ArticlePDF Available
The New Institutionalism: What Can
It
Learn from the
Old?
SANFORD
M.
JACOBY
*
The new institutional labor economics is a promising development, but it
has faults that could be remedied by an infusion of theoretical and
methodological insights from the old institutional approach. This claim is
illustrated by a critical analysis of three
key
concepts: asset specificity,
deferred rewards, and opportunism. The essay concludes with
a
set of
methodological precepts to guide future research.
OVER
THE
YEARS,
economics has persistently expanded the range
of phenomena that it seeks to explain. Among recent expressions of this
theoretical expansiveness is the “new efficiency-oriented institutional labor
economics” (NEO-ILE), of which the papers in this symposium are an
excellent sampling. The NEO-ILE takes efficiency-oriented concepts derived
from neoclassical theory-human capital, agency theory, transactions costs-
and applies them to the analysis of various real-world institutions, chiefly
those that support employment relations in organizations. Yet many of these
institutions had been identified and analyzed by an earlier generation of
labor economists, whose approach can be called simply “institutional labor
economics” (ILE). Thus, the NEO-ILE has
a
mixed parentage. It has roots
in the ILE approach, but it draws more heavily on neoclassical theory and
is more rigorously analytical than was the ILE.
This cross-breeding has resulted in great variation in the kinds of research
produced by those working on NEO-ILE topics. Some of it has
a
strong
empirical, policy, and interdisciplinary orientation. Its use of theory is
eclectic and realistic, as with efficiency wage theory. This strand of the
*
Anderson Graduate School of Management, University of California
at
Los
Angeles. The
author is very grateful for comments and criticism received from
Phil
Beaumont, Peter
Cappelli, Stanley Engerman, Bruce Kaufman, Paul Osterman, Michael Reich, and Keith
Sisson. All usual disclaimers apply.
INDWTRLU
RELATIONS,
Vol.
29,
No.
2
(Spring
1990).
C
1990
Regents
of
the University
of
California
0019iS676i~i525i~~~~10.00
316
The
New
Institutionalism
t
317
NEO-ILE has generated research that is at least as good as (and in some
cases better than) that produced by the ILE approach. Other parts of the
NEO-ILE research corpus are more problematic. There is
a
tendency to
rationalize employment practices in functional or efficiency terms without
regard to their historical-causal (how did the practice arise?) or empirical
(how do we measure theoretical constructs and how much variation do they
explain?) complexities. Noneconomic factors are either ignored or derogated
as transitory or secondary phenomena. As elsewhere in the discipline,
modeling
is
favored over facts and empiricism, which can and sometimes
does lead to theories that are based on wrong or garbled facts, and are of
little use for policy.
In this essay, I argue that the institutional tradition still has much to offer
modern labor economics. Some labor economists know that and borrow in
various ways from the tradition. They include economists who work on
NEO-ILE issues as well as others whose research is less efficiency-oriented
and closer, in various respects, to the older ILE tradition. In this latter
group are such scholars as Katharine Abraham, Francine Blau, Clair Brown,
William Brown, William Dickens, Richard Freeman, Robert Frank, Harry
Katz, Daniel Mitchell, Paul Osterman, Peter Philips, Michael Piore, David
Soskice, Myra Strober, and many others. Members of this group need to
be more assertive about what it is that makes their work distinctive and
differentiates it from both the neoclassical and the NEO-ILE approaches.
Other economists, including some whose work falls under the NEO-ILE
rubric, are ignorant or contemptuous of institutionalism. This essay is
intended to goad them
a
bit. The discussion begins with an analytic history
of institutionalism and then focuses specifically on the ILE approach and
its intellectual concerns. Three succeeding sections examine key NEO-ILE
concepts-asset specificity, deferred rewards, and opportunism-and subject
them to a critical evaluation using ideas derived from the ILE approach.
The final section offers some methodological suggestions to guide future
work in the NEO-ILE area.
Early
Ins
ti tu tionalism
As
Hausman
(1989)
and others show, the view of methodology held by
most orthodox economists today is quite similar to that espoused by
nineteenth century classical and neoclassical economists like
J.
S.
Mill and
J.
N.
Keynes:
A
set of assumptions is generated by inductive observation
or by introspection; from these assumptions are deduced probabilistic
statements (or laws,
ceteris pan'bus)
that have predictive and explanatory
consequences. Empirical analysis then determines whether the consequences
318
/
SANFORD
M.
JACOBY
are correct and
if
causal factors have been omitted. This view has had
numerous critics over the years, including Friedman
(1953).
Here,
I
focus
on the criticisms made by the institutional economists.
Institutionalism swept American economics in two waves: the first appeared
in the eighteen seventies and eighties, and included the work of Ely,
Seligman, and Patten; the second-and more influentialrrested during the
two decades after
1900,
and included the work of Veblen, Mitchell, and
Commons. In those days, neoclassical economics was only weakly committed
to empirical research,
a
stance which the institutionalists roundly criticized.
Drawing on the inductivism of the German historical school and on the
pragmatist credo that the truth of an idea is to be judged by its consequences,
the institutionalists called for
a
stronger factual and empirical orientation in
economics. Although institutionalism itself was criticized for excessive
description and fact-gathering, at its best it combined the pragmatic emphasis
on facts and chance to produce-as in Mitchell’s work-a statistical
approach to economic research that has its modern counterpart in empirical
econometrics as carried out
at
the
NBER
(founded by Mitchell) and
elsewhere.
The institutionalists were driven by an abiding concern with public policy.
They found the neoclassical approach unhelpful in formulating policy and
often hostile to the idea that economic and social problems could be
ameliorated by the state. Indeed, it was out of their diverse policy concerns-
the “labor” question, industrial regulation, municipal ownership, trade
protection, public administration-that the institutionalists developed their
two main criticisms of orthodox theory: first, that its predictions diverged
from empirical reality (they saw the real world as a place where state
policies contributed to economic growth; trade unions raised wage levels;
unemployment was pervasive; and trade protection facilitated economic
development); and second, that these discrepancies arose from
a
lack of
realism in the model’s assumptions. Against those assumptions, the
institutionalists offered their own: indeterminancy; endogeneity; behavioral
realism; and diachronic analysis.
Indeterminacy.
Whereas the orthodox model assumed perfect competition
and unique equilibria, the institutionalists pointed to pervasive market power
and to indeterminacy even under competition. Commons
(1924),
for example,
The American Economic Association was founded in 1885 by Ely (1886) and other early
institutionalists, who angered their more orthodox colleagues by including a phrase in the
AEA’s opening statement that called the state “an agency whose positive assistance is one
of
the indispensable conditions of human progress.” For a more detailed analysis of
institutionalism and its origins, see Jacoby (1988).
The
New
Institutionalism
I
319
argued that in the labor market, the employer possessed superior information,
resources, and bargaining power as compared to individual workers; the
Webbs (1920tborrowing from Edgeworth-claimed that the labor contract
was indeterminate regardless of the degree of competition, leaving its terms
to be specified by custom, bargaining, or law.
Endogeneity.
In the orthodox view, an individual’s wants were taken as
given and exogenous to the realm of economic need satisfaction dominated
by efficiency forces. The institutionalists held instead that market exchange
was mediated by social institutions that determined, and were
at
the same
time determined by, individual wants and behaviors.
As
Commons (1934)
put it, the transaction between two persons, the basic unit of economic
analysis, was subject to the working rules of going concerns, but these rules
were expressions of collective action and thus subject to change through
human volition. Veblen (1919, p. 239) expressed the same idea when he
said that economic conduct is “hedged about and directed by [social]
relations.. .of an institutional character,” while
at
the same time, “the growth
and mutations of the institutional fabric are an outcome of the conduct
of
individual members of the group.”
Behavioral realism.
The orthodox view took from utilitarianism the
assumption that
homo economicus
was guided by rational self-interest, whereas
the institutionalists derived from pragmatism and other sources their belief
that economic theory had to be based on psychological facts rather than on
assumptions about economic behavior. From the new empirical disciplines
of cognitive and social psychology, they developed their emphasis on habit
as
a
component of rationality. Not only do habits have an historical dimension
(they cause past choices to constrain a person’s present ones), they also have
social and cultural origins and can promote collective action. Commons
(1950) saw the latter as a key feature of modern capitalism.
Diachronic analysis.
In the orthodox view, economic theory was synchronic:
an abstraction from reality that isolated its transhistorical and universal
aspects. Following Max Weber, the institutionalists asserted that the
abstractions of economic theory were neither timeless nor placeless but
instead were an ideal type-an enhancement of features unique to modern
Western capitalism.’
As
such, economic theory was “not
so
much an account
*
Commons (1934) acknowledged his debt to Max Weber, and sociologists again are
beginning to consider the relations between economy and society (e.g., see Granovetter,
1985).
320
I
SANFORD
M.
JACOBY
of how men do behave as an account of how they would behave if they
followed out in practice the logic of the money economy” (Mitchell,
1937,
p.
371).
The institutionalists realized that nothing was wrong with this
approach to theory
so
long as the distance between abstraction and reality
was not great. But their empirical research led them to view this distance
as substantial, at least for some parts of orthodox theory, and also as
historically variable, because the economy perpetually was “a moving,
changing process” (Commons,
1924,
p.
376).
Hence, they insisted that
diachronic analysis-how the economy acquired its features and the conditions
that cause those features to vary over time and place-had to be part of
economics, alongside synchronic abstraction.
It
is an error of long standing to charge that the institutionalists discarded
economic abstraction and the deductive method. Possibly the accusation
could apply to Veblen and some of the early institutionalists, but it was not
true of Commons, Hoxie, Mitchell, and others who followed them. Yet it
is
true that the institutionalists were unable to resolve their dispute with
orthodoxy. Commons thought that the way to proceed was to base economic
theory on realistic assumptions revealed by empirical research within and
outside of the discipline. “The problem now,” he explained, “is not to create
a different kind of
economic+‘institutional’
economics divorced from
preceding schools-but how to give collective action, in all its varieties, its
due place throughout economic theory” (Commons,
1934,
p.
5).
Mitchell
had much the same thing in mind when he criticized his friend Veblen for
failing to state his ideas in testable terms and said that economic methodology
should involve a “passing back and forth between hypothesis and observation,
each modifying and enriching the other”
(1937,
p.
302).
But neither man
produced even a partially modified version
of
neoclassicism, although both
left a wealth of ideas and data that questioned some of its foundations.
The institutionalist critique had its greatest impact on the field of labor
economics, a specialty that emerged in the twenties (McNulty,
1980).
Not
only were labor markets the site of numerous theoretical anomalies, but they
had features that overlapped the concerns of other disciplines to a greater
extent than other applied subjects in economics. Both the traders and the
commodity traded in labor markets-human effort-were more likely to be
affected by psychological and social factors than, say, brokers and bonds
(but see Shiller,
1984).
Also, labor economics was a field in which practical
issues were salient and unavoidable. The “labor” question, broadly defined,
was arguably the most pressing problem in American society through the
fifties, long after other nations had adapted to trade unionism and the welfare
state.
The New Institutionalism
I
321
Institutional
Labor
Economics
There was considerable intellectual continuity between the institutionalists
and the institutional labor economists (the ILEs) who succeeded them in
the twenties and thirties (e.g., Douglas, Lescohier, Slichter, Wolman) and
in the forties and fifties (e.g., Kerr, Dunlop, Lester, Reynolds, Ulman).3
Like their forebears, the ILEs deprecated the lack of empiricism in orthodox
economics and its limited relevance to public affairs. They were deeply
concerned with policy problems and actively involved in solving them: the
first generation cut its teeth on industrial relations and labor market issues
during the First World War; the second generation did much the same
thing during the thirties and the Second World War. The ILEs were struck
by discrepancies between orthodoxy’s predictions and the realities that they
observed from their posts in government and through their more systematic
empiricism.
The ILEs traced the anomalies they observed-including nonclearing
labor markets, wage rigidity, persistent wage differentials, labor immobility,
administered (internal) labor markets, and bargaining power-to the same
flaws in the orthodox model that the institutionalists had earlier identified.
And they championed many of the same alternative assumptions:
endogeneity
(the ILEs thought that market behavior and social relations were mutually
determined, as in the ILE emphasis on custom and equity in wage setting);
diachronic analysis
(the
ILEs
focused on time- and place-specific elements in
economic theory and conducted theoretically informed comparative and
historical studies, as in the works published under the aegis of Kerr and
Dunlop’s Inter-University Study of Labor project); and
indeterminacy
(because
of the theoretical advances made in the thirties-by Keynes, Robinson,
Chamberlin, and others-the second generation of ILEs gave more attention
than did their predecessors to imperfect and nonprice competition, yet the
effect was the same: an emphasis on indeterminacy in the orthodox model,
leaving room for the exercise of discretion, power, and social norms in the
labor market).4
Some have recently asserted a disjuncture between the institutionalists and the
ILEs
of
the forties and fifties, calling the latter “postinstitutionalists” (Segal,
1986)
or “neoclassical
revisionists” (Kerr,
1988).
As
I
explain below, there
were
differences, but they were not
sharp.
The
ILEs
had relatively less to say about behavioral realism, in part because the academic
division of labor began to change during the thirties as new fields like personnel psychology
and management split off from labor economics. But these fields became part of the new
interdisciplinary umbrella of industrial relations, under which most of the
ILEs
also chose
to gather.
From
here, the
ILEs
closely followed developments in those fields and participated
in
their debates. See note
10.
322
/
SANFORD
M.
JACOBY
Methodology.
Where the ILEs, especially the second generation, differed
from their forebears was in their methodological approach to resolving these
anomalies. Rather than espouse (much less attempt) the creation of a new
body of theory based on realistic assumptions, the ILEs built a wide range
of eclectic, middle-level theories that bridged the abstractions of pure theory
and the institutional realities of American labor markets. When received
theory failed to predict with a high degree of precision, the ILEs added
more realistic elements-diachronic, nonprice, social, political-into their
explanations. Critics faulted the ILEs for failing to develop a substitute for
neoclassical theory. The middle-level approach was, however, a reasonable
and fruitful way to proceed, given the previous generation’s failures in this
area, and, especially, given the growing realization that neoclassicism’s
theoretical core was surrounded by impenetrable bands of conditional clauses
(see Hutchison,
1938;
Latsis,
1976).5
Nor was this approach unique to labor
economics. Industrial organization economics of the forties and fifties, both
in the
U.S.
and the
U.K.,
generated empirical findings that were at variance
with neoclassical predictions. Like the ILEs, the industrial organization
economists developed a range of eclectic, middle-level,
or
bridging theories
to account for anomalies like rule-of-thumb pricing and the shape
of
long-
run cost curves (e.g., Boulding,
1952;
Heflebower,
1952;
Walters,
1963).
The ILEs’ approach was not invincible, of course.
A
disinclination toward
methodology left them vulnerable to Friedman’s
(1953)
clever attack on
realism and descriptive accuracy and to Machlup’s
(
1946)
claim that middle-
range theories dealt with mere deviations
or
disturbances from long-run
equilibrium. On the heels of those criticisms came other developments that
undermined the ILE approach. By the late fifties, the public concerns that
had animated the ILEs-remedying labor market inequalities through
industrial relations and macro-welfare policies-were less pressing. Moreover,
the mathematization
of
economics cut the ILEs off from the generation of
labor economists trained in the sixties and seventies. The ILEs were
perceived as dated, and their eclectic, middle-range theories proved difficult
to model in formal terms.
At
the same time, labor economics experienced
a neoclassical resurgence with the rise of human capital theory. Increasingly,
the thrust was to reconcile labor economics with orthodox theory.
Nevertheless, during the seventies and eighties a new trend emerged in
labor economics. Various anomalies from the ILE tradition became the
subject of theoretical concern: internal labor markets, wage differentials,
In a recent elegiac essay, Clark Kerr
(1988,
p.
21)
faults his generation for failing
to
produce “an integrated statement”
or
“consistent theory.” Kerr is
too
hard on himself and
his colleagues. True, better modeling could have been done, but the
ILEs
made the right
decision by not attempting a full-blown alternative to neoclassical theory.
The
New
Institutionalism
I
323
wage rigidity, and unionism. Economists again began to study employment
institutions, giving rise to the “new efficiency-oriented institutional labor
economics” (NEO-ILE). As previously noted, the NEO-ILE is a motley
body
of
research ranging from work that is avowedly institutional to more
neoclassical and even Marxian approaches. But throughout the NEO-ILE
literature there are theoretical problems which could be rectified by a more
conscious application of insights from the ILE tradition. Below, I substantiate
this claim by critically examining three central NEO-ILE concepts-asset
specificity, deferred rewards, and opportunism.
Internal Labor Markets and Asset Specificity
Asset specificity in various forms plays a key role in the NEO-ILE
literature on internal labor markets and related topics. Doeringer and Piore’s
(1971)
now-classic book on internal labor markets was one of the earliest
attempts to provide an efficiency-oriented explanation for a phenomenon
first identified and discussed by the preceding generation of ILEs. Borrowing
from Becker
(1964)
and focusing on the stability of employment in internal
labor markets, Doeringer and Piore gave causal primacy to skill specificity
and the informal process by which firm-specific skills are acquired. Skill
specificity leads employers to stabilize employment and reduce turnover
so
that they can recoup their firm-specific investments in employees.
At
the
same time, stability facilitates the informal process of specific-skill acquisition,
which occurs on the job through learning-by-doing. Doeringer and Piore
also made an historical argument that employment had become increasingly
stable due to growing use of mass production technology, whose detailed
division of labor necessitated firm-specific skills.
While subsequent research has confirmed an historical shift over the last
century in the stability of industrial employment (Carter,
1988),
there is
little evidence that the shift resulted from a growing reliance on firm-specific
techniques or skills. In fact, the evidence suggests that the opposite was
true: that technology and job skills became less, rather than more, firm-
specific over time. By the early nineteen hundreds, the
U.S.
was developing
a capital goods industry that allowed firms to purchase identical machinery
from national vendors instead of having to craft their own, as in the
nineteenth century (Jacoby,
1979).
As
one manager said in
1917,
all firms
“can buy the same kinds
of
machinery if they know where to get it; or they
can design the same kind. Processes cannot now be kept entirely secret”
(Hopkins,
1917,
p.
2).
Moreover, as contemporary economists were well
aware, this standard technology had the advantage of lowering training costs
by narrowing job content and de-emphasizing skill specificity (Douglas,
324
/
SANFORD
M.
JACOBY
1921).
To escape from this paradox, Williamson (1975) expanded the sources of
skill specificity beyond Doeringer and Piore’s “technological specificity” to
include a host of organizational factors that make incumbents valuable
to
employers: team accommodations, informal process innovations, and
knowledge of codes and procedures
.6
But similar historical problems adhere
even to this expanded definition. Take, for example, a firm’s codes and
procedures. One of the major findings
of
modern industrial sociology is that
work organizations today are more rationalized and bureaucratic than in the
past (Bendix, 1956; Jacoby, 1973). As a result, employees (whether in steel
mills or universities) find other organizations in the same industry to be
more similar-in terms of their codes and procedures-than was true one
hundred years ago. Nineteenth-century firms were stamped by the identities
of their owners, workplace routines were unstandardized, and workers had
to learn the idiosyncracies of foremen who ran their departments like
independent fiefs.
The specificity concept also has empirical problems, chiefly the fact that
idiosyncratic elements can be found in most jobs. Restaurant work provides
a prime example of an unstructured labor market (one lacking internal labor
markets). Yet this work entails process, team, and communications
idiosyncracies as unique and elaborate (from restaurant to restaurant) as
those found in semi-skilled auto assembly jobs, the
focus
classicus
of a
structured internal labor market (Whyte, 1948). Surely then, to paraphrase
Williamson (1979, one cannot place primary reliance on idiosyncracy to
produce structure, for if idiosyncracy is a common employment feature,
how is the coexistence of structured and structureless labor markets to be
explained?
The preceding claim admittedly lacks detailed documentation, but the
same is true of the claims made by adherents of idiosyncracy. Indeed, the
major problem with the concept is its vagueness and lack of empirical
content.
No
one has devised a way to measure skill specificity directly.
At
best, one must make the heroic assumption that the amount of specific skills
a worker possesses is proxied by years of on-the-job experience. But that
raises even more difficulties than the analogous assumption that years of
schooling is a proxy for general human capital. At least in the case of
schools, indirect evidence exists (such as test scores) to assess the plausibility
of the notion that additional schooling augments general human capital.
There is no similar evidence to show that additional job tenure raises one’s
Asset specificity is central to Williamson’s approach.
Its
absence, he says, would “vitiate
much
of
transaction cost economics”
(1985,
p.
56).
The
New
Institutionalism
I
325
stock of firm-specific human capital. In fact, the evidence-which shows
that tenure is rewarded independently of productivity-runs in the opposite
direction (Abraham and Medoff, 1983).
These criticisms do not come directly out of the ILE tradition, but they
are consistent with its emphasis on facts, empiricism, and historical
specificity. That tradition needs to be reckoned with, rather than dismissed,
by those seeking to rationalize internal labor markets in terms of asset
specificity. Still, as economists are fond of saying, it takes more than facts
to kill
a
theory; it takes another theory. In this case, at least, other NEO-
ILE theories of internalization are available that are more consistent with
the facts. However, those theories, too, could benefit from an infusion
of
the ILE approach, especially its attention to the social and historical contexts
in which employment institutions are rooted.
Internal Labor Markets and Deferred Rewards
An efficiency-oriented alternative to idiosyncracy is Goldberg’s (1980,
1984) work on internal labor markets. Using an eclectic combination of
agency theory (Lazear, 1979), implicit contract theory of the “invisible
handshake” variety (Okun, 198
l),
radical theories of bureaucratic control
(Edwards, 1979), and elements from business history (Chandler, 1977),
Goldberg argues that, as firm size and throughput rates rose in the late
nineteenth century, worker turnover and indiscipline became more costly to
employers. At the same time, larger firm size made traditional methods of
control (using foremen to monitor and dismiss refractory workers) less
feasible and more costly. The search for alternatives led employers to the
idea of penalizing premature separations and promoting self-monitoring
through the use of deferred compensation (i.e.
,
internal promotion, pensions,
and other plans that tie pay to seniority by tilting age-earnings curves, much
like posting
a
bond). Use of these less visible control structures stabilized
employment: directly, through reductions in voluntary mobility; and
indirectly, through employer-initiated rules restricting unfair, premature
dismissal. By imposing rules on themselves that curbed their own opportun-
ism, employers ensured that workers would not discount the value of their
deferred compensation and that the latter would retain its incentive effect.
Goldberg’s interpretation generates internal labor markets without idiosyn-
cracy.
It
also has the advantage
of
being studded with large chunks of
historical reality. Nevertheless, at several key points the story does not
fit
the
historical record. These disjunctures point to
a
more complex interpretation of
internal labor markets, one that is less supportive of Goldberg’s (and others’)
view that economic efficiency incentives are sufficient to account for observed
326
/
SANFORD
M.
JACOBY
institutional outcomes. First, the notion that an increase in average firm size
caused internalization is a fallacy of the
post
hoc,
ergo
propter
hoc
variety.
Undoubtedly there was some relationship in some cases, but, on average,
the two events had only a loose and widely varying temporal correlation.
Giant manufacturing firms were a common feature
of
the industrial landscape
by 1890, yet in many of them internalization-whether measured by means
(pensions, internal promotion lines, employer restraints on dismissal) or by
ends (geographic mobility, labor mobility, average tenure) did not occur
until four
or
five decades later (Ross, 1958). One is left wondering why
some large employers were
so
slow to realize the efficiency advantages of
internalization and deferred rewards, especially when size-driven changes
occurred so rapidly in marketing, production, accounting, and other spheres
of management. Moreover, medium-sized firms (those with fewer than 1,000
workers) were often among the first to abolish the traditional system of
factory labor administration.
Second, key methods by which internalization supposedly was effected
did not become prevalent until the fifties, which is
after
internalization had
occurred and the modern pattern had been set for employment stability,
tenure, and mobility. Three examples: (i) The proportion of wage-earners
eligible to receive severance pay stood at less than 9 per cent in 1946 but
rose
to 25 per cent by 1964. (ii) Only 2 per cent of wage-earners received
pensions in 1929, and by the end of World War
I1
this had risen to only 8
per cent. Again, the big increase came after the war, when the proportion
rose to 73 per cent in 1964. (iii) Employer-initiated restraints on dismissal,
as measured by the proportion of firms having defined dismissal rules, stood
at only 16 per cent in 1940 but rose
to
43 per cent in 1964 (Jacoby, 1986;
Gordon, Edwards, and Reich, 1982). In other words, deferred rewards and
the structures governing them were a consequence, rather than a cause, of
internalization. They had less to do with employer control than with
employee preferences (deferred rewards are valued by stable employees) and
shifts in power and social norms (dismissal rules protect, and severance pay
liquidates, vested job rights).
Third, internal labor markets did not spontaneously develop in an ever-
growing number of firms. Typically, they were adopted during two periods
of crisis for the older system of labor administration-the late nineteen tens
and the decade from 1935 to 1945, years during which employee bargaining
power was high as a result of labor shortages and/or union threat effects
(Jacoby, 1985). This pattern suggests that external pressure was critical in
the shift to more structured employment practices, whose putative efficiency
incentives many managers remained skeptical
of.
Some managers were
reluctant to stabilize employment out of fear that it would raise labor costs
The
New
Institutionalism
I
327
by forcing them to give regular pay increases to incumbent employees.
Others held tight to the conviction that intimations of job security eroded
discipline and that only close supervision would keep employees in line.
The steady increase from 1900 to 1960 in the ratio of foremen to production
workers in American manufacturing firms is consistent with this belief
(U.S.
Bureau of the Census).
Normative
efficiency.
It would be wrong to suggest that efficiency was
irrelevant
or
that internal labor markets were a random development. Size
did
sometimes influence outcomes, along with other economic variables like
product demand stability, inventory accumulation costs, and product
diversification (Piore, 1975; Jacoby, 1989). But these variables take
us
only
part of the way. One critical element missing from Goldberg’s story
is
worker preferences, in the form of a “second generation effect.” Workers
born in the U.S. were more likely than mobile immigrants to value
attachments to particular employers and communities. That effect was
already felt in the late nineteenth century and then gradually increased over
time. It was given major impetus by post-1920 changes in immigration law,
which halted much of the reverse-migration to Europe. Workers who might
have emigrated before the change in the law now sank roots in the United
States. As well, the law caused a drop in immigration, which brought a
decline in the proportion of rootless newcomers in the labor force.
Consequently, the children of the immigrants who flocked here between
1880 and 1915 made up a growing proportion of the labor force. Presumably,
they wanted to remain in the workplaces and communities where their
parents had finally settled. Hence, workers’ overwhelming concern with job
security during the twenties and thirties stemmed from a desire to protect
new forms of communal stability from the disruption caused
by
job loss
(Jacoby, 1983).
More was involved in the second generation effect than an inexplicable
shift in time-discounting preferences. Immigrants and their children were
affected by the norms and attitudes of American society, which emphasized
equality, rights, and lack of deference to authority. The industrial workplace,
however, sharply contradicted those values, with its visible exercise of the
foreman’s power and authority; its lack of due process in labor allocation
and dismissal; and its rankling status differences between blue- and white-
collar workers, with the latter, but not the former, receiving implicit
or
explicit promises of job security (Jacoby, 1982). The tension between social
norms and industrial realities built
up
and then erupted during the nineteen
tens and again (in more organized fashion) during the thirties and forties.
Out of these episodes emerged a more stable, impersonal, and rule-bound
328
/
SANFORDM.
JACOBY
system of employment.
It is possible to rationalize all of this in efficiency terms, as when Stiglitz
(1987, p. 49) argues that “it is in the interests of employers to construct
employment relationships that are attractive to workers. In competitive
markets firms will be ‘forced’ to provide such efficient contracts.” But
because what is attractive to workers is historically and socially specific,
Stiglitz’s linkage between what he calls the “positive” and “normative”
aspects of contracting entails a concept of efficiency that also is historically
and socially specific. It can be termed “normative efficiency” and is akin to
Leibenstein’s (1976) x-efficiency concept.
It
is different from the kind of
efficiency implied or expressed in synchronic NEO-ILE theorizing about
labor market institutions, although the two efficiency concepts are best
thought of as complements or joint constraints. Thus, in the case at hand,
internal labor markets (and employment practices more generally) are a
blend of adaptations: to forces that produce allocative or transactional
efficiency and to specific historical and cultural forces that produce normative
efficiency.’ Moreover, if Stiglitz’s “competitive markets” condition is not
met, then there is room for more complexity in the model-including, on
the one hand, employer sluggishness in responding to worker preferences;
on the other, the use of collective action via unions or government to speed
up the response. Without getting into what is usually an abstract and fruitless
debate over whether or not power relations shape the organization of work,
suffice it to say that most NEO-ILE models either ignore the exercise of
unilateral power or doubt that the concept can explain very much (Macneil,
198 1; Williamson, 1985). If, however, labor markets are imperfectly
competitive, then power-in the form of collective action-has a role to
play. It can speed up the process of “efficient” institutional change or cause
the adoption of institutional forms that are inefficient. Collective action can
be pursued by workers-as in the wholesale replacement during the thirties
of millions of individual at-will employment contracts by collective contracts
requiring just cause-or by employers, as in the recent political jockeying
over dismissal and notice legislation.
Even if labor markets are competitive, externalities across firms can act
to deter the provision of a sufficient supply of attractive jobs. These
externalities reveal the endogeneity of the employment relationship, taking
us into the realm of social costs, social benefits, and social policy, issues not
often considered in NEO-ILE models. One example of such an externality
was given by Slichter (1919, pp. 426-431), based on his observation of firms
The median-voter model
of
unionism (Freeman and Medoff,
1984)
can be classed as a
normative efficiency concept.
The
New
Institutionalism
I
329
and labor markets during the nineteen tens. Slichter was concerned that
market forces would fail to end the prevalent “drive policy” of close
supervision and unstable jobs because the policy was “profitable” and
“practical” for employers “dealing with some classes of workers.” Yet, he
said, what was privately rational had undesirable consequences for “the class
interests of [all] employers” as well
as
for “the general social welfare.” The
drive policy, explained Slichter, “intensifies class consciousness of labor”
and “wears workers out to such an extent that it is equivalent to a decrease
in the general supply of labor.”
To
encourage wider pursuit of “a liberal
and broadly conceived policy,” he urged governments, universities, and
employers’ associations to take a more active role.
Indeed, around this time state and federal agencies first began to promote
“socially responsible” employment practices, many of which fostered
internalization. The government took steps such as training hundreds of
personnel managers during the First World War, legislating anti-layoff
incentives during the mid-thirties, and reforming private dismissal practices
during the Second World War. The Wagner Act, too, was partly motivated
by the idea that unions would bring industrial stability not only by curtailing
strikes but by hastening the demise of drive practices (Casebeer,
1987).
Diachronic
specificity.
Normative efficiency factors, externalities, and mar-
ket imperfections add to the specificity of the employment relationship
and enmesh it in a web of institutions (unions, legislation, schools) that vary
over time and place. Given this specificity, one would expect to find
substantial cross-national variation in the structure of internal labor markets
and in the labor market outcomes to which they give rise. This is precisely
what recent studies have uncovered. For example, as compared to the
U.S.
or
the
U.K.,
Japan and Germany have internal labor markets that are less
finely structured and less market- (versus organization-) oriented, while their
external labor markets have lower mobility levels (Dore,
1973;
Cole,
1979;
Osterman,
1988).
Others have found that large French and German firms-
although neighbors and trading partners-nevertheless exhibit systematic
variations in their internal wage structures and mobility patterns. These
variations are traceable to differing national educational systems and how
the latter are linked to firms (Maurice
et
al.,
1986).
Orthodox economists assert that such variations can safely be abstracted
from because they have but “negligible effect”
as
compared to the more
fundamental outcomes produced by long-run forces of competition (Reder,
1989,
p.
457).
In other words, it is believed that one can neatly disentangle
economic/universal from noneconomiclspecific factors, and that the former
will have causal primacy, at least in the long run. But when it comes
to
330
/
SANFORDM.
JACOBY
studying labor markets, it would seem safer or, to use an old term, more
pragmatic, to hold an agnostic position on these issues and to take each
analytical problem as it comes. Otherwise, one runs the risk of making
theory a rationalization for, rather than an explanation of, observed
phenomena.
*
To make this point concrete, consider the various theories of wage rigidity
that appeared during the seventies as an offshoot of work on internal labor
markets. The theories explained wage rigidity as the result of optimizing in
the face of risk aversion or turnover costs. These were elegant models, but
they were framed solely in functional-synchronic terms without regard to
empirical or diachronic factors. The risk-shifting version eventually broke
apart on hard factual reefs: the limited incidence and scope of real wage
protection (e.g., COLAS) and of income replacement (e.g., SUBS and UI)
for American workers on layoff (Oswald, 1987; Brown, 1988). Other research
pointed to substantial historical and international variations in wage rigidity,
which neither theory could explain as well as models that incorporated
specific institutional factors. Thus, with respect to the substantial increase
in
U.S.
wage rigidity since the twenties, Mitchell (1985, p. 36) said, “Perhaps
risk aversion and turnover costs increased after the 1920s, but the most
significant changes that occurred involved unionization, new public policies,
and changing social expectations.” As for international variations in wage
rigidity (with the
U.S.
an outlier on this dimension), the best explanation
again came from theories that emphasized the institutional structures affecting
wage determination: the level, timing, and extent
of
bargaining; union
ideologies and norms; and measures of corporatism, such as implicit social
contracts or the political integration of organized labor (Bruno and Sachs,
1985; Flanagan, Soskice, and Ulman, 1983; Gordon, 1982; Helliwell, 1988).
The preceding research on wage rigidity demonstrates that economists
working on NEO-ILE topics do not always adhere to an orthodox position.
But it is fair to say that an eclectic, interdisciplinary approach-blending
abstract economic and specific social and historical factors-remains the
exception rather than the rule. Among NEO-ILE economists, the dominant
view of other social science perspectives is that they are “rivals” (Williamson,
*
The criticism advanced here is similar to Dow’s
(1987,
p.
26)
argument that neoclassical
and transactional theories are functional rather than causal explanations of institutional
outcomes. Belief in the efficiency of observed structures, says DOW, leads to functionalist
explanatory statements of the sort: “governance structure
X
exists because efficiency
requirements dictate
X
for transactions of type
Y...
Causal explanations, on the other hand,
describe how later structures have emerged out of earlier ones.”
The
New
Institutionalism
I
331
1988, p. 183) to economic orthodoxy-substitutes rather than complements-
and, judging by citations in NEO-ILE publications, inferior goods to
Opportunism and Shirking
Unlike wage or employment rigidity, opportunism is a topic that most
ILEs failed to consider. Nevertheless, the problem of reducing opportunistic
behavior in the workplace is interesting and provides
a
potential link between
the NEO-ILE and organizational research in other disciplines. The problem
lies
at
the heart of
a
variety of NEO-ILE models, including efficiency wage
theory (Katz, 1986; Shapiro and Stiglitz, 1984), agency theory (Pratt and
Zeckhauser, 1985), and transactions-cost theory. Opportunism is “self-
interest seeking with guile. This includes but is scarcely limited to more
blatant forms, such as lying, stealing, and cheating” (Williamson, 1985, p.
47).
In the workplace, it often takes the form of shirking, which is associated
with incomplete or distorted disclosure of information regarding employee
effort.
To
mitigate opportunism in employment requires either hierarchical
monitoring and governance procedures or less visible control mechanisms
like deferred rewards and supra-market clearing wages that increase the
worker’s cost of job loss when malfeasance is detected. Some Mamian models
of employment rely on similar notions: that as the cost of job loss decreases,
opportunism and the need for control both increase (Weisskopf
et al.,
1983).
For some, the similarity is cause for hopes of reconciliation between
neoclassical and Marxian economics (e.g., see Putterman, 1986), although
Bowles (1985) asserts that the Marxian model differs fundamentally from
the NEO-ILE’s Hobbesian approach because it specifies opportunism as a
feature of capitalist society rather than
of
human nature.
This is
a
key difference and Bowles is correct in criticizing the NEO-ILE
approach for failing to endogenize the workplace in
a
social context. But
the Marxian model shares with the NEO-ILE approach some dubious
features, chiefly an overemphasis on one side of human nature in capitalist
workplaces (shirking and conflict) and
a
tendency to ignore other sides-
altruism, morality, trust, cooperation-or to disparage them as false
consciousness or disguised self-seeking. Both the NEO-ILE and Marxian
models tend to assume certain invariant and fundamental features of human
behavior or of capitalism, an approach that pragmatic institutionalists like
Mitchell (1969, p. 596) rejected as “a schematic and superficial view of
human nature.” Thus, with respect to conflict and shirking,
a
more suitable
The result is what Nelson and Winter (1982,
p.
405)
term “the intellectual autarky
of
economics.”
332
/
SANFORD
M.
JACOBY
approach would endogenize workplace relations, while recognizing the rich
variety of those relations and not constraining them
to
a single essence
or
modality.
lo
Socialization.
In detaching the employment relationship from its social
and historical moorings, NEO-ILE opportunism models miss the fact that
socialization
is
a prerequisite
for
,
and alternative
to,
organizational control
structures (Parsons, 1949). First, there is primary socialization-in families
and school-which usually supports what employers do inside workplaces
and gives meaning to the incentives and penalties found there. Then there
is secondary socialization at the workplace itself, which relies on consensual
methods of inculcating norms and goals, such as ideologies of authority that
must be seen as legitimate if they are to be persuasive. Both types
of
socialization are required to make workplace controls effective; both also
offer alternatives to those controls.
For
example, studies show that primary
schools prepare young people for the disciplinary regimen they will encounter
at work. The
fit
between school and work is not perfect, however. Graduates
of professional programs attach high value to professional rewards that are
not given by the employer. At the same time, employees imbued with a
strong work ethic
or
a strong sense of corporate loyalty are less likely to
require supervision than those who have neither (Andrisani, 1978; Baldamus,
1961; Kohn, 1977; Ouchi, 1980). Hence, raising penalties for shirking is not
the only
or
even the most efficient way to economize on monitoring, just as
stiffer sentencing for criminals is not the only
or
most effective way to deter
crime and reduce the number of police.’’
In modern workplaces that do not rely on corporal force, controls are
usually a blend of the utilitarian (economic) and normative (social) types
(Etzioni, 1961). Yet NEO-ILE models focus exclusively on utilitarian
controls, perhaps because these are relatively less dependent on socialization
and can be more easily divorced from a social context. Problems then arise
when analysis turns
to
prescription,
as
when the NEO-ILE is offered as
practical human resource management,
lo
Given their disdain for “human relations” and Elton Mayo (cf. Kerr and Fisher,
1957),
the ILEs might seem a poor source from which to draw this critique; and, in fact, at times
the ILEs were more economistic and conflict-oriented than their forebears. Still, the behavioral
realism the
ILEs
inherited from the institutionalists led them to criticize Mayo
on
empirical
grounds and to regularly emphasize moral and social sides
of
human behavior, as, for
example, in analyses of union solidarity, cooperative bargaining, and business leadership.
I’
Hirschman
(1985)
offers an interesting explanation of why economists deal with
undesirable behaviors by proposing to raise their costs rather than relying on norms and
sanctions. Leonard’s
(1987)
empirical results question the predictions of efficiency wage
models that incorporate shirking.
The New Institutionalism
I
333
Compensation provides a case in point. First, the NEO-ILE literature
overemphasizes its efficacy and centrality to human resource management.
Recent studies show that the impact of financial incentives on productivity
is less than that of factors like employee involvement in goal setting (Guzzo
and Katzell, 1987). Second, rather than being purely a utilitarian control,
as it is portrayed in the NEO-ILE literature, organizational pay setting is
enmeshed in normative processes, a point the ILE tradition emphasized
(Marsden, 1986). Recent research on equity theory, for example, shows that
workers gauge pay differentials according to cultural and organizational
norms of fairness (Weick
et
al.,
1976; Carrel1 and Dittrich, 1978). When
wage dispersion exceeds these norms, effort is reduced and turnover increases
(Mowday, 1983; Pfeffer and Blake, 1988). Even the pay levels of corporate
executives are affected by social factors, such as the reinforcement of
hierarchical authority in large firms (Simon, 1957)
or
the maintenance
of
equity between CEO and board member salaries (O’Reilly
et al.,
1988).
Although some NEO-ILE pay models incorporate these factors
or
recognize
the existence of normative controls (e.g., Akerlof, 1982), most do not.
That is not unexpected, given that few economists are exposed during
their training to social and behavioral sciences other than economics.
l2
The
standard modeling metaphor learned by graduate students in economics
remains that of Robinson Crusoe, alone on his island without a community,
a history,
or
even an employer. The NEO-ILE has given
Crusoe
a boss but
not a great deal more. The imagery in efficiency-wage and principal-agent
shirking theories remains that of Crusoe. The metaphor is derived from the
classical model of contract, with its mutual suspicion, atomistic individualism,
weak ties, and pervasive fear of opportunism. That model is “infused with
the spirit of restraint and delimitation; open-ended obligations are alien to
its nature; arms-length negotiation is the keynote.” The classical
or
purposive
contract stands in sharp contrast to the status contract intended to govern
ongoing relationships such as marriages or long-term employment. These
contracts are laden with cultural specificity, full of “general and diffuse
commitments.. .[and] bonds and obligations” derived from social relations
and cultural norms (Selznick, 1969, pp.
5455).
While Williamson (1979)
and some of his followers have grasped the inappropriateness of using
purposive contract concepts to describe status relationships, other NEO-
ILE theorists have not. And even Williamson remains weak
on
a key point:
that both status and purposive contracts draw on prior social bonds, customs,
IL
Colander and Klamer’s
(1987)
survey
of
graduate students in economics found that only
13
per cent had intellectual interactions with students
or
scholars in other disciplines; well
more
than half thought that a thorough knowledge
of
the economy was unimportant for
professional success.
334
/
SANFORD
M.
JACOBY
and mores-what Durkheim
(1933,
p.
206)
called the “noncontractual
relations of contract”-without which any contract would be impossible to
negotiate or to enforce. Abstracting from these elements creates the risk of
seriously misunderstanding how markets, including labor markets, work
(Hirsch,
1976).
Even guile-an important feature of Williamson’s concept
of opportunism-presupposes the existence and subsequent violation of
contractual principles that have their origins in ongoing social relations and
social norms (Macneil,
198 1).
Trust.
Ongoing social relations build trust, another concept whose
importance has not been widely recognized in the NEO-ILE literature,
except by Williamson
(1975)
and his followers. Even for Williamson,
however, trust is an add-on rather than an integral part of the analysis
because his operating assumption-like that of most other NEO-ILE
theorists-remains that of pervasive opportunism. Yet, as industrial studies
have repeatedly shown, the presumption of innate opportunism is fatal to
trust (Gouldner,
1954;
March and Simon,
1958).
It leads to a proliferation
of control structures-supervision, rules, and deferred rewards-intended
to inhibit opportunism. These create resentment and distrust among
employees, who correctly perceive the controls as expressions of their
employer’s distrust. Expectations fulfill themselves when worker resentment
breeds opportunism and the employer is forced to implement additional
controls, now with the conviction that his initial beliefs were justified. Fox
(1974)
terms this the low trust syndrome: low trust begets low trust in an
ever-downward spiral of resentment, opportunism, and control. Imagine an
employment system embodying the low trust syndrome; imagine modern
Britain, says Fox. Alternatively, imagine vast segments of American industry,
filled with complex governance structures intended to attenuate opportunism
and distrust but paradoxically producing an opposite result. While there is
risk here of being a Pollyanna, NEO-ILE adherents rarely have considered
the risks inherent in their own approach.
Concededly, some economists have built trust models which do not assume
that opportunism is pervasive but instead treat its occurrence as a probability
determined by the cost of its consequences (Klein
et
al.,
1978).
For employers
and workers, these costs are said to be higher when growth is expected to
be rapid than when growth is stable or de~lining.’~ Thus, in high-growth
l3
Under rapid growth, the employer’s cost of opportunism is future increased wages for
an expanding work force; knowing this, employees are more likely to trust the employer not
to
take advantage
of
them. Opportunistic workers stand to lose the income and promotion
opportunities associated with high growth; knowing this, employers are more likely to trust
them.
The New Institutionalism
I
335
firms-like those
of
Silicon Valley or Japan-opportunism costs are high,
and with that comes higher trust, less rigid prices andlor wages, and fewer
explicit contracts. This is an interesting approach, although the model tries
too hard to rationalize trust in terms of an economic variable-growth
rates-without considering the possibility that what contributes to those
high growth rates is the prior existence
of
high levels of trust. As suggested
above, it is likely that trust reduces firms’ operating costs and raises their
growth rates. With trust, employers have less need for control devices and
cumbersome governance structures that are costly and can interfere with
their ability to adapt quickly to changing market circumstances. Similarly,
in the case
of
suppliers and customers, trust and goodwill can speed up
innovation, reduce the cost of restructuring, and obviate costly or cumbersome
contractual devices intended to lessen opportunism (Dore,
1987;
Perrow,
1986).
An understanding
of
the economy or of the employment relationship thus
requires that we analyze the social relations that underpin them. It is here
that economists can learn much from noneconomists about the production
of trust (and related conduct like cooperation and altruism) and about the
social conditions that promote those behaviors (Etzioni,
1988).
Consider the
possibility that the NEO-ILE assumption
of
rampant opportunism is an
unwitting time- and place-specific abstraction from modern American society
(e.g., see Titmuss,
1970;
Landau,
1984).
Therein lies the current fascination
of American managers with foreign social arrangements and workplace
practices. It is more than a little ironic that those managers are presently
seeking to promote trust and shrink bureaucratic controls (through decentra-
lization, employee involvement, and union cooperation), while at the same
time NEO-ILE theorists are devising models that rationalize in efficiency
terms such phenomena as metering, hierarchy, and unilateral management
(Alchian and Demsetz,
1972;
Jensen and Neckling,
1979;
Singh,
1985;
Williamson,
1980).
Behavioral realism.
The institutionalists’ skepticism of
homo economicus
(Veblen’s “homogenous globule of desire”) has been validated by recent
findings in cognitive psychology that question some of economic theory’s
behavioral precepts. Researchers have shown that expected utility maximiz-
ation models do not accurately forecast behavior and that economic decisions
are often only quasi-rational (Kahneman
et
al.,
1982;
Russell and Thaler,
1985).
People ignore high consequence events
of
low probability (e.g., few
purchase flood or earthquake insurance) unless they have previously been
exposed to them (Kunreuther
et
al.,
1978).
These findings raise serious
questions about the psychological adequacy
of
key NEO-ILE concepts such
336
/
SANFORDM.
JACOBY
as reputation effects in agency theory or the self-monitoring consequences
of efficiency wages. In each case, even if some minority of workers was
treated unfairly
or
caught malingering, an employee who has never been
mistreated or caught shirking (the vast majority) will think it unlikely that
the employer will treat
her
unfairly (hence an unfair employer’s reputation
can remain unsullied) or that the employer will detect
her
malfeasance and
dismiss her (hence shirking will continue). Thus, even with one’s premium-
wage job or reputation at stake, there is a threshold level (conceivably quite
high) below which those stakes have no deterrent effect.
The new psychology also stresses that search and choice under uncertainty
are not entirely rational but instead are guided by heuristic “frames” that
are likely to have social and cultural origins. These can explain various
behaviors of interest to labor economists, including bargaining, arbitration,
tipping, and money illusion. In the case of wages, money illusion is thought
to result from social norms regarding the “fairness” of real but not nominal
wage cuts (Bazerman,
1985;
Kahneman
et
al.,
1986;
Dawes and Thaler,
1988).
Again, the NEO-ILE’s mixed parentage is reflected in the fact that
some NEO-ILE researchers have incorporated these findings into their work
(e.g., Akerlof,
1980;
Kaufman,
1989),
although most remain insensible or
indifferent to them.
Conclusions:
Some
Thoughts
on
Methodology
As the papers in this symposium demonstrate, modern institutional labor
economics is complex. It includes those who carry on in the ILE tradition
and assert the necessity of middle-range theory and eclectic approaches in
labor economics, and others who believe that a neoclassical efficiency (rational
actor) approach can and should be extended to cover the field. In some
respects the tendencies are similar. Both are in contact with and borrow
from neoclassical theory, and both operate at a lower level of generality than
neoclassicism’s hard core. But they diverge on the issue of whether labor
economics’ distance from the core should be minimized or maintained. This
essay has argued that distance as well as contact between the two is inevitable,
necessary, and valuable. Although distance creates theoretical tension, it is
something that institutional labor economists, whether NEO
or
not, should
acknowledge and maintain, rather than rationalize or minimize.
In closing,
I
would like to offer some methodological precepts that can
be gleaned from the ILE tradition and from the analysis presented here.
These precepts-realism, empiricism, and eclecticism-are not arcane points
from the past. They reflect the approach of a sizable group of economists
at work today both within and outside of labor economics. Greater
The
New Institutionalism
I
337
consciousness of methodological issues may improve the quality of work
presently being done in labor economics and in economics more generally.
Realism.
Although Friedman’s
(1953)
dictum on realism of assumptions
was the methodological mother’s milk of many
U.S.
economists, philosophers
of science now reject his emphasis on prediction. Explanation, they say,
matters too. Friedman claimed that a theory’s assumptions do not have to
be realistic, but realism-what Freeman
(1988)
calls “informed priors”-
enhances a theory’s explanatory power. It allows us to know
why
a prediction
is valid and to choose among competing theories (Caldwell,
1982;
Hausman,
1989).
Realism is also necessary in dealing with phenomena that are fixed
in time and space, such as the institutional practices that are a focus of the
NEO-ILE; and, at a lower level of abstraction, realism is important in
analyzing and formulating policy, which includes the managerial actions and
decisions that are of concern to the NEO-ILE.
At
that level, cautioned
Knight
(1951,
p.
83),
“account must be taken of many details of situations
that fluctuate from case to case and that must be learned from empirical
observation at the scene of the action.”
Knight’s point cannot be overemphasized. For example, when analyzing
the labor market factors that determine competitive advantage among
advanced industrial nations, details of institutional structure are extremely
important. Evidence from comparative studies of firms in the same industry
using identical technology shows significant national productivity differences,
which are traced to particular and fine-grained institutional features: relations
between workers, employers, and unions; linkages between social and
industrial structures; and employment customs and practices (Dore,
1973;
Lincoln and Kalleberg,
1985;
Melman,
1958).
Indeed, advanced nations
that are successful at international trade are found to have “institutional
structures widely different from those prescribed by.. .textbook models
of
competition” (Soskice,
1989,
p.
2).
Empiricism.
Knowing the facts about labor markets and keeping theory
close to empirical reality are key parts of the ILE tradition that should be
preserved.
l4
Facts not only are needed for the verification of theoretical
l4
Economics today is more empirical than it was
50
years ago, when reliable economic
data were scarce and sparsely used by economists. Yet Wesley Mitchell’s jab at the armchair
theorizing of his colleagues (which he politely aimed at the classical tradition) still rings true.
Classical economics, said Mitchell, “erred sadly in trying to think out
a
deductive scheme
and then talked of verifying
that.
Until a science has gotten to the stage of elaborating the
details of an established body
of
theory, say
...
filling in
a
gap in the table
of
elements-it is
rash to suppose one can get an hypothesis which stands much chance of holding good except
from
a
process of verification, modification, fresh observation, and
so
on” (quoted in Gruchy,
1947,
p.
268).
338
/
SANFORDM.
JACOBY
predictions but are elemental substances from which new theories are
derived. Often the “stylized facts” that theorists work with are old or faulty.
Without fresh infusions of facts the frontiers of scientific discovery remain
frozen
or
mired in sophisticated explanations of the obvious. Too often
NEO-ILE (and other) economists rush into print theories based on partial
or incorrect factual knowledge; only later do the empiricists come to clean
up the mess.15
Admittedly, some of those problems stem from the unavailability of facts.
We need more data generated from a variety of sources and less reliance on
the same limited sets of government or panel data. More data could permit
testing of the myriad and contradictory theoretical claims in the NEO-ILE
literature (e.g., Are wage structures stable because of efficiency wages,
human capital, or noneconomic factors?
Do
age-earnings profiles tilt due to
human capital
or
to incentive effects?) But as Leontiefs
(1982)
study
suggests,
it
is more convenient and prestigious for economists (including
NEO-ILE economists) to “do theory” than to gather and study new facts.16
The ILE literature bulged with factual material-case studies, surveys,
industry studies, and labor market studies-but could have used more
theory. Today there is an opposite problem.
Eclecticism.
Economists are adept at arguing the irrelevance of other
disciplines
or
of evidence that does not
fit
optimization models. They can
always claim that market competition
or
learning effects will eliminate any
noneconomic intrusion
or
any short- to medium-term departure from strict
rationality. The strength of orthodox theory-and also its weakness-is its
ability to rationalize or dismiss anomalous findings and facts.
l7
Economists
do
have
a
point when they argue that other social sciences could benefit by
paying more attention to the analytic methods and theoretical precision
found in economics. But cross-disciplinary enlightenment flows in both
directions. This is particularly true of topics that fall under the NEO-ILE
rubric (e.g., work motivation, organizational structure, unionism), which
Is
For example, some able theorists (e.g., Hall and Lillien, 1979; Aoki, 1984) have
repeatedly confused the causal relationship between the NLRB’s contract bar rule and union
contract durations. Since the Second World War, the former has determined the latter, not
the reverse (Jacoby, 1987).
j6
Leontief found that two-thirds of the articles published in the American Economic
Review between 1977 and 1981 contained no data, a share that makes economics unique
among the natural and social sciences (Morgan, 1988).
As
demonstrated
by
a variety of critics (e.g., Nelson and Winter, 1982; Thurow, 1983),
these rationalizations depend on a host of subsidiary and questionable assumptions: that
markets are competitive and in equilibrium, that strong selection pressures weed out inefficient
(including noneconomic) factors, and that long-run equilibrating processes are immune from
shocks that would prevent an equilibrium position from being reached.
The
New
Institutionalism
I
339
have received considerable attention from noneconomists. Incorporating
perspectives from other disciplines would make NEO-ILE research more
powerful by widening its range of evidence.
At
the same time, it would
make NEO-ILE research more attractive and plausible to noneconomists.
Akerlof
(
1984) compares the unwillingness of neoclassical economists to
derive assumptions from other disciplines to French chefs whose unwritten
rules forbid the use of certain ingredients like raw fish or catsup. Below,
I
offer a sampling of ingredients from psychology, history, and sociology that
would make for a more well-rounded, tasty, and nutritious intellectual
cuisine. Of course, the chef might object that these additives have no place
in his or her kitchen, but there is a long tradition in economics (including
the ILE approach) that rejects this as excessive parsimony (Hirschman,
1985).
Cognitive and social psychology offer a rich body of empirical research
that questions the strict rationality assumptions of the orthodox model.
Previous efforts to go beyond that model-with concepts like Simon’s (1957)
bounded rationality and Leibenstein’s (1976) selective rationality-never led
economists very far into the domain of psychological research. Recently,
however, there has been a welling up of research at the boundaries between
the disciplines, although only a bit
of
it has trickled back into labor
economics. Future work remains to be done that would ground rationality
assumptions in empirical research on individual decision-making
,
the purpose
of seemingly nonrational factors in utility functions, and the processes
by
which those functions derive their terms.
At
the very least, psychologists
could teach NEO-ILE economists to make their behavioral assumptions
explicit, especially any unconscious or unarticulated views about human
nature and work behavior.
As labor economists return to the study of institutions, historical analysis
should bulk larger in their research. Unlike markets, institutions and
organizations have specific origins and histories; indeed, those origins often
determine the course
of
institutional development (David, 1985). Historical
and comparative analyses also reveal what is and is not unique about
particular institutions and thus deter the use of inappropriate analytical
generalizations. As one economic historian recently asked, how many
economists know specifically what is modern or American about “modern
American labor markets”? (Wright, 1987). Finally, more attention to
historical methods and reasoning can help steer economists away from the
kind of teleology that occasionally surfaces in the NEO-ILE, as when history
is portrayed as the successive development of more efficient institutional
forms.
Sociology has long been avoided by neoclassical economists, in part because
340
/
SANFORDM.
JACOBY
holistic concepts like society contradict the normative and methodological
individualism that lies at neoclassicism’s core. But a variety of sociological
concepts nevertheless are implicit in the
ceteris
paribus
conditions that link
the core to the real world, such as the set of social institutions that permit
market exchanges to occur (everything from contract laws to the police to
trust). This is what economic sociologists mean when they describe the
economy as “embedded” in structures of social relations and social norms
(Granovetter, 1985; Lowe, 1935).
As
shown in this essay, when economists
abstract from these social terms or reduce them to maximizing behavior,
they create economic man, who, says Sen (1982, p. 99), “is close
to
being
a social moron.” Social moronism might be a good approximation to reality
in some instances, but it is a misleading oversimplification when applied to
labor market behavior. More
is
involved in taking account of these social
terms than simply inserting them into an individual’s utility function.
This solution can cause analytic problems, especially if preferences are
interdependent and require individuals to act against immediate self-interest
in the pursuit of social goals (Elster, 1982; Sen, 1985).
It
is
one thing to advocate for the NEO-ILE a greater reliance on eclectic,
middle-level theories grounded in empirical reality. It is an entirely different
matter to achieve it. Joining synchronic and diachronic analysis is not easy;
nor is it easy to bridge theory and reality, economics and other disciplines,
or what
R.
A.
Gordon (1976) called “rigor and relevance.” Striking the right
intellectual balance depends on the topic at hand, on the uses of
our
knowledge, on judgment, and on an open mind. In understanding these
matters, the ILE tradition-whatever its defects-is an old dog that still has
tricks to teach
us
all.
... In this article, on the basis of the Marxist-Leninist view and Ho Chi Minh's thought, inherits the interdisciplinary approach to labor relations of domestic and foreign scientists, of the authors: Jacoby (Jacoby, 1990), Kaufman (Kaufman, 1993), Godard (Godard, 1994), Kimmel (Kimmel, 2000). Especially in The History of Industrial Relations as a Field of Study, Carola Frege (2007) the author inherits many parts of the history of industrial relations. ...
... After the end of World War II (in 1945), industrial relations became increasingly popular at many American universities (Carola Frege, 2007: 36). Many US industrial relations researchers have approached multiple disciplines, such as: Jacoby (Jacoby, 1990), Kaufman (Kaufman, 1993), Godard (Godard, 1994), Kimmel (Kimmel, 2000). In which, Jacoby and Kaufman are institutional economists who have studied industrial relations in an interdisciplinary approach, such as behavioral science, pragmatic science, and public policy-oriented studies, labor market institutions. ...
... The theory of endogenous institutions (e.g. Jacoby, 1990) suggests that routines in organizations (for instance, related to age management) are the result of institutionalized leadership practices. ...
Article
Purpose This work was undertaken in order to develop a conceptual model for identification of the capability of an organization to implement age management measures. Design/methodology/approach Barriers to delayed retirement were reviewed; observations retrieved from a research consortium study were used to identify main attributes that needed to be taken into consideration in the development of the model. Findings The capability of organizations to react to the demographic challenge by introduction of age management measures can be classified operatively as “proactive”, “reactive”, “passive” or “chained”, depending on their resources and preferences. Practical implications The model may be useful to HR as a point of departure in the development of a business case for age management and a didactic tool to be used in internal marketing. Originality/value The concept “organizational capability” was developed as a corollary to the individual aspects of capability; recognizing preference and resource as main drivers made it possible to develop a typology that is new and is easy to understand and apply.
... Institutions derived from an organisation's environment are referred to as exogenous institutions (Scott 2008). Here, the term endogenous institutions (Jacoby 1990) are used for these intraorganisational institutions. Theories about endogenous institutions can help us understand the dynamics between the taken-for-granted rules and actors' behaviour when organisational actors challenge difficult-to-change practices and norms (Kaghan and Lounsbury 2011). ...
Article
This study examines the support, control and strategic functions of an executive hospital political board from the perspective of how institutionalised internal procedures influence board functions. Board members and leading hospital managers were interviewed about politicians’ ways of working, managers’ preparation of decision issues and board work under the presidency model. The politicians strengthened the decision-making processes by requesting additional supporting documents, and by defending the hospital from public criticism. However, the board exercises weak strategic and control functions; it absorbs the responsibilities of the hospital managers and the upper political levels as it reaffirms its executive authority. Nonetheless, the study suggests that abolishing or replacing the board with a non-political board may not improve organisational governance. The study gives insights in politicians’ contribution to strategic public management, an aspect lacking in existing literature.
Article
Full-text available
Are sources of economic rent, as defined by the prevalent business strategy paradigm, sufficient to attain and maintain superior returns? The perspective developed within the conceptual framework of the Institutional Theory may offer managers a contribution towards understanding the strategy process and its potentialities, particularly by stressing the leading role played by legitimacy, the influence of many institutional spheres, the isomorphic pressures, ceremonial behavior and decoupling, among other elements, that mainstream business strategy fails to address directly, but which may have a significant effect on firm performance. We advance that these elements must be accounted for in the pursuit and acquisition of economic rents, even if the ability to articulate them purposefully is constrained by rationality, agency conditions and the manager's social embeddedness.
Article
This article discusses the historical trajectories of employment research in the U.S. and Germany. Both countries exemplify very different research traditions, which have their origins in the 19th and early 20th century and are shaped not only by the field of inquiry (different national employment institutions and practices), but also by different social scientific traditions and political philosophies, for example the different interpretation of what constitutes industrial democracy or employees' rights at work. I conclude that current research still differs across countries and that we would do well to endorse this plurality rather than assuming an inevitable convergence to US research styles.
Book
The meso-level realm of social reality is structured by corporate and categoric units, along with their respective cultures. Unlike the macro and micro realms of social reality, the meso-level does not reveal its own unique forces. Rather, the dynamics of meso-structures and cultures are driven by macro- and micro-level forces pushing on individual and collective actors as they build corporate units and develop parameters defining membership in particular social categories. © Springer Science+Business Media, LLC 2012. All rights reserved.
Article
This work offers a deeper evaluation of entrepreneurship through institutional theory understanding, in which the technical-economic sphere is not only complemented by a social perspective, but also structured and governed in an ethical manner. Based on a hermeneutic approach, it argues there are primarily three distinct logics or rationalities (technical-economic, psycho-social and ethical anthropological) which are nonetheless compatible and united in one sole agent and in one unique human action. This proposal is an essential complement to institutional theory findings related to entrepreneurship, allowing a sound dialog between technical, social and ethical aspects of entrepreneurship where profit is not excluded, but, rather, considered as a means for achieving human and social ends—thereby radically eliminating the purely instrumental perspective of the human being. The review of the category of entrepreneurship, based on institutional theories, evidences the need to incorporate anthropology and ethics which is explicitly done through an open dialogue that contemplates and integrates all of them.
Article
Full-text available
The notion of capital has enjoyed a rich career in the social sciences, its use across a range of subjects and in diverse academic and professional contexts having served to establish its conceptual status as 'given'. With particular attention to human and social capital - including cultural capital - this book traces the roots of this theoretical and conceptual trend to economics, revealing the proliferation of various forms of capital to be based upon an encroachment of the conceptual apparatus of economics into other social sciences. Offering an in-depth, critical analysis of the concepts of human and social capital, as well as their surrounding theories, Anti-Capital: Human, Social and Cultural proposes an alternative theoretical framework, whilst better explaining the realities that they mask in economic terms. A rigorous exploration of the most popular forms of 'capital' in the contemporary social sciences, this book will be of interest to scholars and students of sociology, political and social theory, demography and economics.
ResearchGate has not been able to resolve any references for this publication.