Full text of Economic Review (Federal Reserve Bank of Cleveland) : November 1966

View original document

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

NOVEMBER 1966

THIS

ISSUE

A Note on Interpreting
Monetary Variables

. 2

Regional Trends
in Steel Production . . 9

FEDERAL RESERVE

B A N K OF C L E V E L A N D

ECONOMIC REVIEW

A NOTE ON INTERPRETING
MONETARY VARIABLES
As the nation's central bank, the Federal

of monetary policy. One dominant ch aracter

Reserve System has responsibility for manag

istic pervades all these discussions—there is

ing the monetary system in a way that helps

complete agreement on very few things. What

achieve the broad goals of economic policy.

this implies is that the state of economic

These goals are well known and widely a c

knowledge, while steadily advancing, has

cepted: economic growth (maximum pro

not come sufficiently far to provide answers

duction and maximum employment), stable

to some basic questions.

prices, and reasonable balance in the nation's
international transactions.

SOME MONETARY VARIABLES

While the general nature of the role of the

It is in this spirit that this article briefly

Federal Reserve in monetary management is

discusses some of the general problems in

not difficult to explain, it is difficult to explain

volved in interpreting monetary statistics,

the specifics of how that role should be per
formed, for example, how monetary policy

that is, the behavior of the variables that the
Federal Reserve is generally considered to be

should be designed, how the variables to be

able to influence, directly or indirectly. Mon

influenced should be selected, and how the

etary statistics are difficult to interpret be

results should be measured. Thus, it is not

cause there is no precise and complete specifi

surprising that economic literature is replete

cation of either (1) which monetary variable (s)

with academ ic and nonacademic, technical

is (are) most important in terms of the ultimate

and n o n te ch n ica l, q u an titative and non-

effect on policy goals such as economic growth,

quantitative (qualitative) discussions on the

stable prices, and balance of payments equi

specifics of monetary policy—ranging from

librium, or (2) the actual process (linkages)

issues involving definitions and measurement

by which change (s) in a monetary variable is

to issues concerned with the policymaking

(are) transmitted to ultimate goals. In addition

process, the implementation of policy, the

to these seemingly overwhelming consider

variables to be affected, and the ultimate goals

ations, there are some practical problems

NOVEMBER 1 9 66
feel that total reserves (nonborrowed plus

that compound the task of interpreting shortrun changes in monetary variables.

borrowed reserves) are a better measure (or

From the standpoint of Federal Reserve in

target) to be used when evaluating or im

fluence, a host of monetary variables have
been suggested as being the most important,

plementing monetary policy.1
There is similar lack of agreement beyond

either to be used when evaluating monetary

the reserve measures. For example, there is

policy or to be relied on when implementing

the fundamental question of how responses to

policy. For example, since one of the major

changes in monetary policy—to the reserve

instruments of monetary policy is open market

variables—are to be measured. In the case of

operations, that is, the purchase or sale of

banks, for example, should policy be measured

Government securities which adds to or

in terms of total deposit growth, or on the

subtracts from bank reserves, some economists

other side of the balance sheet—by bank

have suggested that the amount of reserves

credit growth? Is a particular type of deposit,

supplied (or withdrawn) through open market

such as demand deposits (or more broadly,

purchases (sales) of securities should be the

demand deposits and currency—money sup

major measure or target of monetary policy.

ply) the relevant magnitude, or is a particular

This is so perhaps because the Federal Reserve

type of bank credit such as business loans the

is able to offset by open market transactions,

key variable? Are

the other factors that affect bank reserves, and

availability and cost of credit—the important

whether the Federal Reserve supplies or does
not supply reserves, it is suggested, can be

variable as opposed to the quantity of bank
credit (or total credit)? In this connection,

construed to represent the intent of monetary

interest rates, which are an element of credit

policy.

(Reserves

supplied

through

credit conditions—the

open

conditions, are considered by a number of

market operations—assuming constant re

analysts to be a significant variable, in terms

serve requirements—are identified as non
borrowed reserves.) At the same time, banks

of both evaluating and implementing mon
etary policy. It goes without saying that be

may obtain borrowed reserves, under certain

yond the foregoing ''intermediate'' stage of

circumstances, by discounting or securing

the monetary process, there is as much lack

advances from the Reserve banks, that is, by

of agreement, if not more, in evaluating the

borrowing. Some analysts feel that the level

influence of the monetary variables on broader

of borrowings, since it presumably reflects

objectives and goals, to say nothing of select

the extent of pressure on banks to make re

ing the variable(s) to watch.

serve adjustments, should be taken as the

Ironically, even if there were clear-cut and

major measure or target of monetary policy.

definitive evidence of the appropriate mone

Still others feel that the net reserve position

tary variable (s) to control and watch, there

of banks (excess reserves less member bank

would still be a fundamental and practical

borrowings) is a better statistic to consider

1 The m easures mentioned are intended only to suggest

when evaluating or implementing monetary

the variety of measures available and the lack of agreem ent

policy. On the other hand, some economists

as to which is (are) the most significant m easure(s).

E C O N O M IC

R E V IE W

problem involved in the presentation, use,

First
Second
Quarter Quarter
1966
1966

and interpretation of basic statistical infor
mation—especially for assessments of shortrun changes in monetary variables with re
spect to desired policy changes. In this article,
for purposes of illustration, attention is pri
marily devoted to some of the more basic
monetary or reserve variables.

MEASURING MONETARY VARIABLES
There are many ways to measure monetary
or reserve variables, and each method or

Based on quarterly averages
of monthly averages of daily
figures:
(1) Annual rate of change
from preceding quarter
(2) Annual rate of change
from same quarter in
preceding year . . .
Based on monthly averages of daily
figures:
(3) Change from three months
earlier, at annual rate:
December 1965 to
March 1966 and . . . .
March 1966 to
June 1966 .....................

+ 6 .0 %

+ 5 .5 %

+ 5 .3 %

+ 6 .0 %

+ 5 .0 %
+ 4 .3 %

approach usually yields different results. For
example, data that are used may be on a daily,
weekly, monthly, quarterly, or annual basis.

Perhaps the most perplexing aspect of the

W hen the appropriate time unit is decided

above figures is not the different rates of

upon, it must be decided whether the data

change generated by the different methods,

should be analyzed on the basis of levels,

but the fact that two computations (1 and 3)

absolute changes, or rates of change. And,

indicate a deceleration in the rate of growth

then, it is necessary to determine the appro

during the second quarter while the third (2)

priate time period over which to consider

indicates an acceleration. Since the same

levels or changes. In using rates of change,

variable (money supply) is being measured,

which are based on terminal values, it should

the differences obvioulsy reflect the fact that

be remembered that both the time span and

the computations are based on different time

the unit of the terminal period (for example,

spans and different average lengths of the

week, month, or quarter) will influence the

terminal periods. At this point, a question can

resultant computed rate of change. Changing

legitimately be asked, which set of statistics

either the time span or the terminal units (for

is the most meaningful in terms of interpret

example, to a subsequent period such as

ing the behavior of the money supply in re

from April to May, or to a quarterly as opposed

lation to an evaluation of monetary policy?

to a monthly average) can produce substan
tial differences in rates of change based on
those terminal values.

In answering this question, the figures need
to be put into a more proper perspective. For

As an illustration, the following rates of

example, the money supply (quarterly aver

increase can be associated with the conven-

ages based on monthly averages of daily

tionally-defined money supply for the first

figures) increased substantially in both the

two quarters of 1 9 6 6 :2

third and fourth quarters of 1965; therefore,
the money supply for the first two quarters of

2 These and subsequent figures are seasonally adjusted.

1966, even with no further increases during

NOVEMBER 1966
those two quarters, would still show year-overyear increases from the first two quarters of

both January and March 1966, the rate of in
crease from December to March became
quite substantial even though it was less than
the increase in the preceding period. As

Changes in the Money Supply
from the Preceding Quarter

another example, growth of the money supply

(annual rates of growth)
1965
1st. Q.
+ 2 .0

2nd. Q.
+ 3 .0

Changes in the Money Supply
Between Last Months of the Quarters

1966

3rd. Q.
+ 5 .0

4th. Q.
+ 6 .9

1st. Q.

+ 6.0

(at annual rates)

2nd. Q.
+ 5 .5

19 6 5 (3.8 percent and 3 .0 percent, respec
tively). Since the money supply actually in
creased substantially in the first two quarters

1965
1st. Q.
+ 1.5
(12/643/65)

2nd. Q. 3rd. Q.
+ 3 .5
+ 5 .9
(3/65- (6/656/65) 9/65)

1966
4th. Q.
+ 7 .6
(9/6512/65)

1st. Q. 2nd. Q.
+ 5 .0
+ 4 .3
(12/65- (3/663/66)
6/66)

of 1966, year-over-year gains from 1 9 6 5 also

in April 1 9 6 6 was almost as great as the

were substantial.

December 1 9 6 5 increase; that growth in fact
accounted for almost all of the net increase

Actual Year-Over-Year Rates of Gain
in the Money Supply
1965
1st. Q.
+ 4 .0

2nd. Q.
+ 3 .9

in the second quarter since money supply
1966

3rd. Q.
+ 3 .7

4th. Q.
+ 4 .3

1st. Q.
+ 5 .3

changes in May and June were virtually off

2nd. Q.

+ 6.0

setting. The March to June growth, however,
was still below the December to March growth.

The change from three months earlier (be
tween the last months of the quarters) re

PROBLEMS OF MONTHLY DATA

flects the peculiarities of the terminal months

Generally, the shorter the time span of

involved. For example, the substantial in

basic data, the more volatile rates of change

crease in the money supply that occurred in

are likely to be (due of course to the behavior

December 1 9 6 5 contributed substantially to

July

Aug.

1965
Sept.
Oct.

+ 5 .2

+ 4 .4

+ 8 .1

Jan.
+ 5 .7

Feb.

Mar.

— 1.4 + 7 .8

+ 8 .0

Nov.

Dec.

+ 2 .9

+ 1 1 .6

1966
Apr.
May
+ 1 1 .3

— 4.9

June
+

6.3

of the underlying series). Thus, while monthly
rates of change in various monetary variables

Monthly Changes in the Money Supply
at Annual Rates

(expressed as annual rates) are helpful in
indicating the pace and direction of short-run
changes, the nature of the variables under
consideration is such that monthly rates of
July

— 10.5

Aug.

0.0

change will exhibit considerable volatility,
to the extent that their underlying significance
is indeterminable.

a high rate of growth for the fourth quarter as

This is clearly revealed in Table I, where a

a whole. Any rate of increase based on Decem

number of monetary variables are presented

ber would thus be expected to exhibit mod

for purposes of illustration. Putting the question

erate increases. However, since the money

most simply, how should monetary policy be

supply actually increased considerably in

evaluated during the period shown in the

ECONOMIC REVIEW
TABLE I
Changes in Selected Monetary Variables
Percent Change From Preceding Month at Annual Rates
1965
December
Total Reserves . . .
Nonborrowed Reserves
Money Supply . . .
Bank Credit . . . .

+
+
+
+

1 4.5%
10.5
1 1.6
11.9

1966
January
+ 6.1%
+ 9.0
+ 5.7
+ 12.2

February
+ 4 .3 %
+ 3.1
— 1.4
+ 0 .4

March
+
—
+
+

April

2 .4 %
5.2
7.8
11.3

+ 13 .5 %
+ 11.1
+ 11.3
+ 9.6 p

May

June

+ 0 .1 %
— 0.3
— 4.9
+ 6 .3 p

— 0 .7 %
— 1.8
+ 6 .3
+ 4 .3 p*

July
+
+
—
+

8 .9 %
4.4
10.5
n -0 p

August
— 1 3 .7 %
— 8.8
0.0
+ 6.2 p

p Preliminary.
* Beginning June 9, about $1.1 billion of balances accumulated for payment of personal loans were deducted as a result
of a change in Federal Reserve regulations.
NOTE: Data are monthly averages of daily figures, except for bank credit data which are for the last W ednesday of the month.
All the series are seasonally adjusted.
Sources: Board of Governors of the Federal Reserve System and Federal Reserve Bank of Cleveland

table? It is apparent from the table that the

and conversely. While some of the monthly

growth of total reserves decelerated from

fluctuations were due to technical factors, for

December through March, and then fluctu

example, changes in seasonal patterns associ

ated widely in subsequent months, including

ated with shifts in U. S. Treasury deposits,

some

there is still a large element of unexplained

serves increased at a slackening pace early

variation in monthly values. Bank credit also

in 1966, declined in March, and then also

exhibited large fluctuations in the period
shown in the table although all changes were

negative

values.

Nonborrowed

behaved erratically.

positive. (Monthly swings in bank credit were
As long as successive changes are in the

also influenced by shifts in U. S. Treasury

same direction, they are relatively easy to

deposits, particularly as the latter affected

interpret; if they are not in the same direction,

seasonal adjustment of the bank credit data.)

that is, if monthly rates of change alternate

Despite the fact that monthly changes may

between positive and negative changes, they

suggest, over some time span, whether a vari

become extremely difficult to evaluate. A

able is rising, falling, or remaining about the

decline (negative value) in one month fol

same, it is often impossible to adjudge the

lowed by an increase (positive value) in the

underlying movements of the particular series,

next month raises a question as to the net rate

especially since monthly changes often offset

of change over the two months, since at least

one another. Accordingly, short-term mea

a portion of the increase is offset by the de

sures should be supplemented by longer run

cline in the previous month.

measures, for example, averaging or looking

Such a situation is apparent in the behavior
of the money supply in the time period covered
in Table I; month-to-month changes in the

for periods of sustained change.

USE OF LONGER TIME PERIODS

money supply were substantial, with large

Unfortunately, it is easy to run into similar

increases being followed by large decreases,

problems when selecting longer run mea-

NOVEMBER 1966

would seem to be nothing particularly sig

sures. For example, on a p r io r i grounds, there

Table I are presented in Table II. Year-to-year
changes exhibit more stability than monthly

nificant or sacrosanct about calendar quarters,

changes

fiscal years, or other contrived time periods

smoothed out over the longer time period.

in interpreting movements in time series

The

covering monetary variables—or any other

changes makes it easier to discern longer

since

greater

short-run

stability

fluctuations

of the

are

year-to-year

economic time series for that matter. Perhaps

term patterns. It is noticeable that for total

a three-month moving average would be ap

and nonborrowed reserves, and to some ex

propriate so that there would be a "quarterly"

tent bank credit, the growth rates tend to be

average for each month of the year. Perhaps

come lower as the year progresses. The rates

not.

of change still have to be interpreted with

On the other hand, and merely as an illus

care since they are based on two arbitrarily

tration, another measure that could be used

selected values, and do not reflect the inter

is year-over-year changes for each month of

vening patterns. Nevertheless, a sequence of

the year (the change in a variable—absolute

year-to-year changes may be helpful as a sup

or percentage ch an ge—from a given month

plement to short-run measures in assessing

in the preceding year to the same month in

changes in monetary variables.

the current year). Such a measure appears to

Use of year-to-year percentage changes

provide continuous year-to-year perspective

could overcome many seasonal adjustment

for each month, that is, a longer term per

problems. Since changes would be between

spective to supplement shorter term measures.

the same month (or quarter) in respective

Moreover, it is not tied to calendar or fiscal

years, the basic data could be used in un

year periods, which have little economic

adjusted form and would not have to be sea

significance.

sonally adjusted by procedures that are fre

Year-to-year changes, in percentage terms,

quently open to question. Problems would

for the selected monetary variables used in

still arise, however, where seasonal patterns

T A B LE II

Changes in Selected Monetary Variables
Percent Change From Same Month in Preceding Year
1965

Total Reserves . . .
Nonborrowed Reserves
Money Supply . . .
Bank Credit . . . .

1966

December

January

February

March

April

May

June

July

+ 5.3%
+ 4.3
+ 4.7
+ 10.2

+ 5.3%
+ 4.9
+ 5.2
+ 10.3

+ 4 .7 %
+ 5.4
+ 5.3
+ 9.3

+ 4 .4 %
+ 3.8
+ 5.6
+ 8.9

+ 4 .8 %
+ 4 .1
+ 6.1
+ 9.3 p

+ 4 .8 %
+ 3.9
+ 5.9
+ 8.9 p

+ 3.9%
+ 3.4
+ 5 .8
+ 8.4 p*

+ 4 .3 %
+ 3.4
+ 4 .4
+ 8.8 p

August
+
+
+
+

3.2%
2.5
4.Op
8.3 p

p Preliminary.
* Beginning June 9, about $1.1 billion of balances accumulated for payment of personal loans were deducted as a result
of a change in Federal Reserve regulations.
NOTE: Data are monthly averages of daily figures, except for bank credit data which are for the last Wednesday of the month.
All the series are seasonally adjusted.
Sources: Board of Governors of the Federal Reserve System and Federal Reserve Bank of Cleveland

ECONOMIC REVIEW
are changing rather rapidly; such a case, for

uid assets?). The reaction time could also be

example, would be the changing seasonal

expected to vary according to the type and

pattern of U. S. Government deposits as a

size of bank in that, for example, larger money

result of accelerated tax payments.

market banks would be expected to respond
more rapidly than smaller banks that are out

The problems involved in measuring and

side the central money market. Unfortunately,

interpreting short-run changes really reflect

not enough is known about the nature and

more fundamental difficulties. As suggested

speed of the response pattern of banks to

earlier, the fundamental problem involves the

determine any quantifiably precise pattern.

question of what type of change really con

In the absence of knowledge about short-

stitutes an economically meaningful change.

run response patterns, it appears that several

In the present context, how long must a given

different measures are necessary to interpret

change in various reserve aggregates (for

changes in monetary variables. Reliance on

example, total or nonborrowed reserves) per

any one measure may give rise to incomplete

sist before a bank (the banking system) re

or false interpretations. Longer term mea

sponds by expanding deposits and credit, or

sures should be used to supplement short

how large a change must actually occur; more

term measures so that basic patterns may be

than likely, the larger the change, given other

discerned, and so that the intervening con

factors, the shorter the reaction time on the

tours giving rise to longer term changes can

part of the banks. In other words, there is
probably a variable reaction time on the part

be determined.
Finally, since persistent changes over a

of the banks depending on the duration and

longer period of time, say six months to a

magnitude of the change as well as expecta

year, take on a permanent quality that over

tions concerning future changes (is the change

rides shorter term variations, longer term

likely to be permanent, and what will be the

measures of changes in monetary variables

effects on the level of interest rates, the de

may be particularly useful in making any

mand for bank funds, and the volume of liq

judgment about monetary policy.

http://fraser.stlouisfed.org/
8
Federal Reserve Bank of St. Louis

NOVEMBER 19 66

REGIONAL TRENDS
IN STEEL PRODUCTION
Four of the nation's 11 steel producing

chart—has been due to the fact that the com

centers are located in the Fourth Federal

bined output of mills in the District posted an

Reserve District. These four centers produce

infinitesimally small average annual rate of

a substantial portion of the nation's steel,

increase (0.1 percent) as compared with that

accounting for more than a third (50 million

for the U. S. as a whole (1.5 percent) during

tons) of the nation's record 131 million ton

the 1 9 4 7-65 period.1 Indicative of the de

steel output in 1965. Nevertheless, steel out

clining importance of the District in total steel

put of Fourth District centers accounts for a

output is the fact that the combined output of

smaller share of the U. S. total than earlier,

District centers in 1 9 6 5 —even after three

even though steel production in the District

years of rapidly expanding production—was

has increased appreciably since 1961, a de
velopment that has served to interrupt the

hardly more than it had been in 1955, the
previous all-time high (see Table I). In con

steady attrition in the District's share of total

trast, steel output in the U. S. in 1 9 6 5 (and

output. The current share of total steel out

1964) had gone much beyond the previous

put accounted for by mills in the District

high (1955). As can be seen from Table I,

represents a significant loss of position as

the nature of the smaller contribution of Dis

compared with, say, 1 9 4 7 when the Fourth

trict output—as well as its poor growth re c

District produced almost half of the nation's

ord—is pointed up by the large number of

steel. In short, during the postwar period the

years since 1 9 4 8 (eight) that annual steel

Fourth District, in relative terms, has declined
as a steel producing region.
The relative loss of position is reflected in

1 The average annual rates of ch an g e used in this article
are derived by using the compound interest formula
based on logarithms of the data (log Y = log A + [log

the gradual but persistent widening of the

B] X). The use of average annual rates of ch an ge allows

gap between total steel production in the

the first and final years of the overall time period to in

nation and the combined output of centers in
the Fourth District. The widening gap be
tween the respective series—as shown in the

fluence the statistical results to the sam e extent as any
interim year. Thus, the first and final years do not, by
virtue of their position, determine either the rate of
ch ange or the direction of trend.

STEEL INGOT PRODUCTION
U.S. and Major G e o g r a p h i c Centers

INDEX 1957-59=100
180

UNITED STATES

DETROIT

IN DEX 1957-59=100

M i l l i o n s o f t ons

140

1947 - 3.12
1965 - 9.66

M i ll i o n s o f tons

1947 - 8 4 .8 9
1965 - 131.19

FOURTH FEDERAL
RESERVE DISTRICT

WESTERN

M i ll i o n s o f tons

194 7
1965

M i l l i o n s o f t on s

40.19
5 0 .2 9

1947 - 4.33
1965 - 8.41

OTHER THAN
FOURTH DISTRICT

SOUTHERN

M i l l i o n s o f t on s

79 4 7 - 4 4 . 6 0
1965 - 8 0 .8 9

M i l l i o n s o f t on s

1947 - 4.01
1965 - 7.70

CHICAGO
M i l l i o n s of t o ns

1965 - 2 6 .3 9

1947 - 4.21
1965 - 6.73

PITTSBURGH

CLEVELAND

M illio n s o f tons

1947
1965

22 .30
25 .9 7

M i l l i o n s o f tons

194 7 - 4.01
1965 - 7.11

NORTHEAST COAST
CINCINNATI

M i llio n s o f tons

1 9 4 7 - 10.31
1965 - 18.20
M i l l i o n s o f t ons

1947
1965

2.71
6 .24

YOUNGSTOWN

M illio n s o f tons

194 7
1965

ST. LOUIS

11.17
11.35

M i l l i o n s o f t ons

194 7 - 1.45
1965 - 3 .42

A v e r a g e a n n u a l rate of growth
NOTE:

G e o g r a p h i c c e n t e r s r a n k e d a c c o r d i n g to i n g o t o ut p u t
in 1 9 6 5 .

19 4 7

S o u r c e s o f d a t a : A m e r i c a n I r o n a n d S t e e l I ns ti tute a n d F e d e r a l R e s e r v e B a n k of C l e v e l a n d

TABLE I
Steel Ingot Production, 1947-65
U. S. and Major Steel Producing Centers
(in millions of tons)

United States

1947

1948

1949

1950

1951

1952

1953

1954

84.89

88.64

77.98

96.84

105.20

93.17

111.61

88.31

956

1957

1958

1959

1960

1961

1962

1963

1964

1965

+ 1.5%

112.72

85.26

93.45

99.28

98.01

98.33

109.26

1 27.08

131.19

49.35 48.53

45.09

32.56

38.34

37.85

36.43

37.07

41.24

48.89

50.29

+ 0.1

51.95

67.64 66.41

67.53

52.76

55.13

61.18

61.63

61.37

67.97

77.49

80.89

+ 2.5

22.04

18.92

23.60 22.63

22.24

18.34

17.94

20.68

21.07

23.02

25.94

26.39

+ 1.8

26.72

19.99

26.30 25.67

24.83

18.31

19.99

19.15

19.57

21.67

25.25

25.97

— 0.3

11.70

16.17 16.44

16.30

12.40

13.10

14.35

14.07

13.65

14.77

17.69

18.20

+ 2.4

12.67 12.32

Fourth District

40.19

41.27

35.36

43.95

48.02

42.35

49.14

36.47

Other Than Fourth District

44.60

47.09

42.46

52.94

57.18

50.71

62.37

Chicago

17.16

17.75

15.60

19.12

20.48

17.55

26.30

23.36

Pittsburgh

22.30

22.94

19.78

24.20

Northeast Coast

10.31

11.01

9.75

12.12

13.38

11.56

14.35

Youngstown

11.17

11.52

9.40

12.05

13.29

11.43

12.85

8.69

3.12

3.45

3.34

4.65

4.81

4.60

5.14

4.10

Detroit

1955

Average
Annual
Rates of
Growth

117.04 15.22

10.46

7.09

9.04

8.33

7.80

7.98

8.93

10.90

11.35

— 1.5

6.02

6.24

4.54

5.63

6.51

6.67

7.11

8.42

9.41

9.66

+ 5.6

Western

4.33

4.69

4.27

5.43

6.16

5.73

6.64

5.36

6.46

6.64

7.01

5.73

5.55

6.16

6.77

6.10

7.01

7.80

8.41

+ 2.6

Southern

4.01

4.35

3.96

4.92

5.03

4.58

5.88

5.20

6.22

5.43

6.67

5.09

5.20

5.65

5.71

6.22

7.23

7.70

+ 2.7

Buffalo

4.21

4.37

4.06

4.87

5.38

4.82

6.04

4.77

6.55

6.29

6.45

4.06

4.72

5.18

4.72

4.77

5.43

6.04

7.11

+ 1.4

Cleveland

4.01

4.06

3.67

4.59

4.92

4.49

5.99

4.68

6.08

5.75

5.79

3.86

4.83

5.55

5.12

5.36

5.84

7.19

6.73

+ 2.4

Cincinnati

2.71

2.75

2.52

3.11

3.50

3.07

3.58

3.11

4.29

4.80

4.01

3.30

4.48

3.97

4.36

4.17

4.80

5.54

6.24

+ 4 .0

St. Louis

1.45

1.48

1.84

1.95

1.86

2.27

1.89

2.63

2.74

2.63

2.60

2.99

2.66

3.01

2.96

3.10

3.37

3.42

+ 4 .9

Sources: American Iron and Steel Institute and Federal Reserve Bank of Cleveland

TABLE II
Shares of Total Steel Ingot Output
Produced by Major Steel Centers, 1947-65
1947

1948

1949

1950

1951

1952

1953

1954

1955 1956

1957

1958

1959

1960

1961

1962

1963

1964

1965

United States

100.0%

100.0% 30.0%

100.0%

Fourth District

47.4

46.7

45.4

45.6

45.5

44.1

41.2

42.2

12.2

40.0

38.2

41.0

38.2

37.2

37.7

37.8

38.7

38.3

Other Than Fourth District

52.6

52.3

54.6

54.4

54.5

55.9

58.8

57.8

60.0

61.8

59.0

61.8

62.8

62.3

62.2

61.3

61.7

Chicago

20.2

20.1

19.7

19.5

18.9

19.7

21.4

20.2

19.7

19.8

21.5

19.2

20.9

21.1

21.4

21.1

20.5

20.1

Pittsburgh

26.3

26.0

25.4

25.0

25.1

24.0

22.6

22.5

22.3

22.0

21.5

21.3

20.2

19.5

19.9

20.0

19.8

14.5

14.0

14.5

14.3

13.9

13.5

14.0

13.9

9.7

8.4

8.0

8.1

8.2

8.6

Northeast Coast

12.1

12.5

12.7

12.4

12.9

13.3

13.8

14.3

14.5

Youngstown

13.2

13.0

12.1

12.5

12.6

12.3

11.5

9.8

10.8

10.7

9.3

8.3

Detroit

3.7

3.9

4.3

4.8

4.6

4.9

4.6

5.1

5.4

5.5

5.3

6.0

6.6

6.8

7.2

7.7

7.4

Western

5.1

5.3

5.5

5.6

5.9

6.2

6.0

6.1

5.5

5.8

6.2

6.7

5.9

6.2

6.9

6.2

6.4

6.2

6.4

Southern

4.7

4.9

5.1

4.8

4.9

5.3

5.9

5.3

4.7

5.9

6.0

5.6

5.7

5.8

5.7

5.9

5.0

5.1

5.2

5.4

5.6

5.5

5.7

4.8

5.1

5.2

4.8

5.0

4.8

5.4

Buffalo

5.0

4.9

5.2

Cleveland

4.7

4.6

4.7

4.8

5.4

5.3

5.2

5.0

5.1

4.5

5.2

5.6

5.2

5.5

5.3

5.7

5.1

3.2

3.3

3.2

3.5

3.7

4.2

3.6

3.9

4.8

4.0

4.5

4.2

4.4

4.8

1.9

1.8

2.0

2.1

2.2

2.4

3.1

3.2

2.7

3.1

3.0

2.8

2.7

2.6

Cincinnati

3.2

3.1

3.2

St. Louis

1.7

1.9

Sources: American Iron and Steel Institute and Federal Reserve Bank of Cleveland

ECONOMIC REVIEW
output in the District failed to match 1 9 4 7 -4 8

As shown in Table I as well as in the chart,

levels, particularly in the late 19 5 0 's and

of the four steel centers in the Fourth District,

early 1960's. The situation in the nation as a

only in the Cincinnati area did production

whole was conspicuously better as can be

set a new record in 1965. Outside the Fourth

seen by comparing the first two lines in Table I

District, production in all seven steel areas

with the respective figures in each case for

reached new highs in 1 9 6 5 —in fact, in most

19 4 7 and 1 9 4 8 .

cases

considerably

higher than

Not only did the growth path of steel out

records. While record production had been

put in Fourth District centers differ from that

achieved in Cleveland in 1964, output in

of the nation during 1947-65, particularly

1 965 receded slightly. In the cases of Pitts

since 1 955, but the combined output of Dis

burgh and Youngstown, despite sizable in

trict centers was also characterized by much

creases in recent years, steel output did not

wider fluctuations during periods of change

return to the record levels achieved in 1 9 5 3

in economic activity. That is to say, the swings

and 1951, respectively.

in steel output from peak periods to trough

Average Annual Rate of Growth. While the

periods, and from troughs to peaks, were rela

combined output of steel centers in the Fourth

tively larger for the Fourth District than for

District was moving upward at an almost im

the U. S. as a whole. In short, then, combined

perceptible rate of 0.1 percent per year be

steel output for Fourth District centers gener

tween 1 9 4 7 and 1965, output of the combined

ally grew less and fluctuated more than else

steel centers outside the District was advanc

where in the nation. These patterns are

ing at an annual rate of 2 .5 percent (see Table

clearly visible when production figures for

I). Of the four centers located within the

the District—both total and individual centers

Fourth District, in only the Cincinnati area

—are compared with figures for steel centers

did production register a growth rate (4 per

located outside the District.

cent) that exceeded the growth rate of the

STEEL OUTPUT: VOLUME, GROWTH,
AND CYCLICAL SWINGS

combined centers outside the District (2.5
percent), although Cleveland was not far be
hind (2.4 percent). In marked contrast to

Volume of Steel Output. Steel output for the

trends elsewhere, Pittsburgh and Youngs

nation as a whole was a record 1 3 1 .2 million

town had average annual rates of decline

tons in 1965. Correspondingly, steel pro

during the 1 9 4 7 -6 5 period (—0 .3 percent

duction in the Fourth District, as well as out

for Pittsburgh and — 1.5 percent for Youngs

side the District, was also at record levels

town), reflecting the influence of the many

(see Table I). Mills within the District pro

years when steel production in those centers

duced 5 0 million tons of steel in 1965, or 2 5

badly lagged output elsewhere. All centers

percent more than they had in 1 9 4 7 ; mills

outside the Fourth District scored long-term

outside the District produced a total of 81

gains (positive growth rates) during 1947-65.

million tons in 1965, or 81 percent more
than in 1947.

In terms of average rate of growth, Cincin
nati ranked high among the 11 steel centers.

NOVEMBER 1 9 66
In fact, only two centers—Detroit with a 5 .6

which had amounted to 3 .2 percent of total

percent average annual rate of increase and

output in 1947, increased to 4 .8 percent in

St. Louis with a 4 .9 percent rate of increase—

1 965; and Cleveland's share, which had been

surpassed Cincinnati. The 2 .4 percent growth

4 .7 percent in 1947, improved to 5.1 percent

rate of steel production in the Cleveland area,

in 1965. On the other side of the ledger, two

while less impressive than that of Cincinnati,

District centers accounted for smaller shares

was still larger than the growth rates of the

of total output in 1 9 6 5 than in 1947. Steel

Chicago (1 .8 percent) and Buffalo (1 .4 per

output in the Pittsburgh area amounted to

cent) areas, among steel centers outside the

1 9 .8 percent of total U. S. production in 1 9 6 5

District. As the only two steel producing

compared with 2 6 .3 percent in 1947, and

regions in the nation to experience a decline

Youngstown accounted for 8 .6 percent in

in trend during

19 6 5 compared with 13.2 percent in 1947.

1947 -6 5 , Pittsburgh and

Youngstown are of course at the bottom of the

In the case of Pittsburgh, the reduction in

list—tenth and eleventh, respectively— in a

share of total output was fairly evenly spread

ranking of growth rates for the 11 steel centers

over the 1 9 4 7 -6 5 period, although about 6 0

of the nation.

percent of the reduction in share occurred in

Changes in Share of Production. The growth

the first half of the overall period. Also, Pitts

pattern and cyclical swings experienced by

burgh has not lost any ground since 1961,

each steel producing center during 1 9 4 7 -6 5

and in fact has even recouped somewhat.

were reflected in the changing proportion of

The pattern was fairly similar in the Youngs

total steel output accounted for by each cen

town area during 1947-65, as shown in Table

ter. As shown in Table II, in 1 9 4 7 the four

II, although Youngstown lost relatively more

steel centers in the Fourth District accounted

ground than Pittsburgh in terms of share of

for 4 7 .4 percent of total steel production. In

national total.

contrast, in 1 9 6 5 the District's share of total

The losses in share of total production ex

production had declined to 3 8 .3 percent.

perienced by Pittsburgh and Youngstown

This is not surprising in view of the com para

over the years have been turned into gains in

tive growth rates for the Fourth District and

proportions of total output accounted for by

the other-than-Fourth District areas. Never

most of the other steel producing centers.

theless, because of the relative spurt in steel

Detroit has been the largest gainer, moving

output in recent years at District centers, the

from 3 .7 percent of total output in 1 9 4 7 to

proportion achieved in 1 965 remained a

7 .4 percent in 1965. On the other hand,

shade above the lowest point in share of total

Chicago, now the leading steel producing

output accounted for by the District—3 7 .2

area, accounted for almost the same pro

percent in 1961.

portion of the nation's steel output in 1 9 6 5 as

Two of the steel producing areas within the

it did in 1947.

Fourth District increased their proportionate

Cyclical Swings. The cyclical swings in steel

shares of total steel production during 1947-

output invariably have been wider in the

65. Thus, steel output in the Cincinnati area,

Fourth District than in the other-than-Fourth

E C O N O M IC

R E V IE W

TABLE III
Cyclical Swings in Steel Output
Percent Change From Peak to Trough
Business
Cycle Periods
(Recessions)
1948-49
1953-54
1955-58
1960-61

United States

Fourth District

— 12%
— 21
— 27
— 1

— 14 %
— 26
— 33
— 4

Other Than
Fourth District
— 10%
— 17
— 22
+

Cincinnati

Cleveland

Youngstown

Pittsburgh

— 9%
— 13
— 23
+ 10

— 10%
— 21
— 37
— 8

— 19 %
— 33
— 44
— 6

— 14%
— 23
— 30
— 4

Percent Change From Trough to Peak
Business
Cycle Periods
(Expansions)
1949-53
1954-55
1958-60
1961-65

United States

Fourth District

Other Than
Fourth District

Cincinnati

Cleveland

Youngstown

Pittsburgh

+43%
+ 33
+ 16
+ 34

+ 39%
+35
+ 16
+ 38

+47%
+ 30
+ 16
+ 31

+42%
+ 38
+20
+43

+63%
+30
+44
+ 31

+37%
+46
+ 18
+45

+35%
+32
+ 9
+35

NOTE: Percentage changes are based on annual data, with the terminal dates of each period representing the years in
which peaks or troughs in steel output occurred.
Sources: American Iron and Steel Institute and Federal Reserve Bank of Cleveland

District areas, and as a result, in the U. S. as

11 percentage points larger. During 1960-61,

a whole.2 Thus, as shown in the upper part

the decline in the District amounted to 4

of Table III, in the four recession periods

p ercen t, while the steel output of other-

since 1947, the peak to trough drop in steel

than-Fourth District areas actually rose by 1

production was relatively greater on each

percent. As can be seen from the data for

occasion in the District than it was outside the

individual centers, the Youngstown area was

District (see columns 2 and 3). It is also ap

the "biggest swinger" in the first three re

parent from Table III that during the first

cessions, and Cincinnati the most moderate.

three recession periods, the difference be

The Cleveland area showed the sharpest de

tween percentage declines in steel production

cline in 1960-61, while Cincinnati moved

in the Fourth District and all other steel cen

completely contrary to the District pattern,

ters widened with each recession. Thus,

actually posting a 10-percent increase.

during 1 9 4 8 -4 9 , the decline in Fourth Dis

Looking at the trough to peak changes,

trict production was 4 percentage points

only during the 1 9 4 9 -5 3 expansion did steel

larger; during 1953 -5 4 , the decline was 9 per

production in the Fourth District fail to rise by

centage points larger; and during 1955-58,

at least as much as, or more than, outside the
District. However, even with the greater per

2 B ecause of the problems involved in adjusting monthly

centage increases in production in two of the

steel production for seasonal variation, annual data are

periods and an equal percentage increase in

used in this analysis. P ercen tage changes are based on
swings in annual data, using as terminal dates the years

one period, steel output in the District lost

in which peaks and troughs in steel output occu rred .

ground during 1947-65, indicating that these

Digitized for16
FRASER

NOVEMBER 19 66
advances were not strong enough to com

in Pittsburgh steel production have not been

pensate sufficiently for the preceding down

as wide as in Youngstown but generally more

turns.

volatile than in the other two steel producing

PATTERNS IN
FOURTH DISTRICT CENTERS

business cycle periods, as shown in the chart,

Pittsburgh. For many years the Pittsburgh

centers in the Fourth District. Disregarding
during 1 9 5 9 -6 3 steel output in Pittsburgh
showed relatively little year-to-year change

area represented the leading steel producing

compared with other steel centers. In fact,

region in the U. S.

between 1 9 5 9 and 1 9 6 0 steel production in

Since 1 960, however,

Pittsburgh has ranked second to Chicago in

Pittsburgh remained unchanged, while it rose

volume of output, as can be seen from Table

sharply in seven steel districts and dropped

I. The shift of positions of the two areas was

almost as sharply in the other three districts.

the inevitable result of their individual growth

Youngstown. The Youngstown district rank

patterns since 1 9 4 7 . In the late 19 4 0 's and

ed fourth in production of steel in 1965, hav

early 1950's, annual steel production in Pitts

ing dropped out of the number three spot in

burgh exceeded that in Chicago by 5 to 6

the U. S. as long ago as 1949. Unlike other

million tons. However, with production, on

districts, Youngstown reached its postwar

an average, growing in Chicago and declin

peak in production as early as 1951, and each

ing in Pittsburgh, the 5-million ton difference

surge since that time (except for 1965) has

was eventually wiped out. In 1965, steel out

fallen short of that peak by an increasingly

put in the Chicago area exceeded output in

larger amount (see the chart). In 1947, out

the Pittsburgh area by almost half a million

put of the Youngstown area amounted to 11.2

tons. In 1 9 6 5 , Pittsburgh accounted for only

million tons. In 1951, which was Youngs
town's re co rd year, production am ounted to

1 9 .8 percent of steel output in the nation as
compared with 2 6 .3 percent in 1947.

13.3 million tons. In 1953, 1955, 1964, 1965,

The long-term downtrend of the Pittsburgh

all of which were years when new national

steel district probably reflects the gradual

records in steel production were being set,

change over the years in the technology of

production in the Youngstown area amounted

making steel and marketing the finished

to 1 2 .9 million tons, 1 2 .7 million tons, 10.9

product. Regardless of the reason for the sec

million tons, and 1 1 .4 million tons, respec

ular decline, as shown in the chart, record

tively. (Thus, only in 1 9 6 5 did production in

steel production was achieved in Pittsburgh

crease over the previous peak; but it still

in 1953, and output failed to reach that level

failed to return to the 1951 record level.)

again in any of the succeeding steel peaks in
1955, 1964, and 1965.

Production dropped by a greater percent
age in Youngstown than in any of the other

Fluctuations in steel output in the Pittsburgh

Fourth District centers in three out of the four

district generally have been less than for the

cyclical downturns shown in Table III. In two

U. S. and the Fourth District as a whole dur

of the four advances in output, production in

ing both recessions and expansions. Swings

Youngstown increased by a greater percent

ECONOMIC REVIEW
age than in any of the other Fourth District

the United States, the Cincinnati district

centers. For example, during 1961 -6 5 , pro

ranked tenth in volume of steel produced in

duction was up 4 5 percent in Youngstown

1 965. Two key features—the growth rate and

compared with 4 3 percent in Cincinnati, 3 5

cyclical behavior—have distinguished Cin

percent in Pittsburgh, and 31 percent in

cinnati from the other steel centers in the

Cleveland. Nevertheless, over the long term

Fourth District—Cleveland, Youngstown, and

it is apparent from the — 1.5 percent average

Pittsburgh. That is to say, steel production in

annual decline that the periods of expansion

Cincinnati has grown more and fluctuated

have been insufficient to compensate for the

less than in the other District centers.

periods of decline.

Steel production in Cincinnati grew at an

Cleveland. As one of the smallest steel dis

average annual rate of 4 percent during 1947-

tricts in the U. S., Cleveland ranked ninth in

65. In the rest of the nation, only Detroit and

1 9 6 5 in volume of steel produced. The Cleve

St. Louis surpassed Cincinnati in growth of

land district produced 6 .5 million tons of

steel production. In 1947, steel output in the

steel in 1 9 6 5 (slightly more than 5 percent of

Cincinnati area amounted to 2 .7 million tons;

the nation's total output), an increase of 6 8

in 1 9 6 5 it amounted to 6 .2 million tons, an

percent from the 4 million tons produced in

increase of 130 percent from the 1 9 4 7 volume.

1947. The average annual growth rate of 2 .4

Cincinnati's better-than-average growth rate

percent for Cleveland during 1 9 4 7 -6 5 was

resulted in a steadily higher proportion of

higher than that for the Fourth District as a

total steel output between 1 9 5 4 and 1965.

whole (0.1 p ercen t) as well as for the nation

If recent trends were to continue, Cincinnati

(1.5 percent), but lagged such fast growing

might well nudge Cleveland out of ninth

steel centers as Detroit, St. Louis, and Cin

position in the ranking of steel producing

cinnati. As can be seen from the data in Table
I, as well as from the chart, the bulk of the

centers.
A comparison of percentage decreases in

growth in steel production in the Cleveland

steel output from peak to trough in four post

area took place during the first half of the

war downturns indicates that Cincinnati's

1950's.

strong growth pattern resisted the downward

Steel production in Cleveland over the

pressure of cyclical declines. In each of the

years has tended to be very volatile. Unlike

first three downswings, Cincinnati's percent

in Youngstown, however, the sharp plunges

age decline was less than that of any of the

in Cleveland production have been more

other steel centers in the District, and during

than offset by increases in periods of eco

1960-61 Cincinnati posted an increase in

nomic expansion, so that the area's output

contrast to declines in the other three centers.

has registered some secular growth. In 1965,

In fact, when a similar comparison is made

the experience of Cleveland was different in

between Cincinnati and the combined steel

that steel output decreased while it was in

centers outside the District, the figures show

creasing in the other ten districts.

that, as a general matter, Cincinnati fared

Cincinnati. Among the 11 steel districts in
Digitized for18
FRASER

quite well.

NOVEMBER 1966

CONCLUDING COMMENTS

production in centers outside the District,
particularly in the four cyclical downturns

The steel industry in the Fourth District has

between 1 9 4 7 and 1965. There are probably

declined in relative importance during the

a number of reasons why steel output in the

postwar period, although the percentage in

Fourth District fluctuates more widely over

crease in District production since 1961 has

the business cycle than does output in the

matched that of the nation. The combined

nation, as well as why District output has de

output of the four steel centers located within

clined in relative

the District chalked up only an average annual

trends and cyclical behavior of the various

rate of growth of 0.1 percent during 1 9 47-65,

steel regions are determined by mill location,

with better-than-average growth rates of steel

steel product mix, shifts in demand, and im

importance.

Long-term

output in Cincinnati and Cleveland being

port-export relationships, among other things.

virtually canceled by declines in rates of

But it is no simple matter to quantify these

change in both Pittsburgh and Youngstown.

factors. If research presently going on is suc

With the exception of Cincinnati, production

cessful in isolating some of these reasons,

in steel centers located within the District has

they will be reported on in a subsequent

evidenced wider cyclical swings than has

article.

APPEN D IX
11 M ajor Steel Producing Districts in the United States
C H IC A G O D ISTRICT
Chicago, Illinois
Chicago Heights, Illinois
Lemont, Illinois
Morton Grove, Illinois
Sterling, Illinois
East Chicago, Indiana
G ary, Indiana
Kokomo, Indiana
Fort W ayne, Indiana
New Castle, Indiana
Duluth, Minnesota
PITTSBU RG H D ISTRICT
Johnstown, Pennsylvania
Washington, Pennsylvania
Donora, Pennsylvania
Braeburn, Pennsylvania
Latrobe, Pennsylvania
Monessen, Pennsylvania
Aliquippa, Pennsylvania
Ambridge, Pennsylvania
Beaver Falls, Pennsylvania
Midland, Pennsylvania
Monaco, Pennsylvania
Butler, Pennsylvania
Duquesne, Pennsylvania
Braddock, Pennsylvania
McKeesport, Pennsylvania
Munhall, Pennsylvania
Clairton, Pennsylvania
Brackenridge, Pennsylvania
Pittsburgh, Pennsylvania
Bridgeville, Pennsylvania
Carnegie, Pennsylvania
West Homestead, Pennsylvania
Oakmont, Pennsylvania
Weirton, West Virginia
Steubenville, Ohio
Toronto, Ohio
N O R T H E A S T C O A S T D ISTRICT
Bridgeport, Connecticut
Phillipsdale, Rhode Island
W atervliet, New York
Harrison, New Jersey

http://fraser.stlouisfed.org/
Source: American Iron and Steel Institute
Federal Reserve Bank of St. Louis

Roebling, New Jersey
Philadelphia, Pennsylvania
Coatesville, Pennsylvania
Phoenixville, Pennsylvania
Ivy Rock, Pennsylvania
Fairless Hills, Pennsylvania
Reading, Pennsylvania
Bethlehem, Pennsylvania
Harrisburg, Pennsylvania
Steelton, Pennsylvania
Milton, Pennsylvania
Burnham, Pennsylvania
Sparrows Point, Maryland
Baltimore, Maryland
Claymont, Delaware
Y O U N G ST O W N D ISTRICT
New Castle, Pennsylvania
Farrell, Pennsylvania
Campbell, Ohio
Youngstown, Ohio
Lowellville, Ohio
Warren, Ohio
Canton, Ohio
Massillon, Ohio
Mansfield, Ohio
D ETR O IT D ISTRICT
Dearborn, Michigan
Ecorse, Michigan
Ferndale, Michigan
Trenton, Michigan
Warren, Michigan
W ESTERN D ISTRICT
Pueblo, Colorado
Geneva, Utah
Helena, Arizona
Seattle, Washington
Portland, Oregon
South San Francisco, California
Emeryville, California
Niles, California
Pittsburg, California
Los Angeles, California
Torrance, California
Fontana, California

S O U TH ER N D ISTRICT
Newport News, Virginia
Roanoke, Virginia
Atlanta, G eorgia
Tampa, Florida
Knoxville, Tennessee
Ensley, Alabam a
Fairfield, Alabam a
Birmingham, Alabam a
Gadsden, Alabam a
Anniston, Alabam a
Jackson, Mississippi
Sand Springs, Oklahoma
Fort Worth, Texas
Lone Star, Texas
Longview, Texas
Houston, Texas
Pampa, Texas
B U F F A LO D ISTRICT
Cortland, New York
Sycracuse, New York
Lockport, New York
Buffalo, New York
Lackawanna, New York
Tonawanda, New York
Dunkirk, New York
Irvine, Pennsylvania
Erie, Pennsylvania
C LE V E LA N D D ISTRICT
Cleveland, Ohio
Lorain, Ohio
C IN C IN N A T I D ISTRICT
Huntington, West Virginia
Cincinnati, Ohio
Middletown, Ohio
Portsmouth, Ohio
Ashland, Kentucky
Newport, Kentucky
Owensboro, Kentucky
ST. LO U IS D ISTRIC T
Peoria, Illinois
Alton, Illinois
Granite City, Illinois
Kansas City, Missouri