The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Economic
Review
FEDERAL RESERVE BANK OF ATLANTA
MAY 1986
JOB GROWTH Tracing Employmen^^ains
m
u
w
o f p m l a d e u w a
SPENDING CUTS Balancing the Budget
DEFICITS Is Accommodation Inevitable?
President
Robert P. Forrestal
Sr. Vice President and
Director of Research
Sheila L Tschinkel
Vice President and
Associate Director of Research
B. Frank King
Financial Institutions a n d Payments
David D. Whitehead, R e s e a r c h Officer
Larry D. W a l l
Robert E Goudreau
Macropolicy
Robert E Keleher, R e s e a r c h Officer
Thomas J. Cunningham
Mary S. R o s e n b a u m
Jeffrey A R o s e n s w e i g
J o s e p h A. Whitt, Jr.
Pamela V. Whigham
Regional Economics
G e n e D. Sullivan, R e s e a r c h Officer
Charlie Carter
William J . Kahley
J o e l R. Parker
W. G e n e Wilson
Publications and Information D e p a r t m e n t
Bobbie H. McCrackin
Public I n f o r m a t i o n
D u a n e Kline, Director
Linda Donaldson
Editorial
Harriette Grissom, Publications Coordinator
Melinda Dingler Mitchell
Ann L P e g g
Graphics
Eddie W. Lee, Jr.
Typesetting, W o r d Processing
Cheryl B Birthrong
Beverly Newton
Belinda W o m b l e
Distribution
George Briggs
Vivian Wilkins
Ellen Gerber
The Economic Review seeks to inform the public
about Federal Reserve policies and the economic
environment and, in particular, to narrow the gap
between specialists and concerned laymen Views
expressed in the Economic Review are not necessarily those of this Bank or the Federal Reserve
System Material may be reprinted or abstracted if the
Review and author are credited. Please provide the
Bank's Research Department with a copy of any
publication containing reprinted material. Free subscriptions and additional copies are available from
the Information Center, Federal Reserve Bank of
Atlanta, 104 Marietta Street, N.W., Atlanta, Ga 303032713 (404/521-8788). Also contact the Information
Center to receive Southeastern Economic Insight
a free newsletter on economic trends published by
the Atlanta Fed twice a month. The Review is indexed
online in the following data-bases: ABI/lnform, Magazine Index Management Contents, PAIS, and the
Predicasts group
ISSN 0732-1813
2
VOLUME LXXI, N O . 5, MAY 1986, E C O N O M I C REVIEW
Whaf s Behind Patterns of State Job Growth?
William J. Kahley
Lower wages, inexpensive energy, and market size help to
determine where both foreign and domestic firms will put
people to work, according to this statistical study.
Projecting Federal Deficits and the Impact of the
Gramm-Rudman-Hollings Budget Cuts
I i
.J
Thomas J. Cunningham and
Rosemary Thomas Cunningham
The authors explain how the spending cuts mandated by
this legislation work to achieve a balanced budget by the
required 1991.
19
The Long-Run Outcome of the Permanent Deficit
Thomas J. Cunningham
By examining debt-to-income ratios over the long term, the
author shows that monetary accommodation may not be
an inevitable outcome of government deficits—even large
ones.
25
Statistical Summary
Finance, Construction, General, Employment
35
FEDERAL RESERVE BANK OF ATLANTA
3
What's Behind Patterns of State Job Growth?
Which state characteristics contribute to employment
gains? This analysis suggests that both foreign and domestic
firms are influenced by basic business considerations such as
labor and energy costs
William J. Kahley
The quickening movement of jobs and people
in recent decades from the northeastern and
north-central Frostbelt of the United States to
warmer southern and western locations has
stirred widespread debate over the causes and
consequences of the shifts. Observers generally
agree that locational disparities in costs and
benefits account for differences in regional
growth rates; they see the Sunbelt phenomenon
as the result of several primary forces and many
contributing ones that favor the South. The
The author
is a regional
economist
for the Atlanta
Research
Department
Brent
Croce,
a Research
ment intern, provided
valuable
research
assistance.
4
Fed's
Depart-
precise way these forces are linked and the
impact of the various determinants remains
uncertain, though, because of several conceptual and practical difficulties in disentangling
their separate effects. Understanding these
determinants and how they influence a family's
or firm's choice about location is important in
two ways: first, as a theoretical investigation, it
contributes to our knowledge about economic
decisionmaking; and second, at a more practical level, this understanding can inform economic development planning. Gaining further
insight into what lures people and business can
help communities take action to enhance their
attractiveness.
Recently, media and scholarly attention has
turned to the issue of why foreign investors
MAY 1986, E C O N O M I C REVIEW
have favored certain U.S. locations over others.
The South and West appear to have been
particularly successful at enticing domestic
and foreign capital, according to available data,
although some evidence suggests that the
Northeast has been staging a relatively strong
rebound in this period of economic expansion.
It may not be obvious why a state or area would
want to attract foreign investment and analysts
concerned with the impact of this type of
investment have debated its desirability.1 This
article will not argue that point but simply
assume foreign direct investment (FDI) benefits
a local economy because it helps to create
jobs.
Explaining a foreign investor's choice appears
even more difficult than explaining shifts within
the United States, perhaps because the often
subjective decisionmaking process involves
assigning weights to influences that vary in
importance with the investor's nationality, corporate culture, experience, and concerns. Moreover, some evidence suggests that site selection
criteria and foreigners' investment strategies
can shift dramatically over time. David McClain
has even argued that, "the multinational character of firms can emerge from many different
sets of circumstances that may have little else
in common," so that "approaching an analysis
FEDERAL RESERVE BANK OF ATLANTA
of their behavior by exercises in pure theory or
econometrics may not be a promising research
strategy."2
Theoretical complexities and measurement
difficulties notwithstanding, this article will
present an analysis of the differences in factors
that determine both overall and foreign affiliate
employment shifts. The results of our analysis
suggest that basic business considerations, such
as the varying costs of producing goods and
moving them to markets, as well as business
environment and lifestyle conditions, explain
significant variations in job growth from state to
state due to foreign and domestic investment
activity. Moreover, theoretical economic arguments appear to explain foreign investors'
decisionmaking better than they explain internal job shifts. This rather interesting result will
be discussed in some detail later in this article.
Theoretical Model
Industrial location theory traditionally has
been concerned with the optimal siting of a
plant, often with respect to transportation considerations, based on regional differences in
potential profitability. Profit potential is determined by variations in supply and demand
factors among regions. Costs of production,
5
based on supplies of labor, capital, land, and
other materials, can vary from place to place
due to differences in the relative endowments
of these resources. These price differences
mean that production costs and hence profitability will diverge among regions. Japanese
and domestic auto makers, for example, are
said to have located in Tennessee largely because of the relatively lower labor and transportation costs. The latter are attributable to
In a dynamic
economy,
revenues and costs change as
resources are discovered or
depletedI as new
technologies
are introduced or as consumers'
preferences
develop.
These regional shifts over time can be viewed
as the cumulative result of decisions by producers to expand or contract local area production in response to existing or emerging
differences in profitability. A major historical
shift in the post-World War II period benefited
the southeastern states through the "migration"
of textile and apparel firms to the region from
New England. This movement was sparked
mainly by the lower cost and abundant supply
of labor in the Southeast
Relative rates of employment growth among
states or regions, then, can be traced to the
producers' response to both existing differences and changes in comparative advantage.4
Varying wage rates across regions, for example,
would constitute an existing disparity in profitability among areas, while unequal changes in
wage rates would cause comparative advantage
to shift In either instance, a firm might be
motivated to choose a new location.5
The Shift-Share Technique
the state's central location vis-a-vis parts suppliers and the availability of favorable interstate
highway arteries. The location of foreign chemical manufacturers in Louisiana where abundant
energy supplies and excellent port facilities
create a lower cost environment for these
producers, represents comparative regional
advantage due to supply conditions. Revenue
projections, which reflect demand, will also
differ among areas depending on the size of
the potential local market W h e n these projections are combined with estimates of local
production costs, a projection of profitability
results. Realistically, however, firms first survey
markets broadly across regions or states and
then select a specific site that will maximize
profits. The research reported here, however,
does not deal with specific decisions about
plant sites but with how location choices by
broadly defined industries affect state employment 3 The analysis and results should, however,
be consistent with individual firm or plant
locations.
In a dynamic economy, revenues and costs
change as resources are discovered or depleted,
as new technologies are introduced or as consumers' preferences develop. Within a country,
these revenue and cost changes will cause the
regional location of economic activity to shift
6
The shift-share technique, originated by
Daniel Creamer (1942), is used to isolate and
identify precisely the sources of local employment change vis-a-vis relative changes nationally (see Box). For each area shift-share separates
actual employment change over a period of
time into three components, measuring how
each affects overall local employment fluctuation. The first factor is the reference area component which measures local employment
change that would have occurred had the local
area gained its proportionate share of overall
national economic growth. The second factor,
the industrial mix component considers a
local area's distribution of fast- or slow-growing
industries relative to the national industrial
growth mixture. The remaining element called
the local share effect accounts for changes in a
locality's competitive position in a particular
industry. A positive number for the local share
effect means an area's comparative advantage
in that industry has improved in relation to
other areas. If Florida's services sector employment growth, for instance, exceeds that of the
nation as a whole for a given period, then
Florida will have a positive local share effect for
that period, because it has improved compared
with other regions in providing employment in
that sector. In this instance, the above-average
MAY 1986, E C O N O M I C REVIEW
growth may be the result of a favorable weather
climate that attracts retirees and tourists who
spend large fractions of their incomes on services.
Box
Mathematical Statement of
Shift-Share Components
Following John R. Gordon and others (1980), the
shift-share components can be stated mathematically for industry i in region j as follows:
where,
E = employment
R = reference area effect
M = industrial mix effect
L = local share effect
E
t
= total employment in the reference area
(such as the United States as a whole)
during the terminal period,
E q 0 = total employment in the reference area
during the base period,
Ejj = employment in industry i in area j (such as
a particular state) during the terminal
period,
Em = employment in industry i in area j during
the base period,
e| q = employment in industry i in the reference
area during the terminal period, and
e£> = employment in industry i in the reference
area during the base period
(1) A Ejj = (Ejj - eJj) = R + M + L
A cr t
' o o
. Eb
^oo
(2) R =
l
(3) M
E oo
^
j
"
/V
Eb\
S o -CIO
E io
V
(
E
y
LE-b\
LU
-0:=r
(4) L =
Ì
FEDERAL RESERVE BANK OF ATLANTA
f cF*
00
V
f
V
E
Et
b V
- cE00
°°
.
Eb
E io
y
V
)
Application of the Analysis
Using the general shift-share framework to
explain specific divergence in state employ
ment due to comparative advantage factors
requires two empirical procedures. First shiftshare analysis measures the components of the
total change in state employment described
above. The positive or negative amount of
employment that is due to a state's comparative
advantage position (that is, the amount of U
see Mathematical Statement of Shift-Share
Components) is calculated at this stage of the
analysis. The question is what determines employment change due to comparative advantage? This employment figure, the amount of L
is used as the dependent variable in a regression
model that includes comparative advantage
factors as independent variables expected to
explain the employment change.6
The dependent variables in this study refer
to (1) manufacturing, non manufacturing, and
total employment change in multi-establishment foreign-owned firms between 1975 and
1982, and (2) manufacturing, non manufacturing, and total employment change in all U.S.
establishments in those same years.7 The years
1975 and 1982 were chosen because they
represent comparable positions in the business
cycle, and 1982 is the most recent year for
which relevant data are available. Table 1
shows the original employment data and calculated local share effects. States are ranked in
the table according to how much comparative
advantage contributed to employment growth
between 1975 and 1982.
Table 1 and the accompanying maps confirm
the accuracy of media news headlines that for
years have trumpeted the shift of job and
population growth toward the Sunbelt and
away from Frostbelt states. Even a cursory
glance at the table shows the relative gains to
warm-weather southern and western states
and losses to colder states, particularly in the
industrial heartland. This shift is reflected in
employment growth attributed to both foreign-owned and U.S. firms. A close statistical
association or movement between the two
7
Table 1
Comparative Advantage Employment
Part 1. Total Establishment (Domestic and Affiliate) Employment
(Thousands)
Local S h a r e Effect
Total, Manufacturing, and Nonmanufacturing
1975
State
Texas
California
Florida
Louisiana
Colorado
Oklahoma
Arizona
Georgia
Washington
Nevada
Virginia
N e w Hampshire
Wyoming
New Mexico
South Carolina
Utah
Mississippi
North Dakota
Vermont
Alaska
North Carolina
Kansas
Idaho
Hawaii
Minnesota
Oregon
Arkansas
Montana
Maine
Maryland
Alabama
Delaware
South Dakota
R hod e Island
Connecticut
Nebraska
Tennessee
W e s t Virginia
Kentucky
Wisconsin
Missouri
Massachusetts
New Jersey
Iowa
Indiana
Michigan
Ohio
Pennsylvania
Illinois
New York
Total
1161.8
695.1
505.0
210.3
209.5
204.5
182.2
165.1
150.3
105.7
71.6
58.7
53.5
47.1
46.9
36.7
19.6
18.9
15.6
14.3
14.1
4.6
3.2
2.4
0.6
0.2
0.0
-5.8
-7.9
-9.9
-10.9
-11.0
-12.6
-13.7
-17.0
-42.2
-42.6
-48.8
-62.2
-74.0
-95.9
-97.7
-106.1
-122.1
-227.6
-427.0
-554.1
-589.0
-594.4
-824.9
United States
—
MFG
263.7
286.5
119.3
19.0
45.5
34.9
52.3
60.6
36.3
6.7
18.9
26.3
1.1
5.2
30.7
15.4
3.8
-1.4
7.6
1.7
25.3
5.4
-0.6
-2.4
20.5
-0.5
18.4
-3.3
9.0
-30.3
15.5
-1.0
5.0
4.2
4.1
-0.7
5.2
-26.8
-16.4
-25.6
-0.5
20.3
-53.1
-35.3
-78.2
-101.3
-203.1
-212.4
-250.4
-124.8
858.3
364.4
332.5
174.8
149.8
160.0
118.1
105.7
106.8
91.4
43.9
34.9
48.4
32.7
30.6
15.4
20.9
15.6
8.2
8.3
27.1
-4.9
1.1
-4.4
-25.3
-2.8
-14.8
-7.9
-15.4
4.2
-20.4
-8.1
-22.2
-13.8
-4.2
-48.5
-33.5
-23.8
-45.0
-31.2
-97.7
-110.8
-36.1
-87.3
-120.4
-293.2
-302.6
-330.8
-315.9
-732.2
—
S o u r c e : B u r e a u of the Census, County
8
NONMFG
—
Business
Patterns,
1982
Total
MFG
NONMFG
Total
MFG
NONMFG
4367.8
7748.0
2797.3
1198.3
932.3
875.6
723.6
1705.8
1197.4
252.0
1726.2
790.3
1570.3
329.5
161.5
6332.4
9867.2
3816.4
1105.0
1958.2
470.1
212.2
186.0
192.3
156.4
503.5
292.0
19.4
398.1
113.7
8.8
33.6
369.4
87.2
206.3
14.9
49.2
9.2
5227.4
7909.0
3346.3
1416.6
1127.2
1048.7
882.4
1680.9
1275.8
384.6
1717.0
273.3
204.4
437.3
798.9
470.4
591.5
229.9
152.1
6777.0
19.0
106.7
396.9
85.5
443.0
121.5
251.2
499.3
389.5
590.8
743.0
230.1
638.5
940.8
1254.6
1336.6
1229.0
1415.7
3577.5
6177.7
2467.8
1016.8
800.3
727.7
625.8
1289.8
957.2
240.1
1370.0
195.2
127.7
331.3
629.1
372.6
467.2
175.5
117.8
136.1
1253.9
622.6
213.1
316.3
1125.3
637.3
439.5
209.1
251.7
1174.7
813.9
159.4
183.0
231.6
814.6
459.8
1027.2
436.6
776.7
1131.9
1304.5
1698.8
1912.3
742.9
1259.6
2126.9
2701.5
3029.3
3163.2
5361.3
75410.3
18026.6
57381.7
277.3
134.9
358.0
947.3
440.0
657.4
190.8
156.9
143.1
1968.4
783.8
257.2
339.0
1434.4
808.1
607.2
229.6
343.5
1407.3
1119.6
225.8
202.0
338.3
1211.5
545.3
1470.2
558.1
1027.9
1631.2
1694.0
2292.6
2655.3
973.0
1898.1
3067.7
3956.1
4365.9
4392.2
132.0
147.9
97.8
416.0
240.2
11.9
356.2
82.1
7.2
26.7
318.2
67.4
190.2
15.3
39.1
7.0
714.5
161.2
44.1
22.7
309.1
170.8
167.7
20.5
91.8
232.6
305.7
66.4
1628.8
1313.2
1241.0
1038.8
2184.4
1567.8
404.0
2115.1
387.0
213.2
470.9
1168.3
557.6
797.8
244.8
201.3
183.7
2344.3
932.5
307.7
403.7
1698.6
956.8
718.8
266.0
398.7
174.5
1558.4
755.5
261.4
381.9
1349.0
775.5
521.9
247.5
292.0
1438.8
973.6
186.6
201.3
269.0
990.6
513.0
1221.0
509.4
903.6
1351.1
1495.4
1963.9
2299.3
1656.1
1314.5
256.3
226.5
386.8
1417.2
603.3
1697.8
611.9
1154.6
1857.0
1909.5
2616.3
3037.2
1029.7
2019.4
3204.5
4129.1
4579.3
4605.1
7197.7
785.9
177.0
46.3
21.8
349.6
181.3
196.9
18.5
106.7
217.3
340.9
69.7
25.2
117.8
426.6
90.3
476.8
102.5
251.0
505.9
414.1
652.4
737.9
209.7
601.5
900.2
1132.5
1210.5
1057.9
1382.3
820.0
1417.9
2304.3
2996.6
3368.8
3547.2
5815.4
89270.6
19192.5
70078.1
1975 and 1982.
MAY 1986, E C O N O M I C REVIEW
Table 1 (continued)
Comparative Advantage Employment
Part 2. Foreign-Owned Multi-Establishment Employment
Local S h a r e Effect
Total, Manufacturing, and Nonmanufacturing
1982
1975
State
Texas
California
Florida
Georgia
Tennessee
South Carolina
North Carolina
Arizona
Oklahoma
Maine
Utah
Virginia
Missouri
Washington
Minnesota
Louisiana
R hod e Island
Kansas
Mississippi
New Mexico
Colorado
Nevada
Alaska
Wyoming
Oregon
Alabama
South Dakota
W e s t Virginia
Delaware
North Dakota
Montana
Nebraska
Iowa
Vermont
Idaho
N e w Hampshire
Indiana
Maryland
Hawaii
Massachusetts
Arkansas
Connecticut
Kentucky
Michigan
Wisconsin
Illinois
Pennsylvania
Ohio
New Jersey
New York
51897
26278
20065
19276
17260
16457
13002
12060
11883
11172
10902
8916
8587
6769
6112
5462
5367
3450
3160
3088
2483
2324
1475
958
953
406
232
7901
17329
8308
-1325
7236
6898
-5767
812
4251
6576
7514
-617
3427
747
11097
-2683
4710
-152
449
—
-3538
731
-904
—
1343
-683
—
—
—
—
—
-432
-546
-705
-855
-915
-1375
-4008
-7374
-7471
-7537
-10661
-10805
-13801
-17844
-18852
-19134
-21586
-27015
-27357
-53941
-55970
United States
NONMFG
MFG
Total
—
280
-511
-1846
796
-1810
-3841
-3845
-4694
-894
1173
-11014
-536
-18765
-10300
10693
-5390
-5896
-4867
-35698
-2243
—
—
43438
-1551
10014
21743
13207
14278
24297
10343
7570
4999
3104
11474
4970
6135
-6979
7526
981
4121
3179
—
5897
1413
2413
—
-765
2806
—
12358
3781
—
-1043
-202
1677
-2060
722
591
-1203
-1734
-8832
-11893
2489
-13869
4354
-6220
-32509
-18864
-18686
-20976
-12780
-70670
—
Total
MFG
NONMFG
Total
MFG
NONMFG
39788
80319
17046
17757
12354
15314
25184
5366
5101
1920
1542
14530
10819
6155
8671
13794
20024
19764
53438
10639
6810
1288
479
4012
3864
172606
269950
71779
73147
54740
62917
89405
28339
27358
16997
15580
52997
62046
90016
25633
28276
37159
47012
51482
4873
11233
10870
8587
110560
179934
46146
44871
17581
15905
37922
23466
16125
6127
6993
25354
23408
15539
16832
1609
3327
2575
1050
6953
891
1618
800
3532
6833
442
6896
3474
544
814
2009
6007
1763
1403
5413
15959
15263
7111
22937
8139
16334
12893
26091
24200
49116
47312
42094
60303
96939
783619
26881
6407
10947
11066
14835
21172
1502
2582
1588
397
10443
5390
3386
1660
6327
1310
2492
2024
2519
332
1145
4087
5429
2769
41410
25442
32418
47310
10248
13543
10972
6273
23577
27621
18002
24226
39463
26274
723
D
2223
776
51
2054
614
D
702
970
1870
1345
253
1485
4244
5766
6551
11013
583
8654
1184
9061
15404
27257
19028
17868
20840
70665
14036
58925
13887
35753
21271
60303
54284
127423
116521
100348
129007
238124
9903
15586
14425
8252
6586
5922
1007
5821
1291
1516
617
4883
15695
D
D
D
D
583
2298
9341
1926
1300
6780
27833
20986
620
33416
9418
20231
12896
35750
34478
53717
70585
60641
71011
68803
410255
373364
2377350
1109339
D
3461
207
895
D
1309
6057
391
4842
2860
D
112
1039
4137
418
1150
3928
11715
9497
560
11924
7556
7680
11709
17030
8796
21859
28284
7011
7467
299
835
551
D
3492
684
5027
6384
3385
11668
21136
1573
D
D
1218
1924
5390
17369
4434
2881
12414
41043
38834
32885
1996
6957
5050
5266
17756
3736
4868
2768
6785
5441
D
19334
5866
D
1341
3092
8028
2508
1581
5634
13210
17848
13416
25509
4469
15522
8375
24553
19806
73706
45936
39707
57996
169321
1268011
D = Data Withheld
S o u r c e : Bureau of the Census, Selected
FEDERAL RESERVE BANK OF ATLANTA
Characteristics
of Foreign-Owned
Firms, S e r i e s F O F , Nos. 1 and 6 , 1 9 7 5 and 1982.
9
Gains in Employment Due to Comparative Advantage
P a r t 1. T o t a l E s t a b l i s h m e n t ( D o m e s t i c a n d Affiliate) E m p l o y m e n t
M
I
i
I
/
/
i
i M
M
\
ü
^
li
ä
r
i M
'
O
Manufacturing Employment Gains
19
Nonmanufacturing Employment Gains
fc^
Total Employment Gains
Source: Bureau of the Census, County Business
Patterns, 1975 and 1982.
'
M
**N
H
Gains in Employment Due to Comparative Advantage
Part 2. Foreign-Owned Multi-Establishment Employment
Manufacturing Employment Gains
Nonmanufacturing Employment Gains
Total Employment Gains
Source: Bureau of the Census, Selected
Characteristics
of Foreign-Owned
Firms, Series FOF, Nos. 1 and 6, 1975 and 1982.
sets of employment gains suggests that foreign
and domestic firms are influenced by similar
locational decisionmaking.8
Numerous elements can be considered for
inclusion as independent variables to explain
state employment changes in a multiple regression statistical model. Indeed, factors included in analysts' checklists used to evaluate
potential sites for plant locations occasionally
number over 100.9 Of course, many factors
that pertain to a particular investor concern may
be grouped together—for example, numerous attributes are associated with the work
force (such as cost availability, suitability, and
productivity) or with a state's fiscal climate
(such as its tax revenue structure and spending
patterns). The multitude of potentially relevant
factors suggests that empirical research in this
area involves a degree of experimentation,
particularly since direct measures of some
theoretically "correct" variables, such as labor
and transportation costs, are unavailable In
addition, it is difficult to interpret certain "proxy"
Factors included in analysts'
checklists used to evaluate
potential sites for plant locations
occasionally number over 100.
measures unambiguously. This fact limits their
statistical usefulness as well. For example, the
level of state taxes is a factor used in almost all
locational preference studies, yet its influence
can be interpreted at least two ways. Negatively,
high taxes may reflect a high cost burden; in a
positive vein, they also may represent a high
level of available public services, such as good
schools and transportation systems.
After some experimentation that eliminated
ambiguities of this sort, a group of useful
variables emerged. Table 2 defines these variables and summarizes their expected influence
These factors are generally considered to be
among the most basic determinants of the
locational decision. Variables dealing with supply include characteristics of a state's work
force, fuel cost, and waterborne transportation
activity. An agglomeration variable (which accounts for a larger mass of an industry, sector or
12
Table 2
Estimated Model*
Dependent Variables
Six equations are estimated that partially explain the
amounts of manufacturing nonmanufacturing, and
total comparative advantage employment for foreignowned firms and for all U.S. firms. The form of the
dependent variables is the ratio of the local share
effect columns in Table 1 to the national employment
changes that occurred in the 1975 to 1982 period.
(The underlying employment information refers to
first quarter levels in 1975 and 1982 and is from the
following U.S Department of Commerce publications;
Bureau of the Census, County Business Patterns and
Selected Characteristics of Foreign-Owned U.S Firms)
Independent Variables
Table 3 reports results for the following variables,
chosen from a larger set of possible factors after
some experimentation:
LABOR
Skilled Workers - ratio of state population over 25
years of age with a high school education to the
comparable U.S figure. Represents the state's share
of the U.S trainable work force, with the impact of the
variable dependent upon the needs of firms for either
"skilled" or "nonskilled" labor. ("Bureau of the Census,
Statistical Abstract of the U.S)
Union Membership - ratio of percent unionization in
a state to the percent in the United States Represents the relative proportion of unionization in the
state. Since it implies other possible labor considerations such as wage rates and labor relations, it is likely
to be a negative influence on investors—that is, the
relationship should be inverse and the sign negative
(Leo Troy, Union Sourcebook)
ENERGY
Fuel Cost-ratio of state's average fuel cost per
1,000 Kilowatt-hour equivalent to the comparable
U.S average fuel cost Represents an often important
production cost factor and therefore should have a
negative influence on investors (Bureau of the Census
Annual Survey of Manufacturers.)
aggregate of industries, or for the proximity of
firms), measured by a state's share of U.S.
employment in 1975, is also included, as are
variables dealing with market size, climate (an
amenity variable), and a state's promotional or
marketing effort Most of the estimated coefficients for this model, as discussed below, are
stable, have the anticipated signs, and are
generally significant The model is stable in that
coefficients have the same signs and similar
values over multiple applications, and these
MAY 1986, E C O N O M I C REVIEW
Climate-ratio of a state's normal seasonal heating
plus cooling days to the comparable national norm.
Represents the severity of a state's climate and is
expected to be a negative amenity influence on
investors (U.S. National Oceanic and Atmospheric
Administration, Climate of States.)
INFRASTRUCTURE
Trade Volume-ratio of state's waterborne trade to
total U.S waterborne trade Represents state's experience in, and infrastructure for, international trade
and should positively influence investment (Bureau
of the Census, U.S. Waterborne Exports and General
Imports, FT985.)
State Development Effort- ratio of the number of
state economic development organizations to the
U.S. total. Represents the state's desire and effort to
attract investment implying a possibly more favorable
investment environment; it is therefore expected to
be a positive influence on investors. (Conway Data
Inc, Industrial Development and Site Selection Handbook.)
CENTRALIZATION P R E F E R E N C E
1975 Employment Share- state's share of U.S total
or affiliate employment Represents the relative concentration of employment in a state, where large
shares are generally found in the most industrialized
states (for example, the Northeast and Midwest). The
impact of this factor depends on whether an investor
seeks the benefits of agglomeration or the benefits of
decentralization. Foreign investors may be positively
influenced by a larger mass of an industry, sector or
aggregate of industries or by proximity of firms; they
may also be repelled by mounting congestion with
increase in size (Bureau of the Census Selected
Characteristics of Foreign-Owned U.S Firms.)
MARKET
Personal Income - ratio of a state's total personal
income to the U.S total. Represents the size of the
consumer market of the state and is expected to be a
positive influence on investors (Bureau of the Census
Statistical Abstract of the U.S.)
^Sources for data are in p a r e n t h e s e s
directions of influence are generally as expected.
Moreover, the explanatory variables "belong"
in the model based on their statistical contribution to predicting values of the employment
series.
Discussion of Model and Results
As a practical matter, the equations estimated
and shown in Table 3 focus on the notion that
FEDERAL RESERVE BANK OF ATLANTA
locational disequilibrium characterized the
employment situation in 1975. That is, both
foreign and domestic firms were in the process
of adjusting to differences in levels of profitability associated with the various states in 1975.
This is a reasonable scenario given the high
national unemployment rate during 1975 and
the period just before, a rate that roughly
doubled from late 1973 to mid-1975, rising
from 4.6 percent to 9 percent Economists
often argue that structural changes accelerate
from economic recession to recovery. It also
seems reasonable to trace employment change
between 1975 and 1982 to differences in the
levels of profit-determining variables in 1975,
especially those affected by the dramatic jump
in energy cost between 1973 and 1975, to
which firms then adjusted over the next several
years.
This line of reasoning is supported by the
results of preliminary equation estimations in
which regression coefficients on the variables
expressed as changes in levels were not signifi-
Both foreign and domestic
firms were in the process of
adjusting to differences in levels
of profitability associated with the
various states in 1975.
cant If the location decisions made in the 1975
to 1982 period represented movements from
one equilibrium to another, employment
changes would have been statistically related
to changes in the profit-determining variables
over the period, not their levels in 1975. Concretely, this means that firms responded, for
example, primarily to differences in the extent
of unionization in 1975, rather than to changes
in the extent of unionization over the 1975 to
1982 period. Empirically, a better story is told
when the determinants refer to levels in one
state relative to others in 1975; that is, when
the changes in variables were included in an
equation with the variables in levels form, the
explanatory power of the model was not improved or it was diminished. Usingthese variable
changes also created statistical problems; estimation of equations with variables expressed
both in level changes and relatives presented
13
the statistical problem known as multicollinearity. This phenomenon occurs when explanatory variables are so interrelated that it
becomes difficult to disentangle their separate
influences and obtain reliable estimates of
their relative effects.
The specific model results for all foreignowned U.S. firms is shown in equation 1 of
Table 3. As expected, a state's comparative
advantage employment gain or loss relative to
U.S. job growth in the 1975 to 1982 period is
inversely related to its relative cost of energy,
average number of heating and cooling days, and
union membership. State employment gains are
positively associated with the levels of international trade shipments, with personal income,
and with the number of economic development
agencies, as expected.
The negative regression coefficient on the
1975 employment share variable suggests that
foreigners are attracted more by decentralization benefits, such as less local competition,
than by the benefits of agglomeration, such as
locating in the proximity of existing businesses.
Also, the negative influence of the education
variable suggests that investors may actually be
more attracted by an unskilled work force than
by a skilled o n e (If states' shares of the U.S.
population lacking a high school education is
used as an independent variable, resuJts indicate that firms are attracted by the unskilled
worker pool.) This implies that firms, both
domestic and foreign, are basing a portion of
their site-selection decision on the availability
of unskilled, and possibly low-wage, workers.10
Taken together, these eight variables explain
just over two-thirds of the variation in affiliate
employment growth among states (R 2 = .69,
see Table 3). These results are very good in
explaining the pattern of employment change,
given the length of time under consideration
and the 1979 to 1980 energy price shock.
In analyzing employment gains traceable to
foreign-owned manufacturing versus nonmanufacturing industries, one might expect to find
variables that operated positively in one sector
but negatively in the other. This was not the
case. However, at least one variable had an
impact on nonmanufacturing but not on manufacturing: adverse climate seems to repel nonmanufacturing affiliates while not having a
statistically discernible effect on manufacturing
affiliates, perhaps because nonmanufacturing
firms are more footloose. An amenity factor
14
such as climate thus is likely to take on greater
importance than it does for manufacturing (see
note 4). Manufacturing and nonmanufacturing
sectors also may not attach equal collective
importance to the factors. Generally, the sectoral equations—2 and 3—are less robust than
equation 1 in terms of both the equations'
overall explanatory power and the degree of
statistical significance of individual regression
coefficients.
State employment gains are
positively associated with the
levels of international
trade
shipments, with personal income,
and with the number of economic
development
agencies.
Equations 4 through 6 show the model results
for total (foreign plus domestic) employment
changes due to comparative advantage and for
overall manufacturing and nonmanufacturing
employment change Generally, the results are
similar to those for foreign-owned employment change. However, as shown by the values
for the coefficient of determination (R 2 ), the
equation pertaining only to all foreign-owned
firms (equation 1) provides a much better fit
than the same equation estimated for foreign
plus domestic employment (equation 4). This
suggests that foreigners are more sensitive to
the economic variables included in the equations.
It is reasonable to expect foreigners to be
more sensitive to these factors? The tentative
answer is yes. This response is based on differences in the nature of the employment growth
processes of foreign-owned and domestically
owned firms. Foreign firms new to the United
States must choose a location, while many
domestic firms are bound by inertia or other
factors to their state of residence. As a consequence, the growth of foreign-owned firms
may be explained better by these basic economic variables than a state's share of overall
employment growth due to comparative advantage. (This will be the case for both overall
employment change and employment change
due to comparative advantage since they are
positively correlated.)
MAY 1986, E C O N O M I C REVIEW
Table 3
Estimated Equations 1
Affiliate
Variable
Skilled Workers
Fuel Cost
Climate
Union Membership
Trade Volume
State Development Effort
1975 Employment Share2
Personal Income
Constant
R2
P
N (Number of Observations)
Domestic and Affiliate
Total
(1)
MFG
(2)
NONMFG
(3)
Total
(4)
MFG
(5)
NONMFG
(6)
-0.6449
-0.5598
-0.3326
-1.551
-6.376
-0.9112
(-2.12)*
(-1.21)
(-0.61)
(-1.85)**
(-2.08)*
(-1.45)
-0.0043
-0.0005
-0.0065
-0.0146
-0.0323
-0.0119
(-1.54)**
(-0.13)
(-1.3)**
(-1.94)*
(-1.18)
(-2.12)*
-0.0075
0.0081
-0.0264
-0.0223
-0.0848
-0.0160
(-1.6)**
(1-14)
(-3.12)*
(-1.7)*
(-1.77)*
(-1.63)**
-0.0125
-0.0067
-0.0145
-0.0178
-0.0709
-0.0123
(-3.97)*
(-1.41)**
(-2.55)*
(-2.02)*
(-2.2)*
(-1.86)*
0.0553
0.0396
0.0497
0.1831
0.4062
0.1453
(1.61)**
(0.76)
(0.81)
(1.91)*
(1-16)
(2.02)*
0.1650
0.1225
0.2419
0.2285
-0.0145
0.2678
(1.95)*
(0.96)
(1.59)**
(0.98)
(-0.02)
(1.52)**
-0.0081
-0.0085
-0.0069
(-6.83)*
(-4.76)*
(-3.24)*
1.307
1.446
0.6132
1.120
5.655
0.5201
(4.24)*
(3.09)*
(1.1)
(1.54)**
(2.13)*
(0.96)
0.0224
-0.0057
0.0500
0.0555
0.1972
0.0399
(3.75)*
(-0.63)
(4.66)*
(3.42)*
(3.33)*
(3.29)*
0.69
13.96
(48)
0.30
3.48
(48)
0.59
9.32
(48)
0.34
4.39
(46)
0.27
3.47
(42)
0.35
4.65
(44)
:
—
—
'Regression coefficients are shown for each variable, with t-statistics in parentheses. Those t-values that are significant at the
.05 level by the appropriate one- or two-tail test are indicated with an asterisk (*), while those significant at the .1 level are
indicated by a double asterisk (**). Observations are for all continental states, except those states for which data were
suppressed.
2 The
domestic and affiliate figures for this variable were not entered by the S P S S / P C regression procedure used to estimate
this equation b e c a u s e of the degree of its collinearity with other independent variables
3 F values for all equations are significant at the .99 level
S o u r c e s for the data used in this table are listed in Table 2.
FEDERAL RESERVE BANK OF ATLANTA
15
Foreign and domestically-owned firms may
also respond differently in terms of direction or
magnitude to individual factors that determine
firm location. Comparing the two sets of three
equations suggests that both foreign and domestic producers respond in the same direction
to variables included in the model, though
their degree of response to some variables
differs. One example is the impact of a change
in a state's share of the nation's pool of high
school graduates. This ratio is calculated to be
more than twice as great for the universe of
U.S. firms as for foreign-owned firms, as shown
Foreign and domestic
investors
alike seem to respond to
differences in potential
profitability
levels as reflected by differences
among states in comparative
advantage positions.
by differences in the magnitude of this variable's regression coefficients, which also are
sensitivity estimates.11
In a few cases the related statistical results
are somewhat disquieting. For example, the
regression coefficients for fuel costs are more
robust for nonmanufacturing than manufacturing, an unexpected result (Lower t-values indicate lesser degrees of confidence that the
coefficients are significantly different than zero.)
This finding, which is difficult to explain, somewhat undermines our confidence in the overall
model. So do a few of the findings from similar
comparisons in the magnitudes of the regression
coefficients, which represent the degree of
influence exerted by the individual factors.
Fuel cost and union membership have a greater
impact in the nonmanufacturing equations than
in those for manufacturing, raising suspicion
concerning the model specification. All in all,
though, the sensitivities to the variables are
relatively close for the model reported in Table
3.
One method for testing directly whether
foreign and domestic investors' preferences
differ is to create a new dependent variable,
the ratio of affiliate and total employment
gains. This variable is then regressed on the
16
determinants of location listed in Table 2.
Using this procedure and estimating the model
shown in Table 3 yielded an equation in which
none of the explanatory variables generated
significant regression coefficients. Moreover,
the percentage of variation explained (R 2 ) was
.1 and the 3.7 F value indicated that the equation's
explanatory power was statistically insignificant
W e concluded that there are no significant
differences in either direction or magnitude for
the factors tested in this model. In fact using
the regression procedure for stepwise selection
of independent variables (adding variables
one at a time to ascertain their marginal contribution in explaining the dependent variable),
all of a large group of potential explanatory
variables (including some variables in addition
to those already discussed) failed to meet
entry requirements up to the 25 percent level
of significance12
Comparison With Other Researchers'
Findings
Numerous studies have looked into foreign
and domestic investors' reasons for locating in
particular states separately, but most of these
have been investor surveys. Typically, results
echo this report by Frank L Dubois and Jeffrey
S. Arpan last yean " I n terms of specific cost
factors critical in the investors' location decision,
two were mentioned by 80 percent of the
respondent firms: (1) salary and wage levels,
and (2) cost of utilities." As a consequence,
there is no way to summarize comparisons between the specific findings of our study and the
myriad subjective survey studies, although our
general results are consistent with other survey
expectations and compatible with their actual
findings.
Not much statistical work has been attempted on
issues like those discussed and reported in this
article. Jane S. Little used regression analysis in
1978 to gain insight into "which locational
characteristics are weighted relatively heavily
by foreign investors in comparison to U.S.
manufacturers." That analysis partially explained
the ratios of states' shares of foreign versus U.S.
constructions and expansions in 1975 and
1976. Little found foreigners gave relatively
more importance to state wage differentials
and the availability of large port facilities than
did U.S. investors, while U.S. manufacturers
MAY 1986, E C O N O M I C REVIEW
appeared comparatively more sensitive to state
variations in fuel costs. Other variables—personal income, degree of unionization, state
industrial development incentives, the number
of trade missions dispatched or development
offices established overseas, state and local
government tax revenues per capita, manufacturing employees, annual average output per
man-hour, and the unemployment rate—appeared to be of relatively minor importance in
explaining the ratios of foreign versus U.S.
constructions and expansions in different states
during 1975 and 1976.
Little's model basically attempts to explain
shares of investment in 1975 and 1976 as a
function of differences in the levels of explanatory variables in 1974. By contrast our study
partially explains shares of employment gains
between 1975 and 1982 that can be attributed
to differences in the levels of comparative
advantage factors in 1975. Thus, the results of
the two models are not directly comparable
However, we estimated an equation analogous
to Little's, using employment as the dependent
variable and the same explanatory variables as
in Little's equation. None of the variables in the
employment equation were significant and the
entire equation provided a poor fit with F = .57
and R2 = .05. While not negating Little's work,
these results raise related questions about
possible differences in factors that explain
employment gain and those that explain actual
plant investment as well as about possible
changes in the relative importance of locational
determinants over time.
Conclusions
It is difficult but we hope not impossible, to
make progress in identifying the links and
quantifying the separate effects of factors that
influence a firm's site location. This type of
study necessarily involves a certain amount of
experimentation. Even so, collinearity among
the explanatory variables can easily cloud the
relationship under study. A few variables were
included in preliminary estimating equations
but were found to be insignificant For example,
a state's interstate highway mileage, expenditure and revenue growth experience, population and work force size, and amount of exported manufacturing shipments failed to explain differences in state employment gains
FEDERAL RESERVE BANK O F ATLANTA
statistically, even though plausible economic
arguments can be made to support their economic importance. Statistical or measurement
problems could account for their lack of significance.
Despite difficulties, we can draw a few tentative conclusions from this study. Foreign and
domestic investors alike seem to respond to
differences in potential profitability levels as
reflected by differences among states in comparative advantage positions. Both types of
investors are attracted by low-skill and nonunionized labor, and the cost of energy matters
Both types of investors are
attracted by low-skill and nonunionized labor, and the cost of
energy matters
The size of the
market also seems to influence
site location, as do efforts to
attract employers.
Besides these cost factors (low-skill and nonunionized labor may proxy for lower-wage
labor), the size of the market also seems to
influence site location, as do efforts to attract
employers. In summary, there is some statistical
support for the argument that economic forces,
based on economic theory, operate for foreign
and domestic investors.
This study also found no significant differences between the factors influencing foreign
employers when they make plant location
decisions and those that affect domestic employers. While a few studies have suggested
that factors influencing the locational decisions
of the two groups may vary—for example, that
foreign firms may be more influenced by international transportation networks and workers'
wage levels than domestic firms, these findings
suggest otherwise. Among the variables considered here, little difference was observed in
the degree of importance placed on resource
or other production costs by the two investor
types. Both domestic and foreign producers
seem to seek out places that enable them to
earn maximum profits, and the profit-determining factors are apparently common to both
producers.
17
NOTES
1A
discussion of impacts, both positive (benefits) and negative
(costs), c a n be found in U.S. Department of Commerce,
"Foreign Direct Investment in the Southeast United States:
A Comparative Analysis," The Costs and Benefits of Foreign
Investment
from a State Perspective,
Cedric L Suzman,
editor, Washington, D.C, 1982.
2 David McClain (1983) p 279.
It should also be noted that some of the statistical differe n c e s found between foreign and domestic firms may
actually be reflecting differences between the average
domestic firm and the average foreign firm that decides to
locate in the United S t a t e s rather than reflecting inherent
"foreignness" or "domesticness." There is a subtle selection
problem with this type of analysis since many foreign firms
will evaluate all regions of the United States and decide on
none of t h e m
E m p l o y m e n t is o n e of the broadest measures of the activity
and importance of industries, but it is not a direct measure of
investment The dollar value of investment arrived at by
using the measuring rod of money, is the yardstick of the
relative importance of investment in different industries The
spacial distribution of investment by industry can vary
significantly from that of employment a s c a n their determinants S e e William J . Kahley, "Foreign Direct Investment—A
Bonus for the S o u t h e a s t " Economic
Review, voL 70, n o 6
( J u n e / J u l y 1985), p p 4-17 for a discussion of differences in
the Southeast's industrial concentrations of employment
and investment assuming that investment is similar to the
industrial distribution of gross book values of property, plant
and equipment
4 As Chalmers and Beckhelm (1976) put it
To the extent that locational equilibrium is achieved in
an industry, a differential shift could occur in that
industry only if the profit topography shifts. Locational
disequilibrium, however, would be consistent with
differential shifts in response to levels of relative
profitability. S i n c e the extent to which an industry is in
locational equilibrium at any point in time cannot be
specified a priori, both c h a n g e s in the absolute levels
of space costs and revenues and their levels at one site
relative to all others should be examined a s determinants of the differential shift
It should also b e noted that "footloose industries," those for
which input cost structures or transport costs are unimportant in determining profits, use other criteria in choosing a
location. Amenities such a s climate and recreational activities thus c a n be important factors in location decisions t o o
5 Notice that w e use the concept of comparative advantage in
providing jobs and in industrial production interchangeably.
While there probably is a high degree of correlation between
the two, they are not the same. A similar remark is applicable
to a comparison of amounts of investment and employment
changes among industries These distinctions are not explored or analyzed in this paper, largely b e c a u s e of a lack of
data
6 A table showing the complete results of the shift-share
analysis is available from the author.
7 ldeally, w e would like to estimate equations using separate
data for foreign-only and domestic-only industry employment growth rather than foreign multi-establishment and
foreign plus domestic single- and multi-establishment industry employment While data are not available in the
preferred form, this limitation does not appear to be of major
importance, since multi-establishment employment accounts for about 98 percent of all such employment and
since no compelling reason suggests that state patterns of
total and multi-establishment employment are likely to
diverge.
Unfortunately, data on foreign affiliate employment are
also not available at the sub-state level or with more
industrial detail (such as the two-digit Standard Industrial
Classification System level) for states These data limitations
thus necessitate analysis at highly aggregated geographic
and industrial levels
8 The correlation coefficient between the two sets of comparative advantage employment gains is 0.81, a statistically
significant association at the .001 probability level
9 S e e , for example "Industrial Growth Charts U p d a t e " published annually by Conway D a t a Inc, in Industrial
Development and Site Selection
Handbook.
1 0 B e c a u s e these variables are positively correlated, the education variables may actually be serving a s a proxy for w a g e
data that are unavailable
11 The concept of sensitivity, in this instance called "elasticity"
by economists refers to the degree to which the dependent
variable c h a n g e s in response to changes in an independent
variable with the levels of other independent variables held
constant
1 2 These variables were a state's interstate highway mileage,
expenditure and revenue growth experience population
and work force s i z e and amount of exported manufacturing
shipments
REFERENCES
Chalmers J a m e s A, and Terrence L Beckhelm. "Shift and
S h a r e and the Theory of Industrial Location,"
Regional
Studies, voL 10, n o 1 (January/February 1976), p p 15-23.
Creamer, Daniel. "Shift of Manufacturing Industries" Chapter
4 in Industrial Location and National Resources. Washington, D.C.: Government Printing Office, 1942.
D u B o i s Frank L , and Jeffrey S Arpan. "Foreign Investment in
South Carolina" Business and Economic
Review, voL 32,
n o 1 (October 1985), p p 30-33.
Gordon, J o h n R , and others " U s i n g the Shift-Share Technique in Economies with Widely Varying Sectoral Growth
R a t e s Observations and a Suggested Model Modification,"
The Review of Regional
Studies, voL 10, n o 1 (Spring
1980), p p 57-67.
18
Little J a n e S "Locational Decisions of Foreign Direct Investors in the United S t a t e s " New England Economic
Review,
July/August 1978, p p 43-63.
McClain, David. "Foreign Direct Investment in the United
S t a t e s Old Currents ' N e w W a v e s ' and the Theory of
Direct Investment" in The Multinational
Corporation
in the
1980's, C P . Kindleberger and D . B Audritsch, e d s Cambridge, M a s s a c h u s e t t s MIT P r e s s 1983.
T o n g Hsin-Min. Plant Location
Decisions
of Foreign
Manufacturing Investors
Ann Arbor, Michigan: University Microfilms International R e s e a r c h P r e s s 1979.
MAY 1986, E C O N O M I C REVIEW
1
s
A i
a Iiii
W
H
T
WSBÊm*
m *
Projecting Federal Deficits
And the Impact of the
Gramm-Rudman- Hollings
Budget Cuts
Spending cuts made now will reduce the
amount of subsequent cutbacks needed to
balance the budget by lowering interest obligations Forecasting future deficits remains a
problematic undertaking, nevertheless
f
r
a
ft i T i
>
m
Thomas J. Cunningham and
Rosemary Thomas Cunningham
Economic forecasting has never been an easy
task. The further one looks into the future, the
more difficult it becomes to predict the course
of events reliably, as one must depend more
and more on assumptions about long-run behavior that may or may not hold true. Such is
the case in projecting federal budget deficits.
By the very nature of the problem, short term
forecasts for one or two quarters are not particularly useful. The relevant horizon for government budgets may be several years, and this is
precisely the period of time that poses the
greatest forecasting difficulties.1 Nevertheless,
current events—in particular the Gramm-Rudman-Hollings balanced budget law—have generated a great deal of interest in deficit estimates. This article sheds some light on how the
Gramm-Rudman-Hollings legislation could affect government spending over the next few
years.
If the Gramm-Rudman-Hollings deficit targets are to be met without raising taxes, given a
reasonable economic scenario, a 4 percent
reduction in programmatic government spending would be necessary for fiscal year 1987,
followed by a slightly less than 2 percent
reduction for fiscal 1988 and an approximately
1 percent reduction in spending in fiscal years
1989 through 1991. Increasing the size of the
cut now will lower future interest obligations,
progressively reducing future budget cuts necessary to achieve the budget bill's final objective
The authors
are, respectively,
a macropolicy
economist
for
the Atlanta
Fed's Research
Department
and an
assistant
professor
of economics
at Agnes Scott College
in
Decatur,
Georgia.
FEDERAL RESERVE BANK OF ATLANTA
19
However, certain caveats must be issued concerning the effort to predict the impact of the
Gramm-Rudman-Hollings legislation.
A Somewhat Murky Crystal Ball
To estimate future deficits, both income and
expenditures must be forecast Estimates of
expenditures are not as easy to make as they
might first appear, though in some areas the
process is mechanical and straightforward. Estimating costs of future construction projects is
one example. If the contracts have been signed,
the time and amount of nominal disbursements are already specified. In other areas,
forecasting outlays is extraordinarily difficult
Anticipating the costs associated with the Strategic Defense Initiative ("Star Wars"), for instance, is problematic, because it is hard to
gauge how much Congress will be willing to
appropriate and because new technologies
may present expensive unforeseen problems
to overcome before the system can function.
The difficulty of predicting costs on most budget items generally lies somewhere between
Inflation can serve to
reduce...
the deficit though the effect is
not as pronounced as with real
growth.
20
these extremes. Medicare expenses, for example, depend on the number of people eligible, a figure known with some certainty, and on
changes in the cost of medical care, which are
more difficult to anticipate.
Both the Congressional Budget Office (CBO)
and the Office of Management and Budget
have large staffs devoted to such estimation
problems, and this article does not presume to
better their judgment Total federal expenditure,
with some notable exceptions, is rather well
understood.
The greatest challenge in forecasting deficits
is anticipating changes in economic conditions.
A higher-than-expected real rate of economic
growth, for example, will reduce income maintenance expenditures by raising the level of
income. Programs other than income maintenance are less influenced by changes in real
economic activity so their expenditure level
remains largely unchanged. At the same time,
real growth means a real increase in tax revenue
The net effect of real economic growth, then, is
a reduction in the size of deficits. If, however,
the economy slides into a recession, these
effects would reverse themselves to create a
budget shortfall much larger than expected.
About one half of all government spending is
indexed to inflation, as are income taxes. As a
result inflation can serve to reduce the real
size of government expenditure commitments,
and hence reduce the deficit though the effect
is not as pronounced as with real growth.
Table 1 presents current C B O projections of
federal deficits through 1991, the relevant
timetable for Gramm-Rudman-Hollings.2 The
most striking feature of these revised deficit
projections is their time trend. Baseline budget
deficits are now projected as consistently declining3 By the year 1991, the C B O projects a
deficit of $104 billion, down from a fiscal year
1986 deficit of $208 billion.
The baseline budget projections make two
crucial assumptions, each of which deserves
some discussion. First they presuppose no real
growth in defense spending. Second, they take
for granted several aspects of economic performance over the course of the next five years.
The supposition that no real growth in defense expenditure will occur seems somewhat
misleading. The budget submitted by the Reagan administration for fiscal year 1987 calls for
real growth in defense outlays of approximately
M A Y 1986, E C O N O M I C REVIEW
Table 1
CBO Baseline Budget Projections
(Billions of dollars)
1986
1987
1988
1989
1990
1991
Revenue
778
844
921
991
1,068
1,144
Outlay
986
1,025
1,086
1,135
1,188
1,248
Deficit
208
181
165
144
120
104
Source: C B O , The Economic
Review
and Budget
Outlook
Fiscal
3 percent and real growth in the military
budget authority of almost 8 percent 4 While
no one knows how much of this real growth
Congress will underwrite, the baseline budget
projection is not designed to accommodate
any program changes not already mandated by
Congress, and a significant increase in real
military expenditure represents a program
change.
Anticipating economic growth is, as discussed
above, a substantial problem in forecasting
deficits. The baseline budget projections rest
on a real rate of growth varying from 3.1 to 3.5
percent and averaging 3.3 percent over the
next five years. Inflation is assumed to range
between 3.4 and 4.4 percent over the same
period, averaging slightly less than 4.2 percent
Growth at a higher rate than this would serve to
reduce the deficit at a much faster rate. Sustained real growth over 4 percent per year
would, the C B O projects, easily eliminate the
deficit within the next five years. The CBO's
lov^growth scenario sees deficits rising substantially with a recession but then resuming a
downward path as the economy returns to its
historical long-run growth rate of slightly over 3
percent Even though revenues are sensitive to
changes in economic performance, the baseline
budget indicates that the economy will eventually outgrow the deficit regardless of fluctuations in the economy.
Cutting Spending From Here to Eternity
Despite these hazards to prediction, an estimate of the size of the budget cuts needed to
FEDERAL RESERVE BANK OF ATLANTA
Years 1937-1991,
February 1986.
achieve the targets of the Gramm-RudmanHollings amendment is desirable. Table 2 presents the baseline budget series and the size of
spending reductions necessary to balance the
federal budget by 1991, following the deficit
time path specified in the amendment Some
explanation of the table's categories is in order:
The deficit series in Table 2 reports both the
actual baseline deficit as in Table 1, and the
same deficit divided between the "basic deficit"
and interest payments on the outstanding federal debt 5 The basic deficit represents the
difference between spending on programs and
tax revenue; it is the total deficit minus interest
This distinction is useful for several reasons.
First the government has only indirect control
over current interest expenses. Interest rates
are determined in money and capital markets,
and the amount of debt outstanding is determined by previous spending and taxation
decisions. Interest outlays represent commitments on bonds already sold; these expenditures cannot be reduced unless the government were to default on its debt an extremely
unlikely eventuality. Interest expenditures can
be reduced only by borrowing less or borrowing
at reduced rates. The former depends heavily
on program spending and implies a reduction
in the basic deficit while the latter is largely out
of the government7 s control.
The government does, however, have a great
deal of control over how much is spent on
various programs. To reduce deficits below
their baseline levels, spending on programs
must be reduced. The result will be a reduction
in the basic deficit
21
Table 2
Cuts Needed to Meet Gramm-Rudman-Hollings Targets
(Billions of dollars)
1986
1987
1988
1989
1990
1991
Revenue
778
844
921
991
1,068
1,144
Outlay
986
1,025
1,086
1,135
1,188
1,248
Deficit
208
181
165
144
120
104
144
108
72
36
0
37
57
72
84
104
Gramm-Rudman-Hollings
Target
Baseline Deficit Less
Gramm-Rudman-Hollings
—
—
Net Interest
139
145
154
158
159
160
Basic Deficit
69
36
11
-14
-39
-56
37
16
12
8
13
4.2
1.7
1.1
0.7
1.0
37
53
65
73
86
Cuts Needed to Meet
Gramm-Rudman-Hollings
Size of Program Cut
(in percent)
—
Cumulative Cut
Source: C B O , The Economic
—
—
and Budget
Outlook:
Fiscal
Years 1937-1991,
In Table 2, the deficit is represented as a
positive number, and surpluses are represented
as negative quantities. Note that in 1989 the
basic budget shows a surplus of approximately
$14 billion. After 1989, the deficit consists
solely of interest expenses on debt already
outstanding.
To meet the Gramm-Rudman-Hollings targets, baseline budget expenditures must be
reduced by $104 billion by 1991. Table 2
presents the size of the cuts needed to meet
this goal, and also expresses them as a fraction
of basic budget spending. (Again, spending on
interest payments could not be reduced unless
the government were to default on interest
payments on its debt) As stated in the introduction, the budget-balancing goals can be
met through overall spending reductions of 4.2
percent in 1987, 1.7 percent in 1988, and
approximately 1 percent for the next three
years, if the C B O baseline budget assumptions
discussed above are correct 6 The actual size of
22
February 1986.
the cut on a year-by-year basis is shown as Cuts
Needed to Meet Gramm-Rudman-Hollings,
while the Cumulative Cut row shows a running
total of the yearly spending reductions. Cumulative cuts of $86 billion ultimately result in the
$104 billion reduction in the size of the deficit
necessary to meet the target
The importance of separating the basic deficit
and interest expenditures can now be illustrated.
To have a balanced federal budget by 1991,
overall government spending must be reduced
a total of $104 billion from baseline estimates.
The cuts outlined in Table 2 achieve that result
but with cuts that total less than $104 billion,
because reductions in the deficit occurring
before 1991 result in smaller borrowing needs
and hence lower future interest expenses. A
reduction in the deficit now lowers future
deficits through both the cut in spending and
the smaller future interest liabilities. By meeting
the deficit reduction targets, which call for
substantial cuts now in the basic deficit future
MAY 1986, E C O N O M I C REVIEW
interest expenses are reduced, requiring progressively smaller future cuts. Thus, total program cuts of $86 billion could result in a $104
billion decrease in expenditure by 1991.
Taking this relation between the change in
expenditure and the change in interest obligation to an extreme, an expenditure reduction
of $80 billion occurring in fiscal year 1987
would be sufficient to bring about a balanced
federal budget by the year 1991. This would
represent an across-the-board reduction of 9
percent in government spending on programs.
Waiting until 1991 would require the full $104
billion to be cut at once.
The Gramm-Rudman-Hollings legislation
does more than simply specifying deficit targets
for a balanced budget in 1991; it also names
areas in which cuts will be made automatically
in the event that the deficit targets (plus a
margin for error) are not met (The automatic
spending reduction mechanism is the subject
of the legal challenge to Gramm-Rudman-Hollings Act based on the constitutional separation
of powers.)
Chart 1 represents the current fiscal yeaKs
budget, divided into various expenditure categories.7 Under the balanced budget legislation,
only 35 percent of total government expense
including interest is subject to automatic cuts.
Regardless of the constitutionality of GrammRudman-Hollings, if it is viewed in any sense as
a potential guideline for reductions in outlay,
the difficulty of achieving a balanced budget is
easily perceived. A comparatively small portion
of overall spending is subject to cuts, which, if
confined to these areas, would result in substantial reductions in the levels of services
affected.8
Chart 1. The Fiscal Year 1986 Budget
Divided Into Gramm-Rudman-Hollings Categories
About 65 percent of the 1986 budget is exempt from cuts and 3 5 percent is subject to across-the-board (A-T-B) or partial c u t s
•Programs in these areas, including some retirement medicare and student loans, m a y b e eligible forcuts but are also exemptfrom
cuts in certain a r e a s
S o u r c e C B O a n d Office of Management and B u d g e t December 1985.
FEDERAL RESERVE BANK OF ATLANTA
23
Taking the cuts now not only lessens the
absolute level of government spending but
also diminishes the necessary size of future
cuts by putting future outlays on a more rapidly
declining time path. Lowering future interest
expenditures will significantly reduce the size
of future cuts needed to balance the budget
NOTES
10ver
Conclusion
Recent concern over the size of federal
deficits has sparked a general interest in the
eventual elimination of these deficits. C B O
baseline budget projections see the federal
deficit falling to roughly half its current size
over the next five years, given a seemingly
reasonable set of economic assumptions. If the
deficit is to be eliminated by then without
raising taxes, substantial cuts in expenditures
are essential.
The sooner these curtailments are made,
however, the smaller the need for overall reduction from current baseline spending levels.
24
a much longer horizon, the importance of business
cycles diminishes Forecasting output growth for the next
few d e c a d e s may be done without reference to temporary
swings in economic activity.
C o n g r e s s i o n a l Budget Office, The Economic
and
Budget
Outlook: Fiscal Years 1987-1991, Government Printing Office,
February 1986.
' " B a s e l i n e budgets" are constructed on the assumption that
current expenditure programs remain in their current state.
With programs neither added nor deleted, and following
economic assumptions discussed below, this is the course
that the budget would take by itself with no c h a n g e s in
programs.
4 From Budget
of the United States Government
Fiscal Year
1987, G P O , 1986. Budget outlays differ from budget authority
in that budget authorization (authority) must occur before
outlays c a n take p l a c e Not all budgeted funds authorized in
a fiscal year will be spent in that year. For a more complete
description of the budget process and terms, s e e Lisa
Rockoff, "The Federal Budget Process: How It Works,"
Economic Review, voL 70, n o 5 ( M a y 1985), p p 38-40.
5 T h e implications of this division are discussed at length in
"The Long-Run O u t c o m e of a Permanent Deficit" this issue.
8 The implied average rate of interest on government debt
outstanding is assumed to continue following C B O estimates,
even though the level of debt outstanding is reduced slightly
from baseline projections The C B O presupposes that the 3month Treasury bill rate will monotonically decline from 6.8
percent to 5.4 percent over the next five y e a r s S i n c e the
average length to maturity of all publicly held federal debt is
now 5 years ( T r e a s u r y Bulletin, Winter issue, first quarter
fiscal year 1986), t h e s e interest rate and expenditure assumptions apply to half the debt outstanding through the
relevant Gramm-Rudman-Hollings period
7 T h e source of the numbers used is the Office of Management
and Budget and the C B O , December 1985, a s reported in the
New York Times, J a n u a r y 16, 1986.
8 l n 1987 alone, the required cut in spending shown in Table 2
represents a reduction in spending of approximately 10.3
percent on programs that are subject to automatic reductions
(both across-the-board and special rule).
MAY 1986, E C O N O M I C REVIEW
The Long-Run Outcome of a
Permanent Deficit
This study shows that whether the central
bank must finance the government depends
more on the government debt-to-income ratio
than on the size of the debt
Thomas J. Cunningham
The size of recently reported deficits, running
on the order of $200 billion, is unprecedented.
To exceed the current deficit as a percentage
of gross national product—now fluctuating in
the range of 4 or 5 percent—one must go back
to World War 11, when the deficit was as large as
28 percent of the GNP. 1 The sheer magnitude
of the shortfall has prompted concern over
problems that arise with prolonged periods of
deficit
One issue is whether a government can
continue on a path of large fiscal deficits without assistance from the monetary authority; or, does the financing of a "permanent
deficit' necessarily imply some accommodation of the shortfall in the future?
The answer to this question, which
will be explored here, depends
upon the eventual stability of
the debt-to-income ratio.
A second, separate issue raised
by the prospect of a large deficitone which will not be addressed
here—is the potentially adverse impact
on the rest of the economy. The "crowding
out' debate, as it is referred to, revolves
around how much the deficits stimulative effects
are offset by declines in private investment or
changes in the level of net exports. Whatever
The author
Research
FEDERAL RESERVE BANK OF ATLANTA
is a macropolicy
economist
for the Atlanta
Feds
Department
25
crowding-out problems exist are clearly exacerbated as government deficit financing increases,
suggesting a loose link between the problem of
stability addressed in this article and the concerns
raised in the literature about the crowding out
effect
The Deficit and the Monetary
Authority
Recently renewed interest in the Real Bills
doctrine has sparked a related concern about
the ability of the fiscal authority to operate
independently of policies undertaken by the
monetary authority.2 In the United States, the
spending and taxing (fiscal) authority is best
thought of as a shared function of the President
and the Congress, and the monetary authority is
the Federal Reserve. The Real Bills doctrine has
seen many different interpretations. In the
version now receiving attention, the value of
the assets backing the money supply is seen as
being ultimately responsible for the value of
money itself. If the monetary authority purchases bonds from the fiscal authority because
the fiscal authority cannot sell them to anyone
else, then the money created in the process of
purchasing these worthless bonds will, itself,
be worthless.
Even if this is not the case, should the fiscal
authority find it impossible to continue its
policies without aid from the monetary authority, the economy could well be affected in the
present by events that might not actually occur
for several generations. Anticipation of situations expected to unfold over the course of
several generations might then influence current spending and saving behavior.
Long-run workability, then, is more important
than it first appears. Robert Barro, in a pathbreaking article, argues that there is an equivalence between bond and tax financing of government spending, so long as the bonds issued
to finance spending now are expected to be
paid off in the future 3 If there is an intergenerational transfer of wealth (bequests), then
leaving government bonds to future generations also means leaving an equivalent tax
liability to the same generation in which the
bond is to be paid off. As a result government
bonds that are expected to be paid off do not
represent net wealth, and the economy will
alter its savings behavior accordingly. Thus,
26
events expected to take place in future generations may affect current economic decisions.
In discussing policies undertaken by the
fiscal authority two interrelated concepts are
helpful. The first concept is the "basic budget"
sometimes called the "primary budget" which
represents government spending on all noninterest items, less tax revenue received. If
spending on goods, services, and transfers
exceeds tax revenue, the basic budget is in
deficit if tax revenue exceeds programmatic
spending, the basic budget is in surplus. The
reported deficit on the other hand, represents
the basic deficit plus whatever borrowing is
necessary to service the debt already outstanding. If no debt were outstanding, and hence no
interest payments were to be made, the basic
budget balance would coincide with the reported budget Otherwise the reported deficit
will equal zero only when the basic budget has
a surplus equal to the interest payments due
on the debt outstanding.
The "fiscal regime," the second important
concept is the set of spending and taxation
policies to which the government is committed.
Anticipation
of situations
expected to unfold over the
course of several generations
might...
influence
current
spending and saving behavior.
Growth in national defense spending, Medicare
benefits, and different tax benefits are some of
its components. The fiscal regime usually is
responsible for the basic budget balance: it
dictates how much is to be spent on various
programs and what the rate of taxation will be.
In deciding the tax rate, the government also
decides how much tax revenue will be collected
at any given level of income, and thus results
the basic budget balance. With the exception
of a change in the fiscal regime, the basic
budget balance can change only if the level of
income, and hence the amount of tax revenue,
changes.4
MAY 1986, E C O N O M I C REVIEW
Strictly speaking this definition of the fiscal
regime is too narrow. It could also be the case
that the government sets specific targets for
the actual budget deficit thus making the basic
budget responsive through spending, taxation
or both, to the amount of interest it must pay
on its accumulated debt
The current U.S. budget process provides an
excellent example of the fiscal regime in operation. Clearly the goal is to reduce the reported
deficit The stumbling block, however, comes
at the basic budget level, where it is difficult to
agree which programs should be cut or whose
taxes should be raised. The problem is that the
fiscal regime results in a basic budget deficit
which, combined with the interest payments
on the debt already outstanding, leads to a still
larger actual deficit
From the perspective of those interested in
the Real Bills doctrine, one concern about the
deficit is whether a given fiscal regime will lead
to a situation that requires the monetary authority to actively accommodate the fiscal authority.
Whether or not the monetary authority is
"forced" to help finance the deficit depends
As long as the government
debt-to-GNP ratio is stable; the
government should have no
problem marketing its debt
regardless of the nominal size
of the debt to be sold.
on the Treasury's ability to sell to the public the
bonds necessary to finance the deficit The
public's willingness to assume the additional
government debt depends, among other things,
on the amount of government debt it already
holds. The greater the amount of debt outstanding, the harder it may be to borrow more. A
fiscal regime that runs a substantial basic budget deficit must issue a large amount of debt
The issuance of debt means interest payments
must be made, and, since the basic budget is in
deficit these new interest payments can be
made only by selling additional bonds. Thus,
bonds must be sold to pay interest on bonds
FEDERAL RESERVE BANK O F ATLANTA
already outstanding So long as the fiscal regime
fails to produce a basic budget surplus large
enough to cover interest payments on the
outstanding debt completely, the debt outstanding will grow.
The willingness of the public to hold additional government debt is not jeopardized,
however, simply by the size of the outstanding
debt or by the fact that it is growing in absolute
terms. What is important is whether the level of
debt is high and growing in relation to the level
of income. (For simplicity's sake, the level of
income is assumed to be equal to GNP.) If the
level of debt is high, but the additional debt
being marketed is a trivial proportion of income growth, the government should not have
any difficulty marketing it and should not be
forced to resort to the monetary authority for
aid.5 As long as the government debt-to-GNP
ratio is stable, the government should have no
problem marketing its debt regardless of the
nominal size of the debt to be sold.
On the other hand, if the fiscal regime requires that the government debt-to-income
ratio grow continuously, the fiscal authority
eventually will be unable to sell enough of its
bonds to finance its deficit because investors
will become reluctant to add additional government bonds to their portfolios. The fiscal authority will then have to turn to the monetary
authority for financing if the fiscal regime is to
continue
Similarly, the fiscal regime may result in a
stable debt-to-income ratio but at a level too
high for the economy to tolerate. Even though
the debt-to-income ratio is mathematically
stable, at some point before stability is achieved
the public may simply refuse to purchase
additional bonds. Thus, the fiscal authority
would have to turn to the monetary authority
for financial aid. Exactly what constitutes a "too
high" debt-to-income ratio is difficult to divine.
At the end of World War II, the public held
government bonds amounting to 1.15 times
the gross national product 6 Other economies
have held a substantially greater proportion of
debt
Even though the fiscal authority's financial
condition places a theoretical constraint on the
monetary authority, the constraint is not necessarily binding. For example, if the financing
required by the fiscal authority implies a rate of
growth in the money supply of, say, 2 percent
27
arid the monetary authority is expanding the
money supply by 4 percent the 2 percent
constraint imposed on the monetary authority
by the fiscal authority, while real, is also trivial.
On the other hand, if the real after-tax rate of
interest is greater than the real growth rate of
the economy, then a deficit regime is infeasible
without monetary accommodation, since the
debt by itself, will grow as a proportion of GNP.
The Standard Model
Growth Rates Versus Interest Rates. The
question of when a fiscal deficit regime is
feasible without the aid of monetary intervention has recently received a great deal of
attention in the form of a debate concerning
current U.S. fiscal policy between Thomas J.
Sargent, Neil Wallace, and Preston J. Miller on
one side and Michael Darby on the other.7
Whether or not a deficit regime eventually will
result in a stable debt-to-income ratio hinges
on the relationship between the real after-tax
rate of interest and the real growth rate of the
economy.
This relationship becomes clear if we examine
how debt service contributes to the borrowing
needs of government Debt service equals the
real after-tax rate of interest times the amount
of debt outstanding. (The term "rear eliminates
illusions caused by inflation and "after-tax"
designates what the government actually, net
pays.) Thus, the real size of the debt outstanding
will grow by itself, independently of the borrowing needs imposed by the the basic budget
at the real after-tax rate of interest
If the rate of interest is less than the growth
rate of the economy, then the debt will grow
more slowly than the economy. As a proportion
of GNP, then, the debt would be shrinking and
would eventually approach zero, even though
the level of debt might be quite large and
growing If the fiscal regime calls for a basic
deficit that is a constant proportion of GNP, a
constant debt-to-income ratio will evolve, since
the only significant factor contributing to debt
in the long run will be the debt issued to pay for
the basic deficit Debt service will eventually
shrink to a trivial portion of GNP, although this
process may take considerable time. Since the
basic deficit is assumed to be a constant proportion of GNP, the addition to debt needed to
finance the basic deficit is also a constant
proportion of GNP. The total debt-to-GNP
ratio will approach a stable maximum asymptotically, that is, the ratio will constantly approach
its stable level though it will never finally reach
it
28
If the rate of interest is less
than the growth rate of the
economy; then the debt will grow
more slowly than the economy.
Unlimited growth of the debt-to-GNP ratio (an
"exploding" ratio) is something that cannot
continue forever. An infinite amount of debt
cannot be purchased.
The Deficit and Debt with Inflation. Discussions of stability usually revolve around real, as
opposed to nominal, interest rates largely for
the sake of tractability. If there were no inflation
in the economy, then real rates of growth and
interest would correspond directly to the nominal rates we actually observe. Inflation complicates the picture in several ways. First is the
familiar problem of discriminating between
nominal changes that occur because of real
changes in variables and changes that occur
because the value of money has changed. For
example, the 1970s saw periods of combined
inflation and recession, when the nominal level
of income rose but the real level of income fell.
The second problem, again a familiar one, is
that inflation is difficult to forecast and the
difficulty increases with the time horizon. Since
the deficit fiscal regime may continue (in both
real and nominal terms) for generations, a
discussion in real terms avoids this complication. The final problem concerning inflation
relates specifically to government finance.
Since financing a deficit is a nominal problem, inflation may make deficit financing easier
in two ways. First if income taxes are progressive
and unindexed, "bracket creep" will increase
the government's real tax revenue. If real wages
remain unchanged, nominal wages will rise by
the rate of inflation. The rise in nominal wages
will push the earner into a higher tax rate,
thereby increasing the proportion of income
M A Y 1986, E C O N O M I C REVIEW
given up in taxation even with an unchanged
real wage U.S. income taxes are not fully
indexed, though there is considerable movement in that direction. Second, and perhaps
more interesting, inflation represents a partial
repudiation of the debt already outstanding.
Since the outstanding bonds are denominated
in nominal terms, the real value of the debt
previously issued decreases as the price level
rises. To the extent that inflation decreases the
real burden of servicing the already outstanding
debt the government is better off financially.8
Thomas J. Sargent argues that government bonds
should be indexed to the price level so that they
would retain their real value as the price level
changes, thus helping to make a government's
anti-inflation plan credible by removing a substantial motive to inflate.9
Applications to the United States
Whether or not the debt-to-income ratio will
eventually become stable can be seen through
a simple set of simulations showing the longrun results of a given basic fiscal regime under
various economic conditions. These simulations are based on historical experience in the
United States strictly for the sake of familiarity.
Historical Growth and Interest Through the
post-World War II era, the United States has
enjoyed a real growth rate that has averaged
between 3 percent and 4 percent a year,
though swings have been considerable. At the
same time, real after-tax rates of interest on
U.S. Treasury securities have been on the order
of 2 to 3 percent The federal budget deficit is
currently around 4 percent of GNP, historically
quite high for a non-war, non-recession year.10
The figures presented here show several
debt dynamics assuming various average interest rate and growth trends.11 Of course,
short-run economic fluctuations will obscure
long-run trends, making the short run look very
different from these pictures. Nonetheless,
eventually the long run arrives (we may be
dead, but not our heirs) and with it come the
long-run results.
In the legends of the figures, "g" stands for
the real growth rate of the economy, G/Y
represents basic government spending as a
proportion of GNP, T/Y stands for the percentage of G N P collected as tax revenue net taxes
on interest income from government bonds.
FEDERAL RESERVE BANK OF ATLANTA
The difference between G/Y and T/Y is the
basic (not actual) deficit as a proportion of
GNP. The symbol
represents the real aftertax rate of interest and "mg" is the rate at
which the real money supply grows. All values
are expressed in 1985 dollars. Inflation, which
would make the financing process easier by
reducing the values of the bonds outstanding,
is not considered here As a result all simulation
figures are in 1985 dollars.
The two different average rates of money
growth used in this exposition are not meant to
imply anything about actual or preferable
monetary policy. A zero percent rate of money
growth obviously is not realistic It is, however,
of substantial theoretical interest for it establishes clearly if the deficit regime, without any
monetary accommodation whatsoever, is feasible. The 4 percent money growth figure is
frequently used in discussions of possible
growth rates under a "fixed rule" monetary
regime, generally advocated by monetarist
economists.
Figure 1 illustrates the debt-to-income ratio
for the next 25 years assuming a 3 percent
basic budget deficit a real rate of growth of 3
percent an after-tax interest rate of 2.5 percent, and a money growth rate of 4 percent
Figure 1. The Debt-to-Income Ratio,
1985-2010
g = 3 % , G/Y = 2 2 % , T/Y = 19%, i = 2.5%, mg = 4 %
"g" = real growth rate of the economy
G/Y = basic government spending a s a proportion of G N P
T/Y = percentage of G N P collected a s tax revenue
T = real after-tax rate of interest
"mg" = rate at which the real money supply grows
29
Since the real rate of growth in this scenario is
greater than the real after-tax interest rate, the
deficit regime is inherently feasible, as the
debt-to-income ratio eventually will stabilize.
Since short-run business cycle
fluctuations will swamp longterm trendsthey
make the
underlying growth and interest
rates difficult to discern.
Even if the money growth rate were zero
percent, or if the feasibility assumption were
violated with a growth rate for the economy of
2 percent and an after-tax interest rate of 3
percent, in none of these cases would the
debt-to-income ratio be higher by the year
2000 than it was at the end of World War II
(again in 1985 dollars). Regardless of the ultimate feasibility of the fiscal regime, the net
impact on the economy, in terms of the government debt-to-income ratio, would be rather
similar. As Robert Barro argued, however, the
economy may alter its behavior with regard to
savings and bequests depending on the longrun outcome of the fiscal regime. Even though
the difference in debt-to-income ratios might
not be apparent within a generation, the ultimate feasibility of such regimes can be determined simply by comparing growth and interest
rates. Given the basic deficit, crowding out and
other problems associated with large fiscal
deficits would exist in similar magnitudes (at
least for the next 100 quarters) regardless of
the long-run independent workability of the
fiscal regime.
Figure 1 is drawn on the basis of long-run
historical trends in the economy, something
we never really get to observe in a short period
of time. In fact since short-run business cycle
fluctuations will swamp long-term trends, they
make the underlying growth and interest rates
difficult to discern. The apparent relationship
between the real rate of growth and the real
rate of interest will change constantly, making
the overall trend clear only after a comparatively
long time.
30
This problem is exacerbated by the continually changing nature of the fiscal regime.
Changes in taxation and spending occur constantly, which may or may not represent a real
change in the fiscal regime. Since the fiscal
regime may affect real interest rates as well as
real growth rates, it could take some time to
evaluate the "new" fiscal regime's feasibility,
and by then the fiscal policy might again be
changed.12
In the Long Run
Observing the feasibility of a new fiscal
regime, then, is a long-run matter. Figures 2 and
3 illustrate the outcome of the permanent
basic deficit regime pictured in Figure 1 after
200 years. Figure 2 depicts the long-run outcome assuming a deficit regime in basically
feasible economic conditions: 3 percent real
growth and a real after-tax interest rate of 2.5
percent The debt-to-income ratio stabilizes at
around six to one without monetary accommodation. While the deficit would still be a
significant proportion of G N P and the crowdingout problems associated with the deficit would
persist the fiscal regime by itself is not infeasible
Figure 2. The Debt-to Income Ratio,
1985-2185
g = 3 % , G/Y = 2 2 % , T/Y = 19%, i = 2.5%, mg = 0 %
"g" = real growth rate of the economy
G/Y = basic government spending a s a proportion of G N P
T/Y = percentage of G N P collected a s tax revenue
T = real aftertax rate of interest
"mg" = rate at which the real money supply grows
MAY 1986, E C O N O M I C REVIEW
Figure 3. The Debt-to-lncome Ratio,
1985-2185
g = 2 % , G/Y = 22%, T/Y = 19%, i = 3 % , mg = 0 %
The Debt-to-Income Ratio
"g" = real growth rate of the economy
G/Y = basic government spending a s a proportion of G N P
T/Y = percentage of G N P collected a s tax revenue
V = real after-tax rate of interest
"mg" = rate at which the real money supply grows
W e r e the fiscal regime deemed important
enough politically, it could continue without
necessarily impinging upon the monetary authority, regardless of the associated financing
problems.
Figure 3 illustrates the result of a deficit
regime under fundamentally infeasible conditions: a 2 percent growth rate and a 3 percent
real after tax rate of interest Instead of asymptotically stabilizing as in Figure 2, the debt-toincome ratio is large and increasing at an ever
faster rate. (It's blowing up!)
government debt to expand the money supply,
accommodation, to a greater or lesser degree,
is taking place.
Consider a monetary authority that is following a basically monetarist policy of expanding
the money supply by 4 percent per year.14
Figure 4 assumes the same economic conditions as those depicted in Figure 3, though now
with the monetary authority following a 4
percent money growth rule and the viewing
time stretched to 300 years.
Since the money supply is growing faster
than the interest rate, eventually the expansion
of the money supply will be great enough not
only to service the debt outstanding but also to
finance the deficit itself. The debt-to-income
ratio will drop to zero, and the monetary
authority will need to purchase bonds from
sources other than the fiscal agent to continue
the 4 percent rate of monetary growth. This
monetary regime may, however, imply a modest rate of inflation, since the money supply is
growing faster than real GNP. Since infeasible
deficit regimes occur because the after-tax
interest rate is greater than the growth rate of
the economy, and this monetary accommodation occurs by expanding the money supply at
Figure 4. The Debt-to-lncome Ratio,
1985-2285
g = 2 % , G/Y = 2 2 % , T/Y = 19%, i = 3 % , mg = 4 %
The Debt-to-lncome Ratio
The Power of Monetary Accommodation
Monetary accommodation, to a greater or
lesser degree, takes place whenever the monetary authority purchases the fiscal authority's
bonds, the usual process for expanding the
money supply.13 Of course, at any time the
monetary authority could purchase all the fiscal
authority's bonds, in effect printing money to
finance the deficit but such a step might prove
inflationary. For a monetary authority, continually "monetizing" an out-of-control deficit is
generally considered imprudent Nevertheless,
whenever the monetary authority purchases
FEDERAL RESERVE BANK OF ATLANTA
"g" = real growth rate of the economy
G/Y = basic government spending a s a proportion of G N P
T/Y = percentage of G N P collected a s tax revenue
"i" = real after-tax rate of interest
"mg" = rate at which the real money supply grows
31
a rate greater than the rate of interest, the
money supply is growing faster than the economy, raising the mone^to-GNP ratio and thus
possibly implying some inflation. An infeasible
deficit regime cannot be accommodated unless
the money supply grows faster than the economy.
Varying the Deficit The basic fiscal regime
that drives the actually reported deficit need
not remain constant Historically large deficits
may create pressure to reduce the borrowing
necessary to finance government operations,
even if the basic fiscal regime is by itself
feasible. (Again, simply because a fiscal regime
is feasible does not necessarily make it desirable.)
Figure 5 depicts the results over the next 25
years of one such basic-deficit-elimination
scenario. This case makes the same assumptions as depicted in Figures 1 and 2, but now
real spending is held constant (in 1985 dollars)
until the basic budget is balanced. Tax revenues
remain at 19 percent of GNP, and government
expenditures are set at 19 percent of GNP,
after the basic budget is in balance to assure
continued balance.
A deficit even a comparatively
large one, need not necessarily
impinge upon the monetary
authority.
Continuing the assumptions of 3 percent
real growth and 2.5 percent real after-tax interest rates, tax revenue would grow sufficiently
to balance the basic budget by around the year
1991. Once the basic budget is in balance, the
only source of additional debt is the service
payments due on the bonds already outstanding. Since the economy is growing faster than
debt service funding requirements (g is greater
than i), the debt-to-income ratio will then
shrink toward zero without monetary accommodation, even though there may be a substantial reported deficit remaining.
The scenario depicted in Figure 5 is not
intended to reflect upon current efforts to
control the deficit It does, however, amply
32
Figure 5. The Debt-to-Income Ratio,
1985-2010
g = 3 % , G/Y = T/Y = 1 9 % (1991), i = 2.5%, mg = 0 %
The Debt-to-Income Ratio
"g" = real growth rate of the economy
G/Y = basic government spending a s a proportion of G N P
T/Y = percentage of G N P collected a s tax revenue
" P = real after-tax rate of interest
"mg" = rate at which the real money supply grows
illustrate the nature of the feasibility argument
by isolating the debt service component of the
reported deficit In this scenario, once the
basic budget is balanced, debt service as a
proportion of G N P continuously declines.
The Moral
The feasibility of a fiscal regime does not
necessarily depend on the spending and taxation decisions associated with that regime.
Indeed, while the basic budget contributes to
both outstanding debt and the additions to the
debt so long as the basic deficit is a constant
fraction of GNP, feasibility of the deficit is
determined by the relationship between the
growth rate of the economy and the real rate of
interest the government pays on its outstanding
debt
If the growth rate of the economy exceeds
the after-tax rate of interest a constant basic
deficit regime will result eventually in a stable
government debt-to-GNP ratio without any aid
from the monetary authority. Thus a deficit
even a comparatively large one, need not
necessarily impinge upon the monetary authority. Varying the historically based assumptions
MAY 1986, E C O N O M I C REVIEW
monetary authority at some future date (and
the economy's reaction in the present to such
future constraints), undesirable effects of large
deficits, such as crowding out, may exert considerable political pressure upon the monetary
authority. Yet these pressures could be seen as
completely independent of the actual financial
needs of the fiscal authorities. Despite the
undesirable side effects of the deficit it may
not necessarily require any expansion of the
money supply that is not desired by the nation's
central bank
concerning growth and interest rates does not
significantly change this conclusion over the
course of the next quarter century, provided
the basic deficit remains unchanged. Changes
in the basic deficit can, of course, affect the
path of the debt-to-income ratio significantly in
a comparatively quick period of time; freezing
the real level of government expenditure results
in a rapid turnaround in the debt-to-income
time path as the basic deficit disappears.
Beyond the purely mathematical requirement that fiscal deficits may impose on the
APPENDIX
Drawing the Pictures
The following equations were used to generate the debt-to-income ratio:
GNP t = GNP t _i x (1 + the assumed growth rate)
Debtt = [Debtt_-] x (1 + the assumed rate of interest)] + (the basic deficit)t - (Money t - Moneyt.-|)
Money t = Moneyt.-j x (1 + the assumed rate of money growth),
where " f represents the present period, and "t-1" the previous period The simulations begin with
end of the second quarter, 1985 levels of GNP, government debt outstanding, and monetary base.
NOTES
' T h e s e are calculated from The Economic
Report of the
President
February 1983, Government Printing Office 1983,
and The Budget of the United States Government
FY 1986,
(GPO 1985).
2
A wide body of literature much of it coming from researchers
associated with the Federal R e s e r v e B a n k of Minneapolis, is
summarized in Thomas J . Cunningham, "The Asset S i d e of
Money Liabilities" (Ph.D. dissertation, Columbia University,
1985).
3 Robert J .
Barro, "Are Government Bonds Net W e a l t h ? "
Journal of Political Economy, voL 82 (1974), pp. 1095-1118.
4 0ff budget expenditure is, for simplicity, assumed to be " o n
budget" Also spending on unemployment compensation
and other programs that depend upon the business cycle or
vary with the level of income is held constant since, in the
present long-run analysis, any sort of cyclical movement is
ignored Similarly, c h a n g e s in the composition of income
(wages vs. profits) are ignored
s T h e simplifying assumption being made here is that the
public is willing to continue to hold the s a m e proportion of
government debt to income it w a s willing to hold in the p a s t
®Computed from Economic Report of the President
February
1983 ( G P O , 1983).
7 S e e Thomas J . Sargent and Neil W a l l a c e " S o m e Unpleasant
Monetarist Arithmetic," Federal Reserve B a n k of Minneapolis, Quarterly Review, voL 5, n o 3 (Fall 1981), pp 1-17.
Michael R. Darby, " S o m e Pleasant Monetarist Arithmetic,"
Federal R e s e r v e B a n k of Minneapolis, Quarterly Review, voL
8, no. 2 (Spring 1984), p p 15-20, and Thomas J . Sargent and
Preston J. Miller, "A Reply to Darby," Federal R e s e r v e
B a n k of Minneapolis, Quarterly Review, voL 8, n o 2 (Spring
1984), p p 21-27, all reprinted in the Quarterly Review of the
Minneapolis Federal R e s e r v e Bank, Winter 1984. An assumption generally made in the current debate but relaxed
in this article is that the deficit regime specifies a basic
budget deficit that is a constant proportion of G N P .
8 E v e n if people anticipate some inflation at the time they
purchase the bond and therefore demand a higher rate of
FEDERAL RESERVE BANK OF ATLANTA
return at the time of the initial sale, inflation will still erode
the real value of the b o n d T h e extent of the bondholder's
loss depends on how much higher inflation is then anticipated
"Thomas J. S a r g e n t "Stopping Moderate Inflation: The Methods of Poincare and Thatcher," Conference on Indexation
and Economic Stability in the World Economy, Escola de
Pos-Graduacao em E c o n o m i c do Fundacao Getulio Vargas,
Rio de Janeiro, December 1981.
10 Wars, of c o u r s e push up spending. A s discussed earlier, the
deficit-to-GNP ratio reached its high during World W a r II
(1943) at 28 percent of G N P . Recessions, on the other h a n d
decrease tax revenue a s economic activity declines and
increase government transfer payments in the form of
income protection p a y m e n t s The recession starting in
1974 brought about a peak quarterly deficit of close to 6
percent of G N P .
" E q u a t i o n s used in t h e s e simulations are contained in the
appendix.
1 2 Sargent and Darby a g r e e that a change in fiscal regime may
influence the relationship between the interest rate and the
growth rate A s the financing n e e d s of the government
increase, interest rates may rise a s the government is forced
to offer higher rates of return in order to sell additional d e b t
eventually pushing the real after-tax rate of interest a b o v e
the real rate of growth. Sargent argues that the current fiscal
regime has produced such an infeasible result while Darby
argues that the current fiscal regime has left the growth
rate/interest rate relation fundamentally u n c h a n g e d in a
feasible condition
13 ln the United States, the monetary authority
purchases
government bonds on the open market and does not p u r
c h a s e securities directly from the Treasury. The monetary
authority is independent of the fiscal authority.
, 4 The "high-powered" or "base" money supply, not M1.
33
The Federal Reserve Bank of Atlanta Research Department
provides a working paper series to stimulate the professional
discussion and exploration of economic subjects W e welcome
readers of the Economic
Review to fill in the coupon below and
return it to receive new technical papers as they are published.
If you would like copies of any of the current papers listed
above, simply check the subject that interests you and return
this page and coupon. All papers are provided free of charge.
Return to:
Information Center
Federal Reserve Bank of Atlanta
104 Marietta Street N.W.
Atlanta Georgia 30303-2713
Please start my subscription to the
Working Papers Series
Name
Address
City
34
State
Zip.
MAY 1986, E C O N O M I C R E V I E W
FINANCE
r i r i f i f
i
APR.
1986
MAR.
1986
APR.
1985
ANN.
%
CHG.
APR.
1986
MAR.
1986
APR.
1985
ANN.
%
CHG.
704,881
28,745
155,636
520,294
MAR.
653,534
64,214
694,541
26,107
155,645
512,654
FEB.
652,508
63,136
N.A.
N.A.
N.A.
N.A.
MAR.
513,518
57,610
+2/
+11
91,726
4,769
19,977
67,056
MAR.
92,474
5,295
90,895
4,367
19,765
66,678
FEB.
93,061
4,776
N.A.
N.A.
N.A.
N.A.
MAR.
N.A.
N.A.
5,843
278
1,020
4,585
MAR.
5T8ÎÔ
347
5,771
256
1,014
4,541
FEB.
5,880
318
N.A.
N.A.
N.A.
N.A.
MAR.
4,327
291
+34
+19
60,101
3,183
13,753
42,870
MAR.
56,064
3,519
59,458
2,901
13,600
42,636
FEB.
56,638
3,200
N.A.
N.A.
N.A.
N.A.
MAR.
46,333
3,087
+21
+14
7,741
593
1,682
5,572
MAR.
10,941
528
7,712
549
1,686
5,542
FEB.
10,798
452
N.A.
N.A.
N.A.
N.A.
MAR.
9,164
419
+19
+26
9,866
352
2,101
7,583
MAR.
10,101
253
9,864
324
2,066
7,577
FEB.
10,176
252
N.A.
N.A.
N.A.
N.A.
MAR.
10,277
N/C
1,778
86
268
1,418
MAR.
2,715
253
1,748
80
264
1,403
FEB.
2,721
262
N.A.
N.A.
N.A.
N.A.
MAR.
2,565
N/C
6,397
277
1,153
5,028
MAR.
6,843
395
6,342
257
1,135
4,979
FEB.
6,848
292
N.A.
N.A.
N.A.
N.A.
MAR.
6,173
322
$ millions
Commercial Bank Deposits
Demand
NOW
Savings
Time
C r e d i t Union Deposits
Share D r a f t s
Savings & T i m e
1,584,570 1,552,866 1 ,418,379
326,316
304,147
344,800
113,067
94,555
119,953
439,941
449,723
390,026
667,829
711,139
714,012
N.A.
52,902
51,229
N.A.
7,395
6,739
45,329
43,823
N.A.
+12
+13
+27
+27
+15
Mortgages Outstanding
Mortgage C o m m i t m e n t s
Commercial Bank Deposits
Demand
NOW
Savings
Time
C r e d i t Union Deposits
Share D r a f t s
Savings & T i m e
187,147
39,777
16,771
51,176
84,316
6,247
721
5,250
181,896
37,671
15,483
50,086
83,351
6,105
656
5,109
166,494
35,979
12,762
44,600
77,277
N.A.
N.A.
N.A.
+12
+11
+31
+15
+ 9
Commercial Bank Deposits
Demand
NOW
Savings
Time
C r e d i t Union Deposits
Share Drafts
Savings & T i m e
18,599
4,118
1,582
3,822
9,627
967
146
713
18,095
3,883
1,504
3,755
9,941
940
131
692
16,846
3,565
1,188
3,440
9,148
N.A.
N.A.
N.A.
+10
+16
+33
+11
+ 5
FLQRIDA
mmercial Bank Deposits
Demand
NOW
Savings
Time
edit Union Deposits
Share Drafts
Savings & T i m e
70,081
15,117
7,209
23,653
26,034
3,237
371
2,675
6 ,526
14,,048
6 ,,644
2 2 ,,984
2 5 .,552
3,,162
343
2,,603
61,311
13,475
5,431
21,291
22,594
N.A.
N.A.
N.A.
+14
+12
+33
+11
+15
Commercial Bank Deposits
Demand
NOW
Savings
Time
C r e d i t Union Deposits
Share Drafts
Savings & T i m e
29,397
7,948
2,225
8,112
12,494
1,171
113
1,079
28,504
7,579
2,060
7,946
12,380
1,156
100
1,050
25,063
7,081
1,632
6,644
11,027
N.A.
N.A.
N.A.
+17
+12
+36
+22
+13
Commercial Bank Deposits
Demand
NOW
Savings
Time
C r e d i t Union Deposits
Share Drafts
Savings & T i m e
29,051
5,404
1,908
7,549
14,655
28,780
5,316
1,834
7,443
14,730
27,018
5,360
1,619
5,892
14,652
+ 8
+ 1
+18
+28
+ 0
*
*
*
*
*
*
*
*
*
Commercial Bank Deposits
Demand
NOW
Savings
Time
Credit Union Deposits
Share Drafts
Savings & T i m e
13,646
2,501
1,135
2,713
7,541
13,341
2,441
1,097
2,684
7,426
12,352
2,338
889
2,417
6,999
Commercial Bank Deposits
Demand
NOW
Savings
Time
C r e d i t Union Deposits
Share Drafts
Savings & T i m e
*
*
*
*
*
*
*
*
*
26,373
4,689
2,712
5,327
13,965
872
91
783
25,650
4,404
2,344
5,274
13,772
847
82
764
23,904
4,160
2,003
4,916
12,857
N.A.
N.A.
N.A.
Savings & Loans**
Total Deposits
NOW
Savings
Time
Savings & Loans**
Total Deposits
NOW
Savings
Time
M o r t g a g e s Outstanding
Mortgage Commitments
Savings & Loans**
Total Deposits
NOW
Savings
Time
M o r t g a g e s Outstanding
Mortgage C o m m i t m e n t s
Savings & Loans**
Total Deposits
NOW
Savings
Time
Mortgages Outstanding
M o r t g a g e Commitments
Savings & Loans**
Total Deposits
NOW
Savings
Time
Mortgages Outstanding
Mortgage Commitments
Savings & Loans**
Total D e p o s i t s
NOW
Savings
Time
M o r t g a g e s Outstanding
Mortgage Commitments
+ 7
+28
+12
+ 8
Savings & Loans**
Total Deposits
NOW
Savings
Time
M o r t g a g e s Outstanding
Mortgage Commitments
+13
+35
+ 9
Total Deposits
NOW
Savings
Time
M o r t g a g e s Outstanding
Mortgage C o m m i t m e n t s
- 2
+ 6
+11
+23
N o t e s : All deposit data are extracted from the Federal Reserve R e p o r t of Transaction A c c o u n t s , other Deposits and V a u l t Cash ( F R 2 9 0 0 ) , and
are reported f o r the average of the week ending the 1st M o n d a y of the m o n t h . This d a t a , reported by institutions with over $25 million in
deposits and $2.4 m i l l i o n of reserve requirements as of June 1 9 8 5 , represents 95% of deposits in the six state a r e a . T h e annual rate of
change is based on m o s t recent data over c o m p a r a b l e y e a r - a g o d a t a . The major differences between this report and the "call report"
are s i z e , the treatment of interbank d e p o s i t s , and the treatment of f l o a t . T h e data generated from the R e p o r t of Transaction A c c o u n t s is f o r
banks over $25 m i l l i o n in deposits as of June 1 9 8 5 . The total deposit data generated from the R e p o r t of Transaction A c c o u n t s
eliminates interbank deposits by reporting the net of deposits "due to" and "due from" other depository i n s t i t u t i o n s . T h e R e p o r t of
Transaction Accounts subtracts cash in process of collection from demand d e p o s i t s , while the call report does n o t . Savings and loan m o r t g a g e
data are f r o m the Federal Home Loan Bank Board Selected Balance S h e e t D a t a . The S o u t h e a s t data represent the total of the six s t a t e s .
Subcategories were chosen on a selective basis and do not add to t o t a l .
* = fewer than four institutions r e p o r t i n g .
N . A . = S&l and credit union deposit data are in the process of being revised d u e to reporting c h a n g e s .
C o m pfor
a r aFRASER
b l e y e a r - a g o data not available at this t i m e .
Digitized
N/C = Data not collected this m o n t h by original s o u r c e .
35
CONSTRUCTION
APR
1986
MAR
1986
APR
1985
ANN.
% .
CHG.
$ Mil.
62,887
8,776
16,058
11,619
2,302
1,104
64.,743
8.,775
16.,487
11.,540
2,,157
i;,087
64.,746
9,,143
16.,123
9.,949
2,,022
1,,147
- 3
- 4
- 1
+17
+14
- 4
Residential Building Permits
Value - $ M i l .
Residential Permits - Thous.
Single-family units
Multifamily units
Total Building Permits
Value - $ M i l .
10,632
1,180
2,512
2,339
390
159
10,906
1,192
2,638
2,291
381
154
9,788
989
2,337
1,992
356
106
+ 6
+19
+ 7
+17
+10
+50
Resident i a l B i m ^ n ^ e r m U s
Value - $ M i l .
Residential Permits - T h o u s .
Single-family units
Multifamily units
Total Building Permits
Value - $ M i l .
Mil.
649
65
167
169
18
17
635
65
155
163
16
17
658
99
101
141
51
7
- 3
-35
+65
+20
-65
+143
Residential B u i I d i n g P e n n U s
Value - $ M i l .
Residential Permits - Thous.
Single-family units
Multifamily units
Total Building Permits
Value - $ M i l .
n d e n t i a l Building Permits - $ Mil.
:al Nonresidential
5,455
Industrial Bldgs.
469
Offices
1,144
Stores
1,238
Hospitals
213
Schools
54
5,,604
492
1.,194
1,,228
218
49
4,,948
557
1,,087
1,,111
161
48
+10
-16
+ 5
+11
+32
+13
Residential Building Permits
Value - $ M i l .
Residential Permits - Thous.
Single-family units
Multifamily units
Total Building Permits
Value - $ M i l .
$ Mil.
2,008
349
528
382
36
21
2,029
332
556
352
38
21
1,834
200
540
310
32
18
+ 9
+75
- 2
+23
+13
+17
Residential Building Permits
Value - $ Mil.
Residential Permits - Thous.
Single-family units
Multifamily units
Total Building Permits
Value - $ M i l .
Mil.
1,051
39
376
222
35
45
1,133
49
405
230
46
46
1,219
47
307
238
72
25
-14
-17
+22
- 7
-51
+80
Residential Building Permits
Value - $ Mil.
Residential Permits - Thous.
Single-family units
Multifamily units
Total Building Permits
Value - $ Mil.
n i.
307
33
65
74
17
7
312
32
67
74
18
7
242
12
43
45
6
3
+27
+175
+51
+64
+183
+133
Residential Building Permits
Value - $ Mil.
Residential Permits - Thous.
Single-family units
Multifamily units
Total Building Permits
Value - $ Mil.
1,194
223
261
245
45
14
886
73
258
146
33
5
+31
+208
-11
+73
+11?
+200
Residential Building Permits
Value - $ M i l .
Residential Permits - Thous.
Single-family units
Multifamily units
Total Building Permits
Value - $ M i l .
12-month cumulative rate
Nonresidential Building Permits
Total Nonresidential
Industrial Bldgs.
Offices
Stores
Hospitals
Schools
Nonresidential Building Permits
Total Nonresidential
Industrial Bldgs.
Offices
Stores
Hospitals
Schools
Nonresidential Building Permits
Total Nonresidential
Industrial Bldgs.
Offices
Stores
Hospitals
Schools
Nonresidential^Building
Total Nonresidential
Industrial Bldgs.
Offices
Stores
Hospitals
Schools
Nonresidential Building Permits
Total Nonresidential
Industrial Bldgs.
Offices
Stores
Hospitals
Schools
Nonresidential Building
Total Nonresidential
Industrial Bldgs.
Offices
Stores
Hospitals
Schools
1,162
225
232
253
70
15
APR
1986
MAR
1986
APR
1985
ANN.
%
CHG.
87,135
85,282
74,701
+17
992.9
782.8
973.3
783.0
900.1
738.2
+10
+ 6
150,022
150,025
139,447
+ 8
15,537
15,414
13,674
+14
203.1
162.5
200.7
167.5
186.1
167.7
+ 9
- 3
30,676
30,424
21,635
+42
595
589
463
+29
9.9
8.5
9.6
8.6
8.8
6.6
+13
+29
1,244
1,224
1,121
+11
8,711
8,715
7,816
+li
106.6
97.1
105.9
101.0
99.6
99.4
+ 7
- 2
12,466
14,319
12,764
+11
3,480
3,411
2,847
+22
49.6
28.3
48.8
28.4
43.5
25.5
+14
+11
5,488
5,440
4,681
+17
7,277
7,326
9,247
-22
11.0
5.3
11.2
5.4
13.2
10.5
-17
-50
1,779
1,865
2,143
-17
352
351
367
- 5
5.8
2.9
5.9
3.0
6.5
3.6
-11
-20
658
662
610
+ 8
1,672
1,615
1,256
+33
20.1
20.4
19.4
20.8
14.7
22.2
+37
- 8
3,044
3,018
2,142
+42
NOTES:
Data supplied by the U . S . Bureau of the Census, Housing Units Authorized By Building Permits and Public Contracts C-40
Nonresidential data excludes the cost of construction for publicly owned buildings. The southeast data represent the total of the si>
states.
36 FRASER
Digitized for
M A Y 1986, E C O N O M I C R E V I E W
m
GENERAL
UNITED STATES
Personal Income
(Sbi1. - S A A R )
T a x a b l e Sales - S b i l .
Plane P a s s . A r r . (000's)
Petroleum P r o d , (thous.)
Consumer Price Index
1967=100
K i l o w a t t Hours - m i l s .
SOUTHEAST
Personal Income
($bi1. - SAAR)
T a x a b l e Sales - Sbil.
Plane P a s s . A r r . (000's)
Petroleum P r o d , (thous.)
Consumer Price Index
1967=100
Kilowatt Hours - mils.
ALABAMA
Personal Income
(Sbi1. - SAAR)
T a x a b l e Sales - S b i l .
Plane P a s s . A r r . (000's)
Petroleum P r o d , (thous.)
Consumer Price Index
1967=100
K i l o w a t t Hours - m i l s .
FLORIDA
~
Personal Income
(Sbil. - SAAR)
Taxable Sales - Sbil.
Plane P a s s . A r r . (000's)
Petroleum P r o d , (thous.)
Consumer Price Index
1967=100
MIAMI
Kilowatt Hours - m i l s .
GEORGIA
Personal Income
(Sbil. - S A A R )
Taxable Sales - S b i l .
Plane P a s s . A r r . (000's)
Petroleum P r o d , (thous.)
Consumer Price Index
1967=100
ATLANTA
Kilowatt Hours - m i l s .
LOUISIANA
Personal Income
(Sbil. - SAAR)
Taxable Sales - S b i l .
Plane P a s s . A r r . (000's)
Petroleum P r o d , (thous.)
Consumer Price Index
1967=100
K i l o w a t t Hours - m i l s .
MISSISSIPPI
Personal Income
(Sbil. - S A A R )
T a x a b l e Sales - S b i l .
Plane P a s s . A r r . (000's)
Petroleum P r o d , (thous.)
Consumer Price Index
1967=100
K i l o w a t t Hours - m i l s .
TENNESSEE
Personal Income
(Sbil. - S A A R )
T a x a b l e Sales - S b i l .
Plane P a s s . A r r . (000's)
Petroleum P r o d , (thous.)
Consumer Price Index
1967=100
Kilowatt Hours - mils.
ANN.
%.
CHG.
LATEST C U R R .
DATA PERIOD
PREV.
PERIOD
YEAR
AGO
3,211.6
N.A.
N.A.
8,790.5
3;,109.7
N.A.
N.A.
9.,031.8
+ 5
MAY
3,268.1
N.A.
N.A.
8,848.0
MAY
MAR
326.2
187.7
325.3
193.2
321.3
184.9
+ 2
+ 2
391.3
N.A.
6,040.4
1,490.0
378.5
N.A.
5.,248.3
1.,517.0
+ 5
APR
MAY
398.3
N.A.
5,268.5
1,417.0
MAR
N.A.
27.9
N.A.
29.8
N.A.
25.3
APR
MAY
43.1
N.A.
134.8
62.0
42.2
N.A.
145.2
59.0
40.9
N.A.
124.9
57.0
+ 8
+ 9
MAR
N.A.
3.7
N.A.
4.2
N.A.
3.7
0
4Q
154.5
151.2
145.5
+ 6
2,699.6
31.0
MAY
173.0
8.1
3,197.2
31.0
MAR
174.5
8.6
2 ,598.2
36.0
MAY
171.0
7.5
+ 4
-14
74.3
N.A.
1,869.8
N.A.
APR
334.9
4.8
73.2
N.A.
2,152.1
N.A.
FEB
336.9
4.6
70.0
N.A.
2 ,019.1
N.A.
APR
324.6
4.4
+ 6
49.4
N.A.
308.8
1,238.0
48.7
N.A.
293.4
1 ,335.0
+ 1
APR
MAY
49.4
N.A.
296.0
1,240.0
MAR
N.A.
4.0
N.A.
4.3
N.A.
4.4
APR
MAY
24.1
N.A.
35.7
84.0
23.2
N.A.
38.5
190.0
23.0
N.A.
34.9
88.0
MAR
N.A.
1.9
N.A.
2.1
N.A.
1.9
52.9
N.A.
232.6
N.A.
52.1
N.A.
198.6
N.A.
50.5
N.A.
177.8
N.A.
N.A.
5.5
N.A.
6.1
N.A.
6.2
4Q
4Q
4Q
APR
MAY
MAR
4Q
APR
MAR
4Q
4Q
4Q
APR
MAR
+ 2
+ 0
- 7
+10
+ 5
+ 1
+ 8
- 7
+ 3
+ 9
+ 1
- 7
- 9
+ 5
+ 2
- 5
0
+ 5
+31
-11
ANN.
MAY
%
1985 C H G .
MAY
1986
APR (R)
1986
Agriculture
Prices Rec'd by Farmers
Index (1977=100)
124
Broiler Placements (thous.)
85,331
Calf Prices ($ per c w t . )
58.40
Broiler Prices ({ per lb.)
30.90
Soybean P r i c e s (S per bu.)
5.18
Broiler Feed Cost ($ per ton) (Q2)189
121
84,740
58.90
29.90
5.22
(Q1J189
130
90,266
65.60
30.00
5.70
(Q2)204
- 5
- 5
-11
+ 3
- 9
- 7
114
35,525
54.84
29.78
5.28
181
109
35,386
55.03
28.17
5.27
181
123
34,965
60.91
27.81
5.84
204
- 7
+ 2
-10
+ 7
-10
-11
Agriculture
Farm Cash Receipts - S m i l .
(Dates: F E 8 , F E B )
298
Broiler Placements (thous.)
12,186
54.50
Calf Prices (S per c w t . )
Broiler Prices (t per lb.)
30.00
Soybean Prices (S per bu.)
5.33
Broiler Feed Cost ($ per ton)
181
-
11,930
54.20
28.00
5.20
179
274
11,774
59.20
28.00
5.89
195
+ 9
+ 3
- 8
+ 7
-10
- 7
Agriculture
Prices Rec'd by Farmers
Index (1977=100)
Broiler Placements (thous.)
Calf Prices ($ per cwt.)
Broiler Prices (t per lb.)
Soybean Prices (S per bu.)
Broiler Feed Cost ($ per t o n )
Agriculture
Farm Cash Receipts - S m i l .
(Dates: F E B , FEB)
Broiler Placements (thous.)
Calf P r i c e s ($ per c w t . )
Broiler Prices ($ per lb.)
Soybean Prices (S per b u . )
Broiler Feed Cost ($ per ton)
••
777
2,349
62.50
29.00
5.33
181
2,388
57.10
27.00
5.20
230
Agriculture
Farm Cash Receipts - $ m i l .
(Dates: F E B , FEB)
376
Broiler Placements (thous.)
14,230
Calf P r i c e s ($ per c w t . )
49.60
Broiler Prices ({ per lb.)
29.00
5.20
Soybean P r i c e s (S per bu.)
Broiler Feed Cost ($ per ton)
181
14,308
51.90
27.50
5.23
180
-
-
Agriculture
Farm Cash Receipts
- S mil
(Dates: F E B , FEB)
Broiler P l a c e m e n t s (thous.)
Calf Prices (S per c w t . )
Broiler P r i c e s (i per lb.)
Soybean P r i c e s (S per bu.)
Broiler Feed Cost ($ per ton)
310
N.A.
54.00
31.00
5.22
181
N.A.
55.10
29.50
5.23
245
Agriculture
Farm Cash Receipts - $ m i l .
(Dates: F E B , FEB)
Broiler Placements (thous.)
Calf P r i c e s ($ per cwt.)
Broiler P r i c e s (£ per lb.)
Soybean Prices (S per bu.)
Broiler Feed C o s t (S per ton)
341
6,760
55.60
31.20
5.23
181
6,760
55.60
30.10
5.28
157
Agriculture
Farm Cash Receipts - $ m i l .
(Dates: F E B , F E B )
Broiler Placements (thous.)
Calf Prices (S per c w t . )
Broiler Prices (i per lb.)
Soybean Prices (S per bu.)
Broiler Feed Cost (S per ton)
276
N.A.
51.50
27.50
5.41
189
N.A.
55.40
26.00
5.41
176
-
-
-
721
2,174
65.30
27.00
5.89
235
+ 8
+ 8
- 4
+ 7
-10
-23
393
14,216
59.10
27.00
5.83
225
- 4
+ 0
-16
+ 7
-11
-20
264
N.A.
62.50
30.00
5.58
250
+17
414
6,801
59.70
29.00
6.04
160
-18
- 1
- 7
+ 8
-13
+13
348
N.A.
59.70
26.00
5.80
183
-21
-14
+ 3
- 6
-28
-14
+ 6
- 7
+ 3
NOTES:
Personal Income data supplied by U . S . Department of C o m m e r c e . T a x a b l e Sales are reported as a 12-month cumulative total. Plane
P a s s e n g e r A r r i v a l s are collected from 26 a i r p o r t s . Petroleum Production data supplied by U . S . Bureau of M i n e s . Consumer Price Index data
supplied by Bureau of Labor S t a t i s t i c s . A g r i c u l t u r e data supplied by U . S . D e p a r t m e n t of A g r i c u l t u r e . Farm Cash Receipts data are reported
as c u m u l a t i v e for the calendar year through the month s h o w n . Broiler placements are an average w e e k l y r a t e . T h e Southeast data represent
the total of the six s t a t e s . N . A . = not a v a i l a b l e . T h e annual percent change calculation is based on m o s t recent data over prior y e a r .
R = revised.
RESERVE BANK O F ATLANTA
Digitized forFEDERAL
FRASER
37
EMPLOYMENT
1986
MAR
1986
APR
1985
Civilian Labor Force - t h o u s .
Total Employed - thous
Total Uemployed - t h o u s .
Unemployment R a t e - % SA
Insured Unemployment - t h o u s .
Insured U n e m p l . Rate - %
M f g . A v g . W k l y . Hours
Mfg. Avg. Wkly. Earn. - S
116,317
108,201
8,115
7.2
N.A.
N.A.
40.5
392
116,309
107,643
8,667
7.2
N.A.
N.A.
40.7
395
114,325
106,175
8,150
7.3
N.A.
N.A.
40.1
380
Civilian Labor Force - t h o u s .
Total Employed - thous
Total Uemployed - t h o u s .
Unemployment R a t e - % SA
Insured U n e m p l o y m e n t - t h o u s .
Insured U n e m p l . Rate - %
M f g . A v g . W k l y . Hours
Mfg. Avg. Wkly. Earn. - $
15,637
14,448
1,188
7.9
N.A.
N.A.
41.2
357
15,587
14,350
1,237
7.8
N.A.
N.A.
41.1
355
15,255
14,087
1,164
8.0
N.A.
N.A.
40.6
338
"flip®
1,697
109
9.4
N.A.
N.A.
40.7
357
1,673
179
8.9
N.A.
N.A.
40.6
355
B
Ü
M
l
H
H
Civilian Labor Force - t h o u s .
Total Employed - thous
Total Uemployed - t h o u s .
Unemployment Rate - % SA
Insured Unemployment - t h o u s .
Insured U n e m p l . Rate - %
M f g . A v g . Wkly. Hours
Mfg. Avg. Wkly. Earn. - $
5,445
1,697
169
9.4
N.A.
N.A.
40.7
355
5,439
1,673
179
8.9
N.A.
N.A.
40.6
356
Civilian Labor Force - t h o u s .
Total Employed - thous
Total Uemployed - thous.
Unemployment Rate - % SA
Insured Unemployment - t h o u s .
Insured U n e m p l . Rate - %
M f g . A v g . W k l y . Hours
Mfg. Avg. Wkly. Earn. - $
2,934
2,774
160
5.8
N.A.
N.A.
44.1
381
1,981
1,719
262
13.2
N.A.
N.A.
40.8
430
Total Employed - thous
Total Uemployed - t h o u s .
Unemployment Rate - % SA
Insured Unemploynent - t h o u s .
Insured U n e m p l . R a t e - %
M f g . A v g . W k l y . Hours
Mfg. Avg. Wkly. Earn. - $
Total Employed - thous
Total Uemployed - t h o u s .
Unemployment R a t e - % SA
Insured Unemployment - t h o u s .
Insured U n e m p l . R a t e - %
M f g . A v g . W k l y . Hours
Mfg. Avg. Wkly. Earn. - $
Total Employed - thous
Total Uemployed - t h o u s .
Unemployment R a t e - % SA
Insured Unemployment - t h o u s .
Insured U n e m p l . R a t e - %
M f g . A v g . W k l y . Hours
Mfg. Avg. Wkly. Earn. - $
Total Employed - thous
Total Uemployed - t h o u s .
Unemployment R a t e - % SA
Insured Unemployment - t h o u s .
Insured U n e m p l . Rate - %
M f g . A v g . W k l y . Hours
Mfg. Avg. Wkly. Earn. - S
NOTES:
" f l P
ANN.
% .
CHG
2
+ 2
2
+
-
+ 1
+ 3
+ 3
+ 3
+ 2
+ 1
+ 6
1,642
153
8.5
N.A.
N.A.
40.8
338
+ 3
+10
5,293
1,642
153
8.5
N.A.
N.A.
40.8
345
+ 3
+ 3
+10
2,921
2,753
167
6.2
N.A.
N.A.
43.2
370
2,852
2,682
-170
6.7
N.A.
N.A.
40.9
319
—
+ 3
+ 4
- 6
1,983
1,721
262
13.2
N.A.
N.A.
41.1
435
1,982
1,754
228
11.4
N.A.
N.A.
41.4
344
1,021
124
11.2
N.A.
N.A.
40.2
430
1,010
1.27
10.6
N.A.
N.A.
40.4
435
—
998
113
10.1
N.A.
N.A.
39.8
344
2,266
2,087
177
7.9
N.A.
N.A.
41.0
346
2,256
2,069
187
7.5
N.A.
N.A.
41.1
342
2,222
2,041
177
8.0
N.A.
N.A.
40.8
332
— r—
"fHfP®
.
APR
- 2
+ 3
+ 8
+19
- 2
+15
- 2
+25
lipasi
+ 2
+10
+ 1
+25
+ 2
0
+ 0
+ 4
M
APR
1985
ANN.
%
CHG
Nonfarm Employment - thous.
Manufacturing
Construction
Trade
Government
Services
Fin., Ins. & Real. Est.
Trans. Com. & Pub. Util.
99,817
19,269
4,751
23,715
16,865
22,873
6,180
5,309
98,922
19,262
4,471
22,818
16,871
21,630
6,118
5,282
96,909
19,375
4,451
22,797
16,475
21,766
5,833
5,243
+
+
+
+
+
+
+
1
7
4
2
5
6
1
Nonfarm Employment - t h o u s .
Manufacturing
Construction
Trade
Government
Services
Fin., Ins. & R e a l . Est.
Trans. Com. & Pub. Util.
13,010
2,310
771
3,232
2,303
2,749
750
717
12,956
2,303
770
3,221
2,301
2,741
747
720
12,6/1
2,311
759
3,132
2,251
2,647
718
721
+
+
+
+
+
+
-
3
1
2
3
3
4
4
1
347
70
307
306
246
68
71
1,423
358
71
301
299
242
65
72
+
+
+
+
-
4,565
521
337
1,234
697
1,192
330
244
4,569
522
337
1,235
700
1,193
329
244
4,420
513
330
1,193
684
1,134
312
243
+
+
+
+
+
+
+
+
3
2
2
3
2
5
6
0
Nonfarm Employment; - t h o u s . 2,613
Manufacturing
557
Construction
150
Trade
663
Government
457
474
Services
Fin., Ins. & Real. Est.
141
163
Trans. Com. & Pub. Util.
2,607
562
149
659
457
469
140
164
' • • "
2,549
552
142
639
452
461
135
161
+
+
+
+
+
+
+
+
3
1
6
4
1
3
4
1
^ 7 5 5 3
168
98
384
328
322
85
112
• • • • • •
1^598
180
106
383
329
320
85
115
- 7
-13
0
0
0
+ 1
thous.
Manufacturing
Construction
Trade
Government
Services
Fin., Ins. & Real. Est.
Trans. Com. & Pub. Util.
- 2
+ 6
MAR
1986
• ¡ • H
T7444
356
71
309
307
248
68
71
...
Nonfarm Employment - t h o u s .
Manufacturing
Construction
Trade
Government
Services
Fin., Ins. & Real. Est.
Trans. Com. & Pub. Util.
fhousP^T
Manufacturing
Construction
Trade
Government
Services
Fin., Ins. & Real. Est.
Trans. Com. & Pub. Util.
^^lorffSmÌSB^BP^
Manufacturing
Construction
Trade
Government
Services
Fin., Ins. & Real. E s t .
Trans. Com. & Pub. Util.
^^^SfSPRPÜyH!?^
Manufacturing
Construction
Trade
Government
Services
Fin., Ins. & Real. Est.
Trans. Com. & Pub. Util.
H
T
169
93
383
329
320
86
108
^ M I S "
900
222
35
179
193
134
37
39
834
220
35
175
190
131
35
39
^899
483
81
457
318
376
87
90
1,847
489
76
441
298
359
87
90
221
36
181
193
136
37
39
I
1918
486
83
462
320
380
88
91
-
- 1
0
3
3
2
5
2
+7
+
+
+
+
+
+
0
3
3
2
4
6
0
+
+
+
+
+
+
1
9
5
7
6
1
1
I
All labor force dara are from Bureau of Labor Statistics reports supplied by state a g e n c i e s .
O n l y the unemployment rate data are seasonally a d j u s t e d .
T h e Southeast data represent the total of the six s t a t e s .
38
MAY 1986, E C O N O M I C R E V I E W
Federal Reserve Bank of Atlanta
104 Marietta St, N.W.
Atlanta, Georgia 30303-2713
Address Correction Requested
Bulk Rate
U.S. Postage
PAID
Atlanta, Ga.
Permit 292
@FedFRASER