Discussion Post Unions & Minorities

tobasj

unions_and_social_inclusiveness.pdf

Home >Human Resource Management homework help >Discussion Post Unions & Minorities

Reprints and permission: http://www. sagepub.com/journalsPermissions.nav

DOI: 10.1177/0160449X11405340 http://lsj.sagepub.com

LSJ405340LSJ36310.1177/0160449X11405340LeymonLabor Studies Journal

1University of Oregon, Eugene, OR, USA

Corresponding Author: Ann Shirley Leymon, 1291 University of Oregon, Eugene, OR 97403, USA Email: [email protected]

Unions and Social Inclusiveness: A Comparison of Changes in Union Member Attitudes

Ann Shirley Leymon1

Abstract

Unions have a well-known history of social conservativism, especially regarding exclusion of women and racial minorities prior to the 1960s. Restricting membership in this way led to a smaller labor movement than was possible, reducing the potential strength of labor unions. Recent labor scholarship has hailed a union revitalization, with new progressive leadership and new organizing strategies, but the issue of growing social inclusiveness has become a secondary issue. One can observe a shift toward progressive inclusiveness in high-level union organization, but unless this shift is occurring in the membership, the impact on labor union strength will be minimal at best. This article uses longitudinal American National Election Survey data to determine if members of unions have developed increasingly favorable attitudes toward various races as well as other political categories of people over time, compared to nonmembers. Findings suggest that union members are growing increasingly inclusive at a faster rate than nonmembers.

Keywords

labor and labor movements, race/class/gender, political sociology, attitudes, explor- atory methods

Union scholarship has been in a resurgence during the last several years with special attention paid to union revitalization, sometimes called new unionism or social move- ment unionism (Bronfenbrenner and Hickey 2004; Clawson and Clawson 1999). One characteristic of this new unionism is a focus on social justice, more broadly defined

Leymon 389

than just workplace rights; it values expanded democracy and rights to social services (Fantasia and Voss 2004). In order to support these as legitimate union goals, union members must be able to identify their common economic position with other work- ers, regardless of race, gender, and citizenship status. Thus, we need more research on the inequalities that have plagued the unions for much of their existence. These divi- sions inhibit all workers in real and significant ways, but they especially hurt those who most need the protection and security that unions offer.

Historically, membership in most US labor unions was exclusive, especially before the civil rights era. In the earlier part of the twentieth century, many union constitutions restricted membership to white, English-speaking, US-born men. The extent of the racialized and gendered exclusion varied; some unions accepted black men to their membership, while others maintained highly discriminatory practices. Those practices prevented many workers from obtaining the economic gains and job security that unions offer. It also weakened the US labor movement by limiting the number of labor union members. In recent years, unions have ceased many of their overtly discrimina- tory practices, and membership of nonwhite, female, and immigrant workers has increased. However, have the attitudes of union membership shifted as well?

If unions are changing their policies while race and gender continue to divide union members, then the new policies will have little positive effect on minorities’ access to stable, well-paying union jobs. This article will examine whether changes in union inclusivity, observable in union image and union leadership, are reflected in the views of the membership, where they would have the most impact on society. If there has been a shift in the attitudes of union members toward various groups, this will indicate that the move toward inclusiveness in labor unions has occurred deeply in the labor movement rather than among leadership or image only. Racism and sexism have only served to divide unions, so developing more inclusive attitudes is an essential step in building solidarity of the working class, enabling workers to advocate more success- fully for better wages and working conditions. This would especially benefit workers from historically excluded groups.

A Brief History of Labor Unions and Diversity Race, ethnicity, and gender often influence the occupation and the wages of a worker (Reskin 2000). This often leads to divisions in the labor force, created and reproduced by both workers and employers. Competition, a central feature of capitalism, compels the working class to subdivide and create boundaries as a way of self-preservation (Silver 2003, 22). During the early twentieth century, some workers were willingly divided by race and gender as a way for certain groups to gain some economic security at the expense of other groups, rather than uniting against their employers (DeVault 2004; Hill 1977). Unions such as the International Longshoremen’s Association segre- gated black workers into their own local during the 1940s.1 Although they were union- ized, the white locals prevented them from gaining control over a pier and thus the black union members could not maintain steady work (Davis 2002).

390 Labor Studies Journal 36(3)

In the 1930s, the CIO worked to eliminate these divisions. They prohibited segregated locals, and their unions had more women in leadership than AFL unions (Stepan-Norris and Zeitlin 2003, 268; Zeitlin and Weyher 2001, 432-33). The AFL, however, looked the other way even when white workers committed crimes against black workers (Zeitlin and Weyher 2001). The CIO fought for contracts specifying equal pay for equal work, and as a result CIO-dominant areas had less black-white inequality than AFL-dominant areas. Once the CIO merged with AFL, progress on race relations in the labor unions slowed. Black representation was reduced in all levels of leadership, and many of the AFL-CIO leaders avoided civil rights issues altogether for fear of appearing sympathetic to communists (Stepan-Norris and Zeitlin 2003, 275).

After the AFL-CIO merger, some unions joined in the civil rights movement, but many continued to struggle with internal racial divisions. Leaders of the Detroit-based United Auto Workers (UAW) in the 1960s opposed racism, but chose to fight it by raising living standards to remove feelings of competition between white and black workers. Led by Reuther, these white liberal leaders opposed black self-organization in the labor movement, believing that it was a communist strategy to gain power (Lewis-Colman 2008, 46-47) and that it would be as detrimental to solidarity as white racism (Lewis-Colman 2008, 3). They focused on reform instead. Black autoworkers argued that racial divisions were exploited by the capitalist class and that the labor move- ment’s survival depended on the elimination of racism2 (Lewis-Colman 2008, 116). The Civil Rights Act forced the railroad industry to integrate, giving new opportunities for black men to gain entry into previously off-limits occupations and unions. White workers resisted through harassment and by not training black workers in their new responsi- bilities (Arnesen 2002, 231-32, 243).

Many union contracts had seniority clauses that effectually kept black workers in undesirable jobs even after the Civil Rights Act. When it became clear that preventing future discrimination would not be enough to undo decades of injustice, individuals sought affirmative action to secure opportunities for advancement for women and minorities. This meant union contracts with seniority would need to be altered to ensure equal access. Although some white members objected, unions such as the United Steelworkers cooperated to adjust contracts, allowing more direct channels for histori- cally disadvantaged groups to move into better jobs (Zieger 2007, 181). For building trade unions, like the United Brotherhood of Carpenters and Joiners of America (UBCJA), the battle to increase opportunity for minorities was intense. White union members and leaders resisted the removal of their traditional advantages; some boy- cotted projects at which black workers had been hired (Zieger 2007, 183-85). Although initially some white workers opposed affirmative action, eventually many labor unions came to support it. In 1985, when Reagan wanted to weaken affirmative action laws, several unions opposed his plan, including the Association of Federal, State, County, and Municipal Employees (AFSCME), Newspaper Guild, International Union of Electrical Workers, the AFL-CIO, and even the UBCJA (MacLean 2006, 309).

A broad look at the labor movement shows that race and gender relations have improved over time, with substantial variation among unions along the way. But the

Leymon 391

issue still needs attention. Chen and Wong (1998) point out that identity caucuses and constituent groups such as the Coalition of Black Trade Unionists (CBTU) and the Coalition of Union Women (CUW) experienced growth during the 1960s and 1970s. The growth of these groups, which serve to articulate and refine goals, train leaders, and present unified demands, was a sign that the needs of the people they represented were not being met through mainstream union activity (Chen and Wong 1998, 191). At the time of the split between Change to Win and the AFL-CIO, black leaders asked both federations to consider black employment issues, without success. And although representation of blacks and minorities in leadership positions has increased, those leaders have not succeeded at creating change in those positions or through coalitions (Zieger 2007, 224-28). Unions also overlook women as targets for organizing campaigns, even though research shows that women are more likely to vote for a union and successful organizing cases are concentrated in bargaining units where women are the majority, especially black women (Bronfenbrenner 2005; Cobble and Michal 2002). By failing to take aggressive action to ameliorate inequality in the unions, labor unions are diminishing their own success and effectiveness.

Labor Unions and Diversity Today Overall union density stands at 12.3 percent, but 15.4 percent of black men and 12.7 percent of black women were union members in 2009 (BLS 2010). Although in abso- lute numbers, white men are still the largest group of union members, the proportional growth of women and nonwhite men is a significant demographics shift for unions. A glance at the AFL-CIO and Change to Win websites shows that unions are working to be more inclusive, or at least working to have a more inclusive image. Both web- sites are offered in English and Spanish, and include sections on immigrant rights. Both federations support health care coverage for all. AFL-CIO has sections of the website dedicated to women workers, lesbian and gay workers, workers with disabili- ties, and workers of color, all of which are attempts to make unions more welcoming to groups traditionally discouraged from participating in the union movement.

Labor unions can express a commitment to diversity and equality in several differ- ent ways. UNITE HERE at their constitutional convention in 2009 created a commit- tee and an elected Vice President position to work on diversity issues (UNITE HERE 2009). The Service Employees International Union’s (SEIU) Code of Ethical Practices and Conflict of Interest Policy states that SEIU is “committed to pursuing justice for all, and in particular to bringing economic and social justice to those most exploited in our community” (SEIU 2009). The United Electrical (UE) workers have passed reso- lutions at their conventions to end discrimination based on sexual orientation, fight racism, advance women’s rights, encourage international solidarity, and stand up for the rights of immigrant workers (UE Union 2009). The Boilermakers include a work- shop on diversity in unions on their schedule of educational events (International Brotherhood of Boilermakers 2009). A commitment to diversity can come in the form of educating members, encouraging equal representation among leaders and staff of

392 Labor Studies Journal 36(3)

the organization, passing resolutions that inform policy agenda, assigning a union offi- cial the task of advancing equality in the organization, or promoting contracts that protect and advance rights of minorities.

Current Research on Inequality in Labor Unions Current research on diversity in unions has noted that race and gender continue to be salient identities for union members. During the California United Food and Commercial Workers (UFCW) strike of 2003, strikers racialized nonwhite strike- breakers (both customers and workers) but did not criticize or comment on the race of white strikebreakers, or the mostly white executives who were responsible for bringing in scabs (Wilson 2008). The Democratic primary for the 2008 presidential election offered liberal voters the choice of a black man (Barack Obama) or a white woman (Hillary Clinton), but it also offered researchers a unique opportunity to examine if and how race and sex prejudices affected union members’ votes. Overall, union voters preferred Clinton over Obama 41.6 percent to 32.8 percent, but white unionists preferred Clinton 52.5 percent to Obama’s 32.8 percent while 94.4 percent of black unionists chose Obama (Boris and Bruno 2010, 101). Additionally, women voters were more likely to prefer Clinton. Although union voters preferred Obama’s stance on most issues, Clinton received more support because they per- ceived her as a fighter willing to engage in conflict with Republicans to protect working-class interests (Boris and Bruno 2010). Studies of organizing immigrant workers reveal that few unions are exploring organizing possibilities in immigrant- dominated occupations or sectors, but those that are have found some success along with some defeats (Delgado 1993, 2000; Milkman and Wong 2006; Zabin 2000).

Unions Adjust to Needs of New Constituents Changing demographics have compelled labor unions to adjust their strategies for orga- nizing and bargaining. In order to serve immigrant workers, unions have hired staff who are multilingual and who represent a variety of nationalities and ethnicities. Unions have worked with community organizations that provide assistance to immigrants. They have removed citizenship requirements for holding office, enabling locals with large immi- grant membership to more freely choose their leadership representation (Wells 2000). Unions have also adapted bargaining styles to new constituents. The Harvard Union of Clerical and Technical Workers (HUCTW) is a female-dominated union that chose to take a less combative, more inclusive approach to bargaining. They focus on empower- ing and including workers in bargaining and in the workplace, preferring a bargaining style that utilized several small groups working out compromises on specific interests (Cobble 2004). In these cases, the unions understood what the workers needed and helped them to get those needs met. This is how unions should seek to appeal to immi- grant workers—by finding out what problems the workers face and working toward solutions, just as they should do with all workers (Sherman and Voss 2000).

Leymon 393

US Attitudes toward Diversity

Shuman, Steeh, and Bobo (1985) found that in principle, white attitudes toward equal treatment of blacks have grown increasingly favorable over time. However, whites are less supportive of the implementation of those principles (measures such as affirma- tive action, equal housing laws, schools, etc.). In addition to marked differences between abstract questions and concrete questions, their study showed that implemen- tation questions do not show the upward trend that abstract attitude questions show. For some of the specific measures, it is possible that a lag effect is present, but not for all (Shuman, Steeh, and Bobo 1985, 136-37).

Present Study Following the above research, I expect to find that over time, union members will grow more tolerant at a faster rate than nonmembers. Although many would argue that our society has become less racist over time or that people tend to conceal their racist attitudes now because it is not socially acceptable, if union members’ attitudes change at a faster rate than nonmembers, this trend shows something exceptional is happen- ing among union members.

According to Zieger, black unionists today tend to have more progressive attitudes toward illegal immigration than black nonunionists, opposing harsh sanctions and sup- porting paths to legal citizenship (Zieger 2007, 230). I argue that the organizational culture of unions can influence racial attitudes, yielding membership with noticeably different positions from the nonunion population. The question at hand is not whether union members were more or less racist than nonmembers in the past or future. Comparing the rate of change of these groups separately over time will indicate whether the changes in union policy as discussed above have impacted union members, result- ing in attitude patterns that are markedly divergent from the general population. Whether unions are explicitly excluding nonwhites and facilitating segregation at work through institutional and informal means, or educating and encouraging their members about worker solidarity, union policies could have the effect of legitimizing or illegitimizing racism and sexism among their members. Controlling for race, gender, and class vari- ables will indicate whether differences are due to the changing demographics of union members, particularly changes in the race and gender compositions, instead of their union status. My hypotheses are as follows:

Hypothesis 1a: Over time, union members will report increasingly favorable attitudes toward minority groups.

Hypothesis 1b: The rate at which union members’ attitudes toward minority groups becomes more favorable will be faster than that of nonmembers.

Hypothesis 2a: Over time, union members will report increasingly favorable attitudes toward people promoting rights of minority groups.

394 Labor Studies Journal 36(3)

Hypothesis 2b: The rate at which union members’ attitudes toward people pro- moting rights of historical minority groups will be faster than that of non- members.

The historically disadvantaged groups3 included in this analysis are blacks, Asian Americans, Chicanos/Hispanics, gays and lesbians, and illegal immigrants. People promoting rights of historical minorities are operationalized as civil rights leaders, feminists, women’s libbers, and black militants. When union members’ median thermometer ratings of these groups increase over time, hypotheses 1a and 2a will be supported. If the rate at which union members’ attitudes increase is faster than the rate at which nonmembers’ attitudes toward these group increase—in other words, if the slope of the line is steeper—then hypotheses 1b and 2b will be supported.

Next, to determine if these attitude changes have an impact on the behavior of union members, I examine whether attitudes are a predictor of voting choice. Because behavior measures are more reliable than attitude measures, comparing the two mea- sures could reinforce the attitude results. Union members vote Democrat 60.1 percent of the time, while nonmembers vote Democrat 44.8 percent of the time (ANES 2005), but do attitudes toward minority groups and the people who work to promote the rights of minority groups influence vote choice?

Hypothesis 3: Attitudes toward minority groups and toward people who pro- mote the rights of minority groups will predict voting behavior.

If these attitudes do not help explain vote choice, any findings that support increas- ing progressivity of union members will be less important. Changes in abstract atti- tudes without any accompanying change in behavior are essentially meaningless.

Methods Data for this analysis comes from the National Election Survey Data. This survey, which has been conducted biannually since 1948, collects data on respondent voting behavior, perceptions of various strengths and weaknesses of political candidates and parties, knowledge of public figures, and attitudes toward various groups in society (such as the elderly, environmentalists, radical students, and many others). The pri- mary dataset for this analysis combines each year that the survey was administered, omitting any questions that were asked fewer than three years (ANES 2005). The secondary dataset is the 1996 version of the study, which was the only year that included a series of questions on groups to which the respondent may belong, and how often any of those groups discussed politics (ANES 2001).

The main variables in this analysis are union membership,4 voting behavior,5 group membership,6 political discussions in groups,7 and thermometer ratings. Thermometers are a series of questions that ask respondents how warm or close they feel to a particu- lar group on a scale of 0 to 100, with 0 being completely cold or distant, 100 being

Leymon 395

completely warm or close, and 50 being neutral or having no information. I compare the union members against nonmembers in the following thermometer ratings to rep- resent social progressivity: Asian Americans, blacks, black militants, Chicanos and Hispanics, civil rights leaders, gays and lesbians, feminists, illegal aliens, whites, and women’s libbers.8 Shuman, Steeh, and Bobo affirmed the usefulness of thermometer questions, which deal with abstract principles, by comparing their correlations with relevant policy questions. For example, they found that individuals giving blacks higher thermometer ratings also reported favoring desegregation. However, respon- dents’ answers probably are driven by contemporary controversies rather than general feelings toward a group.

The lack of temporal change despite the relation of the question to other volatile issues suggests that people respond to these issues over time in terms of their policy content, but do not generalize them to influence overall feelings toward blacks. (Shuman, Steeh, and Bobo 1985, 121-22)

Data are analyzed using OLS regression including a variable for time and controls for demographics to determine statistical significance. Graphs are used to visually display changes in the union member attitudes over time.

When studying attitudes, the researcher presumes that a survey respondent’s response does not directly measure the attitude; rather, the attitude underlies the response. Surveys do not perfectly measure attitudes (Shuman Steeh, and Bobo 1985, 43-44). This is why it is important to corroborate analysis with data measur- ing behaviors. Most of this data focuses on principles—general attitudes without discussion on policy practices or individual practices. To make the analysis stronger, I conclude by examining how well those attitudes predict voting behavior in presi- dential elections. The value of this latter analysis is the internal comparison between changes in union member attitudes and voting behavior against changes in non- member attitudes and voting behavior. The analysis of voting behavior serves as a validity check on the attitudes research. Using the method of King, Tomz, and Wittenberg (2000) to estimate quantities of interest,9 I compare predicted likeli- hood of voting Democrat for union members and nonmembers at different values of thermometer ratings measuring attitude, using the same groups as are in the pri- mary analysis.

Results First, I used the 1996 ANES to see how labor union members compare to members of various other groups in time spent discussing politics. This evidence is necessary to determine if labor unions are a site where political activity and dialogue occurs. If not, any historical shift in union member attitudes is due to some factor other than the member’s participation in the labor movement. The graph depicted in Figure 1 shows that labor unions are quite politically engaged compared to other groups; only civic

396 Labor Studies Journal 36(3)

and political organizations had a higher percentage of members report political dis- cussions. This indicates that there is enough political activity occurring within labor unions to continue with the analysis.

Regressions on the thermometer ratings of interest10 show that union member atti- tudes toward Democrats, feminists, women’s libbers, black militants, and civil rights leaders were all statistically significant and positive even after controlling for race, gender, income, and education, indicating that union membership influenced per- ceptions of these groups. Relationships between union status and thermometers for Republicans were statistically significant and negative with the same controls, mean- ing union members rated Republicans consistently lower than nonmembers. Union membership was statistically insignificant for all other thermometers. The survey did not include each thermometer every time the survey was administered, so the number of time points is included in the following table. The forthcoming graphs show exactly when and how often each question was included in the survey.

The graph depicted in Figure 2 showing differences between union member and nonmember thermometer ratings of blacks provides additional information. The rela- tionship was statistically insignificant in the OLS regression analysis. The graph shows that union members rated blacks lower than nonmembers in earlier years of the study, with the gap closing over time. This indicates that the history of racism, at least explicit and transparent racism, of union members is passing.

The next graph in Figure 3 reinforces this finding by comparing attitudes toward feminists with attitudes toward women’s libbers, both of which were statistically sig- nificant and positive. For both categories, later years show a higher rating by union members compared to nonmembers. This indicates that union members are becoming more friendly toward women’s issues. It also shows that the OLS regression analysis

La bo

B us

in es

C hu

rc h

R el

ig io

P ol

iti ca

C iv

P ar

ty o

r C

an di

da te

E ld

er ly

P er

ce nt

o f R

es po

nd en

Figure 1.

397

T a b

le 1

. C

o m

pa ri

so n

o f U

ni o n

M em

be rs

’ a nd

N o nm

em be

rs ’ A

tt it ud

es t

o w

ar d

V ar

io us

G ro

up s,

an d

D es

cr ip

ti ve

S ta

ti st

ic s

C o ns

ta nt

U ni

o na

T im

e R

ac eb

G en

de rc

Ed uc

at io

nd In

co m

ee O

bs er

va ti o ns

ea r)

M ea

n (S

D )

D em

o cr

at s

75 9

4. 32

** (

0. 53

) -.

34 **

( .0

3) -1

4. 2*

* (. 59

) -3

.1 **

( .3

7) -4

.8 **

( .4

0) -1

.8 **

( .1

6) 12

,7 20

( 9)

66 .1

5 (2

0. 87

) R

ep ub

lic an

s 61

9 -4

.4 3*

* (0

.5 5)

-. 29

** (

.0 3)

13 .7

** (

.6 3)

-1 .3

9* *

(. 39

) 0.

19 (

.4 2)

0. 38

* (. 17

) 12

,6 81

( 9)

59 .8

7 (2

1. 30

) B

la ck

s -3

3 -0

.0 7

(0 .3

7) .0

6* *

(. 01

) -2

2. 2*

* (. 37

) -4

.1 4*

* (. 25

) 3.

55 **

( .2

6) -0

.0 3

(. 11

) 26

,0 58

( 20

) 64

.6 5

(2 0.

76 )

W hi

te s

56 7

-0 .1

6 (0

.3 6)

-. 25

** (

.0 1)

6. 1*

* (. 38

) -2

.7 3*

* (. 24

) -3

.4 4*

* (. 26

) -1

.0 7*

* (. 11

) 22

,9 54

( 18

) 74

.6 7

(1 8.

49 )

A si

A m

er ic

an s

-1 ,1

27 -0

.0 4

(0 .9

8) .6

0* *

(. 06

) -1

.7 (

.8 8)

-1 .8

0* *

(. 60

) 4.

91 **

( .6

5) 0.

45 (

.2 8)

3, 56

8 (4

) 62

.9 9

(1 8.

74 )

W o m

en ’s

lib be

rs -1

,3 86

2. 98

** (

0. 56

) .7

3* *

(. 02

) -1

3. 5*

* (. 58

) -2

.4 5*

* (. 38

) 2.

66 **

( .4

0) -0

.7 5*

* (. 18

) 17

,4 71

( 12

) 56

.3 2

(2 5.

59 )

Fe m

in is

ts -3

07 5.

64 **

( 0.

99 )

.1 9*

* (. 05

) -8

.2 **

( .9

4) -2

.6 6*

* (. 63

) 0.

89 (

.6 6)

-0 .5

9* (

.3 0)

4, 87

3 (5

) 53

.7 6

(2 1.

78 )

C hi

ca no

s an

d H

is pa

ni cs

-5 25

0. 53

( 0.

54 )

.3 0*

* (. 02

) -7

.9 **

( .5

5) -2

.3 5*

* (. 36

) 4.

78 **

( .3

7) 0.

56 **

( .1

7) 11

,3 38

( 10

) 60

.5 7

(1 9.

63 )

C iv

il ri

gh ts

-1 ,3

07 3.

01 **

( 0.

70 )

.7 0*

* (. 04

) -3

2. 5*

* (. 76

) -2

.9 **

( .4

8) 9.

42 **

( .5

1) 0.

58 **

( .2

3) 10

,2 07

( 7)

48 .8

6 (2

6. 33

) B

la ck

ili ta

nt s

-1 ,9

00 1.

20 (

0. 64

) .9

9* *

(. 04

) -2

3. 9*

* (. 71

) -2

.2 8*

* (. 44

) 3.

12 **

( .4

7) -1

.1 6*

* (. 21

) 10

,4 25

( 7)

22 .9

5 (2

4. 01

)

G ay

s an

d le

sb ia

ns -1

,9 23

1. 21

( 0.

82 )

.9 8*

* (. 05

) -4

.4 **

( .7

9) -8

.4 9*

* (. 53

) 10

.7 7*

* (5

6) 1.

21 **

( .2

5) 10

,0 88

( 9)

38 .8

0 (2

7. 82

)

Ill eg

al a

lie ns

-4 33

0. 44

( 1.

00 )

.2 4*

* (. 06

) -1

1. 5*

* (. 95

) -3

.3 **

( .6

5) 4.

86 **

( .6

7) -0

.1 7

(. 30

) 5,

32 2

(4 )

35 .8

8 (2

4. 34

)

N o te

: S ta

nd ar

d er

ro r

in p

ar en

th es

es .

a 0 =

n o nu

ni o n,

1 =

u ni

o n

b 0 =

n o nw

hi te

, 1 =

w hi

te c 0

= fe

m al

e, 1

= m

al e

d 0 =

h ig

h sc

ho o l o

r le

ss , 1

= m

o re

t ha

n hi

gh s

ch o o l

e 1 =

lo w

in co

m e,

5 =

h ig

h in

co m

e *

si gn

ifi ca

nt a

t 5%

; * *

si gn

ifi ca

nt a

t 1%

. +N

o te

: T he

q ue

st io

n as

ke d

w as

: O n

a sc

al e

o f 0

-1 00

, w ith

1 00

b ei

ng e

xt re

m el

y cl

o se

a nd

w ar

m a

nd 0

b ei

ng e

xt re

m el

y di

st an

t an

d co

ld , h

o w

d o y

o u

fe el

a bo

ut _

__ _?

So ur

ce : T

he A

m er

ic an

N at

io na

l E le

ct io

n St

ud ie

s (2

00 5)

398 Labor Studies Journal 36(3)

M ea

n A

tt itu

de S

co re

o n

S ca

le o

f 0 to

1 00

1960 1970 1980 1990 2000 Year of Study

Union Members toward Blacks Nonmembers toward Blacks

Figure 2.

M ea

n A

tt itu

de S

co re

o n

S ca

le o

f 0 to

1 00

1960 1970 1980 1990 2000 Year of Study

Union Members toward Feminists Nonmembers toward Feminists Union Members toward Women's Libbers Nonmembers toward Women's Libbers

Figure 3.

obscures these trends. The feminist regression has a higher level of significance because in all years of this analysis, union members rated feminists roughly four points higher. The women’s libbers regression had a lower level of significance and a smaller coefficient because although union members rated women’s libbers roughly four

Leymon 399

M ea

n A

tt itu

de S

co re

o n

S ca

le o

f 0 to

1 00

1960 1970 1980 1990 2000 Year of Study

Union Members toward Asian Americans Nonmembers toward Asian Americans

Figure 4.

points higher than nonmembers in later years, there was little difference between union members and nonmembers in earlier years of the analysis. Because this research is focused on the trend, I argue that the graph indicates that union members are becom- ing more friendly to women’s issues even though the regression only weakly supports that assertion.

The graphs in Figures 4 and 5 also provide support for the hypothesis. They both show union members’ attitudes growing more favorable at a faster rate than nonmem- bers’ attitudes. In each case the union members rate the group lower in early years, then increase over time until they surpass the nonmembers’ average rating of the group. The low number of years does weaken the impact of this graph, but also could explain why neither Asian Americans nor illegal aliens showed significance in the statistical analysis. With a much smaller N, it is difficult to achieve significant results in spite of the clear graphical trend.

The graph shown in Figure 6 is ambiguous. The overall trend clearly shows increas- ingly positive ratings of gays and lesbians, but differences between union members and nonmembers vary considerably. Union members and nonmembers rated gays and les- bians similar in most years of the study, though there is evidence that union members are rating gays and lesbians higher in later years of the study. In the first four years, union members gave gays and lesbians mostly lower or equal ratings, while in the last three years union members gave gays and lesbians higher or equal ratings, compared to nonmembers.11

Graphs for the remaining four thermometers (whites, Chicanos, civil rights leaders, and black militants) are not included here because the thermometer trends

400 Labor Studies Journal 36(3)

M ea

n A

tt itu

de S

co re

o n

S ca

le o

f 0 to

1 00

1960 1970 1980 1990 2000 Year of Study

Union Members toward Illegal Aliens Nonmembers toward Illegal Aliens

Figure 5.

M ea

n A

tt itu

de S

co re

o n

S ca

le o

f 0 to

1 00

1960 1970 1980 1990 2000

Year of Study

Union Members toward Gays and Lesbians Nonmembers toward Gays and Lesbians

Figure 6.

do not clearly differ between union members and nonmembers over time, though the latter two of those groups were statistically significant in the OLS regression analysis.

Leymon 401

In order to add more impact to this analysis, I conclude by analyzing how well the thermometer ratings predict self-reported voting behavior. The conceptual nature of thermometer ratings can make it difficult to decipher the meaning of their variations. Voting behavior, however, is more concrete. If thermometer ratings strongly predict voting behavior, it adds import to analysis of thermometer ratings. Regression analysis shows that both union membership and all of the selected thermometer ratings are statistically significant predictors of voting behavior, even controlling for income, race, gender, rural/urban, and south/nonsouth.12 By displaying simulated quantities of interest, it is clear how union members and nonmembers differ in their voting behav- ior, and which attitudes have the biggest political impact.

This table shows that as attitudes toward these social groups become more positive, the likelihood of voting Democrat increases. It also shows that union members are much more likely to vote Democrat than nonmembers who hold similar attitudes toward these social groups. For example, comparing union members and nonmembers who feel negatively toward black militants (simulating a thermometer rating of 25), a union member is predicted to vote Democrat 54.3 percent while a nonmember is pre- dicted to vote Democrat 38.8 percent. Additionally, nonmembers who feel favorably toward illegal aliens (simulating a thermometer rating of 75) are only about five per- centage points more likely to vote Democrat than union members who feel unfavor- ably toward illegal aliens. In this table, union members are predicted to vote non-Democrat when they feel unfavorably toward feminists, civil rights leaders, Democrats, women’s libbers, blacks, or Chicanos and Hispanics, or neutral toward Democrats. In all other instances they are predicted to vote Democrat. Conversely, nonmembers are more likely to vote Democrat when they feel favorably toward black militants, gays and lesbians, feminists, illegal aliens, civil rights leaders, Democrats, women’s libbers, or blacks, or neutral toward black militants, gays and lesbians, ille- gal aliens or Republicans, or negative toward Republicans. This table shows which prejudices union members retain that trump class and economic needs in voting decisions.

Discussion and Conclusions Statistical findings indicate that union member attitudes toward Democrats, Republicans, feminists, women’s libbers, civil rights leaders, and black militants were significantly different from nonmembers over time, though sometimes only slightly. Additionally, exploratory graphs indicate that union member attitudes toward Asians, illegal aliens, and possibly gays and lesbians were also growing more positive at a faster rate than nonmembers. The finding that attitudes toward blacks, whites, and Chicanos/Hispanics were not significant for union members could mean people iden- tify more with the values held by feminists, women’s libbers, civil rights leaders, and black militants; the titles of these groups indicate voluntary membership and a united goal, compared to the racial-ethnic groups.

402 Labor Studies Journal 36(3)

Alternatively, these findings combined with the findings from Table 2 could indi- cate which prejudices are “mainstream” prejudices versus extreme prejudices. For black militants, gays and lesbians, and illegal aliens, union members reporting nega- tive attitudes were predicted to vote Democrat, while nonmembers reporting negative attitudes were predicted to vote non-Democrat. Although union voters admitted to negative attitudes, they were less attached to those particular prejudices at the polls. Negative attitudes toward feminists, civil rights leaders, blacks, and women’s libbers could be classified as more extreme, because union members with negative attitudes toward those groups were predicted to vote non-Democrat, as were nonmembers with neutral attitudes toward those groups. Nonmembers who had positive attitudes toward those groups, however, were predicted to vote Democrat. These attitudes strongly influenced vote behavior, indicating more extreme commitment to these prejudices.

Since the evidence also suggests that attitudes influence voting behavior, along with union membership, increasing inclusive attitudes by union members could have a real political outcome, in addition to improving union strength. However, there are problems with the data that raise questions for the conclusions. The biggest limitation of the analysis is that we do not know which union any of these union members belonged to. Thus, we cannot determine if the shift is causally related to increasing political involvement and education by their union. Also, most of the thermometer questions were not asked during the entire life of the survey. Because the relevant

Table 2. Estimated Percentage to Vote Democrat Using Regression Simulations

75 50 25

Union Nonunion Union Nonunion Union Nonunion

Black militants** 78.5 66.1 67.6 52.7 54.3 38.8 Gays and lesbians** 76.7 63.3 66.7 51.1 54.8 38.8 Feminists** 73.7 65.1 53.2 42.9 31.5 23.3 Illegal aliens** 73.2 60.3 64.7 50.6 55.2 40.7 Civil rights leaders** 69.8 55.3 54.5 39.0 38.3 24.9 Democrats** 68.8 57 31.6 21.7 8.8 5.5 Women’s libbers** 68.1 55.3 52.1 38.7 35.6 24.3 Blacks** 65.7 50.2 53.8 38.0 41.5 27.2 Chicanos/Latinos** 65.7 49.7 58.0 41.6 49.8 33.9 Asians** 61.9 49.0 58.7 45.7 55.4 42.4 Whites 60.1 45.4 60.8 46.1 61.5 46.8 Republicans** 38.0 28.9 68.2 58.8 88.2 83.3

**Regression coefficient is significant at <.01 Note: This graph compares the likelihood of voting democrat for union members and nonmembers, based on attitudes toward various groups of people. For each group, this graph shows the prediction based on warm (75), neutral (50), and cold (25) thermometer ratings. Source: The American National Election Studies (2005).

Leymon 403

questions do not start until later years, the time period of comparison is more limited than is ideal for statistical analysis. In spite of this, the regression analysis revealed strong results for some of the measures. Combined with the graphical evidence, this analysis indicates that an increasing level of racial, ethnic, and gender inclusion among union members is present, and the relationship between union status and changing social attitudes is worth evaluating.

Another alternative explanation for the change in attitudes is the demographic shift that has occurred in labor unions over the last fifty years. Membership of white men is declining proportionally, while membership of women and nonwhites is increasing. Immigrants are also joining unions in larger numbers. Though there is of course much diversity in political beliefs within each of these groups, historically discriminated- against groups are more likely to have inclusive attitudes on average (Boris 2010). Perhaps it is related to the overall decline in union density; maybe the most conserva- tive members have left the movement while the most progressive ones remain. Or along the same lines, the older union members who held less socially progressive attitudes may have retired, while younger, more progressive workers joined the movement. A combination of these factors is most likely. All of these explanations have intellec- tual merit and deserve more attention from those who study labor movements.

If union member attitudes toward the groups in this research continue to trend upward, it could mean that the labor movement will continue to grow more inclusive and cohesive and thus be a stronger force for improvements in working conditions. On the other hand, if union diversity continues to develop and multiply, this could present an extra challenge for unions and their organizers, requiring adjustments in organizing strategy, bargaining style, and organizational characteristics. If unions cannot meet the challenge, density could continue to fall, possibly getting to levels low enough to ren- der the labor movement ineffective. On the other hand, the challenge could force unions to innovate, continually designing new methods and creatively figuring out how to meet more workers’ needs.

The finding that union members tend to vote Democrat more often than nonmem- bers with similar attitudes on social measures indicates that when union members vote, other issues trump conservative social attitudes. This suggests that an increase in union density could have a valid political impact if this trend held for new union mem- bers. Coordinated efforts in electoral campaigns among labor unions or between labor unions and other organizations could overcome the gap in a close race, or could pos- sibly even accomplish widespread support for a third party in the United States. However, it is possible that this effect could be influenced by other factors, such as the campaign efforts labor unions conduct. More research should be done to test this effect, including testing other groups such as the National Rifle Association or other conservative groups and in-depth studies to determine the mechanism that causes union members to vote Democrat in spite of having conservative social views.

If union members’ attitudes, values, and belief systems have changed over the years, and the unions have failed to revise their organizational structure and values, this could be a contributing factor to the decline in union density. While the changing

404 Labor Studies Journal 36(3)

structure of the economy, the union-hostile legal and political climate, and the well- documented employer offensive clearly take much of the blame for the decline, all potential reasons must be considered if the unions will ever be able to reverse the trend. Still, many scholars are enthusiastic about the possibility of unions for broad change. “No force in our society has more democratic potential (or radical possibility) than the labor movement. Its base and focus is the large working-class majority under- represented, or outright neglected, by many other social movements” (Clawson 2003, 196). Scholars must continue to research the labor movement, illuminating its flaws and possibilities for its improvement, to help it achieve this potential.

Acknowledgments

I wish to thank Caleb Southworth, Seth Crawford, Angie Beeman, Mark G. Harmon, and three anonymous reviewers for helpful comments and advice on earlier drafts of this article.

Declaration of Conflicting Interests

The author declared no potential conflicts of interests with respect to the authorship and/or publication of this article.

Funding

The author received no financial support for the research and/or authorship of this article.

Notes

1. Other segregated unions include International Brotherhood of Boilermakers, International Association of Machinists, Seafarer’s International Union, United Association of Plumbers and Pipefitters, United Brotherhood of Carpenters, International Brotherhood of Electri- cal Workers, Teamsters, Brotherhood of Locomotive Engineers, Brotherhood of Railroad Trainmen, Brotherhood of Locomotive Firemen and Enginemen, and Order of Railroad Conductors.

2. A genuine tension exists between race-based and class-based organizing. See Lewis-Colman (2008) for a nuanced discussion of this issue through the case study of UAW in Detroit dur- ing the mid-twentieth century.

3. Choice of groups was limited by data availability. These measures are discussed in more detail in the Method section.

4. The question asked was: “Are you or another member of this household a member of a labor union?” I recoded this variable 1 if the respondent indicated that they were a union member and 0 if they indicated that they were not a union member.

5. This question was: “How did you vote in the most recent presidential election?” and only included respondents who reported that they remembered voting in the last presidential election. I coded this response 1 if the respondent voted for the Democratic candidate and 0 if the respondent voted for any other candidate.

6. This question was included only in 1996 and asked respondents about different groups they belonged to, including labor, business, religious, political, and many others.

Leymon 405

7. This question, also only included in 1996, asked respondents whether or not they discuss poli- tics in the groups that they had just reported belonging to. I coded these as a percentage; the number of people who affirmed discussing politics in a given group divided by the total number of people who affirmed belonging to that group. The relevant graph includes only those groups with high levels of political participation or groups that are explicitly political in nature.

8. Terminology in the article is the same as that which was used in the ANES survey. 9. Estimated quantities of interest is a technique that allows the researcher to insert values into

the regression equation to see what the outcome would be; for example, what would the voting outcome look like for someone who was a union member and had positive ratings for Asian Americans? King, Tomz, and Wittenberg (2000) include an adjustment for standard errors.

10. Because the thermometer ratings are bounded between 0 and 97, I ran tests to ensure that the model did not predict scores higher or lower than those values to ensure accurate estimation. Analysis indicates that results fall between 0 and 97.

11. I plotted the fitted regression line and confirmed that although it was not statistically signifi- cant, the slope of the union member was steeper than the slope of the nonmember line.

12. These regression tables are available from the author by request.

References

The American National Election Studies. 2001. The 1996 National Election Study [MRDF]. Ann Arbor: University of Michigan (producer and distributor). http://www.electionstudies.org.

The American National Election Studies. 2005. The 1948-2004 ANES Cumulative Data File [MRDF]. Stanford University and the University of Michigan (producers and distributors). http://www.electionstudies.org.

Arnesen, Eric. 2002. Brotherhoods of Color: Black Railroad Workers and the Struggle for Equality. Cambridge MA: Harvard University Press.

Boris, Monica Bielski. 2010. “Fighting For Equal Treatment: How the UAW Won Domestic Part- nership Benefits and Discrimination Protection for Lesbian, Gay, and Bisexual Members.” Labor Studies Journal 35 (2): 157-80.

Boris, Monica Bielski, and Robert Bruno. 2010. “‘Not Ready to Make Nice’: The Politics of Identity and Why Union Voters Wanted a Class Champion in 2008.” Labor Studies Journal 35 (1): 94-115.

Bronfenbrenner, Kate. 2005. “Organizing Women: The Nature and Process of Union-Organizing Efforts among US Women Workers Since the mid-1990s.” Work and Occupations 32: 441-63.

Bronfenbrenner, Kate, and Robert Hickey. 2004. Changing to Organize: A National Assessment of Union Strategies. In Rebuilding Labor: Organizing and Organizers in the New Union Movement, ed. Ruth Milkman and Kim Voss, 1-17. Ithaca, NY: Cornell University Press.

Bureau of Labor Statistics. 2010. “News Release: Union Members—2009.” Washington, DC: US Department of Labor. http://www.bls.gov/news.release/union2.t01.htm.

Chen, May, and Kent Wong. 1998. “The Challenge of Diversity and Inclusion in the AFL-CIO.” In A New Labor Movement for the New Century, ed. Gregory Mantsios, 185-201. New York: Monthly Review Press.

406 Labor Studies Journal 36(3)

Clawson, Dan. 2003. The Next Upsurge: Labor and the New Social Movements. Ithaca, NY: Cornell University Press.

Clawson, Dan, and Mary Ann Clawson. 1999. What Has Happened to the US Labor Movement? Union Decline and Renewal. Annual Review of Sociology 25:95-119.

Cobble, Dorothy. 2004. The Other Women’s Movement: Workplace Justice and Social Rights in Modern America. Princeton, NJ: Princeton University Press.

Cobble, Dorothy Sue, and Monica Bielski Michal. 2002. “On the Edge of Equality? Working Women and the US Labour Movement.” In Gender, Diversity, and Trade Unions: Interna- tional Perspectives, ed. Fiona Colgan and Sue Ledwith, 232-56. London: Routledge.

Davis, Colin J. 2002. “‘Shape or Fight?’: New York’s Black Longshoremen, 1945-1961.” Inter- national Labor and Working-Class History 62:143-63.

Delgado, Hector. 1993. New Immigrants, Old Unions: Organizing Undocumented Workers in Los Angeles. Philadelphia: Temple University Press.

Delgado, Hector L. 2000. “The Los Angeles Manufacturing Project: An Opportunity Squan- dered?” In Organizing Immigrants: The Challenge for Unions in Contemporary California, ed. Ruth Milkman, 225-38. Ithaca, NY: Cornell University Press.

DeVault, Ileen A. 2004. United Apart: Gender and the Rise of Craft Unionism. Ithaca, NY: Cornell University Press.

Fantasia, Rick, and Kim Voss. 2004. Hard Work: Remaking the American Labor Movement. Berkeley: University of California Press.

Hill, Herbert. 1977. Black Labor and the American Legal System. Washington, DC: Bureau of National Affairs.

International Brotherhood of Boilermakers. 2009. Education and Training Services Center. Kansas City, KS: International Brotherhood of Boilermakers. http://www.boilermakers.org/ structure/departments/education_training/event_calendar.

King, Gary, Michael Tomz, and Jason Wittenberg. 2000. “Making the Most of Statistical Anal- yses: Improving Interpretation and Presentation.” American Journal of Political Science 44 (2): 341-55.

Lewis-Colman, David M. 2008. Race against Liberalism: Black Workers and the UAW in Detroit. Urbana: University of Illinois Press.

MacLean, Nancy. 2006. Freedom is Not Enough: The Opening of the American Workplace. New York: Russell Sage Foundation.

Milkman, Ruth, and Kent Wong. 2006. “Si, Se Puede: Union Organizing Strategies and Immi- grant Workers.” In L.A. Story: Immigrant Workers and the Future of the US Labor Move- ment, by Ruth Milkman, 145-86. New York: Russell Sage Foundation.

Reskin, Barbara F. 2000. “The Proximate Causes of Employment Discrimination.” Contempo- rary Sociology 29 (2): 319-28.

SEIU. 2009. SEIU Code of Ethical Practices and Conflict of Interest Policy. Washington, DC: Service Employees International Union. http://www.seiu.org/a/ourunion/seiu-code-of-ethical- practices.php.

Sherman, Rachel, and Kim Voss. 2000. “‘Organize or Die’: Labor’s New Tactics and Immi- grant Workers.” In Organizing Immigrants: The Challenge for Unions in Contemporary California, ed. Ruth Milkman, 81-08. Ithaca and London: Cornell University Press.

Leymon 407

Shuman, Howard, Charlotte Steeh, and Lawrence Bobo. 1985. Racial Attitudes in America: Trends and Interpretations. Cambridge, MA: Harvard University Press.

Silver, Beverly. 2003. Forces of Labor: Workers’ Movements and Globalization since 1870. New York: Cambridge University Press.

Stepan-Norris, Judith, and Maurice Zeitlin. 2003. Left Out: Reds and America’s Industrial Unions. Cambridge: Cambridge University Press.

UNITE HERE. 2009. Unite Here’s 2009 Convention. New York: UNITE HERE. http://www .unitehere.org/convention_2009/.

UE Union. 2009. UE Policy 2009-2011. Pittsburgh, PA: United Electrical, Radio, and Machine Workers of America. http://www.ueunion.org/policy_tc.html.

Wells, Miriam J. 2000. “Immigration and Unionization in the San Francisco Hotel Industry.” In Organizing Immigrants: The Challenge for Unions in Contemporary California, ed. Ruth Milkman, 109-29. Ithaca, NY: Cornell University Press.

Wilson, Jake B. 2008. “The Racialized Picket Line: White Workers and Racism in the Southern California Supermarket Strike.” Critical Sociology 34:349-67.

Zabin, Carol. 2000. “Organizing Latino Workers in the Los Angeles Manufacturing Sector: The Case of American Racing Equipment.” In Organizing Immigrants: The Challenge for Unions in Contemporary California, ed. Ruth Milkman, 150-68. Ithaca, NY: Cornell University Press.

Zeitlin, Maurice, and L. Frank Weyher. 2001. “‘Black and White, Unite and Fight’: Interracial Working-Class Solidarity and Racial Employment Equality.” American Journal of Sociol- ogy 107 (2): 430-67.

Zieger, Robert H. 2007. For Jobs and Freedom: Race and Labor in America since 1865. Lexington: University Press of Kentucky.

Bio

Ann Shirley Leymon is a PhD candidate in sociology at the University of Oregon. Her research focuses on two main strands: inequality within the labor movement and labor markets, and an organizations-centered approach to labor union behavior. She is currently completing her dis- sertation examining union political behavior in the context of economic crisis.

or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.

However, users may print, download, or email articles for individual use.