Unnatural Extinction: The Rise and Fall of the
Independent Local Union
Unnatural ExtinctionSANFORDM. JACOBY
SANFORD M. JACOBY*
This article analyzes what happened to independent local unions (ILUs), also
known as company unions, since 1935. After providing a statistical analysis of
ILU membership since 1935, the article looks at the factors that shaped member-
ship trends: changes in labor law, the characteristics of ILUs, worker attitudes
toward ILUs, and employers’ industrial relations policies. New evidence is pre-
sented that suggests that even those employers who still favored ILUs in the
1950s were orienting them away from collective bargaining and toward the
“new nonunion model” of the 1960s and 1970s.
INDEPENDENT LOCAL UNIONS ARE SINGLE-EMPLOYER UNAFFILIATED
UNIONS that represent employees of a plant or company. Most of the inde-
pendent local unions in existence in the 1940s started out as employee
representation plans or “company unions,” which, as Daniel Nelson
(1982) has shown, were popular in the 1920s (when they were associated
with progressive welfare capitalism) and once again in the mid-1930s
(when employers rapidly established them in the wake of the 1933
National Industrial Recovery Act). Few of these company unions sur-
vived, however. Either they became part of national unions affiliated with
the American Federation of Labor (AFL) and Congress of Industrial
Organizations (CIO) or they were disestablished by the National Labor
Relations Board (NLRB), which in its early years implemented the
*Management, History, and Policy Studies, University of California, Los Angeles. E-mail:
email@example.com. An earlier version of this article was presented at the Conference on Non-
union Forms of Employee Representation sponsored by Georgia State University and the University of
Calgary, September 1997, Banff, Alberta. Thanks to Bruce Kaufman, Daphne Taras, John Fossum, and
three anonymous reviewers for helpful comments.
INDUSTRIAL RELATIONS, Vol. 40, No. 3 (July 2001). © 2001 Regents of the University of California
Published by Blackwell Publishers, 350 Main Street, Malden, MA 02148, USA, and 108 Cowley Road,
Oxford, OX4 1JF, UK.
Wagner Act’s company union ban with Draconian stringency. Precisely
how many company unions disappeared in the period between 1935 and
1940 has never been calculated precisely; this article provides an
The company unions that survived gradually evolved into relatively
autonomous organizations that were financially independent of the
employer; these became known as independent local unions (ILUs). The
ability of these former company unions to withstand both NLRB scrutiny
and raiding attempts by affiliated (national) unions encouraged new
groups of employers and workers to establish ILUs during the war. Later,
when the Taft-Hartley Act of 1947 signaled that the NLRB would take a
more lenient approach, there was an increase in the rate of ILU formation,
a spurt that lasted into the 1950s. A number of ILUs became feisty organi-
zations during those years, successfully fending off raids by affiliated
unions and even engaging in some successful raids of their own. ILUs
could be democratic organizations, with internal debates and contested
elections. They appealed to white-collar workers and to employees in
smaller firms outside the big cities. Because these constituted the main
sectors where unionism did not reach in the 1950s, observers forecast a
rosy future for ILUs: that they would “more than hold their own” (Troy
In fact, however, after the late 1950s, ILUs went into a decline from
which they never recovered. Few new ILUs appeared, NLRB
disestablishments persisted, and affiliated unions became more adept at
raiding ILU members. While ILUs never completely disappeared, by the
1970s they had become a relatively insignificant feature of American
After providing a statistical analysis of ILU membership between 1935
and 1967, this article takes a closer look at the factors that shaped mem-
bership trends: changes in labor law, the characteristics of ILUs, worker
attitudes toward ILUs, and employers’ industrial relations policies. On
this last point, new evidence is presented that suggests that even those
employers who still favored ILUs in the 1950s were orienting them away
from collective bargaining and toward something resembling the “new
nonunion model” of the 1960s and 1970s. Whether this same reorienta-
tion would have occurred if Congress and the NLRB had been more toler-
ant of ILUs remains an open question. Whatever the answer, today
Congress and the NLRB have an opportunity to moderate their
long-standing hostility to ILUs and other employer-assisted organizations
by providing a space for experimentation with representational forms.
378 / SANFORD M. JACOBY
Company and Independent Union Trends, 1935–1967
According to a survey done by the Twentieth Century Fund
(1935:79–80), there were about 2.5 million workers represented by com-
pany unions in early 1935, shortly before passage of the Wagner Act. If
the average size of a company union was the same in 1935 as it had been
in 1932 (i.e., 1645 workers per union), then there were about 1515 com-
pany unions in 1935 (Millis and Montgomery 1945:835). Many of these
company unions disappeared after passage of the Wagner Act, which
made it an unfair labor practice for an employer to dominate, interfere
with, or provide financial assistance to a labor organization.1
Prior to passage of the Wagner Act, the Roosevelt administration and
its friends were divided on the question of company unions. On the one
hand, NRA administrators like Hugh Johnson and Donald Richberg
wanted to preserve company unions because they thought such unions
could coexist with national unions (as was the case with some shop com-
mittees established during World War I) and because they believed that
company unions provided representation to workers whom the labor
movement had been unable or unwilling to organize. On the other hand,
the AFL took the position that company unions were not a complement to
but purely a substitute for national unions and stood in the way of inde-
pendent organizing activity (Dubofsky 1994).
A third position was taken by Senator Wagner, who was sympathetic to
the cooperative and self-governing ideals of some shop committees and
representation plans established prior to the 1930s (Barenberg 1993). Yet
Wagner thought that the company unions hastily established by employ-
ers in the wake of the National Industrial Recovery Act had few of these
redeeming features. And he shared with other underconsumptionists the
view that only massive wage increases could end the Depression and that
company unions never could raise wages as effectively as national unions
(Jacoby 1993; Kaufman 1996; Mitchell 1986). The senator’s position led
to the compromise contained in the Wagner Act: The act did not ban com-
pany unions, but it could be interpreted to mean that any company union
that had been created with employer support—nearly all of them—was
presumptively tainted and liable for disestablishment. In fact, this was
Unnatural Extinction / 379
1The Wagner Act also specified that a labor organization was one that “dealt” rather than “bargained”
with the employer. Had Congress adopted the latter terminology, as proposed by the Secretary of Labor in
1935, it would have legitimized employer-supported organizations that did not engage in bargaining. The
result might have been something like the two-tiered industrial relations systems presently found in
Europe, where works councils coexist with industrial unions and industrial bargaining (Millis and Brown
1950; Cabot Carbon 1959).
how the early NLRB interpreted the act, both because it observed numer-
ous instances in which company unions thwarted fair-representation elec-
tions and because the political proclivities of the board’s early members
caused them to harbor a suspicion not only of company unions but of
many craft unions as well (Gross 1981).
Initially, many employers paid little attention to the Wagner Act’s Sec-
tion 8(2), fully expecting that the Supreme Court would find the act
unconstitutional. In 1937, however, the Supreme Court upheld the act.
Now the NLRB picked up the pace of disestablishments and applied a
harsh rule: A disestablished company union was not permitted to appear
on an election ballot even if the employer withdrew his support. To the
annoyance of company-union supporters, the NLRB did not apply this
same stricture to tainted locals of an affiliated union, as in a “sweetheart”
situation. In these cases, an affiliated union could appear on the ballot if it
could prove the employer had ceased his support (Crager 1942). On top of
all this, a growing number of company unions were taken over by
national union affiliates, as in the automotive, steel, and agricultural
equipment industries. Company unions made tantalizing targets for
national unions because it was easier to take over an established organiza-
tion than to organize an inchoate group of nonunion employees.
In response to these developments, some company unions transformed
themselves into new organizations—independent local unions
(ILUs)—that were financially and otherwise independent of employers
and which had other features modeled after trade union organization and
procedure (Saposs 1936; Rosenfarb 1940). Although employers gave
implicit winks and nods—and frequently more—to the loyal employees
who formed these ILUs, employers could not legally involve themselves
in the process.
National union leaders were skeptical that any worker would willingly
choose to create an ILU in the absence of employer support or coercion.
An Oil Workers’ (OCAW-CIO) organizer made employee loyalists look
like zombies: In the oil industry, he said, the “concept of company union-
ism had become so well grounded that certain employees, as if in a trance,
went through the motions of creating [a new] series of company unions”
(O’Connor 1950:97). Early NLRB officials took this view as well, lead-
ing to an extremely hostile environment for ILUs.
Membership Estimates, 1935–1940. Of the roughly 1500 company
unions in existence in 1935, how many survived to 1940? To answer this,
we would have to know the number of disestablishments that occurred
from 1935 to 1940 and also the number of successful raids by affiliated
380 / SANFORD M. JACOBY
unions. The number of disestablishments is relatively easy to obtain:
almost 700 through fiscal year 1940, followed by 502 disestablishments
in 1940–1941, the peak year for all time, making a total of 1200
disestablishments from 1935 to 1940–1941 (Millis and Montgomery
1945:850–1; NLRB Annual Reports). If all these were disetablishments
of company unions in existence in 1935 and none involved double count-
ing, then by mid-1941 the NLRB had disestablished about 80 percent of
all company unions in existence 6 years earlier—a colossal achievement.
However, neither of these conditions obtained. Some disestablishments
included company unions formed after 1935, whereas others, as at
Thompson Product’s main plant in Cleveland, constituted multiple incar-
nations of the same original company union (Jacoby 1989). While the lat-
ter occurrence probably was not of great quantitative significance, it is
difficult to say how prevalent was the former since, as mentioned, a num-
ber of new company unions appeared between 1935 and 1939. Even with
these qualifications, it is still the case that a vast number of company
unions were eliminated by the NLRB from 1935 to 1940.
As for raids, company and independent unions participated in 1101
elections between 1935 and 1940–1941, and of these, they lost half, or
556 elections (Millis and Montgomery 1945:853). Unfortunately, we
cannot simply add this figure to the 1200 NLRB disestablishments to get
a total number of company union “deaths” because in a few instances an
independent union defeated during an election refused to concede or was
in the midst of legal proceedings and subsequently was disestablished by
the NLRB. What we can say is this: Administrative decisions by the
NLRB eliminated twice as many ILUs as were eliminated by NLRB elec-
tions from 1935 to 1940, and of those ILUs in existence in 1935 or created
after that date, a very large number had disappeared by 1940.
There is, however, another way of determining what happened to com-
pany unions between 1935 and 1940: Estimate the number of ILUs in
existence in 1940 by counting backward from survivor data available for
1967. We can form this estimate by looking at survivors of the 1940 ILU
population cohort still alive in 1967 and adding to them the number of
cohort deaths that occurred between 1940 and 1967. While the calcula-
tions that follow depend on several assumptions, the adjustments are
made conservatively so as to bias downward the final estimate.
The starting place is a comprehensive Bureau of Labor Statistics (BLS)
survey of all 884 ILUs in existence in the United States in 1967. The BLS
asked these ILUs for their founding dates. Of those specifying a date (12
percent did not say), 171 had been formed prior to 1940. (These 171 ILUs
had 192,000 members in 1967.) If those not reporting a founding date
Unnatural Extinction / 381
contained the same proportion (22 percent) of pre-1940 foundings as the
reporting group, then in 1967 there were 194 ILUs with 204,000 members
who were survivors of the 1940 cohort (BLS 1969:14).
The chief causes of death for the 1940 cohort were disestablishments
and raids. Again, disestablishments are easy to determine: There were
2167 disestablishments between 1940 and 1967, with the bulk of
these—1554—coming between 1940 and 1947 (NLRB Annual Reports).
Of course, not all these disestablishments were of ILUs from the 1940
cohort; some were ILUs “born” after that date. Using the BLS founding-
date data, we can infer that ILUs formed prior to 1941 comprised about
40 percent of all ILUs in the 1940s, 24 percent of all ILUs in the 1950s,
and 22 percent of all ILUs in the 1960s, through 1967. If
disestablishments of the 1940 cohort were proportional to their weight in
the total ILU population over time, then about 700 of the
disestablishments occurring between 1940 and 1967 involved ILUs from
the 1940 cohort.
How many workers belonged to these disestablished ILUs? In 1967,
the average unit size of ILUs founded before 1941 was 1122 members.
These survivor ILUs probably were larger than the average ILU in exis-
tence in 1940. If we conservatively estimate average unit size at 700
members, which is the average number of workers per unfair labor prac-
tice case in 1947, then about 490,000 workers were affected by
disestablishments of the 1940 cohort.2
Next we need to estimate how many of the ILUs alive in 1940 were
raided successfully by affiliated unions. To get this figure, I first calculate
the total number of successful raids by affiliated unions and then estimate
how many of these raids affected ILUs from the 1940 cohort. The chief
source of raiding data is Krislov (1954, 1955, 1960), whose studies
combine ILUs and “left wing” unions for the period 1940–1948 and sepa-
rately identify ILUs for the period 1948–1960.
According to Krislov, the number of raids on independent and
left-wing unions rose and then fell in the 1940s: from 16 (1940) to around
200 (1943 and 1946) and then down to 98 (1948). In the early 1950s,
ILUs made up about half the raids on ILUs and left-wing unions com-
bined. Applying this ratio to the 1940s and smoothing Krislov’s data to
derive annual figures results in an estimated 624 raids on ILUs from 1940
to 1948. How many of these raids were successful? In the 1950s, ILUs
lost 46 percent of all raids initiated against them (Troy 1960). If this
382 / SANFORD M. JACOBY
2The average number of workers per unfair labor practice case heard by the NLRB fell from 1706 in
1943 to 746 in 1947 (NLRB Annual Reports).
proportion was the same in the 1940s—a conservative assumption—then
raids accounted for 287 ILU “deaths” from 1940 to 1948, a period during
which the 1940 cohort comprised 33 percent of the total ILU population.
The bottom line: There were 95 “deaths” from the 1940 cohort due to
raids in the 1940s, with a total loss of members of 47,500 (based on the
average unit size of all units affected by raids in the 1940s of 500
We have better data on raids for the period 1948 to 1958, when ILUs
lost 308 of the 671 raids initiated against them. Applying the 1940 cohort
population weights of 0.46 in 1948 and 0.25 in 1958 gives a total loss due
to raids of 111 units from the 1940 cohort. The average unit size in all ILU
raids, successful or not, during this period was 300, making the total loss
due to raids about 33,000 members.
Therefore, if we add to the 194 cohort survivors in 1967 the sum of
losses due to disestablishments (700) and losses due to raids (206), we
arrive at an estimated number of ILUs in 1940 of 1100, which is 73 per-
cent of the total existing in 1935. Following the huge number (502) of
disestablishments in 1940–1941, there would have been in 1941, ceteris
paribus, 598 ILUs in existence, which is about 40 percent of the popula-
tion in 1935. As for membership, if we add to the 192,000 cohort survi-
vors in 1967 the losses due to disestablishments (490,000) and to raids
(80,500), we arrive at an estimated membership in 1940 of 762,500,
which is 31 percent of the total from 1935. Following the
disestablishments of 1940–1941, there would have been, ceteris paribus,
411,000 ILU members. In short, the number of company unions and their
membership shrank enormously from 1935 to 1941, with much of this
shrinkage due to actions by the NLRB and, to a lesser extent, to raids by
affiliated unions (see Table 1).
The remainder of this section presents an overview of ILU membership
trends for the subperiods 1941–1947, 1947–1953, and 1953–1967.
1941–1947. Net membership did not change very much, if at all, from
1941 to 1947. Troy (1960, 1961) calculates that in 1947 there were about
469,000 ILU members in 222 ILUs. The stabilization of ILU membership
during and immediately after the war was due, in part, to a steady decline
in the number of disestablishments—from 283 in 1942 to 101 in 1944 to
an average of only 47 per year from 1945 to 1947. Based on the number of
workers per unfair labor practice case in these years, total losses due to
disestablishments are estimated to equal 511,000 members for the
Unnatural Extinction / 383
This period witnessed the initiation of raids by ILUs against AFL and
CIO unions, although such activity remained “sporadic” (Krislov
1954:20). On the other hand, raids by affiliated unions against ILUs con-
tinued at a steady pace; about 96,000 ILU members switched to affiliated
unions during and after the war.
Disestablishments and raids against ILUs (accounting for 600,000 ILU
member “deaths”) were roughly offset by ILU “births” during these
years. The BLS data show that of the ILUs in existence in 1967, 146 had
been founded between 1940 and 1945 and an additional 115 between
1946 and 1950. The first cohort in 1967 had 127,000 members; the sec-
ond cohort had 56,000 members. If we interpolate the latter cohort, we
come up with a figure of 139,000 new ILU members joining in
1940–1947 and surviving to 1967. We know that the 1940 population
cohort had a survival rate of 0.25 to 1967. If we conservatively apply the
same survival rate to the 1940–1947 birth cohort, it suggests that 556,000
new members joined ILUs during this period.
Thus from 1941 to 1947, net ILU membership either increased slightly
(based on the difference between my 1941 figure and Troy’s figure for
1947) or decreased slightly (based on my rough comparisons of “deaths”
minus “births”). However, this appearance of stasis in fact masks the
intersection of two sizeable gross flows: “deaths” and “births” of about a
half million individuals each.
1947–1953. According to Troy (1960, 1961), ILU membership rose
between 1947 and 1953 from 469,000 to 706,000, a huge net gain of
237,000 members. Although Troy’s figure may be on the high side—my
estimates show a net gain of only 136,000—it is plausible because, as
during the preceding period, the ILU “death” rate continued to drop due
to a decline in disestablishments, whereas the “birth” rate rose.
The post-Taft-Hartley years saw a brief acceleration in the rate of ILU
formation. The BLS (1969) survivor figures show an average of 24 ILUs
being formed each year from 1940 to 1950 and 29 ILUs formed each year
from 1951 to 1960. Averaging these figures yields an estimate of 179
ILUs with 63,000 members established in the period 1947–1953 and sur-
viving to 1967. Applying a survival rate of 0.25 yields an estimate of
252,000 members joining 700 new ILUs during this period.
What is distinctive about postwar ILU growth is that it occurred at
smaller establishments than previously. According to the survivor data,
the average size of a unit formed prior to 1940 was 1122 employees; new
unit size fell in the late 1940s to 487 and dropped in the 1950s to 211
384 / SANFORD M. JACOBY
Another postwar characteristic was the greater ability of ILUs to win
elections. ILUs did well against national unions, winning a majority of all
representation elections and a majority of votes from 1948 to 1958 (1953
was the one exception). ILUs also prevailed in the majority of raids
against them; their win rate averaged 54 percent for the 1948–1958 period
A sign of heightened ILU confidence was the raids launched by ILUs
against national unions, of which ILUs won about two-thirds (Krislov
1954). Belying their reputation for quiescence, a number of ILUs also
engaged in organizing activity during this period. These efforts remained
modest—the number of elections for unorganized workers in which ILUs
participated peaked at 22 in 1958—but it was yet another indication “of
increasing strength of local independents in competition with ‘legitimate’
unions” (Krislov 1960:219).
“Deaths” during this period occurred chiefly through disestablishments
and raids by national unions: 358 such raids occurred between 1948 and
1953, and national unions won 163 of them. With an average unit size of
about 300 members, such raids accounted for a loss of about 49,000 ILU
members (Troy 1960). As for disestablishments, they continued on a
steady downward trend. There were 134 disestablishments from 1948 to
1953, accounting for ILU losses of about 67,000 members.
In brief, my estimates show a net increase of 403 ILUs during this
period (higher than Troy’s figure; see Table 1) and a net membership gain
of 136,000 (smaller than Troy’s).
1953–1967. The vitality demonstrated by ILUs in the early 1950s did
not last. The total number of elections involving ILUs fell from 294 in
Unnatural Extinction / 385
COMPANY UNIONS AND ILUs, 1935–1967
1935 1515 2,500,000
1940 1100 762,500
1941 598 411,000
1947 222 469,000*
1953 625 605,000
1956 385 709,000*
1961 1277 452,000†
1967 884 475,000†
These figures come from *Troy (1961) and †BLS (1969) or are estimated in this article.
1953 to 205 in 1958; the number of voters dropped from 68,000 to
35,000. To put this another way, the average size of an ILU election unit
fell from 231 to 170 (Troy 1960). Because the ILUs created in the 1950s
were relatively small, they were less stable entities than in earlier years.
Many were one-shot organizations created in response to a union organiz-
ing drive, which then slowly dwindled away (Shostak 1962). When the
BLS (1962) conducted its national census of ILUs in 1961, it could find
only 452,000 members belonging to 1277 ILUs, the majority of which
had fewer than 100 members.
During the 1960s, the formation of new ILUs slowed to a crawl in
nearly all industries and geographic areas. The survivor data show that
only 56 ILUs with 11,000 members were created between 1961 and 1967
(BLS 1969). Moreover, the disestablishment rate rose: from about 27 per
year under the Eisenhower NLRB to 37 per year under the Kennedy-
Johnson board. Other signs of weakness included membership losses in
the telephone industry, where large ILUs in New York and California
succumbed to raids by the Communications Workers of America (CWA),
and in the electrical machinery industry, where the International Union of
Electrical Workers (IUE) raided large ILUs at Stromberg-Carlson,
Sprague, and Wagner Electric (Shostak 1962:67; BLS 1969:12).
Changes in Labor Law
The shrinkage of ILUs after 1935 and their subsequent brief revival
were due to a confluence of causal factors, chiefly the interaction between
ILU characteristics, the evolving labor law, worker attitudes, and
employer strategies. Without doubt, the rapid decline of ILUs from 1935
to 1940 was caused principally by the Wagner Act’s restraints on
employer activity and the way these were interpreted by the early NLRB.
At the 1939–1940 hearings on the Smith bill (a proposal to reform the
Wagner Act that anticipated Taft-Hartley), employers bitterly com-
plained about the NLRB’s hostility to independent local unions and its
practice of disestablishing them even when they no longer received
employer support. The Smith committee later excoriated the NLRB for
having “consistently pursued a policy aimed at the extermination of these
nationally unaffiliated organizations” (Millis and Brown 1950:176,
348–53; Gross 1981).
Although Pearl Harbor curtailed legislative reform efforts in Congress,
employers kept beating the drums for more lenient NLRB treatment of
independent local unions. At a Conference Board roundtable held in New
York in 1942, the head of the National Association of Manufacturers,
386 / SANFORD M. JACOBY
Frederick Crawford, urged the audience to “find a way of putting into the
field a greater awareness of independent unions, Wagner or not,” and then
assailed the NLRB for its “prejudiced attitude” (New York Times, May
21, 1942). The NLRB was sensitive to these criticisms, to charges aired at
the Smith hearings, and to the prospect of renewed legislative activity
when the war ended. Under Harry Millis, who became chairman in
November 1940, the NLRB steered a more centrist course. As noted, the
number of disestablishments dropped steadily: from 502 in 1941 down to
54 in 1945. Doctrinal changes also occurred. In cases where the only
charge was that the ILU had grown out of an employer-supported com-
pany union and where the parties had been bargaining for at least 2 years,
the NLRB now permitted the ILU to appear on the ballot if employer
domination had ceased—the same rule the NLRB had been applying to
affiliated “sweetheart” locals (Gross 1981; Millis and Brown 1950).
When the Taft-Hartley Act was passed in 1947, its company union pro-
visions codified what had already become NLRB procedure. The act for-
bade the NLRB from discriminating between affiliated and unaffiliated
unions. Now, in contrast to the situation under the Wagner Act, the NLRB
interpreted the law in ways favorable to ILUs. When judging cases of
employer support to unions, it no longer paid attention to affiliation
status. Instead, the distinction was between employer domination (the
remedy for which still was disestablishment) and lesser forms of illegal
employer interference. If the employer ceased the latter, the NLRB would
allow the ILU to appear on the ballot. Taft-Hartley also forced the NLRB
to apply weaker fracture rules: Unfair labor practices committed more
than 6 months prior to the filing of an election petition by an independent
local union were not allowed to bear on the petition. These changes effec-
tively grandfathered the legal status of existing company unions.
Another provision of Taft-Hartley was interpreted by some employers
(and courts) to permit the formation of employee committees to discuss
grievances, wages, and working conditions in the absence of a certified
union. Language specifically permitting such committees was contained
in the House version of the act [Section 8(d)(3) of the Hartley bill], but the
Senate version contained no such provision. Instead, the Senate bill added
to the existing Section 9(a) (which read that “any individual employee or
group of employees shall have the right at any time to present grievances
to their employer”) the words “and to have such grievances adjusted with-
out the intervention of the bargaining representative, as long as the adjust-
ment is not inconsistent with the terms of a collective-bargaining contract
or agreement in effect.” While the revised Section 9(a) seems to apply to
situations where bargaining agents are in place, it was capable of being
Unnatural Extinction / 387
interpreted as applying to employee committees in unorganized
workplaces. Indeed, after a conference with the Senate, the House confer-
ees reported to their colleagues that Section 8(d)(3) would be unnecessary
because the revised Section 9(a) “permits individual employees and
groups of employees to meet with the employer.” With Taft-Hartley thus
appearing to sanction committees and with the NLRB meting out milder
remedies for employer interference, the stage was set for an upsurge in
ILU formation. Between 1947 and 1953, ILUs grew at a more rapid pace
than either AFL or CIO membership (Note 1957; Millis and Brown 1950;
Cabot Carbon 1959:215–17).
Rulings favoring ILUs continued to appear in the 1950s, as the first
Republican-appointed board since the NLRB’s inception steered a more
conservative course than its predecessors. However, the Eisenhower
board was reluctant to seek disestablishments, preferring instead to order
employers to cease illegal support. (Disestablishments under the Eisen-
hower board averaged only 26 per year.) As one appellate court said
approvingly, the NLRB no longer had to “baby along employees in the
direction of choosing an outside union as their bargaining representative”
(Coppus Engineering 1957:2315).
The federal courts generally approved of the NLRB’s more lenient
treatment of ILUs. In Chicago Rawhide (1955), which current NLRB
Chairman William Gould (1996) recently termed a “landmark case,” an
appellate court upheld the legality of an employee association formed at a
new nonunion plant in the early 1950s. Straining to find a formula that
would permit the association’s legality—even though the employer had
given assistance of various kinds to it—the court drew a distinction
between “support” and “cooperation.” Support was impermissible
because “even though innocent, [it] can be identified because it consti-
tutes at least some degree of control or influence.” Cooperation, however,
“only assists the employees or their bargaining representative in carrying
out their independent intention.” Assistance, in other words, “does not
always mean domination ...,andtheBoard must prove that the employer
is actually creating company control over the union.” Hence the test of
whether an employer action was support (unlawful) or cooperative assis-
tance (lawful) was not clear-cut. As the court admitted, using language
from another case, it was “not an objective [test] but rather subjective,
from the standpoint of the employees.” From the early 1940s through the
late 1950s, then, the legal atmosphere increasingly was tolerant of ILUs.
388 / SANFORD M. JACOBY
One feature common to ILUs was their identification with a single
company. Affiliated locals were similarly oriented, but they had other,
more diffuse concerns: with a craft, occupation, or industry and with the
national union itself. Inside large companies the unitary link between the
ILU and the employer reinforced personnel policies intended to foster a
career-type employment relationship. Hence there was a strong associa-
tion between the presence of ILUs and enterprise-oriented policies such
as systematic personnel management and fringe benefits (Nelson 1982;
As compared with industrial unions, ILUs were conservative organiza-
tions, suspicious of “big” government and “big” labor. After the war, this
conservatism helped ILUs to proliferate in smaller companies outside the
big cities, places that traditionally were wary of outsiders. Because of
their focus on a single firm and their isolation from other unions, many
ILUs revolved around the personalities of their leaders. The ILU’s
person-based culture meshed smoothly with—indeed, reinforced—the
family-oriented and charismatic type of management existing in compa-
nies dealing with ILUs. Yet ILU informality had its downside: It some-
times created oligarchies that were less tolerant of dissent than the more
bureaucratic but also more democratic leadership of affiliated unions.
When Shostak (1962) investigated ILUs in the late 1950s, he found that
roughly half the ILUs in his study were “weak”: organizations that pre-
pared poorly for negotiations and rarely challenged the employer’s
authority. Many of them were small, semirural, and inexperienced. Few
were survivors dating back to the 1920s or the 1930s; instead, most had
been created in the 1940s and early 1950s. Many had been started in
response to a union organizing drive, usually by the employer acting with
support from conservative employee loyalists. In one plant, for example,
the ILU prospered because the employees were fearful of the Teamsters;
in another, the ILU was led by a former small businessman who attributed
his company’s failure to the Ladies’ Garment Workers Union.
On the other hand were Shostak’s “strong” ILUs, which were more
independent than weak ILUs but less militant than most national unions.
While strong ILUs rarely went on strike, they had other ways of pressur-
ing the employer, such as rejecting a contract or threatening to affiliate
with a national union. Strong ILUs often were found in urban areas,
where they were exposed to regular raids from national unions. Because
of this—and because of their internal structure—these ILUs often did a
better job of “servicing” their members than did business agents who had
to deal with numerous affiliated locals. For example, at Thompson
Unnatural Extinction / 389
Products, which had both types of unions, the grievance rate inside one
ILU plant was lower than inside an affiliated-union plant represented by
the United Autoworkers (UAW), but the ILU pressed a greater number of
these grievances through to arbitration. At Swift & Company, which also
had both types of unions, the ILU plant had a higher grievance rate and
steward ratio than the AFL-affilated plant (the Meatcutters) but lower
than the plant represented by the CIO’s Packinghouse Workers (Jacoby
and Verma 1992; Purcell 1960).
A selling point for ILUs was that they offered their services for less
than the dues charged by national unions. Yet the bargain here was more
apparent than real, since ILUs provided fewer services—recreational,
educational, and informational—than most national unions. On the other
hand, ILUs did possess some real advantages as a result of their financial
and organizational autonomy—what Strauss (1991) calls “union decen-
tralization.” This permitted a vigorous, almost Jeffersonian kind of
democracy that gave members more influence than in larger and more
bureaucratic unions. This was especially true of “strong” ILUs that prided
themselves on their members’ active participation. Many such ILUs had
started out looking like their weaker counterparts, but competition with
national unions and situational factors caused them to evolve over time.
In fact, the strong unions in Shostak’s sample were, on average, older
than the weak ones (Shostak 1962; Jacoby 1997; Personnel 1955).
This same evolution occurred at Thompson Products (today TRW)
when what once had been quiescent company unions initiated by manage-
ment in the early 1930s were transformed in the 1940s and 1950s into
relatively independent, though cooperative ILUs (Jacoby 1989, 1997).
The largest of Thompson’s ILUs was the Aircraft Workers Alliance
(AWA), which represented workers at Thompson’s factories in the
Cleveland area. Over time, the AWA developed a cadre of leaders who
were adept at contract administration and bargaining. These leaders were
rarely hostile to management, although they were far from obsequious.
Unlike, say, a “sweetheart” union, the AWA disagreed with management,
stubbornly pressed grievances, and persuaded the company to change its
policies and plans. Nevertheless, the AWA’s leaders rarely displayed the
belligerence typical of many affiliated-union locals. Some of this can be
traced to the use of joint council meetings, which permitted a regular air-
ing of differences rather than letting them build up to eruptions at contract
expirations. The council meetings had a collaborative, collegial, and ami-
able atmosphere. (Not without reason had the old IWW slogan averred,
“Feet which meet under a mahogany table don’t kick.”) Then there was
the fact that the AWA, in contrast to an affiliated local, offered no
390 / SANFORD M. JACOBY
opportunities for retiring officers to move into union staff positions.
Leaders returned to their old jobs when their terms expired unless they
were promoted into management, and promotion was not likely if they
had been too contentious.
A striking feature of the AWA was its isolation from other unions. It
had few contacts with and no formal ties to local unions at Thompson
Products plants outside Cleveland. Within Cleveland, the local AFL and
CIO were contemptuous and wary of the AWA, which repaid these feel-
ings in kind. Too small to train its own officers and cut off from main-
stream labor education programs, the AWA had to rely on management to
meet some of its basic organizational needs. The company held classes
for newly elected stewards, where they received information on Thomp-
son’s personnel practices, financial situation, and markets. Company
managers sought to “integrate” new AWA officers by taking them to con-
ferences of the American Management Association and by holding train-
ing sessions for them on technical subjects such as job evaluation. Given
all this, it is not surprising that when the UAW reappeared at Thompson’s
Cleveland plants in the 1960s, its organizers castigated the AWA’s “cozy
relationship” with management. From a trade union perspective, there
was something odd about the AWA’s isolation and its dependence on
management. However, the AWA seems less peculiar when viewed from
within, that is, when set in the context of Thompson’s own evolving
species of progressive welfare capitalism.
Not every ILU was as isolated as the AWA. Throughout the nation
were federations comprised of ILUs from the same company, city, geo-
graphic region, or industry. Examples of single-company federations
included the National Brotherhood of Packinghouse Workers (NBPW) at
Swift and the Central States Petroleum Union, which in the 1950s repre-
sented about 5000 employees belonging to several ILUs at Standard Oil
of Indiana. Multiemployer federations were prominent in the oil refining
industry, where one could find organizations such as the Federation of
Independent Oil Unions, started in 1955 by six ILUs facing organizing
drives from the CIO-affiliated Oil, Chemical, and Atomic Workers
(Brandt 1960; Marshall 1961; Purcell 1960; Shostak 1962). The largest
multiemployer federations, however, were in the telephone industry. Dur-
ing the late 1930s and 1940s, the National Federation of Telephone
Workers (NFTW)—the industry’s main national union—was comprised
of ILUs from the various Bell operating companies. After the CWA
emerged out of the NFTW in 1947, some of the Bell ILUs—notably those
representing white-collar office personnel and high-wage employees on
the Pacific Coast and in the Northeast—chose to remain independent,
Unnatural Extinction / 391
while forming new federations such as the Alliance of Independent Tele-
phone Unions and the New England Federation of Telephone Operators.
There also were several multiemployer federations of white-collar and
professional ILUs, including the National Federation of Salaried Unions,
which claimed 40,000 members in the early 1950s, and the Engineers and
Scientists of America (ESEA), comprised of ILUs representing engineers
and draftsmen at Boeing, Douglas, Sperry, Honeywell, and other firms
(Schacht 1985; Barbash 1952; Troy 1958; Waters 1954; Fortune
1956:203; Walton 1961).
Resembling the AFL and CIO were the multi-industry ILU federations,
of which the largest was the National Independent Union Council, started
by the NBPW in 1950 and also including ILUs at companies in Illinois,
Kansas, and Missouri. Similar organizations included the Confederated
Unions of America, started in 1942, and the National Federation of Inde-
pendent Unions (NFIU). The three federations merged in 1960 to form
the Congress of Independent Unions, which, in addition to the Swift
locals, included ILUs at Zenith Electronic, ARMCO Steel, Elgin Watch,
and other firms. Among the services these federations provided for their
members was a staff of skilled negotiators, information on contracts and
wages, a monthly magazine, and the “advantage of membership in a bona
fide labor federation which protects them from the dangers of isolation;
NFIU affiliates cannot be labeled ‘company unions’ or be accused of
being under management or any other domination” (Union Labor News
Because most ILUs were too small to hire a permanent staff, they
depended heavily on advice given to them by lawyers who specialized in
representing ILUs. Occasionally, these attorneys amalgamated their
clients into a federation, although these “federations” usually were noth-
ing more than an entrepreneurial gimmick to garner new clients. Milton
Roemisch of Cleveland was one of the new breed of lawyers who
emerged during the 1930s and 1940s as advisers to and often de facto
leaders of ILUs. Roemisch represented independent unions at a number of
Cleveland area firms, including Ohio Tool, Sherwin-Williams Paint,
Thompson Products, and Ohio Crankshaft. In the early 1940s, Roemisch
affiliated 26 of these independents into a federation called the National
League of American Labor (Jacoby 1997).3
392 / SANFORD M. JACOBY
3Also see Calco Chemical, 12 NLRB 275 (1939) on the Hamilton Labor Bureau.
Worker and Management Attitudes
That ILUs could win a majority of the NLRB elections in which they
participated during the 1950s suggests that workers had some consistent
reasons for preferring ILUs over both national unions and no unions. An
important factor was the desire of ILU members to preserve compensa-
tion premiums that might otherwise be eliminated by national unions.
Union organizers readily admitted that ILUs frequently paid at or above
the union rate; an OCAW official made the acid remark that ILU mem-
bers “have everything that union men do in wages and working condi-
tions, everything—except freedom” (O’Connor 1950:97). These
premiums were endogenous—they were due to preexisting skill or geo-
graphic wage differentials—but they were threatened by the wage stan-
dardization policies pursued by national unions. Troy (1960) cites
examples drawn from the Bell System—where ILUs were found in
high-wage cities such as Chicago, New York, and Los Angeles, whereas
the CWA had its locals in lower-wage cities such as Atlanta, Cleveland,
and Philadelphia—and from other industrial unions such as the Transit
Workers and the Automobile Workers, which periodically experienced
defections of their skilled trades into ILUs. Hostility of the NLRB to craft
severance, despite protests from the AFL prior to its merger with the CIO,
thwarted what might otherwise have been a source of additional ILU
members (Tomlins 1985).
Another reason that ILUs had support from their members was the sim-
ple fact that they charged low dues. The ratio of ILU–to–national union
dues ranged from one-quarter at Swift to one-sixth at Thompson Products
and even less in some small companies (Purcell 1960; Jacoby 1997).
Third, many ILUs offered a more direct kind of union democracy than
national unions could provide. All an ILU’s officers—from the president
on down—were from the plant and well known by the members. This
familiarity, together with the fact that ILUs constantly were being chal-
lenged by national unions, made ILU officers responsive to member
demands, a key ingredient in union democracy (Robinson 1968; Shostak
1962; Sayles and Strauss 1953). At Thompson Products, for example,
ILU elections were hotly contested and highly publicized. Unlike an affil-
iated local of a national union, the ILUs were masters of their own fate.
They did not have to answer to a higher authority or conform to policies
that were not of their choosing. Because they had no staff and only a sim-
ple hierarchy, the ILUs tended to be less formal than a national union, and
although the UAW faulted Thompson’s ILUs for failing to hold regular
meetings, the latter’s small size and numerous social activities kept its
officers in close touch with the “middlemass” of rank and file members
Unnatural Extinction / 393
(Strauss 1977). There were no outside experts or business agents telling
members how to run the union. The ILU’s claim to be “your union” con-
jured up a more appealing image than the UAW’s promise of providing
“one big union, coast to coast.” Realizing this, the AWA referred to the
UAW as “the octopus” and urged Thompson workers not to feed it
Because ILUs often were found in small firms outside urban centers,
their members tended to be conservative and pro-management. Members
nurtured dreams of self-employment that caused them to identify with the
producerist values of the small businessmen who employed them and
who were a daily presence on the shop floor and in the community. Loy-
alty to management made ILU members reluctant to strike, and in fact,
strikes by ILUs were a rare occurrence. At Swift, the ILUs four times
refused to join in strikes called by the national meatpacking unions in the
late 1940s and 1950s, complaining that the nationals relied too heavily on
the strike as a bargaining tactic. National unions were disliked for other
reasons: their liberal proclivities, Communist affiliations, or simply the
fact they were—like chain stores—large organizations from the “out-
side.” Alongside the story of the CIO’s victories in Chicago, Detroit, and
Pittsburgh, there is a different tale to be told about the conservative and
isolationist values that united managers and workers outside the cities,
even in labor’s Midwestern redoubt. Think, if you will, of the five ILUs
formed by the Christian Labor Association, all of whose members
belonged to the Dutch Reformed Church and lived in the vicinity of Hol-
land, Michigan (Purcell 1960; Shostak 1962; Millis and Montgomery
1945:869; Nelson 1995).
Finally, ILUs drew support from groups who felt that national unions
were insensitive to their needs and concerns. Despite national unions’
claim to be inclusive—to possess what one historian terms a “culture of
unity” (Cohen 1990)—the climate inside the new CIO unions often was
similar to that of the AFL’s, that is, dominated by the concerns of white,
male, manual workers. The leadership of the CIO unions often con-
demned racial discrimination, but matters were different at the plant level,
where white workers viewed the union as their special preserve and tried
to keep blacks out. This was the case at Thompson Products, where white
UAW members discriminated against black workers, turning the latter
into stalwart supporters of the ILU and also of management, which went
out of its way to hire black workers during the 1940s. At Swift, black
workers much preferred the ILU to the AFL’s Meatcutters Union, which
had a long history of racial discrimination (Purcell 1960; Jacoby
1997:175–6; Sugrue 1996).
394 / SANFORD M. JACOBY
Female workers were another group who gave disproportionate sup-
port to ILUs. Data from the late 1960s show that the percentage of women
in ILUs was nearly twice as high as in national unions (BLS 1969). While
there is little research on the subject, it is likely that the strength of ILUs
in the Bell System owed something to the fact that women preferred these
organizations to national unions. Sixty-five percent of the phone compa-
nies’ employees were female, and although many explanations exist for
the correlation between ILUs and female employment, surely one of them
is the fact that women were turned off by the “macho” take-charge atti-
tudes of national union organizers, who could be insensitive to the needs
of women workers. For this as well as other reasons, women were rela-
tively loyal to top management. When Purcell (1960:67) surveyed Swift
employees, he found that women in the ILU plants demonstrated the
highest level of corporate loyalty of Swift’s employee subgroups. Tele-
phone industry ILUs sometimes made explicit recognition of gender, as
with the Federation of Women Telephone Workers, which had nearly
13,000 members in California in the 1950s (Troy 1958).
Reinforcing gender was the fact that women, like those in the telephone
industry, were disproportionately employed in white-collar occupations
—yet another group that was relatively loyal to management. The BLS
(1969) data show that the proportion of white-collar workers belonging to
ILUs was more than twice as high as for national unions. One of the
better-known instances of this phenomenon was the various ILUs repre-
senting engineers and technicians at companies such as General Electric,
Lockheed, Boeing, RCA, Westinghouse, and other firms. Most of these
unions were formed after the war to avoid inclusion in larger industrial
unions dominated by blue-collar workers. The professionals and
semiprofessionals in these ILUs found themselves between a rock and a
hard place: subject to occasional raids by national unions while under
constant criticism from the national engineering societies, who asserted
that professionalism and unionism—even of the ILU variety—were
incompatible (Walton 1961; Shostak 1962).
Management Attitudes. How did managers regard ILUs? This question
is difficult to answer because management’s usually positive appraisal of
ILUs was based on consideration of the next worst alternative, that is, an
affiliated union. In a world without affiliated unions, management would
have been far less tolerant of organizations as independent and unpredict-
able as ILUs. However, they still might have sought (as they did in the
pre-Wagner era) analogous institutions for handling employee griev-
ances, for securing employee participation in enterprise affairs, and for
Unnatural Extinction / 395
organizing employee-related activities. In fact, this combination of
functions—voice, governance, and welfare—was the basis for the first
employee representation plans that appeared at the turn of the century,
many of them at companies that had little to fear from trade union organi-
zation but instead were seeking a more participatory employment system
To the extent that ILUs sustained this participatory logic, managers
were happy with them. And even when an ILU displayed some mettle—
in bargaining and grievance handling—managers were tolerant if they
believed that the ILU helped to repel national unions, especially those
affiliated with the dreaded CIO. This was the situation for most of the
period from the mid-1930s to the mid-1950s. However, several events
changed managerial preferences. First, the CIO appeared less threatening
after it shed its Communist wing in the late 1940s and merged with the
AFL in 1955. Second, this same period saw the capacity of the labor
movement to attract new members noticeably weaken. And third, as we
will see, employers developed an alternative to ILUs that was less
“unionesque” while still providing a mechanism for employee participa-
tion. As a result of these events, management’s interest in ILUs gradually
Managers of the 1940s and 1950s preferred an ILU to an affiliated
union because of the ILU’s conservative orientation, its divorce from pol-
itics, and the fact that it harbored no left-wing “eggheads” on its staff
(Shostak 1962). Moreover, the ILU was more inclined to take a coopera-
tive, problem-solving approach in its relations with management. Part of
this was due to the simple fact that an ILU’s top officers were always
available to make adjustments on the spot. There also were cooperative
incentives arising from the existence of a bilateral monopoly: The ILU
had no other company to bargain with, and management preferred it to
any other union (Jacoby 1989). Evidence of cooperation could be seen in
many dimensions, from the absence of strikes to a more flexible organiza-
tion of work. For example, in the oil refinery industry of the 1950s, only
ILUs—not affiliated unions—had craft consolidation plans for reducing
the complexity of craft demarcations. And at TRW (the successor to
Thompson Products), the company’s ILU plants had fewer job classifica-
tions than its affiliated union plants (Brandt 1960:148; Jacoby and Verma
However, managers also had reasons to be skeptical of ILUs, often
regarding them as a species of Trojan horse. ILU members made tempting
organizing targets for national unions because, unlike workers in non-
union firms, they already were a collectivity and were familiar with the
396 / SANFORD M. JACOBY
arcana of contracts and grievance procedures. Even though national
unions were not always successful in raiding ILUs, they tried repeatedly.
Raids against ILUs increased in the 1950s as the number of easily orga-
nized nonunion enterprises (i.e., large blue-collar operations outside the
South) steadily shrank. Ten percent of all ILUs experienced an organizing
raid each year in the 1950s, surely an annoyance to management. A single
ILU plant at Swift experienced no fewer than seven raids from 1943 to
1959, even though it won all of them by comfortable margins (Krislov
1960; Shostak 1962; Purcell 1960).
Not only raids but labor costs, too, made ILUs a costly labor relations
strategy. ILU labor costs were equal to—and in some instances higher
than—those of a comparable affiliated union. This premium reflected the
aforementioned skill or geographic premiums that ILUs were seeking to
preserve, or it constituted rents that ILUs extracted from management in
return for not affiliating with a national union. There is considerable
evidence of ILUs strategically playing on management’s affiliation fears
at Thompson Products from the 1940s all the way through the 1970s. The
increase in raiding activity in the 1950s no doubt gave ILUs greater lever-
age in this regard (Jacoby 1997).
Faced with these immediate problems—and convinced that unions,
especially militant CIO unions—were a declining threat, some companies
that earlier had pursued an ILU strategy began to abandon it. Take, for
example, the experience of DuPont. Throughout the 1930s and the war,
DuPont established company unions at its U.S. facilities; by 1946, 85 per-
cent of its employees were represented by them. After the war, however,
DuPont made a significant shift. Over the course of the next decade it
opened 25 new plants, many in the South and all without ILUs. By 1960,
ILUs represented only 59 percent of DuPont’s employees, and the per-
centage has fallen steadily since then (Rezler 1963).
Not every company was as quick as DuPont to shed the ILU strategy.
Thompson Products stuck with it after the war, establishing ILUs (or
something akin to them) at all the new plants it built or acquired in the late
1940s and 1950s. Unlike DuPont, Thompson’s plants were concentrated
in the Midwest and Northeast, where organizing drives remained a threat.
Also, Thompson’s ILUs were less militant than DuPont’s. Between 1944
and 1959, Du Pont’s ILUs carried out 4 strikes and filed 16 unfair labor
practice charges against management, whereas not one of Thompson’s
ILUs struck or complained to the NLRB. This record, as well as its ILUs’
willingness to cooperate in boosting efficiency, convinced Thompson
management that ILUs raised productivity sufficiently to offset their
wage costs (Rezler 1963; Jacoby 1997).
Unnatural Extinction / 397
This is not to say that Thompson management was entirely satisfied
with its independent local unions, however. Wage costs were high, and
raids—and the legal expenses that went with them—were a perennial
issue. Thus, in the early 1950s, Thompson began to experiment with an
alternative to ILUs—the committee system—at three small plants
acquired after the war. While the impetus for seeking an alternative to
ILUs may have been internal, the fact that the Taft-Hartley Act now
appeared to allow the formation of employee committees was a permis-
sive, if not directly causal, factor. Employee committees were more
decentralized and informal than ILUs, making it harder for organizers to
transform them into national union affiliates. Also, the committee system
explicitly ruled out collective bargaining, thus eliminating the labor-cost
pressures associated with ILUs. On the other hand, management hoped
that the new system would replicate what it saw as the positive features of
company unions and ILUs: two-way communication, rapid dispute reso-
lution, and a cooperative approach to problem solving. Thus Thompson
was trying to create new institutions to preserve the communicative fea-
tures of ILUs without an excess of independence or bargaining power.
In keeping with Thompson’s diversification into technology-intensive
products, two of the three “new system” plants were electronics factories:
Bell Sound Systems, a Columbus, Ohio, manufacturer of amplifiers and
communications equipment, and Dage Television, which made
closed-circuit televisions for the military in Michigan City, Indiana. (The
third plant was an automotive valve factory in Fruitport, Michigan.) Over
a 3-year period, from 1953 to 1956, Thompson management introduced
virtually identical organizations in all three plants: the Voice of the Plant
(Bell Sound), the Frontiersmen Association (Fruitport), and the Dage
Employees’ Association. Neither the Bell nor the Fruitport organization
met directly with management. Instead, their elected officers spent their
time directing the plants’ numerous recreational and welfare activities.
Separate from these organizations, however, were “human relations com-
mittees” that met with management to discuss problems and complaints.
In order to avoid coming under NLRB purview, the purpose of the com-
mittees was officially described as “the exchange of ideas on matters
affecting human relations...andgood living within the company”
The Dage plant, the smallest of the three, did things a bit differently. As
at the other two plants, acquisition by Thompson brought a rapid expan-
sion of employee recreational activities and regular use of advanced per-
sonnel techniques such as attitude surveys. However, local management
felt that the Dage employees would not be satisfied with the vagaries of
398 / SANFORD M. JACOBY
the Bell-Fruitport model and pressed company officials for something
that would “provide as many advantages of a union-type organization as
possible in order to avoid possible outside interference.” The end result,
the Dage Employees’ Association, was a hybrid combining some features
of an ILU (regular meetings with management) and others taken from the
new committee system (handling welfare activities autonomously and
carefully refusing to “deal” with management on wages or working con-
ditions) (Jacoby 1997).
The new approach received sanction from the courts. The more lenient
legal environment prevailing after Taft-Hartley, as reflected in decisions
such as Chicago Rawhide (1955) and Coppus Engineering (1957), no
doubt encouraged Thompson to stay the course and caused other firms to
imitate Thompson. For example, General Electric in the 1950s pioneered
the use of employee associations known as sounding boards. These
brought together engineers and managers in a single association to con-
sider mutual problems. The National Society of Professional Engineers
gave its imprimatur to these associations, viewing them as more suitable
for engineers than a bargaining group like the ESEA (Shostak 1962). It is
difficult to say how prevalent these committees were in the 1950s,
although one study done in the late 1950s of two dozen nonunion firms in
an unidentified Midwestern city found that nine of them (40 percent) had
some kind of representation plan, and only one of these was an ILU. The
other eight were employee committees whose rotating membership was
appointed by and met with management on a regular basis to discuss
workplace issues (Conant 1959).
All this activity came to a sudden halt after 1959, however, when the
Supreme Court—in its Cabot Carbon decision—ruled emphatically that
an employee committee that does not bargain with management neverthe-
less is a labor organization governed by Section 8(a)(2). After analyzing
the legislative history of Section 9(a) as amended, the Court concluded
that Congress had never intended to authorize an employer to deal with an
employer-supported committee (Cabot Carbon 1959).
The consequences of Cabot Carbon were swift and wide. Employers
who used the committee system immediately proposed a federal law that
would explicitly exclude committees and “sounding boards” from the
law’s definition of labor organizations; the parallels to recent events
(“Electromation” and the proposed TEAM Act) are striking (Shostak
1962:85). However, legislative efforts were fruitless, and the NLRB and
lower courts followed the Supreme Court’s lead. One of the first compa-
nies affected was TRW: A 1960 organizing drive at Dage had resulted in
unfair labor practice charges being filed against the plant’s committee
Unnatural Extinction / 399
system. In a decision later upheld by the Seventh Circuit, the NLRB ruled
that the Dage committees were a labor organization and that they were
dominated by management (TRW 1961, 1962). Although the company
asserted that the committees never bargained and merely “expressed
views and conveyed information to management,” the board rejected the
claim and ordered TRW to get rid of them. Other decisions in the 1960s
were equally inimical to employee committees, including Walton Manu-
facturing (1960) and Dennison Manufacturing (1967)—a firm that was
one of the first to experiment with company unions in the 1910s.
However, not every post-Cabot decision covering employee commit-
tees held them unlawful. In cases such as Modern Plastics (1967) and
Hertzka & Knowles (1975), the courts found committees to be permissi-
ble, usually because there was insufficient evidence that the committees
were dominated by the employer. And even in cases where committees
were held unlawful, one heard dissenting opinions such as that of an
appellate justice who argued that “an inflexible attitude of hostility
toward employee committees defeats the Act. It erects an iron curtain
between employer and employees, penetrable only by the bargaining
agent of a certified union, if there is one, preventing the development of a
decent, honest, constructive relationship between management and labor”
(Walton Manufacturing 1961:182).
Despite these exceptions, for most employers, prudent conduct was to
abandon employee committees. With its resounding defeat in Dage, TRW
shifted gears and gave up on both ILUs and the new committee system.
During the 1960s, it built or acquired 39 new plants and created ILUs at
none of them (Verma and Kochan 1985). The 1960s saw the introduction
at TRW of yet another hybrid, this one combining a new “small group”
approach (sensitivity groups and other team-building programs) together
with a bevy of communications programs whose roots stretched back to
the 1930s (attitude surveys, “sensing” sessions, and leadership training).
The new approach was more individualized, emotionally engaging,
and rooted in the behavioral sciences than ILUs or the committee system.
Combined with attitude surveys and other psychological techniques, it
became the basis for the “new” nonunion model of the 1960s and 1970s
(Kochan, Katz, and McKersie 1986). However, the small-group approach
also was more passive, less democratic, and less of a threat to managerial
prerogatives than ILUs had been. Still, few in either management or orga-
nized labor mourned the passing of the ILU. In fact, managers saw the
new approach as consistent with the thrust of earlier efforts. As TRW’s
vice president for personnel told Business Week (1966:92), “Openness,
leveling, listening therapy, conflict resolution—all this talk by behavioral
400 / SANFORD M. JACOBY
scientists. Hell, we’ve been doing these things for 30 years. They’re just
elegant terms for principles we practiced years ago—old wine in new
Company unions represented over 2 million workers in 1935, but by
the end of World War II, most of these company unions had disappeared.
One of the driving forces behind the demise of company unions was the
NLRB, which, especially in its early years, worked hard to disestablish a
large number of them. Others were taken over by national unions, which
offered a more militant and innovative form of representation.
Yet company unions did not entirely disappear from the scene. Several
hundred of them evolved into relatively independent organizations
—ILUs—that offered an alternative to national unionism in the 1940s and
1950s. For employers, the benefits of ILUs were clear: They were less
aggressive, strike-prone, and politicized than many affiliated unions. For
ILU members, the story is more complex. Some members no doubt were
misled or coerced into supporting ILUs, which is the traditional union
perspective. There was some of that, to be sure, but there was more. ILU
members were a relatively conservative group—loyal to their union and
to the companies that employed them. And the ILUs they belonged to
were dynamic institutions, which, due to situational incentives and to
pressure from government and national unions, evolved into organiza-
tions that met their members’ needs. Compared with national unions,
ILUs were less independent and powerful; on the other hand, they also
were less bureaucratic and remote.
By the 1950s, however, a paradox began to emerge: Just when the law
was making it easier for employers to start ILUs, companies that already
had them began to lose interest. The explanation is straightforward. With
the threat from organized labor fading, the benefits of a company union
faded too; meanwhile, as ILUs grew more independent, they became
more costly to deal with. Managements eventually found a cheaper alter-
native, one that channeled employee participation through committees
and, later on, through small groups instead of elected bodies. Thus the
ILU was a precursor to some of the more innovative participatory
schemes developed by behavioral scientists in the 1960s and 1970s
(Strauss 1968, 1974).
Nevertheless, something was lost during the transition from ILUs to the
“small group” approach. Independent local unions, however imperfect, at
least relied on formally democratic methods that empowered union
Unnatural Extinction / 401
members. ILUs also were able to influence management at the corporate
and strategic levels. On the other hand, the behavioral sciences—in both
their early (human relations) and modern (small group) manifesta-
tions—have a tendency to discourage self-organization, ignore union
interests, and restrain employee influence to the workplace level. For
these reasons, employee involvement (EI) remains anathema to many
U.S. trade unionists (Juravich 1998).
These trade unionists are even more suspicious of ILUs and of legisla-
tion like the proposed TEAM Act. This is not the place to analyze the pros
and cons of the TEAM Act (see Gely 1997). Suffice it to say that the leg-
islation provides insufficient protection to organizing activity; on the
other hand, it offers the possibility of a superior alternative to nonunion
EI plans. In large, high-wage, nonunion companies—places where the
labor movement has largely ceased organizing—something akin to ILUs
would provide employees with a more democratic form of participation
than contemporary EI plans and would function very much like the works
councils that exist across continental Europe. In fact, the United States
today might very well have a European-style industrial relations system
had the Wagner Act taken a different approach to company unions. Euro-
pean governments, employers, and unions long ago reached a rapproche-
ment with their works councils. The time has come for their U.S.
counterparts to consider why they are unable to do the same.
Barbash, Jack. 1952. Unions and Telephones: The Story of the Communications Workers of Amer-
ica. New York: Harper and Bros.
Barenberg, Mark. 1993. “The Political Economy of the Wagner Act: Power, Symbol, and Work-
place Cooperation.” Harvard Law Review 106:1381–496.
Brandt, Floyd S. 1960. “Independent and National Unionism in the Oil Refining Industry.” Ph.D.
dissertation, Harvard University.
Bureau of Labor Statistics (BLS), United States. 1938. Characteristics of Company Unions. Bulle-
tin No. 634. Washington, DC.
_____. 1962. Unaffiliated and Single-Employer Unions, 1961. Bulletin No. 1348. Washington,
_____. 1969. Unaffiliated Intrastate and Single-Employer Unions, 1967. Bulletin No. 1640. Wash-
Business Week. 1966. “Where Diversity Is the Tie that Binds.” September 24, 1966, p. 92.
Cabot Carbon et al. v. NLRB. 1959. 360 U.S. 203.
Chicago Rawhide Mfg. v. NLRB. 1955. 221 F. 2d (7 CCA).
Cohen, Lizabeth. 1990. Making a New Deal: Industrial Workers in Chicago, 1919–1939. Cam-
bridge, UK: Cambridge University Press.
Conant, Eaton H. 1959. “Defenses of Nonunion Employers: A Study from Company Sources.”
Labor Law Journal 10(February):100–9.
Coppus Engineering v. NLRB. 1957. 240 F. 2d 564 (1 CCA).
Crager, Burton. 1942. “Company Unions under the National Labor Relations Act.” Michigan Law
402 / SANFORD M. JACOBY
Dennison Manufacturing and United Papermakers and Paperworkers. 1967. 168 NLRB 1012.
Dubofsky, Melvyn. 1994. The State and Labor in Modern America. Chapel Hill: University of
North Carolina Press.
Fortune. 1956. “Independents Begin to Huddle.” 53(March):203–6.
Gely, Rafael. 1997. “Whose Team Are You On? My Team or My TEAM?” Rutgers Law Review
Gould, William. 1996. “Remarks on Workplace Cooperation.” 208 DLR (October 28):E1–3.
Gross, James A. 1981. The Reshaping of the National Labor Relations Board: National Labor
Policy in Transition, 1937–47. Albany: SUNY Press.
Hertzka & Knowles v. NLRB. 1974. 503 F. 2d 625.
Jacoby, Sanford M. 1985. Employing Bureaucracy: Managers, Unions, and the Transformation of
Work in American Industry, 1900–1945. New York: Columbia University Press.
_____. 1989. “Reckoning with Company Unions: The Case of Thompson Products, 1934–1964.”
Industrial and Labor Relations Review 43(October):19–40.
_____. 1993. “Reflections on Labor Law Reform and the Crisis of American Labor.” Chicago-
Kent Law Review 69(1993):219–27.
_____. 1997. Modern Manors: Welfare Capitalism Since the New Deal. Princeton: Princeton
_____ and Anil Verma. 1992. “Enterprise Unions in the United States.” Industrial Relations
Juravich, Tom. 1998. “Employee Involvement, Work Reorganization, and the New Labor Move-
ment: Toward a Radical Integration.” New Labor Forum 1(Spring):84–91.
Kaufman, Bruce. 1996. “Why the Wagner Act? Reestablishing Contact with Its Original Purpose.”
Advances in Industrial and Labor Relations 7:15–68.
Kochan, Thomas, Harry C. Katz, and Robert B. McKersie. 1986. The Transformation of American
Industrial Relations. New York: Basic Books.
Krislov, Joseph, 1954. “Raiding Among the ‘Legitimate’ Unions.” Industrial and Labor Relations
_____. 1955. “The Extent and Trends of Raiding Among American Unions.” Quarterly Journal of
_____. 1960. “Organizational Rivalry Among American Unions.” Industrial and Labor Relations
Marshall, F. Ray. 1961. “Independent Unions in the Gulf Coast Oil Petroleum Refining Industry.”
Labor Law Journal 12(September):823–40.
Millis, Harry, and Emily Clark Brown. 1950. From the Wagner Act to Taft-Hartley. Chicago:
University of Chicago Press.
_____ and Royal E. Montgomery. 1945. Organized Labor. New York: McGraw-Hill.
Mitchell, Daniel J. B. 1986. “Inflation, Unemployment, and the Wagner Act: A Critical Reap-
praisal.” Stanford Law Review 38(April):1065–95.
Modern Plastics v. NLRB. 1967. 379 F. 2d 201.
NLRB. Various years. Annual Report of the National Labor Relations Board. Washington, DC.
Nelson, Daniel, 1982. “The Company Union Movement, 1900–1937: A Reexamination.” Business
History Review 56:335–57.
_____. 1995. Farm and Factory: Workers in the Midwest, 1880–1990. Bloomington: Indiana
Note. 1957. “Section 8(a)(2): Employer Assistance to Plant Unions and Committees.” Stanford
Law Review 9(March):351–65.
O’Connor, Harvey. 1950. History of the Oil Workers International Union (CIO). Denver: OWIU.
Personnel. 1955. “The Case for the Local Independent Union.” 32(November):226–33.
Purcell, Theodore V. 1960. Blue Collar Man: Patterns of Dual Allegiance in Industry. Cambridge,
MA: Harvard University Press.
Rezler, Julius. 1963. “Labor Organization at DuPont: A Study in Independent Unionism.” Labor
Unnatural Extinction / 403
Robinson, James W. 1968. “Structural Characteristics of the Independent Union in America.”
Labor Law Journal 19(July):417–37.
Rosenfarb, Joseph. 1940. The National Labor Policy and How It Works. New York: Harper & Bros.
Saposs, David J. 1936. “Organizational and Procedural Changes in Employee Representation
Plans.” Journal of Political Economy 44(December):803–11.
Sayles, Leonard, and George Strauss. 1953. The Local Union: Its Place in Industrial Plants. New
York: Harper & Bros.
Schacht, John. 1985. The Making of Telephone Unionism, 1920–47. New Brunswick, NJ: Rutgers
Shostak, Arthur B. 1962. America’s Forgotten Labor Organization. Princeton, NJ: Industrial Rela-
tions Section, Princeton University.
Strauss, George. 1968. “Human Relations—1968 Style.” Industrial Relations 7(May):262–76.
_____. 1974. “Job Satisfaction, Motivation, and Job Redesign.” In Organizational Behavior
Research and Issues, edited by G. Strauss et al. Madison, WI: IRRA.
_____. 1977. “Union Government in the U.S.: Research Past and Future.” Industrial Relations
_____. 1991. “Union Democracy.” In The State of the Unions, edited by George Strauss, Daniel
Gallagher, and Jack Fiorito. Madison, WI: IRRA.
Sugrue, Thomas. 1996. The Origins of the Urban Crisis: Race and Inequality in Postwar Detroit.
Princeton, NJ: Princeton University Press.
Thompson Ramo Wooldridge (TRW) and Teamsters Union. 1961. 132 NLRB 995.
Thompson Ramo Wooldridge (TRW) v. NLRB. 1962. 50 LRRM 2759.
Tomlins, Christopher L. 1985. The State and the Unions: Labor Relations, Law, and the Organized
Labor Movement in America, 1880–1960. Cambridge, UK: Cambridge University Press.
Troy, Leo. 1958. “The Course of Company and Local Independent Unions.” Ph.D. dissertation,
_____. 1960. “Local Independent and National Unions: Competitive Labor Organizations.” Jour-
nal of Political Economy 68(October):487–506.
_____. 1961. “Local Independent Unions and the American Labor Movement.” Industrial and
Labor Relations Review 14(April):331–49.
Twentieth Century Fund. 1935. Labor and the Government. New York: McGraw-Hill.
Union Labor News Review. Various issues.
Verma, Anil, and Thomas A. Kochan. 1985. “The Growth and Nature of the Nonunion Sector
within a Firm.” In Challenges and Choices Facing American Labor, edited by Thomas A.
Kochan. Cambridge, MA: MIT Press.
Walton, Richard E. 1961. The Impact of the Professional Engineering Union. Boston: Graduate
School of Business Administration, Harvard University.
Walton Mfg. v. NLRB. 1951. 289 F. 2d 177 (5 CCA).
Waters, Elinor. 1954. “Unionization of Office Employees.” Journal of Business 27(October):
404 / SANFORD M. JACOBY