ArticlePDF Available

Sanford Jacoby and Daniel J.B. Mitchell, "Origins of the Union Contract: Development of Contractual Features of the Union-Management Relationship," Labor Law Journal, August 1982.

the Union Contract
Development of Contractual Features
Of the Union-Management Relationship
University of California, Los Angeles
NEGLECTED TOPIC in labor history has heen the day-to-day
workings of lahor-management relations.' Much has been made
of broad conflicts between labor and management, such as the open-
shop movement of the 1920s and the CIO organizing campaigns of
the 1930s. But little attention is paid to the kinds of contractual features
that were negotiated. The contents of a contract signed after a long
struggle for recognition lack the drama of the struggle
since obtaining a contract was the ostensible goal of most such struggles,
lack of interest in the final product cannot be justified.
Recent developments in economic theory suggest that a linkage
between labor history and modern labor economics could be established.
Wage determination has long been a weak area in macro- and micro-
economic theory. Although there is division between Keynesians and
monetarists on various issues, wage "stickiness" plays an important
part in the explanation of unemployment in both approaches.
The macro puzzles have led to the development of new micro theories
of implicit contracting in the labor market. Various versions of these
theories exist. A common thread is that all seek to find foundations for
wage stickiness and other "anomalies" in the labor market (labor hoarding,
procyclical productivity, career ladders, and other internal labor market
arrangements) in terms of the traditional objectives of profit and utility
maximization. So far the relationship between implicit contracts (basically
nonunion) and explicit contracts (union) has not been widely explored.
We suggest that a good place to start in the study of labor-market
contracting is the development of explicit union agreements. Union
contract practices are more readily documented than implicit nonunion
arrangements, precisely because they are formalized and written. In
to space limitations, only limited citations are provided. Complete refer-
ences are available from the authors.
August, 1982 Labor Law Journal
what follows, differences between pre-
World War II and postwar contracting
are discussed. It is asked whether the
changes during that interval can be
ascribed to "rational" incentives of the
type featured in the implicit contracting
literature or whether events related to
the war are the primary explanation.
Finally, some implications for public
policy are discussed.
Although references can be found to
"contracts" between employers and
unions prior to the Civil War, collec-
tive bargaining agreements are gen-
erally a post-Civil War phenomenon
and became increasingly popular dur-
ing the late 19th century. By fixing
wages and working conditions for a
definite period, each party assured
the other that it would not take ad-
vantage of seasonal or cyclical swings
in production to secure more favor-
able rates than those negotiated.
Unionized workers were guaranteed
a steady wage and the employer could
guarantee the product prices adver-
tised to customers. Most importantly,
by minimizing the risk of a strike, the
employer avoided a costly disruption
of production and possible loss of market
share. The union's no-strike promise
was the quo it offered for the employer's
Collective labor contracts were not
legally enforceable in most states until
after World War II. To secure the
employer's good faith, the union had
to make good on its promise not to
strike during the term of the agree-
ment. This was accomplished by the
strict controls that national unions placed
on strikes by their locals, especially the
refusal to provide financial support to
an unsanctioned strike. Thus, the spread
of labor contracts after 1880 owed much
to the development of a national union
capable of enforcing its agreements.
The prevailing wisdom is that modern
collective bargaining provisions are
traceable to the World War II and
postwar period. To explore this as-
sumption, we have collected a large
number of contracts negotiated before
the establishment of the National War
Labor Board (NWLB).2 The table,
based on those contracts that have so
far been coded, shows that many "mod-
ern" provisions in fact existed in the
prewar period. These include wage
reopeners, deferred wage adjustments,
union security provisions (including
maintenance of membership), grievance
aribitration, and benefits such as health,
welfare, and pension provisions. It is
clear that the parties did not require
the civilizing influence of the NWLB
to invent these features. An important
question is why these innovations devel-
oped. Five hypotheses can be suggested.
Hypothesis (1): an agreement of ex-
tended duration is a logical consequence
of the employer's desire to minimize
negotiation and strike costs over time.
a lengthy contract allows the
employer more accurately to predict
his labor costs, which facilitates plan-
ning and satisfactory customer rela-
But three factors may make
employers reluctant to sign two- or
three-year agreements.
First, employers will not sign these
agreements unless they are convinced
of the union's integrity with regard
to its no-strike promise. Second, em-
ployers will be unwilling to commit
themselves to an extended agreement
° The contract file consists of nearly 600
pre-NWLB agreements gathered from library
and other collections around the country. Con-
tracts from the 1935-1942 period represent
two-thirds of the sample. The sample is rough-
ly consistent with the industrial distribution
IRRA Spring Meeting
of trade union members over time. However,
contracts from the apparel and printing indus-
tries are overrepresented, while contracts
from the construction industry are under-
represented, in the sample.
until they have accepted the union as
a permanent feature. In the early years
of a union-management relationship,
an employer might seek a short agree-
ment in the belief that the union no
longer will be around when the contract
expires. As the relationship matures,
the employer might switch from a short-
run strategy of eliminating the union
to a long-run strategy that minimizes
the cost of what is now perceived as
a quasi-permanent relationship. Third,
an extended agreement only is possible
after the parties have worked out the
rules that will govern their day-to-day
relations. Negotiations then can occur
less frequently. These factors suggest
that long-duration contracts will be
found where the parties have been bar-
gaining with each other for a number
of years.
Hypothesis (2)
contractual union-
security provisions also may be related
to the age of the bargaining relation-
An employer only will grant rec-
ognition to a union which has been
accepted as a fixed feature of the en-
vironment. Alternatively, a union secu-
rity clause may be a quid pro quo for
union acceptance of an extended agree-
ment. Union leaders would rather not
renounce strikes for an extended period
because the strike is the ultimate weapon
for achieving union objectives. They
may fear that, if the right to strike is
limited, members no longer will perceive
the union as a militant organization.
A union security clause can assuage
these fears.
Hypothesis (3): once an agreement of
extended duration has been signed,
mechanisms may be needed to adjust
wages for future changes in economic
conditions. A wage reopener clause
provides some protection against un-
events; a deferred wage ad-
justment clause allows for a flexible
response to
future conditions.
Use of these clauses should be posi-
tively related to the agreement duration.
Hypothesis (4): the use of third-party
neutrals to arbitrate rights disputes may
he a learning phenomenon unrelated to
contract duration. Most agreements
provide some mechanism for resolving
intracontractual disputes. Yet neither
party will be eager to grant outsiders
the power to resolve these disputes
until they have ibecome convinced that
outsiders can adjudicate effectively.
Arbitration will become more prevalent
as the parties learn that it has worked
well in other industries. Also, arbitra-
tion will become more predictable and
acceptable as the number of experienced
arbitrators and a body of arbitral norms
increase over time.
Hypothesis (5): in the absence of ex-
ternal incentives (such as those now
provided by the tax code) there is no
reason for workers to prefer that em-
ployers provide them with benefits "in
kind" which they could otherwise pur-
chase. Some unions, drawing on the
earlier tradition of beneficial societies,
might work out arrangements with
employers whereby benefits such as
life insurance would be provided. Or
they might seek to provide benefits
not available externally (such as un-
employment pay before 1935). However,
some employers might provide such
benefits for "paternalistic" reasons or
to reduce turnover among senior workers
outside the union contract framework.
Supporting Evidence
As expected, the agreement duration
was positively related to the age of the
bargaining relationship. The longest
agreements, including many of two- or
three-years' duration, were found in
industries where there was a long his-
tory of contracting, such as printing,
apparel, mining, and construction. The
proportion of agreements of less than
thirteen months' duration was highest
among those signed during the pre-
1910 and 1935-1942 periods. These were
August, 1982 Labor Law Journal
(percentage of contracts)
Union security clause
Union shop
Maintenance of membership
Closed shop
Deferred wage adjustment
Reopener clause
Conditional on inflation
Escalator clause
welfare benefits
Pension plan
1-12 months
13-24 months
25-36 months
Over Z7 months
Indefinite or open
" Contracts after February 1942 were excluded.
" 1948: "Survey of Contracts Under Taft-Hartley Act," Labor Relations Reference Matmal,
Vol. 22 (Washington, D. C: Bureau of National Affairs, 1948); "Labor-Management Con-
tract Provisions, 19S0-S1," Bureau of Labor Statistics Bull. No. 1091 (Washington, D. C:
U. S. Government Printing Office, 1952), p. 19; 1954: "Collective Bargaining Agreements:
Expiration, Reopening and Wage Adjustment Provisions of Major Agreements," BLS Report
75 (Washington, D.C.: U.S. Government Printing Office, 1954), p. 3.
periods when bargaining was first being As predicted, a positive and significant
introduced to various industries. relation was found between the presence
Union-security clauses were least °^ ^'^her a reopener or deferred wage
prevalent during these two periods; 25 Provision and contractual duration. Both
percent of pre-1910 and 48 percent of Provisions rarely were found in the
post-1935 agreements contained union ^^"^^ contract, which indicates that
security clauses as opposed to 69 per- ^^^y ^ere alternative mechanisms for
cent during the intervening years. This dealing with future circumstances. It
finding supports the hypothesis of a
''^^ "^^^d that escalator clauses
relation between age of the bargaining ^^''^ ^^^^ but can be found in a few
relationship and presence of a union- prewar contracts,
security clause. But there is also a The provision that a neutral arbi-
positive and significant relationship be- trator be used to settle rights disputes
tween presence of a union-security clause had a positive and significant relation
and contract duration. It appears that to the year in which a contract was
such clauses were an employer con- signed. This supports, as does the
cession to obtain union acceptance of table, the hypothesis that grievance
a longer contract. arbitration grew steadily over time, ir-
IRRA Spring Meeting 515
respective of a contract's duration or
the age of the bargaining relationship.
Employer-provided "fringe" benefits
date back to the late 19th century but
were not found with any frequency
until after World War II. The bene-
fits provided were meager; in 1929 they
averaged only two percent of total
annual payrolls. These benefits almost
never were part of a collective agree-
ment, except for the unemployment
insurance plans found in clothing con-
tracts during the 1920s. Each party
preferred unilaterally to provide fringes
rather than negotiate them. The "tech-
nology" of supplementary benefits was
well-known but not widely implemented
until the 1940s.
the general hypothesis that
"rational" factors determined the con-
tent of prewar agreements is supported.
Most features of the modern agreement
can be found in older contracts. Bar-
gaining relationships established after
1935 were likely to have matured into
something closely akin to the modern
contract even without external interven-
tion. While older agreements rarely were
as detailed as current contracts, they
provided a wealth of experience and
innovations upon which new bargain-
ing parties could draw.
The Way It Became
The NWLB traditionally has been
viewed as a key institution that created
the framework for postwar bargaining.
The Board reputedly was a major im-
petus for the adoption of contractual
fringe benefits, grievance arbitration,
and union security (maintenance of
membership) clauses. While the NWLB
was a contributing influence to the
modem contract, we think that its ulti-
mate importance has been overstated.
The NWLB's extensive wage sta-
bilization program sought to control in-
flation by establishing limits for various
types of wage adjustments. Approval
of the NWLB was not required when
the parties negotiated "reasonable"
group health, welfare, and pension plans.
This policy decision was an obvious
stimulus to the growth of these plans
and some have claimed that it "prob-
ably more than any other single event
triggered the dramatic expansion of
employee benefits... ."*
However, the number of trade union
members covered by these plans did
not begin to approach modern levels
until well after the war. In part this
was due to the Supreme Court's 1949
Inland Steel* decision, which brought
such benefits into the scope of bargain-
ing. But a learning phenomenon also
is suggested by the data. The number
of union members covered by some type
of negotiated plan more than doubled
from the close of the war to early 1947,
when about 1.5 million members were
covered. The number doubled again
to three million in mid-1948. By mid-
coverage had grown to 7.6 mil-
lion and reached 11.3 million in 1954.
It is impossible to know what these
figures would have looked like had it
not been for the NWLB. Yet we think
that much the same growth pattern
would have been observed once changes
in tax incentives are considered.
The role of the NWLB in establish-
ing modern grievance arbitration is
well-known. The Board "insisted, as a
matter of paramount importance, upon
arbitration as the final step of grievance
procedure."" Its directives provided
a framework for the handling and ar-
bitration of employee grievances. But
it would, perhaps, be more accurate
' Robert M. McCaffery, "Employee Bene-
fits : Beyond the Fringe
Personnel Ad-
ministrator (May 1981), p. 26.
* (US SCt, 1949), 336 US 960.
* The Teniwiatioii Report of the National
War Labor
Vol. 1 (Washitigton, D.C:
U. S. Government Printing Office, 1947), p.
August, 1982 Labor Law Journal
to say that it was the wartime no-strike
agreement which forced the parties to
rely more heavily on arbitration, a pro-
cedure that had been used in a variety
of industries—coal, apparel, printing,
and construction—since the 1910s.
Admittedly, many older contracts pro-
vided for arbitration only if a joint
committee could not reach a decision.
Often arbitration was used only with
discretion. But the outlines of modern
grievance arbitration were clear long
before the war. The successful use of
arbitration in older bargaining rela-
tionships influenced the newer, post-
Wagner Act relationships.
By 1940, both of the major contracts
in the auto and steel industries provided
for a multistep grievance procedure
ending in arbitration. While there was
reluctance in these new industries to
entrusting contract interpretation to
inexperienced outsiders, the pressure
to avoid strikes during the war eroded
this resistance. Also, the heavy war-
time use of arbitration created an army
of experienced labor arbitrators. One
of the few issues that participants in
the 1945 Labor-Management Conference
could agree upon was the value of griev-
ance arbitration. The NWLB provided
additional pressure and guidance without
which arbitration might have grown
more slowly. But the prewar history
of arbitration suggests that a learning
phenomenon also was present.
It is sometimes said that mainte-
nance-of-membership (MOM) clauses
were "devised by the public members
of the NWLB."8 It is true that the
Board was responsible for the wider
use of these clauses, but it hardly in-
vented or "devised" this form of -union
Variants of these clauses can be found
in a few pre-1920 agreements that com-
bined them with a closed shop provi-
sion. During the 1920s the clauses
appeared in some apparel and street
railway contracts; they were a stan-
dard provision in some meatpacking,
chemical, and paj>er industry contracts
during the 1930s (see the table). In
fact, it was the NWLB's short-lived
predecessor, the National Defense Medi-
ation Board, that first ordered the in-
sertion of these clauses into disputed
contracts. Here as elsewhere, the war-
time authorities drew heavily on pre-
war contractual innovations.
The importance of the NWLB's re-
peated orders to establish MOM pro-
visions should not be underestimated.
Fifty-eight percent of the NWLB cases
in which union security was an issue
involved the new CIO unions, whereas
only 34 percent involved the AFL. Many
of the new unions did not have what
they viewed as adequate union security
at the start oi the war and the NWLB
helped overcome employer resistance
on this issue. A counterfactual ques-
tion may be posed here
had it not been
for the NWLB, how long vrould it have
taken the new CIO unions to obtain
union security? We do not have a
definite answer. But it is possible that
employers eventually would have be-
come resigned to the new unions as a
fait accompli. Union security would
have been granted or made a trade-
off for multiyear contracts.
In some ways, the wage controls of
World War II may have retarded cer-
tain modern contractual features. Con-
trols discourage multiyear contracts.
Hence, they discourage the parapher-
nalia of such contracts such as wage
reopeners, deferred wage adjustments,
and escalators. As noted, escalator
clauses were known before World War
But the widespread use of such
° Harold S. Roberts, Roberts' Dictionary
of Industrial Relations (Washington, D. C.:
Bureau of National Affairs, 1966), p. 233.
IRRA Spring Meeting517
clauses began with the 1948 GM-UAW
two-year agreement. Such agreements
might have come earlier in the absence
of wartime disincentives to long-dura-
tion contracts.
There have been recent suggestions
that union contracting makes it dif-
ficult for monetary policy to reduce
inflation. The argument is that long-
term contracts "lock in" prior inflation
and inflation expectations. Our analysis
suggests that, even if long-term con-
tracting has dysfunctional macroeco-
nomic effects, its forms stem from
micro-level incentives, mainly the avoid-
ance of dispute costs. Thus the banning
of such contracts—a proposal that is
sometimes heard—would be strongly
resisted by the parties.
If there is a macro-level need to build
wage sensitivity to demand pressures
into contracts, however, it might be
possible to change the micro incentives
by offering tax benefits to "gain-shar-
ing" plans. Such plans include every-
thing from conventional profit-sharing
arrangements to Japanese-style profit-
related 'bonuses. Gain-sharing would
add the desired wage sensitivity to long-
term contracts.
While our reading of the NWLB
period does not suggest an ability o.f
government agencies to impose features
for which no need is felt by the parties,
such agencies can "educate" the parties
in the adoption of devices for which
incentives exist. There is little doubt
the NWLB assisted in speeding the
spread of grievance arbitration. If in-
centives were created for innovations
like gain-sharing, it might be possible for
existing agencies (such as the FMCS)
to play such an educational role.
Finally, the recent rash of wage con-
cessions has raised the issue of whether
a "turning point" has been reached in
industrial relations. If "turning point"
simply means lower wage settlements
in the face of mass layoffs, obviously
such a point has been reached. But, if
"turning point" is taken to mean a
new way of contracting, involving a
permanent end to long-term agreements
with their escalator clauses and deferred
wage adjustments, our analysis sug-
gests that such a turn is very unlikely.
That is because we are evolutionists
rather than creationists in our view
of how such devices developed in the
first place. [The End]
Seniority Provisions
Collective Bargaining*
Rhode Island
ENIORITY RULES create hier-
archies of precedence based on length
of service among employees in a spec-
ified seniority district. Seniority is a
kind of institutional age according to
which "older" workers are entitled to
preference over "younger" ones in pro-
and other conditions
of employment. Negotiation of such
* This inquiry
by a
grant from
American Philosophical
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.