rates, Inflation Expectations
and Spurious Elements in Measured Real Income,"
American Economic Review, December 1980.
Krugman, Paul R. "Is the Strong Dollar Sustainable?," in
The U.S. Dollar-Recent Developments, Outlook, and
Policy Options. A Symposium Sponsored by the Federal Reserve Bank of Kansas City, August 21-23,
1985.
_ _ _ _ . "Sustainability and the Decline of the Dollar,"
in Ralph Bryant, Gerald Holtham, and Peter Hooper,
eds., External Deficits and the Dollar: The Pit and the

44

Pendulum, Washington D.C.: Brookings Institution,
1988.
_ ___..-__ and Richard E. Baldwin. "The Persistence of
the U.S. Trade Deficit," Brookings Papers on Economic Activity, No.1, 1987.
Marston, Richard C. "Stabilization Policies in Open Economies," in Ronald W. Jones and Peter B. Kenen, eds.,
Handbook of International Economics, Vol. II. Amsterdam: Elsevier Science Publishers B.V., 1985.
Mundell, Robert. "Capital Mobility and Stabilization Policy
Under Fixed and Flexible Exchange Rates," Canadian Journal of Economics and Political Science,
November 1963.
Price, Robert W. R. and Patrice Muller. "Structural Budget Indicators and the Interpretation of Fiscal Policy
Stance in OECD Economies," OECD Economic StudieS,Autumn 1984.
Sachs, Jeffrey D. 'The Dollar and the Policy Mix: 1985,"
Brookings Papers on Economic Activity, 1985, NO.1.
Shafer, Jeffrey R. and Bonnie E. Loopesko. "Floating
Exchange Rates After Ten Years," Brookings Papers
on Economic Activity, No.1, 1983.
Siegel, Jeremy J. "Inflation-Induced Distortions in Government and Private Saving Statistics," Review of
Economics and Statistics, February 1979.
Tanzi, Vito, ed. Taxation, Inflation, and Interest Rates.
Washington: International Monetary Fund, 1984.
_ _ _ _ , Mario Blejer, and Mario O. Teijero. "Inflation
and the Measurement of Fiscal Deficits," IMF Staff
Papers, December 1987.
Throop, Adrian W. "A Macroeconometric Model of the
U.S. Economy," Working Papers in Applied Economic
Theory, No. 89-01, Federal Reserve Bank of San
Francisco, March 1989 (Revised).
Warner, Dennis and Mordechai E. Kreinin. "Determinants
of International Trade Flows," The Review of Economics and Statistics, February 1983.

Economic Review I Summer 1989