Economic History Association
American Law and the Marketing Structure of the Large Corporation, 1875- 1890
Author(s): Charles W. McCurdy
Source: The Journal of Economic History, Vol. 38, No. 3 (Sep., 1978), pp. 631-649
Published by: Cambridge University Press on behalf of the Economic History Association
Stable URL: http://www.jstor.org/stable/2119473 .
Accessed: 17/11/2013 12:56
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
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 support@jstor.org.
.
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
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
THE JOURNAL OF ECONOMIC
VOLUME XXXVIII
HISTORY
SEPTEMBER 1978
NUMBER 3
AmericanLaw and the MarketingStructureof the
LargeCorporation,1875-1890
By CHARLESW. MCCURDY
This paper employs the techniques of legal history to explore the relationship
between the rise of big business and the size of the American market. It
emphasizes law as a determinant of market size, and it analyzes judicial construction of the Constitution's commerce clause over time to delineate the role of
integrated corporations in generating legal change. Specifically, the paper suggests that if the American market is defined as a free trade unit, enlargement of
the market was a result of, rather than a prerequisite for, the post-Civil War
revolution in business organization.
HE "secret"of Americaneconomicgrowth,Englishlegal scholar
Sir Henry Maine wrote in 1886, lay in "the [constitutional]
prohibitionagainstlevying duties on commoditiespassingfrom State
to State. . . . It securesto the producerthe commandof a free market
over an enormousterritoryof vast naturalwealth, and thus it secondarily reconciles the Americanpeople to a tariffon foreign importations as oppressiveas ever a nationhas submittedto."LThe debate on
T
Journal of Economic History, Vol. XXXVIII, No. 3 (Sept. 1978). ? The Economic History
Association. All rights reserved.
The author is Assistant Professor of History, University of Virginia. An earlier version of this
paper was read in March 1978, at the Twenty-Fourth Annual Meeting of the Business History
Conference. A printed version of this paper will also appear in Business and Economic History,
second series, Volume 6. The author is indebted to Fred V. Carstensen, William H. Harbaugh,
Jonathan Lurie, and Harry N. Scheiber for comments that proved helpful in revising the essay
for publication.
1 Sir Henry Maine, Popular Government (New York, 1886), p. 247.
631
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
632
McCurdy
the tariff'scontributionto industrialdevelopment has not been resolved any more satisfactorilyby modern scholars than it was by
nineteenth-centurypoliticans.But virtuallyeveryonehas long agreed
that a divisionof the United States"intoa numberof smallermarket
areasseparatedfromeach other by tariffwalls,"as in Europe,would,
as Stuart Bruchey stated it, "have abridgedthe possibility . . . of
large-scaleproduction."For Bruchey, as for Maine, "of the many
contributionsto growthmade possible by the adoptionof the Constitution, perhapsthe most fundamentalwas that it laid the foundations for a nationalmarket."2
Yet on close inspection the sweeping statements of Maine and
Bruchey claim too much. Neither the PhiladelphiaConventionnor
the Supreme Court under Chief Justices John Marshalland Roger
Taney had, in fact, deprived the states of all power to interpose
obstacles to the movement of productsthroughoutthe nation. Although the framers of the Constitution proscribed outright tariff
barriers,state governmentsretainedample authorityto devise more
subtle formsof protectionthroughoccupational-licensing
and inspection laws. Before the last quarterof the nineteenth century, moreover, appellatecourtsroutinelysustainedsuch statutes.On the eve of
the post-CivilWarrevolutionin the structureof Americanmarketing,
there remaineda host of barriersto free intercourseamongthe states.
This paper attemptsboth to describe the late-nineteenth-century
emergence of the "free trade"doctrine in Americancommerce law
and to offer some generalizationsabout economic growth and the
dynamicsof legal change. It furtherendeavorsto shed some new light
on AlfredD. Chandler,Jr.'scontentionthat the growthof a national
market was the chief prerequisite for the rise of large-scale,
vertically-integratedcorporationsin the manufacturingsector.3Specifically,it suggeststhat if the nationalmarketis definedin termsof a
free-tradeunit, rather than in terms of an integratedtransportnetwork,then the post-CivilWarpioneersin businessorganizationwere
instrumentalin the creationof the market.Insteadof respondingto
an existingfree marketof continentaldimensions,producersof sewing machinesand dressedbeef actuallyignoredlegal barriersdevised
by state governmentsand instructedtheir local marketingagents to
invite arrestand conviction.At that point, the companies'headquar2 Stuart Bruchey, The Roots of American Economic Growth, 1607-1861 (New York, 1965),
pp. 96-97.
3 Alfred D. Chandler, Jr., "The Beginnings of Big Business in American Industry," Business
History Review, 33 (1959), 1-31.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
MarketingStructure
633
ters mobilized the substantialfinancialresourcesnecessaryto press
the Supreme Court for relief, and hired counsel who succeeded in
persuadingthe Court that existingcanonsof constitutionalconstruction had to be modified to accord with the changing structure of
business enterprise. Earlier effortsto evoke similarmandatesfrom
the judiciaryhad failedbecausepriorplaintiffshad not been affiliated
with big business firms.They had lackedthe resourcesto engage in
protractedlitigation;equallyimportant,their interest in a free-trade
unit had not been sufficientlycompelling to induce an innovative
response from appellatejudges. From the legal historian'sperspective, in short, the rise of big business was a prerequisitefor the
emergence of a nationalmarket.
Amongthe chief motivesforcallingthe PhiladelphiaConventionin
1787 was the "interferingand unneighborly[state] regulations"that
had created"animosityand discord. .. between differentpartsof the
confederacy"under the Articlesof Confederation.4The power of the
severalstates"toimposedutieson importsand tonnage,"JusticeLevi
Woodburyacknowledgedin 1849, "had caused so much difficulty,
both at home and abroad,that it was expressly and entirely taken
awayfromthe States."5On questionsof internalcommerce,however,
the framershadbeen "characteristically
Delphic."6They vested Congress with plenarypower over "Commercewith foreignnations,and
among the several States," and AlexanderHamilton, in particular,
viewed that provisionas an instrumentto facilitate"an unrestrained
intercoursebetween the states"thatwould"advancethe tradeof each
by an interchangeof their respective productions,not only for the
supply of reciprocalwants at home, but for exportationto foreign
markets."7But the permissiblescope of state activityin the silence of
Congressneitherattractedattentionnor fomentedinstructivedebate
at the PhiladelphiaConvention.8
Participantsin the pamphlet war generated by the ratification
Clinton Rossiter, ed., The Federalist Papers (New York, 1961), p. 144.
Passenger Cases, 7 How. 283 (U.S. 1849) at 545.
6 John P. Roche, "Entrepreneurial Liberty and the Commerce Power: Expansion, Contraction, and Casuistry in the Age of Enterprise," University of Chicago Law Review, 30 (1963),
682.
7 Rossiter, ed., Federalist Papers, p. 89.
8 Albert S. Abel, "The Commerce Clause in the Constitutional Convention and in Contemporary Comment," Minnesota Law Review, 25 (1941), 432-94; Charles Warren, The Making of
the Constitution (Boston, 1928), pp. 567-89.
4
5
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
McCurdy
634
controversyalso devoted scantspace to the internalcommercequestion. When commentatorsdid treat the commerceclause, moreover,
they gave no indicationthatthe states'police andtaxpowerswouldbe
affectedunless Congressoccupied the field with its own legislation.
Hamilton, for example, flatly promised state governments that "with
the sole exception of duties on imports and exports," they would
retain their power to tax "in the most absolute and unqualified
sense."9 James Madison, too, accorded wide policy latitude to state
governments. "The regulation of commerce, it is true," he stated in
The Federalist, "is a new power [of Congress]; but that seems to be an
addition which few oppose, and from which no apprehensions are
entertained." Indeed, Madison explained, "the powers reserved to
the States will extend to all the objects which, in the ordinary course
of affairs, concern the lives, liberties, and prosperity of the
State[s]."10
The commercial-policy objectives of state governments were, in
fact, extraordinarily broad. Throughout the nineteenth century,
Americans looked primarily to state and local officials to promote
internal improvements and regulate commercial traffic, tended to
regard "each State as a community of interest . . . operating in a
hostile environment of rival State communities," and expected governmental agencies closest at hand to be responsive to their particularistic interests.11 In an era when constitutional scruples and
regional power groupings forestalled vigorous congressional action
under the commerce clause, state and local interventions did play an
important role in overcoming the physical obstacles that for generations had circumscribed inter-regional trade within narrow limits.12
State governments facilitated the expansion of interstate transactions,
generally, and also protected shippers, passengers, and consumers
against negligence or fraud on the part of carriers and merchants.13
But the state legislatures also spun an effective web of barriers to
internal commerce. Measures designed to protect consumers or to
promote inter-regional transactions joggled incongruously with statutes frankly adopted to impede the introduction of out-of-state prod9 Rossiter, ed., Federalist Papers, p. 198.
10 Ibid.,
pp. 292-93.
11 Harry N. Scheiber, The Condition of American Federalism: An Historian's View (Washington, D.C., 1965), p. 5.
12 Carter Goodrich, Government Promotion of American Canals and Railroads, 1800-1890
(New York, 1960), pp. 3-165.
13 Albert S. Abel, "Commerce Regulation Before Gibbons v. Ogden: Interstate Transportation Facilities," North Carolina Law Review, 25 (1946), 121-71.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
635
MarketingStructure
ucts on a bargainingparitywith local goods. State and local officials
prescribedmarketingpractices, enacted discriminatoryschemes of
mercantilelicensing and taxation,proscribedthe entry of unfavored
articles of commerce, and devised inspection laws to improve the
competitive position of their citizens relative to producersin other
states.14 In short, state governments acted freely on all matters
concerning commercial traffic-whatever their interstate ramifications-as-if they were unaware, or at least unconcerned, that
the commerce clause might have divested them of powers they had
exercised under the Articles of Confederation.
Before 1875, the federal courts said nothing that disturbed the
states' impulse to intervene on behalf of local interests. In the landmark cases of Gibbons v. Ogden (1824) and Brown v. Maryland
(1827), the Marshall Court established two principles of profound
importance: "Commerce is intercourse" and it includes transportation
and traffic, which comprise "its essential ingredients."15 In both
instances, however, the Chief Justice deftly avoided a direct confrontation with the question, as he put it, "whether this [commerce]
power . . . is surrendered by the mere grant to Congress,
or is
retained [by the states] until Congress shall exercise the power. "16
Although he handed down dicta which looked toward a "dormant"
theory of the commerce clause, Marshall preferred to invoke unlikely
federal statutes-a federal coasting-license act in the one case, and
national tariff laws in the other. As a result, the Court virtually
ignored the broad policy issues raised by counsel, and held that the
"sole question
[was], can a state regulate commerce
.
.
.
while
Congress is regulating it. "17 Anchoring his opinions as much upon the
supremacy and import-export clauses as upon the commerce clause,
Marshall flatly stated that state laws must give way once Congress
occupied the field with its own legislation.
The legitimacy of state and local commercial interventions in the
silence of Congress was too vital an issue to be long suppressed -by
Marshall's penchant for "esoteric statutory construction," and between 1837 and 1851 the Taney Court split into a bewildering arrayof
shifting factional alignments as the Justices attempted to devise a
14 Albert S.
Abel, "Commerce Regulation Before Gibbons v. Ogden: Trade and Traffic,"
Brooklyn Law Review, 14 (1947-48), 38-77, 215-43; Stanley C. Hollander, "Nineteenth Century
Anti-Drummer Legislation in the United States," Business History Review, 38 (1964), 479-500.
15 Gibbons v. Ogden, 9 Wheat. 1 (U.S. 1824) at 189; Brown v. Maryland, 12 Wheat. 419
(1827) at 466.
16 Gibbons v. Ogden, 9 Wheat. 1 (U.S. 1824) at 200.
17 Ibid.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
636
McCurdy
workable canon of constitutional construction.18 Not until Cooley v.
Board of Wardens (1851) was the protracted intra-Court controversy
over whether the regulation of internal commerce belonged exclusively to Congress or admitted of a concurrent power in the states
ultimately stilled. Justice Benjamin Curtis there pointed out that
commerce embraced a great variety of subjects, some of such a nature
as "imperatively" to require a uniform, national rule whereas others
admitted of local control until such time as Congress occupied the
field.19 The Cooley rule, the Court later noted, was "as satisfactory a
solution as perhaps could be obtained ... [on a] question which had
so long divided the judges."20 As an adjudicatory mechanism, however, it was virtually useless. As Kent Newmyer observed, "the
Cooley decision was less a doctrinal clarification than it was an agreement to stop looking for one. "21
For the purposes of the postwar Court, the most important aspect
of Cooley was its unarticulated premise: when Congress remained
silent, the Court might supply its voice. In exercising the enormous
discretion inherent in the Cooley rule, moreover, the Court had
ample room to resolve disputes on the basis of franklyinstrumentalist,
extra-constitutional criteria. Between 1851 and 1875 the question of
how the Court might employ that self-created power, and for what
purposes, remained uncertain. Then the revolution in the structure of
American marketing generated a period of extraordinaryferment that
culminated with the creation and systematic application of the "freetrade" doctrine.
II
Through the first five decades of the nineteenth century the independent, "sedentary" merchant integrated the American marketplace. Urban-based wholesalers supplied manufacturerswith capital for building plants, purchasing equipment, and paying wages; they
also managed the flow of finished goods to retailers. Direct contacts
between manufacturers and consumers were rare. Indeed, as late as
1860 the word "drummer"-which later became the popular term for
18 Felix Frankfurter, The Commerce Clause Under Marshall, Taney and Waite (Chapel Hill,
1937), p. 20. See also Carl Brent Swisher, Oliver Wendell Holmes Devise History of the
Supreme Court of the United States, vol. V, The Taney Period, 1836-1864 (New York, 1974), pp.
357-422.
19 Cooley v. Board of Wardens, 12 How. 299 (U.S. 1851).
20 County of Mobile v. Kimball, 102 U.S. 691 (1880) at 701.
21 Kent Newmyer, "History Over Law: The Taney Court," Stanford Law Review, 27 (1975),
1378.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
MarketingStructure
637
travelingsalesmen-was used to refer to the men wholesalersplaced
in hotel lobbiesto greet the hinterlandbuyerswho madeannualvisits
to eastern marketingcenters.22But changes generatedby a swiftly
developingnew technology,by the expansionand integrationof the
nation'stransportnetwork, and by Civil War financialinnovations
andcommercialbankwhichcreatednew ties between manufacturers
of American
the
structure
all
contributed
to
a
revolution
in
ers
marketing.23
of new, expensive,and techThe pioneerswere the manufacturers
nologicallycomplex productssuch as sewing machines.By 1860, I.
M. Singer & Companyhad learned that existing wholesalerswere
unable to provide consumer credit or handle demonstrationand
repairservices;consequentlyit moved into the wholesalers'domain
and began to create its own distributionnetwork.24It was not an
instantaneousprocess, however. Operationof company-ownedretail
outlets entailed high overheadcosts, and fewer than 100 such stores
were opened as late as 1872. Not until 1873, when Singer's enormous new Elizabethport,New Jerseyfactorynearedcompletion,did
the central office commit itself to expanding the firm's marketing
organizationto the entire domesticmarket.By 1879the companyhad
severedits relationswith all independentmerchants,andits distribution networkconsisted of 530 retail stores which also served as the
base of operationfor a still largerforce of door-to-doorsalesmen.25
Legal barriers posed immediate problems for the architects of
Singer'saggressivesales organization.State governments,prodded
by the local merchants and manufacturerswhose interests were
threatened, not only stepped up enforcement of long-established
licensinglaws for peddlers but also enacted new revenue statutesto
buttressthe competitivepositionof local businessmen.The very size
of Singer'smarketingorganizationafter 1873 meant that there were
substantialprofitsto be made by breakingdown these barriers.As a
result, Singer coupled its final drive to integrate its manufacturing
and distributionoperations with a determined effort to challenge
protectioniststate legislationin the nation'scourts.
In choosingto mount a legal assaulton state tradebarriers,Singer
22
Hollander, "Nineteenth Century Anti-Drummer Legislation," p. 481.
23 Glenn Porter and Harold C. Livesay, Merchants and Manufacturers: Studies in the
Changing Structure of Nineteenth-Century Marketing (Baltimore, 1971).
24 Andrew B. Jack, "The Channels of Distribution for an Innovation: The Sewing Machine
Industry in America, 1860-1865," Explorations in Entrepreneurial History, 9 (1957), 11341.
25 The data on Singer's post-Civil War marketing strategy were generously supplied by Fred
V. Carstensen.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
638
McCurdy
faceda formidabletask. Statutesthat requirednon-residentsalesmen
to pay higher licensing fees than local merchantsviolated the privileges and immunitiesclause of the Constitution;thus they had been
subjected to the judicial veto in a long line of cases culminatingin
1870 with an authoritativeruling by the Supreme Court.26When
state and local governmentsdiscriminatedagainstout-of-stateproducts ratherthan the salesman'sdomicile, however, tax and licensing
laws had been sustainedroutinelyby the courts.
Throughthe mid-1870s,the state judiciariesregularlytreated the
1827case of Brownv. Marylandas the Court'sfinalwordon state and
local taxationof commercialtraffic.There the MarshallCourt had
invalidateda state law that requiredwholesalersof foreignmerchandise, and only dealers in foreign merchandise, to pay an annual
license tax. Speakingthroughthe Chief Justice, the Courtheld that
Congresshad alreadyforced importersto pay tariffduties on their
wares, and had thereby conferredupon them the right to sell in an
unfettered market-a right which, if abridgedby state law, would
have made the right to import of little value. As Felix Frankfurter
later observed, "the circumstancesof the case furnished a ready
opportunityfor curbing state taxationdiscriminatingagainst interstate commerce."27 Nevertheless, Marshall focused the bulk of his
discussion on the import-export clause, and he suggested that the
states might tax all commodities imported from abroad, or from any of
the several states, once the "original package" had been broken and
the goods had "become incorporated with the general mass of property."28
The state judiciaries readily discerned "an immeasurable difference" between the act nullified in Brown and discriminatory taxation
of goods offered for sale by peddlers.29 Statutes of the latter variety,
the Indiana court ruled in 1835, did not impede the operation of the
federal revenue laws, for the commodities thus taxed had already
"become incorporated with the great mass of property in the state."
Moreover, the Indiana court asserted, the power to tax "is inseparable from sovereignty, essential to its existence, and one which all the
expounders of the Constitution admit to have been reserved" by the
26 Ward v. Maryland, 12 Wall. 418 (U. S. 1870). See also the long line of state court decisions
collected in Thomas M. Cooley, A Treatise on the Constitutional Limitations Which Rest Upon
the Legislative Power of the States of the American Union (Boston, 1868), p. 487.
27 Frankfurter, Commerce Clause, p. 36.
28 Brown v. Maryland, 12 Wheat. 419 (U.S. 1827) at 447.
29 Beall v. State, 4 Blackf. 107 (Indiana 1835) at 109.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
MarketingStructure
639
states.30 It was an inexorable corollary, an Alabama judge added a
generation later, that a state legislature might tax all merchandise sold
within its jurisdiction, whatever its state of origin, while encourag[ing] manufacturersin its [own] borders, by exempting the articles so
manufactured from taxation for a time, or altogether."31
The line of reasoning pioneered in Indiana comported with
the particularistic needs of local merchants and manufacturers
everywhere, and other state courts quickly adopted a similar position.32 In an often-cited opinion in the License Cases (1847), moreover, Justice Woodbury observed that "it is perfectly competent for
[the states] to assess a higher tax or excise, by way of license or direct
assessment, on articles of foreign rather than domestic growth belonging to her citizens; and it has ever been done, however it may
discourage the use of the former."33When Thomas Cooley published
the first edition of his Constitutional Limitations in 1868, then, the
principle of unrestricted state taxation of commercial traffic, in the
silence of Congress, had already attained the status of a settled rule.
"The states may unquestionably tax the subjects of commerce,"
Cooley wrote in his influential treatise, "and no necessary conflict
with that complete control which is vested in Congress appears until
the power is so exercised as to defeat or embarrass the congressional
legislation. Where Congress has not acted at all upon the subject, the
state taxation cannot be invalid on this ground."34
Despite the weight of precedent, Singer persevered with remarkable success. In the fourth edition of his Constitutional Limitations,
published in 1878, Cooley had already begun to note exceptions to
the rule he had formulated so confidently a decade earlier; when the
sixth edition appeared in 1890, the passage on state taxation of commercial traffic had been excised altogether. The leading case in this
crucial doctrinal transformation came up from Missouri in 1875.
Thirty years earlier the Missouri legislature had enacted a revenue
measure that defined peddlers as persons selling commodities "not
the growth, produce, or manufacture of th[is] State," and required
them to pay a license fee for the privilege of engaging in local
30
Ibid.
Seymour v. State, 51 Ala. 52 (1874) at 54.
32 Raquet v. Wade, 4 Ohio 107 (1829);People v. Coleman, 4 Cal. 46 (1854); State v. Pinckney,
10 Rich. L. 474 (South Carolina 1857); Davis v. Dashiel, 61 N.C. 114 (1867); State v. Hogdon,
41 Vt. 139 (1869).
33 License Cases, 5 How. 504 (U.S. 1847) at 622.
34 Cooley, Constitutional Limitations, p. 486.
31
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
640
McCurdy
business.35 M. M. Welton, an agent of I. M. Singer & Company, had
been convicted under the statute and the law had been sustained by
the state's highest court.36
On appeal to the Supreme Court, counsel for Missouri simply stood
on the precedents in a terse, five-page brief. Singer, on the other
hand, hired two luminaries from the Missouri bar who compiled
lengthy briefs that spoke directly to the policy issues involved in the
dispute. Stated simply, they argued that existing doctrines were "not
practical in this case. "37Missouri's equation of peddlers with hawkers
of out-of-state goods. they contended, was such "linguistic legerdemain" that the legislature might as well have "define[d] a peddler to
be one who deals in boots and shoes manufactured in Lynn, or salt
produced in Syracuse."38In short, counsel emphasized that the statute was simply a protective tariffdisguised as a licensing law. Because
peddlers of local products were exempt, the Missouri law was "not a
tax on the occupation of selling, but a burden on the goods themselves."39 It followed, counsel concluded, that "[i]f this is a valid
exercise of the taxing power, the legislature may wholly exclude . . .
products of sister states; for a lawful exercise of a power knows no
limitation except such as are to be found in the discretion of the
lawmakers.
"40
The Supreme Court concurred. Speaking through Justice Stephen
Field, the Court readily conceded that under the Brown doctrine the
Singer machines had lost their interstate character before their sale
had become subject to the Missouri licensing law. Nevertheless,
Field flatly asserted that Marshall's"guarded language" could not be
expected to control disputes arising in an integrated national economy.41 Paraphrasingcounsel's brief, Field asserted that it was unnecessary to consult an economist to discern that "where the business or
occupation consists in the sale of goods, the license tax required for its
pursuit is in effect a tax upon the goods themselves."42 It was equally
clear, Field added, that unless the "original package" doctrine were
" Evans Casselberry, ed., The Revised Statutes of the State of Missouri (St. Louis, 1845),
pp. 404-05.
36s State v. Welton, 55 Mo. 288 (1874).
37 James S. Botsford, M. M. Welton,
Plaintiff in Error vs. The State of Missouri: Brieffor
Plaintiff in Error (Jefferson City, Mo., 1875), p. 9.
a S. M. Smith, M. M. Welton, Plaintiffin Error vs. The State of Missouri: Brieffor Plaintiff
in Error (St. Louis, 1875), p. 5.
39 Botsford, Brief, p. 12.
40 Ibid., p. 13.
41 Welton v. Missouri, 91 U.S. 275 (1876)at 281.
42
Ibid., p. 278.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
MarketingStructure
641
modified, a barrageof interstate tariffwars, like those which had
"depressed[the] conditionof commerceand [created]obstaclesto its
growthpreviousto the adoptionof the Constitution... mightfollow,
and the experienceof the last fifteen years showswould follow, from
the action of some of the States."43Because of new developments in
the structure of American marketing, he concluded, it had become
necessary to extend Marshall'sBrown doctrine, and "to hold now that
the commercial power [of Congress] continues until the commodity
has ceased to be the subject of discriminating legislation by reason of
its foreign character.""
The protectionist impulse in the states was not easily curbed, and
the Missouri case established only a beachhead for large-scale,
vertically-integrated firms. As Harry N. Scheiber has demonstrated
for the antebellum era, state governments were extraordinarilyadept
at initiating successful "counterthrusts" to the Supreme Court's nationalistic doctrines.45 After the postwar marketing revolution, however, the ingenuity of local lawmakers rarely went unchallenged.
Firms such as Singer whose interests were national in scope were
quick to muster test cases in response to each new statutory innovation. Thus, in 1880 Singer's counsel was back in Washington to
challenge a Virginia law enacted five years earlier.
The measure disputed in Webber v. Virginia (1880) apparently was
designed to compel Singer and all other out-of-state firms to disband
their sales forces and deal exclusively with local wholesalers. Under
the act, all salesmen who peddled "manufacturedarticles or machines
. . . of other states or territories" were required to pay the state a
license fee of $25 and an additional $10 fee in every county where
they did a local business. Only distributors who actually owned
products manufactured outside Virginia were exempt.46 For the
Singer Company, whose Richmond agency supervised salesmen
working door-to-door in a dozen surrounding counties, the Virginia
law had a potentially disastrous effect. Nevertheless, the Virginia
Court of Appeals sustained the statute, distinguishing the Missouri
case on the ground that there was no discrimination against out-ofstate products as long as the manufacturerhad the option of distributing his wares through local wholesalers.47
Ibid., pp. 280-81.
" Ibid., p. 281.
Law
45 HarryN. Scheiber, "Federalismand the AmericanEconomicOrder, 1789-1910,"
and SocietyReview, 10 (1975),84.
46 Virginia,Acts of Assembly,1875-1876,p. 184.
47 Webber-v.Commonwealth,
33 Gradt.898 (Va. 1880).
4
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
642
McCurdy
The Supreme Court, after having been briefed about Singer's
organizational structure, voted unanimously to reverse the decision
rendered below. Speaking again through Field, the Court disposed in
a single sentence of the issue emphasized by the Virginia bench.
"Sales by manufacturers," Field proclaimed, "are chiefly effected
through their own agents. "48 Once the postwar marketing revolution
had been thus ratified, the result flowed inexorably from the principles announced in Welton v. Missouri:
It matters not whether the tax be laid directly upon the articles sold or in the form of
licenses for their sale. If by reason of their foreign character the State can impose a
tax upon them or upon the person through whom the sales are effected . . . she may
place the tax at so high a figure as to exclude the introduction of the foreign article,
and prevent competition with the home product. It was against legislation of this
discriminating kind that the framers of the Constitution intended to guard, when
they vested in Congress the power to regulate commerce among the several States.49
The last sentence above merits special attention, for it reveals in
disarming fashion the degree to which I. M. Singer & Company had
succeeded in fomenting a doctrinal revolution. As Field conceded,
the framers of the Constitution had vested Congress-not the
Court-with the authority to regulate interstate commerce. Nevertheless, the Court believed that the idea of a unitary national market
would be nullified if large-scale firms were required to press Congress
for relief each time the states disguised protectionist legislation in the
form of licensing laws. As Justice Robert Jackson later observed, "the
balkaniz[ing]" policies of state governments were just "too petty, too
diversified, and too local to get the attention of a Congress hard
pressed with other matters."50
The Court's decisions in the license-tax cases also marked a decisive
break with prior doctrinal formulations. In Gibbons and Brown, the
Marshall Court had curbed the states in order to protect rights that
Congress had conferred on persons engaged in interstate and foreign
commerce. Then, during the Taney era, the majority's concerns had
shifted from protecting the prerogatives of Congress to maintaining
the territorial integrity of the states. Consequently, the Taney Court
tended to classify powers-taxation, police, commercial regulationand then assign control of public policy to the proper governmental
agencies. The Cooley doctrine looked to the subject matter of state
policies; since regulation of chattel slaves, prevention of disease or
48
49
50
Webberv. Virginia,103 U.S. 344 (1880)at 350.
Ibid., pp. 350-51.
Ducktorth v. Arkansas,314 U.S. 390 (1941)at 400.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
MarketingStructure
643
pauperism,and licensingof liquordealersand steamboatpilots were
all "local"matters,the Taney Courthad held that they were subject
to state laws. Once faced with questionsgeneratedby the rise of big
business, however, the Courtbegan to conceptualizeissues in terms
of free trade and free markets.In Welton and Webber, the Court
looked to the incidence of state laws;if barriershad been erected to
impede the inter-regionalflow of commodities, revenue measures
were held to be invalid despite the fact that the states'power to tax
had been "admit[ted]to have been reserved. .. [by] all the expounders of the Constitution."51
In effect, then, Singer'scounsel prompted the Courtto deduce fromthe commerceclause a new, fundamentally important constitutionalright: the right of American businessmen, even without congressionallicense, to engage in interstate transactionson terms of equality with local merchants and
manufacturers.
As late as 1885, the "freetrade"doctrinehad been appliedonly to
state tax laws. Eventually, however, the rule formulatedin Welton
spilled over and controlledthe Court'sposition on inspectionlaws.
Appropriately,the key agents of the latter development were the
"Big Four"meatpackers.
The developmentof railroadrefrigerationexerted a revolutionary
impact on the Americanmeat business.52For centuries priorto the
1880s,cattleand swine had been drivenon the hoof, and laterby rail,
to highly localizedprocessingplants.When fresh meat was available,
consumers knew it had been slaughterednearby. Beef and pork
preparedfor interstateand foreign commerce had to be salted and
barrelled,or cannedwith preservatives,in orderto prevent spoilage.
The refrigeratorcar not only extended the potential market for
dressed beef but, since unsaleableparts of the animalneed not be
shipped, it also permittedthe processorto save up to 35 percent on
freight costs. By combiningrefrigerationwith mass-processingtechniques and a strategic location amidst the Chicago stockyards,the
"BigFour"packerswere able to ship dressedbeef thousandsof miles
and still undersell local butchers by a substantialmargin.
The combinationof factorsthatenabledChicagopackersto obtaina
virtual monopolyon the dressed-beeftrade is still a matterof some
dispute among scholars.53In testimony taken at St. Louis in 1888,
51
BeaUv. State, 4 Blackf.107 (Indiana1835)at 109.
MaryYeagerKujovich,"Te RefrigeratorCarand the Growthof the Anerican Dressed
Beef Industry,"Business HistoryReview, 44 (1970),460-82.
53 See e.g., RudolfA. Clemen, The AmericanLivestockand Meat Industry (New York,
52
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
644
McCurdy
however, a Senate Select Committee discovered that old-style local
butchers believed almost unanimously that the "Big Four" packers
had conspired to "freeze out" all competitors.54 Chicago packers, the
small-scale butchers testified, had extorted from carriers special rates
for handling refrigeratorcars; they had ordered their wholesale agents
to employ predatory pricing tactics in local markets; and they had
conspired with the stockmen's commission merchants to ensure that
live cattle were sold only by the carload. Witness after witness nevertheless informed the senators that no federal intervention was necessary. A national organization, the Butcher's Protective Association,
had been created, they explained, and it intended to seek relief in the
several state legislatures. One witness, Detroit butcher John Duff,
testified as follows:
Q. What is your remedy for it [collusion among Chicago packers]? -A. Give us a
livestock inspection, and when meats are not inspected do not allow them. to be
sold.
Q. Do you want State or national inspection? -A. Give us State inspection.
Q. Do you think that a State inspection would be all that would be necessary? -A. I
think so. I think it would cover the case.
Q. You think that would cover all the evils. -A. Yes, sir.55
The small packers' faith in the efficacy of state action should not be
surprising: the Butcher's Protective Association's model statute prohibited the sale of dressed beef, mutton, or pork unless it had been
inspected by state officials twenty-four hours before slaughter. In
short, the B.P.A. proposed to banish the "Big Four" packers from all
but the Chicago market.
In 1889, the B.P.A. persuaded lawmakers in Minnesota, Indiana,
and Colorado to enact their panacea for monopoly in the dressed-beef
trade.56 Bills providing for pre-slaughter inspection failed to pass in a
score of other states because, according to one proponent, "of the
presence of a powerful lobby representing the most colossal
monopoly, perhaps, that any government was ever confronted
1923); Richard J. Arnould, "Changing Patterns of Concentration in American Meat Packing,
1880-1963," Business History Review, 45 (1971), 18-34; Robert Aduddelland Louis Cain,
'Location and Collusion in the Meat Packing Industry," Business Enterprise and Economic
Growth: Essays in Honor of Harold F. Williamson, Louis P. Cain and Paul L. Uselding, eds.
(Kent, Ohio, 1973), pp. 85-117.
54 U.S. Senate, Testimony Taken by the Select Committee on the Transportation and Sale of
Meat Products to Accompany Senate Report No. 829, 51st Congress, 1st Session (Serial 2705).
55 Ibid., p. 156.
" Minnesota, Laws, 1889, p. 51; Indiana, Laws, 1889, p. 150; Colorado, Laws, 1889, p. 244.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
MarketingStructure
with."57Where
645
lobbying had proved ineffective, however, the "Big
Four"had no choice but to ignorethe inspectionlaws;thus theirlocal
agents were promptly indicted by state authorities.Within a year
the leading case of Minnesotav. Barber(1890)was on the docket of
the Supreme Court. The lawsuit was so vital to the interests of
Indiana butchers that the state's attorney general asked for, and
received, the Court'spermissionto join his Minnesotacounterpartin
defending the statutes.
Counsel for Minnesotaand Indiana presented compelling arguments. Inspectionlaws had long been used by the states to improve
their producers'competitive position, and the framersof the Constitutionhad deemed inspectionmeasuresto be of such greatimportance to local economiesthat the states'power to enact them, and to
chargefees for their operation,had been expresslyrecognizedin the
Constitution.In Gibbons v. Ogden, Chief Justice Marshallhad accorded furtherlegitimacyto existing practicesby observingthat the
states'powerto enact inspectionlaws had "not[been] surrenderedto
the generalgovernment"despite the fact that they "m[ight]have...
a considerableinfluenceon commerce."58 Moreover,counselargued,
as late as 1878the Courthad quoted Marshall'slanguagein an opinion
that sustained a Kentuckylaw providingfor pre-sale inspection of
illuminatingoil manufacturedin St. Louis.59But in the event longaccepted constitutionalconstructionand the weight of precedent
were not enough, counsel for Minnesotaand Indianaalso implored
the Court to take judicial notice of what they considered to be a
well-establishedscientific fact. Studies of diseased meat, including
one cited in the Senate Select Committee'sown reportto Congress,
demonstratedthat
offreshbeef,veal,mutton,lamborpork...
to tell, by aninspection
it is impossible
thatan
whetheror not it camefromanimalsthatwerediseasedwhenslaughtered;
is
on thehoof,withina veryshorttimebeforetheanimalsareslaughtered
inspection
withcertainty.Y?
the onlymodeby whichtheirconditioncanbe ascertained
57 GordonE. Cole,In the Matterof theApplication
of HenryE. Barberfora Writof Habeas
Corpus:Pointsand Authoritiesfor Appellant(St. Paul, 1889), p. 43.
58
59
Gibbons v. Ogden, 9 Wheat. 1 (U.S. 1824) at 203.
Pattersonv. Kentucky,97 U.S. 501 (1878).
60 LouisT. Michener,Briefin Behalfof Appellantsin the Matterof HenryE. Barberfora
Writ of HabeasCorpus,Filed by Leaveof the Court and Consentof the PartiesHerein, by
Counselfor the State of Indiana (Indianapolis, 1889), pp. 30-31. See also Dr. Henry Behrend,
"Diseases from Butcher's Meat," Nineteenth Century, 26 (1889), 409; U.S. Senate, Report of
and Saleof MeatProducts,SenateReportNo. 829,
the SelectCommitteeon the Transportation
51st Congress, 1st Session (Serial 2705), p. 26.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
646
McCurdy
And science, it was argued, only confirmed what common sense
suggested: "The examinationof the hind quarterof an ox will not
detect tubercles in.his lungs or canceroustumorsupon his neck."61
Counsel for the defendant, an Armouragent, filed a remarkably
candidbrief.They agreedat the outset thatall the inspectioncaseson
the books supported their opponent's position. Having made that
concession,however,they urgedthe Courtto "bearin mind"its prior
decisions in the license-tax cases and to recognize that unless the
principles laid down in Welton were extended, the idea of a freetrade unit would necessarilybe sacrificedat the altarof plenarystate
inspectionpower. "Ifthe State [of Minnesota]can prohibitinterstate
commerce in beef, unless the livestock is first inspected therere"
counsel for Armourcontended,
it mayin fishunlessthey arefirstinspectedwhen caught.It mayin butterandcheese
and milk and leather, unless the cow fromwhich they are drawnis first inspected
therer. It may in wool and all clothing made from it, unless the sheep is first
inspectedtherer. It mayin cottonand clothingmadefromit, unless the cottonand
the groundthatproducesit is inspectedin Minnesotabeforethe cottonis picked;and
there is no productof the agricultureor manufactureof other Statesthat this State
may not thus exclude;none of this State that every other may not exclude.62
The paradeof horribleslikely to proceed from an affirmativeruling
thus laid bare, counsel took up the problem of whether the Court
ought to take judicialnotice of the scientificstudies adducedby the
states'attorneysgeneral. Counsel'stacticson this issue were extraordinary. Rather than attempting to rebut the contention that only
pre-slaughter inspection was effective, counsel emphasized that
"freshmeats consumedby Minnesotanswere never inspected on the
hoofby Stateor city inspectorsbeforeApril, 1889, andyet population
has increasedand the death rate has been low."63 It followed, counsel concluded,that the Minnesotastatutemust have been enactedto
protect the competitive position of local butchers rather than to
promote the public health.
The SupremeCourtconcurredwith Armour'scounsel. The propositions Justice Field had formulatedin Welton, ratherthan previous
cases involving inspection laws, controlledits decision. We cannot
"shut our eyes," JusticeJohn MarshallHarlandeclaredfor a unanimous Court, "to the fact that the act, by its necessaryoperation..
61
Cole, Pointsand Authoritiesfor Appellant,p. 9.
WalterH. Sanborn,In the Matterof the Applicationof HenryE. Barberfor a Writof
HabeasCorpus:Pointsand Authoritiesfor Respondent(St. Paul, 1889), p. 49.
63 Ibid., p. 48.
62
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
MarketingStructure
647
directly tends to restrict the slaughteringof animals . . . to those
engaged in such business in that State."" Moreover, Harlan observed, there was "no real analogy"between the pre-slaughterinspection laws and the Kentuckyinspection statute the Court had
sustained12 years earlier. Pre-saleinspectionof illuminatingoil was
"neitherunusualor unreasonable,"nor did implementationof that
law ineluctably discriminateagainst commodities "because of the
locality of production." As for the "alleged" necessity for preslaughterinspection, the Court concluded disingenuously,"we are
not awarethat such is the view universally,or even generallyentertained."65If governmenthad a duty to protectconsumersagainstthe
dangersof unwholesomemeat, as counsel for Minnesotaand Indiana
had indicated, another strategywould have to be pursued.66
The obviousremedyforotherwiselegitimateconcernsaboutpublic
healthwas, of course,the creationof a federalinspectionforce. But in
Barber, as in the license-taxcases, the Courtcorrectlyperceivedthat
unless the federaljudiciarysupplied the voice of Congress, federal
lawmakerswouldnot move with dispatch,if at all, to displacediscriminatorystate regulationswith a uniformrule. The reportof the Senate
Select Committee, which appeared two weeks before Barber was
decided, lookedto state actionfor protectionof Americanconsumers
and recommendedfederalinspectiononly of meat productsprepared
for export.6f7Not until five weeks after the Court had spoken did
Congressappropriatethe meat inspectionprovisionsof a comprehensive pure food bill destined to die and tack them on to the Select
Committee'sbill.68The CourtthereforeforcedCongressto applythe
same solution for state barriersagainst "Big Four" meat that the
" Minnesotav. Barber, 136 U.S. 313 (1890)at 322-23.
Ibid., pp. 327-28, 321.
" The "BigFour"packers,like I. M. Singer& Co., hadto contendwith "counterthrusts"
by
the state legislatures;counsl for Armour& Co. was backin courta year later to challengea
Virginiastatutethatprovidedfor post-morteminspectionof all dressedbeef slaughteredmore
than100milesfromthe placeof saleandalsorequiredthe processorto paythe costof inspection
(Virginia,Actsof Assembly,1889-1890,p. 63). The Court,againspeakingthroughHarlan,held
that "the heavy [inspection]charge of one cent per pound"was, "in reality, a tax"which
divestedout-of-statepackersof the rightto "competeuponequal termsin the marketsof that
v. Rebman,138U.S. 78 [1891]at 81-82).The lastB.P.A.-sponsored
(BrImmer
Commonwealth"
law to fall was the Coloradostatute(cited at note 56), which was struckdown in Schmidtv.
People, 18 Colo. 78 (1892).
p. 40;CongressionalRecord,51st Congress,
67 U.S. Senate,Reportof the SelectCommittee,
1st Session(1890),pp. 3056-58,5928-31.The internationaldimensionsof the meat-inspection
issue arethoroughlytreatedin JohnL. Gignilliat,"Pigs,Politics,andProtection:The European
Boycottof AmericanPork,1879-1891,"AgriculturalHistory,35 (1961),3-12; RichardPerren,
'Te NorthAmericanBeefand CattleTradewith GreatBritain,1870-1914,"EconomicHistory
Review, 24 (1971),430-44.
68 OscarE. Anderson,TheHealthof a Nation:HarveyW. Wileyand theFightfor PureFood
65
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
648
McCurdy
Select Committee had prepared already for European barriers against
all American meat. The Federal Meat Inspection Service, established
in 1891, was authorized to conduct pre-slaughter and post-mortem
inspections of meat products produced for interstate and foreign
commerce alike.69
IV
The Supreme Court's commerce clause decisions of the 1875-1890
period were of immediate importance to large-scale manufacturersand
had an enduring influence on American economic growth, for they
firmly established the Supreme Court's role as the umpire of the
nation's free-trade network. Even Justice Oliver Wendell Holmes, a
persistent critic of many late-nineteenth-century decisions, ardently
believed that review of state commercial regulations was an essential
judicial function. "I do not think the United States would come to an
end if we lost our power to declare an Act of Congress void," he
announced in 1913.
[But] I do think the Union would be imperiled if we could not make that declaration
as to the laws of the several States. For one in my place sees how often a local policy
prevails with those who are not trained to national views and how often action is taken
that embodies what the commerce clause was meant to end.70
Holmes's suggestion that the commerce clause was "meant to end"
discriminatory state policies was, as we have seen, correct only insofar
as the Constitution empowered Congress to intervene. Before the
Court could establish fully its claim to monitor the free-trade unit in
the silence of Congress, two prerequisites had to be fulfilled. First,
the Court had to be apprised by skillful counsel of the growth-eroding
potential of state laws and to be persuaded that new juridical principles must be forged to preserve free trade among the states. Second,
the legitimacy of protectionist state legislation had to be challenged
by litigants with sufficient resources to finance scores of lawsuits in
order both to secure initial favorable decisions and to combat the
tendency of state governments to mobilize "counterthrusts" against
the Supreme Court's nationalistic doctrines. For students of the
NAACP's operations in the twentieth century, neither of these
(Chicago, 1958), p. 78; Congressional Record, 51st Congress, 1st Session (1890), pp. 5674, 6514,
10191.
69 United States Statutes at Large, XXVI, p. 1089. See also A. D. Melvin, 'The Federal
Meat-Inspection Service," Twenty-Third Annual Report of the Bureau of Animal Industry
(Washington, D.C., 1908), pp. 65-99.
70 Oliver Wendell Holmes, Collected Legal Papers (New York, 1920), pp. 295-96.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
MarketingStructure
649
caveatsis apt to be startling.But their implicationsfor the development of a nationalmarketsupervisedby appellatecourtswere simply
enormous.What the NAACPLegal Defense fund accomplishedfor
blackAmericansunder the FourteenthAmendmentin the twentieth
century,the legal-defensewarchests of I. M. Singer& Companyand
the "Big Four" meatpackersaccomplishedfor vertically-integrated
corporationsunder the commerce clause between 1875 and 1890.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 12:56:36 PM
All use subject to JSTOR Terms and Conditions
Economic History Association
Exclusive Dealing and the Whiskey Trust, 1890-1895
Author(s): Werner Troesken
Source: The Journal of Economic History, Vol. 58, No. 3 (Sep., 1998), pp. 755-778
Published by: Cambridge University Press on behalf of the Economic History Association
Stable URL: http://www.jstor.org/stable/2566623 .
Accessed: 17/11/2013 11:31
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
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 support@jstor.org.
.
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
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Exclusive Dealing and the WhiskeyTrust,
1890-1895
WERNERTROESKEN
This articleuses the historyof the WhiskeyTrustto explorethe competitiveeffects
of verticalrestraintssuch as exclusivedealing.The WhiskeyTrustdistilledalcoholic
spirits and bribed distributorsnot to carry competing brands of spirits. For the
WhiskeyTrust,exclusivedealingwas an ineffectivepredatorystrategy.Despite the
trust's marketdominanceand manifold predatorystrategies,it failed to preempt
entry.The trustfailed,in part,becauseits rivalscouldverticallyintegrateat low cost.
Competitiondisciplinedthe trustmoreeffectivelythandid numerousantitrustsuits.
ome economistsarguethatfirmsuse verticalrestraintsto increasetheir
rivals' costs and deter entry.' Considera frequentexclusive dealing
arrangement:contractsrequiringdistributorsto carryonly the productsof
A largemanufacturer
a particularmanufacturer.
mightuse such contractsto
undercutthe ability of his rivals to compete. If the manufacturercontrols
This becomes
distribution,his rivalshave to opentheirown distributorships.
S
a costly endeavor when distribution outlets are scarce. Other economists,
however, argue that vertical restraints promote efficiency. Among other
things, vertical restraintsmight protect relationship-specific investments and
prevent competitors from free riding on a firm's advertising expenditures.2
Moreover, even if firms use vertical restraints strategically, it is not clear
thatsuchrestraintspromotesubstantialmarketpowerover the long run.The
debate over vertical restraintsand exclusionarypracticesis a subset of a
largerdebateover the abilityof firmsto use strategicbehaviorto createand
sustainmarketpower.
Although the ratio of theory to evidence is high, there is a growing empirical literature on the competitive effects of vertical restraints.3This article
presents a case study on the Whiskey Trust to explore the sources and consequences of exclusive dealing. Exclusive dealing has received scant empiriThe Journal of Economic History. Vol. 58, No. 3 (Sept. 1998). C The Economic History
Association. All rights reserved. ISSN 0022-0507.
Werner Troesken is Associate Professor of History and Economics, Department of History,
University of Pittsburgh, Pittsburgh PA 15260. e-mail: troesken+@pitt.edu
I gratefully acknowledge helpful comments from Patty Beeson, Karen Clay, Dave DeJong, Martin
Gaynor, Jim Lesage, John Murray,Timothy Sass, Lester Telser, seminar participants at the University
of Toledo, and two anonymous referees from this journal. The usual disclaimer applies.
'Krattenmaker and Salop, "Competition"; and Salop and Scheffnan, "Raising Rivals' Costs."
2On general efficiency arguments, see Bork, Antitrust Paradox, pp. 193-207; Posner, Antitrust Law,
pp. 171-205; Lopatka and Godek, "Another Look"; and Masten and Snyder, "United States." See also,
Mathewson and Winter, "Competitive Effects." For advertising and exclusive dealing, see Marvel,
"Exclusive Dealing."
3See, Granitz and Klein, "Monopolization"; Lopatka and Godek, "Another Look"; Masten and
Snyder, "United States"; and Sass and Saurman, "Mandated Exclusive Territories."
755
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
756
Troesken
cal attention, case study or otherwise. The Whiskey Trust distilled alcoholic
spirits. It sold the spirits to rectifying houses who blended them with flavorings to produce various brand-name whiskies. Rectifying houses then sold
the brand-name whiskey to wholesalers who distributed the whiskey to
consumers and retailers. During the early 1890s, the Whiskey Trust conducted an unusual experiment in exclusive dealing. The trust did not contract with the rectifying houses who bought its spirits; it contracted with the
wholesalers who bought the rectifiers' brand-namewhiskies. Starklyput, the
trust bribed wholesalers not to distributerectified whiskey made with spirits
from nontrust distilleries. Less starkly,it offered wholesalers substantial rebates if they would deal only with those rectifiers who purchased trust-distilled spirits.
The experience of the Whiskey Trust suggests that exclusive dealing is an
ineffective predatory strategy. Despite the trust's market dominance and
manifold predatory strategies, its rebate program failed to preempt entry.
The rebate program failed, in part, because the trust's rivals could vertically
integrate at low cost. Competition disciplined the trust more effectively than
did the numerous state and federal antitrust suits brought against the trust
and its rebate program. Although the story of the Whiskey Trust bears primarily on the debate over exclusive dealing, it has implications for other
debates as well. It contributesto the debate over the effectiveness of antitrust
regulation; it contributes to a growing literature on state regulation; and it
sheds light on nagging questions about the late-nineteenth-century trust
movement.4 For example, it explores some of the competitive strategies
employed by the trusts. Also of particularinterest is the article's analysis of
the price of spirits over time. Previous studies of the trusts have been able
to present only limited evidence on the trust movement's effect on prices.
Much of the evidence for this study comes from an investigation conducted by the United States Industrial Commission in 1899. Appointed by
Congress to investigate large industrialtrusts, the IndustrialCommission included prominent economists and academics. Perhapsthe best-known member of the commission was Jeremiah Jenks, who wrote several books and
articles about the trusts and other pressing economic questions.'
4The literature on the desirability of antitrust regulation is voluminous. I cite only a few examples.
Kovaleff(Antitrust Impulse) presents a series of essays that generally agree that antitrust regulation is
effective and necessary. Stigler ("Economic Effects") argues that the antitrust laws have been, at best,
mildly effective. Bittlingmayer ("Stock Market")argues that antitrustregulation is pernicious. See also
the brief survey of empirical studies of antitrustin Troesken, "AntitrustEnforcement."A growing number of economists and economic historians have come to appreciate the need for greater study of state
regulation during the late nineteenth centuty. Several of the essays found in Goldin and Libecap's
Regulated Economy illustratethis. There are a range of interpretations of the trust movement. See, for
example, Chandler, VisibleHand; Lamoreaux, Great Merger Movement; James, "Structural Change";
Libecap, "Rise"; McCraw, Regulation; and Telser, Theory, pp. 19-40.
'See U.S. House, Industrial Commission Reports. Hereafterreferred to as ICR. For a survey of how
economists during this period saw the trusts, see DiLorenzo and High, "Antitrust." Other sources used
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Exclusive Dealing and the WhiskeyTrust
757
THE STRUCTUREOF THE WHISKEYINDUSTRY
Therewere two typesof whiskey,straightwhiskeyandrectifiedwhiskey.
StraightwhiskeyincludedKentuckybourbonandrye whiskey,the latterdistilled almostentirelyin Pennsylvaniaand Maryland.Straightwhiskey was
distilled from corn, rye, and malt. It requiredat least threeyears of aging.
The best whiskiesagedin oakbarrelsfor up to seven years.Aboutone-third
of all whiskeyproducedin the UnitedStateswas straightwhiskey.Cheaper
and poorer tasting than straightwhiskey, rectified whiskey requiredno
aging. It was madeby blendingalcoholic spiritswith water,brownsugar,a
small amountof straightwhiskey, and other flavorings.Industryofficials
claimedthatrectifiedwhiskey did not competewith straightwhiskey.The
two served separatemarkets.The marketsfor both rectifiedand straight
whiskeywere large.Duringthe late nineteenthcentury,per capitawhiskey
consumptionaveragedover one gallon a year.6
The productionand distributionof rectifiedwhiskey can be dividedinto
fourstages.In stageone, distillersmadealcoholicspirits,whichweretheprimaryinputin rectifiedwhiskey.Spiritshad virtuallyno otheruse except as
an inputformakingwhiskey.Distillersfermentedcornintoalcohol,thenran
the undilutedalcohol throughcharcoalto removevariousoils and flavors,
leavingspirits.Spiritswerehomogeneous.As one industryofficialobserved,
"therewas no such thing as a brandof spirits."Besides a small amountof
maltused to initiatefermentation,cornwas the primaryinputinto spirits.7
In stagetwo, rectifyinghousesblendedthe spiritswith flavoringaccording
to theirrespectivetrademarks
andbrandnames.Brandnamesandtrademarks
may have helped assure quality.Rectifying, the IndustrialCommission
argued,was more profitablethandistillingbecausebrandnames differentiatedvariousrectifiedwhiskies.Rectifiersalsoperformedthe importanttask
of removing fusel oil from the spirits.Containingsmall amountsof amyl,
butyl,andpropylalcohol,fuseloil was poisonous.In stagethree,wholesalers
purchasedwhiskeyfromrectifyinghouses andthendistributedthe whiskey
to retail outlets. In stage four,retailoutlets sold the whiskey to final consumers.8
Licensing requirementsincreasedthe costs of integratingacross these
stages. Testifyingbefore a Congressionalinquiryin 1893, J. B. Greenhut,
include newspaper accounts, court reporters, and a Congressional investigation of the trust conducted
in 1893. See U.S. House, Whiskey TrustInvestigation. Hereafter referred to as Whiskey Trust.
6ICR, pp. 75-79, 168, 204-05, 258, and 842-43.
7ICR, pp. 168 and 835.
8ICR, p. 75. Some observers claimed that rectifying houses adulterated their whiskey by watering
it down excessively and by flavoring it with unhealthy chemicals. The truthfulness of such claims
appears dubious. See ICR, pp. 75-76, 204-05, 230-31 and Whiskey Trust, pp. 3-24, 43, 82-98. See
also, High and Coppin, "Wiley." High and Coppin provide much information about the nature of the
whiskey industry and about claims of "impure" whiskey.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
758
Troesken
the president of the Whiskey Trust, explained that if a distiller acted as a
rectifier "in any shape, form, or manner,"the internal revenue would seize
all of its property. Greenhut continued: "There are two kinds of licenses
which the Govenment issues, one a rectifying license, which is more expensive, and the other is a wholesale liquor dealer's license. The ordinary
wholesale dealer does not rectify."9These licensing requirements forced a
firm operating at one stage of the production process to incorporatea separate company to operate at another stage of production. For example, wholesalers often incorporated separate organizations to operate distilleries. Although the wholesaler may have owned and operated the distillery, the distilling and wholesaling operations were legally and organizationallydistinct.
In light of the stages of production described previously, the phrase
"Whiskey Trust"is a misnomer. The trust did not operate at stage two, rectifying brand-name whiskey, as the phrase Whiskey Trust suggests. The
trust's fornal name, the Distilling and Cattle Feeding Company, is more accurate. A combination of distillers, the trust made alcoholic spirits.
Most distilleries located around Peoria, Illinois. By locating around
Peoria, a region rich with corn, distillers avoided the costs of transporting
a key input. Another importantconsiderationwas the region's water supply.
Peoria had a limitless "supply of cold water running at a temperature of
about 54" degrees. Watertemperaturewas a "dominant factor" in distilling
and 54 degrees was near perfect.10
The federal government taxed domestic and imported spirits. The import
tariff was so large that the United States did not import any spirits or rectified whiskey. This insulated distilleries from foreign competition and protected all but the most expensive brands of straight whiskey.11 Only the
highest grades of foreign straight whiskies were imported. The tax on domestic spirits was also large. During the late 1890s, the federal tax on spirits
was $1.10 per gallon. At the time, it cost 8 to 15 cents to produce one gallon
of spirits. A vibrant trade in illicit spirits grew from the tax. There are, unfortunately,no precise data on the production of stills. The InternalRevenue
Service, which collected the tax, only reported the number of illicit stills
seized. By the late 1890s internalrevenue agents were seizing two thousand
stills per year. Tax-paying distillers claimed that the internalrevenue discovered only a "small proportion" of all illicit stills. 12
9Whiskey Trust,p. 43.
'?Information and quotations, from ICR, pp. 201-02.
"During the nineteenth century, many observers argued that tariffs allowed the trusts to form and
raise prices. For example, during the Congressional debate over the Sherman Antitrust Act, many
Democrats advocated lowering the tariff as a means of combating monopolistic trusts. If the tariff were
lowered, only trusts based on genuine cost efficiencies would have formed. See DiLorenzo, "Origins";
DiLorenzo and High, "Antitrust";and Grandy, "Original Intent."
'2ICR, pp. 90-91 and 817-40.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Exclusive Dealing and the WhiskeyTrust
759
Industry observers argued that barriers to entry were low. According to
the Industrial Commission, "the cost of establishing a new distillery" was
"slight." "The ease with which new distilleries" were established made it
"almost impossible" for the trust to hold "control of the business." Only if
distillers "keptprices low," the commission explained, would they not "provoke competition." It is surprising that industry observers described entry
as easy, because there were a number of factors that otherwise would have
inhibited entry, including evidence of scale economies in distilling; licensing
requirements; and brand name loyalty."3There was, however, one type of
entry that clearly was easy and low-cost: the creation of small and illicit
stills. Competition from illicit stills may have constrained the trust.
THE ORIGINSOF THE WHISKEYTRUST
Distillers began forming pools during the early 1880s. Through pooling
arrangements, distillers agreed to limit their production. The pools failed
because of market entry and the absence of effective sanctions for members
who defected. In 1887, after pooling had failed, distillers organized the Distillers and Cattle Feeders' Trust. Unlike many other combinations that were
also called trusts, the Distillers and Cattle Feeders' Trust was a bona fide
trust. Distillers who agreed to join the trust gave their stock to a board of
trustees. The trustees, in return, gave the distillers certificates representing
their shares in the trust. Once a distiller joined the trust, it was supposed to
follow the managerial decisions of the trustees. Of the 86 distilleries that
eventually joined the trust, the trust kept only 10 or 12 operating and shut
down the others. Sometimes the distilleries the trust shut down would reopen and compete with the operatingmembers of the trust. To prevent this,
the trust often leased the ground and plant of member distilleries for up to
25 years. It would then remove or destroy the machinery in the plant, leaving the distillery inoperative.14
During the 1880s state courts raised questions about the legality of trustarrangements.Fearingthat state courtswould eventually disbandits trustagreement, the Distillers and Cattle Feeders' Trust reorganized as an Illinois corporation, the Distilling and Cattle Feeding Company, in 1890. Though the
13See ICR, pp. 81 and 87 for quotations regarding barriers to entry. On economies of scale, see
James, "StructuralChange," p. 445; and ICR, pp. 88-89. For a critical analysis of the concept of entry
barriers, see Demsetz, "Barriers."
14ICR, pp. 76, 168-69, 220, and 828. On the relationship between price wars and pooling, see
Lamoreaux, Great Merger Movement; and Porter,"Study." On pools as antecedents of large industrial
combinations, see Lamoreaux, Great Merger Movement. As to why the trust shut down some distilleries, Charles A. Clarke, who operated a trust-affiliated distillery, said that the Whiskey Trust economized on costs by shutting down or scaling back production from inefficient and poorly located plants.
Samuel M. Rice, a distributerfor the trust, and John McNulta, suggested that the trust wanted to reduce
output and drive up prices. For Clarke's testimony, see ICR, pp. 170. For the testimony of McNulta,
see ICR, pp. 196-97, 203, and 216. For Rice's testimony, see ICR, p. 832. See also, ICR, pp. 88-89.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Troesken
760
people continuedto call
new companyhadabandonedthetrustarrangement,
it the WhiskeyTrust.Afterit reorganizedandthe WhiskeyTrustcontinued
to expandits controlover the whiskey industry.Duringthe early 1890s it
bought six new distilleries,includingtwo Chicago distilleries that were
among the largestfive distilleriesin the world.Withthese acquisitions,the
trustcame to produce95 percentof all the spirits(legally)producedin the
United States.'5
A DESCRIPTION OF THE REBATE PROGRAM
In the summerof 1890 the trustbegan its exclusionaryrebateprogram.
Under the programthe trustpaid wholesalersa large rebateif they purchasedwhiskeyexclusivelyfroma selectgroupof rectifyinghouses. Every
time a wholesalerpurchasedwhiskey from a trust-selectedrectifier,the
wholesalerreceiveda voucher.If, aftersix months,the wholesalerhadpurchasedonly from trust-selectedrectifiers,the wholesalercould redeemthe
voucherfor a rebate.The size of the totalrebatewas based on the amount
of spirits containedin each brandof whiskey. Supposea wholesalerpurchasedten gallonsof whiskeyfroma trust-selectedrectifier.If thoseten gallons of whiskey containednine gallons of spirits,the whiskey would have
contained"nineproofgallons."Between 1890 and 1891 the trustset the rebate at five centsperproofgallon. Between 1891 and the summerof 1894,
it raisedtherebateto sevencentsperproofgallon.Duringthe fall andwinter
of 1894the trustreducedtherebateto two centsperproofgallon.Compared
to the priceof spirits,the rebatewas substantial.At the time, the before-tax
price of spiritsrangedfrom 10 to 30 centsper gallon;the after-taxprice of
spirits ranged from $1 to $1.20.16
Cheating-wholesalers buying from nontrustrectifiers and then redeeming their rebate vouchers anyway-was a potential problem. Monitoring
costs, however, were not prohibitive. The InternalRevenue Service required
wholesalers to keep detailed records of all of their purchases. If wholesalers
failed to keep these records, or kept them inaccurately, they risked having
"5Forthe court cases raising questions about the legality of the trust, see State v. Nebraska Distilling
Company et al., 29 Neb. 700 (1890); People v. The American Sugar Refining Company, 7 Ry. & Corp.
L.J. 83 (1890); State v. American Cotton-Seed Oil Trust, 40 La. Ann. 409 (1888); People v. Chicago
Gas Trust Company, 130 Ill. 268 (1889); and People v. North River Sugar Refining Company, 121
N.Y. 582 (1890). For an analysis of the effectiveness of these suits, see Troesken, "Antitrust
Enforcement." Also, testifying before the IndustrialCommission, Charles Clarke explained: "the New
York courts had declared the sugar trust an illegal combination; and in order to avoid the same thing
... we organized the corporation known as the Distilling and Cattle Feeding Company." Quoted in
ICR, p. 171. For growth of the Whiskey Trust, see ICR, pp. 75-90 and the discussion of the Shufeldt
distillery later in the article.
'6ICR, pp. 84, 171-72, and 241-42; and the following court cases: Olmstead et al. v. Distilling &
Cattle Feeding Co., 67 F. 24 (1895); Olmstead v. Distilling & Cattle-Feeding Company; Graves v.
Same; Bayer v. Same; 73 F. 44 (1895).
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Exclusive Dealing and the WhiskeyTrust
761
their operating licenses revoked. The trustmonitored wholesalers with these
records. 17
Beyond the wholesale rebate, the trust also offered a rebate to rectifiers.
This rebate, however, was not explicitly exclusionary. For every gallon of
spirits they purchased, trust-selected rectifying houses simply received two
cents off the list price of spirits. They received the rebate whether or not
they purchased exclusively from the trust. However, trust-selected rectifiers
faced a strong incentive to continue buying most of their spirits from the
trust. If they stopped, or bought too much from competing distilleries, they
ran the risk that the trustwould remove them from the wholesale rebate program. Rectifiers did not want to be removed from the program, because the
trust gave participatingrectifiers market power by subsidizing distributors'
purchases from them.'8
The Whiskey Trust abandoned the rebate plan when it entered receivership in January 1895. The trust reorganized as the American Spirits Manufacturing Company in August 1895. Instead of reviving the rebate plan,
American Spirits chose a new form of vertical integration. In January 1896
American Spirits organized another,separatecompany, the Spirits Distributing Company. The Spirits DistributingCompany was a combination of rectifying houses. American Spirits organized the company "as a means for
securing the control of the [rectifying] business." According to the Industrial
Commission, "all of the common stock" of the Spirts Distributing Company
"belonged to the American Spirits Manufacturing Company." American
Spirits also under-wrote all of the preferredstock of the Spirits Distributing
Company.19
INTERPRETINGTHE REBATEPROGRAM
How Historical Observers Saw the Rebate Program
The Industrial Commission asked John McNulta, the court-appointed
receiver of the Whiskey Trust, how he thought the rebate program affected
the whiskey business. McNulta answered: "I do not think its effect was
good. It demoralized the trade and created bad feeling among the customers,
so that there was a general disposition to get out of it." McNulta continued,
explaining that the trust startedthe rebate system at a time when it had control over "almost the entire market." "There was," McNulta claimed, "no
alternative for buyers except to start distilleries of their own, and so they
submitted to it, but always regardedit as a great burden" [emphasis added].
17See ICR, pp. 83-84 and 172.
"8Ibid.,pp. 84, 171-72, and 241-42; and the following court cases: Olmstead et al. v. Distilling &
Cattle Feeding Co., 67 F. 24 (1895); Olmstead v. Distilling & Cattle-Feeding Company; Graves v.
Same; Bayer v. Same; 73 F. 44 (1895).
191CR,pp. 78 and 835.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
762
Troesken
When asked, "Did the company institute the rebate system solely for the
purpose of controlling the output?"McNulta answered "Yes." Investigators
then asked McNulta if wholesalers would ever again consent to an exclusive
rebate system. McNulta testified: "I think it hardly possible ever to start the
rebate system again, at least within the memory of the men who had experience with it." Other witnesses before the IndustrialCommission agreed with
McNulta's assessment of the rebate program.20
McNulta's testimony implies that the trust used the rebate program to
increase its marketpower. However, it is not clear from this testimony how
the rebate program would have succeeded in this. The testimony also suggests that wholesalers and rectifiers were dissatisfied with the rebate program and wanted the trust to abandon it. This is puzzling. If wholesalers and
rectifiers did not like the rebate program, why did they consent to it? The
discussion below clarifies these issues.
The Rebate Program in Light of CurrentEconomic Theory
In the usual anticompetitive story, manufacturersuse exclusive dealing to
foreclose scarce distributing outlets. This makes it costly for new firms to
enter. If new companies decide to enter, they must do so as vertically integrated enterprises, operating both as manufacturersand as distributors, because incumbents have tied up all distributingoutlets. The Whiskey Trust's
rebate program went the usual story one better. If new distilleries wanted to
enter, they not only needed to integrateone step forward, into rectifying, but
two steps forward, into rectifying and wholesaling.
Suppose a new distillery opened. Large rectifying houses that participated
in the trust's rebate program would have been reluctantto purchase the new
distillery's spirits. If they did, they ran the risk of the trust removing them
from their list of selected rectifiers. Rectifiers did not want to be removed
from the list because the rebate program gave them marketpower-by subsidizing wholesalers' purchases from specified rectifying houses, the rebate
program made the specified rectifiers more attractive to wholesalers. Even
if the new distiller managed to find a rectifier to purchase its spirits, that rectifier would have had a hard time finding a wholesale outlet for his whiskey.
After all, wholesalers had a strong incentive to purchase only from trustselected rectifiers. With few wholesale outlets and limited demand for its
whiskey, the rectifier would, in turn, have had only a limited demand for the
new distillery's spirits.
In this way the rebate programraised the costs of entry and insulated the
Whiskey Trust from competition. Protected from competition, the trust
20A11of McNulta's statements are taken from ICR, p. 207. For the testimony of others documenting
McNulta's statements, see ICR, pp. 172, 181, 241, and 815. See also, Whiskey Trust, p. iii.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Exclusive Dealing and the WhiskeyTrust
763
could have increasedthe priceof spirits.Dependingon the size of the price
increase,the rebateprogramcould have left rectifyinghouses and wholesalersworse off. An increasein the priceof spiritswouldhave increasedthe
costs of rectifying.Some of the increasewould have manifesteditself in
higher whiskey prices and would have been passed along to wholesalers.
Some of it, however,maynot havebeenpassedalong. In particular,rectifiers would have sufferedlosses from the price increaseif industrydemand
was not perfectly inelastic or if they had marketpower.21Although the
trust'srebateprogramincreasedthe marketpower of its chosen rectifying
houses, it is not clear if the benefits of that increasemore than offset the
losses causedby the increasein the price of spirits.
For wholesalers, the rebatereduced the price of whiskey from trustselectedrectifiers.Initially,the rebatemusthave been largeenoughto offset
any increasein the price of whiskey thatresultedfrom an increasein the
price of spirits. If it were not, wholesalerssimply would have purchased
rectifiers.However,as the amountof whiswhiskeyfromnon-trust-selected
key purchasedrose, wholesalerswould have foundit increasinglycostly to
abandontrust-selectedrectifiersin favorof independents.Therebatevouchers could only be redeemedif wholesalers continuedto patronizetrustselectedrectifiers.Also, it mighthavebeen costlyto purchasefromnontrust
rectifiersbecause,as notedearlier,theywere smallerandmighthave manufacturedthe leastpopularbrandsof whiskey.Overtime,then,the trustmight
have been able to drive up the price of whiskey to a point where it overwhelmedthe rebate.
BrandNames,LicensingCosts,and theEffectivenessof theRebateStrategy
In the long runthe effectivenessof the trust'srebatestrategydependedon
the costs of verticalintegration.If new distilleriescould easily open their
own rectifyinghousesandwholesaleoutlets,the rebateprogramwould not
have deterredentry.Alternatively,if new distilleriescould not easily open
theirown rectifyinghouses,and all existing rectifierswere beholdento the
trust,new distillerieswould not have been able to find a marketfor their
spirits.This would have deterredentry.
Two factors,mentionedpreviously,wouldhave affectedthe costs of vertical integration:brandnamesand licensingrestrictions.To the degreethat
2'Under the first scenario,assumethe marketfor rectifiedwhiskey was perfectlycompetitiveand
industrydemandwas notperfectlyinelastic.In thiscase,an increasein thepriceof spirits(a key input)
would have shiftedthe supplycurveup. Equilibriumpricewould have been higherand equilibrium
outputwouldhavebeenlower.Therewouldhavebeenfewerrectifiersproducingless whiskey.Under
the second scenario,assumethe marketfor rectifiedwhiskeywas not perfectlycompetitiveand that
each rectifierhad some marketpower.(Perhapsthis marketpowerderivedfromthe rectifier'sbrand
nameor froma locationaladvantage.)An increasein thepriceof spiritswouldhave shiftedcost curves
up and reducedeach rectifier'sprofitsandprofitmaximizingoutput.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
764
Troesken
loyalty to existing brandsof whiskey kept consumersand retailersfrom
switching to new brandsof whiskey,it would have been difficultfor new
distilleries to open their own rectifyinghouses and createnew brandsof
whiskey. To the degreethatthe licensingprocess was expensive or politicized, it would have been difficult for new distilleriesto open their own
rectifyinghouses.The cost of licensingis perhapsthe most relevantconsideration.Regardlessof how loyal consumerswere to specific brandsor how
easily new distilleriescould find wholesale outlets, existing rectifiersand
wholesalerscould have openedtheir own distilleriesto competewith the
trust.AlthoughI have no directevidenceon licensingcosts andbrandname
loyalty,the historicaldiscussionbelow shows thatwholesalersfoundways
to open theirown distilleriesandrectifyinghouses. Given this, the costs of
verticalintegrationdo not appearto have been prohibitive.
THE REBATEPROGRAMAND THE PRICEOF SPIRITS:A TIME SERIES
ANALYSIS
Testifyingbeforethe IndustrialCommission,industryofficials suggested
a trend:initially,the trustchargedmoderateto low prices;later,it charged
high prices. MartinR. Cook, a rectifierfromNew YorkCity,testifiedthat
originallythe trust"maintaineda fairprice andwas willing to accepta fair
profit."Only laterwere prices "advanced."Cook also arguedthatthe advancewas not causedby an increasein the price of corn or any otherinput.
Receiver JohnMcNultatestifiedthat "afterthe concernreachedthe point
where [it] controlled a large proportion of the output of spirits," it "tried to
push prices up." Otherindustryobservers,testifying before Congress in 1893,
expressed the identical sentiments.22The data supportthese observations.
DATAAND ANALYSIS
Data Sources
As noted earlier, most distilleries located in Peoria. In its final report to
Congress, the Industrial Commission used data from the Peoria Board of
Trade to construct a time series of the price of corn and spirits. The commission's data are monthly and extend from January 1884 through December 1899. Data on corn prices are especially useful because corn was such
an importantinput into the production of alcoholic spirits. McNulta argued
that spirits were "corn in liquid form" and that their cost was "based upon
the cost of corn."23
22All of the quotations are from ICR: Cook's, p. 241; and McNulta's, p. 207. For other testimony
corroborating these views, see Whiskey Trust, pp. 11-13; and ICR, p. 812.
23Thedata are from ICR, pp. 813-16. McNulta's quotation is from ICR, p. 239.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Exclusive Dealing and the WhiskeyTrust
765
Estimating Procedure
In his study of Standard Oil and its effect on consumer prices, Lester
Telser used data very similar to the data here. Using monthly price data,
Telser regressed the price of refined oil against the price of crude oil, a time
trend, a dummy variable for the StandardOil Trust, and a dummy-time trend
interaction term.24Following Telser, I estimated the following equation to
identify the effects of the Whiskey Trust and its rebate program:
SPIRITSt= a + PI(CORN) + P2(TAXt)+ P3(TRUSTt)+ N4(REBATEt)+
. .. +
P5(TIME)+ P6(REBATEt*TIME)+ Et
The variables are defined as follows. SPIRITS is the natural log of the
price of spirits (net of tax, and after the rebate). CORN is the natural log of
the price of corn. TAXis the natural log of the internal tax on spirits. REBATE is the rebate (in levels) offered by the Whiskey Trust. TIME is a time
trend; and Et is an error term. All dollar values have been adjusted for
changes in the general price level. The base month is January,1884. TRUST,
a dummy variable, identifies how the trust influenced prices during the prerebate period. It assumes a value of one for a period of 36 months between
1887 and 1890 and zero otherwise. During these 36 months the trust operated but did not offer any rebates. The interactionterm, REBATE* TIME, is
included because theory and congressional testimony suggest that the
effectiveness of the rebate would have varied over time. To induce people
to participate in the rebate program, the trust would have had to offer very
low prices initially Later, after it had lured people in, it would have tried to
raise prices. This logic predicts a positive coefficient on REBATE*TIME.
The Data and Implications for Estimation
Figure 1 plots the natural log of the real price of spirits (SPIRITS) and
corn (CORN). This figure suggests that the price of spirits and the price of
corn were highly correlated, a pattern consistent with the comments of industry observers who argued that corn was the primary input in the production spirits. Figure 2 plots the naturallog of the real federal tax on spirits and
the trust's rebate in levels. As the figure suggests, the government changed
the nominal tax on spirits only once, raising the tax from 90 cents to $1.10
in 1894. All other movements in the real tax are generated by changes in the
general price level. The rebate, in contrast, was changed more frequently. It
equaled zero between 1884 and 1890 and between 1895 and 1899. In nominal terms, the rebate varied between two cents and seven cents per proof gallon from 1890 through 1894.
24See Telser, Theory, pp. 36-40.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Troesken
766
0
Corn
0)
0)
-1.5-3. -
z
-2.5Spirits
1884
1886
1888
1890
1892
1894
1896
1898
FIGURE 1
NATURAL LOG OF THE REAL PRICES OF CORN AND SPIRITS, EXCLUDING TAXES
AND REBATES
In a preliminaryanalysis, the data were tested for nonstationarityand
two frequentproblemswhenworkingwith time-seriesdata.
autocorrelation,
Using an augmented Dickey-Fuller test, the hypothesis that the data
(SPIRITS)are stationaryat the 5 percentlevel of significance cannot be
rejected.25Preliminarytests, however,indicatedsecond orderautocorrelation. Autocorrelation
mighthavestemmedfromimproperfunctionalspecification.To controlfor this possibility,the equationwas estimatedwith and
withoutlogarithmictransformations.
Functionalformdid not matter;autocorrelationappearedregardless.Only the resultsfor the logarithmicmodel
arereported.(The otherresults,which are identicalin substance,are available uponrequest).Omittedvariablesare anothercommonsourceof autocorrelation.But it is unlikely that this would have been a concern here.
Based on the testimonyof witnessesbeforethe industrialcommission,corn
prices,the internaltax, andthe rebateappearto havebeen the primarydeterminantsof the priceof spirits.All of thesewere includedin the regressions.
Nonetheless,to controlfor the remotepossibilitythatsomethingimportanthadbeen omitted,I estimateda secondmodelthatincludeda lag of the
25The test results are not included here, but are available from the author upon request. The
estimating procedure followed Pindyck and Rubinfeld, Econometric Models, pp. 459-65. They report
the relevant critical values in table 15.1.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Exclusive Dealing and the WhiskeyTrust
767
0.3
Tax
; 0.15
~~~~~~~~Rebate
0
0
-0.15
,
1884
1886
1888
1890
1892
1894
.
.
1896
.
1898
FIGURE 2
NATURAL LOG OF REAL TAX AND LEVEL OF REAL REBATE
dependent variable. I also surveyed the New YorkTimes Index for events
that might have affected the price of spirits. Based on this survey, two dummy variables were coded, PRICEWARtand SPECULATIONt.PRICEWARt
controls for price wars that occurred during the mid-1880s. SPECULATION controls for changes in prices induced by speculators who anticipated the large increase in the internaltax on spirits. I then estimated a third
equation that included these dummy variables. Including SPECULATION
and PRICEWARSimproved the fit, but did not eliminate the autocorrelation.
Because these remedial steps did not eliminate the serial correlation, I controlled for it using two different estimating procedures. For those models
without a lagged dependent variable, Cochrane Orcott was used. For the
model with a lagged dependent variable, Dhrymes' estimating procedure
was used. (Cochrane Orcott is inappropriate when the model includes a
lagged dependent variable).26
Results
Table 1 reportsthe regression results. Because of the logarithmic transformations, the coefficients on the price of corn and the internaltax are elasticities. At around two, the price elasticity of spirits with respect to the tax
26See Pindyck and Rubinfeld, Econometric Models, pp. 141-43.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
768
Troesken
TABLE 1
REGRESSION RESULTS
DependentVariable= SPIRITS
Independent
Variables
CORN
TAX
Mean
(Variance)
-0.86
(0.05)
0.06
(0.02)
TIME
TRUST
0.19
REBATE
0.02
(0.001)
REBATE*TIME
SPIRITS(lagged)
(1)
Coefficient
(t-statistics)
0.24**
(2.89)
- 1.91**
(3.89)
(2)
Coefficient
(t-statistics)
0.26**
(3.14)
- 1.98**
(4.12)
(3)
Coefficient
(t-statistics)
0.20**
(2.99)
- 1.97**
(4.79)
0.001
0.001
0.001
(1.03)
-0.10
(1.28)
- 8.43*
(1.95)
0.07*
(1.90)
(1.46)
-0.11
(1.46)
- 8.50**
(2.07)
0.07*
(1.95)
0.09
(1.60)
(1.02)
-0.19**
(2.53)
- 11.75**
(3.03)
0.10**
(2.87)
-1.78
(0.09)
0.55**
SPECULATION
(8.10)
PRICEWARS
-0.23**
- 1.54**
(13.55)
192
0.86
155.20
Constant
N
AdjustedR2
Log of
LikelihoodFn
- 1.42**
(11.38)
192
0.86
155.78
(6.05)
- 1.53**
(14.42)
192
0.91
196.55
** = Significant at 5 percent.
* = Significantat 10 percent.
Note: See the text for variabledefinitions.
appears large. This reflects the fact that when the tax was raised from 90
cents per gallon to $1.10 in August 1894, the price of spirits fell from 27
cents per gallon to 14 cents in the following months. That the increase in the
tax caused the net price of spirits to decline by roughly the same amount
suggests the presence of a nontaxed substitute with a highly elastic supply,
namely illicit spirits. The coefficient on the trust dummy is always negative
but only statistically significant in the third model. The t-statistic on the time
trend (TIME) is low. This might result from collinearity with REBATE*
TIME.27
In all of the specifications, the coefficients on the rebate terms are statistically significant. The rebate variable is negative, whereas the rebate interacted with the time trend is positive. Although these results are suggestive,
27jIn a seriesof regressions
thatI do not reporthere,I also interactthe trustand rebatedummieswith
thepriceof corn.Thatis, I add TRUST*CORN
andREBATE*CORN
to the firstmodel.Addingthese
interactionstermsdoes not alterthe results.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Exclusive Dealing and the WhiskeyTrust
Z
0.25
Trust Formed, Rebates Introduced
Formed, No Rebates
....Trust
-
769
0.23
--
^
No Trust, No Rebates
Trust
Organized
~44 0.21-
0
0.19-
,~0.17
Rebate
Introduced
i
0.15
1884
1
1886
1888
1
1890
1892
1894
1896
1898
FIGURE3
PREDICTED BEFORE-TAX PRICE OF SPIRITS UNDER ALTERNATIVE REGIMES
Figure 3 provides a clearer sense of the economic effect of the rebate. (In
Figure 3, taxes are held constant). Using the estimates from the first regression, the heavy line plots the predicted price of spirits given actual events.
That is, the heavy line plots the fitted price. The pattem is striking. The
estimated price falls steadily until the introduction of the rebate program
when there is a sharp one time drop in price. Presumably the trust cut prices
to lure potential customers into the rebate program. However, shortly after
the rebate is introduced, the downward trend stops and estimated prices
begin to climb. According to the coefficient estimates from the first regression, the trust had increased prices to the point where they overwhelmed the
rebate by December 1893.28 This pattern corroborates the testimony before
the IndustrialCommission: initially the trust charged low prices but it raised
prices after the rebate program began.
In Figure 3, the lighter solid line plots what prices would have been if distillers had not organized the trust and continued setting prices as they had in
the years before the trust.Note that the line begins in June 1887, the first full
month that the trust operated. Comparing the two lines, one sees that the
price of spirits would have been about 10 to 15 percent higher without the
trust. This pattern is consistent with the hypothesis that, before the rebate,
the trust exploited scale economies and brought consumers lower prices.
28Using the estimates from the first regression, equate the expression (8.43*REBATE) with the
expression (0.07*REBATE*TIME). With this, (TIME) = (8.43)/(0.07) a120. The 120th observation
is December 1893.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
770
Troesken
Without direct evidence that the trust actually did things to reduce costs,
however, it is hard to put much stock in this. A more plausible hypothesis,
also consistent with the low prices before the rebate, is that the trust was
practicing predatory pricing during this period. The statements of industry
observers and the trust's own managers-all of whom claimed that the trust
practiced predatory pricing-provide additional corroboration.
The dotted line shows what would have happened if, ratherthan introducing the rebate, the trust had continued with its prerebate pricing regime.
Note that the line begins in June 1890 when the trust introduced the rebate
program. Comparingthe dotted line and the heavy line shows the predicted
effect of the rebate program. It is clear that predicted prices fell by 10 percent when the trust first introduced the rebate program. However, as noted
above, predicted prices rose steadily thereafter.Eventually they were 10 to
15 percent higher than they would have been had the trust continued its prerebate pricing regime.
THE TRUST'S OTHERPREDATORYSTRATEGIES
Potential distilleries faced low entrybarriers.The trusthoped that its rebate
programwould make it more difficult for competitorsto startnew distilleries.
It also used predatorypricing and threats of violence to discourage entry.
Violence
On 11 February 1891, Chicago police arrestedGeorge Gibson, the secretary of the Whiskey Trust.Thomas Dewar initiated Gibson's arrest.According to Dewar, several weeks earlier Gibson had offered him $25,000 to blow
up a Chicago whiskey distillery. Dewar, an agent of the Internal Revenue
Service, worked with other treasury agents to lay a trap for Gibson. When
Gibson arrived at the GrandPacific Hotel to meet Dewar, police were waiting. State and federal authoritiescharged Gibson with, among other things,
attempted arson and conspiracy to commit murder.29
Dewar could have fabricatedthe whole story. Before joining the Internal
Revenue Service, Dewar worked for the Whiskey Trust, where Gibson had
been his boss. Dewar vacated his position with the trust after only a few
months. Gibson said he was fired; Dewar said he quit. Perhaps Dewar was
a disgruntled employee, bent on destroying his former boss.30
More likely, Dewar told the truth. There were means. When police arrested Gibson they confiscated his grip and found inside an explosive com29See,Dewar's testimony before Congress in Whiskey Trust, pp. 16-30; East, "Distillers and Cattle
Feeders' Trust;" and the Chicago Tribune, 5 June 1891, p. 1.
301t is interesting to compare Dewar's testimony before Congress to that of J. B. Greenhut, the
president of the Whiskey Trust. For Dewar's testimony, see Whiskey Trust, pp. 17-30; for Greenhut's
see ibid., pp. 31-57.
This content downloaded from 130.126.32.13 on Sun, 17 Nov 2013 11:31:11 AM
All use subject to JSTOR Terms and Conditions
Exclusive Dealing and the WhiskeyTrust
771
pound. After searchingGibson's other baggage, they found a detonating
device-what one investigatorcalledan "infernalmachine."Therewas also
motive. For years the Whiskey Trusthad been tryingto get the Shufeldt
whiskeydistilleryto join the trustbutthe Shufeldthad always refused.The
Shufeldtwas the world's second largestwhiskey distillery.It was also the
distillery Gibson allegedly asked Dewar to destroy.Finally, there was a
suggestivepattern.On 10 December1888, someonethrewdynamiteon to
the roof of the Shufeldtdistillery.The explosion causedconsiderabledamage, but it couldhavebeen muchworse. If the dynamitehad fallen through
a skylight,which it scarcelymissed, it would have destroyedthe distillery.
Mysteriousfires also damagedor destroyedthree distilleriesin southern
Illinois.Eachdistilleryhad refusedto join the trust.Anothertime, workers
at an independentChicagodistillerydiscovered"one of the trust'sagents"
walkingaroundthe plantandtakingnotes. Plantworkersreportedlyplaced
a ropearoundthe agent'sneck andwould have lynchedhim had othersnot
intervened.3"
On 4 June 1891, fourmonthsafterGibson'sarrest,the Shufeldtdistillery
and the Calumetdistillery,anotherlargeChicagodistillery,sold out to the
trust.Everyoneassociatedwith the deal expressed"goodfeelingsand sentiment."The ownersof the trustbelieved thatall these warmfeelings would
"materiallyeffect [sic] the Gibsoncase."Aroundthe sametime, the Tribune
reportedthat"pressurefrominfluentialquartersha[d] been broughtto bear
to have the case 'settled."'Withina year, the state and federal cases
againstGibsonwere resolvedandGibsonresignedas secretaryof the Whiskey Trust. The trust continued to pay Gibson $7,500 a year after he
resigned.32
PredatoryPricing
The IndustrialCommissionwrote:
It was the practiceof the Distillers' Trustand the latercombinationsto send agents
into special localities to undersell competitors,their particularcustomersbeing
approachedand offered open...
Purchase answer to see full
attachment