The Puerto Rican Economy and the United States Economic Fluctuations (1962)

EXCESO DE COPll'-PARA DONA~t CPR UPFi RP

THE PUERTO RICAN ECONOMY ND UNITED STATES CONOMIC FLUCTUATIONS

Werner Baer

.\

EXCESO pE

C~P~~

C R PRNRP

r

THE PUERTO RICAN ECONOMY AND UNITED STAPES ECONOMIC FLUCTUATIONS

WERNER BAER Yale University

THE PUERTO RICAN ECONOMY AND UNITED STATES ECONOMIC FLUCTUATIONS

SOCIAL SCIENCE RESEARCH CENTER University of Puerto Rico Río Piedras - Puerto Rico

Printed in Spain

Impreso en España

Imprime: EDICIONES RVMBOS - Ramblas, 23 - Barcelona

TO MY PARENTS

ACKNOWLEDGEMENTS This study would been impossible to undertake without help from many sources. My deepest gratitude goes to Prof. J. K. Galbraith, without whose guidance, inspiration and trust I would never have begun .or finished this .project. I am especially in debt to Robert C. Repetto, of Harvard, for his intelligent and devoted assistance. The study was sponsored by the Social Science Research Center of the University of Puerto Rico. Dr. Millard Hansen, its director, provided invaluable help in getting the project under way, in making important contacts, and in providing advice and encouragement as the project progressed. I also wish to thank Carlos Lastra of the Center for his constructive advice and criticisms, and also Miss Margot P. de la Cruz for her painstaking work as editor. My gratitude goes to a number of economists in various government agencies for their generosity in giving me their time and the facilities of their departments. I would especially like to mention Hu Barton of EDA (it was thanks to his efforts that the material for chapter IV was collected), Alvin Mayne, the director of the Bureau of Economics and Statistics of Puerto Rico Planning Board, Vernon P. Esteves of the Government Development Ban, Felipe S. Viscasillas of the Department of the Treasury, Ruben A. Vilches of the Department of Labor and Juan Labadie Eurite. I wish to thank my friends. in Cambridge for their advice and critical comments on various parts of this study: Professor Arthur Smithies, Barbara Berman, Frank Fisher, Michel Hervé, Kenneth Kauffman. My thanks go also to the Harvard Economic Research Project for use of its facilities.

VII

TABLE OF CONTENTS Page

LIST of TABLES ....

...

...

LIST OF CHARTS

XI

... XIII

FOREWARD BY PROFESSOR J. K. GALBRAITH

...

...

XV

Chaptet

I. INTRODUCTION

1

II. CHANGES IN THE STRUCTURE OF THE PUERTO RICAN ECONOMY

9

III. THE POST-WAR RECESSIONS ... ... ... ... ... ... ... ... A Descriptive Analysis

... ... ... ... ... ... ...

33 33

The Behavior of Puerto Rican Financial Institutions over the Cycle ... ... ... ... ... ... ... The Sensitivity of Puerto Rican Exports and Investment to U'. S. Mainland Fluctuations

75

IV. THE BEHAVIOR OF PROMOTED FIRMS OVER THE CYCLE ...

83

V. PUERTO RICO'S ECONOMIC STRUCTURE AND THE POSSIBLE IMPACT OF UNITED STATES MAINLAND RECESSIONS ... ...

119

VI. SUMMARY OF FINDINGS AND RECOMMENDATIONS FOR COUNTER cYCICAI. PoLicIEs ... ... ... ... ... ... ... ... ...

141

G1

BIBLIOGRAPHICAL NOTE

152

BIBLIOGRAPHY

153

IX

LIST OF TABLES Table

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. .14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

Page

Gross Fixed Investment in Puerto Rico ... ... ... ... ... Puerto Rico's Income ... ... ... ... ... ... ... ... ... ... Puerto Rico's Deficits and Its Principal Finance ... Puerto Rico's Dependence on External Trade ... ... ... Puerto Rico's Export and Import Structure National Income by Industrial Origin Distributive Shares as Proportion of Net Income Average Hourly Earnings in Selected Puerto Rican In... dustries Unemployment in Puerto Rico ... ... Employed Workers by Industry Groups Sources and Uses of Capital Funds ... The Recession of 1953-54 — Comparative Data U. S. Industrial Production and Puerto Rican Exports — 1953.54 ... ... ... ... ... ... ... ... ... Employment, Unemployment, and Migration Labor Force and Employment by Industry The Recession of 1957-58 — Comparative Data ... ... ... U. S. Industrial Production and Puerto Rican Exports Employment, Unemployment, and Migration — 1957-58 Employed Workers by Industry Group ... ... ... Growth of Banking and Financial Institutions in Puerto ... ... Rico ... ... ... ... ... ... ... ... Puerto Rican and Mainland Bank Rates Quarterly U. S. Statistics Fluctuations During the Cycle Measurement of Sensitivities ... Effects of Mainland Variables on Puerto Rican Investments Tabulation of Plant Characteristics ... ... ... ... ... ... XI

13 16 16 18 19 21' 22 22 24 24 30 36 38 41 45 49 50 52 54 64 72 78a 79 80 81 87

Table

Pagc

27. Type of Ownership Distribution, by Industry Groups ... 28. Market Orientation According to Ownership Type, and Market Orientation According to Industry Group ... 29. Plant Size by Type of Ownership, and Plant Size by Industry Groups ... ... ... ... ... ... ... ... ... ... ..: 30. Decline in Monthly Man Hours by Industry Groups 31. Decline in Monthly Man Hours by Form of Ownership 32. Decline in Monthly Man Hours by Market Orientation ... 33. Decline in Monthly Man Hours by Size of Firm ... ... 34. Input Coefficients of Selected Industries with Divisions of Coefficients into Import and Domestic Categories for ... ... ... ... ... ... :..: ... ... ... Puerto Rico 35. Measuremant of the Import Content of Puerto Rico's Exports ... ... ... ... ... ... ... ... ... ... ... ... 36. Measurements of the Sensitivity of Labor Earnings to Fluctuations in Puerto Rican Exports ... .... ... ... 37. Deviations from Trend for Exports and Labor Earnings During 1953-54 and 1957-58 for Selected Products

88

XII

91 92 95 98 98 99 123 124 126a 137

LIST OF CHARTS Chart

PaBe

~

I. Seasonally Adjusted Migration ... ... II. Loans and Discounts ... ... ... ... III. Total Deposits Around the Trend IV. Exports: Fluctuations V. Earnings: Fluctuations Around the Trend

XIII

43 68 70

...

131 ...

133

FOREWARD It is a commonplace that flúctuations in the economically more advanced countries fall with multiplied violence on the developed lands, and especially on those that supply food and. raw materials. Puerto Rico during the period of the Great depression provided a grievous example of this tendency. Even in the much milder recessions of the fifties, it cast the major shadow on the economic relations of the United States with the other Latin. American countries. No result of industrial development and diversification is more to be welcomed than the prospect of reduced vulnerability to such fluctuation. In the case of the developing country, there is the companion question of the effect of the industrial fluctuations on the rate and reliability of economic advance. Together, these questions make the problem of the process and effects of the transmission of the business cycle from. more to less advanced economy one of much practical importance_ Those of us who have been associated over the years with the Social Science Research Center of the University of Puerto Rico. have been much interested in this kind of question. Poor countries. must husband their research as well as their other resources. One. way of doing this is to bring sound techniques of analysis on issues of unquestioned relevance to the community in question. In a poor country, it may better to fail in answering important questions than to succeed in answering unimportant ones. There is aplace for research in the social sciences that is less severely guided by the seeming importance of the problem. But it should. be undertaken in the largest amount by institutions that are most. amply endowed with research funds. Dr. Baer, in this study, has tackled the effect of recent economic fluctuations in the United States on the Puerto Rican economy and the meaning of the results for practical policy in the Commonwealth. The evidence has not been. easy to handle, for effects, in Puerto Rico have been masked by the growth of that economy,, and the task would have been simplified, though at a cost no one: would wish to meet, had the recessions studied been more severe. These difficulties notwithstanding, he has cast important light on Xv

the effect of economic fluctuations as between the two economies. It will also be evident that he has brought both qualification and imagination to the analysis of his data. The study adds important1'y, I think, to what is beginning to be — certainly as compared with the situation a decade ago — a very good understanding of the Puerto Rican economy. Jo)rn Kenneth Galbraith

XVI

CHAPTER I INTRODUCTION It is the purpose of this study to analyze the past, present, and possible future effects of economic fluctuations in the United States on the Puerto Rican economy. We are interested in this relationship especially in the light of the industrialization which Puerto Rico has undergone in the last decade. Since 1947 the real gross product of the island has increased at a rate close to six per cent a year, while per capita income has increased almost as rapidly. In the same period the structure of the economy has undergone and is undergoing fundamental changes. For example, in 1947 the contribution of agriculture to national income amounted. to 26% in 1958 it amounted to 14%; manufacturing contributed 17% in 1947, while in 1958 if contributed 22%. A concomitant of these developments has been a substantial change in the «foreign trade» dependence of the island, which is conducted primarily with the U. S. Mainland. The ratio of exports of goods and services to gross product, which had increased from 32% to 44% in the 1940-48 period, increased by another 10% in the subsequent decade, and this was accompanied by a drastic change in the commodity composition of exports. Thus the principal questions to which we shall- address ourselves are as follows: does the change in the structure of the Puerto Rican economy and its closer relationship with the U. S. Mainland make it more sensitive to U. S. fluctuations? Through what channels are these fluctuations passed on to the island? But to answer these questions we shall have to carry our analysis still further, even in its broader context. Puerto Rico has been, and still is, a rapidly growing and changing economy, and this makes our kind of study more difficult and complicated. Due to 1 2

2

INTRODUCTION

in

the strong growth trend, «normal» reactions to the Mainland cycle will hardly be discernible. The underlying growth factor considerably softens any impact. This leads our study into two broad channels. On the one hand, we shall have to examine the effects Mainland fluctuations have on the rate of growth itself, and on the other hand, we shall have to estimate the impact on the island as if most of the growth rate were to be eliminated. The latter type of approach is of crucial importance to the policy maker. Since the development policies of Puerto Rican authorities have been such as to make the island's growth dependent primarily on external capital and its productive capacity dependent increasingly on external trade, (1) it is of great interest for the policy maker who is in charge of promotional activities for attracting capital, to know about the investment-behavioral patterns of Mainland entrepreneurs in various types of industries over the cycle, and especially to know through what channels and with what impact Mainland fluctuations will be felt when the growth factor will no longer be dominant. Only with such knowledge can effective countercyclical policies be pursued, e. g., can expenditure plans be made so that the total mix provides the most effective countercyclical palliatives. It might even be discovered that in order to shape Puerto Rico not only into a high-income, but also into a stable economy, the promotional arm of the government will have to put special emphasis on the attraction of industries and types of firms whose exports, production, and employment policies vary least over the cycle. In other words, whereas up to now the principal concern of the policy-maker has been the generation of maximum growth, he the promomay in the near future become equally interested tion of the type of structure the island economy should assume at higher levels of production in order to ensure maximum stability. General Remarks About Business Cycle Propagation Fluctuations in economic activity are spread from region to region or country to country principally through balance of (1) This is actually a development not in line with the programs of most presentday underdeveloped countries which try to de-emphasize foreign trade. The Economic Commission for Latin America especially has worked under the assumption that foreign trade =is doomed to lag behind domestic growth, partly because of the frailure of developed countries to buy raw materials as their development proceeds, and, partly because of the necessity of underdeveloped countries to buy capital goods from developed countries. Charles P. Kindleberger, Economic Development (New York: McGraw-Hill, 1958), p. 246.

INTRODUCTION

3

payments on current account (especially changes in exports) and through changes in the flows of capital, especially flows of capital for long-term direct investment. The more a country or region is. dependent on foreign trade, the more it will be sensitive to economic fluctuations in countries or regions with which it has close trade relationships. The Puerto Rican economy should accordingly be extremely sensitive to outside economic fluctuations, since it is to a very large extent dependent on external trade. Although the abovementioned ratios of export to gross product leave no doubt about a fairly high sensitivity of the commonwealth economy to changes in its external trade position, and, therefore (taking into account that «external trade» refers almost exclusively to trade with the U. S. Mainland), to occurrences in the Mainland economy, they do not reveal anything about the exact «degree» of this sensitivity. This will have to be found by examining the commodity structure of Puerto Rican exports in an attempt to discover the income and price elasticities of demand for the principal exports commodities. Since the inception of the development program a dozen years ago, not only has the structure of the Puerto Rican economy changed drastically, i.e., in the sense of substantial changes in the relative importance of sectors from which income is generated, but over the same period of time these changes have been reflected also in the commodity structure of exports. In the immediate post-war period more than three-quarters of Puerto Rico's exports to the U. S. Mainland consisted of agricultural products and products closely related to agriculture, of which sugar, rum, and tobacco were more prominent. When one considers that even at that time total exports accounted for more than 40% of gross product, and that these products were very sensitive to the Mainland business cycle, especially via great fluctuations in prices, there can be no doubt about the almost helpless exposure of the island economy. (Witness the severe repercussions of the depression of the thirties on the island economy.) It should be remembered that until the post-war period these existed no pronounced development program, and hence no substantial growth of investment which would have acted to minimize reactions to external fluctuations. The industrialization program has brought about a most striking relative decline in the importance of agricultural goods and a growth of manufactured goods in the commodity export

4

INTRODUCTION

structure. The former rely only to the extent of 50% o1 their production on the external market, while the latter rely from 75% to almost 100% on the external (mainly U. S.) market. In other words, those industries which are constantly gaining in importance in domestic production and in exports depend for their existence mainly on the U. S. market. Unless the income and price elasticities of demand for the new products are smaller than the older ones, fluctuations in economic activity on the U. S. Mainland, which are passed on to the island via fluctuations in its exports, will have a greater impact on the island than before. A greater degree of dependence on external trade, and especially on trade with the U. S. Mainland, does not necessarily mean that the Puerto Rican economy will be exposed to the full brunt of the U. S. cycle. Reactions could appear in three different ways: a) the export commodities could be among those variables which fluctuate more than the U. S. general cycle, and in that case a relative magnification would take place on the island; b) the export commodities could be among those which fluctuate in the same way as the average Mainland indicators; or c) they could be among those economic variables which fluctuate least over the U. S. cycle, and in that case Puerto Rico would be relatively immune from the Mainland cycle. There are a number of additional factors which could mitigate the effects of increased exports to gross product ratio. First, a greater import content of exports, a change which one would expect with a substantial rise in exports of goods largely dependent on external supplies of raw materials and/or semi-finished products, could be of such a nature as to reduce the multiplier effects of export fluctuations. A relationship similar to that observed in Canadian external relations might also apply to Puerto Rico. Canada's balance of payments on current account tends to experience deficits in periods of prosperity and surp'uses in periods of depression. One interpretation of this phenomenon was that domestic investment in machinery, equipment, and industrial construction have high import contents which leads to import surpluses in high investment periods and export surpluses during depressions. This relationship «is facilitated by the fluctuations in long-term capital imports that have financed a high proportion of Canadian investment in boom periods. To

the extent that imports are cyclically unstable, they act as sta-

INTRODUCTION

5

entirety

bilizers of domestic production. Thus a high export and import ratio may involve a low multiplier.» (a) Secondly, it is important to take into account the type of enterprises the development program' has been attracting. These can be divided into two broad categories: firms which have been set up in or have resettled in the island in their (retaining, of course, marketing outlets on the Mainland), and firms or plants which are subsidiaries of larger Mainland enter prises. The former's rate of production will depend on general U. S. market conditions. Their operations will thus be closely related to demand conditions for their type of products on the Mainland market, the demand conditions depending on the income elasticity for the product. The second types of firms, the subsidiaries, present some possibilities in addition to the simple dependence on Mainland variables. They can be divided into two subcategories: those firms representing horizontal expansion of Mainland firms and those representing vertical extensions. The former are simply the extension of capacity of existing corporations for producing a finished prodúct. The second type are plants set up by Mainland corporations either to produce parts of the product which is assembled on, the Mainland, or to perform an operation on a product shipped to the island for that specific purpose and then shipped back for further processing. The horizontal plants' operations over the Mainland cycle are difficult to predict. They depend on the attitudes and/or policies of each corporation. If a downswing in economic activity reduces' a firm's sales and hence forces a decrease 1n production, how will this come about? Will all the plants of the corporation curtail their production in the same proportion, or will certain plants be shut completely while others continue to produce at full capacity? If the latter is the case, which plant would be closed, the Puerto Rican plant or a Mainland plant? To what extent does this decision depend on cost and to 'what extent on noneconomic considerations? It will be obvious that depending on the answer to these considerations, the establishment of such firms could have intensifying or moderating influences in passing economic fluctuations along to Puerto Rico. Similar assumptions could be made about plants forming (2) G. Rosenbluth, aChanges in Canadian Sensitivity to United States Business Fluctuations», The Canadian Journal of Economics and Political Science, November, 1957, p. 482.

b

INTRODUCTION

parts in the vertical link of the production process of Mainland plants. On .the one hand, the plant might be just one of a number of the corporation producing an intermediary step in the production process. In that case, considerations similar to the ones mentioned above become relevant. If the Puerto Rican plant is the only vertical link in the production process, its operations might be more stable, reflecting simply general fluctuations in the output of the firm. So far we have considered trade in goods and services as a channel though which fluctuations are spread. Fluctuations in the flow of capital will also have a major effect in the propagation of the cycle. Capital flows from one region or country to another because of differences in remunerations and risks. Actual and expected remuneration on capital which will induce flows are based on known or expected profit rates, which in turn depend on factor cost conditions and product market con •ditions. An underdeveloped region like Puerto Rico has to offer a certain minimum earning ratio above the remuneration which capital gets in the known Mainland markets in order to attract It. This was achieved by a program of tax exemption, of training labor which would be available to the Mainland entrepreneur at rates substantially below U. S. wage rates,. and of other activities by development agencies which tend to produce cost benefits to the new entrepreneur. But ultimately, what determines the profit rate, or at least its size over the cycle, will be market conditions. Since Puerto Rican firms and U. S. Mainland firms compete for sales on the same market, there is a possibility that investment decisions over the cycle will affect both places in a similar manner. In a downswing there will be a reluctance to expand productive capacity and hence investment plans might be postponed or scrapped altogether. This might not affect Puerto Rico as much, if the decisions to invest there are based to a substantial degree on cost advantage and tax advantage factors rather than on simple considerations of expansion of productive capacity. There might thus be some hope that capital flows to Puerto Rico will not fluctuate as violently over the cycle as general investment expenditures on the Mainland. The net effect of changes in investment activities of Mainland y capital depends on the foreign capital content of invesment. It can be assumed that most of the machinery used is imported, while the contribution of construction net of imported machinery — labor and contructión material — comes from island

INTRODUCTION

7

in

sources. (There is definite evidence that in construction activities Puerto Rico more labor intensive methods are used than on the Mainland.) Thus the higher the proportion of island sesources used in investment activities, the more will the impact of fluctuations 'in investment activities be felt in the island. There are a number of other channels through which the Mainland cycle can make itself felt. A downswing in U. S. economic activity decreases the net emigration of Puerto Ricans. This can contribute considerably to total unemployment on the island and cause hardships, since in order to gain fully from the industrialization program on a per capita basis, the birth rate has to decrease and/or net emigration has to continue at high rates. With a large number of Puerto Ricans living and working on the Mainland, the amount of money sent back has substantially increased in the last decade. These can also be expected to fluctuate over the cycle and thus cause additional swings in the disposable income of many island families. And finally, there is the effect on tourism, which has also flourished in the last decade. Total multiplier repercussions on the island depend to a large extent on the consumption behavior of the population. The marginal propensity to consume has been extremely high on the upswing, but there is evidence to show that this is not necessarily so in the downward phase of the cycle. Once people have come up from an extremely low standard of living, once they have become familiar with the credit system of a modern society, these is a good expectation that they will try to maintain their consumption expenditures close to the newly reached levels. If such a «Duesenberry effect» holds true in Puerto Rico, the secondary repercussions of a cycle which is passed on from the Mainland might be relatively dampened. Form of the Investigation The various hypotheses stated above will be tested in the following manner. Chapter II will consist of a short analysis of the changes in the structure of the Puerto Rican economy since the 1930's, with an even shorter discussion of policies that have led. to .them. Since we are interested in the sensitivity of the island economy to the Mainland cycle under its newly emerging character, our chief interest will be the recessions of 1953-54 and 1957-58. Hence Chapter III will be devoted to a detailed

8

INTRODUCTION

descriptive study of the recessions as they affected the economy in toto and its development program, and also a discussion of countercyclical policies which were followed consciously or unconsciously by the island authorities. Chapter V will be more analytical in content; in it we shall concentrate on the use of all quantitive evidence available in order to ascertain the impact of future recessions apart from the effects superimposed by secular growth, and effects on the growth trend itself. Chapter IV will concentrate on the policies of different types of firms which have settled in Puerto Rico over the cycle. We hope that the data analyzed will throw some light on some of the hypotheses advanced above about the «degree» of fluctuations passed on being dependent on the actions of larger firms which have subsidiaries in Puerto Rico. Chapter VI, the concluding chapter, will evaluate the total evidence, and then concentrate on policies to be pursued by the development agencies in fostering the type of growth thar will lead to a relatively stable economy and policies which the Puerto Rican government could pursue to minimize the impact of the cycle, in the light of this evidence.

CHAPTER II CHANGES 1N THE STRUCTURE OF THE PUERTO RICAN ECONOMY Puerto Rico's Structure in the Pre-Industrialization Era. (1) Until the 1940's the island of Puerto Rico was a typical example of a stagnant backward economy, with no raw material base and a very low per capita income. The principal portion of the island's income was derived from a few agricultural crops: (principally sugar), whose prices acted in an erratic manner_ From the beginning of U. S. administratión of the island until the late 1920's, a large volume of mainland capital flowed into. Puerto Rico in order to develop the sugar industry, which soon. replaced coffee and tobacco as the principal staple crop. In thirty years over $ 120 million of private capital flowed into the eco nomy to irrigate new lands, build sugar factories, develop transportation systems for the new crop. At the same time, the insular government contributed funds to develop social capital needs- roads, municipal utilities, power plants, etc. Although the development of the sugar industry increased the island's income and provided more job opportunities, this growth was not fast_ enough to keep up with the population explosion which was tak— ing place. At the end of the 1920's, the average Puerto Rican_ could find work only four days of the week, and received ant. income of less than three dollars a week. (2) The development of the 1920's also emphasized the cyclical disavantages of basing growth in the development of a handful! (1) Substantial portions of this chapter were published in the QJE, November, 1959. (2) Harvey S. Perloff, Puerto Rico's Economic Future: A Study in Planned Develop—. ment (Chicago: The University of Chicago Press, 1950), pp. 27-30.

9

10

CHANGES IN THE STRUCTURE

'of agricultural crops. It created a large class of seasonally unemployed workers and thus an internally generated business cycle. In the 20's, over sixty per cent of the working population was <employed in agriculture, mostly in sugar, coffee, and tobacco. The extent of seasonal unemployment in those branches of ;agriculture during the middle of the twenties can be gauged by the following available measurements: monthly employment in the sugar industry varied in the course of a year between 60% and 125% of a year's average employment; in coffee, the variation was between 65% and 140%; and in tobacco, between 20% and 200%. Unfortunately, no data is available on the number of months worked by the average worker' in those fields for that period of time. In the year March 1946, February 1947, the average worker in the sugar industry was employed 8.7 months, in tobacco 5 months, and in coffee 6.8 months. (3) These extremes are somewhat mitigated by the fact that these seasonal cycles do not overlap. Considering the multiplier reactions emanating from such yearly employment cycles, the picture of a perennially unstable economy emerges. In addition to all the disadvantages of seasonal unemployment, this mono-cultural economy was also highly sensitive to outside fluctuations. The value of total exports to the U. S. fell from a high of $ 133 million in 1920 to $ 66 million in 1922, slowly rising again to $ 99 million in 1927, and then sinking again to $ 73 million in 1933. In the same period the index of Puerto Rican exports (1935-9=100) fell from a peak of 325 in 1920 to 128 in 1922, rising to 180 in the following year, and then gradually sinking to a low of 94 in 1933. This bleak picture should be counterbalanced to a certain extent by simultaneous fluctuations in the average price of imports. The terms of trade on a number of occasions also turned in Puerto Rico's favor as seen in the following ratios (terms of trade-import prices/export prices) : 1918-19 1919-20 1920-21 1921-22 1922-23 1923-24

81.1 51.9 55.5 87.9 63.4 65.7

1924-25 1925-26 1926-27 1927-28 1928-29 1929-30

90.0 86.6 85.9 93.0 109.6 108.7

1930-31 1931-32 1932-33 1933-34 1934-35

100.8 92,3 87.9 98.0 104.8 (4)

(3) Statistics derived from charts and tables reproduced in Perloff, op. Cit. Even • when workers were employed, at certain times of the year as many as 25% worked tiess than 30 hours a week. (4) Perloff, op. cit., p. 135.

CHANGES IN THE STRUCTURE

11

In the 1930's, the economy plunged into the doldrums. Two hurricanes, which destroyed most of the sugar crops of 1928 and 1932, followed by the depression on the Mainland which resulted in substantial curtailments of markets and low price levels, drastically slashed the island's income, so that per capita net income dropped from a level of $ 122 in 1929 to $ 86 in 1933. Only through wlolesale relief by the Puerto Rican Emergency Relief Administration was utter disaster prevented. Stagnation continued through the 1930's, and not until 1940 did per capita income again reach the 1930 level. The Early Industrialization Program Serious concern about systematic economic development was expressed by Muñoz Marín and his party which came to power in the legislature in 1940. Although his administration came into office with a development program based primarily in agricultural reform, which was supposed to lead to a stricter enforcement of a federal 500-acre limitation law, the creation of a Land Authority and an Agricultural Development Company for the purpose of increasing agricultural productivity and diversifying agriculture (e. g., introduction of pineapple and dairy production), it was soon realized that the over-populated island could not raise its per capita economic well-being by relying solely on an agricultural development program. Even the best techniques of increasing agricultural productivity and of diversifying agricultural production in an optimal way would not be enough to raise national income at a rate faster than population growth. The average yearly rate of growth of the population in the 1920's was 1.88%, and it rose to 2% in the 1930's. (5) A substantial rise in the standard of living would thus require a rate of growth of gross product substantially above this rate. For example, with a program designed to double the per capita standard of living within twenty years it would require a yearly growth rate of the gross product of 7% (assuming the population continues to increase at the 2% rate). In the 1930's the gross product hardly rose at all, and in certain years in the midst of the depression it actually fell. It was thus fairly plain that any program with (5) This is a very high rate, of growth when one compares it to the Indian rate of around 1%, the average rate of poor countries being around 1.8% and the average European rate .5% in the 1935-55 period. See G. Meier and R. Baldwin, Economic Development, Theory, History, Policy.

12

CHANGES IN THE STRUCTURE

ambitions to raise substantially the standard of living of the population could never rely on agriculture to do the principal job. This soon became obvious to the new administration, and it was decided that if any development program were to be successful it must concentrate on industrialization, in spite of the fact that the island had no natural resources which could support and dictate the direction of an industrialization program. In 1942 the island government created its key development agencies: the Puerto Rico Industrial Development Company (PRIDCO), the Puerto Rico Planning Board, and the Government Development Bank. Between 1941 and 1945 a number of complementary agencies were established: the Water Resources Authority, the Transportation Authority, the Communications Authority, and the Aqueduct and Sewer Authority. (6) PRIDCO was most directly concerned with the creation of new industries, having a large amount of discretion in developing resources, from the planning and research stage to the operation of businesses. Its administratiors were at first convinced that industrialization could best be achieved by creating government-owned and operated plants in certain basic industries. Once done, this was supposed eventually to attract private industry. PRIDCO constructed and operated plants in the glass, shoe, paperboard, and clay products industries. It also took over a cement plant constructed earlier by the Puerto Rico Reconstruction Administration. This government-sponsored and operated development program was a failure. It failed to create the number of jobs hoped for, and it discouraged U. S. Mainland interests from investing in new island industries. It soon became apparent that Puerto Rican authorities did not have funds, the management, and markenting Know-ow to engage in full-scale industrialization by themselves. The initial program was unsuccessful in that but little private capital was attracted, and government operations were not profitable. This was reflected in the relatively small growth of private investment between 1940 and 1947 as opposed to public investment. (See Table 1.) In that period the percentage of private to total investment fell from 78% to 55%. The initial program was not a complete failure, however, in the sense that the real gross product increased by 59% from (6) For further details, see William H. Stead, Fomento: The Economic Development

of Puerto Rico (National Planning Association), March, 1958.

13

CHANGES IN THE STRUCTURE TABLE 1 GROSS FIXED INVESTMENT IN PUERTO RICO (millions of dollars or index number 1947-100) Total Gross Fixed Investmerit

1940 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958

Private

Public

Private

Index Numbers of Columns

of Total

1

2

3

4

1

2

3

23 63 100 116 109 127 150 159 163 196 207 256 274

18 35 60 61 49 76 77 74 88 123 139 158 167

5 28 40 55 60 51 74 85 75 73 68 98 107

78 55 60 52 45 60 51 46 54 63 67 62 60

36 100 159 184 173 201 238 252 259 311 329 406 435

51 100 171 171 140 217 220 211 251 351 417 451 477

18 100 143 196 214 1.82 264 303 268 261 243 350 410

Operational Disbursement of Federal Agencies Total Index

48 48 47 60 51 46 54 63 68

100 100 98 125 221 264 225 217 194 171

SouRca:

Percentage change from: 1940-7 174 94 460 1947-58 335 377 310 Puerto Rico — Statistical Yearbook, 1958, Puerto Rico Planning Board; Index number and percentages computed from yearbook data.

1940 to 1947, while per capita income increased by 22%. But even this advance has to be downgraded. During the Second World War there occurred a tremendous expansion of Federal Government activities in connection with defense and war. Thus the income derived from federal activities as' a percentage of total income increased from 9% in 1939-40 to 24. 7% in 1943-44. Between 1939-40 and 1944-45, net income derived from Federal Government activities increased by almost 550%, while net income derived from all other island sources rose by only 108%. (7) In the post-war period, income derived from federal (7) Economic Development of Puerto Rico: 1940-1950; 1951-60 (Puerto Rico Planning Board), January 29, 1951, p. 9.

14

CHANGES IN THE STRUCTURE

activities decreased drastically, contributing only 6% to the total island income in 1947-48. One could still consider that the industrialization program had some effect, since with the drastic slashes of federal expenditures on the island a stagnation and not a decline actually seemed to be setting in. Government leaders realized in the early post-war years that in order to continue and to increase the rate of industrialization, in order to raise per capita income at a faster rate, something would have to be done to attract private capital, which had more resources and better know-how to create and expand industries in a profitable manner. Revision of the Program The new approach adopted in the late 1940's was to leave the «direct» job of industrialization in the hands of private capital, i.e., of creating and operating industrial firms, while the principal task of the agencies created for the purpose of conducting a development program would be to do everything possible to create an atmosphere conducive to the attraction and development of private industry. This would be done by collecting and evaluating relevant data for projection and promotion purposes, by training workers, by introducing relevant legislation, and by providing all kinds of services, including the building of overhead capital facilities. One of the key features of the new program (also called «Operation Bootstrap») was the introduction of a law in 1947 which exempted new industries from taxation (income and property taxes) for a period of ten years. This is possible, since Puerto Ricans do not pay federal taxes (under the principle of «no taxation without representation»), and all income and indirect taxes are commonwealth taxes. The eligibility for this. exemption was extended to all those manufacturing industries which did not exist in Puerto Rico prior to January 2, 1947, to• those specifically mentioned in the Industrial Incentives Act (consisting mainly of those industries which already existed and which the government wanted to promote further), and to commercial and tourist hotels. Other measures taken in the 1948 reversal of policy were the sale of the PRIDCO factories to private interests and the creation of EDA (Economic Development Administration) in 1950. The latter, which incorporated PRIDCO and other agen-

r

CHANGES IN THE STRUCTURE.

15'

cies, became the chief economic planning and coordinating organ. Its basic function is three-fold: it does research, promo-• tional and service work. It collects data, provides background information for potential investors, and conducts extensive promotional Work crite Mainland, Through PRIDCO in buils and leases factory buildings (sometimes made to order for at. particular type of enterprise), it screens and trains workers,. and provides other technical services (e.g., instalation of machinery and equipment). The Impact of the New Policy There can be no doubt that this policy has paid off. Since the beginning of the program, more than 600 plants have been promoted and assisted. Their combined investment amounts to more than half a billion dollars, and they provide direct employment for more than 40,000 people. The types of firms: which have settled range from textile and apparel products, shoe, furniture, and similar concerns, to' those producing aircraft parts, machine tools, electric machinery, scientific products, and oil refinery. Some were set up by modest entrepreneurs,• and some by well-known Mainland corporations. The success of the Puerto Rican development venture is. impressive no matter what data are used to measure it. (See Table 2.) From 1947 to 1958 the Gross Commonwealth Product more than doubled (real gross product rising by over 70%)„ while per capita income rose by over 80% (59% in real terms),. which makes the island's per, capita income higher than any of the Latin American countries. Total investment, which rose by 174% in the 1940-47 period, increased by 335% in the following decade. Private investment's share of total investment, which had fallen from 78 % in 1940 to 55 '% in 1947, rose to only 60% in 1958. Failure to regain the pre-war share could be attributed to the fact that the industrialization program is of necessity dependent upon a large and continuing measure of public investments in social overhead capital. Sources of Capital Over 50% of investment fund's come from sources outside the island, the most important being direct investments, o

16

CHANGES IN THE STRUCTURE TABLE 2 PUERTO RICO'S INCOME (in millions dollars or indexes 1954-100) Gross Comm. Product

1930 1933 1940 .1947 1948 1949 1950 1951 1952 7953 1954 1955 1956 1937 1958

286 609 647 712 753 828 973 1,053 1,081 1,116 1,152 1,204 1,286

Real * GCP

Per Net

'499 704 718 797 865 928 1,015 1,072 1,081 1,114 1,145 1:151 1,196

Cap. Inc.

122 86 121 254 256 272 278 318 373 400 410 411 421 446 469

GCPipop. in 1954 prices

GCP

269 327 330 364 393 418 455 483 487 509 517 519 530

26 56 60 66 70 76 90 97 100 103 106 111 119

Real GCP

Net

46 65 66 74 80 86. 94 99 100 103 106 106 111

Inc.

30 21 30 62 62 66 68 77 91 97 100 100 103 109 114

Real Net Inc.

55 67 67 75 81 86 93 99 100 102 104 104 107

(*) This looks like GCP/Pop. in 54 prices. Change in real per capita income: 1940-47 : 22% 1947-58 : 59% Change in real gross product: 1940-47 : 41s%o 1947-58 : 70% SOURCE: Puerto Rico — Statistical Yearbook, 1958, Puerto Rico Planning Board. Index numbers and percentages computed from yearbook data.

TABLE 3 PUERTO RICO'S DEFICITS AND ITS PRINCIPAL FINANCE (millions of dollars)

1948 1949 1950 1951 1952 1953 1954 1955 1956 1957

Balance of Payments of Goods & Services

Flow of Long Term Capital *

Unilateral Private Transfers

Unilateral Federal Transfers

—157 -133 -107 -151 -133 -92 -124 -180 -201 -265

18 13 45 92 69 42 74 102 79 125

—2 2 2 7 11 17 18 16 17 19

80 71 73 51 44 51 62 85 95 113

* Net change in external assets in Puerto Rico. Souacs: Statistical Yearbook, 1957; 1958; Economic Report to the Governor.

CHANGES IN THE STRUCTURE

17

which in the recent past appoximately 70% consisted of new equity capital and the rest of undistributed profits. The reliance on Mainland capital for the post-war development spurt is particularly striking when one glances at the long-term capital flows which increased from $ 18 million to $ 161 million between 1948 and 1958. Much of this capital helped to finance the growing deficit in the balance of payments of goods and services between Puerto Rico and the Mainland. The great rise in imports was, of course, generated by direct investments of Mainland firms which took place in those years. (In the import structure, for example, the percentage of Machinery and Vehicles rose from 14.8% in 1948 to over 20% in 1958.) There was also a considerable inflow of capital via government borrowing on the Mainland and, finally, often more than half of the deficit was made up by Unilateral Federal Government payments, - which have fluctuated considerably, and private unilateral transfers, which have risen from $ 2 million in 1949 to almost $ 20 million in 1958. (8) Since 1948 capital flows have covered between a half and a quarter of the deficit on goods and services.. As can be seen in Table 3, capital flows to the island have usually slackened in times. of Mainland recessions; this has been accompanied by a narrowing of the trade gap, and also by an increase in the percentage contributed by the federal government and private unilateral transfers. Structural Changes

in

The main attraction of capital to Puerto Rico has been the cost advantages, cheap labor, attractive overhead capital facilities, tax exemption (if this can be counted as part of a lower cost element), while free access was given to the traditional and well-known Mainland market. It should therefore not be surprising that industrialization, which is usually associated with a decline in the relative importance of trade (especially on the export side), should be accompanied by a substantial growth the relative importance of external economic relations. A rough measuring rod of this dependence on external trade is the ratio of exports of goods and services to gross product. (8) Trade deficits are a relatively new phenomenon. Until the late 1930's surpluses. were the norm. See Perloff, op. cit., pp. 122-25. 3

18

CHANGES IN THE STRUCTURE

The computed ratios in Table 4 indicate a traditional high ratio, which has been rising continuously in the last two decadés, ~mtil exports of goods and services account for more than half of the gross commonwealth product. The height of this ratio is even more dramatic when one considers that only one of the many trade-oriented countries of Europe, the Netherlands, reached the fiftypercent mark, only a few the thirtypercent ratio, while such important trading nations as Germany and the United Kingdom had ratios in the low twenties. A brief glance at Table 4 wil also show that Puerto Rico's extreme dependence on external trade is based almost exclusively on relations with the U. S. Mainland, to which 95% of its exports were directed. The ratio of imports to gross product is even greater than the export ratio, reflecting the persistent and growing balance of payments disequilibrium on current account mentioned above, and possibly a substantial growth in the import content of exports. Since the inception of the accelerated development program ten years ago, the structure of the Puerto Rican economy, and hence also the commodity structure of its exports, has undergone some drastic changes. There has been a most striking decline in importance of agricultural goods and growth of manufactured TABLE 4

PUERTO RICO'S DEPENDENCE ON EXTERNAL TRADE Exports'

GCP

1939-40

32

1947-8 1948-9 1949-50 1950.1 1951-2 1952-3 1953-4 1954-5 1955-6 1956-7 1957-8

44 41 44 44 45 51 51 50 53 53

of,, of Total

Exports to US

Imports

CGP

°J, of Imports from US

37 97 96 89 94 92 95 95 97 95 95

68 60 58 61 58 60 62 66 71 74

94 93 92 91 92 91 91 91 90 89

* Goods and Services. SOURCE: Computed from data given in Statistical Yearbook — Puerto Rico, 1958.

TABLE 5

(in

PUERTO RICO'S EXPORT AND IMPORT STRUCTURE percentages) Exports of Exports to Total Exports, by Commodity Group 1947-8 48-9 49-50 50-1 51-2 52-3

1 2 3 4 5 6 7 8 9 10 ii

53-4

54-5

55-6

56-7

57-8•

0.1 0.4 0.6 0.6 1.2 1.4 0.6 0.7 2.1 3.0 2.8 3.4 3.2 3.2 3.4 3.8 70.3 63.5 60.0 56.7 52.0 45.4 42.0 37.7 32.8 28.5 5.1 6.4 4.3 5.9 5.3 6.3 6.6 7.5 6.0 6.1 18.5 22.3 25.3 23.5 24.1 25.3 26.0 26.2 27.0 25.8 0.2 0.4 1.0 1.0 1.5 1.3 1.5 1.4 0.7 0.8 1.5 1.2 1.7 1.7 1.5 3.4 5.8 9.8 0.2 0.2 0.1 0.3 1.0 0.9 0.7 1.4 1.1 1.3 1.9 0.3 0.6 0.4 0.4 0.8 2.7 4.1 0.6 8.0 8.0 0.5 0.4 0.7 1.7 2.2 1.3 1.5 1.8 1.1 1.4 1.8 1.6 2.1 2.8 2.3 3.0 6.5 7.3 8.3 8.1 9.4 9.4

0.6 70.9 4.9 18.1 0.1

Imports °I, of Imports to Total Imports, by Commodity Group

1 2 3 4 5 6 7 8 9 10 11

12.1 11.8 3.6 3.9 21.4 20.4 2.7 2.9 15.3 14.1 5.9 5.1 4.9 6.0 7.4 7.4 14.8 15.1 7.2 7.7 4.4 5.2

12.0 4.3 19.3 3.1 17.4 4.7 6.0 5.9 13.9 7.5 5.8

10.1 4.5 17.3 32 18.8 5.1 5.8 6.9 15.3 7.8 5.0

11.5 4.7 17.8 2.9 17.0 5.5 6.5 6.8 14.3 7.7 5.0

10.5 4.6 18.5 2.9 16.9 4.9 6.6 6.3 15.8 6.8 6.0

11.4 4.3 7.1 3.1 16.8 4.4 6.8 5.7 17.2 6.3 6.7

10.7 10.0' 4.0 3.9 6.0 152 3.4 4.1 16.2 16.8 4.5 4.1 6.4 5.3 7.4 7.9 18.6 17.9 6.8 7.6 6.2 6.6

10.1 10.9 3.8 3.6 14.4 28.5 4.4 6.1 16'.1 25.8 4.5 0.7 4.7 9.8 8.5 1.9 19.5 8.0 7.4 1.8 6.5 9.4

* '57-8 data estimated from monthly statistical reports. Accuracy of import figures 15 doubtful. SoURcs: Computed from Puerto Rico Statistical Yearbook, 1958; Current Business Statistics, Puerto Rico Planning Board, Bureau of Económics and Statistics, OctobeFNovember, 1958. 1. Animal and Animal Products (edible) a a a a 2. (inedible) 3. Vegetable and Food Products and Beverages 4. Vegetable products, inedible, except fibres and wood 5. Textile fibers and manufactures 6. Wood and Paper 7. Nonmetallic Minerals 8. Metals and Manufactures, except machinery and vehicles 9. Machinery and Vehicles 10. Chemicals and related products 11. Misc. Items. Note: Above categories leave out two items: US merchandise returned and Foreign merchandise— hence percentages do not add up to 100. Above are percentages of total exports and imports to US; do not include transactions with foreign countries which are minimal.

19

20

CHANGES IN THE STRUCTURE

to

goods. (See Table 5.) For example, the category Vegetable Food Products and Beverages alone accounted for 71% of total exports in 1947-48, but in 1957-58 it had shrunk to 28%; while the category Textiles increased in the same period from 18% to 26%, Machinery and Vehicles from almost nothing to 8%, Miscellaneous items, which consist mainly of light manufactured products, rose from 1.6% to 9.4%. It should again be emphasized that this drastic change in the commodity structure of Puerto Rican exports occurred during a time of increased dependence of the economy on external trade, which is not only revealed by the high export to gross product ratio, but which is also reflected in the relative increase of net income originating in manufactures. (See Table 6.) Industrialization has brought about substantial changes in the sources from which income originates. Especial note was made of the more than 10% drop in the importance of agriculture, balanced by a rise in manufacturing, construction, and investment expenditures by the Commonwealth government. Estimates have been made which show that by 1975, if the present trend in development should continue, national income will rise from $ 1,079 million to over $ 3,400 million, and that in 1975 over 40% will originate in manufacturing and less than 8% in agriculture. Also, while national income will triple, total employment will only rise from 555 thousand to 780 thousand; and during that time the percentage employed in agriculture will drop from 27% in 1958 to 11%, while in manufacturing it will rise from 14% to 32%. (9) A change has also taken place in the distributive shares of income (see Table 7), the proportion going to Compensation of Employees rising from 56% in 1940 to 61% in 1947 and 67% in 1958, while income from business profits declined from 35% to 31%, and a noticeable decline also took place in the proportion of rental incomes. This tends to indicate that the spurt of economic development contributed to a rise in the standard of living not only by increasing the absolute size of the national product, but also by making the increment going the labor force of greater proportion than the increment going to the other sectors. (9) Economic Report to the Governor, Mid-year 1958 (Puerto Rico Planning Board), pp. 52-53:

CHANGES IN THE STRUCTURE

21

TABLE 6 NATIONAL INCOME BY INDUSTRIAL ORIGIN (percentages)

Agriculture . . . .. . . . Manufacturing . . . .. Contract Construction Transportation and other Public Utilities Wholesale Trade . . .. . . Retail Trade . . . . .. . . Finance, Insurance and Real Estate. Services (personal and business). Services (other) Commonwealth Government . Rest of World

1947

1952

1956

1957

1958

26 17 2

23 14 5

17 19 4

15 21 4

14 21 4

7 5 12 7 2 4 10 8

7 4 10 8 2 4 10 13

8 5 12 9 3 4 12 6

9 5 12 9 3 4 13. 3

9 5 12 9 3 4 14 3

SOURCE: Net Income and Gross Product of Puerto Rico - 1940 and 1946 - 1955 and Supplements 1956-7, 1957-88, Puerto Rico Planning Board. Computed from absolute figures.

Wages and Productivity Much of the rise in the average living standard is due not only to increased employment opportunities and 'government projects, but also toc the considerable rise in the average wage rate, from $ .42 in 1948 to $ .84 in 1958. In 1948 Puerto Rican average hourly earnings of manufacturing production workers were about 20% of the U. S. Mainland average, while in 1958 this proportion had risen to 40%. Since cheap labor was one of the «natural» advantages of the island, the emphasis was at first on the promotion of labor intensive industries. It is true that lower wage rates are partly balanced by lower productivity rates, but apparently productivity has been high enough to make the lower wages a real cost attraction: Mainland minimum wage legislation does not apply directly to Puerto Rico. It is enacted in a modified form by the Puerto Rico Minimum Wage Board in order to adjust for the special economic circumstances of the island. That this body has much influence is fairly obvious when one considers that the average wage rate had doubled in

CHANGES IN THE STRUCTURE

22

TABLE 7 DISTRIBUTIVE SHARES AS PROPORTION OF NET INCOME Compensation of Bmployees

Business Profits

Net Interest

Rental Income

Profits and Dividends Received Iron NonResidents

1940

56

35

2

9

-3

1947 1950 1953 1956 1958

61 63 67 66 67

35 31 29 32 31

1 1 1 1 1

4 6 5 6 5

-2.5 2.4 -2.4 -3.9 -4.7

SOURCE: Computed from Net Income and Gross Product of Puerto Rico, op. cit.,, see Table 6.

TABLE 8 AVERAGE HOURLY EARNINGS IN SELECTED PUERTO RICAN INDUSTRIES (in dollars per hour)

Food & Kindred Products Tabacco Textile Mill Prod. Apparel Paper Chemicals Petroleum Leather Lumber Stone, Clay, Glass Metal Products Machinery Electrical Machinery Professional and Scientific Equipment

1952

1958

.424 .337 .465 .340 .613 .554

.819 513 .851 .754 .970 1.056 1.142 .641 .663 .960 1.026 1.166 1.089 .906

377 .443 .602 .633

1952- October; 1958 - September. SOURCE: Employment, Hours and Earnings in the Manufacturing Industries in Puerto Rico, series published by Department of Labor of The Commonwealth of Puerto Rico.

CHANGES IN THE STRUCTURE

23

ten years. One source has claimed productivity increases have also kept pace with wage increases, so that in 1958 Puerto Rican labor productivity in certain industries was supposed to have ranged between 80% and 115% of the U. S. Mainland standard. Even though this estimate might be a trifle over-optimistic, the rate of increase of labor productivity has been much higher in Puerto Rico than on the U. S. Mainland. It is estimated that in the coming decade the average yearly increase in productivity per man hour in Puerto Rico will be 5%, as compared to 2t % 'for the Mainland average. This would make it possible for the island's hourly wage rates to increase to 60% of the U. S. average within the decade. (10) Higher wages and productivity rates also reflect the substantial rise in recent years of more capital-intensive industries, such as metal products (average hourly wage rate in 1958, $ 1.026), machinery ($ 1.166), petroleum ($1.142). Such industries are expected to develop higher technical skills and ultimately should give Puerto Rico a more basic advantage than low wages. It is also hoped by the development agencies that such giants as the General Electric manufacturing plants, the new oil refineries, and related chemical plants, will provide new sources of «raw materials» from which related industries are generated. This leads to an interesting problem with which the planners of Puerto Rican development are concerned, and which would be of interest to most countries planning development without a natural resource base. It is obvious from Table 9 that the development program has achieved relatively little in eliminating the unemployment problem. The average yearly rate has dropped from 15.2% in 1952 to 13% in 1957. There has been a substantial decline in the margins of the seasonal swings. Under the circumstances, it is remarkable that any decline took place. As becomes clear from a glance at Table 10, employment in agriculture is declining more repidly than increases which have been taking place in manufacturing. According to Table 10, there actually occurred a slight decrease in the number of people employed in manufacturing. This is due mainly to the decline in the «home-needlework» industry, which employed 55 thousand at its height in 1951, and less than 19 thousand in 1958. It is thus fairly evident that both the rise in new industries and the (10) Puerto Rico's Manpower Needs and Supply (Commonwealth of Puerto Rico, Committee on Human Resources, Department of Labor and Planning Hoard, 1957), p. 20; Economic Report to the Governor, Mid-year 1958, pp. 53-54;;, •Stead, op. cit., p. 79.

24

CHANGES IN THE STRUCTURE TABLE

9

UNEMPLOYMENT IN PUERTO RICO (percent) January

April

July

October

Quarterly Average

19.2 192 18.4 19.1 15.0 15.7 14.9 15.9

11.5 10.0 11.4 10.4 10.2 10.0 10.5 10.6

13.6 12.4 14.4 13.6 12.8 12.8 13.7

16.5 15.8 17.3 14.2 14.1 13.1 16.6

15.2 14.4 15.4 14.3 13.0 13.0 13.9

1952 1953 1954 1955 1956 1957 1958 1959

Source: Employment and Unemployment in Puerto Rico, Quarterly Reports, Bureau of Labor Statistics, Department of Labor, San Juan, Puerto Rico.

TABLE 10

EMPLOYED WORKERS BY INDUSTRY GROUPS (in thousands)

1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 Jan. Apr.

Total

Construction ManoCover- Mining Trade Finan- Public Service Agri culture factoring Private nment ce Utilities Industries

589 585 596 604 571 550 540 539 558 552 554

214 214 214 201 190 170 172 162 159 151 151

103 102 106 111 90 99 98 95 99 99 91

35 34 27 27 33 39 32 34 38 40 37

41 45 45 50 51 50 50 50 55 57

545 574

133 151

88 98

35 40

66 66

(1) (1) (1) (1) 2 3 2 2 (1) 2

86 87 90 93 94 84 83 89 93 93 97

2 4 3 3 4 3 4 3 4 4

97 97

(1) Less than 2000. Source: Puerto Rico, Statistical Yearbook; also same source as Table 9.

30 27 30 28 26 30 31 34 37 37

75 70 77 86 79 71 67 69 71 68

38 38

81 78

CHANGES IN THE STRUCTURE

251

increase in employment in construction, government, and public. utilities have not been enough to absorb the more than sixty thousand people who have lost their jobs in the declining agricultural sector. Decrease in unemployment therefore should be attributed mainly to emigration. Net emigration since 1946 has varied between 21.5 thousand in 1954 and 69.1 thousand in 1953_ Influencing the Direction of Private Development • The essence of Puerto Rican industrialization is that theisland's resources are put to work on unfinished products which are shipped out again as finished products or as products for further processing. The greater the value added, the more the island économy will benefit. The planners of EDA have thus tried to influence the direction of development in such a way as to have a maximum amount of integration, i. e., to have imports come to the island in as «prime» a form as possible„ and to ship the products out as close as possible to marketable form. Since many of the firms settling on the island are too small to become large integrated operations, EDA has encouraged «supplying» industries to move into the island as soon as. the market for their products warrants it. For example, many brassiere firms have moved to the island. As soon as a substantial number had set up operations, the development agencies. tried to encourage supplying firms to move in — e. g., firms supplying buttons or elastic bands, etc. Through «feasibility studies» EDA tried to estimate when the market for a supplying firm would be large enough to warrant setting up operations... (11), This has also been the idea behind the promotion of oil refineries. Although it is quite capital intensive, it is regarded as a source of «raw material» for industries growing up around_ it — petro-chemical industries using much of the by-products. of the refinery. (12) (11) «Industrial feasibility research» is the expression used by EDA for its cost and market research studies, which are used as guides for the direction of its promotional' program. (12) This approach to planning and influencing development has been strongly advanced by H. C. Barton, Jr. Research Director of EDA, in his staff Papers - e. g., «Industrial Development Planning in Puerto Rico,» mimeographed paper presented to. the Economic Planning Seminar of EDA, Oct.-Nov., 1958. Much the same approach to planning has also been treated in a more formal sense by Albert O. Hirschman in his recent book The Strategy of Economic Development (New Haven: Yale University Press, 1958). In chap. 6, «Interdependence and Industrialization,» Hirschman differentiates between two investment inducement mechanisms - backward linkage effects and forward linkage effects. The former refers.

:26

CHANGES IN THE STRUCTURE

.Planning and Agriculture The bulk of government planning and help has gone into .industry, but agriculture has also received some encouragement. The old agricultural staples have suffered both in a relative and .absolute way. On the one hand, mechanization has not been applied with as much rapidity as needed to overcome increasing costs, because of labor pressure against the introduction of labor-saving machinery; and, on the other hand, the rising wage rates in industry have naturally exerted a severe upward pull -for wages in agriculture. The latter resulted in higher costs, since productivity has hardly kept pace with wage increases. Direct encouragement has, however, been given to the type of .agricultural products which are import competing. Dairy and poultry products have been encouraged. Research has also been done on the possibility of establishing a flour and feed mill complex. According to EDA, «the basis of feasibility lay in a much broader conception of an agricultural industrial complex embracing the production of animal feeds, increased cattle, hog, and poultry production, and a modern meat packing establishment.» í13j Not only would this amount to a considerable saving from bulk handling of grain over the cost of transporting bagged flour or refrigerated meat, (14) but establishment of this interrelated complex would also open up a large number of other industrial opportunities. Migration A special advantage of Puerto Rico in its relation with the Mainland has been the opportunity of free migration, which has to the input provision, derived demand, which .will induce attempts to supply through <domestic production the inputs needed in that activity.. The latter means that .every activity that does not by its nature cater exclusively to final demand, will induce :attempts to utilize its outputs as inputs in some new activities.. (P. 100.) .Development policy must attempt to enlist these well-known backward and forward =effects; but it can do so only if there is some knowledge as to how different economic -activities 'score' with respect to these effects. Ordinarily economists have been content with general references to the advantages of external economies, complementarities, cumulative causation, etc. But no systematic effort has been made to describe how rthe development ought to be modified so as to maximize these advantages even though the existence of input-output statistics supplies us with a, few tools of an :analysis of this kind.. (Ibíd.) (13) Annual Report, 1957-58 (Puerto Rico, Economic Development Administration), -page 19. (14) Ibid.: .Operation of such a complex could effect a transportation saving of 'the difference between the ocean freight on bulk grain (3 tons at $12) and refrigerated :meat ($60 per ton).

CHANGES IN THE STRUCTURE

27

substantially helped the economy in controlling its unemployment and in being able to reap the per capita benefits of industrialization. The yearly emigration has varied closely with fluctuations in the Mainland business cycle. In the 195O's it reached a high of 69,124 (1953) and a low of 21,531 (1954), the latter reflecting the Mainland recession. The average has been approximately 40,000 a year. The Impact of Government Expenditures I should like to conclude this survey of the changes in the structure of the economy with an estimate of the quantitative impact of the government in the development of the last decade. Although government receipts as a percentage of gross product have increased slightly since 1950, from 14% to 16% (this does not include public authorities), they are rather small when compared to the same ratio for developed countries. (15) Government expenditures amount to 19% of gross product. It is supposed to be typical for underdeveloped countries that the share of government receipts to gross product is rather small, ranging from 8 % in India to 20 % in Ceylon. The reason for this has been interpreted to be that «the more developed the economy the greater its productivity. The greater its productivity the larger the relative margin of incomes above strictly subsistance, and the larger this margin, the greater, generally speaking, the potential taxable capacity of the economy...» (16). This explanation could satisfy the Puerto Rican economy in the early stages of its development. It seems, however, that the small growth since that time has been partially due to the island's greater reliance on private enterprise for basic development and to the tax exemption system. It should also be noted that less revenue is needed because some essential services such as defense, which takes up a large part of the budget of many sovereign countries, do not have to be provided by the Commonwealth government. Many public agencies which provide overhead facilities (as does the Water Resources Authority), do not fall within the budget, and are financed by borrowing rather than by taxation. (15) See H. T. Oshima, .Share of Government in Gross National .Product,. American Economic Review, June, 1957, pp. 381-90. (16) Ibíd., p. 384.

28

CHANGES IN THE STRUCTURE

Another interesting phenomenon is the revenue side of the Puerto Rican budget. Approximately 52% of total revenues comes from excise taxes, 36% from income taxes, and the remainder from property taxes, licenses, and inheritance and gift taxes. The tax revenue constitutes 67% of total revenues, the balance coming from tax revenues returned by the Federal government (mainly from customs collection), federal grantsin-aid, and non-tax revenues. It is customary for governments of underdeveloped countries to rely to a much greater extent on indirect than on income taxes because of ease of collection, (17) but in the case of Puerto Rico the explanation lies rather in the widespread tax exemption program. Government Expenditures and the Use of Resources It was mentioned above that in the initial stages of its program the government development agencies promoted specially those industries which made use of the island's abundant. resources — especially labor. The same principle_ is also followed in making public expenditures for social overhead facilities, but not always directly. When the government decides to build a road or factory building or hotel accomodations, it asks for bids from contractors. The winners have usually been firms using more labor intensive techniques than do similar firms on the Mainland. The public authorities, which often use their own resources for construction activities, also engage in more labor intensive techniques than do comparable Mainland undertakings. As long as wages are low enough to make a clear-cut cost advantage of labor over capital intensive techniques, there always will exist an automatic tendency to make use of abundant factors of production. The Possibilities of Self-sustaining Growth Once economic growth is achieved, with the help of substantial doses of public and/or private capital, the principal question which arises is: how long will a substantial rate of growth be dependent on foreign savings? At what point will growth have produced an economic structure which can generate enough (17) Charles P. Kindleberger, Economic Development (New York: McGraw-Hill, 1958), pp. 197-?A1.

CHANGES IN THE STRUCTURE

29

savings of its own, so that continued growth will be more and more self-sustaining? Puerto Rico was able to achieve a 6%-7% annual growth rate in the last decade because of substantial inflows of capital. In, recent years capital formation amounted to approximately 20 % of gross product. At the same time, total consumption expenditures of the island were larger than its gross product, and personal savings have for many years been negative. (18). This does not mean that there was no contribution from domestic sources to capital formation. In the last ten years approximately half of the latter has been financed by internal sources — undistributed profits, depreciation, and government savings. (19) That total consumption expenditures were greater than gross product implies that total capital inflows helped to finance not only part of het capital formation, but also a substantial amount of «excess consumption.» This is important to notice, because in order to judge potential self-propelled growth, one has to observe carefully the behavior of internal institutions contributing directly to capital formation. According to projections made by the Planning Board, the great dependence of the island on foreign sources for an average growth of 6% should considerably diminish in the next fifteen years. (See Table 11.) The biggest single source should be the depreciation funds and the second largest are supposed to be personal savings. The latter's impressive growth will be the result of much higher incomes accompanied by higher propensities to save and the increasing availability of many types of savings institutions. Finally, government savings and undistributed profits will also contribute substantially. The dependence on foreign sources will have shrunk to less than 25% and should even be declining in absolute amounts. (18)

1956 1957 1958

Consumption Expenditures *

Gross Commonwealth Product

1,157,705 1,210,239 1,296,699

1,151,606 1,203,537 1,285,526

* Includes households, government, and municipal government. (19) Although personal savings are negative on a net basis, a substantial increase in personal savings has occurred. For example, time and savings deposits in banks have increased from $52 million in 1950• to $162 million in 1958. The First Federal Savings and Loan Association has had a spectacular success since its organization in 1951 in tapping the funds of average wage earners. Its funds amount to approximately $40 million. Souace: Net Income and Gross Product, op. cit.

30

CHANGES IN THE STRUCTURE

Given certain reasonable assumptions (reasonable because of clearly discernible°trends) — that as per capita income rises the savings habit will rise, that institutions for stimulating the latter will continue to grow, that enough incentives will always exert corporations to re-invest substantial portions of their profits, and that local entrepreneurship will develop sufficiently to make effective use of local savings — there appears to be a good probability that Puerto Rican economic growth will become substantially self-generating. These projections should be taken with a large grain of salt. The general assumptions behind them are that the gross product will continue to grow at the rate achieved in the last decade, that by 1975 the Puerto Rican economy will have reached an average living standard equal to that of the United States in TABLE 11

SOURCES AND USES OF CAPITAL FUNDS (in millions of dollars) 1947 1950 1955 1957 Total Sources 46 124 219 14 6 25 22 Undistributed Prof. —38 —38 —35 —17 Personal Savings 40 49 26 34 Government Savings 82 23 37 64 Depreciation Net Inflow of 28 45 104 142 External Capital Statistical Dis—7 39 crepancy 21

1960 1965 1970 1975 38 18 55 91

52 54 74 125

70 113 100 172

95 209 135 238

270

321

214

171

Souaces: Net Income and Gross Product: Puerto Rico - 1940 and 1947-1955, and Proyecciones del Desarrollo Económico de Puerto Rico, Junta de Planificación. Negociado de Economía y Estadísticas, Diciembre, 1957.

1950, and that. relationships between economic variables will be similar to those existing in the U. S. at that time, subject to certain modifications based on peculiar historical trends of the islands. (See second source mentioned in Table 11). For example, depreciation amounted to 6.9% of net income in 1955. For the U. S. in 1950 it was 7.5%. In the projection it is automatically assumed that Puerto Rican depreciation will amount to 7.5% of net income. Personal savings are supposed to climb to 5%, equal to the U. S. percentage in the early fifties. The specific

CHANGES IN THE STRUCTURE

31

assumptions made — e. g., of the savings pattern of the average Puerto Rican family as it moves into higher income brackets, orabout the habits of corporations in determining their profit policies — are never spelled out in the published reports of the Planning Board. Altehough these projections are based on dubious procedures, they are not completely worthless. It is of interest to note that they represent also a policy goal. For example, in describing: projected personal savings, the above-mentioned, publication notes that the 5% savings will not necessarily be the result of automatic forces, but might require government actions to achieve it. These projections thus represent a goal that the government thinks is feasible, and toward which it will work. And a specific hope and aim of the government as revealed in these projections is to make the island's growth more self-propelling.

CHAPTER III THE POST-WAR RECESSIONS.: I - A DESCRIPTIVE ANALYSIS Since most of the analyses, generalizations, and projections in subsequent sections are based on inferences drawn from both the changes in the structure of the Puerto Ricán economy and actual experiences during post-war recessions, it is fitting to review the causes of two Mainland recessions, and to trace their effects on the Puerto Rican economy. We have chosen the 195354 and 1957-58 recessions. The reason for omitting the 1949 downturn is that the economic development program had then just. gotten underway, and 'the structure of the economy had not changed drastically enough to afford observations which might be. used readily for generalizing purposes. The 1953-54 Recession Causes and Characteristics of the Recession on the Mainland The 1953-54 Mainland recession lasted thirteen months, according to the business cycle chronology of the National Bureau of Economic Research; it extended from a peak in July, 1953, to a trough in August, 1954. It originated in two main developments: a decrease in defense expenditures, and a decline in the rate of increase of private consumption expenditures. (1). Absolute increases of the latter in the last quarter of 1952 and the first two quarters of 1953 were 6.8, 3.5, and 2.8 billion dollars. One reason alleged for this was the great bulge in consumer expenditures in (1) Most of the review of the U. S. recession is based on the analysis in Economic Report of the President, January, 1955, and Bert G. Hickman, «The Contraction of 1953-54,» The Review of Economics and Statistics, February, 1958, pp. 36-48.

33 4

34

THE POST-WAR RECESSIONS

the latter part of 1952, the aftermath of a prolonged steel strike which had kept consumption previously at relatively lower levels. Another interesting phenomenon was the relatively increased expenditure for services and automobiles out of disposable income, and the relatively diminished expenditure for other goods, which meant that the rate of increase in the purchase of nonautomotive durable goods and non-durable goods lessened even more. With the leveling of sales in early 1953, there was a pronounced increase in the inventories to sales ratio, which was reenforced by sizable decreases in defense expenditures. The latter declined by $ 3.6 billion. The combined decrease in expenditures on defense and on the declining sector of consumer goods amounted to $ 7.5 billion in the last half of 1953; but with offsetting increases in other principal categories of final demand, reduction in final expenditures amounted to only $ 2 billion, and therefore «had production declined only in proportion to final sales, the drop would have been small indeed.» (2). But declines in all stimulated a drastic reduction in inventory investments of $ 8 billion, causing the GNP to drop by $ 10 billion instead of $ 2 billion. These events were accompanied by a drop in the production of consumer durable and nondurable goods of about 25%, and among the sectors of industrial production «reductions of 10 % or more were recorded for textile, apparel, leather, and rubber products in the non-durable group and for primary and fabricated metals, electrical machinery, and furniture and fixtures in the durable group.» (3). The recession entered a second, stage in 1954. Consumer expenditures which had declined through the 1953 phase now became an expansionary force, and were accompanied by a rise in residential construction, sustained rises in state and local expenditures, and in net foreign investment. All these forces overcame the deflationary pressure of falling federal expenditures. The rise in consumer expenditures could be attributed to a rise of disposable income, despite the fall in personal income from production. This 'was caused by the usual automatic changes in tax receipts and transfer payments induced by the recession, and by reduced rates of personal income taxation which became effective on January. 1. However, observers do not simply attribute the rise in consumer expenditures to changes in disposable income. They also account for it to a large (2) Hickman, op. cit., p. 40. (3) Ibíd., p. 41.

THE POST-WAR RECESSIONS

35

extent by a rise in the consumption function; e. g., although disposable income increased by only $ 3 billion between the last quarter of 1953 and the third quarter of 1954, consumption expenditures rose by $ 7.5 billion, and savings out of income dropped from 8.3% to 6.4% (4).. The early revival of consumer expenditures is important when compared to the drop in federal expenditures by $ 10.5. billion from the fourth quarter of 1953 to the second quarter of 1954, and a drop •of a billion dollars in business fixed investment. With a $ 5 billion rise in consumer expenditures, and a $ 1.5 billion rise in residential construction, net foreign investment and state and local expenditures, final expenditures fell by only $ 2.5 billion. GNP actually rose by one 'billion, mainly because substantial decreases in inventory decumulation occurred. (See Table 12.) In the last half of 1954 recovery, started to take place, with aggregate final expenditures turning upward in the third quarter, and with an abating rate of decline of federal expenditures. A temporary spurt of inventory liquidation provided a short setback to the expansionary forces, and GNP rose by only $ 1 billion in the third quarter. But in the last quarter of 1954 a vigorous expansion set in. Effects of the Recession on Puerto Rico With our knowledge .of the structure of the Puerto Rican economy, we would expect the above events to have had the following effect: decreases in the exports of textile products, apparel, electrical machinery, furniture, and fixturés. These products are of increasing importance in the export structure, and with inventory decumulation on the Mainland and decreases in industrial production one would expect such exports to be substantially affected. The second reaction one would expect because of decreasés in the opening of new plants, caused principally by. Mainland market shrinkages. But this might be mitigated by the relatively small decreases, in gross private fixed investments, the maximum decrease of which was only $ 2 billion between the third quarter. of 1953 and the second quarter of 1954, compared to a shrinkage of $ 9 billion in total gross (4) This rise in the consumption function has been attributed mainly to improved expectations resulting from tax cuts, capital gains in the stock market, intensified selling efforts, price reductinos, and more liberal credit terms.

36

THE POST - WAR RECESSIONS TABLE 12 THE RECESSION OF 1953-4 - COMPARATIVE DATA U. S. Data (billions of dollars) * GPN

Consumption Durable NonDurable

1 2 3 4

361.6 367.4 366.3 357.5

30.2 30.6 30.5 28.0

118.9 119.8 118.9 118.6

1 2 3 4

358.1 358.7 360.0 367.7

28.5 29.2 29.4 30.4

1 2 3 4

379.0 387.7 397.0 402.8

34.7 35.3 37.2 35.4

Gross Priv. Investment

Changes in Business Inventories

Index of Industrial Production (1947-9 = 100)

51.5 53.5 51.8 44.5

2.0 3.1 1.1 -5.2

135 136 133 130

118.9 120.1 121.3 122.0

46.5 48.3 46.8 51.9

-2.5 -1.5 -4.1 0.5

125 124 121 129

122.4 124.8 1'27.4 129.2

55.6 59.7 61.4 65.4

2.7 4.6 3.3 6.3

135 138 137 134

1953 .

1954

1955

Puerto Rican Data (millions of dollars) ** Total Exports

Total Imports

Unemploy- New Plante Plants Disment (°/°) Starting continuing Operations Operations (No. of Plants)

1 3 2 4

70.4 86.2 91.7 80.6

115.6 116.5 113.5 137.9

19.2 10.0 12.4 15.8

15 32 19 17

2 5 3 5

82.6 93.6 102.0 109.4

89.1 98.6 104.6 112.1

65.6 80.9 95.3 102.6

1 2 3 4

76.5 83.7 81.2 94.1

116.2 137.0 129.7 133.3

18.4 11.4 14.4 17.3

22 16 12 11

6 9 15 7

109.6 115.5 112.4 1282

109.2 116.1 112.4 122.6

110.5 114.0 112.5 142.5

1 2 3 4

67.0 100.1 103.0 93.5

134.3 137.8 123.8 150.7

19.1 10.4 13.6 14.2

18 15 il 25

6 4 4 6

134.6 158.1 153.2 160.2

132.4 157.9 148.4 153.7

140.5 158.5 165.5 177.2

Index of Industrial Production (1952-54 = 100) Total Non-Durable Durable

1953

1954

1955

*

All data seasonally adjusted, except Industrial Production. :r Data not adjusted for seasonal variation.. SOURCE: Business Statistics, 1957, U. S. Dept. of Commerce; Puerto Rico, Statistical Yearbook; EDA Reports; U. S. Dept of Commerce, Bureau of the Census, U. S. Trade in Merchandise and Gold and Silver with U. S. Territories and Possessions.

THE POST - WAR RECESSIONS

37

investment between the second quarter of 1953 and the fourth quarter of 1954. As is obvious from a glance at Table 12, the Puerto Rican data are not seasonally adjusted, and therefore must be examined very carefully. One must also consider the strong growth trend which pervades all Puerto Rican data. There can be no doubt of a slackening in the export sector in 1954. This is especially true of the second and third quarters, which are usually the seasonal peak. Based on export figures, it would seem that the brunt of the effect of the Mainland cycle on exports fell during the third quarter when exports were over $10 million less. than in the previous year; which also represented the trough in the mainland. There is, however, a definite lag of three quarters, since decrease in consumer sales on the Mainland and peaks in inventory decumulation were reached in the last quarter of 1953. There seems to be a lagged relationship of about two quarters .when one compares total Puerto Rican exports to overall industrial production on the Mainland. Here the trough was reached in the third quarter, which was also the low point in Puerto Rican exports. The latter are examined in a more detailed fashion in Table 13, where each of the principal export items is compared to Mainland industrial production in the same industry. Production of textile products on the Mainrand started to decline after the second quarter of 1953 and reached a trough in the third quarter of 1954, having fallen 19 points; apparel products, which bulk large in Puerto Rican exports, started to decline after the first quarter of 1953 and also reached a trough in the third quarter of the following year, having declined 18 points. Puerto Rican exports in this category seem to have been affected in the last quarter of 1954, and the first quarter of 1955, when they reached a trough. There thus occurred a lagged relationhip between industrial production of textiles and apparel on the Mainland and textile product exports from the island. While industrial production vas falling drastically on the Mainland, exports continued to rise, although mostly at a smaller rate, and only when production started to recover on the Mainland did exports decline, reaching a trough two quarters after the Mainland trough. Explanations, for this phenomenon are varied. First, there is the growth trend; the rate of openings of new plants during the last half of 1953 and the first half of 1954 was very strong in textiles and apparel

38

THE POST - WAR RECESSIONS TABLE 13 a)

U.S. INDUSTRIAL PRODUCTION AND PUERTO RICAN EXPORTS 1953-4 (For U.S. Index of Industrial Production 1947-9 = 100; P.R. Reports in Millions of Dollars) Iilon-Electrical Machinery

Machinery & Vehicles

Textile Mill Products

Apparel

Machinery

Textiles (Total)

Paper & Allied Products

P. R.

U.S.

U.S.

P. R.

U.S.

U.S.

P. R.

U.S. P. R

1 2 3 4

136 .7 139 1.3 137 1.5 131 :5

150 148 137 137

208 191 189 189

.6» 1.1 1.1 2.1

110 111 100 94

118 110 108 102

18.5 17.6 20.1 21.6

137 134 1.' 130 1.2 131 1.3

1 2 3 4

123 .9 121 .8 121 1.2 125 1.8

132 126 119 122

172 163 170 201

2.8 3.3 3.3 4.0

94 94 92 100

112 103 100 104

21.7 20.7 23.6 22:9

131 1.7 135 1.1 131 1.3 137 1.1

1 2 3 4

126 133 137 140

129 135 133 143

194 184 188» 209

2.5 4.2 4.5 6.3

107 108 102 110

117 113 107 113

20.5 21.7 25.9 26.6

146 154 150 157

Fabricated Metal Products U. S.

1953

Electrical

1954

1955 1.1 1.1 1.1 1.0

1.2 1. 1. 2.

b) SELECTED P. R. IMPORTS Chemicals

1953 '

U.S.

Food & Non-Metalic Beverages Minerals

P. R. U. S.

Textile Vegetable Animal & Fibers Food Products Animal & Migs. & Beverages Products

Machinery & Vehicles

Chemicals & Related Products

P. R.

1 147 2 148 3 143 4 148

1.1 98 40.2 i.l 103 45.0 2.1 118 48.8 1.9 110 30.4

7.6 8.1 8.1 8.0

18.4 20.5 20.4 22.4

21.9 21.0 20.1 20.9

12.6 11.9 13.3 14.5

19.8 19.4 18.8 20.4

6.6 7.7 7.6 7.8

1 148 2 145 3 144 4 156

1.2 97 33.4 1.8 103 40.9 1.8 115 32.0 1.8 109 41.2

6.0 9.8 7.7 7.2

17.4 20.7 18.6 22.7

20.0 20.5 21.8 19.6

12.8 13.9 13.7 13.3

185 24.1 21.2 25.8

6.1 8.1 7.8 7.7

1 162 1.7 98 19.7 2 165 1.3 105 51.9 3 164 1.3 117 47.0 4 177 .9 114 28.0

9,4 9.0 7.3 8.7

20.0 23.5 20.0 26.5

21.7 20.8 21.1 21.4

15.4 13.7 13.3 14.2

23.7 26.7 20.4 27.4

9.2 11.3 10.2 10.4

1954

1955

Souxca: S'ee Table 12.

THE POST - WAR RECESSIONS

39

in

(e. g., eleven new firms starting operations in the first quarter of 1954). Their contribution to exports, even in a declining market, was bound to bolster total exports. Second, many EDApromoted firms work en a contract basis, as subsidiaries of Mainland firms. While sales and production were declining on the Mainland, island plants would keep on delivering for the period of the contract, thus maintaining exports. The pattern most industries usually was a decline in .the rate of growth of exports during most of 1954, with absolute declines taking place in the last quarter of 1954 or the first quarter of 1955. The proportional decline in most cases was larger for Puerto Rican exports than for the respective industrial productions on the Mainland, though the duration of the decline was much shorter for Puerto Rican industries. The most important export categories of Puerto Rico fall into the type of U. S. industries which experienced a greater than average decline in industrial production than the general average index. (Compare the decline in total industrial production in Table 12 to individual declines in Table 13.) There was a definite lag in the reactions to the Mainland recession of the number of new plants starting operations on the island. If we discount the extraordinary spurt in the second quarter of 1953 (Table 12), it would seem that actual reactions set in only in the second quarter of 1954, three quarters behind the Mainland indicators, reaching a trough in the last quarter of 1954. The reason for such a lag is that plants opening operations were promoted a number of months before. Thus when the Mainland recession set in there was probably a backlog of promoted plants which engaged in investment activities even after Mainland investment had gone dówn. The recession is felt much earlier when measured by the number of firms promoted. (5) According to the best authority on the subject matter, «Promotion seems to. lag somewhat behind production on the way up but to respond quickly to a decline». (6) It is also evident from Table 12 that the recession had some effect on the number of plants discontinuing operations during 1954, the peak being reached in The third quarter; which was also the trough point of the Mainland recessión. The (5) Pirms promoted= have no direct effect on Puerto Rican economic activities. Only fluctuations in plants actually opened will have a direct impact on island economic indicators. (6) H. C. Barton, =Background Data for Program Conference on Industrial Development,' July 16, 1957 (unpublished).

40

THE POST - WAR RECESSIONS

greatest number of discontinuances occurred in textiles; especially apparel. Of the latter, Barton writes that «changes in output are, to a considerable extent, the result of openings and closings of entire plants. This is generally true in industries where capital per worker is low and where individual plants are small in relation to the total size of the industry. Also expansion or closing plans are not made as far in advance as they are in more heavily capitalized industries». (7) It is true that quite a number of plants close for reasons other than marketing difficulties arising from a Mainland recession — reasons such as managerial difficulties, costs higher than at first estimated, labor costs too high in relation to productivity, etc.; nevertheless, the increased number of plant closings in times of recession can mostly be attributed to that recession. For example, in a closed plant survey made by the Economic Development Administration, various reasons for the closing o~ each plant surveyed were listed, and the three most essential reasons were graded: U (underlying cause), C (contributing cause), and P (proximate cause). In 1952, eight of thirteen plants listing «marketing» as a principal reason for closing chose it as the «underlying» cause; in 1953, there were 7 U out of ten, and in 1954, 14 U out of 16. (8) Although there are no monthly or quarterly data available of Puerto Rico's GNP, investment and balance of payments items, decreases in new plants opened, and smaller exports were refleeted in annual national accounts. The gross product, which had increased by almost $ 80 million from the fiscal year 1952 to 1953, increased by $ 28 million in the following year, and by $ 35 million in 1954. to 1955; investment declined from an increase of $ 9 million in 1952-53 to $ 4 million in 1953-54, but recovered substantially in the following year and long-term capital inflow declined from $ 83 million in 1953-54 to $ 64 million in the following year. The latter rose to $ 73 million in 1955-56, and $ 152 million in 1956-57. Here we obviously have the lag that appeared in the openings of new plants. Effects on Employment The brunt of the unemployment which the recession brought along was felt on the Mainland in the first three quarters of 1954 (7) Ibid. (8) «closed Plants Survey; 1947-54,» (Economic Development Administration, Office of Economic Research, San Juan, Puerto Rico). (Mimeographed:)

THE POST - WAR RECESSIONS

41

TABLE 14 EMPLOYMENT; UNEMPLOYMENT AND MIGRATION, 1953-5 (Seasonally Unadjusted)

a) Total (Percentages) U. S. Unemployment

P. R. Unemployment

Net Migration ( 'OOOs)

1953 1 2 3 4 1954 1 2 3 4 1955 1 2 3 4

2.8 2.3 2.1 2.8

19.2 10.0 12.4 15.8

-69.1

5.3 5.2 5.0 4.4

18.4 11.4 14.4 17.3

21.5

5.2 4.1 3.4 3.3

19.1 10.4 13.6 14.2

-45.5

Source: U. S. Dep't of Commerce, Business Statistics, 1957; P. R. Dep't of Labor, Bureau of Labor Statistics, Employment and Unemployment in Puerto Rico, (series):

b) Selected Industries (Employment in 'OOOs) Food & Kindred Products Total Average Employment Weekly Hrs.

Paper & Allied Products

Textite Mill Products

Apparel & Related Products

T. E.

firs.

T. E.

Hrs.

T. E.

firs.

T. E.

H rs

Chemicals& Allied Products

1953 1 2 3 4

18.5 19.7 13.7 13.8

37.8 37.1 34.7 35.3

3.3 39.1 3.1 392 3.5 38.7

15.1 16.0 16.9 17.7

34.1 34.3 33.7 33.1

1.7 1.7 1.8 1.8

43.2 42.7 40.9 40.3

1.3 1.4 1.4 1.2

30.9 36.0 35.8 32.4

1 2 3 4

18.4 20.0 13.3 142

36.6 36.9 34.5 34.5

3.5 3.0 3.1 .3.4

38.1 35.6 37.6 37.3

17.8 17.1 16.2 17.2

34.6 35.9 34.2 33.8

1.8 1.7 1.7 1.8

40.8 40.1 40,4 40.5

1.3 1.5 1.4 1.3

33.1 34.6 33.2 32.4

1 2 3 4

18.4 15.5 13.3 13.6

37.5 34.0 35.1 34.9

3.3 3.6 3.6 4.1

365 36.8 37.8 38.6

16.5 17.5 17.4 18.9

33.6 32.5 33.8 34.0

1.9 1.8 1.9 2.0

39.9 41.2 41.5 40.7

1.4 1.5 1.7 1.6

32.0 35.9 33.8 34.2

1954

1955

42

THE POST - WAR RECESSIONS TABLE 14 (continued)

b) Selected Industries (Employment in 'OOOs) Lumber & Wood Products Total Average Employment Weekly Hrs.

Stone. Clay & Glass Products

Metal Products

Leather Products

T. E.

Hrs.

T. E.

Hrs.

T. E.

Hrs.

1953 1 2 3 4

2.9 3.0 2.9 3.1

38.4 38.2 38.7 38.6

2.9 3.0 3.2 3.2

40.2 40.8 39.2 38.9

2.6 3.0 3.3 3.9

39.2 38.8 38.5 38.9

1.8 ' 35.6 2.1 34.6 22 35.4 2.1 35.4

1 2 3 4

2.8 2.7 2.6 3.0

37.3 37.6 39.1 38.7

2.9 2.9 2.9 2.9

38.6 36.9 37.7 37.7

3.9 3.8 3.7 4.2

38.2 38.1 39.5 39.6

1.8 1.7 1.6 1.7

36.0 33.7 37:0 37.6

1 2 3 4

2.9 2.9 2.9 32

37.0 36.4 37.8 39.1

3.0 3.3 3.6 3.9

38.3 39.1 40.1 39.5

4.3 4.4 4.6 4.7

38.8 38.3 39.1 38.6

1.8 2.2 2.2 2.3

36.1 35.9 35.7 35.6

1954

1955

Source: P. R. Bureau of Labor Statistics, Dep't of Labor, Employment, Hours and Earnings in the Manufacturing Industries of Puerto Rico, (series).

when unemployment was generally between 2.5 % and 3 % above the average of the previous year, hovering around 5% (See Table 14). Unemployment in Puerto Rico, which at the time still varied between 19% and 10% during the year, reflected the impact of the Mainland recession in the last three quarters of 1954, reaching a peak in the third quarter. The rates above «normal» unemployment were a little smaller than on the Mainland, varying between 1.5 % and 2 %. An important factor contributing to the rise of Puerto Rican unemployment was the decline in net emigration from 69 thousand in 1953 to 21 thousand in 1954. There was no lag in the net emigration. It began to drop off sharply in the second quarter of 1953, the beginning of the Mainland downturn, and there was no real pick-up again until the last quarter 1954, the beginning of the Mainland recovery. This simply reflects the acute sensitivity of migration to Mainland business conditions — Puerto Ricans being among the first to lose their jobs in a dowuturn, but with the type of work they do, finding it easy to re-enter the job market in an upturn. That a decrease in net emigration

CHART 1 SEASONALLY ADJUSTED MIGRATION (ANNUAL RATES) CALENDAR YEARS 1953-1960 Thousands 750

I

I

I

1

I

1

I

600

Departures

450

Arrivals

300

150

l

—150 1953

l

0

1954

9955

Net Balance I I 1956 1957

I

I 1958

1959

Calendar Year Source: Current Business Statistics, Bureau of Economics and Statistics, Puerto Rico Planning Board.

43

1960

44

THE POST - WAR RECESSIONS

should contribute substantially to an increase in unemployment during a mainland recession should be clear when one recalls. that employment in the declining agricultural sector is decreasing faster than the growing industrial sector can employ them, and that total unemployment can be kept from rising or can be: decreased only by a substantial emigration. Not only was theré a great drop in the net outflow of workers in 1954, as revealed by statistics in Table 14, but the drop was even greater because in 1954, according to a government publication, «most of the net outward movement of population is attributed to women and to children under labor force age. Thus only 8 per cent of the' excess of departures over' arrivals was attributable to men of labor force age». (9). In addition, the ranks of the unemployed were also supplied by an increasing amount of discharges from: the army. Between the outbreak of the Korean war and ear ly 1954, inductions and enlistments in the armed forces exceeded discharges by 44 thousand. This meant that the army absorbed potential and actual members of the labor force at the rate' of almost 1.000 per month. But in the period from March to October, 1954 «the number of discharges from the armed forces. exceeded inductions and reenlistments by over 9.000, thus increasing the number of entries into the labor force». (10). The contribution of the manufacturing sector, and especiallyof the chief export industries, to unemployment was slight. (See' Table 14, section b.). The greatest decline took place in apparel' products; it amounted to about 1,500 workers between the last quarter of 1953 and the third quarter of 1954 (while hours worked increased). In most industries there was a decrease in the rate of increase of employed workers rather than absolutedeclines. It would seem that the chief contributions to unemployment came from the Home Needlework industry, which has been a continuously declining industry, where employment dropped by over 5,000 between 1953 and 1954, and the construction industry,. where employment declined by 6,600 in the same period. (See Table 15.) The latter had nothing to do with the Mainland recession, since private construction activities actually increased between 1953 and 1954 by over $10 million. It was public construction which declined by over $6 million, due to a reduction (9) «External Passenger Movement,. Current Business Statistics (Puerto Rico. Planning Board, Bureau of Economics and statistics), October, 1955. (10) Economic Report to the Governor, 1954 (Puerto Rico Planning Board, Bureata of Economics and Statistics).

THE POST - WAR RECESSIONS

45

TABI,i; 15 LABOR FORCE AND EMPLOYMENT BY INDUSTRY (in thousands) 1952

1953

1954

1955

1956

All Industries .

571.1

550.0

539.5

539.3

557.7

Agriculture Forestry and Fishing . Manufacturing Home Needlework. Mining Construction

190.2 2.0 56.9 33.6 1.7 33 3

170.1 1.8 62.7 36.7 2.7 38.5

171.8 1.7 66.8 31.2 2.1 31.9

162.4 1.3 66.2 28.7 1.5 34.0

159.5 1.6 74.4 25.0 1.3 37.6

Souaca: Economic Report to the Governor, 1956, Puerto Rico Planning Board, Bureau of Economics and Statistics.

in total construction expenditures of the Housing Authorities and the completion of several other public projects. The Commonwealth government engaged in no apparent countercyclical fiscal policy. Revenues from taxes met regular government expenditures, and the net debt continued to increase at about the same rate as in the past, the receipts from the sale of bonds going to capital improvement projects. (11). Puerto Rico has a strong instrument for counter-cyclical policy with its high credit rating on the Mainland, and its low debt to annual revenue. ratio. With low interest rates usually prevailing on the Mainland capital markets during a recession, it would be easy and cheap for the government to engage in increased borrowing for stepped-up public works. Conclusion The effects of the 1953-54 Mainland recession on Puerto Rico were very mild. The principal channel through which it spread was a substantial decline in net emigration, causing a rise in unemployment. Most manufacturing industries which rely to a (11) In 1954 the Commonwealth and municipal debt of Puerto Rico was small relative to the means available tó secure .the means for servicing. The total debt was only 21% of annual revenues, and 3.5% of net income. This is one of the lowest ratios of public debt to annual government, receipts in the world. (See Economic Report to the Governor, 1954, p. 21.)

46

THE POST-WAR RECESSIONS

large extent on exports were hardly affected by declines in exports which took place with a lag of approximately nine months. This can be explained by the backlog of newly promoted firms which were opening their plants in 1954 and thus contributing to total exports and general employment in the respective industries. The sharpest decline in employment took place in the needlework industry, but this was due to a long-established declining trend, and in public construction. The principal effect was thus a mild decline in the rate of growth of the economy. The 1957-58 Recession Causes and Characteristics of the MainlanS Recession The recession of 1957-58 lasted nine months, starting with a peak in the second quarter of 1957, and reaching a trough in the first quarter of 1958. Although it was shorter than the previous recession, it was in many ways a sharper setback for the economy. GNP, gross private investment; and industrial production declined more both in absolute and relative terms;' inventory liquidation was greater and unemployment was higher than in 1953-54. (Compare data in Table 16 to that in Table 12.) The downturn of 1957-58 originated in changes in a number of sectors «developing at the same time mainly by coincidence... Each exerted a growing downward pressure on production and employment, and ultimately their joint effect was to tip the balance... toward a general reduction in economic activity». (12), First of all, there was a drastic decline in investment activities (gross private investment fell by $ 16 billion from peak to trough compared to $ 7 billion in 1953-54), which can be attributed to a much faster growth of capacity than its utilization. For example, in manufacturing, expenditures on new plant and equipment between 1953 and 1957 had increased capacity by about 25%, but production had risen by only 7%. In the same period the cost of new capacity had risen substantially more than prices of industrial output; and in 1957 the financing of expansion programs was becoming more costly (while corporate liquidity was very low). (13). Secondly, there were some sharp reductions in U. S. exports. This was partially a reaction to the (12) Economic Report of the President, 1959, p. 10. (13) Ibid., p. 12. Most of what follows on the U. S. recession is based on information contained in the President's Report.

THE POST-WAR RECESSIONS

4T

a

high volume of exports in the latter part of 1956 and early 1957 to which the Suez crisis had given impetus. The rising volume of exports spurred on by the crisis actually contributed to the maintenance of an upward trend in the economy at a time when investment demand was already leveling off. But the disappearance of the special Suez situation coincided with actual decreases in investment activity, and the volume of exports decreased by 20% during the contraction. Although the initial export decline was due to smaller exports in wheat, cotton, and petroleum,. it soon spread to metals, machinery and manufactured goods, reflecting a slackening in the expansion of European countries. and their build-up of large stocks. A third contributary factor to the recession was a curtailment of expenditures on consumer durables of 11% (much larger than 'in 1953-54), four-fifths of the decline being accounted for by reduced purchases of automobiles. This is remarkable not because cuts in consumer purchases are unusual in a recession, «but the 1957-58 decline was unusually large in view of the small reduction — about one per cent — in the amount of income available to consumers after payment of taxes. (14) Finally, there was a slight reduction in the purchases of the Fe= deral government. These principally independent changes in demand had a magnified reaction on the .economy because of the induced changes in the inventory policies of businesses. There had been a considerable increase in inventories from 1955 until the middle of '1957. In the third quarter, with uncertainties about future sales and with increased difficulties and costs of financing inventories, the rate of accumulation slackened, and by the last quarter of 1957 actual inventory reduction began and reached peak in the first quarter of 1958 when, the rate of reduction rose to 8.2 billion dollars per year. And thus, «a moderate decline in the pace of actual sales produced a much larger effect on sharp swing from production and employment by provoking inventory accumulatiou to inventory liquidation». (15). Over half of the decline was concentrated in the machinery, .aircraft, and: automobile industries, but substantial cuts were also made is the petroleum, chemical, paper, and consumer household durable goods industries. Althóugh inventory liquidation. was the induced factor in this recession, it ultimately accounted for about 60% of the entire decline in GNP. (14) Ibtd., p. 13. (15) Ibfd., p. 14.

48

THE POST - WAR RECESSIONS

The maintenance of over-all consumer demand seems to have been a principal factor in accounting for the short duration of the recession and its relatively small impact in the face of sharp inventory and investment curtailments. Although GNP fell by 20 billion dollars on an annual rate from the third quarter of 1957 to the first quarter of 1958, personal income declined by less tham 4 billion dollars and aggregate consumer expenditures by less than 2 billion dollars. The small decline in personal income is accounted for mainly by a decline in earned personal income which is small relative to the decline in production, (16) an increase of transfer payments, and a reduction in personal tax payments. Also, consumers, in their turn, reduced their purchases of goods and services by an amount smaller than their fall in income. The slight fall in the sales of consumer non-durable goods, the continued expansion of service industries, the continued rise in expenditures of local and state governments, and the rising earnings of the labor force still employed, all contributed to an early halt in the downturn. By the spring of 1958, Federal government expenditures began to rise again, the decline in business expenditures on plant and equipment came to an end, and there was a considerable expansion 'in home building activities. Inventory liquidation continued, however, but at a diminished rate, until the last quarter of 1958. The recovery was underway in the second quarter of 1958, but at a sluggish rate. It continued throughout the year and picked up speed in the last quarter. .Effects of the Recession on Puerto Rico The impact of the recession on Puerto Rican exports seems to have been rather slight. When taking account of seasonal 'factors, it would seem that the principal effect was in a slackening of the rate of growth of total exports during 1958. (See Table 16.) A more detailed glance at individual export items, some of which are reproduced in Table 17, leaves the impression that (16) For every decline of $1 billion in private output, personal income earned fell only $360 million; in 1929-30, it fell $670 million. According to the President's economic advisors, this can be explained by the fact that a larger proportion of our labor force is now employed in industries and occupations not quickly affected by moderate changes :in business conditions; and even in industries that are more sensitive to business fluctuations, a larger proportion of employees hold positions not usually affected by 'production cutbacks of short duration. Ibid., p. 15.

THE POST - WAR RECESSIONS

49

TABLE 16 THE RECESSION OF 1957-58 - COMPARATIVE DATA U. S. Data (Billions of dollars) *

1957 1 2 3 4 1958w 1 2 3 4 1959 1 2

GNP

Consumption NonDurable Durable

429.9 441.2 445.6 438.9

35.9 39.5 40.4 39.6

137.3 137.1 140.5 138.8

63.6 67.0 66.7 61.5

.0 2.9 2.2 2.2

146 144 142 141

427.1 434.5 444.4 457.1

36.3 36.7 37.1 39.8

139.8 141.5 143.1 143.6

50.9 51.3 54.2 61.3

-8.2 -5.8 -3.4 .8

130 128 133 142

470.2

41.3

145.3

69.8

6.1

146 153

Gross Private Investment

Changes in Business Inventories

Index of Industrial Production

Puerto Rican Data (millions of dollars) **

1957 1 2 3 4 1958 1 2 3 4 1959 1 2

Total Exports

Total Imports

Unemployment

New Plants DisPlants continuing Operations

Index of Industrial Production (1952-54 = 100) Total Non-Durable Durable

90.4 111.4 122.9 105.5

172.5 161.9 131.9 1742

15.7 10.0 12.8 13.1

30 23 28 31

186.2 225.2 216.0 230.3

183.6 215.6 209.0 224.1

193.1 249.9 234.0 246.2

91.6 122.9 125.4 111.8

161.6 151.7 150.1 165.3

14.9 10.5 13.7 16.6

25 20 32 22

221.0 245.6 252.4

215.6 233.7 249.4

234.9 276.4 260.4

111.6

174.9

15.9 10.6

30

5 3 4 6

* Seasonally adjusted, except Index of Industrial Production. ** No seasonable adjustment. SOURCE: See Table 12. 5

iC

50

THE POST - WAR 'RECESSIONS

a)

TABLE 17 U. S. INDUSTRIAL PRODUCTION AND PUERTO RICAN EXPORTS (1957-58) Fabricated Metal Products

NonElectrical Electrical Machinery Machinery

Machinery & Vehicles

Textile Mill Products

Apparel

Textiles (Total)

Paper & Allied Products

U.S.

P.R.

U.S.

U.S.

P.R.

U.S.

U.S.

P.R.

U.S. P.R

137 138 140 137

1.4 1.7 1.5 1.8

159 155 146 141

208 193 205 209

8.5 8.7 9.6 10.2

104 100 96 96

118 112 107 105

28.6 29.0 30.3 31.9

160 1.f 160 1.` 156 1.º 158 .S

124 120 132 136

2.0 2.6 3.6 3.7

132 124 122 130

182 162 180 194

7.0 8.6 10.1 16.5

94 94 97 105

109 106 110 114

26.3 25.4 32.1 33.2

155 •4 156 .4 160 .d 168 1.3

136 2.6 146

141 152

199 202

10.4

111 (118)

126 (127)

28.3

173 1.1 (183)

Chemicals

Veg. Food & Beverages

Non-Met. Minerals

Textile Fibers & Mfgs.

Machinery & Vehicles

Chemicals & Rel. Products

U.S P.R.

U.S. P.R.

1 2 3 4

185 1.5 182 1.3 180 1.6 187 2.7

103 22.1 109 38.4 122 41.8 115 23.4

8.5 7.3 7.3 5.7

25.2 25.6 21.7 19.6

27.8 23.4 18.1 25.0

18.0 14.7 15.4 17.0

32.4 34.4 24.1 31.9

12.2 12.9 1.07 12.7

1 2 3 4

181 1.9 178 1.6 181 1.8 195 1.2

104 19.3 111 44.9 125 38.3 119 16.6

7.1 4.5 5.7 5.3

26.0 242 22.7 27.2

22.4 21.3 21.4 22.9

17.8 16.3 18.0 16.7

31.1 31.6 27.9 33.6

13.1 13.3 13.6 11.1

1 201 1.2 108 27.5 2 (207) (115)

6.2

27.7

24.0

19.1

36,3

14.1

1957 1 2 3 4 1958 1 2 3 4 1959 1 2

hI SELECTED P. R. IMPORTS Vegetable Animal & Food Prod's Animal & Beverages Products

1957

1958

1959

\

Souaca: See Table 12.

THE POST - WAR RECESSIONS

51

they did not follow their industrial production counterpart on the Mainland. While there was a substantial cut in the production of fabricated metal products, Puerto Rican exports continued to rise throughout the period; much of this can be explained by the high rate of opening of new firms in that industry during the first three quarters of 1958. A sharp drop in the production of machinery products on the Mainland was reflected in a relatively mild decrease in export of those products from the island in the first quarter of 1958. The decline in the production of textile products, which was somewhat smaller than the general decline of industrial production, had a definite effect on Puerto Rican textile exports which slumped in the first two quarters of 1958 to a proportionately larger extent. In this case the explanation must lie with a substantial curtailment of production of some plants, since the rate of openings of new plants did not drop to any marked extent in that period. Paper and Allied Products, which experienced only a very .small drop in Mainland production, took a relatively sharp drop in exports in the first half of 1958, declining by more than fifty per cent. Chemical exports declined throughout 1958, despite the relatively small slump in Mainland production. The effects or employment of the Mainland recession are reproduced in Table 18. Unemployment increased substantially above the seasonal average in the last half of 1958,. especially in the fourth quarter. At its height, the recession-induced unemployment was higher than in 1953-54, but this was also the case on the Mainland. A glance at individual industries in part b) of the Table reveals a noticeable decrease in employment in the Textile Mill Products and Apparel industries; they reached their low points, in the first and second quarter of 1958 respectively, but recovery was slow throughout the rest of' the year, though weekly hours worked increased rapidly and were above the 1957 average throughout the latter part of 1958. It seems that as production expanded again, firms were slow in their re-hiring policies, preferring to extend hours worked of employees still on the job; and also that new firms preferred hiring fewer workers, but for longer hours. 'The steady rate of employment in the Paper industry, in spite of the slump in exports, is explainable by the lack of comparability — i. e., employment figures include workers in the printing industry —and also by the fact that close to 70 % of this industry's output is. sold domestically. The prolonged slump in chemical product exports is paralleled by a slump in employment in that industry„

52

THE POST - WAR RECESSIONS TABLE 18 EMPLOYMENT, UNEMPLOYMENT AND MIGRATION, 1957-58 (Seasonally Unadjusted) a) Total (Percentages) U.S. Unemployment

P. R. Unemployment

Net Migration

Seas. Adjusted Yearly Rates

C 'OOOs)

1957 1 2 3 4

4.6 4.1 3.9 4.5

15.7 10.0 12.8 13.1

77 -139 53 +118

-59 -25 -35 -32

1 2 3 4

7.4 7.4 6.7 5.7

14.9 10.5 13.7 16.6

-64 -105 -41 +99

-32 -5 25 -67

1 2

6.8 5.3

15.9 10.6

1958

1959 (-40)

Source: See Táble 14a.

b) Selected Industries Food & Kindred Products Total Average Employment Weekly Hrs.

Textile Mill Products

Apparel & Related Products

T. E.

Hrs.

T. E.

Hrs.

T. E.

Hrs.

T. E.

Hrs.

Paper & Allied Products

Chemicals & Allied Products

1957 1 2 3 4

17.1 18.4 13.4 14.0

36.8 36.4 35.6 35.3

4:4 5.1 5.1 4.8

34.6 38.7 35.2 35.8

19.1 18.2 18.7 18.9

32.7 33.1 33.7 33.3

2.0 1.9 2.0 2.1

40.1 40.6 41.1 40.5

2.3 2.6 1.4 1.3

35.3 37.6 33:7 35.1

1 2 3 4

17.1 19.0 13.7 14.8

37.4 37.6 35.8 35.1

4.2 4.3 4.8 4.8

33.3 35.5 37.0 36.7

18.0 16.8 17.1 17.9

32.5 32.0 33.9 34.1

2.0 2.0 2.1 2.1

39.9 40.6 40.2 40.7

1.3 1.4 1.5 1.4

33.4 36.0 36.1 33.3

1 2

18.8

38.0

4.8 36.0

1958

1959 17.7 33.9

2.3 40.0

1.4 34.9

53

THE POST'- WAR RECESSIONS TABLE 18 (continued) b) Selected Industries (continued) Lumber & Wood Products Total Average Employment Weekly Hrs.

Stone, Clay & Glass Products T. E.

lira.

38.2 38.0 38.5

1.4 37.6 1.6 38.3

37.1 36.7 37.0 36.9

1.7 1.8 1.9 1.9

T. E.

Hrs.

1957 1 2 3 4

3.2 3.3 3.0. 3.3

38.0, • 37.0 38.7

4.2 4.1 3.7 4.3

1 2 3 4

3.2 3.1 3.0 3.2

37.4 36.7 37.6 38.6

4.1 3.8 3.8 3.9

1 2.

3.2

3.9

Machinery (Except Electrical)

Electrical Machinery

1958

1959

T. E.

1957

Hrs.

T. E.

37.7 37.8 37.9 37.8

1.9

Hrs.

Leather Products

Metal Products

T. E.

36.3 37.1 38.4 36.1

2.1 38.4 2.1 38.6 22 39.5 2.0 38.8

25 2.6 2.8 2.8

2.0

2.9

Metal Products & Machinery (Incl. Electrical)

Hrs.

T. H.

Hrs.

37.5 38.3

1.4 1.4

352 38.4

3.3 3.4

392 37.2

1.4 1.4

34.6 36.9

5.8 5.8 6.1 6.4

1 2 3 4

1.4 1.3 1.4 1.4

38.2 36.2 37.8 37.2

2.9 2.7 2.8 3.0

33.5 34.0 365 38.9

1.3 1.2 1.2 1.3

36.0 35.0 35.0 38.8

6.0 5.8 6.1 6.3

1 2

1.4

1959

Souitea: Seo Table 14b.

3.1

Hrs.

2.0 38.7 2.1 39.6

1 2 3 4

1958

T. E.

2.3 37.0 2.4 2.3 37.3 2.5 37.2

Professional & Scientific Equipment T. E.

Hrs.

Petroleum

1.4

~~

54

THE POST - WAR RECESSIONS

while the continued rise in the exports of metal products is accompanied by a very small rise in employment and hours worked, possibly due to increases in productivity. Finally, the decline in machinery exports, was accompanied by a decrease in employment of a relatively mild nature; and a more pronounced decrease .in the number of weekly hours worked. A broader resume of the employment situation is obtained from Table 19. The decline in agricultural employment has continued and its rate was not affected to any degree by the recession (as a matter of fact, it decreased between 1957 and 1958). The rate of increase in manufacturing came temporarily to a halt, while construction declined, and Home Needlework continued its steady decline. The decline in employment bettween 1956 and 1957, and the small increase in 1958, which contributed substantially to the proportionately much greater rise in unemployment because of the usual decline in emigration and the continued natural growth of the labor force, can be traced to a large extent to the manufacturing sector and especially to the export-oriented industries. Although there does not seem to be any obvious decline in that sector, it should be remembered that Puerto Rico relies on continued and increased growth of manufacturing industries to absorb the labor force released from such declining industries as agriculture and home needlework. Seen in this light, the stoppage of .growth in employment in manufacturing, and the actual more than seasonal TanLE 19

EMPLOYED WORKERS BY INDUSTRY GROUP (in thousand's)

Agriculture Manufacturing Home Needlework Mining Construction Trade Services Government Total Source; Economic Report to the Governor, 1958.

1956

1957

1958

160 74 25 1 38' 93 71 55

150 79 20 2 40 94 67 57

149 79 12 2 37 97 72 62

558

552

555

THE POST- WAR RECESSIONS

55

decline which occurred in 1957-58 were important contributory factors of the Mainland recession to the island. This is also, evident in a discussion of the employment situation in the 1958 Economic Report to the Governor, where it is noted that employment in April 1958, was 5000 less than a year earlier, and that 1.7 thousand of this loss was in EDA-promoted manufacturing plants. It is further claimed, however, that «taking into account the new jobs created by plant openings during the year, it is apparent that employment cutbacks in those EDA firms which were in operation at the outset of the fiscal year, as well as declines resulting from plant closings must have totaled more than 4.0 thousand». (17) The Mainland recession also made itself felt through a decline in the openings of new plants, but to a lesser extent than in the previous recession. This is surprising in the light of the sharp decreases which occurred in Mainland investment. activities. The plant promotion program suffered even less (the recession causing delays in actual plant establishment of promoted firms). The picture which emerges in the expanding sector of the economy is of a small decrease in the number of promotions and a somewhat larger decrease in the number of openings of new plants due to the greater recession-induced lags between promotion and openings. Usually, «two thirds of the plants opening in a given year are promoted in the same year. In 1957-58, less than half of those opened were promoted during the year, the majority being carryovers». (18). The net increase in new plants was much smaller, due to the rise in the number of plants discontinuing operations (almost double the number of closings in the previous year). With 36 closings in the fiscal year 1957-58, and 97 actual openings, the net increase of plants was down to 61. Decreases in employment in EDA-promoted plants were for the most part concentrated in older plants. The bulk of the decline occurred in plants established prior to 1955 (3.2 thousand), while only a small decline occurred in plants established between 1955 and 1957 (900). Plants established in the first three quarters of 1958 provided employment for 2.5 thousand workers, while employment losses due to closings in the same period (occurring in random fashion among firms in all age groups) amounted to 1.9 thousand workers. It seems that this (17) Economic Report to the Governor, 1958, p. 19. (18) Ibid., p. 10.

56

THE POST - WAR RECESSIONS

was a pattern similar to 1953-54, when the bulk of the employment decline occurred in plants established prior to 1951. One explanation offered is the tendency «for newly established firms to gradually hire more workers during their first few years of operation than are eventually needed to maintain capacity output levels after peak productivity has been reached». (19) Behavior of Imports There occurred a much stronger decline in total imports during the 1957-58 period than during the previous recession, and as in the previous recession it took place without a lag. Much of the decline was concentrated in textiles and machinery. Machinery had not declined at all in 1953-54, whereas in 1957-58 it accounted for one third of the decrease in imports which took place between the second and third quarter of 1957, obviously reflecting a decrease in investment activities on the island. The 1957-58 recession thus revealed a pattern not discernable previously. There was a stronger cutback in the operation of existing firms than in the previous cycle, which was not offset by continued vigorous expansion of new firms. This leads one to conclude tentatively that the present industrial structure of Puerto Rico is increasingly sensitive to Mainland fluctuations and that a vigorous underlying growth factor would be required as an immunizing agent; and this sensitivity might become even greater as industrialization continues and as a larger percentage of firms come to be three-years-old or more. The lagged relationship between Mainland and Puerto Rican economic fluctuations which appeared in 1953-54 seems to have been considerably reduced. For example, the slack in openings of new firms became obvious in the first quarter of .1958, two quarters after the beginning of the Mainland recession. This was also true in .the behavior of individual industries (see Table 17): declines in exports became noticeable in the first quarter of 1958, about six months after Mainland industrial production began its decline, and the trough was reached in most 'cases after the same lag. It is difficult at this stage to advance any definite reason for the shorter lag. The Economic Report to the. Governor attributes it to «the greater sharpness in 1957-58 of inventory liquidation and of reduction in plant and equipment (19) Ibíd., pp. 10-11.

THE POST - WAR RECESSIONS

57"

expenditures in the United States. Another reason in the increased role in the Puerto Rican economy of investments from the Mainland and of Mainland markets for local products.» (20). But one fails to see factors which should necessarily account for shorter lags. Possibly the greater sharpness induces investors•. to postpone plans sooner or affects Mainland customers' decision to curtail orders sooner. Possibly also the structure of the Puerto'. Rican economy has changed to the extent that a smaller number of firms operate on a contractual basis with Mainland customers• and thus are hit by fluctuations in the market sooner than, before. Migration Just as in the previous recession, there occurred a sharp decline in net emigration. The decline, however, was not as abrupt and .as prolonged as in 1953-54. (See -Chart I.) The bulk of the decrease was concentrated in the first half of 1958. The= general impression which Chart I gives is of a softening in the impact which the Mainland recession had on net emigration,, especially when one considers the shortness of the dip. This could possibly be attributed to greater long-run job security and opportunity which Puerto Ricans on the Mainland find, especially after having been there for over a decade = and hence a lower re-immigration of' the jobless. Also, the 1958 dip was attributed principally to female migrants, who are less often to be found_ in the labor force. Thus the impact on the Puerto Rican labor force was much less severe than it might have been, and a much_ smaller proportion of the rise in unemployment could be attributedto the migration factor than in the previous recession. It should also be remembered from our description of thecharacteristics of. the Mainland recession that its impact was much less severe, or even .absent, in the services industries and non-durable goods sectors which employ a large number of Puerto .Ricans. Actions of the Commonwealth' Government As in the 1953-5'4 period, .the Commonwealth government did not pursue any consciously anti-deflationary fiscal ,policy, al. though its budget turned_ out to be of such a nature. From 1957 (20) Ibíd., p. 6.

~

58

THE POST- WAR RECESSIONS

to 1958 total government receipts (excluding the sums borrowed) increased from $ 198 million to $ 204 million, a relatively small increase compared to previous years, while expenditures increased from $ 231 million to $ 248 million. Most of the increase is attributed to increases in operating expenses, rather than for capital purposes, which remained constant. There was, however, a considerable increase in expenditures for economic .development; i.e., for industrial and agricultural development transportation and communication, and rural electrification — this amounted to $ 10 million in all, and thus increased the proportional share of this type of expenditure in the total budget :from 17% to 20%. The entire expenditure for capital programs was financed from the sale of bonds and the drawing down of cash balances. The net debt of the island rose from $ 49 million in June, 1957 to $ 61 million in June, 1958. The debt limit, which amounts to ten per cent of the assessed value of property at the end of fiscal year rose from $ 116 million to 1 158 million, thus leaving ample room for further deficit financing. But, as was mentioned above, the budget was never con sciously used as an instrument of business cycle policy. KConclusion The 1957-58 recession, which was more severe but shorter rthan the previous one, had on the whole a relatively mild effect on Puerto Rico. The island gross product continued to grow, though at a rate smaller than projected without the recession ~(i. e., growth in current values of 6.8 instead of a potential :10%-13%). Gross domestic fixed investment, which had 'increased by $ 51 million in 1956-57, rose by only $ 11 million in the following year. This was especially influenced by private investment in machinery and equipment which increased by $ 22 million in 1956-57, and decreased by $ 3 million in 1957-58. Much of this decrease in private investment was due to a slowdown in new plant openings, which was reflected in balance of payments figures by a decline in the inflow of long-term capital, -which had increased by $ 80 million in 1956-57 (from $ 72 million -to $ 152 million), and which decreased by $ 3 million in 1957-58 to $ 149 million. There was an induced effect on unemployment, but unlike the 1953-54 experience, when the principal channel through which

THE POST - WAR RECESSIONS

59

the effect was propagated was the slump in net emigration, the .principal causative factor in 1957-58 was a widespread curtailment of production of long-established export-oriented industries. Although the decrease in the number of openings of new firms was not too drastic, the net increase in employment resulting from the opening of industrial plants was not enough to counteract the effects of the recession on older firms.

PART II THE BEHAVIOR OF PUERTO RICAN FINANCIAL INSTITUTIOÑS OVER THE CYCLE Some General Observations In a study which is primarily concerned with interregional economic relationships, monetary relations should not be neglected since they form an integral link in the chain that binds regions into a national economic unit. In the case of Puerto Rico, monetary relations with the Mainland are not necessarily the principal element explaining the basic nature of interactions between the island and the Mainland. They are, however, the links through which fluctuations are passed on, and, in some cases, through which induced fluctuations are magnified or dampened. The simplest approach is to view monetary relationships through the balance of payments of Puerto Rico with the Mainland. This presents the characteristic situation of a regional as opposed to an international payments system, where the region, unlike the country, does not have at its disposal the power to change the value of its currency, to restrict payments. to impose quantitative restrictions on trade, etc., in order to adjust the payments to its liking. The outcome of this is «either continuance of free movements and a nondepreciated dollar or a complete cessation of payments and a disguised deprecia tion.» (21) (21) Seymour E. Harris, International and Interregional Economics (New York:

1957), p. 176.

61

•

62

THE POST - WAR RECESSIONS

Actually one never hears about inter-regional balance of payments difficulties due to the equilibrating flows of capital between regions of the same country. For example, a booming region which imports more than it exports to other regions will receive capital from regions experiencing lesser or no growth, and hence an excess of loanable funds. It is even, claimed that the flow of capital should more than offset the deficit because the entrepreneurs' demands for funds will exceed the amount needed for plant, equipment, and inventories by the capital (circulating) required in order to operate to the new capacity. (2Z) In cases where capital does not fill the gap, or where economic stagnation reduces exports at a faster rate than imports, and there is no attraction for autonomous capital movements to make up the gap, adjustment would probably have to take place via contractions in investment, income, imports, and migration; in other words, the mechanism would be parallel to the gold standard adjustment. (23) A net capital inflow into a region creates magnified repercussions in a decentralized fractional reserve banking system. If the booming region experiences a net capital inflow, banks will become móre, liquid and the resulting increased loans could further accentuate investment activities, or provide funds for running the new capital equipment, or they might be used for further imports, in which case further expansion of credit will be checked. The Changing Structure and Importance Banking System

of the Puerto Rican

The most impressive change that has taken place in the Puerto Rican banking system is the growth of its importance. While gross product has increased by 50%, and per capita by 69 %, since 1950, demand deposits have increased by 73 % and time and saving deposits by 233 %; and while investment has grown by 82 %, bank loans have increased by 160 %. All this compares quite favorably with the U. S'., where bank loans and deposists have increased only 93 % and 29 % respectively in the same period. (24) One principal reason for this phenomenon is (22) Tibor Scitovsky, Economic Theory and Western European Integration, pp. 87-91. (23) Ibfd. (24) Current Business Statistics (Bureau of Economics and Statistics, Puerto Rico

Planning Board), September, 1958.

THE POST - WAR RECESSIONS

63

that as an area industriR1i7esT as per capita income increases,. as the percentage of workers employed in non-agricultural actívities increases, greater use will be made of the banking system as a repository of increased savings and a greater use will be made of checking accounts for transaction purposes. Very often the use of currency relative to checking accounts is used as a measure of development. That currency still plays a relatively large role in Puerto Rico can be seen by the fact that while per capita income is about one-fourth of the U. S. per capita income, total bank deposits amount to one-seventh, and private deposits to one-ninth, of those on the Mainland. (25) There are 11 commercial banks operating in Puerto Rico, with a total of 86 banking offices throughout the island. This is a considerable .growth, when one considers that only 38 offices existed in 1945; and 'S1 in 1950. Although this has been considered to represent a fairly good coverage for an island 100 miles by 35 miles, there are still some areas without banking services. To remedy this situation, a law was passed authorizing «bus banking» or mobile banking units to cover municipalities whose needs do not require full-time banking offices. (26) Basically there are three types of commercial banks which operate on the island: seven locally chartered banks, two branches of Nevi York banks, and two Canadian banks. Of the local banks, three are large and four are relatively small. These local banks account for over 60 % of total assets of half a billion dollars, and for 50 % of total loans made by commercial banks. This represents a considerable degree of expansion for domestic banks when one considers that in 1942 they only controlled 37% and in 1950, 56% of total assets, while their share of private deposits increased from 44% in 1942 to 58% in 1950 to 64% in 1958 (See Table 20). This is not true in the case of loans, their share of which amounted to 47 `% of the total in 1942, 54 % in 1950, and 51 % in 1958, which, according to Wallich, reflects the fact that foreign banks tend to bring «head office funds to Puerto Rico instead of exporting Puerto Rican funds to the head office.» (27) In general terms, the fast growth of banking can obviously be explained by effects of the industrialization program of EDA, (25) Henry C. Wallich, Banking Facilities in Puerto Rico, report to Government Development Bank, September, 27, 1958, p. 3. (Mimeographed.) (26) John S. Deieers, =Instruments of the Capital Market is Puerto Rico,» (Government Development Bank for Puerto Rico), June, 1957, p. 6. (27) Wallich, op. cit., p. 4.

ffi4

THE POST - WAR RECESSIONS

the accompanying increase in capital imports, changes in the banking structure — e. g., the introduction of FDIC, the higher interest rates on savings deposits, etc. Also, the relatively faster growth of domestic banks can be accounted for by their more intimate acquaintance with the economy, which makes the penetration into areas of potential demand for credit easier. One of the principal functions of the banks and other financial institutions seems to be the fostering of more local financing of capital formation. Up to now the bulk of Puerto Rico's rapid progress was financed by massive inflows of capital. The largest proportion of bank loans is still granted to commerce and industry, and this is followed by loans to individuals; these make tup 50% and 20% respectively of total lending. Loans on real estate =have, been increasing fairly rapidly in recent years, making up 16% in 1958, due to increases in real incomes, the availability of Federal Housing Administration facilities, and the marketability .'of FHA-insured mortgages in the U. S. (Z8) Before we' can make any statements as to the behavioral pattern of the banking system in relation to Mainland economic and financial developments, a brief examination of the legal TasI.E 20

GROWTH OF BANKING AND FINANCIAL INSTITUTIONS IN PUERTO RICO [in millions of dollars] 1950

1958

°/, Change

Demand Deposits Time and Saving Deposits Government Deposits Loans Investments

107 51. 84 120 97

185 170 70 312 82

73 233 —17 160 —15

Government Dev. Banks

34.5

62.2

80

Gross Product Per Cap. Inc. Investment

753 278 150

1.286 469 274

59 69 82

Commercial Banks

4

(28) Current Business Statistics, op. cit., p. III. .`SOURCE: Comparative Financial Conditions o} Commercial Banks in Puerto Rico, Commonwealth of Puerto Rico, Department of the Treasury.

THE POST - WAR RECESSIONS

65

in

structure of the system is in order. The extension of federal deposit insurance to Puerto Rican institutions in 1950 considerably strengthened the position of locally chartered banks, and partly explains their phenomenal absolute and relative growth. A special act of Congress 1952 required the two national banks to insure their deposits in Puerto Rico with the FDIC. But the Canadian banks, being foreign corporations, are not eligible for membership. (29) Another important factor in bolstering the banking system was the establishment at the request of Puerto Rico of a United States Treasury cash depot on the island in 1952. This assures a ready supply of coin and currency for the community and considerably lessened the need of banks to carry large reserves of vault cash. (30) Since 1952 improvements have been made in modernizing the island's check-clearing mechanism, thus diminishing the size of the «float», the uncollected balances. (31) A most interesting feature of the Puerto Rican banking system is the reserves of banks. All the commonwealth chartered banks are required by the government to hold minimum legal reserves of 20% against demand liabilities, which have to consist of at least one-third of 20%, i.e., 6 2/3% in vault cash, 3-1/3% in the form of checks on island banks held for collection, and the rest in the form of interbank deposits. Puerto Rico's Commercial. Banking Code gives the island's treasurer the power to change minimum reserve requirements (he may raise it up to 30 %). The two Mainland banks fall under the jurisdiction of the Federal Reserve System. The two Canadian banks fall under the jurisdiction of the commonwealth, with the additional requirement that they must keep in Puerto Rico 75 % of their total deposits in either cash, loans, property, or securities of the federal or local government. The Canadian and national banks have recourse to their respective head offices for funds, the latter also having access to- the Federal Reserve System. Local banks have established lines of credit with correspondents. (32) The Puerto Rican banking system has been traditionally concerned with the «excess r- eserves» problem, which has -been diminishing somewhat in the last few years. Reserves in excess of the amount legally prescribed have been traditionally extremely high in Puerto Rico. In the late 1940's they amounted to

a

(29) Biagio DiVenuti, Banking Growth in Puerto Rico (1955), pp. 9-13. (30) «Puerto's Rico's Monetary System .and Economic Growth,» Monthly Review of .Credit and Business Conditions (Federal Reserve Bank of New York), March, 1958. (31.) See ibíd. (32) Di venuti, op. cit., p. 61.

66

THE POST-WAR RECESSIONS

over 100%; since that time they have been declining steadily, reaching a low of 23 % in the third quarter of 1958, which, however, is still considerably higher than the level in Mainland banks. The reason for traditional excess reserves is that locally chartered banks, not being part of the Federal Reserve System, do not have immediate accesss to a Reserve bank for accomoda tion in case of unexpected drains on their reserves. The principal function of «excess reserves» then, «is not their role in determining the extent to which banks may expand their loans and investments, and thus the volume of their deposits; rather, their function, in the absence of direct recourse to Federal Reserve facilities, is that of insuring at all times the capacity of these commercial banks to meet the withdrawal of demand of customers» (33)• Need for ready access to credit has been especially conspicuous to banks located outside of the metropolitan San Juan area where the United States Treasury's Cash Custody Account is held, hence more reserves are needed for effective day-to-day operations. Another factor, which is of especial importance in the context of this study, for holding excess reserves is «the need of maintaining adequate reserve cushion to absorb the impact in Puerto Rico of any disrupting economic influences originating on the Mainland (34)• The implication of these excess reserves is that multiple credit creation of increased deposits is considerably dampened. There .has been a substantial contraction in excess reserves since the early fifties which resulted from the pressure exerted by expanding credit demand that accompanies the industrialization program and all its ramifications. But it has been made possible also by a number of actions taken since 1950 which strengthen Puerto Rico's banking system. There was the extension of Federal Deposit Insurance to Puerto Rican banks in 1950, which tremendously bolstered the confidence in the latter, partially causing a substantial growth in deposits and a more extensive utilization of bank resources. There was also the establishment of the U. S. Treasury cash depot, at Puerto Rico's request, in 1952, which assured a readily available and adequate currency and coin supply; thus the former need of Puerto Rican banks to hold large reserves of vault cash was considerably reduced. Substantial improvements have also been made in streamlining the island's check-clearing mechanism, substantially (33) DiVenuti, op. cit., p. 63. (34) Ibid., pp. 64-65.

THE POST = WAR RECESSIONS

67

(35) «Puerto Rico's Monetary System and Economic Growth,

n

reducing the volume of the «float», i.e., uncollected balances. By agreement among all banks the Government Development Bank was designated as Settling Agent in the collection of intraisland checks. An important link with the Federal Reserve System was established when the Government Development Bank was allowed to establish a non-member clearing account with the Federal Reserve Bank of New York, a link which facilitates the settlement of balances and their transfer to and from the Puerto Rican banks' correspondents in New York (35)• Finally, on January 1, 1958, Puerto Rico became part of the second Federal Reserve District for purposes of check collection_ This step was taken in conjunction with the establishment of par clearance of checks by Puerto Rican banks. It should enhance the mobility of funds between the Mainland and the island. All these institutional changes do much to explain the steady and considerable reduction which has taken place in the holding of excess reserves, the stepped-up activities in lending of the banking system, and the rapid growth of deposits, especially savings deposits. An examination of quarterly banking statistics from 1953 to 1958 leaves the impression that financial channels are not one of the principal connections to the mainland through which economic fluctuations are passed on, and, as a matter of fact, it would seem that they are hardly affected by Mainland fluctuations. Total private deposits, which are reproduced on Chart II, have been rising steadily without an interruption. Only in 1958 does there appear to be a decline in the rate of growth, attributable principally to decreases in the demand deposits in U. S. banks. There appears to be some reactions to Mainland conditions in total loans made by banks (see Chart III), and this is mainly attributable to Mainland subsidiary banks, which curtailed their lending activities in 1953-54 and again in 1957-58, while locallychartered banks continued to expand their loans steadily throughout those periods. Part of the explanation might lie in the fact that Mainland corporations which curtail some of their operations during a Mainland recession borrow funds for operations from U. S: banks rather than from locally-chartered banks. Furthermore, considering the high excess reserve position of banks, especially local banks, which institutional changes have r` endered obsolete, and the increasing use of the credit system op. cit., p. 3.

CHART II LOANS AND DISCOUNTS Millions of $ 70 50 30 10

310 290 270

250 230 210 190 170 150 40 30

CANADIAN BANKS

20 10 ~

~ I i I i T ~ ~ , i , , 1 I 1 2 3 4 1 2 3 4 1 2 3 4 1 __2 3__J4 1 2 3 4 1 2 34 ---~

1953

1954

~

----

1955 68

1956

J

1957

`

ti

1958

Millions of $ 130 120 110 100 90 80 70 60 50 40 160 150 140 130 120 110 100 90

i• i t 1 , I l I l I ! I I I 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4

I

I

I

(

1

1

1

1 2 3 4 1 2 3 4

1953 1954 1956 1955 1957 1958 Source: Comparitive Financial Conditions of Commercial Banks in Puerto Rico, Department of the Treasury, Commonwealth of Puerto Rico: 69

CHART III TOTAL DEPOSITS Millions of $ 350 330 310 290 270 250 230 210 190 CANADIAN BANKS

20 10

I

1

i

I

I

1

1

I

I

1

1

110 • 90-

U.S. BANKS

70 50 70

I

1

1

I

1

1

1

THE POST - WAR RECESSIONS

71

Millions of $ 220 200 180 160 140 120

1 2 3 4 1 2 3 4 1 2 3 ~

ti

1953

~

\______

1954

\

ti

1955

4

1 ~

2 3 4 Y

1956

1

2 3 4 ----

1957

1

2 3 4 ---

1958

Source: Comparitive Financial Conditions of Commercial 8añks in Puerto Rico, Department of the Treasury, Commonwealth of Puerto Rico. by local merchants, producers, and consumers to whom local banks can more adequately cater, it 'is not surprising that continuing advances have been made in the loans of local banks. The possibility and opportunity of expanding loans throughout the economy was thus stronger than any adverse cyclical reaction. This does not mean that credit tightness on the Mainland has but little effect. Tighter U. S. Mainland credit conditions will be felt in Puerto Rico through the increased difficulty of selling mortgages and through the higher costs of bond issues. For example, in 1953, when for a while credit was very tight, «the impact in Puerto Rico was felt particularly through the difficulty in disposing of FHA mortgages. Residential construction suffered a corresponding setback.» (36) But, on the whole, there has been no severe restraining effect either on the expansion of the banking system or on the expansion of the economy due to credit unavailability. it is very interesting, for example, to note that commercial bank rates in Puerto Rico occasionally even move in a different direction from Mainland rates. (See Table 21.) We are left with the conclusion that the banking system of Puerto Rico has hardly been affected by Mainland fluctuations (36) Vernon R. Esteves, =Statement to the National Monetary and Credit Commission,» October 9, 1958. (Mimeographed.)

72

THE POST - WAR RECESSIONS TABLE 21

PUERTO RICAN AND MAINLAND BANK RATES 1951

Bank rates in 19 U. S. cities N. Y. C. rates Puerto Rico

6.2

1952

1953

1957

1958

6.6

3.69 3.61 3.7 42 4.6 3.46 3.36 3.47 4.03 4.46 6.2 6.4 6.5 6.0 5.9

4.3 4.06 6.1

1954

1955

1956

SOURCE: Federal Reserve Bulletin, »Report of examinations of native and foreign commercial banks supervised by the Secretary of the Treasury of the Commonwealth of Puerto Rico.» (Courtesy of Juan Labadie Eurite, Assistant Secretary of the Treasury in charge of Financial Affairs.)

_since the beginning of the economic development program. There are a number of reasons for this. First of all, banks in Puerto Rico, especially local banks, have always been very liquid, i. e., they have always had a substantial amount of excess reserves, whose importance diminished with increased Mainland banking ties and with the introduction of FDIC. Secondly, liquidity was further increased by a strong rise in deposits "resulting from the inflow of capital due to the development of the island, and the increased use of the banking system by the population at large, whose incomes were rising and who were beginning to appreciate the function of demand and savings deposits. Thirdly, economic development has encouraged the banks and made it easier for them (especially local banks) to increase the use of credit in the economy — i. e., with expanding markets locally for all types of businesses, they found more and more credit-worthy firms which needed and had to be taught the use of loans from the banks for running business with increased turnover. In this sense, then, economic development spurred on by Mainland capital has spread some of its secondary growth effects through the banking system. The chief question which cannot be answered at this time is: what will be the effects of Mainland fluctuations once the island banking system has matured, i. e., once much of the excess reserves have vanished and once the system has fully expanded? Then the possibility might exist that a Mainland recession, which cuts off the flow of capital to the island and which curtails exports, would bring along tight conditions in the banking system. Loans might have to be curtailed, interest rates might increase, and a secondary contractionary wave would thus hit

THE POST - WAR RECESSIONS

73

the island. This, of course, discounts the possibility of countercyclical actions by local authorities — e.g., that reserve requirements, which are set in the island, are not lowered, that the Government Development Bank takes no remedial actions. Although in this sense a Mainland recession might bring on tight conditions in the island, Mainland credit might be loosened as a countercyclical measure. Thus, higher interest rates on the island relative to the Mainland might attract some capital, but, even more important, with lower Mainland rates it might be more attractive for the island government to sell bonds on the Mainland, the return of which would be used for capital projects which might be stepped up during a recession.

PART III

THE SENSITIVITY OF PUERTO RICAN EXPORTS AND INVESTMENT TO U. S. MAINLAND FLUCTUATIONS The exposure of the Puerto Rican economy to the Mainland cycle will to a considerable extent depend on the degree of fluctuations of those variables which directly affect the island's exports and investments. The purpose of this brief chapter is to give the results of some rough measures which were made on the income elasticity of demand of the Mainland for the type of products which Puerto Rico exports and on the Mainland determinants of investment. Income Elasticity of Demand for the Type of Products Exported by Puerto Rico It should be remembered that the changing export dependence of the island economy has to be viewed from two not mutually exclusive points of view: the greater export/gross product ratio and the change in the commodity structure of exports. With the rise in the former ratio (from 32 % before the war, to 44 % in 1947-48, to 53 % in 1956-57) the commodity structure has changed to such an extent that the category Vegetable Food Products and Beverages dropped from 71 % of total exports in 1947-48 to 33 % in 1956-57, while textiles increased from 18 % to 27 %, machinery and vehicles from almost nothing to 8 %, etc. It was also shown previously that the growing industries rely to a much greater extent on the -foreign market than the old agricultural and needle works industries. With this in mind, we tried to discover at first 75

76

THE POST-WAR RECESSIONS

to what extent the key Puerto Rican export items are subject ta= fluctuations on the U. S. market over the cycle. Our procedure was to take the main aggregative categories of Puerto Rican exports and see how such goods fare over the cycle in the U. S. Since Puerto Rican goods do not have generally distinguishing features, I assumed that they form part of the general supply of each industry to which they cater on the market. I therefore used total U.. S. production indexes of these goods and observed their fluctuations. I assumed that the observed sensitivities ought to give us an approximation of the demand. elasticities for Puerto Rican exports. The results are, of course, subject to severe qualifications which I shall corn ment on again below. Since the business cycle is a short-run phenomenon, the rela. tionship of economic variables during the cycle ought to be observed on a monthly or quarterly rather than on a yearly basis. I have decided to use quarterly observations due to difficulties in assembling and computing all the necessary data. The key independent variables in our analysis are U. S. GNP, Disposable Income, Investment, and general industrial production. For the dependent variables, which are supposed to indicate the demand in the general U. S. market for the type of products Puerto Rico exports, I wanted to use quarterly consumption and manufacturers' sales figures. This attempt was discontinued due to the immense amount of work involverd and the dubious results which we would get with restricted facilities. Consumption figures are available only on a limited scale, and manufacturers' sales are available, but uncorrected for price changes. The alternative was to choose U. S. production indexes which are published on a monthly basis and which do not present price adjustment problems. I was also able to obtain figures dating back to 1947 without running into the problem of comparison arising from the change of base years in the early fifties. This was possible because the Federal Reserve Board in its December, 1953, Bulletin conveniently published monthly industrial production indexes dating back to 1947 in terms of the new 1947-49 base, taking into account all the changes of relative weights. There can be no doubt that production indexes are crude measures of demand, but they should give a general indication of the .reaction of individual items to changes in natiorlal aggregates — investment or income — or general industrial production. On the downswing they might exaggerate reactions, especially with consumer goods of a durable variety, since some

THE POST - WAR RECESSIONS

77

.in

of the sales will come from inventories, while on the upswing an exaggerated reaction might be revealed due to the rise consumption and the build-up of inventories. But it could justifiably be claimed that what I am really interested in is how the cycle will affect the factors employed in producing the product, not the total amount sold; in that sense production indexes might be even more meaningful. I began by making a list. of Puerto Rico's principal export categories. Eleven product groups were included: Stone, Glass, and Clay Apparel Textile Mill Products Food and Beverages Machinery Fabricated Metal Products

Petroleum and Coal Products. Chemicals Paper and Allied Products Tobacco Leather Products

From monthly data I then calculated quarterly indexes of U. S. production, adjusted for seasonable variation, for each of the products, and compared their fluctuations in relation to GNP, Disposable Income, Investment, and total industrial production. This has been reproduced in Table 22. I went a step farther in establishing the sensitivity of the products to the cycle. Taking the National Bureau of Economic Research's dates of peaks and troughs in the cycles since the war, I fitted the relevant quarterly production data to approximate the latter. By «approximate» I mean that come data were used with a lead or lag of a quarter, since not all indexes reached their peaks or troughs at exactly the same time. The purpose of this was to observe the degree of fluctuation from peak to trough and trough to peak for each product. I then averaged the upward and downward movement for each product and computed for each product the ratio of its upswing or downswing to the up or downward movement of general industrial production. (See Table 24,.) A ratio of over 10.0 'would indicate an upward or downward movement greater than the general trend of production (or economic activity), while a ratió smaller than 100 would indicate a lesser sensitivity to the cycle than general production. What are the results? The general impression is that Puerto Rican exports are increasingly concentrated on products which experience sharper fluctuations over the cycle. Food products, beverages, and tobacco, which are in the declining category of the commodity export

`

THE POST-WAR RECESSIONS

structure, fluctuate to a much smaller extent than industrial production. This is not surprising since it is usually claimed that foodstuffs are relatively income inelastic. Textile products which represent a growing proportion of Puerto Rican exports are considerably more volatile over the cycle, especially in the downward direction, where they fall more sharply than general industrialproduction, while in the upward movement they rise less than the latter. It should be noted, however, that the apparel component, which is extremely important in the general textile production of Puerto Rico, is less sensitive in both the upward and the downward direction than general industrial production. Machinery and Chemical products which represent a growing item on Puerto Rico's export list reveal a much greater sensitivity in the upward direction than general industrial production. This reflects a considerable growth trend in those industries. In the downward direction Machinery is only slightly greater than our reference series, while Chemicals are more than fifty per cent smaller. Pretroleum products and Leather products also exhibit less sensitivity to the general cycle, while Stone, Glass, and Clay products fluctuate only slightly more than the reference series. It can be concluded that the commodity export structure of Puerto Rico has become increasingly dominated by products which fluctuate to a greater extent over the cycle than agricultural products, which used to be dominant. But on the downswing most of the «growing» products are either equally or less sensitive to the reference cycle, with the exception of metals and textiles (remembering, however, that a great component of the latter are apparel products which are less sensitive than the textiles in general). On the upswing, the structure is less sensitive than the reference magnitudes for the most important export components, but in the fast-growing fields of machinery and chemicals the reaction on the upswing is greater than the reference series. (See Tables 22, 23, 24.) New Investments in Puerto Rico as a Function of U. S. Mainland Variables Just as one looks for some definite relationships between Puerto Rican exports and the general behavior of the type of goods exported on the Mainland, one might also hope to find significant systematic relationships between investment behavior in Puerto Rico (which consist to a large extent of Mainland cap-

.

78

TnsLE 23 FLUCTUATIONS DURING THE CYCLE

NBER Cycle

Us Ind. Prod.

Peak 104 (1948-3Q) -

9

Fabric. Metals

Machin.

Text. Mill prds.

Apparel

104 105 110 104 -14 -27 - 9

95 95 Trough 41 44 (1949-3Q)

100

83

91

28

72

4

Stone, Glass & Clay

Leather prds.

Tobacco

102 107 112 -20 -12 95

16

98

92 40

4

12

Paper & prds. Chemicals

104 104 -13 91

16

7

97 43

Petroleum & Coal Food & prds. Bey. Nfs.

104 -

100 100

98 53

0

6

8

34

116 135 104 134 150, 132 163 114 108 Peak 136 111 139 —14 —5 — 5 -20 —14 — 7 —12 —9 —3 -25 (1953-2Q) -13 -17 Trough 123 122 138 (1954-3Q) 22 18 34

102

91 18

128 11

92 30

12

158 113 105 112 140 172 109 145 Peak —11 —22 — 8 —34 —18 (1957-3Q) —17 —19 Trough 128 (1958-2Q)

121

138

91

102

136

97

129

100 13

116

4

145 31

123 40

105 18

8

160 185 141 113 — 8 —7 —12 152

178

129

1

114

Souxcrs: Dates of peaks and troughs are based on National Bureau reference cycle sese Moore, Geoffrey H., Measuring Recession, Ocasional Paper 61, NBER. Data used from previous lat~le, w iy

80

THE POST - WAR RECESSIONS TABLE 24 MEASUREMENT OF SENSITIVITIES

Industrial Prod. Machinery Fabricated Metals Textile Products Apparel Stone Glass Clay Leather Products Tobacco Paper and Allied Prods. Chemicals Petroleum & Coal Prods. Food and Bev.

Average* Degree of Downturn

Ave.* Rise

—13

32

—24 —15 —22 —10 —14 —13 —4 —8 —6 — 9 0

53 31 23 13 35 12 14 37 46 26 8

Ratio of Individ. Fall to Ind prod. Fall

Ratio of Iudivid. Rise to Rise in Ind. prod.

104 115 169 77 108 100 30 61 46 69

166 97 72 41 109 37 44 115 143 81 25

SOURCE: Computed from previous table. * Averages computed from previous table.

ital) and Mainland investment variables. The discovery of any significant relation of the latter type would help the policy maker in pinpointing the sectors on the Mainland which need careful attention in the formulation and direction of promotional efforts. A test of significánce would also establish whether or not investment in Puerto Rico is subject to the cycle at all. My efforts to discover relationships were concentrated on a number of simple regression analyses. As the dependent variable I chose the quarterly openings of new EDA-promoted plants. There are no quarterly investment figures available for Puerto Rico. As independent variables I chose in turn U. S. industrial production index, investment in new plant and equipment, and profits of corporations. Since the opening of new plants lags behind the original decision to invest, I regressed the dependent variable on the independent variables with lags of one, two, or three quarters. (See Table 25.) Unfortunately no result of any significance was found. After having performed a number of regressions, I chose the relationship with the highest correlation coefficient, new plant and

TABI B 22 P.R. EXPORTS

QUARTERLY US STATISTICS (1947139 = 100)

(1955 = 100)

Append

Leather producto

Chemicele

Stone, Glass & Cloy

Peper& Ailed Producto

Petrol, & Coal Pod.

food & Nev.

105 99 97 104

96 93 98 102

104 105 104 112

95 95 95 100

97 98 97 99

100 100 98 100

93 95 98 101

103 98 100 102

98 102 101 102

108 110 103 99

104 103 102 99

104 100 99 97

102 102 103 104

100 105 107 106

101 103 104 102

102 104 103 106

99 100 99 99

40 71 71 41

98 91 91 91

101 101 99 98

94 83 91 106

101 100 103 100

95 94 95 92

101 97 99 105

100 95 96 98

96 91 99 106

102 100 98 98

99 100 100 101

48 73 74 77

126 124 121 118

98 108 122 128

100 100 103 101

106 107 113 118

103 107 112 111

97 98 105 105

110 116 126 132

103 115 124 130

109 114 124 129

100 107 114 118

99 102 106 105

49 70 93 51

122 122 118 119

126 124 121 118

130 130 126 133

108 107 103 108

117 114 100 98

113 108 100 101

104 95 89 87

135 138 136 134

132 135 131 125

132 135 121 115

123 122 121 124

108 104 104 103

53 123 91 53

57 55 50 51

7 11 10 7

17 16 25 49

45 92 61 55

126 115 123 135

121 119 122 132

118 115 118 132

138 141 147 159

109 109 110 111

100 99 105 109

105 107 109 112

94 98 100 104

135 135 137 141

123 124 125. 129

118 117 120 127

124 112 126 129

104 104 105 106

63 95 88 85

59 57 66 81

4 3 4 11

89 117 127 100

59 73 78 84

122 124 123 123

131 136 131 113

134 136 135 129

136 139 139 130

163 162 163 152

114 106 1114 108

105 111 104 94

111 116 109 103

104 103 97 92

143 150 149 145

132 134 135 131

129 134 134 130

128 131 132 129

107 108 108 106

98 112 122 79

75 72 82 90

15 25 26 48

(117) (100) ( 96) (159)

92 91

128 128 129 132

124 125 125 128

118 123 119 133

124 124 123 128

123 121 122 125

140 138 144 146

100 106 100 101

91 94 95 101

106 107 102 106

94 94 96 97

145 147 149 152

128 129 132 134

129 135 135 138

124 124 123 127

106 107 105 107

83 103 82 109

91 87 99 96

63 76 76 90

92 158 142 136

87 95 89 82

IQ 2O 3Q 4Q

136 139 142 144

129 193 135 148

142 153 156 166

133 138 140 143

127 133 137 139

146 153 159 164

107 107 190 195

103 107 107 109

109 113 113 116

105 105 104 104

158 167 170 174

139 149 154 154

143 154 155 157

133 135 135 139

107 109 108 112

52 137 129 81

89 91 109. 111

76 92 99 132

127 100 100 72

74 110 113 103

10 2O 3O 4O

146 146 146 149

138 140 139 142

164 165 163 170

142 142 141 146

134 133 134 140

163 169 174 175

1l7 16 1d6 1d7

109 103 102 103

111 110 112 111

108 103 102 102

176 177 176 178

155 160 158 157

158 160 160 159

143 :140 138 143

112 112 113 114

94 .106 97 114

103 104 127 129

115 144 157 175

73 86 130 151

103 107 114 124

1Q 2O 3O 4O

149 152 153 151

142 142 146 145

155 165 163 151

145 144 145 138

137 138 140 136

172 168 172 159

113 110 112 108

101 99 99 94

110 112 113 106

105 105 104 102

183 183 185 183

155 157 158 149

157 158 360 158

143 141 141 138

113 113 113 112

60 104 111 62

120 122 127 134

164 167 183 192

115 93 114 193

96 116 129 111

1Q 2Q

145 145

143 143

120 119

130 129

125 121

145 138

112 116

91 93

102 107

99 97

178 179

136 140

152 154

129 129

114 114

49

112

130

141

96

loduslrlal Producaon Mechiuery Tobacco Textiles

GNP

Dlep• Joe.

Íuv.

Gen.

Pal,. Melelo

1Q 2O 3Q 4O

96 97 96 98

98 95 97 97

86 88 81 101

99 99 99 102

101 103 103 104

99 104 102 105

101 95 97 103

10 2Q 3Q 4Q

98 100 100 103

98 101 101 102

107 116 121 125

103 104 104 104

105 105 104 104

105 105 105 103

IQ 2O 3O 4O

102 102 103 103

102 102 102 104

105 91 93 89

99 95 96 96

95 90 95 91

10 2O 3O 4O

108 113 115 117

112 111 112 113

116 149 152 172

100 109 118 121

10 2Q 3O 4O

117 118 120 120

110 112 113 114

155 160 142 127

1O 2Q 3Q 40

121 121 122 127

113 114 116 119

10 2O 3Q 4O

130 132 130 128

1Q 2Q 3O 4O

Vra. Foods Beverages

Textiles

Mechlnery

Chem.

Tal.

1947

1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

1958

Souncts9: US Department of Commerce, Business Statistics (A supplement to the Survey of Current Business) 1955, 1957; Survey of Current Business; Federal Reserve Bulletin, December 1953; Puerid IUco Planniog Board, Bureau of Economics and Statistics, Current Business Statistics. US Department of Commerce, Bureau of the Census — US Trade in Merchandise end Gold and Silver With US Territories and Possessions. Above data were computed from monthly figures adjusted for seasonable varietioos; for GNP, dt other national accounling figures,) we made price adjustments (also for PR exports).

7Ba

( ~

81

THE POST - WAR RECESSIONS TABLE 25

EFFECTS OF MAÍNLAND VARIABLES ON PUERTO RICAN INVESTMENTS * (Correlation coefficients based on quarterly series from 1947 to 1959) Mainland Independent

Correlation Coefficient

Variables 1. U. S. Industrial Production Index a) No Lag b) Two Quarter Lag ** c) Three Quarter Lag d) Four Quarter Lag

.780 .781 .756 .756

2. New Plant and Equipment a) No Lag b) One Quarter Lag c) Two Quarter Lag d) Three Quarter Lag e) Four Quarter Lag

.707 .745 .759 .852 .718

3. Profits a) No Lag b) Two Quarter Lag c) Three Quarter Lag

337 .348 .427

4. Correlation when trend removed with New Plant and Equipment Three Quarter Lag

.009

* Puerto Rican Investments are the dependent variable used for all these simple regression analyses. Since total quarterly Puerto Rican investment data are not available, we used new plant openings as our dependent variable. ** A lag here means that the dependent variable lags one more quarter behind the independent variable. SOURCE: Computed from data supplied by the Puerto Rican Economic Development Administration, Survey of Current Business, Department of Commerce, and the latter's hi-yearly Business Statistics.

7 ~

82

THE POST- WAR RECESSIONS

equipment with a. three quarters lag, and eliminated the trend in the time series on both sides of the equation, i. e., I regressed the dependent and independent variables against time, then projected values for each based on the obtained coefficients; then I obtained the differences between the projected and actual values of the independent and dependent variables, and finally I regressed the latter on the former. As can be seen in Table 25, there exists no relationship between the Mainland variable and Puerto Rican investment which can be said to be of a cyclical nature. It is still of interest, however, to glance at the «insignificant» correlation coefficients. The strongest relationship obtained was between new plant openings and Mainland new plant and equipment with a three quarters lag. The assumption of Puerto Rican authorities that there exists a lag between Mainland and Puerto Rican economic fluctuations would therefore seem to be justified from the point of view of investment behavior. Furthermore, the investment program of Puerto Rico is influenced primarily by general investment decisions on the Mainland. These conclusions are very speculative in character, however, because when the trend is removed, no relationships remain.

CHAPTER IV THE BEHAVIOR OF PROMOTED FIRMS OVER THE CYCLE It has already been mentioned in the introductory chapter that the degree to which the cycle is passed on from the U. S. Mainland to Puerto Rico depends -not only on the income elasticity of demand for the products the island exports, the import content of these exports, and .the flow of investments, but also on the type of enterprises the Fomento program has been attracting. I claimed that these enterprises can be divided into three broad categories: new local enterprises;, firms which have been set up in, or have resettled in the island, in their entirety; and firms or plants which are subsidiaries of larger Mainland enterprises. Excluding the firms which cater principally to the .local market, it can be claimed, that the rate of production of the first two types of firms will depend on general U. S. market conditions. Some additional possibilities are offered by the third type of firms, the subsidiaries. They could be divided into two subcategories.: those firms representing horizontal expansions of Mainland firms and those representing vertical extensions. The former are simply the extension of capacity of existing corporations for producing,, a finished product. The second type ar: plants set up by Mainland corporations either to produce parts of the product which is assembled on the Mainland, or to perform an operation on a product shipped to the island for that specific purpose and then shipped back for further processing. It was further maintained that the horizontal plant's opera. tions over the Mainland cycle are difficult to predict because they depend on the policies of each corporation. If a downswing in 83

84

THE BEHAVIOR OF PROMOTED FIRMS

economic activity reduces a firm's sales, and hence forces a decrease in production, how will this come about? Will all plants of the corporation curtail their production by the same proportion, or will, certain plants be shut completely while others continue to produce at full capacity? If the latter is the case, which plant would be closed, the Puerto Rican or a Mainland plant? Similar questions could be asked about plants forming parts in the vertical link of the production process. In that case, considerations similar to the ones mentioned above become relevant. If the Puerto Rican plant is the only vertical link in the production process, its operation might be more stable because it would follow only general fluctuations in the output of the firm (unless there is a drastic change in inventory policy of certain factors of production). Although it is obviously an impossible task to quantify and test all of these considerations, I was able to get some interesting insights through the cooperation of the Economic Research Staff of Puerto Rico's Economic Development Administration. Due to the confidential nature of much of the information in the files of EDA, I was restricted in the type of information available. Data was gathered in the files of EDA for all promoted plants. The chief dependent variable was the number of man hours per month, of which I got the highest month of the peak. year and the lowest for the trough year. Since I only covered the 1953-54 and 1957-58 recessions, the highest monthly man hours for 1953 and 1957 and the lowest for 1954 and 1958 were obtained for each plant. (The fact that this produces distortions due to seasonal considerations and thus could show declines much sharper than those caused by cyclical consideration will be taken into account below when I shall evaluate the data.) This data was collected on a specially prepared sheet (see Appendix to this chapter). On that sheet each firm was classifield according to the SIC code (in my analysis I concentrated on the first two-digit classification), according to type of ownership, and according to its principal market outlet (Puerto Rico or U. S.). Some remarks are in order for the second type of classification-type of own erhip. There are seven different types of ownerships: 1) Local Autonomous: a firm promoted by EDA which is owned by Puerto Rican residents. (LA)

THE BEHAVIOR OF PROMOTED FIRMS

85

2) Local Contractor: a firm promoted by EDA, owned by Puerto Rican residents, which operates principally on a contract basis, either with local or with Mainland firms. (LC) 3) Mainland Autonomous: a firm which is owned by Mainland interests, but which is not a subsidiary of a Mainland firm. (M Aut) 4) Mainland Affiliated: a subsidiary of a Mainland corporation. (MAH) 5) Mainland Affiliated Contractor: a firm connected to a Mainland concern but operating only on a contract basis. 6) Mainland Independent Contractor: Mainland-owned firms operating under contract with many firms .(in our tabulations this category was merged with 5 because of the very small number available). (MAHC) 7) Foreign: a firm owned by foreign interests. (F) These are the only breakdowns by type of ownership available. It was virtually impossible to obtain a breakdown along the horizontal-vertical line. Although a description of the product each firm produces. is available, it is mostly impossible to tell whether the output of a Mainland affiliated firm will be further worked in a Mainland plant of the same corporation. For example, an affiliated plant might produce tubes for television sets, but it is usually quite unclear whether the Mainland concern will assemble these tubes into television sets or whether it simply sells tubes on the market; or, it is difficult to tell whether the output of a subsidiary producing «rubber fittings for electronics products» will be sold directly by the Mainland parent or will be used for further processing. One might still claim that there is a difference between plants producing a product which is ready for the final consumer market and those producing parts of a product which has to be assembled elsewhere, even if the assembling takes place in another firm. In other words, there might be a considerable difference between a «final» market and a «factor» market — i. e., a market for parts of a product to be assembled. Fluctuations in the sales of the former might be less pronounced than in the sales of «parts», since the latter could easily be affected by important secondary changes in inventory policies for these «parts.» Such ,phenomena

86

THE BEHAVIOR OF PROMOTED FIRMS

will have to be intuitively, taken into account when examining the tabulations below. Table 26 should make it clear why the lack of horizontalvertical data will be no great hindrance. It contains a tabulation from the December 1958 Directory of EDA-promoted plants. From the short description of the product of the firm and from 'the name of the parent firm, I tried to classify firms .according to a horizontal-vertical relationship. Only in the categories of Electric Machinery and Textile Mill Products were there a substantial number of firms producing a product, which, according to the type of parent firms they had, could be assumed to form part of a vertical chain of production. It would seem therefore that our dealing with local, autonomous and affiliated categories will not greatly harm our analysis. For tabulation purposes the absolute number of monthly manhour decline was converted into percentage declines. The tabulations were undertaken according to the two different periods examined, type of firm, type of market, type of product, and size of firm (size measured by maximum man hours worked). A conglomerate table with most of these variables is reproduced in the Appendix to this chapter. Let us now turn to a step-by-step analysis of the tabulation results. Distribution of the Type of Ownership It is quite evident from Table 27 that Mainland Affiliated firms are dominant in all the industry groups which have spearheaded the island's industrialization. As might be expected from descriptions in previous chapters, Apparel and Textile Mill products dominate, followed by Miscellaneous products (which include sporting goods, costume jewelry, plastic products, etc.), leather products, and metal products and machinery, both of which have been looming larger in the second period, reflecting an effort by the development authorities to strive toward greater diversity. A relatively large share of Apparel products are also. produced by Mainland Affiliated Contractors, while firms in the Miscellaneous category, producing light manufactured items, have the Largest contingent among the Mainland independent producers. Locally owned firms operate primarily autonomously and are concentrated on food products, apparel, stone, glass, and clay,, and the miscellaneous category.

87

THE BEHAVIOR OF PROMOTED FIRMS TABLE 26

TABULATION OF PLANT CHARACTERISTICS Mainland Owned Sole Plant * Horizontal Vertical Sub. Subsid.

Food & Kindred Products (20) ** 3 Tobacco Manufactures (21) Textile Mill Products (22) 4 Apparel & Related Products (22) 11 Lumber & Wood Products (24) Furniture & Fixtures (25) 2 Paper & Allied Products (26) 1 Printing & Publishing (27) 1 Chemicals & Allied Products (28) 3 Products of Petroleum & Coal (29) 1 Rubber Products (30) 8 Leather & Leather Products (31) 2 Stone, Clay & Glass Products (32) 4 Primary Metal industries (33) 1 Fabricated Metal Products (34) 9 3 Machinery (except Electric) (35) 5 Electric Machinery (36) 15 Miscellaneous (38-9) ***

14 3 31 108 4 3 10 2 11 2 11 27 6 1 21 12 32 51

8

1 2 2 1 1 1 2 3 3 12 5

Local

Foreign

16

1

8 31 1 13 1 4 7

1 20 3 10 2 4 7

2

1 1 1 1 7

* Mainland owned plants with no Mainland connection.

** sic classification Number. *** Miscellaneous— principally Sporting Goods, costume Jewelry, Plastic Products, Lapidary Work, Artificial Flowers, etc. Som ca: Tabulated from Directory of Fomento Promoted and Assisted Manufacturing Plants, December 31, 1958, Economic Development Administration.

It is interesting to observe developments between the twoperiods. The principal first impression is of a relatively faster growth in Locally Autnonomous and Mainland-owned Autonomous firms, than in other types of firms. This partially reflects efforts made by EDA to attract .more local capital and also a secondary wave of local savings into the industrializing sectors, spurred on by increased income and funds due to the momentum of indus₹rialization and the more suitable climate for island capital once a large number of more experienced enterprises have paved the way. The strongest growth of local firms is found in the food, furniture, stone, clay, and glass industries, in which native firms have a natural advantage and which are partially indúced by

TABLE 27 TYPE OF OWNERSHIP DISTRIBUTION BY INDUSTRY GROUPS * Industry Group

20 21 22 23 24 25 26 27 28 29 30 31. 32 33 34 35 36 38-9 Totals

LA

LC

MAut

4 3 2

1

2 1

1 1

3

1

1 2 3

1

1 3 8

20

2

16

1953-I

MALI

5 1 30 55 1 1 5 1 5 2 8 19 3 1 10 10 25 38 220

MAIÍC

For

LA

1957-8 MAut MAIL

LC

10 2 20

1

4 9 2 9

2

1

10 1 5 2 5 7

2

3 1 1 3 1 6 2 5 1 3 3 5 11

69

12

55

1 2

1

.

2 1 2 1 1 4

1

29

5

* Industry group numbers explained in Table 26.

SopRcs: Tabulated from Conglomerate Table in appendix to this chapter.

3 5

9

4 3 34 78 1 2' 5 2 5 3 13 23 2 4 21 6 33 43 282

MALIc

Por

5 34

2

1

1 l

1 6

1 2

48

7

THE BEHAVIOR OF PROMOTED FIRMS

89

growth of population and income; but there were also marked increases in apparel, where contractors made their appearance, metal products, electrical machinery and the miscellaneous industry group. The growth of Mainland Autonomous firms was. spread fairly evenly over the spectrum of EDA industries, espé-cially notable being the rubber products, apparel, and chemicals industries. The growth of contractors from the Mainland chiefly continued in the apparel, textiles, and miscellaneous products groups. It is obvious that in absolute terms the growth of Mainland affiliates, i .e., the subsidiaries, was the largest of all types. Especially noteworthy here is the large relative growth of the number of firms in metal products and electrical machinery. To sum up, between the two periods under examination there has been an important rise in firms other than the MainlandP Affiliate type, thus bringing a little more balance, though the latter is still supremely domimant; and there has also been a marked growth in firms outside the apparel and textile industries, thus leading to a little more diversity in the growing Puerto Rican industrial structure. We shall return later to assess these developments in the light of future cycle sensitivity after we have examined the evidence on the behavioral patterns of these firms as seen through various types of classifications. Market Orientation Another preliminary matter for the better understanding of the data is the market orientation of firms according to type of ownership and industry group. The market orientation according to type of ownership is reproduced in Table 28, section a. As might be expected, there are a larger number of Local Autonomous plants oriented to the local than to the Mainland market, (1) especially in the second period under analysis. This reflects, a secondary reaction to industrialization — with higher incomes there are an increasing number of domestic market possibilities which are best catered to by local interests. Most Local Contractors are oriented toward the U. S. Mainland. There has been an interesting change in the distribution of Mainland autonomous firms. In the first period, Mainland-oriented plants predominated,. (1) When a firm is oriented toward one market, this means that not necessarily all of its products are sold in that particular market, but rather that over 50% of its sales are made in that .market.

X90

THE BEHAVIOR OF PROMOTED FIRMS

-while in the second period Puerto Rico-oriented plants • outnumbered the others; this also reflects the attraction of a growing local market to Mainland-owned firms. Even in Mainland Affiliated firms one finds a small relative shift toward :locally-oriented plants. Mainland Affiliated Contractors still cater exclusively to. Mainland customers. As one would expect, the orientation of most firms in the exports industries, like Textile Products, Apparel, Electrical Machinery, and Miscellaneous items, is toward Mainland markets. (See Table 28, section b.) Local orientation is primarily in industries which are affected by general industrial development in the island, like Stone, Clay, Glass, or Furniture. Much of the growth in the above-mentioned Puerto Rico-oriented firms took place in the latter industries, and also in the Food Products, Fabricated Metal Products, and Machine industries. .Size of Plants A final matter to examine before observing the behavior of -the firms in the two recessions is the size of the plants by type of ownership and industry groups. Size was measured in each period by the maximum number of man-hours per month worked in the boom year. (See Table 29.) When measured by -type of ownership, it is evident that Local Autonomous plants :are rather small, the bulk being in the less than 1,000 man-hours category; this is true even in the second period, growth having taken place principally in the smaller type of firms. Local contractors are evenly divided in the second period between very small and medium size firms. Mainland Autonomous firms also consist chiefly of small firms, mostly less than 3,000 man-hours, the bulk in the less than 1000-man-hour category. Although the absolute numerical majority of Mainland Affiliated firms was -1n the smallest types, the percentage has decreased, and most of the relative gain was in firms with 2000 to 5000 man-hours. _Among the largest firms, the Mainland Affiliated category had relatively the largest percentage, and these types of firms kept their share over the two periods. Mainland Affiliated contractors -were predominantly in the medium-to-large range, and this was strengthened over the two periods, especially in the 4000-to-5000 =man-hour category. Finally, foreign firms were among the smaller ones.

THE BEHAVIOR OF PROMOTED FIRMS

91

TABLE 28 a) MARKET ORIENTATION ACCORDING TO OWNERSHIP TYPE • Type of Ownership

Local Autonomous Local Contractor Mainland Autonomous Mainland Affiliated Mainland Affiliated Contractor Foreign

1953-4 Puerto Rico U. S.

11 4 26

9 2 12 190 28

1957-8 Puerto Rico

56 1 33 67

U. S.

13 11 22 215 48

7

4

b) MARKET ORIENTATION ACCORDING TO' INDUSTRY GROUP

Industry Group * 202122232425262728293031323334353638-9-

2 6 5 2 4 2 1 6 5 2 6 7

7 1 31 75 1 1 1 4 2 7 20 3 1 7 9 26 47

10 1 7 18 2 - 14 5 2 4 2 6 2 17 4 12 6 12 10

* Industry group numbers explained in Table 26. Souxcs: Tabulated from Conglomerate Table in Appendix to this chapter.

6 2 39 119 1 1 1 3 6. 2 15 24 2 2 17 5 33 61

92

THE BEHAVIOR OF PROMOTED FIRMS TABLE 29 a) PLANT SIZE BY TYPE OF OWNERSHIP (by monthly man hours)

Type of Ownership

1953-4 Local Aut. Local Contr. Mainland Aut. Mainland Aff. Mainland Aff. Cont. Foreign 1957-8 Local Aut. Local Contr. Mainland Aut. Mainland Aff. Mainland Aff. Cont. Foreign

0-1000

1-2000

2-3000

3-4000

10

1

2

2 1

6 55 6 3

6 27 3 1

1 20 7 1

45 3 26 65 5 3

8 2 9 42 3 1

4 2 7 43 6 1

4-5000

5-10,000

10,000

2

14 2

11 4

1 1 1 35 7

3

4 1

2 26 5

4 2 3 41 7

21 9

18 1

2 1 30 4

b) PLANT SIZE BY INDUSTRY GROUP (1953-4/1957-8)

3/8

2/3

2/1

6/7 6/3 2/11 12/25 6/18 10/17 1/1 0/1 0/8 0/1 1/4 1/1 0/1 1/1 1/5 0/1 6/6 1/3 1/2 4/6 1/21 0/2_ 1/2 3/6 2/1 1/3 2/6 1/3 1/3 0/1, 4/16 1/5 2/5 8/10 1/3 14/18 6/5 3/4 15/24 11/13 6/9

0/2

0/1

3/4 8/12

1/6 6/7 8/15 17/25

0/3 1/1

0/1

0/1 1/2 2/5 9/13

2/1 r

Industry Group * 202122232425262728293031323334353638-9

1/3 1/2 1/1 0/1 0/1

2/1 2/2 0/3

0/1 0/1 1/2 7/6 1/0 0/1 1/0

2/3 1/6

0/3 0/3

2/5 7/9

* Industry group numbers explained in Table 26. Souxcs: Tabulated from data gathered by EDA especially for this project.

0/1 0/1 2/2 3/3

2/4 2/4

THE BEHAVIOR OF PROMOTED FIRMS

93

When examined by industry groups, we find a predominance of small plants in the Food, Furniture, Printing, Chemicals, Rubber Products and Machinery industries, although in the second period some large firms make their appearance in the Food industry. In textile products and apparel, we find the smallest firms diminishing over the two periods from almost 50% to 30% or less, the largest declining from a little over 30 % to a little less than 30 %, while medium firms gain some ground from 30 % or less to 40% or over. In the Fabricated Metal Products and Machin ery (except electrical) sectors, small firms have predominated and continue to do so. Although over half of the firms in the Electrical Machinery industry are small firms, there has been a definite downward trend in small firms (from about 69 % to 55%), while there has been an increase in medium .and large firms; this is also the case in the Miscellaneous sector. Behavior of Firms During the Recessions On the basis of the above background information we are now ready to examine the behavior of the firms examined in the 1953-54 and 1957-58 recessions. The procedure will be to examine percentage decline in monthly man-hours between the highest month in the peak and the lowest in the trough years. We shall examine this variable from four different viewpoints — type of industry, form of ownership, market orientation, and size, all of which will be .integrated in a final evaluative section. 1) Variation in Man Hours per Month in Different Industries Our discussion in this section refers to Table 30. We shall assume that a decline of 0-20 % constitutes no cyclically-induced decline. Many firms in this category actually did not decline at all, and for some who actually experienced a decline, seasonal or random factors were responsible, since we took the highest month of the peak and the lowest month of the trough year. A decline in monthly man-hours worked of 20 % to 40 % is rather mild, 40% to 60% we shall call substantial, and 60% and over would be severe. We also included the number of plants which discontinued during this period. Here a word of caution is needed. Discontinuance of operation is not necessarily induced by recessionary forces, even if they occur during a Mainland recession period; it might be due to such factors as higher labor cost than expected, lack of managerial talent, etc.

94

THE BEHAVIOR OF PROMOTED FIRMS

An. examination of the 1953-54 period shows a large. proportion of plants which experienced decreases in monthly man-hours of over 60 % in the Textile Mill Products (50 %), and Apparel Industries (42 %). Declines in the increasingly important Electric Machinery Industry were much milder, with only 27 % of the firms decreasing man-hours by more than 60 %. This probably reflects the very strong growth element in that industry, with many firms just getting started in that period. In the Miscellaneous sector, which contains many light manufacturing industries, we find one of the largest percentage of firms curtailing operations by 80% to 100%. In most other sectors we find a fairly even distribution between light and heavy declines. Proportionately, the highest number of discontinuances of operations .are found in the Textile Mill Products, followed by Leather Products and Apparel. Of greater interest is a comparison between the two periods. In the key Textile Mill Products and Apparel Industry, we observe a sort of polarization. The percentage of firms curtailing manhours drastically (over (60 %) remain the same in Apparel, and increase in Textile Mill Products (32% of the firms in that industry cutting man-hours by more than 80 %), while the ,percentage of firms curtailing production only slightly (20 % 40 %) increases; the decline took place in firms curtailing manhours between 40 % - 60 % and 0-20%. There was a substantial rise in the number of firms curtailing operations over 60 % in the growing Electrical Machinery and Miscellaneous industries; after Textiles, Electrical Machinery had the highest percentage of firms experiencing declines of over 80 %. Although the trend is not pronounced, we also find a bias toward greater declines in the non-Electric Machine, Metal Products, Leather Products, and Furniture industries. Oñly the Food Products industry experienced much milder declines than in the previous recession. 2) Variations in Man-Hours per Month by Form of Ownership The results of this tabulation are contained in Table 31. In the first period, 1953-54, we find that if foreign firms are left out, more Mainland Affiliated firms decreased their monthly man-hours by less than 40 % than any other category, but at the same time the proportion of firms curtailing man-hours by over 60 % is also very large. Most other categories had more firms in the medium range of decline (40 % -60 %) than the

TABLE 30

DECLINE IN MONTHLY MALL HOURS BY INDUSTRY GROUPS (number of firms) * Industry Group *** 0-20

202122232425262728293031323334353638-9-

1 3 • (11) 10 (14)

20-40

Percentage Decline from Peak Month in Peak Year to Trough Month in Trough Year e 1957-8 1953-4 80.100 0-20 20-40 60-80 40-60 60-80 DO** 40-60

2 (22)

1 (11)

1 (11)

5 (56)

6 (21) 9 (13) 1

5 (18) 22 (31)

8 (29) 13 19

6 (21) 16 (23)

9 10

1

2 1 1

2 2

1 4 1 3 . 2 9 (30) 6 (13)

1 3 (8) 15 (12)

1

1 3

2 2 3

1 1 2 3 3

1 2 1

2 6 1

1 3 4 (13) 9 (19)

1 1 9 (30) 13 (27)

3 3 6 (20) 8 (17)

1 1 2 (7) 12 (25)

1 3 1 3 1 2 6

2 2 2 6 2 6 1 5 2 (5) 6 (10)

* Numbers in parenthesis indicate .percentage of total during year in that industry. ** DO— plants having discontinued operations in period under revievr, *** Industry Group numbers are explained in Table 2¢. ~oVxca: See Appendix Conglomerate Tablg.

4 (30)

7 (54)

1 (8)

11 (28), 27 (22)

5 (12) 30 (25) 1 3 2

8 (20) 29 (24) 1 3 1

3 2 1 3 3 2 7 2 10 (25) 16 (26)

1 1 3 8 5 2 6

2 3 3 2 5 5 2 1 3 4 9 (22) 7 (11)

7 (18) 16 (26)

Only used for large industries.

80-100

DO**

1 (8) 2 13 (32) 21 (17)

3

5

6 15 1 1

2 3 4 3

1 3 4

4 4 4 1 12 5 (30) 16 10 (26) ~

96

THE BEHAVIOR OF PROMOTED FIRMS

latter. Mainland Autonomous and Mainland Affiliated Contracting sectors had a very large proportion of firms drastically curtailing their man-hours. It should be noted, however, that the latter contained no firms which discontinued operation in the period, whereas the Mainland Affiliated sector contained 33 such firms. A number of changes took place by the time of the second recession. There was a which substantial increase in Local Autonomous firms seems to have polarized the reaction pattern. The number of firms which did not decrease man-hours at all, 'or decreased them but slightly, increased from 26% to 32%, while the other extreme increased from 37 % to 46 %. The growth in the number of Mainland Autonomous firms also seems to have polarized the pattern; the percentage of firms decreasing man-hours by a relatively small amount had increased by 10 %; there was a small decrease in the percentage of firms cutting hours between 40 % and 60 %, and there was a large increase in the percentage of firms cutting man hours by 80 % to 1.00 %. In addition, there was a noticeable appearance of firms which discontinued operations. The appearance of an increased number of Local Contractors also shifted the weight in favor of large contractions of man-hours. In the large group of Mainland Affiliated firms there was also a slight tendency toward greater contraction, especially with .a decline in the percentage of firms hardly affected .(i. e., 0-20 % contraction), and there was a concentration on the 40 % -80 % contraction group. All this was accompanied by a decrease in the number of firms discontinuing operations. The most notable change occurred in the Mainland Affiliated Contractor group, where the percentage of firms with no or very slight contractions increased frt~m 19 % to 52 %. But the latter group also produced seven firms which discontinued operations, whereas in the previous period there were none. I shall evaluate these findings below, when the entire picture with all the relevant variables has been presented.

.3)

Variations in Man-Hours per Month by Market orientation

Table 32 leaves no doubt of larger contraction of monthly man-hours worked in firms dependent primarily on the Mainland markets than in firms dependent primarily on local markets. More than two-thirds of Mainland-oriented plants in both periods

THE BEHAVIOR OF PROMOTED FIRMS

97

contracted man-hours by more than 40 %, while less than twothirds of the other types of plants followed this pattern. There are no important changes in the second period for Mainlandoriented firms, except for a slight decrease in the extreme types of contractions. In the other firms, however, there is an increase in the extreme types of contractions, but this is accompanied by a relative decline in plants discontinuing operations. 4) Variations in Man-Hours per Month by Size of Plants Due to the difficulties in managing the data, the tabulations in this instance are restricted to three types of decline a relatively small decline is assumed to be a curtailment of from 0-50%, while large declines are in the 50%-90% range, and extremes in the 90 % -100 % range. (See Table 33.) An unusual picture emerges. The proportionately largest number of firms with. the smallest curtailment of man-hours are in the two extreme groups during the first period; in the second period these proportions decline somewhat, with a noticeable increase in extreme declines in the largest types of firms. Over the two periods some interesting shifts took place in the middle group. There was a slight increase in instability in the large group (5,000- 10,000 man-hours), and an even larger increase in the smaller groups (1,000-3,000). At the same time the middle group (3,000- 5,000) experienced a substantial increase in stability, with the 3,000 - 4,000 man-hour firms becoming the group with the highest percentage of decreases of less than 50 %. Evaluation of Tabulations

$

We shall try to tie all these individual breakdowns together from the viewpoint of the ownership type, since we started this chapter by asking what the influence of the latter has been on the behavioral pattern of the firm. The reason for having gone into various directions was that each type of firm's behavior is governed by a number of different variables. Our procedure will be to start by summarizing the behavior of firms in this group, and then to evaluate this behavior in the light of the group's structural characteristics (i. e., type of product, size, market orientation) and the general behavior of firms when viewed through these individual structural angles.

TABLE 31 DECLINE IN MONTHLY MAN HOURS BY FORM OF OWNERSHIP (number of firms) * Porms of Ownership

Loc. Aut. Loc. Contr. M. Aut. M. Aff. M. Aff. C. Foreign

0-20

20-40

1 (5)

4 (21)

1 (7) 37 (19) 1 (4) 1

1 (7) 36 (19) 4 (15) 1

Percentage Decline from Peak Month of Peak Year To Trough Month in Trough Year 1953-4 1957-8 40-60 60-80 80-100 DO 0-20 20-40 40-60 60-80

7 (37) 1 5 (33) 40 (21) 8 (30) 2

3 (16) 7 (47) 34 (18) 7 (26) 1

4 (21) 1 1 (7) 43 (23) 7 (26)

3

33

10 (16) 4 (9) 32 (13) 8 (22) 1

10 (16) 1 7 (15) 55 (22) 11 (30) 1

13 (21) 13 (29) 60 (24) 7 (19) 1

80-100

DO

16 (26) 4 8 (18) 54 (21) 7 (19) 2

12 (20) 6 13 (29) 51 (20) 4 (10)

7

60-80

80-100

DO

64 (19) 26 (20)

65 (20) 20 (15)

38 (12) 12 (9)

1 8 28 7 1

* Numbers in parenthesis indicate percentage of total during year in that group. ** Only used for large numbers. Souacs. See Appendix Conglomerate Table.

TABLE 32 DECLINE IN MONTHLY MAN HOURS BY MARKET ORIENTATION (number of firms) * Market

U.S. Local

0-20

Percentage Decline from Peak Month in Peak Year to Trough Month in Trough Year 1953-4 1957-8 40-60 20-40 40-60 60-80 80-100 DO 0-20 20-40

34 (14) 8 (16)

35 (14) 10 (20)

51 (21) 13 (25)

41 (16) 11 (21)

54 (22) 2 (4)

32 (13) 7 (14)

36 (11) 19 (14)

* Numbers in parenthesis indicate percentage of total during year in that group. Souaca. See Appendix Conglomerate Table.

59 (18) 27 (20)

67 (20) 29 (22)

THE.. BEHAVIOR OF -EROMOTED FIRMS

99

TABLE 33 DECLINE IN MONTHLY MAN HOURS BY SIZE OF FIRM (number of firms) * Size

Percentage Decline from Peak Month in Peak Year to Trough Month in Trough Year

0-1000 1-2000 2-3000 3-4000 4-5000 5-10 000 10 000

0-50

1953-4 50-90

90-100

0-50

1957-8j 50-90

90-100

43 (54) 17 (45) 15 (48) 8 (42) 6 (30) 20 (43) 11 (52)

31 (39) 14 (37) 14 (46) 5 (26) 10 (50) 18 (39) 7 (33)

6 (7) 7 (18) 2 (6) 6 (32) 4 (20) 8 (18) 3 (15)

72 (49) 27 (42) 25 (40) 18 (51) 16 (46) 23 (40) 18 (49)

66 (45) 26 (40) 24 (38) 13 (37) 14 (40) 28 (49) 11 (30)

10 (6) 12 (18ti 14 (22) 4 (12) 5 (14) 6 (11) 8 (21)

* Numbers in parenthesis indicate percentage of total during year in that group. Souncs: Computed from data gathered by EDA for this project.

1) The Local Autonomous Firm It was observed that a large proportion of firms in this category substantially curtailed man-hours (over 60 %), and another large contingent curtailed man-hours from 40 % to 60 % in the first period, and that in the second period there was an increased emphasis on the extremes, i. e., either small or substantially large curtailments. Let us examine this behavior in the light of the group's structural characteristics. a) Market Orientation Although Local Autonomous firms were fairly evenly divided between the U. S. and the local market in the first period (with a slight edge toward the local market), the -increase in the total number of firms by the second period shifted the balance drastic-

100

THE BEHAVIOR OF PROMOTED FIRMS

ally toward the latter. Between the two periods there was a pronounced increase in the relative share of locally-oriented firms which curtailed operations by over 60 %, but a reduction in the other extreme. Thus the market orientation could partially explain the movement toward the larger extreme of Local Autonomous firms but not the encreased share of firms at the other extreme. b) Industry Group During the first period there was a concentration on Food, Textile Mill, Apparel, and Stone, Clay, and Glass products. In the first three classes a large proportion of firms curtailed their output drastically, while the latter had most of its firms curtailing man-hours in a medium fashion. In the second period Food and Stone, Clay, and Glass products, which are still among the leading products in the group, have a much larger proportion of firms curtailing man-hours by a medium amount and by small amounts, while Apparel firms had a tendency to maintain a large proportion in the large extreme, but increased the proportion in the small extreme at the expense of medium declines. Among the products which grew in relative importance, Furniture had relatively more firms in the large category, while Metal products were balanced with a slight emphasis on the larger curtailing side. We thus find that the polarization of the reaction pattern of Local Autonomous firms is also explainable when examined through iñdustry groups, many of the new industries tending toward large man-hour curtailments and the older ones either polarizing or shifting toward smaller man-hour reductions. The balance of this polarization is slightly tipped toward the larger side, which is also the general picture which emerges when one simply examines the over-all behavior of Local Autonomous firms. c) Size Local Autonomous plants are mainly of the very small type, which were found to be among the stabler firms, though losing some of their stability in terms of curtailment of man-hours in the second period. The biggest growth in other sizes has been in the 1,000 -2,000 man-hour category, which has also descreased in relative stability; the same is true of the additional plants in the large category of 5,000 - 10,000. The only increase contribut-

THE BEHAVIOR OF PROMOTED FIRMS

101

ing to greater stability was in the 4,000 - 5;000 category. Thus the change in the structure according to plant size also contributes to an explanation for the relatively larger increase in the percentage of firms having pronounced man-hour declines. d) Conclusion Local Autonomous firms can be considered to be of a type which produce a commodity in its entirety. They are subject to market fluctuations and their reactions in sales depend upon their individual policies. Since their orientation is principally toward the local market, where sales fluctuations are smaller than on the Mainland, one would expect considerably milder reactions in terms of man-hour curtailment. This was true in the first period when the number of firms curtailing man-hours over 60 % was proportionately smaller thari elsewhere. The change in the second period, with a larger proportion at both extremes, could be accounted for by the higher standard of living and continued strong rise in the consumption of non-durable commodities, which include food and apparel, while there was a sagging in the consumption of non-durable commodities in 1957-58 (consumption of the latter falling from $ 102 million in 1956 to $ 94 million in 1957, and rising again to $ 102 million in 1958), which include furniture and metal products. The growth of Local Autonomous firms has tended in the recent recession not to change their previous behavior, which neither de-emphasized nor worsened the fluctuations which they passed on. But the • movement toward greater numbers at the extremes has some interesting portents. The shift toward smaller decreases occurred mainly in older industries, while the newer types — furniture and metal products — strengthened the types of firms with larger fluctuations. If we assume that the latter are the dynamic sector as far as this group is concerned, then we should expect that in the future Local Autonomous firms will pass on the cycle with great amplitudes. 2) The Local Contractor This group was quite unimportant in the 1953-54 period when only two firms operated in that category. By 1957-58, it had increased to eleven firms. Almost all of the group curtailed manhours over 60 %, the majority over 80 %.

THE BEHAVIOR OF PROMOTED FIRMS

102

a) Market Orientation In both periods practically all the firms catered primarily to the Mainland market, where there was a stronger emphasis on firms with large declines in man-hours worked. b) Industry Group Most firms operate in the Apparel industry, with a couple in the Miscellaneous category. In the former there was a slight decline in the proportion of firms having, high curtailment rates and a pronounced increase in the small decrease category, while in the latter there was án increased tendency for large declines to take place. Thus, the industry group explains the firm's behavior only to a small extent. c) Size The majority of firms were in sizes which became less stable 5y the second period, i. e., in the smallest and next to smallest group and the 5,000-10,000 group, but the smallest group was still one of the stablest. d) Conclusion Local contractors had a tendency to curtail man-hours by a proportionately large amount and thus to pass on the cycle in a magnified form. This is possibly due to their newness and lack of contacts on the Mainland — especially when compared to the greater stability of Mainland contractors. (See below.) 3) Mainland Autonomous The emphasis in both periods was on firms with large curtailments of man-hours. In 1953-54 over 50 % of the firms were in the more than 60 % decline category. There was some mellowing in the second period with relatively small increases in the 0-40% curtailment category, but there was also a very large increase in the 80% - 100 % curtailment category.

THE BEHAVIOR OF PROMOTED FIRMS

103

a) Market Orientation There was a shift in emphasis between the two periods. During 1953-54, 75 % of the firms were Mainland-oriented; but with the substantial increase in the number of firms in the industry between the two periods, the emphasis shifted on to locallyoriented firms which in 1957-58 constituted 60 % of all the firms. The heavy accent on drastic declines in the first period is not fully explained by the Mainland orientation, where the general emphasis on firms with strong declines in man-hours was not as heavy. The increase in the proportion of firms with mild declines in the second period can be partially explained by the greater orientation toward the local market, while the heavy increase in the number of firms curtailing operations by 80 % - 100 % could be due to the 40 % still oriented to the Mainland, in whose general category there were relatively more firms at the extreme end. b) Industry group During the first period there was an emphasis on the Miscellaneous group, with a number of firms in electrical machinery and the textiles. By 1957-58 the growth in the number of firms was characterized by a spread over almost the whole spectrum of the industry groups. The most notable growth was in the Rubber, Stone, Clay, and Glass, Fabricated Metal Products, and Furniture industries. The Rubber industry had the largest proportion of firms. in the smallest decline categories, while the other three industries were fairly evenly balanced, inclining only slightly to the heavy curtailment rate. The Miscellaneous sector; which is still the most dominant, had a larger percentage of firms in the heavily declining group than in the previous period. It therefore could be inferred that this category's small increase in the lightly declining group was due to a rise of the Rubber and Stone, Clay, and Glass Products industries, and the strong increase at the otiier-extreme to the Miscellaneous and Electric Machinery .industries. c) Size Mainland Autonomous firms were predominantly small in both periods. In the second period most of the increase .occurred in

/

104

THE BEHAVIOR OF PROMOTED FIRMS

the 0-1,000 man-hour group, with most of the rest following in the next two size groups. Since all these size groups experienced a greater degree of declines in the second period, it would seem that their possible influence on this category of firms was in the growth of the number of firms in the large declining group.

d) Conclusion The policies of Mainland Autonomous firms were such as to pass on cycles in a magnified form in the first period, whereas in the second period the shift was such as to make policies more «neutral», i. e., neither exaggerating nor minimizing fluctuations. If the growth of firms in the newly important industries should continue, the movement toward even smaller general fluctuations emanating from this category should continue. Thus the most important influencing variables are the market orientation and the industry group.

4) Mainland Affiliated This is the largest group and had the most even distribution of firms according to percentage man-hour curtailment. During the 1953-54 period, the balance was slightly inclined toward the larger curtailment group, and this tendency continued with slightly more emphasis in 1957-58. a) Market Orientation

s

In both periods the majority of the firms catered to the Mainland market, although there has been a 10 % increase of the firms selling in the local market in the snd—period; this n—the Mainland-oriented still leaves over 75 % of the fi group. From the tabulation—percentage of the size of man-hour expect a somewhat greater amount of decline, we would— th firrs`iñ the large-decline category in the first period, and a somewhat smaller number in the second period than actually occurred. Hence we suspect that market orientation is not a principal independent variable.

THE BEHAVIOR OF PROMOTED FIRMS

105

b) Industry Group Apparel and Textile Mill Products are the dominant industries in both periods; this is followed by the Miscellaneous• group, machinery, Leather products, metal products, and rubber products in order of importance. The greatest relative growth between the two periods occurred in Metal Products and Electrical Machinery. An examination of the behavioral pattern of the Apparel and Textile Mill Products groups between the two periods will reveal an extremely parallel pattern with the Mainland Affiliated group in general, i. e., a decline in the emphasis on both extremes, but there is a little heavier bias toward the large declining groups. In the most rapidly growing industries there developed a heavier concentration on the large curtailment groups, except for the Petroleum group, where the tendency was in the other direction. It thus seems that industry group has a strong influence on the pattern of behavior of Mainland Affiliated groups and that signs point to a greater proportion of firms with larger man-hour declines. c) Size Again we find the largest absolute number in the smallest group, though proportionately this is much smaller than other categories. The second largest number of firms we find in the two largest groups. Over the two periods the smallest group has declined proportionately, the largest ones retaining their relative' share, while the medium-sized groups have increased rapidly. Because the greatest gains were in groups which slightly increased the number of firms in the higher fluctuating categories (e.g., the largest proportional increase was in the 2.0003000 man-hour group, where the percentage of firms in the small curtailment group went down by the great amount — from 48 % to 40 %), the influence of plant size would seem to be toward a larger degree of fluctuation. d) Conclusion Since most of these firms are only a part of a larger complex of firms or plants in one corporation, some of the observations made at the beginning of this chapter become relevant. Depending on the distribution of the parent plant's curtailment policy„

~ I i06

THE BEHAVIOR OF PROMOTED FIRMS

the Mainland cycle can either be passed on in a magnified, equal, or smaller form to the island. Our interpretation of the data presented is that the behavior of the firms in this category was such on balance as to have a neutral effect in the first period, with a very slight tendency to magnify the Mainland cycle in the second period. But this is only a tendency. If, however, the pattern of growth continues with an increasing number of firms in industries which fluctuate more in their total manhours, there is a chance that in the future the Mainland cycle -will be passed in a substantially more magnified form. The larger declines in 'the second period might also be due to the greater fluctuation of the important textile products and metal products on the Mainland; thus it is difficult to say exactly whether the greater man-hour curtailment was caused. by the market variable •or changes in firm policies.

5) Mainland Affiliated Contractor This category produced the most dramatic change between the two periods. In 1953-54 it contained the largest proportion of firms in the big curtailment group, while in 1957-53 there 'was such a pronounced shift to the other extreme that these contractors had the largest proportion of firms in the smallest curtailment group.

a) Market Orientation One hundred per cent toward the Mainland market. There 'was a small movement toward lesser curtailment, but not of a kind which would have so markedly contributed to the radical shift in this category.

b) Industry Group The dominant industry is Apparel, with a small scattering in the Textile Mill Product and Miscellaneous group. There was .a noticeable movement toward milder declines in Apparel, but not in the latter two groups. Thus Apparel Products might have exercised some influence in the drastic shift.

~

THE BEHAVIOR OF PROMOTED FIRMS

107

c) Size The dominant groups here were the firms of medium size, especially the 3,000-5,000 man-hour groups, which experienced a sharp rise in more stable firms over the two periods. Thus, size group seems to be of relative importance in explaining the drastic change in the behavior of these contractors. There are a relatively large number of firms in the biggest groups, the proportion of these firms being larger than in any other category. d) Conclusion It is difficult to explain the surprisingly favorable developments in this category, especially in the light of large declines experienced by local contractors who also cater principally to the Mainland market. It might be assumed that it takes time for a firm to establish a steady relationship with a Mainland customer, that the first contracts cancelled or not renewed are with relatively new firms. It this is the case, it might explain the difficulties of the local contractors, who were relatively new in the second period. It could also be assumed that once a new contractor has established himself, he might operate with longer contracts, which might tie him over a cycle and keep him operating at close to full capacity. In general, Mainland Affiliated contractors tended to magnify the cycle by their behavior in the 1953-54 period,, and tended to soften the impact by their policies in 1957-58. 6) Foreign Firms These were small firms, oriented mainly toward the Puerto Rican market. Their general tendency was to curtail production in a small to medium way during the first period, with slightly larger declines in the second period. General Evaluation All the evidence presented in this chapter seems to show that if the propagation of the Mainland cycle were dependent only on the actions of already established firms, fluctuations would be passed on in a slightly magnified way. And this tendency has increased somewhat in the second period.

108

THE BEHAVIOR OF PROMOTED FIRMS

It would be difficult to make any definite recommendations to the policy, maker on the basis of this evidence. There are no definite black-and-white trends. In the Mainland Affiliated category, it is the newly-growing industries which had the greatest rates of curtailment. Should the policy maker (i.e., especially EDA promotion personnel) therefore de-emphasize the promótion of such industries? In other words, should he sacrifice some growth for greater stability? In all probality, the answer is no. From a more positive point of view, Mainland Affiliated Contractors have shown a very high degree of stability in the second period. The policy maker should thus investigate the reasons for and possible continuation of such behavior before emphasizing the promotion of such firms. The heavy curtailment of Local Contractors should not be taken as a cue for de-emphasizing such firms because of the possibility of greater stability after an. adjustment period (i. e., better Mainland contacts), which is what happened to Mainland contractors. In the promotion of industry groups, and emphasis on firms in the Apparel, Chemical, Leather, Rubber, and Food Product`s, there would seem to be a contribution to greater stability. There is not much that can be done about market orientation, i. e., more emphasis in the promotion of firms catering to the local market because of their greater stability. But it has also been established that even such firms displayed a greater degree of fluctuation in the second period than in the first. There remains the question of how Mainland Affiliated firms tend to pass on the business cycle in their curtailment of manhours policies. Unfortunately no information is available on the man-hour policy of such corporations with their Mainland plants. The general impression is that Puerto Rican plants are not discriminated against, i. e., that they are made to curtail their activities to a larger extent than their Mainland counterparts. Only in such rapidly growing but still relatively small industries as metal products and electrical machinery is there some evidence that curtailment policies" in the downward phase of the cycle are more drastic than on the Mainland. * * * The material on the following pages constitues an Appendix to this chapter. It consists of: 1. A sample sheet from which tabulations were made. 2. A conglomerate Table from which tabulations in chapter Tables were made.

~_

APPENDIX TO CHAPTER IV 1. Sample sheet from which tabulations were made. 2. Conglomerate table from which tabulations in the tables of chapter IV were made.

109

INDUSTRIALIZATION PROJECT

D.O. Name of plant Name of product

S.I.C. S.I.C.

(note only if there was a change in product) 1957-58

1953-54

Ownership (Mark with uX») Local - autonomous Local -contractor Mainland - autonomous Mainland - affiliated Mainland - affiliated contractor Mainland - independent contractor Foreign Sales (percent distribution) Puerto Rico United States Foreign 954

1953

1957

1958

(Highest month) (Lowe t month) (Highest month) (Lowest month)

Man Hours (1000) (Month) Remarks:

111

BEHAVIOR OF EDA PROMOTE D FIRMS OVER TWO CYCLES (percentage decline from peak '53 month to trough '54 month; same for '57-8) 0-10

10-2

20— Food & Kindred Products 1953-4 LA (pr) * (us) MAff (us) 1957-8 LA (pr) (us) MAut (pr) MAff (pr) (us)

20-30

30-40

40-50

50:60

60-70

1 1

2 1

70-80

1

1

1

90-100

1 2

1 1

DO

1 1

3

2 1

80-90

1 1 1 1

21. —Tobacco Manufactures 1953-4 MAff 1957-8 MAff

(us) (pr) (us)

22 —Textile Mill Products 1953-4 LA (pr) (us) MAut (us) (pr) MAff (us) MAff C (us) 1957-8 LA (pr) (us) MAut (us) MAff (pr) (us) MAff C (us)

1 1 1

1 2

1

1 1 1

1

2

1 3 1 1

1

1

1 2 1

3

2

3

2

2 .7

1 1

1 1

6 1

3

1 2 1

2 3

2

1 1 3 1

6 1

4 1

BEHAVIOR OF EDA PROMOTED FIRMS OVER TWO CYCLES (Contd.) 0-10 23—. Apparel & Related Prds 1953-4 LA (us) LC • (us) MAut (us) MAff (pr) MAff 5 (us) MMf C (us) For (pr) 1 1957-8 LA (pr) (us) LC (us) MAut (pr) (us) MMI (pr) (us) 1 MAff C (us) 4 For (pr) 24— Lumber and Wood Prods. 1953-4 MAff (us) 1957-8 LA (pr) MAff (us) Furniture Fixtures 25— & 1953-4 LA (pr) MAE (pr) 1957-8 LA (pr) LC (pr) Maut (pr) MAff (pr) MAff (us)

10-20

3 1

20-30

5 1

30-40

40-50

2 1

1 1 1 1 3 3

50-60

60-70

70-80

1 7 4

1 3 5

1 3

DO

2 4

9 1

9

1

1 4

1

1 1 3 1

2 1 5 1

3 1

6 2

90-100

1

2

7 3

80-90

1 1

1

1 8 6

1 13 4

2 5 3

3

1

6 2

3 7 3 1

8 5 1

1 1

1 1

1

1 1

1 1 2

1

3

1 1 1 1 1

1

BEHAVIOR OF EDA PROMOTED FIRMS OVER TWO CYCLES (Contd.)

26 —Paper & Allied Prds. (pr) 1953-4 MAff (us) 1957-8 MAut (pr) MAff (pr) (us) 27— Printing & Publishing (us) 1953-4 MAff (pr) 1957-8 LA Maut (pr) MAff (us) MAff C (us)

10-20

1

20-30

30•40

1

1

40-50

50-60

60-70

70-80

80-90

90-100

DO

£

0-10

1 1 1

2

l 1

1 1 1 1

(us)

28 —Chemicals and Allied Prds. (pr.) 1953-4 LA MAff (pr) (us) MAff C (us) (pry 1957-8 LA MAut (pr) (us) (pr) MAff 29 —Prods. of Petroleum & Coat (us) 1953-4 MAff 1957-8 MAut (pr) (pr) MAff (us) 1

1 1 1 1

1 1 1

1 1 1

1

1 2

1

1 1

1 1 1

1 1

xM

BEHAVIOR OF EDA PROMOTED FIRMS OVER TWO CYCLES (Contd.) 0-10

20-30

1

1

1

1 1

30-40

40-50

50-6Ó

60-70

70-80

80-90

90-100

DO

1

1 1 1 1

1

1 1 1

1 1

4

3

2

2

1 3

1

1

1 2

2

3 1,

1

2

2

3

1

2

2

1

2

1 2

5

1

2 1

1

(pr)

30 —Rubber Products 1953-4 MAff (pr) (us) 1957-8 LA (pr) 1 (us) MAut (pr) (us) MAff (pr) 1 (us) 1 31 —Leather & Leather Prds. (us) 1 1953-4 MAff MAff C (us) 1957-8 MAut (pr) (us) MAff 1 MAff C (us), 32— Stone, Clay & Glass Prds. 1953-4 LA (pr) MAut MAff (us) (pr) For 1957-8 LA (pr) 1 MAut (pr) MAff (us) For (pr) 1 33 —Primary Metal Industries 1953-4 MAff (us) 1 1957-8 LA (pr) MAut (pr) MAff (pr) 1 (us)

10-20

1 2 3

1

3

1

1 1

1 1 1

1 1 1

1

1 1

1

BEHAVIOR OF EDA PROMOTED FIRMS OVER TWO CYCLES (Contd.) 0-10 34 —Fabricated Metal Prds. 1953-4 MAff (pr) (us) (pr) For (us) LA (pr) 1957-8 LA MAut (pr) (us) MAff (pr) (us)

10-20

20-30

30-40

40-50

1

1

50-60

2

60-70

70-80

80-90

90-100

DO

1

1

1 1

1

1

1 1

1 1 1

1

1

1 1 2

2 1

35— Machinery (excp. elect.) 1953.4 MAut (us) MAff (pr) (us) 1 1957-8 LA (pr) MAut (pr) MAff (pr) (us)

1

36 --Electrical Machinery 1953-4 LA (pr) (us) MAut (pr)

1

1

1

1 1 1 2

2

2

1

1 1

1 1 1

1 1

1

1

1

1

1 1 1

1 1

•

1

2

2

1

1 1

1957-8 MAut MAff

(us) (pr) (us) MAff C (us) For (pr)

1 1 5

3

1 1

1 1

2 1

3

5

2 1

2

BEHAVIOR OF EDA PROMOTED FI RMS OVER TWO CYCLES (Contd.) 0-10

10-20

20-30

30-40

1 1

1

40-50

50-60

60-70

1

1 1,

70-80

80-90

90-100

DO

1957-8 (pr) (pr) (us) MAff (pr) (us) MAff C (us) For (pr) 38-9 - Miscellaneous (pr) 1953-4 LA (us) LC (us) MAut (pr) (us) MAff (pr) (us) MAff C (us) (pr) 1957-8 LA ' (us) LC (us) MAut (pr) (us) (pr) MAff (us) MAff C (us) (pr) For LA MAut

1 1 2

2

3

1 1

1 4

4

3

2

1 2 1

6

3

1 1 1

1 1 2 1 5

1

5 1

1 1

1 2 4

2 1

1 3 1

2 1

2 1

1 2

1 1 5

1 1

1

1 1

2 1

2

4 1

5 1

1 1

1 3

1 1

1 2 1

3

5

5

1 1 5

1 1 1 4 1 l

6

5

LC - Local Contractor; LA - Local Autonomous; MAut - Main land Owned Autonomous firm; MAff - Mainland Affiliated; MAff C Mainland Affiliated Contractor. (pr) firm sells predominantly in Puerto Rican market; (us) US Mainland Market. DO - Discontinued Operations. Sounca: Computed from data Gathered by EDA especially for this project. "

í;I-iAPTER V

PUERTO RICO'S ECONOMIC STRUCTURE AND THE POSSIBLE IMPACT OF UNITED STATES MAINLAND RECESSIONS Our principal difficulty so far in estimating the effect of the business cycle on an industrializing economy 'like Puerto Rico has been the very presence and momentum of growth. It is of interest, of course, to discover what impact external economic fluctuations will have on the rate of growth itself. But the growth factor's presence will also tend to minimize the impact an external recession will have on the developing economy. This, is a desirable situation, but it tends to obscure relationships which will hold once the impact of the growth factor has receded, i. e., it will not make it clear to the policy maker how exposed the economy is really becoming as it acquires a new industrial structure and how severe internal repercussions will actually be. Although the analysis of the firms' behavior over the cycle, in the previous chapter, gave us some clues, many of these firms are still too new to give us a picture of their possible behavior during future fluctuations. In this chapter we shall .try to gain a more profound picture of the possible vulnerability of the economy by making use of information available on the structure of industries in the U. S. which have settled in Puerto Rico. Puerto Rico's Import Content of Expórts An interesting and possibly useful measure of Puerto Rico's exposure, and especially of the internal repercussions of t1is exposure, to the Mainland cycle would be a measure of the degree of import content in the island's exports. One would expect that the higher the import content of exports, the smaller would be 119

120

IMPACT OF UNITED STATES MAINLAND RECESSIONS

the repercussions of outside recessions which hit the island through smaller exports. It should be kept in mind, however, that this amounts to only a partial analysis. A Mainland recession can also cause a substantially diminished flow of capital to the island, resulting in a slow-down of investment activities which creates considerable unemployment. But, especially, a knowledge of the trend in this ratio will be of interest, because ultimately, when industrialization has «stabilized,» i. e., when the structure of the Puerto Rican economy will have assumed a more permanent form, exports rather than capital flows will become the more .important channel through which a Mainland cycle will be passed on. The following is a description of the method I used to obtain an estimate of the import-content of exports. The first step was to make a list of the principal export categories of Puerto Rico. I then turned to the input-output Table, with 192 categories of 1947 (the only one in existence) in order to get a percentage distribution of each export product's inputs. (1) This raises the question of whether one is justified in using production coefficients which are fairly old and which apply to the U. S. Mainland. A first justification is simply that the U. S. input-output data are the only ones available. It can also be claimed that technological coefficients are basically not too "different in island industries than on the Mainland. A majority of industries in Puerto Rico were set up by Mainland concerns and it seems reasonable that techniques of production similar to the Mainland should be adopted. It might be countered that more labor intensive techniques of production are bound to be applied in Puerto Rico due to the abundant supply of cheap labor. It is doubtful, however, that this is the case in most industries (except maybe the building industries and agriculture), since production functions in such industries as Apparel for mass markets, Electronics, etc. are not too flexible. What happened in Puerto Rico was that industries which are relatively more labor intensive per se settled there in order to take advantage of the abundant factor. As far as the time element is concerned, i. e., the input-output data referring to 1947, one has to make the assumption that change in techniques of production (if they have changed at all) are so small that inferences made from the 1947 data are basically still valid. (1) This input.output information was obtained through the courtesy of the Harvard Economic Research Project.

IMPACT OF UNITED STATES MAINLAND RECESSIONS

121

The Puerto Rican export categories were not directly compar able to the categories of the input-output Table. This was due mainly to the latter's much finer breakdown. This actually served our purpose, since it enabled us to lump together only those items of a particular category which were actually produced in Puerto Rico. For example, in the category Metal Products, we would only include those products which are known to be produced on the island; as a result, we would obtain coefficientswhich approximate more closely to Puerto Rican conditions. Once the input coefficients of the principal export items were obtained, each coefficient was split in two, i. e., into imported and domestic categories. This was done in a fairly arbitrary way,.. and was based partly on specific knowledge or inferences from. my general knowledge of the basic structure of the Puerto Rican economy, and partly on the ratio of imports to exports for each commodity. Table 34 contains two examples of this procedure._ I usually assumed that the labor coefficient, which means direct-: labor, was entirely of domestic origin, while inputs such as «metal products and machinery,» «coad mining,» or «copper,. aluminum and metal rolling» were assumed to be entirely imported. Others were split in a rough manner. For example, forthe input «synthetic fibers» in the Apparel industry, we assumect2 that four-fifths were imported and the rest domestically supplied,.. or, for lack of more adequate knowledge, we assumed that. transportation input was divided in a roughly 50-50 manner, etc... Once this procedure had been followed for each major export. commodity, each column was added, and the result gave us estimates of the percentage of total inputs producéd domestically, and total inputs directly imported. Thus, we obtained an idea of the import content of exports in individual industries. But in order to get an over-all picture of the import-content of Puerto Rican exports, it was necessary to average the ratios of aIM industries in a carefully weighted manner. This was accomplished by multiplying the import content ratio of each industry by thee ratio of the value of that industry's exports to total Puerto Rican' exports. The sum of these operations for each industry, which can be computed for any one year, gives us a rough measure of the import content of total Puerto Rican exports. This simple procedure can easly be summarized in symbolic form. If ICa is the import-content ratio for industry a, if Xa stands for total exports of industry a, and if TX represents total Puerto Ricans=

122

IMPACT OF UNITED STATES MAINLAND RECESSIONS

✓exports in any one year, then the above-described operation can .be represented in the following fashion: Xa Import content ratio for year 19— = E (ICa X

) TX

We tried to estimate the import-content ratio for three dif''ferent years, the Puerto Rican fiscal years 1946-47, 1951-52, and 1956-57. The operations to obtain the ratios and the actually obtained ratios are reproduced in Table 35. 1'he resulting ratios reveal a substantial rise in the import content of exports over a ten-year period. The rise was about ""9%, most of it occurring in the latter half of this period. This is not a surprising result. In those ten years, it will be remembered, the export structure had changed considerably. In 194.6-47, more than three-quarters of the island's exports consisted of such •commodities as sugar, tobacco, and other agricultural products, the import-content of which, is very low (less than 10% for most .of them). By 1957, this group made up only about 40% of total exports, while such items as Textile. Fibers and Manufactures (including a high percentage of apparel products), and Machinery and Vehicles, have together almost doubled their share of total exports. As can be seen in Table 35, the import-content ratio of the expanding items on the export list is high; all of them are 4over 20%, and some are as high as 47%. From the point of view of the spread of the cycle through direct trade channels, it would seem at first that the greater the import-content of exports of Puerto Rico, the smaller will be the internal repercussions of the changes in the, value of exports,, if all other factors remain equal. From this point of view, it would also seem that the industrialization of the last decade, which has changed the commodity structure of exports so drasti.zcally, has introduced an element which reduces the island's sensitivity to the Mainland cycle by increasing the import-content 'of export ratio from not quite 1'1% to over 19%. By sensitivity, I do not mean that this particúlar development has changed the reaction pattern to Mainland developments, but rather the degree of such reactions. It should be noted, in passing, that although higher import content ratios of xports might serve as :immunization factors, they also reflect the basis on which Puerto "Rico's industrialization is built — adding value to imported raw materials or semifinished goods which will again be exported for

TABLE 34

INPUT COEFFICIENTS OF SELECTED INDUSTRIES WITH DIVISIONS OF COEFFICIENTS INTO IMPART AND DOMESTIC CATEGORIES FOR PUERTO RICO Total Input Coefficient

Type of Input

A) Apparel: 1. Labor 2: Repairs 3. Advertising 4. Real Estate 5. Banking 6. Trade and Storage 7. Transportation 8. Utilities 9. Plastic Products and Misc. Mfgs. 10. Electrical Appliance 11. Metal Products and Machinery 12. Leather and Mfgs. 13. Misc. Rubber. Petrol & Coke Products 14. Synthetic Fibers 15. Other Chemicals 16. Paper Products and Printing 17. Textile Mfgs. 18. Spinning, Weaving, Dyeing 19. Fishing, Hunting, •Trapping 20. Cotton TOTAL B) Sugar: 1. Labor 2. Repairs 3. Advertising, other Services 4. Real Estate and Rentals 5. Banking 6. Trade and Storage 7. Transportation 8. Utilities 9. Tools and Machinery 10. Concrete, Non-Met. Minerals 11. Petroleum and Coke Products 12: Chemicals 13. Paper Products • 14. Printing 15. Furnishings and Non-Apparel 16. Sugar 17. Misc. Food Products & áll other Agric. 18. Stone, Sand, Clay and Abrasive 19. Coal Mining TOTAL

25.30 .03 1.49 .77 .17 2.74 .68 .40 1.59 .18 .07 .50 .33 1.59 .07 .91 16.56 26.84 2.11 .16

8.81 .12 .78 .30 2.71 .98 2.12 .53 .08 .05 .22 .36 .86 .37 .90 58.01 15.70 .11 .63

Import Coeff.

1.08

.34 1.06 .14 .07 .40 .33 1.06 .07 .64 11.04 8.93 2.11 .16

.90 1.06

Domestic Coeff.

25.3Q .03 .41 .77 .17 2.74 .34 .40 .53 .04 .10 .53 .27 5.52 1'.91

8.81 .12 .78 .30 1.81 .98 1.06 .53.

.08 .15 .18 .57 .45

.05 .07 .I8 .29 .37 .45 58.01 15.70 .11

.63

SOURCE: U.S. Bureau of Labor Statistics, courtesy of Harvard Economic Research Project. input-Output Data for 1947.

123

TABLE 35 MEASUREMENT OF THE IMPORT CONTENT OF PUERTO RICO'S EXPORTS Export Product

Imported Input Content

1946-7 Weight* Weighted Ratio"

1951-2 Weight Weighted Ratio

Misc. Manufactures Toys & Sporting Goods Plastic Products Apparel Misc. Textile Products Alcoholic Beverages Sugar Other Non-Met. Miner. Stone, Sand, Clay, etc. Vegetable Food Products Tobacco Scientific & Prof. Equip. Optical & Photo Equip. Communication Equipment Electrical Appliances Metal Products Drugs. Medicines Industrial Chemicals Paper & Related Mfgs. Footwear

.3579 .3231 .2511 .2704 2304 .2486 .0388 .1299 .0872 .0979 .0572 .4084 .2101 .3028 .4615 .4743 2511 .3091 .3347 .4404

.011

.0039

.230 .001 .047 .528

.0622 .0002 .0117 .0205

.042 .086

.0042 .0049

.024 .002 .005 .205 .037 .019 .519 .005 .006 .028 .056 .001

.0086 .0006 .0012 .0554 .0088 .0047 .0201 .0006 .0005 .0027 .0032 .0004

.004 .010 .013 .001 .010 .004

.0018 .0047 .0033 .0003 .0034 .0018

TOTAL ***

.003

.0009

.1084

* Weight — ratio of value of Product's exports to value of total exports. ** Column 1 times weight for each year indicated. *** Sum of Weighted Ratios is import-content of exports for indicated years. Sovxcs: Imported Input Content — see Table 34 for methodology. Weight — computed from statistics in Puerto Rico Statistical Yearbook, 1957, 1958.

.1221

Weight

1956-7 Weighted Ratio

.048 .012 .023 .250 .068 .014 .296 .035 .015 .011 .037 .004 .001 .005 .074 .012 .013

.0172 .0039 .0058 .0676 .0163 .0035 .0115 .0045 .0013 .0011 .0021 .0016 .0002 .0015 .0341 .0057 .0033

.011 .013

.0037 .0057 .1906

IMPACT OF UNITED STATES MAINLAND RECESSIONS

125

the Mainland market. And thus the higher this ratio, the smaller will be the relative benefits the island will draw in terms of employment of its manpower and other resources. Although Puerto Rico's import-content ratio of exports has increased substantially in the last decade, its implication for decreasing internal repercussions of diminished exports has to be countervailed by two considerations. First, it will be remembered that .in the same decade the ratio of exports to gross product has also increased by about 10% (see chapter II). Thus, it can be said only that although the island economy has become considerably more sensitive to fluctuations in exports, internal repercussions to such fluctuations have also increased, but to a smaller extent because of the import content of export ratio. In other words, such a procedure might give us an idea of the actual magnitude the Mainland cycle will some day have when / or if the basic growth trend will have subsided — especially if there is a substantial difference between the actual and projected values. Secondly, the impact of the Mainland cycle is also felt to a considerable extent through changes in the flow of investments. Therefore, as long as the growth factor is still dominant in the Puerto Rican economy, fluctuations in investment can easily counteract the mitigating effects of the lower import content ratio. The Structure of the Puerto Rican Economy and Predicted Repercussions of Export Fluctuations Upon finishing our work in the measurement of the import content of exports, I decided to, use the input-output data for what is a very tentative procedure. I used the input-output information to predict reactions to declines in exports during the 1953-54 and 1957-58 Mainland recessions; and these were subsequently verified against actual events. Such a procedure is of interest for a number of reasons. It might give us a rough idea about the impact which export fluctuations could have, if they are not offset by increases in output for domestic sales and investment activities; it might further give us some indication whether the greater import-content of exports has a tendency to ameliorate domestic repercussions (this would come forth when the reactions in the two periods are compared); and finally, it might throw some light on the reliability of using U. S. input-

126

IMPACT OF UNITED STATES MAINLAND RECESSIONS

output data in defining the structural relationships of Puerto Rico. I followed a quite simple procedure. After plotting on a chart monthly exports for Puerto Rico's principal export. commodities, I selected the period when a perceptible decline was apparent which coincided with or followed the two Mainland recessions. I measured the decline not by getting the difference between exports in the peak month and the trough month, but rather by getting the difference between the monthly export average in a three-to-four-month period preceding the peak, and the export average of the months of the decline to the trough. This would eliminate erratic extremes. To the value of this decline, I added the value of the product prodúced by the industry which it uses up itself. This could easily be obtained by using the domestic «self-input» coefficient from the inputoutput data and multiplying it by the value of export decline. For purposes of this analysis I was mainly interested in the effects of this decline on labor, since this is usually a principal domestic input, i. e., supplied almost totally by domestic sources. Information from the input-output data was used to «project» the value of the labor input decline; i. e., the total export decline in each industry was multiplied by the labor input coefficient. These projected declines in labor earnings were then compared to «actual» declines of labor earnings in each industry. The latter were obtained by multiplying the average weekly earnings in each industry by the number of people employed in it, and multiplying this product, in turn, by four in order to get monthly earnings of the industry. The results of this operation are reproduced in Table 36. A glance at the Table will make it obvious that the results are not too satisfactory in the sense that the predicted labor declines rarely closely approximate actual labor earning declines. This divergence between predicted and actual results was to be expected in an economy which is undergoing such rapid changes. It is in these divergences, however, that we hope to pick up some clues about the actual structure of the island economy and especially about the degree of instability the economy might face when the growth factor has receded into the background. Table 36 reveals that in almost every case examined, decline in projected earnings was larger than actual declines. One's first suspicion might be that the coefficients used for projection purposes are inapplicable to Puerto Rico. It could justifiably be . claimed that the self-input coefficients were too high for Puerto

TABLE 36 MEASUREMENTS OF SENSITIVITY OF LABOR EARNINGS TO FLUCTUATIONS IN PUERTO RICAN EXPORTS Product I. Wood and Paper (AprilJuly) 1. Export Decline (Jan.-April Av.) * 2. Self-Input (15 %) ** 3. Total Decline (1.-2.) 4. Projected Labor Decline *** (Labor coeff. 20 %) 5. Actual Decline of Earnings @ (From Feb.-July based on D-F Av.)

1954 Period

1957-8 Period (Oct.-Jan.)

—$948,000 $142,220 -$1,090,220 -$218,044 —$144,144

—$964,000 $144,600 -$1,098,670 $221,720 (no lag)

+$296,548

(6 mos, lag)

—$86,496

* Average exports for Jan. - April period minus Av. for April - July period. ** Percent of export decline. *** Percent of total export decline. Labor earnings of projection based on Dec. - Feb. average minus actual earnings in Feb..- July period.

II. Metal Products (March-July '54) 1. Export Decline (March figure) (Jan:March Av.) 2. Total Decline (self-input 5 %) 3. Projected Labor Decline (Labor input 24.16 %) 4. Actual Decline of Earnings (March-July based on Jan:March Av.)

III

—$1,030,000 -$170,000 -$1,081,500 -$178,500 -$261,209

(Oct.-Jan.) (July-Oct. Av:)

+$207,000 +$221,945 +$53,621

—$44,620

11

+$21,691

(Oct.-Jan.-July-Oct. base)

+$175,906 +$39,652

III. Textile Mill Products * 1. Export Decline 2. Total Decline (self-input 19.7 o ) 3. Projected Labor Decline Labor input 25.93 %) 4. Actual Decline of Earnings (Feb.-July '54 = based on Dec.-Feb. Av.)

—$3,426,000 —$4,100,922 $1,063,369

(Oct.-Feb.)

—$311,998

Apparel

-$2,022,000 -$2,419,928 -$627,487

Oct.-Jan. t 6 mos. Jag Textile Mill Products Oct.-Jan. 6 mos. lag Totals Oct.-Jan.

+$152,152 -$444,220 -$277,024 +$147,880 -$124,872

f

-$296,340

With lag for; apparel only

-$721,244

—$383,000 —$388,362 —$108,470

(Oct.-Jan.) (Labor 27.93)

-$1,459,000 -$1,479,426 -$413,204

—$87,300

(Oct.-Jan.) 1)

+$138,252

6 mos. lag

—$108,580

(Labor

IV. Non-Metallic Minerals (April-July) 1. Export Decline (Jan.-Apr. Av.) 2. Total Decline (self-input 1.4 %) 3. Projected Labor Decline coeff. 27.93) 4. Actual Decline of Earnings (Feb.-July based on Dec.-Feb.)

6 mos. lag c#

V. Vegetable Food Products and Beverages (April-July) 1. Export Decline (Jan.-April base) 2. Total Decline (self-input 54.40 %) 3. Projected Labor Decline (Labor input 9.46) 4. Actual Decline of Earnings (Apr.-July based on Jan.-Apr. Av.)

(Oct.-Jan.) —$9,968,000 -$15,390,592 -$1,455,950 -$401,320

'

(self-input 54.4 %) (Labor 9.46 %) (April-July)

—$20,513,004 —$31,672,072 —$2,996,178. —$512,055

7 mos. lag VI. Tobacco Manufactures and Tobacco 1. Export Decline

-$148,000

2. Total Decline (self-input 5.24 %)

-$155,755

3. .Projected Labor Decline (Labor coeff. 17.78 %) 4. a) Actual Decline: Feb.-May based on Dec.-Feb. Av. b) Feb.-July VII. Machinery and Vehicles 1. Export Decline 2. Total Decline 3. Projected Labor Decline (Labor coeff. 30.63 %) 4. Actual Decline of Earnings (using July-Oct. base)

-$27,693

1. a) Oct.-Jan. b)• March-June 2. a) Oct.-Jan. b) March-June 3. a)

+$3,192000' $533,000 +$3,359,260 -$560,929' +$597,276-

—$255,064

b) 4. a) Oct.-Jan.

-$99,733• -$208,128•

b) March-June

+$37,824-

—$31,808 Oct.-Jan. (based on July-Oct. Av.)

—$1,388,000 -$1,481,000 -$453,630-$121,528;

* Including Apparel Products. SouxcEs: a) Exports - Current Business Statistics, Bureau of Economics and Statistics, Puerto Rico Planning Board; U.S. Department of Commerce; Bureau of the Census, U.S. Trade in Merchandise and Gold and Silver with U.S. Territories and Possessions. b) Coefficients: U.S. Bureau of Labor Statistics, courtesy of Harvard Economic Research Project :""Input-Output Data for 1947. c) Actual Statistics of Labor Earnings: Employment, Hours and Earnings In the Manufacturing Industries of Puerto Rico, Bu— reau of Labor Statistics, Department of Labor, Puerto Rico. 126a

IMPACT OF UNITED STATES MAINLAND RECESSIONS

12T

Rican conditions, that a large portion of this factor in many cases is imported and thus has no bearing on declines in employ— ment. This might be a correct objection, but it could hardly account for all of the observed discrepancies. The labor coefficient used might also be regarded with scepticism, but one: would expect it to be larger, considering the labor intensive: conditions on the island. But a larger coefficient would make the discrepancies still wider. A most relevant factor which could explain the discrepancies,. i. e., the larger projected values, is the underlying growth trend. which pervaded the island's economy during both recessions. This trend stems from two specific institutional phenomena: the opening of newly-promoted firms, and increases in and changes in the consumption pattern of the population. The implication of the first phenomenon is that although total exports decline, a certain percentage of the firms in each industry are newly opened firms which have just hired new personnel and which are therefore unlikely to lay off or substantially curtail the: number of hours worked; many, for example, are trying to build up a minimum stock. Some of such new enterprises might not even have started to export products yet, since they are still in the organizational or early production stage. But the fact of their existence means that they have already employed labor, and thus, in toto, they will have increased the number of people employed in the industry. The second factor, increased consumption of the Puerto Rican population as per capita income in— creases, means that domestic sales will have acquired a relatively greater share of total produce, and domestic sales might even, continue to increase during a Mainland slack, especially considering a generally sustained growth through new investments which keep the general level employment and earnings up. Continued high consumption can also de explained by the increasing awareness by the population of the possibility of credits of variousa types. If one accepts these explanations for the mitigation of the effects of declines in exports, then one could take the projected figures as indicating that a substantially more severe impact will be felt at some future date, when the underlying growth trend will have subsided. Let us now examine briefly the seven industry groups on which the above-mentioned operations were. performed. Together they account for over 75 % of total exports..

128

IMPACT OF UNITED STATES MAINLAND RECESSIONS

Wood and Paper Decline of exports in both periods was. almost the same, and the same was therefore true of projected decreases in labor earnings in the industry. Actual declines were smaller in the 1953-54 period, but in the 1957-58 period there was no actual decline at all, but a substantial increase in labor earnings. With ..a six-month lag, a decline in the latter appears, although still less than half as smáll as the projected decline. During the first recession period, two new firms started operations, one in early and one in late 1953; during the 1957-58 period, five new-firms started operations, though some did not start in the exact period •of our measurement of actual declines; one started in March, 1.957, and one in May of 1958. Thus the greater number of openings in the second recession could account for the greater difference between projected and actual changes in labor earnings. There also occurred a considerable increase in the domestic consumption of wood and paper products, partially induced by :a very large number of openings in the Furniture and Fixtures industry. The increase in the household paper product consumption increased at a substantially greater rate in the second recession period than the general increase in average consumption. Further, it should be remembered that over 60 % of the goods produced in this industry are sold in the Puerto Rican market. If all these growth factors were assumed away, one could claim that the decline in labor earnings would have been about 50 % greater in the first recession, and that a decrease of over $ 200 million rather than an increase of over $ 296 million in `labor earnings would have taken place in the 1957-58 period. Metal Products The discrepancies observed in this industry, though not too great for 1953-54, if export declines are calculated from a January to March average, seem to have an easy explanation. In 1953-54, new openings took place with only three discontinuances, while in 1957-58 almost twenty new openings occurred, with four closings. The latter also explains the increase rather than the decline in total exports in the second period. An estimate of the real impact if the growth factors were absent cannot be gauged from

IMPACT OF UNITED STATES MAINLAND RECESSIONS

129

this data. A reference to the previous chapter (Table 30) shows that firms in this category are fairly evenly distributed with regard ₹o the rate of curtailment of man-hours, with a slightly greater emphasis on the larger declines in the second period; this follows the pattern of the average firms in all industries. Thus we can expect that with the growth factor in the background, employment and earnings in this industry will fluctuate in a parallel manner with most industries. Textile Mill Products (including Apparel) The large discrepancies here are also explainable by the large rate of openings of new firms in both periods. In 1953 alone there were over thirty openings, and in the first two months of 1954 there were eleven openings. In the second recession the situation was similar. In the 1953-54 period the impact of the recession without . the mitigating growth factors would have been more than three times as powerful on labor earnings. In the second period, the discrepancy was of a greater or smaller magnitude, depending on the use of actual data. If actual data for labor earnings is taken for the same period as the decline in exports, then projected declines would be five times greater than actual labor earning declines. If we lag the reactions on actual labor earnings, the discrepancy considerably diminishes, projected earning declines only being a little more than twice as large. Or if we juggle our data, i. e., taking a lag for Apparel in order to get a negative number for labor earnings, and not lagging Textile Mill Products for the same reason (it was possible to separate the two only in the second period) then we get a fairly close correspondence between actual and proje_ed figures, actual figures being even larger than the projected ones. The use of lags is a somewhat doubtful procedure in this case because if exports decline, thus causing a decline in labor earnings, there is no reason for any delay. Non-Metallic Minerals The generally close approximation in the first period of the projected to the actual labor earnings decline is due to the very small number of new openings. The discrepancies in the second. period are again due to about ten new openings. lo

130

IMPACT OF UNITED STATES MAINLAND RECESSIONS

Vegetable Food Products and Beverages Many of the large discrepancies could be due to the large increases in domestic consumption of food products which went on unabated by the recession. This is convincing evidence since over half of the products in this category are sold in the Puerto Rican market. Machinery and Vehicles This is a rapidly growing industry, where individual data were not even available for the first recession period. The almost four-fold discrepancy is thus due mainly to the strong growth element in this industry. The general conclusion from this procedure would be that the impact of a Mainland recession which is spread through fluctuations in exports when mitigating growth factors are not present is close to three times as powerful as was actually experienced in the last two recesions. One especialy weak side of this procedure was that we did not take notice of the possible growth trend inherent in exports themselves, i. e., that declines in exports might have been even more severe without the impact of new firms sending out their goods for the first time. But this would affect Only the general impacts, not the relative magnitudes which are of interest here. An Alternative Approach: Deviations from the Trend Before coming to any general conclusions, it would be useful to present the results of an investigation attacking the problem from a different angle. The principal approach here was to observe fluctuations in exports and labor earnings when both seasonal fluctuations and the trend have been eliminated. Once the growth factor was removed in earnings and exports, the recession appeared in the data. This would also show that the effect of the recession manifested itself primarily by a slackening in the rate of growth. For each principal category of goods we plotted the fluctuations of exports and labor earnings, fitted a trend line, and then, for each recession period, we chose the time period when the most pronounced downward deviation from the trend took place. (See Chart IV.)

CHART IV EXPORTS: FLUCTUATIONS AROUND TREND Thousands of $

4000 3500 3000 2500 2000 1500 1000 500 0 Thousands of $ 55000 TOTAL EXPORTS 50000 45000 40000

i

35000 30000

V

I~

25000 20000 15000 10000 Zr

a1

1948 1949 1950

I

I

1951

i I I I I I 1952 1953 1954 1955 1956 1957 1958 131

CHART IV (continued)

Yhouscncls of $ 1004 800METALS AND MANUFACTURES machinery and vehicles) (except 600 400240f

t

Thousands of $ 12 10 8 6 4

1948

1949 1950 1951

1952 1953 1954. 1955 1956 195T 1958

Source: Current Business Statistics, P.R. Bureau of Economics and.Statistics, P.R. Planning Board; U.S. Trade in Merchandise and Gold and Silver with U.S. Territories and Possessions, U.S. Department of Commerce, Bureau of the Census. 132

CHART Y

EARNINGS: FLUCTUATIONS AROUND TREND Million of $ TOTAL EARNINGS OF MANUFACTURING INDUSTRIES (seasonally adjusted)

10

5

I

I

I

I

Thousands of $ 340 320 300 280 260 240 220 200 180 160 140 120100 -1, 80 1953

1954

1955

1956

1957 _

1958

Source: Employment Hours and Earnings in the Manufacturing Industries in Puerto Rico, P R. Department of Labor, Bureau of Labor Stptistics. 133

CHART V (continued) Thousands of $ 580 540 500 460 420 380 340 300

Millions of $ 2,7

APPAREL, TEXTILE MILL PRODUCTS AND RELATED PRODUCTS

2,6

2,5 2,4 2,3 2,2 2,1

2,0 1,9 1,8 1,7 1,6 1,5 /

1,4 1,3 1,2 1,1 1,0 1953

194

1956

1955

1957

1950

134

~

IMPACT OF UNITED STATES MAINLAND RECESSIONS

135

Once the relevant time period for an industry group was chosen, the following procedure was undertaken. A projection of labor earnings during that period was made on the basis of trend values; this was followed by a computation of actual labor earnings during that time, adjusted in order to eliminate the growth factor. (2) The difference between those two figures represents the total deviation from the trend during the chosen time period. The same operation was repeated for exports, except that to the differénce between exports based on trend value and actual exports a self-input factor was added. The deviations from the trend would thus give a measure of the «real» impact a Mainland recession might have on Puerto Rico if the growth factor were not present as a mitigating element. Since this operation was performed for each principal export industry for the two recession periods (1953-54 and 1957-58), it also enabled us to get an idea of the «trend» in the severity a recession might have. These 'computations were put to one further use. In the previous section, we estimated changes in labor earnings by multiplying changes in exports by a labor coefficient — i.e., changes in labor earnings being a function of changes in exports. Our unknown in that case was labor earnings. Subsequently estimated labor earnings were checked against actual labor earnings. In the present. case, the unknown is the labor coefficient — changes in 'labor earnings and exports having been estimated separately. Therefore b dividing the obtained deviation from the trend in exports into the deviation of labor earnings, we ought to get an estimate of the Puerto Rican labor coefficient. This can serve as a rough check with the Mainland labor coefficient which we applied in the previous section. Also, the higher this ratio, the more sensitive is the island to Mainland fluctuations. It is thus also an index of internal sensitivity to Mainland cycles via exports. Especially, the trend in this ratio as revealed by the changes between the two periods examined will give a clue into which direction the economy is moving. It might be objected that the' ratio of these two deviations from the trend do not really reflect a labor coefficient because changes in earnings are also a function of variables other than (2) Labor earnings during that time were obtained by summing up the monthly data below the trend line, once the data had been adjusted to eliminate all growth. Growth was eliminated in such fashion as to make the data representative of fluctuation around the tr ad value applicable on January 1, 1953.

136

IMPACT OF UNITED STATES MAINLAND RECESSIONS

changes in exports, hence a real labor coefficient will be higher than the one obtained in the above mentioned manner. It should be remembered, however, that these deviations from the trend are figures for which growth values have been eliminated. And since variables other than export changes would under present conditions only bolster employment (like investment, domestic market growths, etc.), their influence is already eliminated from changes in earnings and hence they do not have to be included in the denominator, which contains the independent variable. But, keeping in mind that our calculations can not be very adequate because of the nature of the data, we could assume a small upward bias in the ratios. Due to the dearth of relevant data over the two recession periods, our experiment was restricted to four industries. In almost all iñdustries covered, and in the estimates for total manufacturing industries, the fluctuations around the trend have increased in the second recession. The one exception is Non Metallic Minerals. In the Apparel and Textile Products category the earnings to export-deviation-from-trend ratio declined because of a much larger increase in the deviation from trend for exports than for labor earnings. This indicates a much milder reaction to changes in exports than previously. This ratio also decreased substantially for Tobacco manufactures and NonMetallic Minerals. But in Wood it went up, and this was also the case for Total Manufacturing. The latter would indicate a higher ratio for such industries as metal products, machinery, leather products, etc., for which no computations could be made. On the whole, these computations leave the impression of substantially larger declines of exports and earnings when no trend is present, even compared to the projections made on Table 36, which took declines in exports (with no adjustments for trends) as a basis for projections of the decline of earnings. These declines have increased in most cases between the two cycles. (See Table 37.) Of significance are the decline in the ratio of earning to export deviation from the trend for three of the four industry groups examined. This indicates that the Mainland recession has had a greater impact on Puerto Rican exports than on labor earnings in each industry, .and thus a decline in the relative internal vulnerability of Puerto Rico seems to have taken place, i. e., earnings are less vulnerable to fluctuations in exports. The ratio for total manufactures is obviously too low to reflect the average labor coefficient of these industries. The

IMPACT OF UNITED STATES MAINLAND RECESSIONS

13T

TABLE 37 DEVIATIONS FROM TREND FOR EXPORTS AND LABOR EARNINGS DURING 1953-4 AND 1957-8 FOR SELECTED PRODUCTS 1953.4

1957-8

$18,378,000 (4/54-8/55)

$11,231,00c (12/57-1/58)

$17,063,000

$ 9,479,0002

$ 1,315,000

$ 1,752,000'

$85 442 000 (5/54-6/55)

$49,824,000 (12/57-7/58)

$80,292,000

$37,307,000]

$ 5,150,000

$11,517,000

$ 6,164,550

$13,785,850+

21.33%

12.7194

$ 3,426,300

$ 3,807,000'

$ 3,222,800

$ 3,390,100'

$

$

1. Apparel and Textile Products a) Total Earnings on Basis of Trend Values b) Total Actual Earnings Adjusted to Eliminate Growth Factor . Difference: a)-b) d) Total Exports on Basis of Trend Values e) Actual Exports Adjusted to Eliminate Growth Factor f) Difference: d)-e) g) Export Decline plus Self-Input (19.7 %) h) Ratio of Earnings Deviation from Trend to Export Deviation from Trend (Estimated Labor Coefficient) c)-g) 2. Paper and Wood Products a) Total Earnings on Basis of Trend Values b) Total Actual Earnings Adjusted to Eliminate Growth Factor c) Difference: a)-b) . d) Total Exports on Basis of Trend Values e) Actual Exports Adjusted to Eliminate Growth Factor . . . . . f) Difference: d)-e) g) Export Decline (15 %) h) Ratio of c)-g)

plus

203,500

416,900'

$ 3,718,000

$ 3,718,000'

$ 2,647,000

$ 2,123,000.

$ 1,071,000

$ 1,595,000

$ 1,231,650 16.52%

$ 1,834,250' 22.73%a

Self-Input

1138

IMPACT OF UNITED STATES MAINLAND RECESSIONS TABLE 37 (continued) 1953-4

1957-8

$ 388,200 (2/55-6/55) $ 516,900 (5/55-8/55) $ 543,980

$ 674,000 (7/57-1/58) $ 4,762,500

70.2%**

13.44%

3. Tobacco Manufactures a) Decline of Total Earnings b) Decline of Exports c) Export Decline,- Self-Input (5.24 %) . d) Ratio of Earnings Deviation from Trend to Export Deviation from Trend (Estimated Labor Coefficient) a)-c)

$ 5,012,055

4. Non-Metallic Minerals a) Decline of Total Earnings b) Decline of Exports c) Export Decline - Self-Input (1.4 %) d) Ratio of Earnings Deviation from Trend to Export Deviation 'from Trend (Estimated Labor Coefficient) a)-b)

$ 3.178,000 $ 2,659,000 (2/54-5/55) (4/58-12/58) $13,016,000 $11,944,000 (5/54-9/55) (10/57-12/57) $13,198,,200 $12,111,200

24.08%

21.95%

$ 1,199,000 (4/54-1/55) $15,272,000 (5/54-2/55) $24,223,000

$ 3,430,000 (2/58-12/58) $34,602,000 (12/57-2/58) $45,718,000

4.95%

7.50%

5. Total Manufacturing Industries a) Decline of Total Earnings . b) Decline of Exports

.

..

.

c) Export Decline - Self-Input (32.57%). d) Ratio of Earnings Deviation from Trend to Export Deviation from Trend (Estimated Labor Coefficient) a)-c)

* The rest of the table contains only the results — i.e. the difference between figures based on trend values and those based on actual data from which growth factor have been eliminated — therefore by «declines is meant this difference. ** The 1953/54 results for Tobacco are probably distorted by a poor fit of the trend line to the export data for that period. Souaca: See Table 36.

IMPACT OF UNITED STATES MAINLAND RECESSIONS

139

average labor coefficient for all Puerto Rican export industries based on U. S. input-output data (weighted according to the relative importance of each industry in Puerto Rico's export) is 18.66 %, considerably above the one estimated. The explanation cannot be found in a possible lower labor coefficient for other major export industries because most of them have labor coefficients higher than 20 % in the U. S., and there is little chance of drastically lower ones for Puerto Rico. The only plausible explanation is that Total Manufacturing includes a large enough number of industries catering only to the local market, where no decline took place. One could still take the low ratios to indicate that the general vulnerability of earnings to changes in exports is small. In the four industries examined, the ratios as estimates of labor coefficients are fairly close to the coefficients obtained from U. S. input-output statistics. This would seem to support the conclusions we reached in the previous analysis (of Table 36) that the principal factors which accounted for deviations of projected figures from actual ones on the behavior of labor earnings vis-a-vis fluctuations in exports were growth factors. Furthermore, since the error in the estimated labor coefficient is in many cases less than 5 % (i. e., the difference between the estimated coefficient and the one derived from U. S. inputoutput data), the method pursued in this section could be used to predict with some degree of accuracy the effect on total earnings of a certain decline in exports, if the necessary adjustments for growth and seasonal variations were previously made.

,

ws

CHAPTER VI

SUMMARY OF FINDINGS AND RECOMMENDATIONS FOR COUNTERCYCLICAL POLICIES The aim of this study was to obtain an understanding of how the U. S. business cycle has affected the Puerto Rican economy in the past and how it might affect the island in the future. It was shown that such an understanding was of interest because Puerto Rico's industrialization program has made the island increasingly dependent on exports to the U. S. Mainland and because a major portion of the capital for- the island's development has also come from the Mainland. Our descriptive analysis of the two Mainland recessions in the 1950's revealed a very mild reaction in the island. The 1953-4 recession was mainly felt through a substantial decline in net emigration, thus causing a rise in unemployment. In the 1957-8 period the recession was passed on mainly through ,a widespread production curtailment of older firms in the export industries. Although the decrease in the number of openings of new firms not too was drastic, the net increase in employment resulting from the opening of industrial plants was not enough to counteract the effects of the recession on older firms. We assumed in the rest of our study that reactions to these recessions did not reveal the true sensitivity of Puerto Rico to the Mainland cycle, which it was acquiring through its industrialization. The principal reason for this assumption is the impact of the development program, i. e. the substantial backlog of investment commitments which was always present during the recessions of the 1950's. In addition, these Mainland reces141

142

PUERTO RICAN COUNTERCYCLICAL POLICIES

sions have not been sharp and long enough to substantially curtail investments of newly promoted firms. Since the relative impact of the growth trend will not always be as great as in the 1950's, there is the possibility that Mainland recession might be felt more severely in the future. It was therefore our purpose to estimate changes in the «real» sensitivity of the Puerto Rican. economy, i. e. discounting the growth trend as the island industrializes. The results of such estimates would be of use to the policy maker in two broad ways: in the planning of countercyclical policies and in the direction of further promotional activities — with a'view not only toward achieving growth, but also toward achieving a more stable economic structure. Financial Relations Already in the third chapter we tried to get a first glimpse of the potential impact of the Mainland cycle by establishing the financial relations to the Mainland, the sensitivity to the cycle of the type of products Puerto Rico exports to the Mainland, and the sensitivity of Investments in Puerto Rico to Mainland aggregative variables. In the brief invetigation of Puerto Rico's financial institutions, it was found that in the past, financial conditions on the Mainland have had little direct effect on the island's credit conditions. This was due especially to the perennial excess reserve condition of local banks. The future effects of changes in Mainland financial conditions are difficult to predict. If the tendency for a. diminished amount of excess reserves should persist, the possibility might arise in the future that a Mainland recession, which cuts the flow of capital to the island and which curtails exports, would bring along tight credit conditions. Loans would decrease, interest rates would rise, and a secondary contractionary wave would hit the island, unless, of course, countercyclical actions are taken by Puerto Rican authorities. Income Elasticity of Demand for Puerto Rican-type Exports We obtained an estimate of the Mainland income elasticity of demand for the type of goods Puerto Rico exports by computing the ratio of changes in the Mainland industrial production for each of the principal Puerto Rican export commodities to

PUERTO RICAN COUNTERCYCLICAL POLICIES

143.

changes in general U. S. industrial production (for explanation of procedure see pp. 66-8.) These changes were the decline or rise in production during the various NBER cycles since the war. This procedure revealed that Puerto Rican exports are increasingly concentrated on products experiencing sharper fluctuations over the cycle, although some of the newer products, i.e. of industries most recently promoted, on the Puerto Rican export list (e. g. machinery, chemicals) show a smaller sensitivity in the downward direction than general industrial production. All this, however, is not necessarily directly relevant to Puerto Rico, since other variables have also to be taken into account when considering, exports. Investment Behavior Our efforts to establish a relation between investment in Puerto Rico and Mainland variables through regression analysis. were not sucessful. We tried to find a relation between openings of new EDA-promoted plants and a number of Mainland! variables: U. S. industrial production, investment in new plant and equipment, and profits of corporations. In this regression analysis we also experimented with various lags. The highest: correlation was found with new plant and equipment with a three quarter lag. But, unfortunately, once the trend was eliminated, it was found that no relationship of a cyclial nature: could be determined between the Mainland and Puerto Rican variable. The Behavior of Firms In order to ,get a more realistic picture of the island's expo-sure to the Mainland cycle, we took into account the type of enterprises which industrialization has attracted. Although it might be assumed that firms which were set up in or resettled to the island in their entirety could pass on Mainland fluctuations in an unmagnified form, the behavior of subsidiaries of Mainland corporations might be such as to either enhance,. mitigate, or pass on the cycle in neutral form. This depends on whether the parent Mainland corporation cuts production principally in its Puerto Rican plant, whether it cuts production ina the same proportion in all its plants, or whether the cut takes place: principally in its non-Puerto Rican plants.

x144

PUERTO RICAN COUNTERCYCLICAL POLICIES

We obtained some insight into the policy of firms under various aypes of ownership conditions, types of industries, and of different market orientation, by having had access to information on the number of man-hours worked per month by all firms promoted by EDA. We selected the highest and lowest number of man-hours -worked for each firm during the two recession periods of the 1950's and computed the percentage decline from the peak manhour month to the trough man-hour month. The latter mforma tión was used as the dependent variable and was compared to such independent variables as type of ownership, industry, and market orientation. (For detailed procedure and qualifications of method used see pp. 73-6.) The principal findings and conclusions based on this data .are as follows: 1. DISTRIBUTION OF FIRMS BY TYPE OF OWNERSHIP. The dominant group has been and still is the Mainland Affiliated firms, although there has been a marked growtn of firms in other categories (Local Autonomous, Mainland Autonomous, etc...) There has also occurred a substantial growth of firms outside the still dominant textile and apparel industries, leading toward more diversity in Puerto Rico's industrial structure. 2. MARKET ORIENTATION. Mainland Affiliated firms and Local Contractors are oriented principally toward the Mainland, while locally owned firms have found their principal market on ₹he island. Over the period covered there has been a notable .shift of Mainland Autonomous firms from a. Mainland to a more predominantly local orientation. 3. VARIATION IN MONTHLY MAN-HOURS BY INDUSTRY GROUPS. It was found that during the 1953-4 recession a substantial proportion of plants in the textile and apparel industries curtailed man-hours by over 60 % from peak to trough, whereas .curtailments in the newer industries (e. g. electric machinery) -weie hardly noticeable. During the second recession period the .same proportion of firms in the textile and apparel industries -curtailed man-hours by the same amount, but the rest of the firms In this category hardly curtailed their man-hours at all. In most other industries there was a substantially greater curtailment of man-hours worked than in the previous period. This probably reflects the more «normal» reaction of longer established plants. Many of the plants in these industries were so ne* in the

PUERTO RICAN COUNTERCYCLICAL POLICIES

149

previous recession, or were just in the process of establishing themselves, that extensive production declines were hardly contemplated by the parent corporation. It could be assumed that such growth factors are still prevalent today and that in a future recession the total amount of firms curtailing man-hours by over 60 % will be even greater than before. 4. VARIATIONS IN MONTHLY MAN-HOURS BY FORM OF OWNERSHIP. Here we find that in the first recession Mainland Affiliated firms curtailed their man-hours to a very small degree, while Mainland Antonomous firms and Mainland Affiliated Contractors curtailed their operations to a much larger extent. In the second recession the general drift was toward a greater contractionary pattern. In many categories, such as Local and Mainland Autonomous firms, there was a polarization of firms either not curtailing operations at all or curtailing them drastioally. The number or firms which discontinued operations was also greater than previously. Only Mainland Affiliated Contractors, which severely curtailed production in the first recession, were hardly affected by the second recession. 5. VARIATIONS BY MARKET ORIENTATION. Experience during both recessions reveals a greater degree of than-hour contraction of Mainland-oriented firms than of firms catering primarily to the local market. 6. VARIATIONS BY PLANT SIZE. During the first recession both the smallest and largest types of firms curtailed man-hours very little, while in the second period there was a slight increase in man-hour curtailment of the biggest firms. In the medium size firms there was an increase in instability of the large firms and an increased stability of the smaller firms. As viewed from all these different vantage points, it would seem that were the propagation of the Mainland cycle to depend only on the actions of already established firms, fluctuations would be passed on in a more magnified way than was previously the case. There is, however, no clear black-and-white trend which would make things easy for the policy maker. In the Mainland Affiliated category, firms in the newer industries have had the greatest rates of man-hour curtailments. But it is doubtful whether authorities should therefore deemphasize the promotion of firms in such industries and therefore hamper the attainment of maximum rates of growth. On the other hand, 11

146

PITERTO' RICAN COUNTERCYCLICAL POLICIES.

the high degree of stability of Mainland Affiliated Contractors `should be investigated by the policy makers for the possibility of emphasizing the promotion of such firms. Since Apparel, Chemical, Leather, Rubber, and Food products have exhibited a 'greater degree of stability, it might also be worth emphasizing the promotion of firms in those industries. It was impossible to get specific evidence of whether Mainland firms, curtail their Puerto Rican subsidiaries to a greater, 1'esser, or equal extent as their' Mainland plants. The general impression is that Puerto Ricañ plants are not discriminated `against, except in such rapidly growing, but still relatively small, industries as metal products and electrical machinery, where there is some evidence that curtailment policies in the downward phase' of the cycle is more drastic than on the Mainland. The Changing Structure of the Puerto Rican Economy and the Impact of the Mainland Cycle In the final chapter we tried to gain a more profound picture of the possible vulnerability óf the Puerto Rican economy by using information about the structure of the type of U. S. industries which have settled in Puerto Rico. First, we developed a method for estimating the import content of Puerto Rican exports. It was found that the import-content ratio rose substantially over the last ten years (from 10.8 to 19.1 %), most of the increase occurring in the latter part of this period. From the point of view of the spread of the cycle, it would seem that the higher this ratio, the smaller will be internal repercussions of changes in the value of exports, all other factors remaining equal. But all other factors did not remain equal, since in the same period the ratio of exports .to gross product increased by about 10 %. It should also be remembered that the impact of the Mainland cycle is determined considerably through the flow of investments. Secondly, we attempted to determine the effect of declining exports in the last two recessions without the presence of the growth trend factor. This was done by using U. S. input-output information in order, to predict reactions to declines of exports of ley Puerto Rican industries (see explanation of method used op pp. 106-7). In almost every case examined, the predicted declines in labor earnings due to changes in exports were greater than actuál declines. It was estimated that without the mitigating

PUERTO RICAN COUNTERCYCLICAL POLICIES

147

growth factor, fluctuations in exports would. have had an effect on labor earnings three times as powerful as was actually expe• .rienced in the last two recessions. A third approach was to observe fluctuations in exports and labor earnings when both seasonal fluctuations and the trend were eliminated. Once the growth factor was taken out, the recession appeared in the data. Results of this procedure indicate substantially large declines of exports and labor earnings when no trend is present, even larger than was indicated in the second approach above. It was also observed that declines in both exports and labor earnings increased in the second recession. It is also of interest to observe the decline in the ratio of labor earnings to export deviations from the trend for three of four industry groups examined. This would indicate that the Mainland recession has a greater impact on Puerto Rican exports than on labor earnings. The Possibilities of Puerto Rican Countercyclical Policies In the last decade the Puerto Rican authorities have not had a conscious countercyclical policy. It might be claimed that this was not necessary, since the effects of the two Mainland recessions of the 1950's on the island were very small. They manifested themselves mainly in some slackening of growth. But even if the general effect is only on growth, it is not .to be taken lightly. ,With the strong sensitivity of migration to the cycle and with an employment decline in agriculture which is larger than the employment creation in the manufacturing sector, a slackening in the rate of growth at present can have severe repercussions on unemployment. Although growth will undoubtedly continue for a long time, it will not be as dominant in the future as it is now. At that time Mainland recessions will have more pronounced effects on the island. This was shown in the projections made in the last chapter and in the analysis of the behavior of established firms over the cycle. There is a need therefore to assess the tools available to Puerto Rican authorities in counteracting deflationary forces when they hit the. island and in shaping the structure of the economy, making it as depression-proof as possible. There are three agencies which could effectively engage in countercycli cal measures: the Economic Development Administration, the Treasury, and the Government Development Bank.

148

PUERTO RICAN COUNTERCYCLICAL POLICIES

The following suggestions for 'countercyclical policies are based on our findings in this study and are very general in nature. We shall not go into an analysis of their specific implementation, for that would be the subject of another careful study. Promotion of Firms A greater effort should be made to promote firms of an ownership type and of an industry which is relatively stable over the cycle. It would seem from our analysis, for example, that Apparel firms are more stable in their policies than other Textile Mill Product enterprises. The greater stability of Mainland Affiliated Contractors should be further investigated, and if their greater stability in terms of monthly man-hours persists, it might be wise to emphasize the promotion of such firms. Since the bulk of EDA-promoted firms fall in the Mainland Affiliated category, which means that for each enterprise the Puerto Rican plant is only one of a larger number of such plants, the tax exemption status could possibly be used as an incentive for greater stability. For example, an amendment to the present law could be introduced which would state that any firm which has curtailed its operation by over 60% of capacity for six months will lose some of its tax exemption benefits for a year. Or, a provision might be introduced stating that if curtailment of the Puerto Rican plant's operations is greater than the same plants of the corporation on the Mainland, there will be a penalty in terms of tax exemption benefits. Or, again, instead of such a negative attitude, a law could be passed stating that if a plant operates at an average of 70 % capacity for over a period of five years, its tax exemption status will be. extended by one year. There could be many variations on these themes. Finally, EDA should concentrate on the promotion of Mainland firms during a Mainland boom, and concentrate more heavily on the promotion of locally-owned firms in recessions. The latter are usually oriented more toward the local market, which is relatively stable. The opportunity for promotion will thus always be there. Timing of Social Overhead Capital Projects It would be difficult to recommend a simple increase in public capital expenditures in times of recessions and decreases in times of booms. Many of the public investments, undertaken

PUERTO RICAN COUNTERCYCLICAL POLICIES

149

are directly cormected witlr plant promotions, and cannot be suspended easily at the very moment when new firms are established — e.g. construction by PRIDCO of plant facilities for new firms. But certain types of public overhead projects could be pushed in times of reduced private economic activities. For example, expansion of the power facilities of the island could be stepped up during a recession, thus creating some excess capacity for future expansion. The same is true of road construct tion, expansion of harbor facilities. In housing, with the constant need for improvements, it would be rather difficult to slacken activities in boom times. Some degree of variation, however, could be introduced in a school construction program, and the construction of civic auditoriums, ball parks, etc. Stepped-up activities in times of recessions would also provide the government with cheaper methods of financing overhead facilities, since most of the borrowing is done in Mainland financial markets where the cost. of borrowing is smaller during recessions.. Puerto Rico has special advantages over other states in the New York market because of the bigger tax base which backs its debt. Monetary Policy Up to now, with the high degree of excess reserves of local banks, any attempt at monetary policy would have been a futile operation. But the decline of excess reserves and the greater use of the banking system by the economic community have provided an opportunity for experimentation. For example, since reserve requirements for local banks are set by the Treasury, they could be lowered during a recession, thus giving greater stimulus to the expansion of local enterprises which EDA would presumably push harder during such a period. The Government Development Bank or the Treasury might go into the business of lending reserves and thus exercising controls via their rediscount rate. Finally, the government bank could also vary the terms of its loans with changes in economic conditions. There can be no doubt that in a drastic depression most of the burden of countercyclical policies will fall on the Federal government. The power of the Commonwealth Government lies only in ameliorating the effects of a Mainland recession through some of the above mentioned tools and by shaping the development of the island's economy in such a ,way as to make it stable, or as stable as possible viia -vis Mainland fluctuations.

BIBLIOGRAPHICAL NOTE: 1. Statistical information for the analysis of chapter IV was obtained from the files of the Economic Research Staff of Puerto Rico's Economic Development Administration. 2. Information on Puerto Rican Bank Rates in Table 20 was obtained through the courtesy of Juan Labadie Eurite, Assistant Secretary in change of Financial Affairs (1958), who sent me a «Report of examinations of native and foreign commercial banks supervised by the Secretary of the Treasury of the Commonwealth of Puerto Rico.» 3. Input coefficients used in some of the analysis of chapter V were obtained from Input-Output Data for 1947, US Bureau of Labor Statistics — courtesy of Harvard Economic Research ,Project.

151

BIBLIOGRAPHY BARTON, H. C., Jr.: «Industrial Development Planning in Puerto Rico», mimeographed paper written for the Economic Planning Seminar of EDA. BARTON, H. C., Jr.: «Background Data for Program Conference on Industrial Development», unpublished paper. BAER, Werner: «Puerto Rico: An Evaluation of a Successful Development program», The Quarterly Journal of Economics, November 1959. DEBIEERS, John S.: «Instruments of the Capital Market in Puerto Rico», Government Development Bank for Puerto Rico, June 1957. DIVENUTI, Biagio: Banking Growth in Puerto Rico, Puerto Rico, 1955. Economic Development Administration, Directory of Fomento Promoted and Assisted Manufacturing Plants, December 1958. Economic Development Administration, Annual Report, 1957-8. Economic Development Administration, Office of Economic Research, «Closed Plants Survey, 1947-54», San Juan, Puerto Rico, mimeographed. Economic Report to the President, 1955, 1959, United States Government Printing Office, Washington, D. C. ESTEVEs, Vernon R.: «Statement to the National Monetary. and Credit Commission», October 9, 1958 (mimeographed). Federal Reserve Bank of New York, «Puerto Rico's Monetary System and Economic Growth», Monthly Review of Credit and Business Conditions, March 1958. HARRIS, Seymour E.: International and Interregional Economics, New York, 1957. HICKMAN, Bert G.: «The Contraction of 1953-54», The Review of Economics and Statistics, ,February 1958. HIRSCHMAN, Albert O.: The Strategy of Economic Development, New Haven:. Yale University Press, 1958. 153

,.

,c

KINIDLEBERGER, Charles P.: Economic Development, New Ybrk, The McGraw Hill Book Company, Inc., 1958. MEIER, Gerald M. and BALDWIN, Robert E.: Economic Development: Theory, History, Policy, New York, John Wiley & Sons, Inc., 1957. PERLOFF, Harvey S.: Puerto Rico's Economic Future: A Study in Planned Development, Chicago: The University of Chicago Press, 1950. OSHIMA, H. T.: «Share of Government in Gross National Product», American Economic Review, June 1957. Puerto Rico, Commonwealth of, Department of Labor, Employment, Hours, and Earnings in the Manufacturing Industries in Puerto Rico, monthly series. Puerto Rico, Department of Labor and Planning Board, Puerto Rico's Manpower Needs and Supply, prepared by the Committee on Human Resources. Püerto Rico, Department of Labor, Bureau of Labor Statistics, Em ployment and Unemployment in Puerto Rico, Quarterly Reports. Puerto Rico, Junta de Planificación, Negociado de Economía y Estadísticas, Proyecciones del Desarrollo Económico de Puerto Rico, Diciembre, 1957. Púerto Rico, Planning Board, Economic Development of 'Puerto Rico 1940-50; 1951-60, Junuary 29, 1951. Puerto Rico, Planning Board, Current Business Statistics, monthly publication. Puerto Rico, Planning Board, Economic Report, to the 'Governor, yearly.; reports.. Puerto Rico, Planning Board,. Net Income and Gross.Product of Puerto Rico -1940 and 1946-1955,, and Supplements 1956-7, 1957-8. Puerto Rico,• Planning Board, Statistical Yearbook. Püerto -Rico, Department of the Treasury, Comparative, Financial Conditions of Commercial Banks in Puerto Rico, quarterly publication.. ROSENIBLUTH, G.: «Changes in. Canadian Sensitivity to United States Business •Fluctuations», The Canadian. Journal: of Economics and.d Political Science, November) 1957: SCITOVSKY, Tibor: Economic Theory. and Western Europeans Integration„ London, George Allen and Uñwin LTD', 1958. United States, Department of Commerce, Business Statistics •(A supplement to the Survey of Current Business), 1955, ,1957. United., States, Department of Commerce, Survey of Current Business. United States, Department of Cómmerce, Büreau, of the Census, US' Trade. in Merchandise, and Gold and Silver with US"Territories, and Póssessions, monthly. publication.. 15.4.

STEAD, William, H.: Fomento: The Economic Development of Puerto Rico, National Planning Association, March 1958. WA LICH, Henry C.: «Banking Facilities in Puerto Rico», report to the Government Development Bank for Puerto Rico, June 1957 (mimeographed).

155