Varieties in Capitalism, Varieties of Association: Collaborative Learning in American Industry, 1900 to 1925

• Author: Marc Schneiberg,Gerald Berk
• DOI: 10.1177/0032329204272390
• Published date: 01 March 2005
• Date: 01 March 2005
• Subject Matter: Articles

10.1177/0032329204272390POLITICS & SOCIETYGERALD BERK and MARC SCHNEIBERG

Varieties in Capitalism, Varieties of Association:

Collaborative Learning in American Industry,

1900 to 1925

GERALD BERK

MARC SCHNEIBERG

Between 1900 and 1925, the American economy witnessed a remarkably successful

effort to upgrade competition through associations.Unlike the prevailing interpre-

tation of American industrialization, in which associations fell preyto antitrust and

collective action problems, we find many associations that reinvented themselves

from cartels to developmental associations. This transition marked two previously

unrecognized varieties in economic institutions. In the first, associations joined

markets and corporate hierarchiesto create variety in American capitalism. In the

second, associations used deliberation, cost accounting, and benchmarking to

enhance productivity and create varieties of collective governance. This article

explains the origins of developmental associations, outlines their principles, traces

their implementation in the commercial printing industry, and surveystheir distri-

bution and performance effects across 344 industries. Based on these findings, we

revise conventional institutionalist assumptions about order and agency to make

room for institutional diversity and actors’capacities for reflexivity and learning.

Keywords: associations; economic governance; reflexivity; learning; bench-

marking; institutionalism

Between 1900 and 1925, the American economy witnessed a remarkably suc-

cessful effort to restructure and upgrade competition through associations. This

effort began in craft and specialtysectors. But it quickly captured the imagination

For their helpful comments on presentations and previous drafts, the authors thank Rogers

Hollingsworth, Jonathan Zeitlin, Marc Ventresca,Woody Powell, Gary Herrigel, participants in the

POLITICS & SOCIETY, Vol. 33 No. 1, March 2005 46-87

DOI: 10.1177/0032329204272390

© 2005 Sage Publications

46

of a broader coalition of businesses, professionals, and policy makers, who

carried this project into national arenas and into batch and continuous process

industries. Furthermore, in reviving associations, institution builders surpassed

late-nineteenth-century cartels to create something new. These were collabora-

tive learning systems, which used deliberation, cost accounting, and benchmark-

ing to shift competition from volume and cutthroat pricing to innovation and

improvement.

In this article, we analyze the rise, spread, and effects of these collaborative

learning associations. Empirically,we discover new forms of association, and we

find them where conventional accounts would least expect—in a hostile institu-

tional setting, in industries with many firms and diverseinterests, and even in con-

tinuous process sectors. Such findings represent important new anomalies for

uniform institutional images of national economies and American industrial

development.1Moreover, in explaining these anomalies, we revise both conven-

tional images of institutional order and economic agency. Like Thelen, Streeck,

Crouch, and Farell, and Orren and Skowronek, we add tothe growing unease with

“excessive determinism” and the inability of a first generation of neo-

institutionalist theories to cope with novelty,diversity, and change.2Our analysis

affirms the heterogeneous and layered character of institutions, and the ongoing

availability of alternative or abandoned paths. And building on Zeitlin, Sabel,

Stark, and others, we see the relationship between institutions and behavior as

inescapably reflexive—grounded, that is, in experimentation, deliberation, and

learning, rather than calculation within constraints or enactment of taken-for-

granted understandings.3

In conventional institutionalist accounts, American economic development

from 1870 through 1970 is first and foremost the story of the rise of the large,

mass production corporation.4Here, associations are a lost alternative, proving

virtually impossible to sustain.5First, they faced insurmountable barriers to col-

lective action, monitoring, and enforcement. In mass production sectors, high

fixed costs yielded “high powered incentives” for free riding, defection, and

cheating; in specialty sectors, firms were too numerous, dispersed, and heteroge-

neous to cooperate.6To complicate matters, the state not only refused to enforce

interfirm compacts, but also made them positively unlawful through antitrust. As

a result, American business abandoned “loose combinations” for consolidation in

the great merger wave at the turn of the century.7

To be sure, associations reemerged during World War I, the 1920s, and the

early New Deal, experimenting with information sharing and third party enforce-

ment. But like premerger cartels, they are cast as efforts to solve overcapacity

GERALD BERK and MARC SCHNEIBERG 47

SCANCOR Workshop at Stanford University,and the conference, “Constructing Markets, Shaping

Production,” organizedby the Institute of International Business, Stockholm School of Economics,

and the Board of Politics & Society.Research for this article was supported, in part, by the Hagley

Museum and Library, the American Council of Learned Societies, and the Stillman Drake Award.

problems associated with mass production, legitimate only temporarily as expe-

dients for managing war, depression, and decline in stagnant sectors. Here, too,

associations fell prey to opportunism and antitrust. And in the exceptional cases

they succeeded as governance forms, such as in infrastructure, associations are

commonly seen as monopolies, which undermined incentives for efficiency,

innovation, and price reduction.8

While accurate for some sectors, this account mistakes the part for the whole: it

neglects the large portions of American industry that fell outside the orbit of the

corporate consolidations engineered by finance capital at the turn of the twentieth

century. Drawing upon extensive archival research and a data set of 344 indus-

tries, we generate two findings about diversity in American industrialization. We

show first that firms, professionals, and policy makers instituted a viable associa-

tional alternative to mass production in a broad array of industries. As we know,

specialty and batch sectors experienced repeated failures to enforce cartels. But

failure did not always result in open markets. Instead, a number of sectors tried

something different: Printers, bridge builders, and foundrymen, among others,

turned from cooperative price controls to collaborativeefforts to enhance produc-

tivity.They used associations to channel rivalry from cutthroat pricing to product

quality, improvement,and nonvolume cost and price reductions. And as associa-

tions in batch and continuous processing sectors learned from craft innovations,

American business forged a new institutional path, producing previously un-

recognized variety in American capitalism. By 1925, 30 percent of American

manufacturing industries had participated in crafting this path, with 13 percent of

American industries institutionalizing its core forms in whole or in part.

Nor were these associations as conventionally understood. To the contrary,

industries and reformers designed associations to elevate, rather than suppress,

competition through deliberation, benchmarking, and learning. In doing so, they

created qualitativelynew varieties of governance within associationalism. Unlike

corporatist systems, market stabilization schemes, or rent-seeking cartels, these

associations rejected redistribution and age-old efforts to suppress competition.

Moreover, when they turned to information sharing, it was not merely to co-

ordinate prices or production via tacit pricing cues, as applied game theo rists

like Bowman, Culpepper, and Gordon suggest.9Rather, when printers and other

sectors crafted accounting systems, benchmarking schemes, and deliberative

forums, they abandoned monitoring for enforcement or cooperative pricing for

what Sabel calls “learning by monitoring” and Stark and Bruszt refer to as “delib-

erative association.”10 And in making that transition, they retheorized and re-

deployed associations from cartels to what we call developmental associations.

By this we mean governance structures that enhanced members’ capacities to

experiment, learn, and improve productivity and product quality. Developmental

associations had three unique features. They crafted standard accounting sys-

tems, which provided firms with tools to enhance productivity in ways that

48 POLITICS & SOCIETY