Description
In American politics, at least, it is welI known
that government bureaucrats and middle managers
are paid high salaries to shuffle papers.
But we learn from the business pages that public
administrators and business executives process
information, the life blood of the postlndustrial
economy. Readers of academic journals,
especially in accounting, can easily
Pergamon Accounting Organfzatfotas and Sockty, Vol. 20, No. 1, pp. M-92,1995
Copyright 0 1994 Elsevia Science Ltd
Printed in Great Britain. All rights reserved
0361-3682T95 19.50+0.00
INFORMATION CULTURES: A REVIEW ESSAY
THEODORE M. PORTER
Department of Histoty, University of Califomiq Los Angeles
Bemdt, E. R & Iiiplett, J. E. (eds), Ff@y Yems of Economic Measummm k TheJ ubilee of t%e Confelmce
onR-fnfname and wt%mJ (Chicago: university of caicago press, 1990).
Temin, P. (ed. ), Instak the Business Entmprk His-1 Perspectim on the Use of I njxmation (Chicago:
University of Chicago Press, 199 1).
In American politics, at least, it is welI known
that government bureaucrats and middle man-
agers are paid high salaries to shuffle papers.
But we learn from the business pages that public
administrators and business executives process
information, the life blood of the postlndustrial
economy. Readers of academic journals,
especially in accounting, can easily recognize
the element of populist demagogy in the iirst
description. ShuIIling paper is processing
information.
We also tend to believe the converse, that
information flow means precisely the movement
of papers and electronic signals. Knowledge
and information are identified with a highly
formalized, often quantitative, style of expres-
sion. The “information society” is a culture of
logorrhea and numerrhea. This is probably right,
but in a way it is misleading. Early modem
artisans, miners and farmers depended no less
on the transmission of information than do
modem executives and researchers, but they
did not broadcast it indiscriminately. Theirs was
a culture of the face to face. The shift to one of
printed reports and tables, which is and wiIl
never be complete, cannot be understood
simply in terms of the growth of knowledge,
communications technologies, and computing
power. It involved also a change of values - a
move away Tom the personal, and a vast
expansion of the public domain.
THE POWER OF MEASUREMENT
Voltaire once observed that if Cod did not
exist, we would have to invent him. That was
the eighteenth century. In this one, many
entities that cannot be clearly deiined or
delimited must nonetheless be measured. This
may sound cynical, but it is more like boasting.
Fifty Years of Economic Measurement com-
memorates the founding of the Conference on
Research in Income and Wealth, under the
National Bureau of Economic Research (NBER).
Its efforts have helped to give fixity and
coherence not just to vague concepts, but to
vague objects. The volume celebrates these
heroic deeds, this triumph of quantitative clarity
over a fuzzy economy.
The contributors are relatively worldly
members of a decidely unworldly profession,
economics. Of course they do not call for or
applaud the imposition of economic measures
through naked force or scholarly scheming.
Their efforts presuppose that it is possible
to discuss intelligently the advantages and
limitations of various measures, and in this way
to improve them. They are, however, under no
illusion that their scientific difficulties can be
definitively solved through discovery of the true
nature of things. The measurement of inflation
in the currency, for example, has been
vigorously debated by economic statisticians,
83
84 T. M. PORTER
and often scorned by theorists. Still, even well-
founded doubts are not decisive. Questions
about the reality of the things being measured
can be fended off with inverted commas. “How
does one measure something that may not
‘exist’?” asks Charles R. Hulten. Just do it. “Come
hell or high water, many of us will be using and/
or producing measures of aggregate capital
input, even if it does not ‘exist”’ (Berndt &
Triplett, 1990, p. 153). Notwithstanding the
contradictions implicit in any measure of the
value of the currency, write Makoto Ohta and
Zvi Griliches, “the Consumer Price Index (CPI)
‘exists’ and is even of some use” (quoted by
Griliches, ibid., p. 189). There may, as Paul
Pieper argues, be “no best method for deflating
construction” (ibid. p. 260). But much can be
learned here about constructing deflation. The
view that science gets at the true nature of things
is quaintly called “realism” by philosophers. A
rather scholastic argument pitting “realists”
against sociological “constructivists” and other
opponents has been central to much recent
philosophy of science. Even the sociologists
have supposed that scientists will be realists
about the things they study. Realism, they argue,
is a myth that science can scarcely give up.
The researchers represented in this volume,
though, are able somehow to do without
realism. Or rather, their realism pertains
more to a political and bureaucratic than
to a philosophical idiom. While neoclassical
theorists quibble about what concepts can be
made mathematically unambiguous, quantiiiers
respond to the administrative demand for tools
to describe and manage a complex economy.
The skeptical comments quoted above are
concessions to theoretical objections. But the
volume is full of doubts about the value of
mainstream academic economics. Some com-
ments by Martin Feldstein, president and chief
executive officer of the NBER, are typical. He
laments that the economic profession “has
strayed from measurement and from serious
empirical work to a greater emphasis on theory
and to empirical studies that are often
much more interesting methodologically than
substantively” (p. 16).
The split between economic measurement
and deductive theory was already evident in the
nineteenth century. Many of our most important
quantitative tools of economic management
were developed and practised less by econo-
mists than by engineers, bankers, and actuaries
(Etner, 1987; Alborn, 1991; Porter, 1994; Klein,
forthcoming). The National Bureau of Economic
Research was conceived in this spirit. Wesley
Mitchell, who with Edwin Gay was largely
responsible for its founding in 1920, argued
vigorously against the increasing emphasis
in economics on marginal utility theory.
His reasoning drew from interpenetrating
commitments to behaviorism and to universal
measurement:
Economists who practice quantitative analysis are likely
to be chary of deserting the tirmground of measurable
phenomena for excursions into the subjective. If
my forecast is valid, our whole apparatus of reasoning
on the basis of utilities and disutilities, or motives, or
choices, in the i ndi vi dual economy, will drop out of sight
in the work of quantitative analysts, going the way of
the static state (Mitchell, 1925, p. 4).
Perhaps his forecasting was not quite perfect.
But the split he identified has if anything grown
more pronounced. The Econometric Society,
founded in 1931, aimed to overcome the divide
between the empirical and theoretical, between
measurement and mathematics. But the alliance
proved diflicult to maintain. Fifteen years later,
Tjalling Koopmans opened a bitter debate over
“measurement without theory” and “theory
without measurement” when he charged that
Arthur Burns and Wesley Mitchell lacked a
genuine economic theory of the business cycles
they purported to measure (Morgan, 1990).
The NBER was created above all to study
business cycles. By design it was aIRhated
with leading universities, academic societies,
business organizations, and government bureaux.
It was strongly supported by secretary of
commerce, later president, Herbert Hoover. It
thus had a quasi-official status, which to a degree
it maintains (Alchon, 1985; Hawley, 1990). As
such, it has had to respond to a variety of
pressures, and to satisfy a much larger audience
INFORMATION CULTURES 85
than that of economists. Fifty Years ofEconomic
Measu remfmt, contrary to reasonable expecta-
tions raised by the title, does not even attempt
a historical account of the measurement of
income and wealth, nor of the organization it
celebrates. It comes closest to being historical
in the luncheon speeches, in which the grand
old men of the Conference on Research in
Income and Wealth reminisce about their days
with Simon Kuznets.
Most of the papers offer review essays
to support the mixture of prophecy and
proselytism with which they address the future
of economic measurement. But it is not really
possible to use these essays as a basis for
assessing the history of the measurement of
capital and income. That story would be a
fascinating one, and might be as revealing
about the history of economic policy and
governmental power in the United States
as is the rather different history of national
accounting in France (Fourquet, 1980; Miller,
1986). This book is more an effort to survey the
state of the art. To the disciplinary outsider, it
is perhaps most interesting for what it reveals
about economic quantification as a social
achievement, and about how its practitioners
have adapted, and contributed, to their political
culture.
MAKING THINGS WITH NUMBERS
The lirst point, one that much exercises the
economists and statisticians, is that they are
very largely at the mercy of a data-gathering
apparatus that was not designed with their needs
in mind. As a “policy users’ panel” reports with
regret at the end of the volume, the United
States has a particularly decentralized statistical
apparatus. Most of the numbers that eventually
become available for economists to digest are
gathered by diverse agencies for their own
purposes and on their own schedules. Rudolph
Penner remarks that U.S. trade data are collected
by the customs service only as a byproduct of
the administration of trade law. Wall Street has
of late become obsessed with the monthly ups
and downs of the balance of trade, but these
numbers are “of terribly low quality.” Tax data
are provided by the Internal Revenue Service.
It releases statistics of income 2 years after the
fact. Until then, economists cannot separate
payroll taxes from income taxes (Bemdt &
Triplet& 1990, p. 430).
Some quantities assigned great importance by
the economists are not reported directly, but
rather must be inferred somehow from related
data. Low savings rates in the United States have
occasioned much hand-wringing, and are often
blamed for an unacceptably poor American
economic performance in recent decades.
Michael Boskin points out that household saving
in American national accounts is measured
indirectly by subtracting consumer expendi-
tures, taxes, and interest payments from
personal income. Measurement by residuals is
dazzlingly inexact. If unreported income is only
4% of total income, and if it is primarily saved,
this would raise the net private savings rate from
5% to 9% (ibid., pp. 167-168).
The construction of social quantities is in
every case a messy, business. It might well be
said of economic measures, as it has been said
of legislation and sausages, that those who love
them should avoid observing how they are
made. To use them expertly, on the other
hand, is scarcely possible without an intimate
knowledge of the data-gathering process. The
creation and use of economic numbers make up
a fascinating episode in the cultural history
of knowledge. Alain Desrosieres (1993) has
recently published a general history of statistics
with the problem of constructing quantitative
entities at its center. The history of economic
theory has only a bit part in his narrative.
Mathematical statistics and probability theory
are more important for him. But the really
outstanding feature of his book is its attention
to the twin problems of classification and
coding. These activities shed light on economies
and societies, but they also shape them. They
should be understood, he proposes, not
simply in terms of their objectivity, but also
of objectification - the collaboration of
statisticians in the construction of new things.
86 T. M. PORTER
Desrositres’ premier example is the subtly
different categorization in English, French and
German of “professionals”, cadres, and
Angestellte.
“National income” and “assets” or “capital”
are also objects that had to be redefined in order
to be measured, as the authors in Fifty Years
recognize. In other domains, quantifiers have
helped to make up “Hispanics”, “hypertension”,
and “battered child syndrome” (see Petersen,
1987; Hacking, 1990, 1991). It has been
suggested that the low Japanese death rate
from heart attacks, commonly attributed to a
“healthful diet”, may tell more about the cultural
acceptability of various ways of dying than about
public health (Bowker & Star, 1994). Although
doubts are sometimes heard, for most purposes
these constructed objects are treated as
natural categories, and used to explain other
phenomena. Quantifying, accordingly, should
not be understood merely as a descriptive
activity, but, like science itself, as a form of
intervention (Hacking, 1983; Latour, 1987;
Porter, 1992b, 1995).
THE MARKET FOR OBJECTMTY
Statisticians and economists may make their
own categories of analysis, but they do not make
them however they choose. Although it is hard
to say much about nature or the economy
without using the categories created by
researchers, the world does not passively accept
whatever forms of description and explanation
may strike the fancy of some social or natural
scientist. Social and political structures, too,
help to define what can count as acceptable
knowledge. We generally suppose that the
content of science is mainly determined by
disciplinary specialists, and in a sense this is
true. At least physicists and biologists do not
ordinarily have to submit their technical
vocabulary for political approval. Economic
statisticians, since their work is often absorbed
directly into the machinery of policy, operate
under more severe constraints. When, for
example, the Bureau of Labor Statistics wanted
to shift from using house prices to rental values
in determining the rate of inllation, they were
strongly and effectively opposed by the labor
unions (Charles L. Schultze in Berndt & Triple&
1990, p. 426). More generally, the power of
experts is sharply constrained, especially in
the United States. Validity is identified with
impersonal routines, not with the seasoned
judgment of wise elites. Economic quantities are
presumed to be largely independent of human
subjectivity, and this is crucial to their
credibility. Subjective adjustments, however
plausible, are not politically acceptable, even if,
as Griliches grumbles, the official numbers differ
mainly in that the judgments that go into them
“are usually well hidden behind the official
facade of the statistical establishment” (ibid.,
pp.190-191).
American politics nurtures a vigorous strain
of populism, and hence a pervasive suspicion of
experts, sometimes even of scientists. At the
same time, Americans seek whenever possible
to use scientific knowledge as a check on
administrative discretion, and in this respect
they seem often to be uncritically reverential
toward science. To make sense of this seeming
contradiction, it is sometimes argued that
Americans oscillate “between deference and
skepticism toward experts” (Jasanoff, 1990, p.
9; also Balogh, 1991, p. 34). Doubtless there is
some truth to this. Also, reverence and suspicion
are concentrated in different social groups,
and directed to different authorities. Still, the
characteristic American sense of expertise
permits a union of love and hate. For public
purposes, especially, experts are not esteemed
for a nuanced sensibility, but for their command
of a rigorous body of scientific principles and
rules. They are expected to follow them
slavishly. When they can be shown to have relied
instead on their own judgment, or subjectivity,
this is regarded as scandalous. The extraordinary
reliance on quantification in American admini-
strative culture is the purest expression of these
values. Of course it is never possible to eliminate
all subjective discretion, but it must be buried
in the details, not exercised openly by
prominent researchers or agency officials.
INFORMATION CIJ LTURES 87
The firm commitment of economists working
in the public domain to the quantification of
everything possible can be understood partly as
a response to this climate of political suspicion.
Griliches himself insists on the need for “a
decent price index of ‘health’ ” (in Berndt 81
Triplett, 1990, p. 191) using scare quotes to
mock those who would insist that it is above
price, but perhaps also to deny that it can
properly exist before it is measured. Paul Pieper
argues that construction cost indexes must
be adjusted to take account of improved
productivity, but denigrates all previous at-
tempts by calling them “arbitrary or subjective”
(p. 240). Would he deny that a large element
of discretion is required to make any particular
measure of construction productivity officially
valid? The authors in this NBER volume are well
aware that much of what they undertake to
measure does not “exist” unambiguously. The
point of a ftrm rule is not necessarily to make
the measures more accurate, but to make them
more constraining, so that they do not depend
so much on the particular individuals who do
the work, and in particular are less vulnerable
to bias or self-interested manipulation.
Following explicit rules is not the only
criterion of a rhetorically effective measure.
Official sanction by a public agency is generally
even more decisive, though it must be added
that the agencies have often earned that
credibility by putting mechanisms in place that
produce the numbers almost automatically.
Such numbers may be culturally valid, even if
their accuracy is questionable. The availability
of sanctioned economic numbers can have a
huge effect on political decision-making. Ian
Stewart, formerly deputy minister of finance in
Canada, remarks:
I may be wrong, but it struck me as an observer from
the north that the U.S. tax reform process was
enormously fertilized and facilitated by the appearance
of legislators, executives, and policy analysts carrying
the same sheaf of data printouts. Thus, data at least
offered a first-order estimate of the consequences of
changing parameters of the U.S. tax system and at least
permitted the beginnings of a discussion. For example,
lf municipal bonds are to be tax free then the costs are
x billion dollars, and lf that x bllons are to be put back
into the system, then compensating changes must
be made somewhere else. [Ijt permits public
policymakers to begin to argue their dlfTerences, and the
dllerences in their behavioral assumptions, on a broad
data base that starts the discussion some miles past the
starting point (ibid., p. 435).
This reliance on a single sheaf of data
printouts, it should be said, is not the only
possible way to run a nation’s finances. The
British treasury, at least until very recently,
depended less on impersonal information than
on an intricate network of personal knowledge
and trust at the top levels of government.
Negotiations were less open; a broader domain
of shared values and expectations among
the participants could be assumed. Heclo &
Wildavsky (1974) call this “the private govern-
ment of public money”, and it is in many ways
less democratic than the more data-driven style
adopted by the Americans. The contrast, in any
case, is revealing of the form of culture that
depends most on impersonal information. As is
shown by the other NBER conference volume
under review here, Peter Temin’s I nside the
Business Enterprise: Historical Perspectives on
the Use of I nformation, that culture embraces
much of the business world as well as
government.
TRUST-BUSTING
To economists, the large corporation appears
increasingly as the solution to a problem of
information. Daniel Ratf and Peter Temin (in
Temin, 199 1) offer an account of the business
organization in terms of a version of agency
theory. Not surprisingly, this move evokes some
sharp comments in the discussion, but the
paper does at least take organizational cul-
ture seriously. The problem, as they pose it,
is that a system of incentives and surveil-
lance is required to get people to act in the
interest of the larger organization. Within an
organization, wages and salaries should be
structured to encourage employees to work
effectively without creating a huge burden of
88 T. M. PORTER
administration and oversight. Information prob-
lems may also provide incentives to expand the
organization and bring in outsiders, Vertical
integration, for example, can be thought of as a
strategy to minimize transaction costs. It is
important to add, though, that the limited supply
or high cost of information is not alone
responsible for the advantages of organization
over individually negotiated contracts. There is
also a scarcity of information about information,
a ubiquitous problem of deciding whom to trust
(see Casson, 1994).
Indeed, it is evident from I nside the Business
Enterprise that the history of information is also
a history of trust and distrust. Or, more helpfully,
forms of information must be understood in
terms of varying means for establishing trust. All
of the papers in this volume address a transition
from the relatively local, personal, and some-
times secretive to the cosmopolitan, stand-
ardized, and public. This has entailed a shift in
the locus of trust away from individuals,
presumed to be bound by customs and moral
obligations, in favor of impersonal institutions
and vigilant oversight. In some ways, the
transition seems to follow almost inevitably
from economic expansion, but there are clear
cultural differences in its extent. It could be, as
two of the papers imply, that the relative
persistence of informality and personal trust in
Japan relative to the United States is partly
responsible for its superior postwar economic
performance.
In the late nineteenth century, the Scovill
Manufacturing Company deliberately avoided
the centralization of power that would be implied
by a vast net of information. Responsibility was
better left near the scene of action.
We have never had any shop rules printed. There is a
general understanding that ten hours constitute a day’s
work and that the hands are expected to do a day’s work
if they get a day’s pay. Each department is under the
direction of a foreman, in whom we trust and who sees
that the hands are industrious and attend to their
business. If they do not do it, he sends them off and gets
others.
This observation by M. L. Sperry, who later
became company president, is quoted in a fine
book by JoAnne Yates ( 1989, p. 168) and also
in two more recent papers for collections on
the history of information, including the
Temin volume under review. She never treats
this outlook, though, as a viable scheme of
management. While conceding that “new
methods for internal coordination, methods less
dependent on vertical paper-based information
flows, might have been developed’ (in Temin,
1991, p. 132), Yates generally discusses the
centralization of information and power as if it
were inevitable. She writes in I nfomration
Acumen, edited by Lisa Bud-Frierman: “As
managers began to systematize procedures, they
needed to communicate them downward
in written form both to specify them as
unambiguously as possible and to provide a
source of reference on the new procedures”
(Yates, 1994, p. 31).
Yates explicitly rejects the view that the
expansion of information was simply a rational
response to objective conditions. Information
technologies, she points out, had also a symbolic
dimension, and were valued by managers as
emblems of modernity. As other scholars have
observed, their appeal was felt much more
powerfully in the United States than in Europe.
British offices like the Railway Clearing House
processed huge quantities of information, yet
were strikingly reluctant to invest even in
typewriters or adding machines (Campbell-
Kelly, 1994). The differences go beyond
symbolism and trendiness, pointing to deep
differences in business culture and styles
of management. The push by professional
managers, located in offices, to extend their
reach to operations in distant offices or on the
shop floor is part of the story, but this is no
simple history of a usurpation of power by
central management. The culture of printed
information was in many ways more open than
one based on oral communication, and indeed
companies sometimes resisted it in order to
preserve secrecy (Margaret Levinstein in Temin,
199 1, pp. 83,102). Reliance on highly processed
forms of information such as return on
investment (ROI) was conceived as a strategy
to permit the central office to evaluate the
INFORMATION CULTURES
89
success of subordinate units without inter-
vening too forcefully in day-to-day operations
(Miller & O’Leary, 1987). The issues raised
by information culture are very general. As
Desrosieres ( 1994) reminds us, we need to
think about what kinds of business activity are
compatible with a preoccupation with statistics
and other relatively objective forms of infor-
mation. He also urges that the history of
information should not be confined by the
boundaries of the business organization, but
should look to forces outside, and in particular
to the state.
Happily, several of the authors of Instie the
Busi ness Enterpri se are attentive also to its
outside. Among the outsiders who acted most
effectively on this move toward an information
culture were investors, banks, business edu-
cators, and regulatory agencies. Precisely
because they were located outside the boun-
daries of the organization, they generally were
not well placed to take advantage of informal
communications. The culture of paper and
objectivity, it would seem, provided their only
real access to what a company was doing, and
hence their best opportunity to protect their
own interests.
In fact the external affairs of firms were also
handled mostly through informal means until
about the beginning of this century. Banks, far
from having to content themselves with Iinancial
records to decide ifprospective borrowers were
creditworthy, loaned mainly on the basis of
personal and informal knowledge. At least
this was true for nineteenth-century American
banks, as Naomi Lamoreaux shows (Temin,
1991). They made money available almost
exclusively to friends and family of the bank
directors. Most banks loaned most of their funds
to the directors themselves. Access to capital
was the main reason to take the trouble to serve
as a bank director at all. State regulatory
authorities, the timeless enemies of informal
trust, greatly disapproved of all this insider
lending. In Massachusetts they required a vote
of approval by stockholders before directors
could be loaned more than 30% of the banks
capital stock. Stockholders routinely approved
such proposals.
Writes Iamoreaux: “The indulgence with
which stockholders treated insider lending is
easier to understand once one appreciates the
importance of reputation in this information-
scarce economy” (p. 172). The point is an
important one. It needs to be qualified: while
there was little information of the sort that
we Iind now in reports to shareholders and
regulators, the people in the community were
often very well informed about what mattered
most. They were interested above all in the
character of the borrowers. They wanted their
money to be loaned to honorable, trustworthy
men. Doubtless some were fooled by scoundrels,
but the evidence from New England suggests
that the honor of substantial men provided good
assurance of the safety of capital. Lamoreaux
offers the case of one John James Dixwell, who
as president of the Massachusetts Bank arranged
a considerable loan to the Boston Brick
Manufacturing Company. As president of the
latter he felt some obligations even though he
had no personal liability for the loan. When the
company failed, ‘he made the loan good out of
his own resources. The reputations of the
directors were jeopardized whenever one of
their fellows defaulted. On this account, they
watched each other very carefully. “Given the
poor quality of information in the period’,
argues Iamoreaux, “such monitoring of insiders
by insiders may actually have been less risky
than extending credit to perfect strangers”
(p. 173).
This is greatly understated. Whatever the
availability of formalized information such as
accounts, insider lending was potentially safer
than extending credit to strangers. But the
system could not work without the benefit of
frequent face-to-face interactions. This defined
the bounds of effective personal knowledge, and
the domain within which reputation really
mattered. In consequence, the emergence of a
national lending market led to the breakdown
of this informal system of mutual surveillance
and self-regulation. Towards the end of the
century, Lamoreaux points out, a national
market developed for commercial paper. Now,
inside knowledge became inadequate. Many
banks suBered heavy losses. They began to
90 T. M. PORTER
rely on impersonal information about their
borrowers. They subscribed to the reports of
credit agencies like R. G. Dunn and Company.
They tried to work out “agreed-upon-standards
-particularly a set of objective criteria for loans
-that directors could use to monitor managers’
performance” (p. 188). In short, they left the
world of the face to face and entered that of
impersonal measurement.
Lamoreaux tells this story in terms of the
growth of information, even of economic and
institutional progress, but it might well be told
instead in the classic New England tradition as
one of moral declension. To be sure, concern
for reputation did not disappear in the financial
markets. Bradford De Long suggests in another
paper in the Temin volume that the investment
house of J. P. Morgan thrived mainly on account
of its reputation for integrity and competence.
Morgan’s operation was so huge that his
reputation could extend nationally and even
internationally. Significantly, it did not work by
assembling mountains of paper in a New York
office. Morgan participated actively in running
the companies to which he loaned, inspecting
them closely from his place on their boards of
directors. That personal knowledge and control
made it possible for him to commit for the long
term. The great symbol of financial greed seems
now to have perpetuated a kind of moral
economy into the era of big capitalism. It is a
nice irony, De Long concludes, that in our own
day “many of the intellectual children and
grandchildren of the Progressives appear to call
for a return to ‘financial capitalism”’ (p. 233).
This is not exactly what Thomas Johnson
argues in his paper on “Managing by Remote
Control”, but certainly there is a family
resemblance to both the papers on banking. The
theme of decline and moral failure is doubly
emphasized by the grandiloquent title and
subtitle of Johnson and Kaplan’s book (Johnson
& Kaplan, 1987) Relevance Lost: The Rise and
Fall of Management Accounting, thorn which
his paper draws extensively. There was once a
time, the argument begins, when management
accounting was used sensibly for limited
purposes by people who relied extensively on
other kinds of knowledge as well. So long
as ownership and management of American
companies were closely intertwined, there was
no need to back up business decisions with
reams of information. Owners were interested
in financial data, but they kept it mostly to
themselves, and they managed their firms less
on the basis of accounting information than
of knowledge of products, operations, and
customers (Temin, 199 1, p. 43). Before about
1920, “top managers were intimately familiar
with their companies’ customers and techno-
logies. They did not have to hide behind a facade
of accounting information to converse with
subordinates” (p. 49).
The serpent in this retelling of Milton appeared
to businessmen in the form of investors and
banks, offering money. All they asked in return
(to shift to Faust) was the company’s soul. They
wanted objective financial information, which
should be made available to outsiders in order
to increase public confidence and thereby
separate investors from their money. This was
not yet intolerable, so long as the accounts were
kept in their place. And for a time, Johnson tells
us, they were. Financial numbers were used to
promote the confidence of investors, and not to
run the company. Since World War II,
this separation has mostly broken down.
Government regulators, such as the Securities
and Exchange Commission, demanded huge
quantities of information, and prescribed its
form. Business schools succumbed to the dogma
of economics that the firm was a site of
quantitative rationality. They idealized the
knowledge of outsiders, the kind of knowledge
that the professors themselves possessed in
the greatest abundance. Within companies,
management was separated from the scene of
action, confining itselfmainly to a world ofpaper
and numbers. The financial side of companies
gained control over the real business of produc-
tion and sales. In short, the all-important
insider knowledge assumed second place to
outsider knowledge, the formalized, objectified
information of graphs and numbers.
Johnson writes not in despair, but with the
hope that this lamentable postwar lapse can
INFORMATION CIJLTURFS
91
be remedied, perhaps reversed. I would not public purposes, it is often necessary to measure
venture to predict whether this is possible. But according to relatively unambiguous rules
it is worth observing that this triumph of rather than to apply judgment where judgment
financial measurement at the level of corpora- is called for. The business corporation has, in
tion parallels and in some way reflects the many ways, been brought into the public
spectacular growth of quantification in the domain. In America, especially, it is a domain
domain of public policy, as evidenced in Fifty that idealizes objectivity. Even ifbusiness people
Years of Economic Measurement. The authors become convinced, as Johnson urges, that
in that volume, it may be recalled, stress the sensible management can never be reduced to
public need for measurement even of entities the manipulation of financial numbers, it may
whose “existence” was open to doubt. For not be easy to escape their tyranny.
BIBLIOGRAPHY
Alborn, T., The Other Economists: Science and Commercial Culture in Nineteenth-century Economics,
Ph.D. dissertation, Harvard University ( 199 1).
AIchon, G., The I nvfsfbl e Hand of Pl anni ng (Princeton: Princeton University Press, 1985).
Balogh, B., Cbafn Reacti on Expert Debate and Publ i c Partfcfpatfon i n Ameri can Commerci al Nucl ear
Power, 39451975 (New York: Cambridge University Press, 1991).
Barber, W., From New Era to New Deal (Cambridge: Cambridge University Press, 1985).
Berndt, E. R. SK Triplett, J. E. (eds), Fi fty Years of Economi c Measurement: Tbe J ubi l ee of tbe Conference
on Research i n I ncome and Weal th (Chicago: University of Chicago Press, 1990).
Bowker, G. & Star, S. L., Knowledge and infrastructure in international information management, in Bud-
Frierman, L. (ed.), I nformati on Acumen: Tbe Understandi ng and Use of Knowl edge i n Modern
Busi ness, pp. 187-213 (London: Routledge, 1994).
CampbeII-Kelly, M., The Railway Clearing House and Victorian Data Processing in Bud-Frierman, L. (ed.),
I nformati on Acumen: Tbe Understandi ng and Use of Knowl edge i n Modern Busi ness, pp. 51-74
(London: Routledge, 1994).
Casson, J., Economic Perspectives on Business Information, in Bud-Frierman, L. (ed.), I nformatfon
Acumen: Tbe Understandi ng and Use of Knowl edge i n Modern Busi ness, pp. 136-167 (London:
RoutIedge, 1994).
Desrosieres, A., Lapol fti que desgrands nombnx: bfstoi re de l a rafson statfsti que (Paris: Ia Decouverte,
1993).
Desrosieres, A., Official statistics and business: history, classifications, uses, in Bud-Frierman, L. (ed.),
I nformati on Acumen: Tbe Undemtandtng and Use of Knowl edge i n Modern Busi ness, pp. 168-186
(London: Routledge, 1994).
Etner, F., Hfstofre de cal cul kconomfque en France (Paris: Economica, 1987).
Fourquet, F., Les comptes de l a pui ssance (Paris: Encres, 1980).
Hacking, I., Representi ng and I nterveni ng (Cambridge: Cambridge University Press, 1983).
Hacking, I., Tbe Tami ng of Chance (Cambridge: Cambridge University Press, 1990).
Hacking, I., The Making and Molding of Child Abuse, Crfti cal I nqui ry (1991) pp. 253-288.
Hawley, E. W., Economic Inquiry and the State in New Era America, in Furrier, M. 0. & Supple, B. (eds),
Tbe State and Economi c Knowl edge, pp. 287-324 (Cambridge: Cambridge University Press, 1990).
Heclo, H. & Wildavsky, A., Tbe Prfvate Government of Publ i c Money: Communi ty and Pol i cy I nsfde
Bri ti sh Pol i ti cs (Berkeley: University of California Press, 1974).
Jasanoff, S., Tbe Fi ftb Branch: Sci ence Advi sers as Pol i cytnaJ zers (Cambridge: Harvard University Press.
1990).
Johnson, H. T. & Kaplan, R. S., Rel evance Lost: Tbe Ri se and Fal l ofManagementAccounti ng (Cambridge:
Harvard Business School Press, 1987).
Klein, J,, Ti me and tbe Sci ence of Means: The Stati sti cal Anal ysi s of Changi ng Phenomena, 1830-1940
(Cambridge: Cambridge University Press, forthcoming).
Latour, B., Sci ence i n Acti on (Cambridge: Harvard University Press, 1987).
92 T. M. PORTER
Miller, P., Accounting br Progress - National Accounting and Planning in France: a Review Essay,
Account@ Organfza~fons, and Society (1986) pp. 83-104.
Miller, P. & O’Leary, T., Accounting and the Construction of the Governable Person, Accounting,
OrganrZatfons, and Society (1987) pp. 235-265.
Mitchell, W. C., Quantitative Analysis in Economic Theory, American Economfc Review (March 1925)
pp. 1-12.
Morgan, M., A H&x-y of Econometrtc I deas (Cambridge: Cambridge University Press, 1990).
Petersen, W., Politics and the Measurement of Ethnic&y, in Alonso, W. & Starr, P. (eds), The Polftics of
NumBeqpp. 187-233 (New York: Russell Sage Foundation, 1987).
Porter, T. M., Quantification and the Accounting Ideal in Science, So&al Studies of Science (1992a)
pp. 63ti51.
Porter, T. M., Objectivity as Standardiiation: the Rhetoric of Impersonality in Measurement, Statistics and
Cost-Benefit Analysis, Annals of S&o&ship (1992b) pp. 19-59.
Porter, T. M., Rigor and Practicality: Rival Ideals of Quantification in Nineteenth-century Economics, in
Mirowski, P. (ed.), Natural Images in Economic Tbougbt, pp. 128-170 (New York: Cambridge
University Press, 1994).
Porter, T. M., Trust in Numbers The Pursuit of Objectivity in Science and Public Life (Princeton:
Princeton University Press, 1995).
Temin, P. (ed.), I nside ibe Business Enterprise: Historical Perspectives on the Use of I nformation
(Chicago: University of Chicago Press, 1991).
Yates, J., Control Tbrougb Communfcatfon: The Rise of System in American Management (Baltimore:
Johns Hopkins University Press, 1989).
Yates, J., Evolving Information Use in Firms, 1850-1920, in Bud-Frierman, L. (ed.) I nfwmatfon Acu~
The LJ ?kierstunding and Use of Knowledge in Modern Business, pp. 26-50 (London: Routledge, 1994).
doc_552848025.pdf
In American politics, at least, it is welI known
that government bureaucrats and middle managers
are paid high salaries to shuffle papers.
But we learn from the business pages that public
administrators and business executives process
information, the life blood of the postlndustrial
economy. Readers of academic journals,
especially in accounting, can easily
Pergamon Accounting Organfzatfotas and Sockty, Vol. 20, No. 1, pp. M-92,1995
Copyright 0 1994 Elsevia Science Ltd
Printed in Great Britain. All rights reserved
0361-3682T95 19.50+0.00
INFORMATION CULTURES: A REVIEW ESSAY
THEODORE M. PORTER
Department of Histoty, University of Califomiq Los Angeles
Bemdt, E. R & Iiiplett, J. E. (eds), Ff@y Yems of Economic Measummm k TheJ ubilee of t%e Confelmce
onR-fnfname and wt%mJ (Chicago: university of caicago press, 1990).
Temin, P. (ed. ), Instak the Business Entmprk His-1 Perspectim on the Use of I njxmation (Chicago:
University of Chicago Press, 199 1).
In American politics, at least, it is welI known
that government bureaucrats and middle man-
agers are paid high salaries to shuffle papers.
But we learn from the business pages that public
administrators and business executives process
information, the life blood of the postlndustrial
economy. Readers of academic journals,
especially in accounting, can easily recognize
the element of populist demagogy in the iirst
description. ShuIIling paper is processing
information.
We also tend to believe the converse, that
information flow means precisely the movement
of papers and electronic signals. Knowledge
and information are identified with a highly
formalized, often quantitative, style of expres-
sion. The “information society” is a culture of
logorrhea and numerrhea. This is probably right,
but in a way it is misleading. Early modem
artisans, miners and farmers depended no less
on the transmission of information than do
modem executives and researchers, but they
did not broadcast it indiscriminately. Theirs was
a culture of the face to face. The shift to one of
printed reports and tables, which is and wiIl
never be complete, cannot be understood
simply in terms of the growth of knowledge,
communications technologies, and computing
power. It involved also a change of values - a
move away Tom the personal, and a vast
expansion of the public domain.
THE POWER OF MEASUREMENT
Voltaire once observed that if Cod did not
exist, we would have to invent him. That was
the eighteenth century. In this one, many
entities that cannot be clearly deiined or
delimited must nonetheless be measured. This
may sound cynical, but it is more like boasting.
Fifty Years of Economic Measurement com-
memorates the founding of the Conference on
Research in Income and Wealth, under the
National Bureau of Economic Research (NBER).
Its efforts have helped to give fixity and
coherence not just to vague concepts, but to
vague objects. The volume celebrates these
heroic deeds, this triumph of quantitative clarity
over a fuzzy economy.
The contributors are relatively worldly
members of a decidely unworldly profession,
economics. Of course they do not call for or
applaud the imposition of economic measures
through naked force or scholarly scheming.
Their efforts presuppose that it is possible
to discuss intelligently the advantages and
limitations of various measures, and in this way
to improve them. They are, however, under no
illusion that their scientific difficulties can be
definitively solved through discovery of the true
nature of things. The measurement of inflation
in the currency, for example, has been
vigorously debated by economic statisticians,
83
84 T. M. PORTER
and often scorned by theorists. Still, even well-
founded doubts are not decisive. Questions
about the reality of the things being measured
can be fended off with inverted commas. “How
does one measure something that may not
‘exist’?” asks Charles R. Hulten. Just do it. “Come
hell or high water, many of us will be using and/
or producing measures of aggregate capital
input, even if it does not ‘exist”’ (Berndt &
Triplett, 1990, p. 153). Notwithstanding the
contradictions implicit in any measure of the
value of the currency, write Makoto Ohta and
Zvi Griliches, “the Consumer Price Index (CPI)
‘exists’ and is even of some use” (quoted by
Griliches, ibid., p. 189). There may, as Paul
Pieper argues, be “no best method for deflating
construction” (ibid. p. 260). But much can be
learned here about constructing deflation. The
view that science gets at the true nature of things
is quaintly called “realism” by philosophers. A
rather scholastic argument pitting “realists”
against sociological “constructivists” and other
opponents has been central to much recent
philosophy of science. Even the sociologists
have supposed that scientists will be realists
about the things they study. Realism, they argue,
is a myth that science can scarcely give up.
The researchers represented in this volume,
though, are able somehow to do without
realism. Or rather, their realism pertains
more to a political and bureaucratic than
to a philosophical idiom. While neoclassical
theorists quibble about what concepts can be
made mathematically unambiguous, quantiiiers
respond to the administrative demand for tools
to describe and manage a complex economy.
The skeptical comments quoted above are
concessions to theoretical objections. But the
volume is full of doubts about the value of
mainstream academic economics. Some com-
ments by Martin Feldstein, president and chief
executive officer of the NBER, are typical. He
laments that the economic profession “has
strayed from measurement and from serious
empirical work to a greater emphasis on theory
and to empirical studies that are often
much more interesting methodologically than
substantively” (p. 16).
The split between economic measurement
and deductive theory was already evident in the
nineteenth century. Many of our most important
quantitative tools of economic management
were developed and practised less by econo-
mists than by engineers, bankers, and actuaries
(Etner, 1987; Alborn, 1991; Porter, 1994; Klein,
forthcoming). The National Bureau of Economic
Research was conceived in this spirit. Wesley
Mitchell, who with Edwin Gay was largely
responsible for its founding in 1920, argued
vigorously against the increasing emphasis
in economics on marginal utility theory.
His reasoning drew from interpenetrating
commitments to behaviorism and to universal
measurement:
Economists who practice quantitative analysis are likely
to be chary of deserting the tirmground of measurable
phenomena for excursions into the subjective. If
my forecast is valid, our whole apparatus of reasoning
on the basis of utilities and disutilities, or motives, or
choices, in the i ndi vi dual economy, will drop out of sight
in the work of quantitative analysts, going the way of
the static state (Mitchell, 1925, p. 4).
Perhaps his forecasting was not quite perfect.
But the split he identified has if anything grown
more pronounced. The Econometric Society,
founded in 1931, aimed to overcome the divide
between the empirical and theoretical, between
measurement and mathematics. But the alliance
proved diflicult to maintain. Fifteen years later,
Tjalling Koopmans opened a bitter debate over
“measurement without theory” and “theory
without measurement” when he charged that
Arthur Burns and Wesley Mitchell lacked a
genuine economic theory of the business cycles
they purported to measure (Morgan, 1990).
The NBER was created above all to study
business cycles. By design it was aIRhated
with leading universities, academic societies,
business organizations, and government bureaux.
It was strongly supported by secretary of
commerce, later president, Herbert Hoover. It
thus had a quasi-official status, which to a degree
it maintains (Alchon, 1985; Hawley, 1990). As
such, it has had to respond to a variety of
pressures, and to satisfy a much larger audience
INFORMATION CULTURES 85
than that of economists. Fifty Years ofEconomic
Measu remfmt, contrary to reasonable expecta-
tions raised by the title, does not even attempt
a historical account of the measurement of
income and wealth, nor of the organization it
celebrates. It comes closest to being historical
in the luncheon speeches, in which the grand
old men of the Conference on Research in
Income and Wealth reminisce about their days
with Simon Kuznets.
Most of the papers offer review essays
to support the mixture of prophecy and
proselytism with which they address the future
of economic measurement. But it is not really
possible to use these essays as a basis for
assessing the history of the measurement of
capital and income. That story would be a
fascinating one, and might be as revealing
about the history of economic policy and
governmental power in the United States
as is the rather different history of national
accounting in France (Fourquet, 1980; Miller,
1986). This book is more an effort to survey the
state of the art. To the disciplinary outsider, it
is perhaps most interesting for what it reveals
about economic quantification as a social
achievement, and about how its practitioners
have adapted, and contributed, to their political
culture.
MAKING THINGS WITH NUMBERS
The lirst point, one that much exercises the
economists and statisticians, is that they are
very largely at the mercy of a data-gathering
apparatus that was not designed with their needs
in mind. As a “policy users’ panel” reports with
regret at the end of the volume, the United
States has a particularly decentralized statistical
apparatus. Most of the numbers that eventually
become available for economists to digest are
gathered by diverse agencies for their own
purposes and on their own schedules. Rudolph
Penner remarks that U.S. trade data are collected
by the customs service only as a byproduct of
the administration of trade law. Wall Street has
of late become obsessed with the monthly ups
and downs of the balance of trade, but these
numbers are “of terribly low quality.” Tax data
are provided by the Internal Revenue Service.
It releases statistics of income 2 years after the
fact. Until then, economists cannot separate
payroll taxes from income taxes (Bemdt &
Triplet& 1990, p. 430).
Some quantities assigned great importance by
the economists are not reported directly, but
rather must be inferred somehow from related
data. Low savings rates in the United States have
occasioned much hand-wringing, and are often
blamed for an unacceptably poor American
economic performance in recent decades.
Michael Boskin points out that household saving
in American national accounts is measured
indirectly by subtracting consumer expendi-
tures, taxes, and interest payments from
personal income. Measurement by residuals is
dazzlingly inexact. If unreported income is only
4% of total income, and if it is primarily saved,
this would raise the net private savings rate from
5% to 9% (ibid., pp. 167-168).
The construction of social quantities is in
every case a messy, business. It might well be
said of economic measures, as it has been said
of legislation and sausages, that those who love
them should avoid observing how they are
made. To use them expertly, on the other
hand, is scarcely possible without an intimate
knowledge of the data-gathering process. The
creation and use of economic numbers make up
a fascinating episode in the cultural history
of knowledge. Alain Desrosieres (1993) has
recently published a general history of statistics
with the problem of constructing quantitative
entities at its center. The history of economic
theory has only a bit part in his narrative.
Mathematical statistics and probability theory
are more important for him. But the really
outstanding feature of his book is its attention
to the twin problems of classification and
coding. These activities shed light on economies
and societies, but they also shape them. They
should be understood, he proposes, not
simply in terms of their objectivity, but also
of objectification - the collaboration of
statisticians in the construction of new things.
86 T. M. PORTER
Desrositres’ premier example is the subtly
different categorization in English, French and
German of “professionals”, cadres, and
Angestellte.
“National income” and “assets” or “capital”
are also objects that had to be redefined in order
to be measured, as the authors in Fifty Years
recognize. In other domains, quantifiers have
helped to make up “Hispanics”, “hypertension”,
and “battered child syndrome” (see Petersen,
1987; Hacking, 1990, 1991). It has been
suggested that the low Japanese death rate
from heart attacks, commonly attributed to a
“healthful diet”, may tell more about the cultural
acceptability of various ways of dying than about
public health (Bowker & Star, 1994). Although
doubts are sometimes heard, for most purposes
these constructed objects are treated as
natural categories, and used to explain other
phenomena. Quantifying, accordingly, should
not be understood merely as a descriptive
activity, but, like science itself, as a form of
intervention (Hacking, 1983; Latour, 1987;
Porter, 1992b, 1995).
THE MARKET FOR OBJECTMTY
Statisticians and economists may make their
own categories of analysis, but they do not make
them however they choose. Although it is hard
to say much about nature or the economy
without using the categories created by
researchers, the world does not passively accept
whatever forms of description and explanation
may strike the fancy of some social or natural
scientist. Social and political structures, too,
help to define what can count as acceptable
knowledge. We generally suppose that the
content of science is mainly determined by
disciplinary specialists, and in a sense this is
true. At least physicists and biologists do not
ordinarily have to submit their technical
vocabulary for political approval. Economic
statisticians, since their work is often absorbed
directly into the machinery of policy, operate
under more severe constraints. When, for
example, the Bureau of Labor Statistics wanted
to shift from using house prices to rental values
in determining the rate of inllation, they were
strongly and effectively opposed by the labor
unions (Charles L. Schultze in Berndt & Triple&
1990, p. 426). More generally, the power of
experts is sharply constrained, especially in
the United States. Validity is identified with
impersonal routines, not with the seasoned
judgment of wise elites. Economic quantities are
presumed to be largely independent of human
subjectivity, and this is crucial to their
credibility. Subjective adjustments, however
plausible, are not politically acceptable, even if,
as Griliches grumbles, the official numbers differ
mainly in that the judgments that go into them
“are usually well hidden behind the official
facade of the statistical establishment” (ibid.,
pp.190-191).
American politics nurtures a vigorous strain
of populism, and hence a pervasive suspicion of
experts, sometimes even of scientists. At the
same time, Americans seek whenever possible
to use scientific knowledge as a check on
administrative discretion, and in this respect
they seem often to be uncritically reverential
toward science. To make sense of this seeming
contradiction, it is sometimes argued that
Americans oscillate “between deference and
skepticism toward experts” (Jasanoff, 1990, p.
9; also Balogh, 1991, p. 34). Doubtless there is
some truth to this. Also, reverence and suspicion
are concentrated in different social groups,
and directed to different authorities. Still, the
characteristic American sense of expertise
permits a union of love and hate. For public
purposes, especially, experts are not esteemed
for a nuanced sensibility, but for their command
of a rigorous body of scientific principles and
rules. They are expected to follow them
slavishly. When they can be shown to have relied
instead on their own judgment, or subjectivity,
this is regarded as scandalous. The extraordinary
reliance on quantification in American admini-
strative culture is the purest expression of these
values. Of course it is never possible to eliminate
all subjective discretion, but it must be buried
in the details, not exercised openly by
prominent researchers or agency officials.
INFORMATION CIJ LTURES 87
The firm commitment of economists working
in the public domain to the quantification of
everything possible can be understood partly as
a response to this climate of political suspicion.
Griliches himself insists on the need for “a
decent price index of ‘health’ ” (in Berndt 81
Triplett, 1990, p. 191) using scare quotes to
mock those who would insist that it is above
price, but perhaps also to deny that it can
properly exist before it is measured. Paul Pieper
argues that construction cost indexes must
be adjusted to take account of improved
productivity, but denigrates all previous at-
tempts by calling them “arbitrary or subjective”
(p. 240). Would he deny that a large element
of discretion is required to make any particular
measure of construction productivity officially
valid? The authors in this NBER volume are well
aware that much of what they undertake to
measure does not “exist” unambiguously. The
point of a ftrm rule is not necessarily to make
the measures more accurate, but to make them
more constraining, so that they do not depend
so much on the particular individuals who do
the work, and in particular are less vulnerable
to bias or self-interested manipulation.
Following explicit rules is not the only
criterion of a rhetorically effective measure.
Official sanction by a public agency is generally
even more decisive, though it must be added
that the agencies have often earned that
credibility by putting mechanisms in place that
produce the numbers almost automatically.
Such numbers may be culturally valid, even if
their accuracy is questionable. The availability
of sanctioned economic numbers can have a
huge effect on political decision-making. Ian
Stewart, formerly deputy minister of finance in
Canada, remarks:
I may be wrong, but it struck me as an observer from
the north that the U.S. tax reform process was
enormously fertilized and facilitated by the appearance
of legislators, executives, and policy analysts carrying
the same sheaf of data printouts. Thus, data at least
offered a first-order estimate of the consequences of
changing parameters of the U.S. tax system and at least
permitted the beginnings of a discussion. For example,
lf municipal bonds are to be tax free then the costs are
x billion dollars, and lf that x bllons are to be put back
into the system, then compensating changes must
be made somewhere else. [Ijt permits public
policymakers to begin to argue their dlfTerences, and the
dllerences in their behavioral assumptions, on a broad
data base that starts the discussion some miles past the
starting point (ibid., p. 435).
This reliance on a single sheaf of data
printouts, it should be said, is not the only
possible way to run a nation’s finances. The
British treasury, at least until very recently,
depended less on impersonal information than
on an intricate network of personal knowledge
and trust at the top levels of government.
Negotiations were less open; a broader domain
of shared values and expectations among
the participants could be assumed. Heclo &
Wildavsky (1974) call this “the private govern-
ment of public money”, and it is in many ways
less democratic than the more data-driven style
adopted by the Americans. The contrast, in any
case, is revealing of the form of culture that
depends most on impersonal information. As is
shown by the other NBER conference volume
under review here, Peter Temin’s I nside the
Business Enterprise: Historical Perspectives on
the Use of I nformation, that culture embraces
much of the business world as well as
government.
TRUST-BUSTING
To economists, the large corporation appears
increasingly as the solution to a problem of
information. Daniel Ratf and Peter Temin (in
Temin, 199 1) offer an account of the business
organization in terms of a version of agency
theory. Not surprisingly, this move evokes some
sharp comments in the discussion, but the
paper does at least take organizational cul-
ture seriously. The problem, as they pose it,
is that a system of incentives and surveil-
lance is required to get people to act in the
interest of the larger organization. Within an
organization, wages and salaries should be
structured to encourage employees to work
effectively without creating a huge burden of
88 T. M. PORTER
administration and oversight. Information prob-
lems may also provide incentives to expand the
organization and bring in outsiders, Vertical
integration, for example, can be thought of as a
strategy to minimize transaction costs. It is
important to add, though, that the limited supply
or high cost of information is not alone
responsible for the advantages of organization
over individually negotiated contracts. There is
also a scarcity of information about information,
a ubiquitous problem of deciding whom to trust
(see Casson, 1994).
Indeed, it is evident from I nside the Business
Enterprise that the history of information is also
a history of trust and distrust. Or, more helpfully,
forms of information must be understood in
terms of varying means for establishing trust. All
of the papers in this volume address a transition
from the relatively local, personal, and some-
times secretive to the cosmopolitan, stand-
ardized, and public. This has entailed a shift in
the locus of trust away from individuals,
presumed to be bound by customs and moral
obligations, in favor of impersonal institutions
and vigilant oversight. In some ways, the
transition seems to follow almost inevitably
from economic expansion, but there are clear
cultural differences in its extent. It could be, as
two of the papers imply, that the relative
persistence of informality and personal trust in
Japan relative to the United States is partly
responsible for its superior postwar economic
performance.
In the late nineteenth century, the Scovill
Manufacturing Company deliberately avoided
the centralization of power that would be implied
by a vast net of information. Responsibility was
better left near the scene of action.
We have never had any shop rules printed. There is a
general understanding that ten hours constitute a day’s
work and that the hands are expected to do a day’s work
if they get a day’s pay. Each department is under the
direction of a foreman, in whom we trust and who sees
that the hands are industrious and attend to their
business. If they do not do it, he sends them off and gets
others.
This observation by M. L. Sperry, who later
became company president, is quoted in a fine
book by JoAnne Yates ( 1989, p. 168) and also
in two more recent papers for collections on
the history of information, including the
Temin volume under review. She never treats
this outlook, though, as a viable scheme of
management. While conceding that “new
methods for internal coordination, methods less
dependent on vertical paper-based information
flows, might have been developed’ (in Temin,
1991, p. 132), Yates generally discusses the
centralization of information and power as if it
were inevitable. She writes in I nfomration
Acumen, edited by Lisa Bud-Frierman: “As
managers began to systematize procedures, they
needed to communicate them downward
in written form both to specify them as
unambiguously as possible and to provide a
source of reference on the new procedures”
(Yates, 1994, p. 31).
Yates explicitly rejects the view that the
expansion of information was simply a rational
response to objective conditions. Information
technologies, she points out, had also a symbolic
dimension, and were valued by managers as
emblems of modernity. As other scholars have
observed, their appeal was felt much more
powerfully in the United States than in Europe.
British offices like the Railway Clearing House
processed huge quantities of information, yet
were strikingly reluctant to invest even in
typewriters or adding machines (Campbell-
Kelly, 1994). The differences go beyond
symbolism and trendiness, pointing to deep
differences in business culture and styles
of management. The push by professional
managers, located in offices, to extend their
reach to operations in distant offices or on the
shop floor is part of the story, but this is no
simple history of a usurpation of power by
central management. The culture of printed
information was in many ways more open than
one based on oral communication, and indeed
companies sometimes resisted it in order to
preserve secrecy (Margaret Levinstein in Temin,
199 1, pp. 83,102). Reliance on highly processed
forms of information such as return on
investment (ROI) was conceived as a strategy
to permit the central office to evaluate the
INFORMATION CULTURES
89
success of subordinate units without inter-
vening too forcefully in day-to-day operations
(Miller & O’Leary, 1987). The issues raised
by information culture are very general. As
Desrosieres ( 1994) reminds us, we need to
think about what kinds of business activity are
compatible with a preoccupation with statistics
and other relatively objective forms of infor-
mation. He also urges that the history of
information should not be confined by the
boundaries of the business organization, but
should look to forces outside, and in particular
to the state.
Happily, several of the authors of Instie the
Busi ness Enterpri se are attentive also to its
outside. Among the outsiders who acted most
effectively on this move toward an information
culture were investors, banks, business edu-
cators, and regulatory agencies. Precisely
because they were located outside the boun-
daries of the organization, they generally were
not well placed to take advantage of informal
communications. The culture of paper and
objectivity, it would seem, provided their only
real access to what a company was doing, and
hence their best opportunity to protect their
own interests.
In fact the external affairs of firms were also
handled mostly through informal means until
about the beginning of this century. Banks, far
from having to content themselves with Iinancial
records to decide ifprospective borrowers were
creditworthy, loaned mainly on the basis of
personal and informal knowledge. At least
this was true for nineteenth-century American
banks, as Naomi Lamoreaux shows (Temin,
1991). They made money available almost
exclusively to friends and family of the bank
directors. Most banks loaned most of their funds
to the directors themselves. Access to capital
was the main reason to take the trouble to serve
as a bank director at all. State regulatory
authorities, the timeless enemies of informal
trust, greatly disapproved of all this insider
lending. In Massachusetts they required a vote
of approval by stockholders before directors
could be loaned more than 30% of the banks
capital stock. Stockholders routinely approved
such proposals.
Writes Iamoreaux: “The indulgence with
which stockholders treated insider lending is
easier to understand once one appreciates the
importance of reputation in this information-
scarce economy” (p. 172). The point is an
important one. It needs to be qualified: while
there was little information of the sort that
we Iind now in reports to shareholders and
regulators, the people in the community were
often very well informed about what mattered
most. They were interested above all in the
character of the borrowers. They wanted their
money to be loaned to honorable, trustworthy
men. Doubtless some were fooled by scoundrels,
but the evidence from New England suggests
that the honor of substantial men provided good
assurance of the safety of capital. Lamoreaux
offers the case of one John James Dixwell, who
as president of the Massachusetts Bank arranged
a considerable loan to the Boston Brick
Manufacturing Company. As president of the
latter he felt some obligations even though he
had no personal liability for the loan. When the
company failed, ‘he made the loan good out of
his own resources. The reputations of the
directors were jeopardized whenever one of
their fellows defaulted. On this account, they
watched each other very carefully. “Given the
poor quality of information in the period’,
argues Iamoreaux, “such monitoring of insiders
by insiders may actually have been less risky
than extending credit to perfect strangers”
(p. 173).
This is greatly understated. Whatever the
availability of formalized information such as
accounts, insider lending was potentially safer
than extending credit to strangers. But the
system could not work without the benefit of
frequent face-to-face interactions. This defined
the bounds of effective personal knowledge, and
the domain within which reputation really
mattered. In consequence, the emergence of a
national lending market led to the breakdown
of this informal system of mutual surveillance
and self-regulation. Towards the end of the
century, Lamoreaux points out, a national
market developed for commercial paper. Now,
inside knowledge became inadequate. Many
banks suBered heavy losses. They began to
90 T. M. PORTER
rely on impersonal information about their
borrowers. They subscribed to the reports of
credit agencies like R. G. Dunn and Company.
They tried to work out “agreed-upon-standards
-particularly a set of objective criteria for loans
-that directors could use to monitor managers’
performance” (p. 188). In short, they left the
world of the face to face and entered that of
impersonal measurement.
Lamoreaux tells this story in terms of the
growth of information, even of economic and
institutional progress, but it might well be told
instead in the classic New England tradition as
one of moral declension. To be sure, concern
for reputation did not disappear in the financial
markets. Bradford De Long suggests in another
paper in the Temin volume that the investment
house of J. P. Morgan thrived mainly on account
of its reputation for integrity and competence.
Morgan’s operation was so huge that his
reputation could extend nationally and even
internationally. Significantly, it did not work by
assembling mountains of paper in a New York
office. Morgan participated actively in running
the companies to which he loaned, inspecting
them closely from his place on their boards of
directors. That personal knowledge and control
made it possible for him to commit for the long
term. The great symbol of financial greed seems
now to have perpetuated a kind of moral
economy into the era of big capitalism. It is a
nice irony, De Long concludes, that in our own
day “many of the intellectual children and
grandchildren of the Progressives appear to call
for a return to ‘financial capitalism”’ (p. 233).
This is not exactly what Thomas Johnson
argues in his paper on “Managing by Remote
Control”, but certainly there is a family
resemblance to both the papers on banking. The
theme of decline and moral failure is doubly
emphasized by the grandiloquent title and
subtitle of Johnson and Kaplan’s book (Johnson
& Kaplan, 1987) Relevance Lost: The Rise and
Fall of Management Accounting, thorn which
his paper draws extensively. There was once a
time, the argument begins, when management
accounting was used sensibly for limited
purposes by people who relied extensively on
other kinds of knowledge as well. So long
as ownership and management of American
companies were closely intertwined, there was
no need to back up business decisions with
reams of information. Owners were interested
in financial data, but they kept it mostly to
themselves, and they managed their firms less
on the basis of accounting information than
of knowledge of products, operations, and
customers (Temin, 199 1, p. 43). Before about
1920, “top managers were intimately familiar
with their companies’ customers and techno-
logies. They did not have to hide behind a facade
of accounting information to converse with
subordinates” (p. 49).
The serpent in this retelling of Milton appeared
to businessmen in the form of investors and
banks, offering money. All they asked in return
(to shift to Faust) was the company’s soul. They
wanted objective financial information, which
should be made available to outsiders in order
to increase public confidence and thereby
separate investors from their money. This was
not yet intolerable, so long as the accounts were
kept in their place. And for a time, Johnson tells
us, they were. Financial numbers were used to
promote the confidence of investors, and not to
run the company. Since World War II,
this separation has mostly broken down.
Government regulators, such as the Securities
and Exchange Commission, demanded huge
quantities of information, and prescribed its
form. Business schools succumbed to the dogma
of economics that the firm was a site of
quantitative rationality. They idealized the
knowledge of outsiders, the kind of knowledge
that the professors themselves possessed in
the greatest abundance. Within companies,
management was separated from the scene of
action, confining itselfmainly to a world ofpaper
and numbers. The financial side of companies
gained control over the real business of produc-
tion and sales. In short, the all-important
insider knowledge assumed second place to
outsider knowledge, the formalized, objectified
information of graphs and numbers.
Johnson writes not in despair, but with the
hope that this lamentable postwar lapse can
INFORMATION CIJLTURFS
91
be remedied, perhaps reversed. I would not public purposes, it is often necessary to measure
venture to predict whether this is possible. But according to relatively unambiguous rules
it is worth observing that this triumph of rather than to apply judgment where judgment
financial measurement at the level of corpora- is called for. The business corporation has, in
tion parallels and in some way reflects the many ways, been brought into the public
spectacular growth of quantification in the domain. In America, especially, it is a domain
domain of public policy, as evidenced in Fifty that idealizes objectivity. Even ifbusiness people
Years of Economic Measurement. The authors become convinced, as Johnson urges, that
in that volume, it may be recalled, stress the sensible management can never be reduced to
public need for measurement even of entities the manipulation of financial numbers, it may
whose “existence” was open to doubt. For not be easy to escape their tyranny.
BIBLIOGRAPHY
Alborn, T., The Other Economists: Science and Commercial Culture in Nineteenth-century Economics,
Ph.D. dissertation, Harvard University ( 199 1).
AIchon, G., The I nvfsfbl e Hand of Pl anni ng (Princeton: Princeton University Press, 1985).
Balogh, B., Cbafn Reacti on Expert Debate and Publ i c Partfcfpatfon i n Ameri can Commerci al Nucl ear
Power, 39451975 (New York: Cambridge University Press, 1991).
Barber, W., From New Era to New Deal (Cambridge: Cambridge University Press, 1985).
Berndt, E. R. SK Triplett, J. E. (eds), Fi fty Years of Economi c Measurement: Tbe J ubi l ee of tbe Conference
on Research i n I ncome and Weal th (Chicago: University of Chicago Press, 1990).
Bowker, G. & Star, S. L., Knowledge and infrastructure in international information management, in Bud-
Frierman, L. (ed.), I nformati on Acumen: Tbe Understandi ng and Use of Knowl edge i n Modern
Busi ness, pp. 187-213 (London: Routledge, 1994).
CampbeII-Kelly, M., The Railway Clearing House and Victorian Data Processing in Bud-Frierman, L. (ed.),
I nformati on Acumen: Tbe Understandi ng and Use of Knowl edge i n Modern Busi ness, pp. 51-74
(London: Routledge, 1994).
Casson, J., Economic Perspectives on Business Information, in Bud-Frierman, L. (ed.), I nformatfon
Acumen: Tbe Understandi ng and Use of Knowl edge i n Modern Busi ness, pp. 136-167 (London:
RoutIedge, 1994).
Desrosieres, A., Lapol fti que desgrands nombnx: bfstoi re de l a rafson statfsti que (Paris: Ia Decouverte,
1993).
Desrosieres, A., Official statistics and business: history, classifications, uses, in Bud-Frierman, L. (ed.),
I nformati on Acumen: Tbe Undemtandtng and Use of Knowl edge i n Modern Busi ness, pp. 168-186
(London: Routledge, 1994).
Etner, F., Hfstofre de cal cul kconomfque en France (Paris: Economica, 1987).
Fourquet, F., Les comptes de l a pui ssance (Paris: Encres, 1980).
Hacking, I., Representi ng and I nterveni ng (Cambridge: Cambridge University Press, 1983).
Hacking, I., Tbe Tami ng of Chance (Cambridge: Cambridge University Press, 1990).
Hacking, I., The Making and Molding of Child Abuse, Crfti cal I nqui ry (1991) pp. 253-288.
Hawley, E. W., Economic Inquiry and the State in New Era America, in Furrier, M. 0. & Supple, B. (eds),
Tbe State and Economi c Knowl edge, pp. 287-324 (Cambridge: Cambridge University Press, 1990).
Heclo, H. & Wildavsky, A., Tbe Prfvate Government of Publ i c Money: Communi ty and Pol i cy I nsfde
Bri ti sh Pol i ti cs (Berkeley: University of California Press, 1974).
Jasanoff, S., Tbe Fi ftb Branch: Sci ence Advi sers as Pol i cytnaJ zers (Cambridge: Harvard University Press.
1990).
Johnson, H. T. & Kaplan, R. S., Rel evance Lost: Tbe Ri se and Fal l ofManagementAccounti ng (Cambridge:
Harvard Business School Press, 1987).
Klein, J,, Ti me and tbe Sci ence of Means: The Stati sti cal Anal ysi s of Changi ng Phenomena, 1830-1940
(Cambridge: Cambridge University Press, forthcoming).
Latour, B., Sci ence i n Acti on (Cambridge: Harvard University Press, 1987).
92 T. M. PORTER
Miller, P., Accounting br Progress - National Accounting and Planning in France: a Review Essay,
Account@ Organfza~fons, and Society (1986) pp. 83-104.
Miller, P. & O’Leary, T., Accounting and the Construction of the Governable Person, Accounting,
OrganrZatfons, and Society (1987) pp. 235-265.
Mitchell, W. C., Quantitative Analysis in Economic Theory, American Economfc Review (March 1925)
pp. 1-12.
Morgan, M., A H&x-y of Econometrtc I deas (Cambridge: Cambridge University Press, 1990).
Petersen, W., Politics and the Measurement of Ethnic&y, in Alonso, W. & Starr, P. (eds), The Polftics of
NumBeqpp. 187-233 (New York: Russell Sage Foundation, 1987).
Porter, T. M., Quantification and the Accounting Ideal in Science, So&al Studies of Science (1992a)
pp. 63ti51.
Porter, T. M., Objectivity as Standardiiation: the Rhetoric of Impersonality in Measurement, Statistics and
Cost-Benefit Analysis, Annals of S&o&ship (1992b) pp. 19-59.
Porter, T. M., Rigor and Practicality: Rival Ideals of Quantification in Nineteenth-century Economics, in
Mirowski, P. (ed.), Natural Images in Economic Tbougbt, pp. 128-170 (New York: Cambridge
University Press, 1994).
Porter, T. M., Trust in Numbers The Pursuit of Objectivity in Science and Public Life (Princeton:
Princeton University Press, 1995).
Temin, P. (ed.), I nside ibe Business Enterprise: Historical Perspectives on the Use of I nformation
(Chicago: University of Chicago Press, 1991).
Yates, J., Control Tbrougb Communfcatfon: The Rise of System in American Management (Baltimore:
Johns Hopkins University Press, 1989).
Yates, J., Evolving Information Use in Firms, 1850-1920, in Bud-Frierman, L. (ed.) I nfwmatfon Acu~
The LJ ?kierstunding and Use of Knowledge in Modern Business, pp. 26-50 (London: Routledge, 1994).
doc_552848025.pdf