Economic History Association

Railroads as an Economic Force in American Development Author(s): Leland H. Jenks Source: The Journal of Economic History, Vol. 4, No. 1 (May, 1944), pp. 1-20 Published by: Cambridge University Press on behalf of the Economic History Association Stable URL: http://www.jstor.org/stable/2113700 Accessed: 10/02/2010 16:45 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=cup. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Cambridge University Press and Economic History Association are collaborating with JSTOR to digitize, preserve and extend access to The Journal of Economic History.

http://www.jstor.org

THE VOL.

JOURNAL IV

OF

ECONOMIC

MAY

HISTORY

I944

NO.

I

Force Railroadsas an Economic in AmericanDevelopment' I NY attempt to discuss the way in which railroadshave promotedthe rise of the Americaneconomymust assumesome theoryof economic evolution.The followinganalysisis based upon Schumpeter'stheory of innovations.2Brieflythis theoryholds that economicevolutionin capitalistic society is startedby innovationin someproductionfunction,that is, by new combinationsof the factorsin the economicprocess.These innovationsmay center in new commoditiesor new services, new types of machinery,new forms of organization,new firms,new resources,or new areas. As Schumpeter makes clear, this is not a general theory of economic,much less of social, change. Innovation is an internal factor operatingwithin a given economicsystem while the system is also affectedby externalfactors (many of them sociological) and by growth (which means,substantially.changes in populationandin the sumtotal of savingsmadeby individualsandfirms). These sets of factorsinteractin economicchange."The changesin the economic processbroughtabout by innovation,togetherwith all their effects, and the response to them by the economic system" constitute economic evolution for Schumpeter.' A

' This article is an elaboration and extension of a paper delivered at the meeting of the Mississippi Valley Historical Association, Washington, D. C., December 28-3I, I938. 'Joseph A. Schumpeter, Business Cycles (New York and London: McGraw-Hill Book Company, I939), Vol. I, esp. chaps. iii and vii; idem, The Theory of Economic Development (Cambridge: Harvard University Press, I934), chaps. ii and vi; idem, "The Instability of Capitalism," The Economic Journal, XXXVIII (I928), 36i-86. Cf. the theory of Allyn A. Young, "Increasing Returns and Economic Progress," ibid., 527-42. 'Business Cycles, I, 86. I

2

LelandH. Jenks

Railroad development has had three phases or moments which have involved innovation in distinctive ways. I shall consider (i) the railroad as an idea, (2) the railroad as a construction enterprise, and (3) the railroad as a producer of transportation services.' II By the railroad as an idea is not meant the original design of steam locomotion on rails. It pertains to the inception in particular areas of particular projects, conceived as likely to be appropriate opportunities for business enterprise. In this sense the idea of any major innovation, such as the railroad, is a potent economic force. For once railway projects have been conceived and plans for their execution elaborated, it becomes easier for other innovating ideas to be entertained.' On the one hand, the socio-psychological deterrents against entering upon new ways are lowered. On the other, the characteristics of the prospective future are altered; they assume an aspect more favorable to men and firms with new plans than to men and firms whose position is established. Thus early railway projects were attended by a retinue of satellite innovations. The first railway projects emerged in the United States in the thirties in a situation in which the psychological risks had already been appreciably lowered by the general passion for internal improvements displayed in a plethora of projects for canals, turnpikes, plank roads, bridges, banks, and other enterprises.' The earliest railways paralleled, supplemented, or improved transport systems that were already in being.7The real railway revolution dates from the forties, prior to the California gold discoveries, in projects to cross the Appalachians, to link the seaboard with the interior, the Ohio Valley with the Great Lakes, and, breaking away from the contours of 'These distinctions are hinted at but not developed in Business Cycles, I, I30-36. They are not to be construed precisely as stages or periods, although each was relatively more conspicuous in certain decades than in others. 'Three types of obstacles to innovation are distinguished in Business Cycles, I, ioo: hostility to the new idea, absence of facilitating economic functions, and inhibitions against entering upon a relatively incalculable course. Young in The Economic Journal, XXXVIII (1928)X 534, stresses the need to remake human material in terms of new skills and habits and in terms of redistribution of population. 'Carl Russell Fish, The Rise of the Common Man (New York: The Macmillan Company, I927), chaps. iv and v. 'One thinks of the Boston & Lowell, New York & New Haven, Philadelphia & Columbia, Allegheny Portage, the original Baltimore & Ohio, and the lines connecting Albany with Buffalo.

Railroadsas an EconomicForce

3

water transport,to unite distantpoints by moredirectroutes.8It was the determinationto build railroadsin advanceof trafficthat gave the "railroad idea" prolongedforce in Americaneconomiclife. The conviction that the railroadwouldrun anywhereat a profitput fresh spursto Americaningenuity and openedclosedpaddocksof potentialenterprise. Innovationsare the workof enterprisers.For the railroadas idea, the role of entrepreneurshipwas pretty much identical with promotion; and the promoterwas rarelylimited in outlook to the railroaditself. In action, he was omnicompetentand omnipresent.His imaginationleaped readily from the concreteproblemof securingauthorityfor a rightof way to visions of a countrysidefilled with noddinggrain, settlements of industriousfamilies, and other evidencesof progressand civilization.Each railway project involved the sanguinejudgmentof enterprisingindividualsandgroupsin particular, local situations that a certain line would be of direct or indirect pecuniaryadvantageto themselves.It was linked to specificplans for town promotionand real-estatespeculation,to combinationsfor contractingservices and supplies or for exploitation of resources,in anticipation of the actual movementof trafficby rail. But as projectsmultipliedthey collectively acquireda symbolic function, dramatizingbroaderpurposes.The railwayprojectorbecamean exemplificationof the power of steam, of the advantagesof the corporateformof businessorganization,of the ability of man to masterhis environment.The early railwaypromoterwas not only a potential economicagent; he embodiedthe dream of developingcommunities, regions,the continent. Thus, as the barriersto new projectswereperiodicallyloweredby the inception of new railwaysystems, the first momentof the railroadas an economic force was manifestedin a wavelike profusionof new enterprisesof many sorts. Moreover,its effectsin the United States were not exhaustedin a decadeor so, as they were in England.The railroadidea was periodically 'The most dynamic set of American innovations consisted in plans to build railways in anticipation of traffic. Lewis Henry Haney, A Congressional History of Railways in the United States to 1850 (Madison: University of Wisconsin, i908), p. 3I. Congressional land grants were a factor, as in the case of the Illinois Central, the first large system built through sparsely settled territory. Paul Wallace Gates, The Illinois Central Rail-road and Its Colonization Work (Cambridge: Harvard University Press, I934). Canal building had, however, in the old Northwest, anticipated the railroad less successfully in building ahead of population. Frederic L. Paxson, History of the American Frontier 1763-i893 (Boston and New York: Houghton Mifflin Company, I924), chap. xxx. For early systems and projects, cf. Caroline E. MacGill et al., Balthasar Henry Meyer, editor, History of Transportation in the United States Before z86o (Washington: Carnegie Institution of Washington, 1917); J. L. Ringwalt, Development of Transportation Systems in the United States (Philadelphia: The Author, i888).

4

LelandH. Jenks

renewedfor regionafter regionand route after route, as national development, at least facilitatedby the earlierrailroads,widened the horizonsof enterprise. III The second momentof the railroadas an economicforce came with the actual constructionof new lines. The statistics of net mileageaddedin each year from i837 to I937 give a quantitativemeasureof this contributionof the railroadto development,as appearson the accompanyingcharts. Two generalstatementsare strikinglysupportedby these data.9In the firstplace, railwaybuildingproceededin an undulatingpattern,parallelingclosely the generalcontoursof majorbusinesscycles until the First WorldWar. From i850 to the nineties,omittingthe years of the Civil War,the rise and fall in new constructionin fact led by a perceptibleintervalmost other indices of businessconditions.'" In the secondplace, therewas a long-runtrendin new railwayconstruction,which was predominantlyupwardin absolutefigures from the late i840's to about i890. The rate of this upward trend tended to

slackenwith the aggregatemovementapproximatinggraphicallya logistic curve,but, for the wholeperiod,expansionof railwayplant averagedabout io per cent a year. The trend since I890 has been irregularly downward,

bearingthe aspect of a reversedlogistic curve.The early persistentsuccession of freshwaves of railwayconstruction,arisinglargely in the developmentof new areasin the AmericanWest and South,mustbe regardedas one of the basic phenomenain the total economicgrowthof the United States, while the logisticcurveof total experiencepresentsin outlinea pictureof an industrypassing from youth throughadolescenceto maturity. But how did railwayconstructionas such act as an economicforce?How couldit be a pace setter? The answeris broadlythat it operateddirectly to create a demandfor variousfactors of production.In responseto this de9The data for these charts are derived from the United States Treasury Department, Bureau of Statistics, Statistical Abstract of the United States, 1goo (Washington: United States Government Printing Office, igoI); ibid., 1914, p. 637; and ibid., 1937, p. 37g. Chart II is adapted from Simon S. Kuznets, Secular Moventents in Production and Prices (Boston and New York: HoughtonMifflinCompany,I930), pp. I9I, 526-27. 10 This correlation was initially based upon inspection of the mileage data in comparison with the chart in Schumpeter, Business Cycles, II, 465, and the analyses of business conditions in Willard Long Thorp, Business Annals (New York: National Bureau of Economic Research, 1926) and National Bureau of Economic Research, Recent Economic Changes (New York: McGraw-Hill Book Company, I929), II, 892. More decisive support is provided by John E. Partington, Railroad Purchasing and the Business Cycle (Washington: The Brookings Institution, I929). As Partington includes orders for replacements as well as for original basic construction, he finds that orders of railway capital goods led business-cycle changes as late as i907. Throughout this period, he finds, railway earnings followed, instead of preceded, changes in purchases.

5

Railroadsas an EconomicForce

mand there were rises in prices or increasesin supply or both. Increaseof supply could come only from some sort of furtherinnovations,such as the drawingof freshincrementsof land, labor,or capital into economicuses or the transferof such factors to more effective combinations.This process 400,000

I00,OO(

CHART

-

-

-

MILES OF RAILROAD IN OPERATION 1837-1937

1867

1847

1907

1887

1927

CHART It 10,000 CHART IM 1000 500

1000 CV0k_1

-500 -1000 -1500 -2000

100

-2500

NET ANNUAL CHANGE IN U.S. RAILROAD MILEAGE 1917-1937 1919

1923

1927

1931

1935

NET ANNUAL CHANGE IN U.S. RAILROAD MILEAGE 1831-1918 L840

1860

1880

1900

1920

meant the periodicdislocationof the economicstructureas well as the disruptionof the activitiesof individualsandcommunities.At the same time it enhancedthe opportunitiesfor enterprisershaving a high degree of flexibility, pioneeringindividualsand groups,the agents of innumerableinnovating firmsand procedures. The land for railroadconstructionwas largelynewland,previouslynot of economicuse. It cost virtually nothing to the railway companies,and not

6

LelandH. Jenks

very much to anyoneelse.11Socially the land devoted to railroadpurposes morethanpaid for itself by the incrementin productivityof adjacentland. This was so obvious to everyone connected with railway building that periodicland booms came to communitieseven before the rails were laid. The speculativeactivity thus diffusedin anticipationof railroadconstruction may have broughtmany creativeinnovationsin its wake. But, by distractinglaborand enterprisefromproductiveto parasiticactivities, it frequently delayedthe realizationof the plausiblehopes upon which railroad projectswereprimarilybased. The demandfor labor initiated a chapterin the history of immigration and colonization.' It also disciplinedmigratoryand local labor power to co-operativeindustrialeffort.But it had widerrepercussions.Laborerswere paid wagesand the wageswerespent for goods.They went to marketto buy the produceof Americanfarmsand mills.Thus the demandfor laborstimulated the spreadof marketeconomyand the more extensive productionof goods and services for distant markets, and thereby contributed to the spreadof economicspecialization. The demandfor capital functionedin parallel to the demandfor labor. I am speakingof realcapital,of goods,of the picks and shovels,sleepersand steel rails, enginesand rollingstock and bridgeworkand culvertsand ordinary buildingmaterial,whichmakeup the physicalplant of a railroad.The constructionmomentof railwayhistorybroughtan initial demandfor these durablegoods.'3Hencetherewas a chancefor the innovatorin the lumbering industry, in quarries,in iron mills and carriageworks. Indeed these in" Frederick A. Cleveland and Fred Wilbur Powell, Railroad Promotion and Capitalization in the United States (New York: Longmans, Green and Company, i909), pp. 199-200. "In the Southern States, and the Mississippi Valley .... all the real estate required for way, and for depots, stations, etc., are generally gratuity to the roads." American Railroad Journal, XXV (January 3, i852), I3. Cf. James Blaine Hedges, Henry Villard and the Railways of the Northwest (New Haven: Yale University Press, I930), passim. 12 Gates, The Illinois Central Rail-road, pp. 89, 94-8. Despite its crucial importance, the subject of labor supply has been too frequently neglected by railway historians. Adequate data for labor employed in new construction are available only for a few large lines such as the Central Pacific, Union Pacific, and the Illinois Central. On each of these, upwards of i0,000 men were employed at the peak of construction. Probably a thousand men were needed for every hundred miles. Assuming that twice as many miles were in progress as were completed in any given year, the figure of 200,000 men is reached as the maximum employed at any one time in the construction of these railways. This figure was not attained until the eighties, by which time the census reported 250,000 officials and employees of railroads, presumably engaged directly or indirectly in transportation service. 13 Cf. files of railway periodicals for advertisements of manufacturers and dealers in railway materials and supplies. Ringwalt, Development of Transportation Systems in the U. S.,

pp. I32-36,

2I0.

Railroadsas an EconomicForce

7

dustries were hard put to keep pace with railway construction.Until the later eighties, every boom period foundAmericanfactoriesunable to meet the demandfor rails,and therewereheavy importationsfromEnglandand Wales.As late as the nineties,over one fifth of the total outputof pig iron in the United States was beingrolledinto railroadbars." Much of this demandfor durablegoods turnedeventuallyinto a demand for labor in mine and quarryand mill, into wage payments to labor. And these wages too were spent for consumers'goods and meant wideningmarkets, increasedspecialization,and, presumably,greaterproductivity. Thus the initial impetus of investment in railway constructionled in wideningarcs to incrementsof economicactivity over the entireAmerican domain,far exceedingin theirtotal volumethe originalinputsof investment capital. To this feature of moderncapitalism,John MaynardKeynes and othershave appliedthe term "multiplier."'It is believed that for presentday Englandthe efficiencyof the multipliermay sufficeto doublethe impact of a new investmentin construction.For nineteenth-centuryUnited States, its efficiencyseems to have been considerablygreaterthan that. I have spokenof inputs and investment.In our economythe demandfor land and laborand capitalhas meantanotherdemand,a demandnot for an independentfactorof production,but for somethingequallyessential,a demand for moneycapital.'6In fact, without a supply of money capital there couldhave beenno effectivedemandfor any of the real factors,no railways, and no stimulus from them for economic development.Hence it is convenient to think of the building of railroadsas an investment of money capital.To this investmenttherecorrespondedin the long runthe accumulation of savings.That savingcamefirstand investmentin the railroadsafterwardsis a propositionfor whichthereis little historicalevidence,at least in the United States. It is true that the practiceof thrift as an individualand family responsibilitywas built into our social system by the Puritans.But the savings thus madein the middleof the nineteenthcenturywent largely into land,into improvementson the farm,into the mill, the privatebusiness, 1 For details, cf. Statistical Abstract of the U. S., 1902, p. 38o, and corresponding tables in earlier volumes. 16 John Maynard Keynes, The General Theory of Employment, Interest and Money (London, 1936), chap. xi; R. F. Kahn, "The Relation of Home Investment to Unemployment," The Economic Journal, XLI (i93i), i73-98. ' Admittedly "money capital" constitutes merely a vehicle or instrumentality, the means of acquiring command over 'the several factors of production. More commonly it is spoken of as long-term credit or capital funds. But sometimes an instrument becomes so important that it exerts influences by itself and requires consideration on its separate account.

8

LelandH. Jenks

and, in relativelysmall amounts,into public securities.Few railroadswere originallyfinancedby directsubscriptionof the shareholdersat parin ready cash.'

In finalanalysis,the funds for railwayconstructioncame fromthe extension of credit by Americanbanks and from foreign exchangesupplied by Europeaninvestors.This was accomplishedby many devices which called into play the charitablecupidity of contractorsand iron manufacturerson both sides of the Atlantic,and the lively anticipationsof propertyownersin the areawhichthe railroadwas to develop. Someof the sharesweresold at a heavy discountto local residents,but more were given outright for land, for legal and legislativeservices,for bankingaccommodation,or as a bonus to promotethe sale of bonds.Frequentlytherewas a constructioncompany, analogousto the CreditMobilier,which took all the securitiesin payment for the roadand operatedit pendingthe completionof construction.Since the books of these organizationshave been conveniently mislaid, it will always be impossibleto ascertainwhat our railroadsreally cost originally in money capital.The constructioncompaniesturnedover whole blocks of securitiesto manufacturersand contractorsin paymentfor goods and services. These enterprisersusually seem to have pledged the securities with banks for workingcapital in the process of supplying the goods. In New Yorkandelsewhere,speculatorsandspecialistsin railwayfinance,operating also on bankloans, facilitatedthis inflationaryprocessby their dealingsin stocks and bonds and daily risked the credit of the railway companiesin their furiouscontestsof bulls and bears. The Americanbanking mechanismdid not have to bear this periodic strain alone.Every burstof new railwayconstruction,in the thirties,in the fifties, at the close of the Civil War, throughthe eighties, and again from I904 to I907, meant new investmentsfrom abroadby British, Dutch, and Germancapitalists.'9Schumpeterstates that the boom from i866 to i873, 17 These were chiefly railroads built in the thirties and forties. Cf. Frank Walker Stevens, The Beginnings of the New York Central Railroad (New York and London: G. P. Putnam's Sons, I926). Even in these cases, as we know from accounts of the crises of 1854 and i857, the subscribers carried their shares on bank loans. Cf. Schumpeter, Business Cycles, I, 325-30. 18 Cleveland and Powell, Railroad Promotion and Capitalization, is still the most adequate account for aspects before i900. Cf. William Z. Ripley, Railroads; Finance and Organization (New York: Longmans, Green and Company, I915), pp. 10-52; Cleveland and Powell, Railroad Finance (New York: D. Appleton and Company, i9i2), chaps. ii-iv and the very rich bibliography; Charles F. Adams, Jr., "Railroad Inflation," North American Review, CVIII

(i869),

138-44.

This paragraph is based upon original research in London and the United States, made possible by a sabbatical from Wellesley College and a grant from the John Simon Guggenheim Memorial Foundation. An introduction to the subject is available in Cleona Lewis, America's 19

Railroadsas an EconomicForce

9

which doubledour railwaymileage,was entirely financedby an estimated two billion dollars of capital importedduringthose years.2 It is incorrect to suppose,as he apparentlydoes, that any such amountof foreignmoney was at that time investeddirectlyin the railways.British, Dutch, and Germaninvestorswerethen buyingnearlyhalf of the Civil War debt, chieflyin 5-20'S and 10-40's, to the amount of more than a billion dollars par. The railroadsobtaineddirectly only about half a billion. The purchaseof governmentbondsby foreigners,however,releasedsavingsand bank resources for railway,industrial,and commercialpromotionin the United States. In no subsequentperiodwas the impactof foreigncapital as momentous;but it is easy to exaggerateits importance.Althoughsomethinglike one fifth of the nominal value of Americanrailroadswas foreign-ownedin i873, the wholevolumeof foreignclaimsamountedto only 6 or 7 per cent of national wealth.' Whilein the courseof subsequentfluctuationsforeignownershipof railroadsecuritiesmay have reachedthe proportionsof one third in i890 and nearlyas muchjust before I9I4, yet at these later datesit constituteda smallerproportionof the total nationalwealth than it had in i873. According to the estimates,foreigninvestmentsdid not keep pace with the growth of the nationalwealth. It wouldbe desirableto measuremorepreciselythe investmentof money capital at successiveperiods.Availablefiguresof railwaycapitalizationare entirely unsatisfactoryfor historicalpurposes.Apart from the obscurities of early railroadfinancealready mentioned,tabulationsand estimates do not carefully and regularlyinclude net floating debt or exclude intercorporate securities.The pathologyof early stock wateringhas no necessary connectionwith the "overcapitalization" fromwhichmost railroadsystems have sufferedin recentyears. This overcapitalizationis entirelycompatible with real historicalinvestmentas large as the nominalcapitalization.But the availablestatistics give no adequateclue, before the last few decades, whensuch amountsactuallywereinvested. Whateverthe source or timing of the applicationof money capital, the financingof railroadconstructionencouragedinnovationsin financialenterprise: the developmentof stock exchangesand their techniques; the speStake in International Investments (Washington: The Brookings Institution, I938), chap. ii; Ripley, Railroads; Finance and Organization, pp. i-iO; and Leland H. Jenks, The Migration of British Capital to i875 (New York and London: Alfred A. Knopf, I927), chap. iii and pp. i69, 255-59 and notes. Before the Civil War the share of foreign investors was smaller than it became later. In only a few cases was it an initiating factor in railroad development. ' Schumpeter, Business Cycles, I, 335. 2' Lewis, America's Stake in International Investments, p. 560.

Io

LelandH. Jenks

cialization of firms, old and new, in investment banking and in security brokerage; the specialization of banking institutions (especially trust companies) as trustees and registration agents for securities, and as agents for distributing capital and interest payments; the rise of legal firms specializing in corporation law and in adjusting construction activities to the intricacies of the American political system. New financial techniques and innovations in corporate structure were involved when established railway companies became agents in the flow of capital. By the early fifties the Pennsylvania was using its credit to supply funds for the building of western connections which it only informally controlled.22With the establishment of the Pennsylvania Company in i869, the holding company became a permanent feature of the American scene. In many cases initial construction was of the sketchiest sort and by the seventies it was an established practice, of which foreign security holders bitterly complained, for companies to invest their earnings in necessary improvements and extensions. This financing of corporate growth from within may fairly be claimed to be an American innovation in capitalistic technique, which has only recently been diffused to the British Isles. With financial innovation came a transformation of the role of the enterpriser in connection with particular railway systems. In the initial moments of construction, the typical enterpriser was still pretty much the omnicompetent pioneer, the individual of imagination, daring, and energy. Like General W. J. Palmer of the Denver and Rio Grande, he considered himself an agent of civilization, an embodiment of collective purpose.' No aspect of the task of railway building was too technical for his consideration and none too petty. In looking for the enterpriser of particular lines, official titles should not deceive. There was usually one man or a small informal group of unspecialized associates who could get things done, who could deal effectively at the same time with laborers, suppliers, politicians, and the local citizenry, and could command the confidence of sources of credit. At the construction moment, administration of a large formal organization was not necessarily involved. The mechanism of subcontracting provided a pattern for the cooperation of innumerable lesser enterprisers of a similar type. Such enterprisers were rarely able, however, to cope with recurrent finan22

Pennsylvania Central R. R. Co., Annual Reports, passim.

'3 William J. Palmer, The Westward Current of Population in the United States (London,

i874) and Glenn Chesney Quiett, They Built the West (New York and London: D. AppletonCentury Company, 1934), chaps. ii-vi, throw light upon the career of this neglected enterpriser.

Railroadsas an EconomicForce

II

cial involvements.The elaborationof the superstructureof railroadsecurities sooneror later compelleda moreformaldivisionof tasks and responsibilities in the continuanceof construction.In somecasesthis involveda shift of the center of decisionfrom the engineer-promoterto financialand legal experts either within or outside the railroadorganization."The financierenterpriserassumedmany guises, now entering upon new constructionto win stock-exchangebattles, now basing a programof calculatedexpansion upona re-orderingof companyaccounts,nowenteringbelatedly,as did William Rockefellerin Northwestern,the race for competitivebigness.' There was inescapablya narrowingof horizon; the financier-enterpriser could decide freelyonly problemsstated in financialterms,andhe focusedhis attention chiefly on relations with potential intermediariesand rivals for the supplyof capital. Thus the secondmomentof the railroadas an economicforce came with a demandfor the factorsof productionin newconstruction,accompaniedby the rise of new techniquesand institutionsof finance,by the aggregationof capital in mobile forms,and by the gradualdisplacementof the omnicompetent type of enterpriser. IV The thirdmomentto be surveyedis that of the railroadas a goingconcern, a complexof tracksand enginesand cars and managersand employeesengagedin the businessof carryingpassengersand freight. By renderingthis transportationservice,the railroadin operationhas doubtlessaddeddirectly to the real incomeof the United States, and indirectly to economicexpansion.26There appearsto be no satisfactory techniquefor giving a precise 4 N. S. B. Gras, Business and Capitalism (New York: F. S. Crofts and Company, I939), pp. 246-59, 272-75, indicates the "normal" process by which financial capitalists became involved in industry. He is correct, I believe, in implying that the opportunity and need have not been confined to late phases of the construction moment. From the standpoint of innovation, the emergence of the financial enterpriser in the railroads is not to be identified with the rise of special departments within the organization. The latter, or their heads, may be simply parts of a formally established group functioning as management-enterpriser. See section IV below. 2 Max Lowenthal, The Investor Pays (New York: Alfred A. Knopf, I933). Ringwalt, Development of Transportation Systems in the U. S., pp. 382-85 and Henry V. Poor, Influence of the Railroads of the U. S. in the Creation of its Commerce and Wealth (New York, i869) are representative of early discussions. "Our new railroads increase the value of farms and open new markets for their products. They lessen the time and cost of travel. They give a value to commodities otherwise almost worthless. They concentrate population, stimulate production, and raise wages by making labor more efficient. Our existing railroads are computed to create more wealth every year than is absorbed for the construction of new railroads." Commercial and Financial Chronicle, XVI (January II, i873), 4I.

LelandH. Jenks

12

measureto the extentof this contribution.It seemsthat the railwayscarried irregularlyincreasington-milesof freightuntil I929, while the aggregateof passenger-milesexpandeduntil I920. The quantainvolved,said to be from 13 billionsof freightin i870 to 450 billionsin I929, are certainlyenormous.' But the availablefigures,at least beforei890, are neitheraccuratenor complete. There have been importantchangesin the compositionof traffic.As Pigoupointsout, any attemptto measuredifferencesin realincomebetween situationsinvolvingsubstantialvariationsin the use of productivefactors and in the compositionof demandis theoreticallyat least precarious.'For contemporarycomparison,Holmstrom has worked out a technique by which "virtualcosts" (operatingand maintenancechargesplus interest on replacementcost of ways and works plus depreciationand profits) are equatedwith "directbenefits"on the one hand and "consumercosts" plus publicsubsidieson the other.2'In view of the defectivecharacterof the data and the violenceof pricefluctuationsin the UnitedStates,thereis little hope of applyingthese meansof measurementto the historicalproblem. It is commonlyassumed that the great contributionof railroad transportation came from the reductionof shipping costs. As comparedwith pre-motorizedformsof highway transportation,the advantageof the railroad has always been obvious. There is no convincingevidence,however, that railwayshave evercarriedfreightat lowercosts either to shippersor to society than canals or waterways.' The advantages that early railways 7 Attempts to use railway data in connection with the study of changes in real income and "productivity" are exemplified by Arthur F. Burns, Production Trends in the United States since i870 (New York: National Bureau of Economic Research, I934) and Spurgeon Bell, Productivity, Wages, and National Income (Washington: The Brookings Institution, I940). A brief factual summary of the role of the railways in the economic system after the First World War is provided by the Bureau of Railway Economics, The Railways and Economic Progress (Miscellaneous Series No. 50, Washington, I929). The theory there suggested that the "economic contribution" of the railways is measured by the volume of their expenditures of all kinds is, however, at variance with the premises of this paper. Incidentally, this is an unusual place to find a theory popularly associated with New Deal economics. On railroad expenditures, cf. Partington, Railroad Purchasing and the Business Cycle. 2A. C. Pigou, "Comparisons of Real Income," Economica, New Series, X (May, I943), pp. 93-8. 9 J. Edwin Holmstrom, Railways and Roads in Pioneer Development Overseas (London: P. S. King and Son, I934), chap. i. Cf. E. A. J. Johnson, "New Tools for the Economic Historian," The Tasks of Economic History, supplemental issue of The Journal of Economic History, December, 194I, pp. 30-8. 30General treatments of the economic significance of improved transportation are also found in D. Philip Locklin, Economics of Transportation (Chicago: Business Publications, I938), chap. i, and Cleveland and Powell, Railroad Finance, chap. i. On comparative costs of service, cf. MacGill, History of Transportation in the U. S. before i86o, pp. 574-82; Haney, Congressional History of Railways in the U. S., chap. iii; Charles H. Ambler, A History of Transportation in the Ohio Valley (Glendale, California: The Arthur H. Clark Company, I932), pp. 358 ff.; Harold Kelso, "Waterways versus Railways," The American Economic

Review,XXXI

(I94I),

537-44.

Railroadsas an EconomicForce showedover canals,such as speed, flexibility of service, and special adaptability to short hauls, are analogousto those of modernhighwaytransport over the railroad.It was far moreimportantthat the railroadbroughttransportationto areas that without it could have had scarcelyany commercial existence at all. At a later epoch, the motor highway provides means to achieve this result, at least in British colonialareas, at lower initial social cost. But historically,the very existenceof mostAmericancommunitiesand regions,of particularfarmsand industrialfirmsand aggregates,was made possibleby the railroad. Holmstrom'sstudy of the cost characteristicsof various forms of transportationbringsother considerationsto the forefrontof analysis.He shows that the trafficpotentialof the railroadper unit of installationis even now far greaterthan that of any other formof transportationthat he considers. For colonialareasin the early I930's, for example,he computesthat human porterscould carry a maximumof I,450 ton-miles of freight per annum; heavy animals, 3,600; "horsed wagons," ii8,800; tractor trains, I,000,000; and broad-guage railways, 3,6I3,500." Thus an initial and continuing poten-

tial contributionof the railroadhas come from the volume of trafficit has beenable to carry. The converseof this propositionis the fact that the railroadconstitutesa case of increasingreturn,with specialfeaturesthat give a decisivebent to its impact uponeconomicstructure.Its social costs per unit of trafficdecrease rapidlywith trafficdensity."2 A familiarmanifestationof this conditionwas the well-knownshift frompassengersandlight trafficas principalsourcesof revenuein the early railroaddays to bulk traffic.Any isolated railroadsystem would tend to expand along those lines. But as new railroadsin the United States becamelinked to previouslyexistinglines, and as the innovation of freight-carinterchangewas establishedafter the Civil War, a principle of accelerationwas manifestedenablingnewerlines to begin farther alongthe cost curve.Betweeni890 and I94I the averageactualhaul of each ton of freight became 50 per cent longer (increasingespecially duringthe First WorldWar and the 1930's); there was an increaseof more than ioo per cent duringthe sameperiodin the distancetraveledby the averagepassenger.These are revealingdata about the long-runfunctionof the railroad in the economicsystem.' Such expansionis, however,not a measureof in81 Holmstrom, Railways and Roads in Pioneer Development Overseas, p. 56. Palmer, The Westward Current of Population in the U. S., relates that in i866 the stage line from the terminus of the Kansas Pacific in Topeka carried six passengers daily to Denver. Two years later, daily trains carried westward one hundred to five hundred passengers daily. 32Holmstrom, pp. I04-12. 5 United States Interstate Commerce Commission, Statistics of Railways in the United States, I941 (Washington: United States Government Printing Office, 1943), pp. I59-60.

LelandH. Jenks

14

novation; the recent increase reflects to no small degree adjustments by railroads to other innovations in the economic system. What is significant about the principle of increasing return in the railroad is that it indicates directions in which railway transportation affects the economic structure. That the railroad tends to attract factors of production to its right of way needs no comment; this perception lay at the heart of the American railroad innovation. As Holmstrom points out, however, this supply of potential traffic does not distribute itself at random. It is polarized first about line terminals, and secondarily about traffic intersections."4There is a further tendency. Irrespective of rate differentials, the service of the railroad is of greatest advantage to large shippers requiring a fairly regular flow of traffic.' Thus railroad transportation provides a considerable addition to the external economies that firms can realize from large-scale operations. Such phenomena as the ecological structure of wholesale trade, the localization and concentration of primary processing establishments, and the vertical integration of production units in spite of their geographical separation are thus functionally related to railroad transportation service. In more concrete terms, attention may be directed to the initial localization of the textile industry in New England, the development of the factory system in some other industries at points remote from water power and dependent upon rail supply of coal, the establishment of stockyards in Chicago and other terminals, the rise of assembly plants, and generally the concentration, at terminals convenient to the source of supply, of industries processing and reducing the bulk of raw materials. In all these respects, railway transportation has worked in the same direction as, but in different areas from, water transport. It has functioned differently from the realized and probable tendencies of highway traffic. The organization of railway enterprise itself early displayed the same tendencies to differentiation that it encouraged in other industries. On the one hand, the railways transferred to other enterprises part of their business. First in individual railway lines, and gradually on a more national scale, came the innovation of express companies, specializing in the rapid transmission of small items of high value. Opportunity arose for Pullman and other specialists in high-cost passenger service. On the other hand, individual railways themselves engaged in other business activities. If their land departments developed in order to implement construction, they proved of s Holmstrom, pp. 265-66, 2 73. 3

Ibid.,pp.

2 7I-72.

Railroadsas an EconomicForce

I

5

more value in augmentingtrafficdensity to remunerativelevels. Reading and othercompaniesacquiredanthracitefieldsin the interestof controlling the supply of bulk trafficbetween terminals.A great deal of changein the internalstructureof railway organizationswas merely a functionof their expansion,involvinginnovationsof a highly derivativeand adaptivecharacter; but other changesinvolved the positive quest of increasingreturn. The extensionof particularsystemsby purchase,lease,and contractdid not invariablycontemplatedevelopment,but often aimedat controllingfor the benefit of originalmain lines the supply of trafficat terminalpoints. The consolidationmovementand much resistanceto it on the part of particular companiesmay be interpretedfrom this point of view. It must be clearthat to yield realincomeand participatein expansionare not the same as to be a force for economicdevelopment.On the economic structure,the impactof the railwayas a goingconcernwas most decisivein the early years of the expansionof each system and in many respectscame from the networkas a whole ratherthan from any particularpart. In time many other forcesreinforcedthe polarizingtendencyof the railroad.Urban centers tended to generateconditionsthat made for their own growthinto metropolises.The returnsto railwaysfromincreasingdensity tendedto increaseat slackeningrates.Changein the railwaysgraduallybecamemorea matterof adjustmentto externalinnovationsthan a primarysourceof disturbanceto the economicstructure. As early as the eighties, railway systems that had been daringventures only a decade before found themselvesembarkingon extensions and improvements,not as acts of innovatingfaith, but to enable them to handle trafficthat had been offeredthemor to keep somebodyelse fromgetting the business." In region after region development initiated by the railroad out-

ran the plans of the projectors.The businessof the railroadcame increasingly to consist not in starting somethingbut in keeping pace with what othersweredoing.That the railwaywouldcarryfreightat knownrates and with gradualchangein the quality of servicecame to be part of the normal expectationsof every businessfirm,a stable part of an environmentwhich, of course,might still be disturbedby other innovations.37 While the real in' For instance, new financing was sought by the Grand Trunk of Canada in the seventies and the Norfolk & Western in the eighties to make it possible to handle traffic already being offered. It was not always an extension that was involved but more often double-tracks, sidings, rolling stock, and improvements in the right of way. 3 Schumpeter, Business Cycles, I, chap. ii, presents a representative theoretical analysis of this "equilibrium" position to which railway enterprises have been approximating.

i

6

LelandH. Jenks

comeaccruingto society fromrailwaytransportationprobablycontinuedto growuntil I929, the railroadfunctioneddecreasinglyas a pace setter or as an inciting forcein the expansionof whichit was a part. By the time of the financialreorganizationsof the nineties,many American railwaysmanifestedsigns of belongingto an industrythat has reached maturity.' The signs became more widespreadin the first decade of the presentcenturywith the completionof the last clusterof new systems. For enterprisesin general,Oxenfeldtthinks"newnessof economicconsequence" can be assumedto have workeditself out within a year of establishment." This seemstoo shorta periodfor the railroad.Althoughthe bulk of improvement in the early years of Americanrailwaysystems is properlyclassed as "construction,"the leverageof increasingreturnin this field involves such extensiverelocationof productiveforces that opportunityfor major business decisionsmay recurforseveralyearsafter"completion"of the system.' That some innovationshave been made by railroadssince i9io must be conceded.Both technologicalandorganizationalchangesare involvedin the recentrapidincreasein ton-milesof freighthandledper employeeand per unit of capital, in the increasedcapacity of cars, in speed of train units, in locomotiveefficiency,etc. The National ResourcesPlanning Board, however, takes the view that potentialitiesin this directionare thus far morean idea than an actuality." Consolidationlooms as the sourceof the most importantinnovationsin the near future.In 1933 only i6 per cent of the time of a typical freightcar fromshipperto consigneewas consumedin hauling; 37 per cent of the time was attributableto railroadterminalmovement;and a total of 84 per cent was spent in terminals."Co-operationamong carrierscould improve this condition,but changesof innovationalconsequenceseem to wait upon governmentaction. But what has been the role of the entrepreneurin the railroadas a going concern?What is the source of innovationin an enterprisealmost wholly E. G. Campbell, The Reorganization of the American Railroad System, 1893-x900 (New York: Columbia University Press, 1938). 3 Alfred R. Oxenfeldt, New Firms and Free Enterprise (Washington: American Council on Public Affairs, 1943), p. 75. ?The degree to which in recent decades public regulation has restricted this opportunity as far as pricing of services is concerned has been the subject of a suggestive inquiry by the National Resources Planning Board. Transportation and National Policy (Washington: United States Government Printing Office, 1942), esp. pp. 87-I28. Ibid., pp. 60-5. 2 Ibid., p. 4I.

Railroadsas an EconomicForce

I

7

concernedwith renderingtransportationservice?The rise of a line organizationwith few staff featureswas an earlyaspectof railwayoperations,and was well establishedby the eighties. The PennsylvaniaCentralseems to have led the way in the practiceof promotionfrom within, a practice that developedrapidlyinto senioritypolicies at all levels and the establishment of railroadingas a career.For a coupleof decadesafter the Civil War, the training thus affordedmade the Pennsylvaniaan importantsource from which new companiesdrew top executiveswho often developedentrepreneurialtalents as individuals.ThomasA. Scott, who rose fromthe ranks to the presidencyof Pennsylvania,was of pioneeringquality. As horizonsof opportunitynarrowed,however,selectionfromwithin tendedto bringcompetent administratorsof a moreroutinesort to top executivepositions,men whohadspentso manyyearsmasteringthe complexitiesof detailedmanagement along establishedlines that they had little interest in changingthose procedures.This tendencyhas been markedin many railroadsystems, and is associatedwith the shift to adaptivechangeas the principalrelationof the railroadsto economicexpansionin recentyears. Nevertheless,someinnovationhas taken place,and it can occasionallybe tracedto pioneeringleadership.Largeorganizationsas such,however,apart from theirdegreeof maturity,set up certainhazardsto innovation.To continue operationsthey requirethe delegation of specialized authority and responsibilityto a considerablenumberof individuals.An innovationdisturbstheir tasks and theirrelationswith each otherquite as muchas it does economicrelationsand activities outsidethe organization.This disturbance to internalequilibriumis not adjustedthroughmarketmechanismsand bargainingtransactions.It involvesplanningactivity. Decisive importancecan scarcelybe allowed to attach to individualswho conceive new ideas, even when this duty is delegatedto them as a specifictask. The locus of decision tends to spreadto a groupthat includespersonsin a position to know and deal with prospectiveinternaldisturbanceswhich are only partially of an economiccharacter.' It is not clear that this developmenthas explicitly gone far in railroadorganization.As an innovationin the role of entrepreneurshipitself, it is emergentin some newerlarge-scaleindustries.The ex'An introduction to the sociological theory of organization can be found in Chester I. Barnard, The Functions of the Executive (Cambridge: Harvard University Press, I938). Cf. T. N. Whitehead, Leadership in a Free Society (Cambridge: Harvard University Press, I936), chaps. vi and viii. The problem at a lower level of enterprise structure is analyzed in F. J. Roethlisberger and William J. Dickson, Management and the Worker (Cambridge: Harvard University Press, I939), chaps. zxiv and xxv.

18

LelandH. Jenks

tent to which the management-enterpriser type, as we may call it, has actually functioned in railroads informally and without explicit recognition deserves inquiry. V This general interpretation of the role of the railroad as an economic force suggests what might be undertaken in greater detail to apply the innovation theory to the history of particular companies and of the railroad system as a whole. What was the impact of the railroad upon technological, locational, structural, and organizational alterations in particular firms, industries, and regions? Parallel inquiries could be made regarding the part played by other major innovations, such as the more recent rise of the electromotive industries. It is not a question of applying the facts of economic history to verify an economic theory. It is a question of using a theory as a tool to coherent understanding of the facts. Economic historians seem increasingly willing to make use of conceptual aids for this purpose. It is one of the most prominent symptoms of what may be a wider tendency to employ analytical procedures in historical studies. For the study of long-run change, the innovation theory stresses two important aspects of historical process: (i) the distinction between innovating (disturbing, inciting, evolutionary) change and various types of adjustment (including expansion), and (2) the distinctive role of entrepreneurship. The first of these aspects provides the framework for systematic exploration of the relation between changes in several sectors of the economy, in so far as these can be interpreted in economic terms. The breakdown of the railroad innovation into three "moments" is only a convenience that may be peculiar to transportation. In any case, the distinction between innovating and adaptive change is a device that should become more serviceable to the historian as it is sharpened by application to a number of particular situations. It does not necessarily require the economic historian to take into account other than economic events and processes. Indeed, its logical adequacy can only gain from rigorous limitation to the items that are considered to be a part of an economic system. The emphasis upon entrepreneurship as the crucial factor in capitalistic evolution involves both theorist and historian in considerations that go far beyond the limits of economics. Schumpeter is explicitly aware of this fact, and insists that in his conception the economy is not isolated but functions in a larger universe which requires in the first instance sociological analysis for its interpretation. The theory of innovations is neither a "great man" nor

Railroadsas an EconomicForce

I9

a "bettermousetrap"theory of history. The innovatoris a person whose traits are in some part a functionof his socioculturalenvironment.His innovationis a new combinationof factorsand elementsalreadyaccessible.It relatesin everyphaseto previouslydevelopedbusinessandmonetaryhabits, technologicalskills, and variabletastes, none of which can be regardedas functionsof economicactivity alone.Thus Schumpeter'stheoryinvolvesthe questionof the sociologicalfactors favorableto the emergenceof entrepreneurship.In a recentwork he has presenteda partial analysis of such factors.' Furtheranalysis seems to be called for, at least so far as American capitalism is concerned,analysis that will come to closer grips with the special features of Americansocial structure and the various influences whichmadefor a strongentrepreneurial bias in the "socialcharacter"of the nineteenth-centuryAmerican. Despite his sociological sophistication,however, Schumpetertends to think of his entrepreneurpretty muchas a deviantperson-a particularindividual or at most a family. This approachtends to make highly problematicalthe existenceof any entrepreneurship in a bureaucraticenterprise such as the railway,whetherunderprivate or public ownership.It must be recognizedthat innovationsin a socialist economywould work themselves out by mechanismsother than undercapitalism.But not all of such differences would be peculiar to socialism. Practically, large-scaleorganization offersa new type of social resistanceto innovation.At the same time, as Schumpeterhimself vigorously argues, the large organizationoffers real supportto technologicalchange,at least, by mobilizingresourcesfor its systematic planning.' It is possiblethat thereis a real social lag in conceptionsof the entrepreneurial function. The question deserves to be consideredwhether policy formationby groupaction is an obstacle to innovation,not inherently,but only becauseof certain peculiaritiesin our culture. Is the entrepreneurial rolein largeorganizationsincreasinglythe functionof a co-operatinggroup? Is it true that this tendencyis not absolutely new but can be discernedin earlier phases of modernindustry; that it is less important in entrepreneurial studies to single out the contributionsof one individual than to ascertainthe personalcompositionof the groupwith whichhe usuallyintera Joseph A. Schumpeter, Capitalism, Socialism, and Democracy (New York: Harper and Brothers, I942), chaps. xi-xiv. 4 Ibid., pp. 96-8. Schumpeter seems to regard this change as more than adaptational. In so far as it is innovational, however, it functions less to develop capitalist structure than to further its incipient transformation into something else.

LelandH. Jenks

20

acted and the way in which the memberscompensatedfor their respective shortcomingsand wereadjustedto each other?In so far as thereis validity in affirmativeanswersto these questions,a practicalproblemof much importancefalls uponthe largeorganizationsof the presentday, that of cultivating social techniquesfor facilitatinginnovations.But there would be a broadersocial problem, that of developingpersonalitieswhose practical imaginationand responsibilityfor decisionwill be stimulated rather than frustratedby membershipin policy-determininggroups.This would be a task for the family and other educationalinstitutions and for socializing processesin the widersociety. WellesleyCollege

LELAND H. JENKS