The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
IN THIS I SSUE D efense Spending in Fourth District S tates P a rt I I ................................2 Dimensions o f Su b cyclical Fluctuations in S teel and Auto O utput . . . FEDERAL RESERVE BANK OF 16 CLEVELAND ECONOMIC REVIEW DEFENSE SPENDING IN FOURTH DISTRICT STATES PART II An earlier article in the E con om ic Review reviewed selected aggregate measures of the impact of national defense.1 The purpose of that article was to provide background for an evaluation of the impact of defense spending in Fourth District states (Ohio, Pennsylvania, Kentucky, and West Virginia). A brief review of the conclusions of the earlier article will perhaps be helpful as an introduction to this article. For one thing, it was concluded that defense spending has absorbed less than 10 percent of the nation's resources in recent years and that the pro portion had been declining until the Vietnam buildup. It was also pointed out that defense procurement has been heavily concentrated in a few major product areas. Moreover, the direct effects of defense spending were found to influence significantly only a few industries and generally the large firms within those industries. Accordingly, it is likely that the high con centration of defense procurement among products, industries, and firms would also involve a substantial geographical concen 1 See "Defense Spending in Fourth District States Part I: N ational B a c k g ro u n d ," E conom ic Review , F e d e ra l Reserve Bank of Cleveland, Cleveland, Ohio, February 1966. Digitized for 2 FRASER tration of defense impact. But here, unfortu nately, aggregate data are inadequate since they are not broken down by region, in other words, subnational data within a national con text. Therefore, in analyzing the regional im pact of defense spending, other types of data must be used. Only in this way can the defense impact in Fourth District states be assessed. MEASURES OF DEFENSE IMPACT IN FOURTH DISTRICT STATES Military Prime Contract A w ards. A first ap proximation of the impact of defense spending at the regional level is provided by military prime contract awards, which are broken down by states. Prime contract awards are contracts awarded by the Department of De fense (DOD) to firms that supply goods and services such as research and development, construction, petroleum, subsistence, clothing, and operating and maintenance supplies. A major limitation of these awards for regional analysis is that a substantial portion of the work under prime contract awards is sub contracted to firms in the same or other re gions, and the extent of subcontracting is not known. Thus, at best, prime contracts can only give a rough indication of defense work in a state or area. MARCH 1966 TABLE I Military Prime Contract Awards to Firms in the United States and Fourth District States, Fiscal Years 1951-1965 In Millions of D ollars* As Percent o f United Sta te s To tal* Fourth Fourth Fiscal United District District Year S tates S tates S tates O h io P a. Ky. Va. 1951 $ 2 9 ,6 2 0 $ 3 ,7 4 4 1 2 .6 % 7 .5 % 4 .8 % 0 .2 % 0 .1 % 1952 3 8 ,4 7 9 4 ,5 3 4 1 1 .8 6.2 5 .0 0 .3 0 .2 1953 2 6 ,9 9 5 2 ,4 7 5 9 .2 5.2 3 .5 0 .3 0 .2 1954 1 0 ,6 31 588 5 .5 3 .7 2 .6 0 .3 0 .2 1955 1 3 ,9 7 2 1 ,5 9 6 1 1 .4 6.3 4 .5 0 .4 0 .2 1956 1 6 ,4 91 1 ,8 0 8 1 1 .0 6 .3 4 .2 0.3 0.1 W e st 1957 1 8 ,1 4 4 1 ,9 6 8 10 .8 6.8 3.8 0 .2 0.1 1958 2 1 ,0 0 9 1 ,7 5 9 8.4 4 .8 3.3 0 .2 0.1 1959 2 1 ,9 1 9 1 ,7 7 4 8.1 4 .7 3.1 0 .2 0.1 1960 2 0 ,4 0 7 1 ,6 4 7 8.1 4 .4 3 .3 0 .2 0 .2 1961 2 2 ,1 1 2 1 ,9 1 7 8 .7 4 .5 3 .6 0 .2 0 .3 1962 2 5 ,0 3 9 2 ,2 5 8 9 .0 4 .5 3 .8 0 .2 0 .5 1963 2 5 ,2 3 4 2 ,4 5 0 9 .7 5 .3 3 .5 0 .2 0 .7 1964 2 4 ,4 1 7 2 ,0 3 9 8.4 4 .2 3 .6 0 .2 0 .4 1965 2 3 ,2 6 8 1 ,9 8 5 8.5 3 .7 4 .2 0 .2 0 .4 includes contracts distributed b y states with net v alu e o f $1 0 ,0 0 0 or more. Source: O ffic e of S e c re ta ry o f D efense, D epartm ent o f D e fen se, “ Prime C ontract A w a rd s b y S ta te ” Table I presents figures on military prime contracts for the U. S. and the four states wholly or partly in the Fourth Federal Reserve District. Of the four states, the bulk of con tract awards has gone to Ohio and Pennsyl vania. In each year, with the exception of fiscal 1965, Ohio has received 50 percent or more of prime contract awards in Fourth Dis trict states. The decline in dollar volume of prime contract awards in Fourth District states in the past two fiscal years is due to the drop in contracts awarded to firms in Ohio. In terms of shares of prime contract awards, Fourth District states declined in importance in the late 1950's and have remained on a virtual plateau in recent years. Table II pre sents the average share of total contract awards to each of the states and to the four states as a whole for the subperiods 1951-57 and 1958-65. These subperiods were selected because they reflect (are on the two sides of) the downward shift in shares of contract awards that occurred during 1957-58. Thus, average shares dropped for three of the four states during the second subperiod, and signifi cantly for Ohio and Pennsylvania, indicating TABLE II A ve ra g e Sh ares of M ilitary Prim e Contract A w ards to Firms in Fourth District States as a Percent of Total A w ards in the United States Fourth Fiscal Y e a rs W e st District O h io P a. Ky- Va. S tates 19 51-57 6 .0 % 4 .1 % 0 .3 % 0 .2 % 1 0 .3 % 19 5 8 -6 5 4 .5 3 .5 0 .2 0 .3 8 .6 Source: Com putations from T a b le I 3 ECONOMIC REVIEW TABLE III Costs Incurred by the Atomic Energy Commission in the United States and Fourth District States, Fiscal Years 1961-1965 In Millions of Dollars A s Percent of United Sta te s Total Fourth Fourth Fiscal United District District Year Sta te s S tates States O h io 1 2 .7 % 6 .7 % 2 .4 % 3 .6 % 1 0 .5 5 .0 2 .4 3.1 W e st P a. Ky. 1961 $ 2 ,7 9 5 $354 1962 2 ,8 8 9 303 1963 2 ,8 3 0 298 1 0 .5 5.1 2 .5 3 .0 1964 2 ,7 9 5 324 1 1 .6 5 .7 2 .9 3 .0 1965 2 ,7 4 8 296 10 .8 4 .8 3 .6 2 .4 Va. * * * * * * Less than 1.0 percent. Source: Atomic E n e rg y Commission TABLE IV Prime Contracts Awarded by the National Aeronautics and Space Administration in the United States and Fourth District States, Fiscal Years 1961-1965 In Millions o f D ollars1 As Percent o f United S ta te s Total Fourth Fourth Fiscal United District District Year S tates States States 1961 $ W e st O h io P a. 380 $ 20 5 .3 % 2 .1 % 1962 939 37 3 .9 1.1 2 .7 1963 2 ,18 1 63 2.9 1.5 1.4 1964 3 ,4 9 0 105 3 .0 1.5 1.5 1965 4 ,1 0 3 114 2 .8 1.3 1.3 3 .0 % K y. Va. * * * * * * * * * * * Less than 1 .0 percent. 1 N atio n al Aeronautics an d S p a c e Adm inistration a w a rd s cover contracts o f $ 2 5 ,0 0 0 an d over. Source: N atio n al Aeronautics an d S p a c e Adm inistration that Fourth District states have been receiving a smaller portion of a declining volume of de fense contracts. It is interesting—and obviously significant—that the subperiods chosen cor respond roughly to the pre-Sputnik and postSputnik eras. Costs incurred by the Atomic Energy Com mission (AEC) and prime contract awards of the National Aeronautics and Space Adm inis tration (NASA) for the four states are pre 4 sented in Tables III and IV.2 Costs incurred by the AEC in Fourth District states have been declining since fiscal 1962. With the exception of Pennsylvania, Fourth District states re ceived a smaller share of the total in fiscal 1965 than in previous years. In contrast, 2 Costs incurred are not the same thing as prime con tract awards, but they are a fairly close approximation. Costs incurred cover operating, m aintenance, and capital expenditures. MARCH 196 6 MILITARY and NASA PRIME CONTRACT AWARDS, and AEC COSTS INCURRED, FOURTH DISTRICT STATES * P r e v i o u s d a t a not a v a i l a b l e N O T E : M i l i t a r y p r i m e c o n t r a c t s c o v e r a w a r d s with ne t v a l u e of $ 1 0 , 0 0 0 t h a t a r e d i s t r i b u t e d b y s t a t e ; N A S A c o n t r a c t s c o v e r d i r e c t a w a r d s of $ 2 5 , 0 0 0 a n d o v e r ; A E C a w a r d s c o n s i s t of c o s t s i n c u r r e d for o p e r a t i n g , m a i n t e n a n c e , and capital expenditures. S o u r c e s of d a t a : Prim e C o n tra c t A w ards b y S ta te , O f f i c e of S e c r e t a r y of D e f e n s e , D e p a r t m e n t o f D e f e n s e ; Energy Commission; and National Sp a c e Administration. At omic Ae r on a ut ic s and NASA contract awards in the four states have grown substantially in dollar volume, reflect ing increased missile and space activity. Nonetheless, awards in the four states have grown less rapidly than total contract awards in the U. S. and constitute a declining pro portion of the total, particularly in Ohio and Pennsylvania. The accompanying chart portrays contract awards of the DOD and the NASA and costs incurred by the AEC, as well as the combined total for Fourth District states. The combined total provides an overall picture of defenserelated procurement in the four states. As shown in the chart, the combined series ex perienced moderate increases from fiscal 1961 through fiscal 1963, when it reached a peak, and then declined slightly in both fiscal 1964 and 1965. Changes in the combined total have been dominated by changes in DOD contracts. Thus, in general, the data on contract awards do not present an encouraging picture for the Fourth District states. With the exception of NASA contracts, prime contract awards have been declining, and Fourth District states have been receiving a smaller share of a de clining total. The major problem of course is to explain why Fourth District states have been receiving a smaller share. One possible answer, or thesis, involves the fact that there has been a change in the structure of defense procure ment, and that Fourth District states do not have the productive capacity to meet such a shift. Military strategy depends upon many factors, and changes in any of the factors—or the strategy—may considerably alter defense posture and a weapons system, that is, a changed defense strategy may require new ECONOMIC REVIEW types of defense goods. An illustration of this is found in the change in defense posture that had been occurring throughout the 1950's, but particularly after Sputnik in 1957. As a result of this change, more emphasis was placed on massive nuclear retaliation with missiles, which meant a step-up in space and missile activity and a shift in the types of re quired equipment. Regions of the country with the basic capacity to produce the new types of defense products would be expected to gain DOD contracts (as well as those of NASA and AEC), while those areas lacking in that capacity would be expected to receive a smaller share of contract awards. When tanks and trucks were major items of procurement during the Korean War and the first half of the 1950's, Fourth District states fared well in terms of share of military procurement, since much of the basic indus trial capacity of these states was readily adapt ed to produce the vehicles and ordnance of conventional warfare. The shift to missile and space equipment and electronic components, among others, required industrial capacity not extensively found in the four states. (It should also be noted that the weather and location of the four-state area are not neces sarily conducive for missile testing.) One way to test the thesis regarding the conformance of the Fourth District states' pro ductive capacity to U. S. defense needs is to compare the rank of procurement programs in the four states with that for the U. S., re membering that the U. S. ranking indicates the priority of programs in overall defense posture. It was shown in the earlier article (see page 12, Table VII) that the top three U. S. programs (missile and space systems, airframes and related assemblies and spares, and electronics and communication equip- ment) accounted for over 50 percent of total military procurement during fiscal years 1962-64. Assuming that prime contracts awarded for various programs in individual states reflect the types of industrial capacity in those states, a close correspondence of program ranks in the various states with the ranking of programs in the U. S. would indi cate an ability to produce the required defense goods. Similarly, lack of correspondence would indicate an inability to supply the de manded defense goods because of the absence of adequate production facilities. Table V presents the various procurement programs by rank for the nation and each of the Fourth District states. The programs are arranged in order of descending magnitude of dollar volume of prime contracts in the U.S. Thus, missile and space systems is the top program for the U. S., but ranks tenth in Ohio, second in Pennsylvania, nineteenth in Ken tucky, and second in West Virginia. A statis tical technique can be used to determine the relationships of the ranks between each of the states and the U. S .3 The results of the statis tical analysis (rank correlation coefficients) s The statistical technique is rank correlation. The form, tor * this test is ... 6 ( 2rf2) ula V = 1, ------------------ n (n2 — 1) W h ere Y' = the rank correlation coefficient = the sum of the squares of the num erical differences between the rank of a proqram in the U. S. and its rank in a state. n = the number of ranks T' can vary between + 1 and — 1; if it is close to 0, this would indicate n o relationship betw een the program rankings in the state and the U. S. As V' approaches + 1 or — 1, it would indicate a close correspondence b e tween U. S. and state program s. A value of + 1 would indicate that the program s in the state are ranked in the same order as the U. S. program . An r ' of — 1 would indicate that the state program s are inversely related to U. S. program s, so that the 25th program in a state is the first program in the U. S. MARCH 196 6 TABLE V Rank of Military Procurement Programs for Fourth District States as Compared with United States Programs (Ranks b ase d on th re e -y e a r total o f d o lla r volume o f m ilitary prim e contracts, Fiscal Y e a rs 1 9 6 2 - 1 9 6 4 ) Program Rank W e st United S tates Program O h io P a. Va. Ky. M issile an d S p a c e S y s t e m s ................................................................................................................ 1 10 2 19 2 A irfram es an d R elated Assem blies and S p a r e s ...................................................................... 2 2 4 20 19 Electronics an d Communication E q u ip m e n t..................................................................................... 3 5 1 4 11 S e r v i c e s .......................................................................................................................................................... 4 9 6 5 16 5 11 3 9 7 C o n s t r u c t i o n ................................................................................................................................................ 6 13 11 3 A irc ra ft Engines and R elated S p a r e s ........................................................................................... 7 1 18 21 A m m u n itio n ..................................................................................................................................................... 8 7 7 12 3 P e t r o l e u m ..................................................................................................................................................... 9 14 14 1 14 All O th e r S upp lies an d Equipment ................................................................................................ 6 — 10 6 5 7 4 ...................................................................................... 11 8 13 15 12 S u b s is t e n c e ..................................................................................................................................................... 12 16 10 2 15 C o m b at Vehicles ..................................................................................................................................... 13 3 8 16 1 N oncom bat V e h i c l e s ................................................................................................................................ 14 4 16 8 20 Textiles, Clothing, an d E q u i p a g e ..................................................................................................... 15 21 9 6 10 W e a p o n s ..................................................................................................................................................... 16 12 12 14 8 Construction E q u ip m e n t .......................................................................................................................... 17 17 19 11 22 O th e r A irc ra ft Equipment and Supp lies Production E q u ip m e n t................................................................................................................................ 18 15 22 17 9 M e d ic a l an d Dental S u p p lies and E q u i p m e n t ........................................................................... 19 20 15 13 17 Photographic Equipm ent an d S u p p l i e s ........................................................................................... 20 22 23 22 23 M a te ria ls H andling E q u ip m e n t ........................................................................................................... 21 18 17 18 13 O th e r Fuels an d L u b r i c a n t s ................................................................................................................ 22 23 20 10 5 M ilitary Building S u p p l i e s ...................................................................................................................... 27 19 21 6 18 24 24 24 — 21 25 25 25 — — Transportation E q u ip m e n t ...................................................................................................................... . S e p a ra te ly Procured C ontainers an d Handling E q u ip m e n t ................................................. Source: M ilita ry Prim e C o n tra c t A w a rd s b y R e g io n a n d S ta te s , F is c a l Y e a rs 1 9 6 2 , 1 9 6 3 , 1 9 6 4 ,, O ffic e of S e c re ta ry o f D efen se, D epartm ent of D efen se, August 1 9 6 5 are presented in Table VI. The rank corre lation coefficient for program ranks of Penn sylvania compared with the U. S., and for Ohio compared with the U. S., are the highest, indicating that programs in Ohio and Penn sylvania are closely related to U. S. programs. Program ranks for both Kentucky and West Virginia compared with the U. S. also dis played positive coefficients, although lower than in the case of Ohio and Pennsylvania, indicating a somewhat less significant re lationship between programs in these states and U. S. programs. TABLE VI Rank Correlation Coefficients for Procurement Program Ranks in Fourth District States Com pared with Rank of United States Programs Rank C o rrelatio n Coefficients B ase d O n S ta te C o m p a re d with United S ta te s P e n n s y lv a n i a ..................... Kentucky ........................... W e st V irg inia . . . . Top 2 5 Top 1 0 Program s Program s .7 7 .1 6 .8 5 .6 4 .4 5 — .2 9 . 43 — .0 6 Source: Com putations b a s e d on T a b le V It is possible, however, that the statistical relationships are due to the less significant 7 ECONOMIC REVIEW defense programs, which do not involve large dollar magnitudes. Consequently, a compari son was made of only the ten major U. S. pro grams with their corresponding ranks in Fourth District states. (The top ten U. S. programs accounted for nearly 90 percent of U. S. pro curement.) The results (rank correlation coef ficients) of this comparison are also reported in Table VI. In all cases the relationship less ens considerably, although it is still significant for Pennsylvania.4 The closer relationship (the higher y ') for Pennsylvania can be ex plained by referring to Table V, where it is seen that for Pennsylvania the programs rank as follows: electronics, first; missile and space systems, second; ships, third; and airframes, fourth. All of these programs are top U. S. programs. The relationship between Ohio and U. S. programs is reduced substantially when ten programs are used, suggesting that Ohio does not produce the types of products 4 Since primary interest is in whether Y' is significantly g reater than 0 , a one-tailed test is appropriate (Hn = 0; H ! > 0 ) . The critical values of Y' for n (number of ranks) equal to 2 4 , 2 6 , and 10 at the .0 5 and .01 levels of sig nificance are reproduced below (the significance table used did not contain values for n = 2 5 ). Number of ranks Significance level (one-tail) .0 5 .01 10 .5 6 4 .7 4 6 24 .3 4 3 .4 8 5 26 .3 2 9 .4 6 5 It is seen that at the 5 percen t level for the top 2 5 pro grams, all of the coefficients for Fourth District states are significant, while at the 1 percent level only those for Ohio and Pennsylvania are. For the top 10 program s only Pennsylvania is significant at the 5 percen t level. (See Sidney Siegel, Nonparam etric Statistics for the Behavioral Sciences (New York: M cGraw-Hill Book C o., Inc., 1 9 5 6 ), pp. 2 0 2 -1 3 and p. 2 8 4 .) Digitized for8 FRASER required by the U. S. defense system. Even if missile and space systems were eliminated and the remaining nine programs ranked, the results ( y ') for Ohio are still low (only 0.3), again indicating little relationship between the types of productive capacity in Ohio, as represented by procurement programs in Ohio, and types of output requirements for national defense, as represented by U. S. procurement programs. As in the nation, the concentration of pro curement is much in evidence in the individ ual states, with the top ten programs of both Ohio and Pennsylvania accounting for about 87 percent of total procurement in these states for fiscal years 1962-64. (The top ten programs in the U. S. also accounted for 87 percent of all U. S. procurement even though the pro grams are not the same.) In Ohio, aircraft engines, airframes, combat vehicles, and non combat vehicles—all conventional forms of military goods—comprised slightly over 60 percent of contract awards to Ohio firms dur ing 1962-64. For three of the four programs— aircraft engines, combat vehicles, and non combat vehicles—Ohio received over onefifth of total U. S. awards for these programs. More specifically, Ohio by itself received al most one-fourth of all U. S. contract awards for aircraft engines and related spares, and one-fifth of all U. S. contract awards for com bat and noncombat vehicles. On the other hand, while electronics awards comprised 7.2 percent of all Ohio contract awards, they accounted for only 2.7 percent of total U. S. contract awards to this program; missile and space systems accounted for 3.1 percent of all Ohio awards, and only 0 .6 percent of U. S. MARCH 1966 contract awards to this program. Thus, it is apparent that Ohio has derived the bulk of its Government contracts for conventional war goods which have not been, with the exception of airframes, the most significant programs in terms of U. S. priority as determined by the ranking of programs by dollar volume. The top four Pennsylvania programs also accounted for about 60 percent of all contract awards to Pennsylvania during 1962-64. Elec tronics contracts and missile and space con tracts each accounted for about 17 percent of total awards to Pennsylvania, and 5.0 percent and 2.4 percent, respectively, of total U. S. contract awards to these programs. In Penn sylvania, 13.2 percent of contract awards were for ships, constituting 7.4 percent of U. S. contract awards for this activity. Thus, Penn sylvania's top programs have been much more similar to the top programs in the U. S. than those of Ohio, and probably accounted for that State receiving a greater dollar vol ume of contract awards than did Ohio in fiscal 1965. Defense-Oriented Industries. As indicated earlier, neither prime contract nor procure ment data gives an indication of subcontract ing, which could be substantial and possibly outweigh changes in prime contracts. One way to evaluate subcontracting is to consider the data on employment and payrolls in de fense-related industries—ordnance and ac cessories, aircraft and parts, communication equipment, electronic components, and ship and boat building and repairing (see Table VII). It should be kept in mind that not all of the output of these industries necessarily goes to defense, so that not all employment is defense-related. But some indication of the size and relative importance of these industries in Fourth District states is gained from the figures, which in turn suggests the capacity for defense work. TABLE VII Employment in Five Defense-Related Manufacturing Industries, United States and Fourth District States, 1958-1963 In Thousands of Persons Year As Percent o f United Sta te s Total Fourth Fourth United District District States States States W e st Pa. O h io Ky. Va. 1958 1,551 138 8 .9 % 4 .6 % 3 .9 % 0 .3 % 0 .1 % 1959 1 ,6 2 4 145 8 .9 4 .6 4 .0 0 .3 0.1 1960 1 ,6 0 5 137 8.5 4 .0 4 .2 0 .3 0.1 1961 1 ,6 4 3 143 8 .7 4 .0 4 .4 0.3 0.1 1962 1 ,7 5 5 157 8 .9 4 .0 4 .6 0 .2 0.1 1963 1 ,7 4 7 158 9 .0 4.1 4 .5 0 .3 0.1 1964 1 ,6 7 7 156 9 .3 4.1 4 .8 0 .3 0.1 N O T E: Figures include all w a g e an d sa la r y em ploym ent in the five industries (not only that portion e n g a g e d in d e fen se w ork), but e xclu d e proprietors, se lf-em p lo yed , an d unpaid fam ily w orkers. U. S. em ploym ent fig ures a r e establishm ent d a t a , while figures for Fourth District states reflect em ploym ent co v ered b y state unem ploym ent insurance program s. Sources: U. S. D epartm ent of Lab o r, Bureau o f Lab o r Statistics; D epartm ent o f Employm ent Security, W e st V irg in ia ; D epartm ent of Eco nomic Security, Kentucky; Division o f Research an d Statistics, O h io Bureau o f U nem ploym ent C om pensation; Bureau of Employm ent Secu rity, Pennsylvania 9 ECONOMIC REVIEW While both Ohio and Pennsylvania have considerable employment in the five defenserelated industries, total employment in these industries has grown in Pennsylvania but not in Ohio. Accordingly, as shown inTableVII, the proportion of defense-related employment in the U.S. has grown for Pennsylvania but fallen for Ohio.5 Broadening the scope to cover employment in the five defense-related industries plu s DOD military and civilian employees in each of the Fourth District states, it is found that larger numbers of the latter make Pennsyl vania by far the leading defense employer in the Fourth District, with Ohio second and Kentucky a distant third. DOD military and civilian employment has been substantial in each of the states, ranging from somewhat less than 50 percent of defense employment (five industries plus DOD) in Ohio to much more than half for Kentucky.6 On balance, between 4.0 and 4 .5 percent of total nonagricultural employment in Ohio and Pennsylvania has been accounted for by defense employment (see Table VIII). The higher proportion of nonagricultural employment accounted for 5 The dominant defense-related industry in Ohio is air craft and aircraft parts, which has been plagued by the decline in the demand for conventional aircraft. From 1 9 5 8 to 1 9 6 4 , employment in the aircraft industry in Ohio declined by over 2 3 ,0 0 0 (from 5 8 ,1 0 0 to 3 4 ,8 0 0 ). Approximately 2 5 percen t of the decline, however, resulted from a reclassification of employees to the ord n ance industry in 1 9 6 1 . The dominant defense industries in Pennsylvania since 1 9 5 8 , in terms of employment, are electronic components and accessories, followed by aircraft and parts and communication equipment. 6 A breakdown of employment by defense-related in dustry and DOD for 1 9 5 8 to 1 9 6 3 is available upon request to the Research Department of the Federal Reserve Bank of Cleveland, Cleveland, Ohio. Digitized for 10FRASER TABLE VIII Employment in Defense-Related Industries Plus Department of Defense Military and Civil ian Employment, Fourth District States, 19581964 As Percent of N onagricultural Employm ent in Each State W e st Year O h io P a. Ky. Va. 1958 ... 4 . 4 % 4 .1 % 9 .8 % 0 .6 % 1959 ... 4 .4 4.1 9 .3 0 .6 1960 ... 4 .0 4 .0 9 .6 0 .6 ... 4.1 4 .2 9 .4 0 .5 196 1 1962 ... 4 .3 4 .5 1 0 .6 0 .6 1963 ... 4 .3 4 .4 9 .3 0.8 1964 ... 3 .9 4 .3 9 .4 0 .7 N O T E: Employm ent in five d e fe n se -re la te d industries represents em ploym ent co v ered b y state unem ploym ent insurance program s. Total nonagricultural em ploym ent b y states is b a se d on establishm ent d a ta . Sources: O ffic e of S e c re ta ry o f D e fen se ; N atio n al Aeronautics an d S p a c e Adm inistration; D epartm ent o f Employm ent S ecurity, W e st V irg in ia ; D epartm ent o f Economic Secu r ity, Kentucky; Division o f R esearch an d Statistics, O h io B ureau o f Unem ploym ent Com pen sation; Bureau o f Em ploym ent Security, P ennsylvania by defense employment in Kentucky reflects the large military bases in that State. Payrolls for employment in the five defenserelated industries and DOD military and civil ian employment as a percent of personal in come in each of the Fourth District states are presented in Table IX. On average, in recent years about 3.5 percent of personal income in Ohio and 3.3 percent in Pennsylvania can be attributed to defense activities. A somewhat higher figure for Kentucky reflects the pres ence of the military bases referred to earlier. In general, both employment and income data suggest that, with the possible exception of Kentucky, the direct effects of defense acti vity, while perhaps considerable in an absolute sense, are of a relatively minor order of im portance when compared with total employ ment and personal income in the Fourth Dis- MARCH 196 6 TABLE IX Wage and Salary Payments of Department of Defense and Defense-Related Industries in Fourth District States, 1958-1964 As Percent of S ta te Personal Income in Each Sta te W e st Year O h io Pa. Ky. Va. 1958 .... 3 . 6 % 1959 .....3 .6 2 .7 % 5 .5 % 0 .5 % 3.2 5 .3 1960 .....3 .3 0 .6 3 .2 5 .4 0 .5 1961 .....3 .4 3 .4 5 .0 0 .4 1962 .....3 .5 3.4 5 .6 0 .3 1963 .....3 .5 3 .6 5 .0 0 .4 1964 .....3 .3 3 .6 5.1 0 .4 N O T E : W a g e an d s a la r y paym ents for em ploym ent in five d e fe n se-re late d industries is b a se d on em ploym ent co v ered b y state unemploym ent insurance program s. Sources: O ffic e o f S e c re ta ry o f D efen se; N atio n al Aeronautics an d S p a c e Adm inistration; D epartm ent o f Employm ent Security, W e st V irg in ia ; D epartm ent o f Economic Secu r ity, Kentucky; Division o f R esearch an d Statistics, O h io Bureau o f Unem ploym ent C om pensation; Bureau o f Em ploym ent Secu rity, P en nsylvania; O f fic e o f Business Economics, U. S. D epartm ent o f Com m erce, S u r v e y o f C u rre n t B u sin e ss, V o l. 4 4 , No. 8 (August 1 9 6 4 ), p. 1 6 trict states, as well as with corresponding figures for the U. S.7 Census Survey. Preliminary results from a supplemental survey conducted by the Bureau of the Census during the 1963 Census of Man ufactures sheds some additional light on indi rect effects of Government procurement. The survey covered 30 defense-oriented indus tries that reported on shipments to Federal agencies and their contractors and suppliers in 1 9 6 3 .8 While the preliminary report does not present all the desired data and only covers one year, it does provide an additional, though limited, basis for judging the impact of defense procurement in Fourth District states. Unfor tunately, since the survey was structured to fit established Census categories, a breakout comparable to procurement programs, which would yield some idea of a state's productive ability to meet various defense programs, is not available. Shipments of surveyed defense-oriented firms totaled $40.3 billion for the U. S. in 1963. These firms employed more than two million persons. About $37 billion of the ship ments were accounted for by firms reporting Government shipments of over $100,000, and over half of the total shipments of these firms ($25.3 billion) were prime or subcontract shipments to various Government agencies. Table X presents basic data for the U. S. and the Fourth District states. As the table shows, in 1963, the value of shipments and employment of surveyed industries was greater TABLE X Government Shipments of Defense-Oriented Industries in 1963, by Prime Contract and Sub contract, for United States, Ohio, and Penn sylvania Shipments to G overn m en t A gen cies (in millions o f dollars) 7 See footnote 1. 8 Bureau of the Census, U. S. Department of Com m erce, S h ip m en ts Total C ontract Subcontract Em p loyed o f D efense - O riented In du stries in 1963, b y In d u stry, R egion, and Federal A gen cy, Summary Series, Preliminary Report, 1 9 6 3 Census of N um ber Prime A rea* $ 2 5 ,2 5 4 $ 2 0 ,0 3 0 $ 5 ,2 2 4 1 ,9 3 8 ,0 0 0 . 1 ,1 5 0 845 305 1 0 2 ,0 0 0 Pennsylvania 697 528 169 8 1 ,0 0 0 United S ta te s O h io . . M anufactures. The survey included the five defenserelated industries discussed here. It provides a more com prehensive estimate than just using the five indus tries alone, and also breaks down total shipments by type of customer. * Kentucky an d W e st V irg inia a r e not broken out. Source: Bureau o f the Census, U. S. D epartm ent o f C om m erce, S h ip m e n ts o f D e fe n s e - O r ie n t e d In d u s trie s in 1 9 6 3 b y In d u s try , R e g io n , a n d F e d e r a l A g e n c y (Sum m ary S erie s, Prelim inary Report), 1 9 6 3 Census o f M anufactures 11 ECONOMIC REVIEW in all categories for Ohio than for Pennsyl vania. This is consistent with the prime con tract data, but not with the estimates of defense employment. The Census survey covers more industries and excludes the Department of Defense, though, and this could account for the difference in results. With reference to shipments by customer category, surveyed firms in Ohio and Penn sylvania distributed a smaller proportion of total shipments to the Federal Government than did firms in the U. S. as a whole (not shown in table). Ohio firms did contribute a slightly larger share of shipments to the Gov ernment through subcontracts than was the case for either Pennsylvania or the U. S. In dollar terms, the bulk of Ohio's Government shipments were through prime contracts (73.5 percent) as compared with subcontracts (26.5 percent). Shipments to the Government under prime contracts of all U. S. surveyed firms amounted to 79.3 percent, with the cor responding figure for Pennsylvania at 75.8 percent. Impact Study. A recent study conducted for the DOD provides 1963 estimates of employ ment, value of shipments, and value added of Government shipments for industries that were not covered by the 1963 Census study of defense on the economies of Ohio and Pennsylvania. The available statistics resulting from the two studies are brought together in Table XI. The estimates in the table indicate that $15.8 billion of value added in manufacturing in the U. S. in 1963 was committed to defense uses. This figure represented 8.3 percent of total value added in manufacturing in that year, and was very close to the proportion of GNP accounted for by defense purchases, or 8.6 percent. Of the total value added in manufacturing in Ohio in 1963, 5.3 percent was in defense uses, while in Pennsylvania the correspond ing figure was 4.1 percent of total value added. In both of these cases, the proportion was below the corresponding national figure. Looking at it another way, Ohio's share of total value added in the U. S. used for defense purposes was 5.2 percent, while Pennsyl vania's share was 3.7 percent. Again, the figures were less than the corresponding national figure. shipments.9 Taken together, this study and the Census survey provide a more complete basis for assessing the impact of defense spending Fourth District states. Since the Census survey and the DOD impact study are available only for 1963, they cannot be used as a check on prime contract awards or employment in the five defense industries in other years. Gener ally, however, the data appear to be mutually consistent. Thus, in 1963, firms in Ohio re ceived more prime contract awards and shipped a larger dollar volume and a larger proportion of their total shipments and value 9 E conom ic Im p a ct S tu d y o f 1963 D efen se-R elated P rocu rem en t o f S elected Products and Services, by C-E-I-R, Inc., Technical Report No. 3, Part I, November 1 9 6 5 , and E con om ic Im p a ct Study o f 1963 D efen seR elated P rocu rem en t o f Selected Products and Services: Source and M e th o d o lo g y , T echnical Report No. 3, Part II, November 1 9 6 5 . The read er is referred to these studies for the methodology involved. 12 CONCLUDING COMMENTS In view of the various measures of defense ac tivity (impact) used in this article, it is legitimate to inquire whether any consensus emerges as to the role of defense spending in the TABLE XI Selected Measures of the Economic Impact of Defense in the United States, Ohio, and Pennsylvania, 1963 V a lu e o f Shipments Employment to G overnm ent V a lu e A d d e d (in millions o f dollars) (in millions o f dollars) Census D epartm ent of Census D epartm ent of Census D epartm ent of Su rv ey D efen se Survey Total S u rvey D efense S u rvey Total S u rv ey D efense S u rvey Total . 1 ,9 3 8 ,0 0 0 2 5 1 ,7 0 5 2 ,1 8 9 ,7 0 5 $ 2 5 ,2 5 4 $ 5 ,2 4 6 $ 3 0 ,5 0 0 $ 1 3 ,5 4 3 ° $ 2 ,7 6 4 $ 1 5 ,7 6 4 O h i o ............................ 1 0 2 ,0 0 0 9 ,1 9 3 1 1 1 ,1 9 3 1 ,1 5 0 323 1 ,4 7 3 688° 134 822 5 .2 5 .3 8 1 ,0 0 0 1 2 ,5 1 8 9 3 ,5 1 8 697 391 1 ,0 8 8 420° 158 578 3 .7 4.1 A re a United S ta te s Pen nsylvania . . . . . Percent^ P ercentc 1 0 0 .0 % 8 .3 % a Estim ated b y F e d e r a l R eserve Bank of C le v e la n d on basis of census surveys. Estim ates rounded to millions o f d o llars. b V a lu e a d d e d o f d e fen se shipm ents o f states a s a percent of total v alu e a d d e d of United S tates defen se shipments. c V a lu e a d d e d o f d e fe n se shipm ents as a percent o f total value a d d e d in the United Sta te s or the states a s listed. Sources: B ureau o f the C ensus, U. S . D epartm ent of Com m erce, S h ip m e n ts o f D e fe n s e - O r ie n t e d In d u s trie s in 1 9 6 3 b y In d u s try , R e g io n , a n d F e d e r a l A g e n c y (Sum m ary Series, Prelim inary Report), 1 9 6 3 Census o f M an u factures; and E c o n o m ic Im p a c t S tu d y o f 1 9 6 3 D e fe n s e - R e la te d P ro c u re m e n t o f S e le c t e d P ro d u c ts a n d S e r v ic e s , b y C-E-l-R , Inc., Techn ical R eport N o. 3 , P art I, N ovem ber 1 9 6 5 TABLE XII Selected Measures of the Economic Impact of National Defense on Fourth District States, 1963 Five D efen se -R e lated Industries Census Survey and D epartm ent o f D efen se S urveys an d D epartm ent o f D efen se Employm ent in G overnm ent V a lu e A d d e d of S u rv e y e d Industries Employm ent W ages Contracts D efense Shipments a s Percent o f a s Percent of an d S a la rie s as Percent of as Percent of Total N onagricultural Total N onagricultural a s Percent of V a lu e A d d e d * S tate V a lu e A d d e d Employment Employm ent Personal Income O h i o ................................................. 9 .8 % 5 .3 % 3 .5 % 4 .3 % 3 .5 % P e n n s y lv a n i a ................................. 7.1 4.1 2 .5 4 .4 3 .6 K e n t u c k y ............................................ 5 .7 9.3 5 .0 W e st V i r g in i a ................................. 8 .9 0.8 0 .4 S ta te * Includes m ilitary prim e contracts (including research), N ational A eronautics and S p a c e Administration prim e contracts and Atomic E n e rg y Commission costs incurred. Sources: TA BLES I, III, IV, VIII, IX, an d Bureau of the Census, U. S. D epartm ent o f Com m erce, A re a Series for O h io , Pennsyl v a n ia , Kentucky, an d W e s t Virg inia, 1 9 6 3 C en su s o f M a n u fa c tu re s , Prelim inary Reports ECONOMIC REVIEW added to the Government than did firms in Pennsylvania. There is some conflict in the employment picture when the Census and DOD studies are compared with data on DOD personnel and employment in the five defenseoriented industries. However, at least a por tion of this can be explained by differences in industry coverage and the inclusion of military and civilian personnel employed by the De partment of Defense. The fact that employ ment in Pennsylvania in the five defense in dustries exceeds Ohio employment in these industries is further indication of Ohio's lim ited participation in industries of major sig nificance to the defense posture. Ohio's greater overall defense-related employment implies that Ohio is involved in a number of defense areas, but not necessarily those of top priority as determined by overall spending. Table XII presents a summary of various measures of the impact of defense on the individual states in 1963. Government con tracts (military, NASA, and AEC costs incur red) are related to each state's value added for 1963, and a similar computation is made for the value added estimates of shipments of de fense industries covered in the 1963 Census of Manufactures and the DOD study. While Government contracts were less than 10 per cent of value added in all of the states, these ratios very likely overstate the significance of defense, as suggested by the value added relationships (see column 2). This is due to the fact that prime contract data include sub contracting which may be done in other states. While it is also likely that firms in each of the states receive subcontracting opportunities from other areas, the magnitude of subcon tracting into and out of a state, and hence the 14 net residual, is not known. In addition, although prime contracts and shipments are here re lated to value added, it would seem more appropriate to relate them to the value of shipments for a state. Such data are not avail able, however, on a state basis. If they were available, the shipments figures would be larger than the value-added figures, thereby lowering the percentages shown in the table. Furthermore, prime contract awards extend beyond the manufacturing sectors (services ranked fourth in the U. S. procurement pro grams shown in Table VI), so that a larger base than manufacturing value added should be used, which would further lower the state ratios. (This shortcoming is more pertinent to prime contracts than to the value-added esti mates.) With the exception of Kentucky, which has a large number of military personnel, various measures of defense employment represents less than 5 percent of total nonagricultural employment in the states. It should be noted that, since few defense industries sell more than half of their output to the Government, a comparison with total employment in these industries considerably overstates the direct impact of defense. If Department of Defense military and civilian personnel are added to the employment of defense industries sur veyed in the 1963 Census of Manufactures and the DOD study, a larger number of De partment of Defense civilian employees brings the total in Pennsylvania to slightly more than the comparable figure for Ohio, but only to 4.8 percent and 5.4 percent of total nonagri cultural employment in Pennsylvania and Ohio, respectively. MARCH 1966 Wage and salary payments to Department of Defense personnel and employees of the five defense-related industries are either equal to or less than 5 percent of total personal in come for each of the Fourth District states. In summary, given the limitations surround ing the data and the methods, it is likely that the direct impact of defense activity absorbs less than 10 percent of the resources of each of the states of the Fourth District. A more likely range would seem to be 3 to 7 percent of re sources in Ohio and Pennsylvania, with a possibly somewhat higher range for Kentucky due to the military installations in that State. Defense activity is of minor importance to West Virginia. Indirect and induced effects would expand the significance of the direct defense impact, but little information is avail able on this aspect.10 Finally, given data limitations, any attempt to predict the near-term or future impact of national defense on individual states should be viewed as highly tentative. Nevertheless, it may be appropriate to make some observa tions on the possible influence of the Vietnam buildup on Fourth District states. Since Fourth District states have experienced a relative decline in defense work in recent years, support is found for the thesis that a changed defense posture requiring different 10 See the discussion in the earlier article, including Table III, which presents the indirect employment for certain defense-oriented industries. The ratio of indirect to direct can be considered as a rough indication of the maximum indirect effect. Not all of the industries re corded in the table are of importance to all of the states; types of defense goods was not conducive to increasing amounts of defense work in Fourth District states. In short, the industrial base of Fourth District states is suited to the produc tion of conventional types of defense goods but, with the apparent exception of Penn sylvania, not to missile and electronic com ponents production.11 The advent of limited conventional wars, such as Vietnam, has increased demands for more conventional military goods: rifles, am munition, smaller aircraft, and vehicles. Many of these military goods have traditionally been produced in the Fourth District states. Even though conventional weapons and warfare also undergo changes, for example, increased emphasis on the helicopter, it seems plausible that a shift to more conventional military goods might stem the relative decline in Ohio's share of total defense activity. The Vietnam buildup does not seem to portend any signifi cant changes in the share of defense work of other District states unless various procure ment programs were to be cut to release funds for essential Vietnam requirements. Such an event, for example, might impinge upon Penn sylvania's share of U. S. defense work, al though stepped-up procurement for conven tional war goods could overcome any such decline. The remaining factor that could exert an impact on District states is the future de ployment of military personnel and other DOD employees. The closing of various military bases, plus consolidation, makes it difficult to assess this aspect of the defense impact. much of the indirect employment can be generated in 11 A third and final article will consider selected aspects other regions. of defense-related industries in Fourth District states. 15 ECONOMIC REVIEW DIMENSIONS OF SUBCYCLICAL FLUCTUATIONS IN STEEL AND AUTO OUTPUT The 1965 episode of sharp fluctuations in steel output, which were associated with strikehedge inventory building, has somewhat gent ly receded into history. An opportunity is thus afforded to review the possible meaning and magnitude of that episode in the favored per spective of hindsight. When the 1965 experience is considered alongside earlier, similar episodes, a pattern of repeated marked fluctuations in steel out put is quickly apparent. (See, for example, the 1962 and 1963 experiences as plotted in the top line of Chart 1.) At first glance, the wave like repetitions are so pronounced as to seem almost seasonal in character, even after ad justment for seasonal variation, but of course that is not the case; in a statistical sense, the gyrations are irregular or random, as con trasted to seasonal and cyclical. The inven tory-induced swings in steel output in several recent years perhaps can be best character ized as "subcyclical," in order to suggest a more than passing degree of instability that involves serious consequences for the steel industry itself as well as broader implications Digitized for 16FRASER for the course of economic activity, even though not necessarily of general businesscycle proportions. It should be pointed out that, in the se quences of steel output fluctuation, the auto industry has played quite a significant part. That is so, of course, because the auto indus try is the largest, or second largest, steel cus tomer industry, depending on the measure ment used. Hedge-buying of steel by the auto industry, under circumstances where future supplies are threatened by a crisis arising in labor-management negotiations in steel, is customarily a strategic factor in the repeated episodes of fluctuations in steel output. In 1965, for instance, the decision of auto manu facturers to order considerably enlarged sup plies of steel was made explicit as early as the fall of 1964, or well before the (then) May 1, 1965 deadline of the steel labor contract. With other steel customer industries, in vary ing degrees, subsequently following the lead of auto hedge-buying, the steel industry, as supplier, was of course in no position to turn down the orders, even if it had wanted to. As it MARCH 1966 turned out, and without ascribing any motives, the auto industry last year had a phenomenally good year in output and sales that was achieved at the expense of drastic fluctuations in steel activity, accompanied by complications in the general business and economic outlook.1 Over a long span, including recent years, the auto industry has, of course, had its own share of fluctuations in output. While most of these have been independent of the connec tions between the auto and steel industries, instances of temporary cutbacks in auto out put because of steel shortages stemming from actual steel strikes occurred in the 1950's, with the latest in 1959. In addition, the auto industry has had its own strikes, actual and not just threatened, which have had rever berations on the pace of industrial production generally, the last instance occurring in the late months of 1964. When taken together, subcyclical fluctu ations in auto and steel output have had ob servable effects upon the economy in general and possibly even greater repercussions by way of magnifying the problems encountered by public policymakers and economic fore casters. Fortunately, as events have turned out in recent years, even the most severe of the downward fluctuations, as in steel, have usually been more than offset by the effects of steady growth in other sectors of the economy. (Autos and steel together account for about 10 percent of the Index of Industrial Produc 1 Another interconnection between the fortunes of the auto and steel industries in 1 9 6 5 was by way of the background of w age negotiations in the two industries. The w age settlement that ended the 1 9 6 4 auto strikes is generally understood to have played a part, although an unm easurable one, in the framing of the target demands of steel labor in 1 9 6 5 . tion.) Nevertheless, this does not mean that the fluctuations have been unimportant. If, for example, the sharp cutbacks in steel out put in the autumn of 1965 had happened to occur within a setting of generally sluggish growth of the economy (or perhaps if the auto industry had been simultaneously faltering), the final results would have been serious. Furthermore, in addition to the direct impacts upon domestic production activity and the influence upon business expectations regard ing the economic outlook, the steel fluctuations have stimulated steel imports, with conse quences for the nation's balance of payments problem taking on proportions more than those of a casual side-effect. Against this background, the discussion that follows is intended to help clarify the various dimensions involved. PERSPECTIVE ON FLUCTUATIONS SINCE 1 9 5 7 The sweep of fluctuations in auto and steel output from 1957 through 1965 is shown in Chart 1. A comparison of cyclical and subcyclical fluctuations of the two industries is afforded by considering on the chart the dis tinction between the marked plunges due to the two cyclical recessions of 1957-58 and 1960-61, on the one hand, and the less sharp subcyclical fluctuations, on a rising growth trend, from 1961 through 1965. The line for auto output in Chart 1 reveals a cyclical trough in 1958, another very low level in late 1959 caused by interruption of steel supplies at the time of the long steel strike, a cyclical drop of moderate proportions between late 1960 and early 1961, then an irregular upward movement, interrupted by 17 ECO N O M IC REVIEW 1. PERSPECTIVE on AUTO O U T P U T and STEEL INGOT PRODUCTION Annual rate (m illions of tons) Annual rate (m illions of cars) S o u rce s of d a ta : W a r d ’s ; A m e rica n Iron and S te e l In stitu te an unusually concentrated compression of the model changeover reduction in August of 1963, and a steep drop in October 1964 due to auto strikes in two of the important com panies. In terms of annual totals, as shown by the horizontal bars, a steady climb in auto output after 1961 is apparent, with 1965 out put reaching the extraordinarily high level of Similarly, the checkered course of the line for steel ingot output reveals a sharp cyclical drop in steel output in late 1957 and early 1958, followed by a burst of recovery, includ ing steel inventory buildup anticipating the contract termination in 1959. During the pro tracted steel strike of 1959, output dropped by about 85 percent (not shown in chart). The 9.3 million cars. The latter may be compared with the previous record of 7.9 million cars in 1955 (not shown in chart). very rapid recovery of steel output by Decem ber of 1959 (see high point) was built on a scramble for replenishment of steel stocks, Digitized for18 FRASER MARCH 196 6 carrying the December 1959 output figure up to about a 150-million ton annual rate. The wave broke at that point and the subse quent liquidation of steel inventories in 1960 ran into a cyclical recession that carried out put all the way down to about a 70-million ton annual rate by the end of 1960. That partic ular inventory liquidation episode was part cause and part effect of the recession of 1960-61. From 1961 through 1965, there was a gen eral upward movement in steel output, despite the three serious subcyclical interruptions already mentioned (1962, 1963, and 1965). The basic pattern in all three episodes was the same: strike-hedge buildup, followed by liqui dation after the labor-management settlement. The fact that the 1963 development was in response to a contract re-opener clause, rather than an outright expiration, is of no material consequence, although it helps to explain how the episodes came to be so frequent in the face of a general industry custom of three-year contracts. In terms of annual totals, as shown by the horizontal bars in Chart 1, steel output in 1965 was at the unexpectedly large figure of 131 million ingot tons. The achievement reflected the prolonged inventory buildup lasting until September 1, which was the date of the final contract settlement. The final settlement, it will be recalled, followed the interim agreement that had been reached on May 1 in order to allow an extension of time for negotiations under the conditions of change over in the union administration. (The magni INTRA-YEAR PATTERNS Further perspective on fluctuations in auto and steel output in recent years may be ob tained by focusing attention on changes in output within individual years. Month-tomonth fluctuations in output of autos (see Chart 2) and of steel (see Chart 3) are shown for selected years, without seasonal adjust ment. By this device, it is possible to obtain a view as to the manner in which more-or-less usual seasonal fluctuations take place in the two industries, concurrently with the episodic or subcyclical fluctuations that are under con sideration. (Such a view cannot be attained through use of data that are already seasonally adjusted, as is the case with Chart 1.) Intra-year patterns of auto output in 1955 (the record year until last year), and for 1964 and 1965, are depicted in Chart 2. For all three years, the spring months are outstand ing for auto output, while sharp drops occur 2 . AUTO OUTPUT: Intra-year Patterns Selected Y ears tude of the steel inventory liquidation occur ring in late 1965 is discussed at a later point.) 19 ECONOMIC REVIEW 3. STEEL INGOT PRODUCTION: Intra-year Patterns Selected Years M illio n s o f tons 14 13 12 11 10 9 8 So u rce of d a t a : A m e ric a n Iron a n d Steel Institute in August (or in the case of 1955, in Septem ber) at the time of retooling for the new models. The upsweeps in the last few months of the year, as the new models are swinging into production, do not characteristically bring output levels back to the high spring levels. In the case of 1964, the sharp, counterseasonal drop in October was, of course, ascribable to the auto strikes affecting two of the major manufacturers. The line for 1965 is notable for the sharp climb in output in the later months of the year, reflecting a firm confidence on the part of auto manufacturers that the un usually strong levels of auto sales which had been achieved in the early part of the year would be continued. The risk that year-end stocks of cars in dealers' hands might be larger than usual was apparently accepted by the industry's planners. Correspondingly, intra-year patterns of steel ingot output are shown in Chart 3. The upper panel depicts the monthly course of output in 0 1965 against a backdrop of the course of out put in 196 4 —a year that was not affected by inventory fluctuations accompanying strike possibilities, except to the extent that some inventory building of steel in anticipation of the 1965 contract question was already occur ring in late 1964. The line for 1965 reflects the extraordinarily high levels of steel output in the early part of the year, continuing with hardly any diminution through the period of the "interim contract," with the liquidation phase apparent during September through November, and the beginnings of an output upturn exhibited in December, despite the fact that inventory liquidation was still pro ceeding. The bottom panel of Chart 3 shows the intra year patterns of steel output for the years 1962 and 1963, with reductions in output, follow ing the contract settlements, becoming appar ent in April of 1962 and June of 1963. The contract settlements occurred, respectively, in March 1962 and May 1963. Of the four years depicted in the two panels of Chart 3, not a single year, with the possible exception of 1964, could be regarded as exhibiting any thing like a "normal" intra-year pattern. In each case the upsetting factor was the impact of inventory gyrations upon output rates.2 A direct comparison of the magnitudes of decline in steel output during the episodes of 1962, 1963, and 1965, as measured from a common starting point in the form of output in the month immediately preceding the de 2 Significant seasonal adjustment factors, even though based on moving averages, are dependent on at least some d egree of regu lar seasonal rhythm. W hen the data for several successive years are upset by larg e erratic swings, it becom es impossible to sort out the seasonal influences with any d egree of confidence. MARCH 1966 cline, is shown in Chart 4. In each case, monthly output in ingot tons, unadjusted for seasonal variation, is converted to an index, with the output of the month immediately before settlement taken as the base (100). Declines are then registered in index number form for each of the succeeding months in volved in the downswing, followed by indexes for the first few months of recovery from the trough. It can be seen from Chart 4 that the decline during the 1965 episode was substantial, amounting to 22 percent between August and November 1965. However, the 1965 decline was somewhat less sharp than those of 1 962 and 1963 and, also, the pace of recovery, during the first few months for which data are presently available, was somewhat more pro nounced than in the other two years.3 The question obviously arises as to why the 1965 fluctuations were somewhat less serious than the 1962 and 1963 affairs, remember ing that, while the main events were in process, it was not possible to know that such would be the outcome. Was the relative moderation in 3 The 1 9 6 0 episode (not shown in the chart) was similar in some respects to the showings of the three years de picted in the chart, but guite unlike them in other re spects. In that episode, the month of peak output was January 1 9 6 0 , which did not coincide with the time of the labor-m anagem ent settlement, but rather o ccu rred after the scram ble to replenish steel stocks in late 19 5 9 , following the depletion of such stocks during the long steel strike. The rate of decline in steel output after January 1 9 6 0 was not very different from the rates shown in the chart, but it lasted much longer. Thus, the period of decline stretched out for six months and was followed, not by recovery, but by five months of p ra c tically unchanged output at low levels. The key to this showing, of course, lies in the fact that the 1 9 6 0 episode o ccu rred in the setting of a general business recession. 4. STEEL OUTPUT FOLLOWING THREE CONTRACT SETTLEMENTS Selected Y ea rs INDEX of m onthly output S o u rce of d a t a : A m e ric a n Iron a n d S te e l Institu te 1965 because the strike-hedge buildup of steel stocks had been less pronounced? The answer is no, as is elaborated below. As a general matter, the difference lay in the boom like phase of general economic expansion that characterized the year 1965, and espe cially its later months, with consequences both for enlarged steel consumption and for a turn toward less conservative inventory policies on the part of steel-consuming industries. The full impact of that general expansion can perhaps only be appreciated by hindsight, resting as it did upon an amalgam of a fastmoving capital expansion program, enlarged consumer takings, and a marked turn in late 1965 toward stepped-up defense activity re lated to Vietnam. 21 ECONOMIC REVIEW HOW MUCH BUILDUP IN STEEL STO CKS? How large were the strike-hedge buildups in the 1962, 1963, and 1965 episodes? It should be pointed out candidly that not only was a satisfactory answer to this question un available at the time of the developments in question, but the question cannot even now be answered with any degree of precision, notwithstanding all the advantages of hind sight. In short, there are no a d eq u a te statistics on the quantities of steel held by consumer industries at any one time. There are, of course, some data that go part way toward answering that important question. The Bureau of the Census, of the U. S. Depart ment of Commerce, publishes estimates (some time after the event) of steel stocks in the hands of manufacturers as well as stocks at the steel mills themselves. (The series began 5. INVENTORY of STEEL MILL SHAPES Selected Yeors M illio n s of tons RATIO of STEEL INVENTORIES to STEEL CONSUMPTION in MANUFACTURING 1962 So u rce of d a ta : 1963 1964 1965 1966 B u re a u of C e n s u s , U .S. D e p a rtm e n t of C o m m e rce in late 1960.) These data reach only part way, however, because of the admitted fact that they do not include information on steel stocks in the hands of construction firms, railroads, utilities or pipe lines. Furthermore, the data do not include steel that is in the course of being processed at various manufacturing plants. (The "goods-in-process" reappear in another form in governmental statistics on inventories, but the steel component is not separated out.) Even by the showings of the statistical series just described as being inadequate, however, So u rce of d a t a : B u r e a u of C e n s u s , U .S. D e p artm en t of Co m m e rce Digitized for22 FRASER it now appears quite clear that the buildup of steel stocks in 1965 was a relatively large one. That is shown by Charts 5 and 6, which pre sent the results of the Bureau of the Census series on steel inventories. Chart 5 shows the estimated tonnages of steel-mill shapes held MARCH 1966 both by manufacturers and by the steel mills. Chart 6 shows the estimated tonnages in man ufacturers' hands, expressed as a percent of current steel consumption rates by the same manufacturers. On either count, the buildup in 1965 was relatively large.4 Because of the lag in availability of statistics, which is not unique with these series, even the partial data shown in Charts 5 and 6 were not available early enough to accommodate the needs of business and governmental analysts who were grappling with the question of how large the buildup actually was. A second kind of estimate of the magnitude of steel stocks can be attempted by comparing actual steel shipments with estimates of steel consumption derived from statistics on pro duction activity of steel-consuming industries. Such estimates made during the 1965 episode seemed to point toward an unusually large buildup of steel inventory, at least consider ably in excess of the buildups of 1962 and 1963. However, in forecasting how many months would be required to work down the accumulated stocks during the liquidation phase, and how steep the corresponding pro 4 At first glance, C hart 5 seems to indicate a buildup in 1 9 6 5 that was relatively of about the same proportion as in 1 9 6 3 and 1 9 6 2 , although at higher levels of tonnage. Taking into account, however, the start of the buildup in duction cuts would have to be, analysts who were using this approach were faced with forecasting the general trend of business, as well as possible changes in inventory policy on the part of steel consumers. Estimates of the length of time during which steel output would be seriously curtailed during the inventory-liquidation phase varied all the way from one or two months to three calendar quarters. As events turned out, many of these estimates gave insufficient weight to the strength of demand forces in the economy generally— a problem that was not uncommon to economic forecasts in general for 1965. But the fact re mains that if more adequate statistics on the amount of the buildup had been available, the areas for errors in judgment would have been markedly reduced. EFFECTS ON STEEL IMPORTS One by-product of the phenomenon of strikehedge accumulation of steel inventory is the stimulus that such buying gives to imports of steel, on top of other factors that may be work ing in the same direction. The importance of such a development, both to the steel industry and to the national welfare via the balance of payments problem, is perhaps understated by use of the term "by-product," although the latter is technically correct. the fourth quarter of 1 9 6 4 (not shown in the chart), the 1 9 6 5 buildup should be considered as markedly longer The top panel of Chart 7 shows annual im than the earlier ones and both relatively and absolutely ports and exports of steel, beginning in 1950 and expressed in terms of dollar values. The rise in imports since 1958 is quite striking. There was a small excess of imports over ex ports in 1963 and a very large one last year. larger in the ag g reg ate. Such a view would be confirmed by C hart 6. In that case, however, a discount in the other direction should be mentioned. The sharp peak of the ratio in July and August of 1 9 6 5 should be interpreted partly in the light of the seasonal downturn in steel re quirements by the auto industry during the retooling period. The data in Chart 6 are not adjusted for seasonal variation. The shaded areas of the chart indicate the years in which imports were boosted by a 23 ECONOMIC REVIEW 7. U.S. EXPORTS and IMPORTS of STEEL STEEL S T R IK E B illio n s of d o lla rs TH REAT TH R EA T \/ TH REAT steel mill products reached about 1 0 ^ mil lion tons. The latter amounted to about 10 per cent of total shipments of finished steel, in cluding imports; thus, if it had not been for the imports, the demand for domestic output of steel might have been 10 percent larger. M illio n s o f tons 1950 '52 '5 4 '56 '5 8 '6 0 '62 '6 4 '66 * P r io r d a t a not a v a ila b le S o u rce of d a t a : If steel exports and imports are measured in tonnage, as is done in the bottom panel of Chart 7, the results give less emphasis to the higher-value product lines in steel, and the rise in imports becomes even more striking. On this basis, there has been an import surplus in every year since 1958. In 1965, imports of U .S. D e p a rtm e n t of C om m erce steel strike or a threat of a steel strike, result ing in the inventory building previously dis cussed. In the intervening years, 1960, 1961, and 1964, the chart confirms that some of the import bulge becomes permanent, that is, imports do not recede to their previous posi tions. Thus, the domestic steel industry has lost valuable ground to imports, despite inten sive efforts in recent years to improve its competitive position through large-scale (and expensive) programs of modernization. From the standpoint of the nation's balance of payments, the unfavorable impact upon the commodity balance brought about by the 1965 excess of steel imports over exports, amounting to more than half a billion dollars, is far from negligible. When measured against the target amounts involved in the national program to eliminate (or reduce to manage able proportions) the overall payments deficit, the steel item by itself represents an appreci able proportion. Digitized for 24 FRASER EFFECTS ON INDUSTRIAL PRODUCTION It was stated earlier that subcyclical fluctu ations in both steel and auto output, although significant, have frequently been counter balanced by opposite movements of other sectors of the economy so that there was no visible downpull in the Index of Industrial Production. In Chart 8, total industrial pro duction is plotted on the top line; below indus trial production, with its own scale but with the same scale units, is shown the part of the production index that is called "motor vehicles and parts," which accounts for 4.7 percent of the total index; and further below is shown the iron and steel component of the produc tion index, which accounts for about 53^ per cent of the total index. It can be seen from Chart 8 that in 1959, the steel strike pulled down auto output, and the declines in steel and auto output, taken together, had a marked impact on the total index. The recessionary declines in autos and steel in 1960 and early 1961 were also re flected in a decline in the production index. As soon as the lines are traced well into 1961 and from then on into the present, the MARCH 1966 8. With and Wit hout Autos and Steel INDUSTRIAL PRODUCTION: INDEX 1957-59=100 148.3 178.0 101.2 146.6 104.4 101.0 PRO DUCTION 1190 e x IRON & STEEL 105.6 150.0 148.5 100.9 IN D U STRIAL PRO DUCTION ex MOTOR VEHICLES & P A R T S, and IRON & STEEL 100.8 S E A S O N A L LY A D JU STED R A TIO SCI A LE 1959 '60 I '61 '62 '63 '64 I '65 '66 * S trik e p e rio d S o urce of d a ta : B o a rd of G o v e rn o rs of the F e d e ra l R e se rv e System story changes. The production index has gen erally been able to absorb the shocks of auto and steel fluctuations with little visible change unfavorable impact on the production index. in direction. Thus, auto strikes in the fall of 1961 pulled down the production index a trifle for only one month. The steel-inventory fluctuations of 1962 and 1963 did not pull down the production index at all, although there was a tendency to flatten it. The auto strikes of late 1964 had a short but visible 1965 entry for iron and steel, and that did produce a dip in the production index for that month. In October 1965, the total Index Finally, the beginning of the latest steel-inven tory liquidation is registered in the September of Industrial Production showed a partial re covery; and in November and December the production index pushed rapidly ahead to new highs, despite the steel-inventory liqui25 ECONOMIC REVIEW IMPACT OF CHANGES IN STEEL AND AUTO OUTPUT ON INDEX OF INDUSTRIAL PRODUCTION (Selected periods, 1 9 6 2 - 1 9 6 5 ) M otor Vehicles All O th e r To tal an d Steel and Parts Com ponents In d e x Iron Index, M arch 1 9 6 2 Index, July 1 9 6 2 ..................................................................... ........................................................................... Points change in index of component, 4 months . . . . . . 1 1 8 .5 1 2 6 .8 1 1 7 .0 1 1 7 .5 . . 8 7 .8 1 3 7 .5 11 9 .9 1 1 9 .0 . . Points change in total index, 4 m o n th s................................ Ind ex, M a y 1 9 6 3 .......................................................................... Ind ex, S ep tem b er 1 9 6 3 ........................................................... Points change in ind ex of component, 4 months . . . 3 0 .7 + 1 0 .7 — 1.6 + 0 .5 . . 1 2 9 .3 1 4 2 .5 . . 1 0 0 .0 149.1 . . Points change in total index, 4 m o n th s................................ Index, August 1 9 6 4 — + 2 .9 ■f 2 .6 — + 1.5 1 2 3 .2 1 2 4 .4 1 2 6 .0 1 2 5 .7 — 2 9 .3 + 6 .6 -f 2.8 — 1.5 + 0.3 + 2 .5 — + 1.3 ..................................................................... . . 1 3 3 .8 1 6 2 .4 1 3 2 .5 1 3 4 .0 Index, O c to b e r 1 9 6 4 ..................................................................... . . 1 3 2 .5 9 6 .2 1 3 3 .4 1 3 1 .6 Points chan ge in in dex of component, 2 months . . . — 1.3 — 6 6 .2 + 0 .9 Points chan ge in total in dex, 2 m o n th s................................ — 0.1 — 3.1 + 0.8 Index, July 1 9 6 5 ° .......................................................................... Index, N ovem ber 1 9 6 5 . . ........................................................... Points change in index o f component, 4 months . . . . . Points ch an g e in total index, 4 m o n th s................................ — — 2 .4 152.1 1 7 7 .4 1 4 2 .0 1 4 4 .2 1 1 3 .0 1 7 7 .6 1 4 6 .7 1 4 6 .3 — 39.1 — 2.1 + 0 .2 -0 - + 4 .7 + 4 .2 — + 2.1 a In this e p iso d e, July output o f steel w a s a ctu a lly the p e a k of the subcycle. A ccordin gly, July 1 9 6 5 is used here a s the starting point rather than August, which w a s the “pre-settlem ent month” used a s b a s e in C h art 4. N O T E: W eig h ts used for computing im pacts on the total index a re as follows: iron and steel: 5 . 4 5 % ; motor vehicles an d parts: 4 . 6 8 % ; all other components: 8 9 .8 7 % Source: Ind ex of Industrial Production an d its com ponent parts, B oard o f G o v e rn o rs o f the F e d e ra l R eserve System dation which was still reducing steel output in November, although not in December.5 To illustrate the points just made in another way, three other lines are plotted in Chart 8. One line shows the Industrial Production Index with motor vehicles taken out, another shows the index with steel taken out, and a final line shows the Industrial Production Index with both autos and steel excluded. When both industries are taken out of the 5 It should be borne in mind that it is possible for steel output to rise while liquidation of steel inventories is still proceeding. That o ccu rs under circum stances where current consumption of steel is large enough to require increased mill production, from previously reduced levels, in addition to the steel provided by consuming industry stocks. Such a situation apparently continued beyond D ecem ber 1 9 6 5 into the early months of this year. 26 production index, as in the final line, the re sult is rather mild fluctuations from 1959 to early 1961, and thereafter a truly handsome line depicting uninterrupted growth for five successive years. As a supplement to Chart 8, an accompany ing table shows measurements of the impacts of fluctuations in auto and steel output on the Index of Industrial Production, together with offsetting impacts provided by changes in other components, for selected periods. The four periods shown cover months of subcycli cal drops in steel output in 1962, 1963, and 1965, as well as two months of sharp drop in auto output at the time of the strikes in late 1964. Thus, for example, the table shows that the iron and steel component of the index dropped MARCH 1966 30.7 points between March and July 1962. Taking into account the relative proportion of the iron and steel component in the total index, the steel downdrag on the total index amounted to 1.6 points. Concurrently, how ever, the auto industry was accounting for a 0.5-point upthrust in the total index, while all other components accounted for a 2.6-point upthrust. On balance, the production index rose 1.5 points over the four-month interval. CONCLUSION Insofar as the foregoing has emphasized instability of output in the steel and auto in dustries, the purpose is to clarify the dimen sions of the problem rather than to exaggerate its seriousness. There is no intention to imply that this type of subcyclical fluctuation is cen tral to the problem of stabilization of the U. S. economy, although an attempt could be made to present such a case. It is, however, signifi cant and at the least might be characterized as of an intermediate degree of gravity. Cer tainly there have been times, notably during the year just past, when business and govern mental policies hinged to a considerable extent upon the evaluation of prospective sequences in this type of fluctuation. By their very nature, the fluctuations des cribed here are, at least partly, by-products of the collective bargaining process. It is stability are being reached, or even seriously modified. The present labor-management contract in steel, for example, expires on August 1, 1968. Unless something "turns up," which seems unlikely, at least some degree of repetition of the familiar sequences may be expected. This is said with full realization that many observers have come to expect that the influence of the Federal Government will usually be exerted in such a way as to forestall a strike in the steel industry. An important point to recall, in this connection, is that many business executives in 1965 did expect that a steel strike would be forestalled, but they were still not willing to risk the outcome by failing to take steps of hedge-buying. The inadequacy of available statistical data on steel inventories, in view of the general situation described earlier, is most unfortunate and should be emphasized, even though its correction is no easy matter. It is perhaps ironic that the general satisfaction expressed about recent achievements toward more scien tific and more conservative controls of inven tory in the general economy has been accom panied by recognition of serious inadequacies in inventory statistics in general, which have in turn aggravated difficulties in making sound appraisals of the business outlook. While use of electronic computers for inven area of collective bargaining procedures. But tory control may indeed be a step forward in certain respects, such a development has little or no relation to the problems here under discussion. Finally, to the extent that inventory-induced one observation may be hazarded: Despite public-spirited intentions and efforts in some quarters, there is no visible assurance on the horizon that the underlying causes of the in steel fluctuations have substantially aggra vated the balance of payments problem, there is one more important reason for considering these matters to constitute a problem area. beyond the scope of this article to make any recommendations or forecasts as to possible routes of amelioration of the problem in the 27 Fourth Federal Reserve District