An Empirical Analysis of the Public’s Attitudes toward Labor Unions

Gordon L. Freeman, Jr., H. Ronald Moser
Middle Tennessee State University

INTRODUCTION
     Using survey data on 669 participants, the researchers refined and validated the Attitudes toward Labor Unions Scale (ATLUS) with exploratory factor analysis.  To measure the participants’ attitudes toward labor unions, the researchers derived factor scores from 10 items on the survey.  These scores were divided into high, median, and low ATLUS scores groups.  People with high ATLUS scores tended to view labor unions more positively.  The main objective of this study was to determine which demographic characteristics accounted for any significant differences in attitudes toward labor unions.  Survey data were obtained from seven metropolitan statistical areas in the state of Tennessee.  This study found that the public has a mixture of attitudes regarding labor unions.  The public strongly agrees that companies should be willing to pay their employees more when business conditions are good and that workers should be allowed to join unions and strongly disagrees that workers be required to join a union, be required to pay the equivalent of union dues to be able to keep their jobs, or form their own political party.  The study found very little difference between the attitudes of male and female workers toward labor unions.  However, a substantial difference between the attitudes of white and nonwhite workers and between more-educated and less-educated groups was reported.  The paper also notes that the percentage of U.S. workers belonging to a union is in a steady decline.
General Overview
     Labor unions are facing some stern opposition not only from the general public but also from their own members on questions of union reform such as (1) whether workers should be free to choose or decline union membership, (2) whether union dues are used for political purposes, and (3) whether there should be extensive financial disclosure of union expenditures. According to Denholm (2004), these issues could dramatically affect unions’ influence and future membership. Unfortunately, labor unions are devoting much of their energy to fighting unions’ progress.
     According to information recently released by the Federal Bureau of Labor Statistics (BLS), 2006 was not a good year for AFL-CIO union membership-organizing efforts. The percentage of U.S. workers belonging to a union decreased from 13.3% in 2002 to 12.0% in 2006. In 2007 the BLS reported a slight increase to 12.1%.  However, on a technical note, the BLS states, “Union membership data for 2007 are not strictly comparable with data for 2006 and earlier years because of the introduction of updated population controls with the release of January data.  The effect of the revised population controls on the union membership estimates is unknown.”  The union membership rate has steadily declined from a high of 20.1% in 1983, the first year for which comparable union data are available. With the exception of the non-comparable results of 2007, the decline has been continuous despite the fact that in recent years AFL-CIO unions have been spending more than $40 million a year on worker-organizing activities. The decline has been the subject of numerous scholarly studies analyzing the causes and offering possible solutions. None of the published research has identified a single cause; but most writers agree that the reasons can be grouped into the following five categories: (1) increased individualism of workers and their attitude toward unions, (2) legal support for unions and existence of right-to-work laws, (3) macroeconomic factors, (4) employer opposition initiatives, and (5) labor unions’ opposition to representation elections (Atleson, 1994; Jarley and Kuruvilla, 1994; Karier, 1991; Leap, 1990; Margalioth, 1998; Saporta and Lincoln, 1995). However, a 2004 nationwide Zogby poll shows that 63% of the respondents answered that, in general, they approve of labor unions.
PURPOSE OF THIS STUDY
     The purposes of this study were to investigate the public’s attitudes toward labor unions and determine whether factors such as age, gender, race, total family income, and education accounted for any significant differences in attitudes toward labor unions.  Specifically, the researchers are testing the following hypothesis:
     Hypothesis: There are no differences in the attitudes of the public toward labor unions based on: (a) age, (b) gender, (c) race, (d) total family income, or (e) education.
METHODOLOGY OF THIS STUDY
Data Collection
     The data for this study were obtained through mail questionnaires sent to a random sample of 6,000 residents from telephone books from the seven metropolitan statistical areas in Tennessee: Memphis, Nashville, Clarksville, Chattanooga, Knoxville, Jackson, and Tri-Cities (Bristol, Kingsport, and Johnson City).  Appropriate numbers from each city were drawn according to the ratio of each city’s population to the total population of all seven urban areas. The research instrument was mailed to these residents, and 669 usable questionnaires were received and used in this study.  The questionnaire was divided into two sections.  The first section concerned demographic characteristics of the respondents.  Respondents were asked to select one of the following six categories regarding their age: (1) 18 – 25, (2) 26 – 35, (3) 36 – 45, (4) 46 – 55, (5) 56 – 64, (6) 65 or older.  For sex, female was coded as 1 and male as 2.  Race had three categories: (1) white, (2) black, and (3) other minority.  Total family income had the following categories: (1) $15,000 or less, (2) $15,001 – $30,000, (3) $30,001– $45,000, (4) $45,001 – $60,000, and (5) $60,001 or more.  Education was determined by the following choices: (1) less than high school graduate, (2) high school graduate, (3) some college, (4) college graduate, and (5) advanced degree.  The second section of the questionnaire included 18 statements designed to assess the respondents’ general attitude toward labor unions. These 18 questions were selected from questions included in various opinion polls conducted over the past 30 years.
Demographic Variables
     Among the 669 participants, 143 were female and 468 were male (missing data = 58). Regarding age, 32 (4.8%) were between 18 and 25; 176 (26.3%) were between 26 and 35; 158 (23.6%) were between 36 and 45; 123 (18.4%) were between 46 and 55; 91 (13.6%) were between 56 and 64; 84 (12.6%) were over 65 or older; and 5 (0.7%) did not specify age.  In terms of race, most respondents were white (82.5%).  Regarding total family income, 104 (15.5%) earned $15,000 or less, 285 (42.6%) earned $15,001 – $30,000, 144 (21.5%) earned $30,001 – $45,000, 50 (7.5%) earned $45,001 – $60,000, 59 (8.8%) earned $60,001 or more; and 27 (4.0%) did not specify family income.  In terms of education, 36 (5.4%) had less than high school graduate; 121 (18.1%) were high school graduates, 201 (30.0%) had some college, 200 (29.9%) were college graduates, 103 (15.4%) had an advanced degree, and 8 (1.2%) did not specify level of education.  Participants in the study were mostly male, white, and with low to middle incomes.  Slightly more than half of the study’s participants had less than a college degree and were 45 or younger.  Thereby, participants in this sample did not match the population perfectly.  However, there was no reason to believe that this sample was atypical.
Analysis
     The data obtained from the 669 respondents via the research instrument were first analyzed by tabulating the frequency percentages for each item on the questionnaire.  These percentages, presented in Table 1, illustrate the distribution of the public’s responses to the 18 statements in the questionnaire about labor unions.  The responses to these statements indicate the public has both favorable and unfavorable attitudes toward labor unions.  Most of the respondents (82.1%) agree that workers should be allowed to join unions, and well over half (56.2%) of the respondents agree that workers should be allowed to strike.  Almost half (46.8%) of the respondents agree that the gains workers have made in this country are chiefly due to labor unions.  In contrast, 39.8% of the respondents agree that workers do not need unions to protect their interests.  Most of the respondents (77.1%) agree that labor unions should be closely regulated, and almost half (48%) of the respondents agree that present laws that regulate unions are not strict enough.  In order to gain a general measure of the public’s attitudes toward labor unions that was consistent among as many of the 18 statements as possible, the researchers used exploratory factor analysis to refine and validate the Attitudes toward Labor Unions Scale (ATLUS).  This scale was developed in such a manner that people with high ATLUS scores tended to view labor unions more positively.  The main objective of this study was to determine which demographic characteristics, if any, accounted for significant differences in attitudes toward labor unions.
Exploratory Factor Analysis
     The survey contained 18 statements describing various attitudes about labor unions.  The responses to these 18 statements were coded so a higher value reflected a more positive attitude toward labor unions.  A test for internal reliability was performed using SPSS.  This test, applied to the responses to the 18 statements, revealed a value of 0.846 for Cronbach’s alpha.  However, for statement 14, “Even without a union, companies should be willing to pay better wages and salaries when business conditions are good,” 14 of the inter-item correlations were negative.  By reversing the coding of the responses to statement 14, the value of Cronbach’s alpha increased to 0.850, but the removal of statement 14 resulted in increasing the value of Cronbach’s alpha to 0.855.  A Cronbach’s alpha value of 0.855, well above the minimum acceptable value of 0.70, indicated that this set of data had a high level of internal reliability.  Furthermore, all inter-item correlations were positive.
     A random sample of 333 was selected from the survey data, and 17 items (item 14 was omitted) were analyzed using a varimax rotation within SPSS.  Because 53 cases contained missing values, the analysis was performed for n = 280 cases.  The fifth eigenvalue was above 0.98, and the five-factor solution accounted for 61.2% of the variance.  However, many items had high cross-loadings.  By eliminating the items with high cross-loadings, the researchers selected 10 items for additional study.  Because 36 cases contained missing values, the analysis was performed for n = 297 cases.  The three significant factors identified using the varimax rotation were (1) positive attitudes toward labor unions (five items, percent of variance = 25.272); (2) attitudes about general worker strikes (two items, percent of variance = 16.908); and (3) attitudes about labor union involvement in the political process (three items, percent of variance = 15.702).  The three factors together explained 57.882 percent of the variance.  Next, we applied the three-factor solution to the full sample (n = 669).
     The three-factor solution applied to the full sample did not perform as well as it did on the random sample.  Because of high cross-loading, the researchers removed item 9.  Subsequently the researchers removed item 12, because it was almost solely responsible for the third factor in the next solution.  The three significant factors identified using the varimax rotation were (1) positive attitudes toward labor unions (four items, percent of variance = 30.306); (2) attitudes about general worker strikes (two items, percent of variance = 21.228); and (3) attitudes about labor union involvement in the political process (two items, percent of variance = 16.534).  The three factors together explained 68.069 percent of the variance.  A test for internal reliability on these eight items using the full sample was performed using SPSS.  It revealed a Cronbach’s alpha of 0.770, still above the minimum acceptable value of 0.70.  The score created by the sum of the three factor scores was labeled the Attitudes toward Labor Unions Scale (ATLUS) score.  Table 2 shows the individual items, their loadings on each factor, their total rotation sums of squared loadings, and explained variance (68.069 percent) for the full sample (n = 602).
Cross-Tabulation Analysis
     The ATLUS scores were recoded as low, median, and high for a cross-tabulation analysis on the scores against the various demographic variables.  Subjects in the low group had the lowest opinion of labor unions, whereas the subjects in the high group had the highest opinion of labor unions.  Those in the median group had an opinion of labor unions that was between that of the low and high groups.  To ensure valid chi-square tests, categories with expected frequencies less than five were eliminated from the analysis and the tests were repeated.  All cross-tabulation tests reported in this paper were for categories having expected frequencies of at least five. 
     As expected, the cross-tabulation tests of the three ATLUS score groups against the union membership (P-value less than .001) and parent-union membership (P-value less than .01) variables were significant.  About 77.5% of the union members were in the high ATLUS score group.  Because only two respondents classified themselves as other minority, race was recoded as (1) white and (2) nonwhite.  The significant results, shown in Table 3, were for race, total family income, and education.  Hypotheses 1a and 1b, that there are no differences in the attitudes of the public toward labor unions based on age (P-value = 0.410) or gender (P-value = 0.912), are not rejected.  In contrast, hypotheses 1c, 1d, and 1e, that there are no differences in the attitudes of the public toward labor unions based on race (P-value = 0.000), total family income (P-value = 0.012), or education (P-value = 0.000), are rejected.
DISCUSSION
     Using exploratory factor analysis on the 18 statements expressing an attitude about labor unions, we identified three factors used to create the Attitudes toward Labor Unions Scale (ATLUS) score.  The eight-item ATLUS that was derived had high internal reliability (Cronbach’s alpha = 0.770) when applied to the full sample.  The ATLUS scores were then recoded into high, median, and low groups.  A cross-tabulation of the ATLUS groups against various demographic variables showed a significant result for race, total family income, and education.  However, there was insufficient evidence to suggest a difference in attitudes toward labor unions based on either age or gender.  The significant results of this cross-tabulation analysis are presented in Table 3.
     We note that 50 of the total 553 respondents for the cross-tabulation analysis of race versus the three ATLUS groups were from the nonwhite group (9%), whereas 15.6% of the high ATLUS scorers were from the nonwhite group.  From another perspective, we see that about 32.5% of the total 553 respondents were high ATLUS scorers but 56% of the nonwhite group were high ATLUS-scorers.  To determine if this effect was related to union membership, the nonwhite group of respondents was cross-tabulated by union membership and ATLUS group.  As previously noted, about 77.5% of all union members were in the high ATLUS score group.  However, when restricted to the nonwhite group of respondents, 92.9% of the union members and 40% of the nonunion members were in the high ATLUS score group.  Even the nonunion members of the nonwhite group of respondents had a higher percentage of high ATLUS scores than expected.  These results suggest that nonwhites are more likely to have a positive attitude toward labor unions.
     The lowest total family income level was $15,000 or less.  We note that 88 of the total 579 respondents for the cross-tabulation analysis of income versus the three ATLUS groups are from the lowest income group (15.2%), whereas 20% of the high ATLUS scorers were from this lowest income group.  From another perspective, about 33.7% of the total 579 respondents were high ATLUS scorers but about 44.3% of the lowest income group were high ATLUS scorers.  To determine if this effect was related to union membership, the group of respondents with the lowest income level was cross-tabulated by union membership and ATLUS group.  As previously noted, about 77.5% of all union members were in the high ATLUS score group.  However, when restricted to the group of respondents with the lowest income level, 71.4% of the union members and about 42.0% of the nonunion members were in the high ATLUS score group.  Even the nonunion members of the group of respondents with the lowest income level had a higher percentage of high ATLUS scores than expected.  These results suggest that people in the lowest income level ($15,000 or less) are more likely to have a positive attitude toward labor unions.
     The highest total family income level was over $60,000.  It is noted that 55 of the total 579 respondents for the cross-tabulation analysis of income versus the three ATLUS groups are from the highest income group (9.5%), whereas about 6.2% of the high ATLUS scorers were from this highest income group.  From another perspective, about 33.7% of the total 579 respondents were high ATLUS scorers but about 21.8% of the highest income group were high ATLUS scorers.  Because only 2 of the 55 respondents in the highest income level were union members, the percentage of nonunion members in the highest income level that were high ATLUS scorers is 20.8%, very close to 21.8%.  There are fewer high ATLUS scorers in the highest income level than expected.  These results suggest that people in the highest income level (over $60,000) are less likely to have a very positive attitude toward labor unions.
     The second highest total family income level was $45,001 – $60,000.  We note that 48 of the total 579 respondents for the cross-tabulation analysis of income versus the three ATLUS groups are from the second-highest income group (8.3%), whereas about 5.1% of the high ATLUS scorers were from this second-highest income group.  From another perspective, about 33.7% of the total 579 respondents were high ATLUS scorers but about 20.8% of the second-highest income group were high ATLUS scorers.  Because only 1 of the 48 respondents in the second-highest income level was a union member, the percentage of nonunion members in the second-highest income level that were high ATLUS scorers is 19.1%, very close to 20.8%.  There are fewer high ATLUS scorers in the second-highest income level than expected.  These results suggest that people in the second-highest income level ($45,001 – $60,000) are less likely to have a very positive attitude toward labor unions.
   We note that 29 of the total 598 respondents for the cross-tabulation analysis of education versus the three ATLUS groups were from the least-educated group (4.8%), whereas 9.5% of the high ATLUS scorers were from the least-educated group.  From another perspective, about 33.4% of the total 598 respondents were high ATLUS scorers, but about 65.5% of the least-educated group were high ATLUS scorers.  We note that 100 of the total 598 respondents for the cross-tabulation analysis of education versus the three ATLUS groups were from the high school graduate group (16.7%), whereas 24.5% of the high ATLUS scorers were from the high school graduate group.  From another perspective, about 33.4% of the total 598 respondents were high ATLUS scorers but 49% of the high school graduate group were high ATLUS scorers.  These results suggest that less-educated people are more likely to have a positive attitude toward labor unions.  However, about 37% of these two less-educated groups were union members who tend to have a more positive attitude toward labor unions than nonmembers.  Therefore, this result is not unexpected.
     This research has provided several contributions.  First, the Attitudes toward Labor Unions Scale (ATLUS) was refined and has been determined to be useful in measuring people’s attitudes regarding labor unions.  Second, this research has identified segments of the public that have a more positive impression of labor unions than their percentage membership in the sample suggests.  Third, this research has identified segments of the public that have a more negative impression of labor unions than their percentage membership in the sample suggests.
     Some limitations of the application of the findings of this research should be noted.  The respondents in this study were located in a single state in the southeastern U.S. and were mostly male and white with low to middle income.  A little more than half of these respondents were 45 or under, and fewer than half had one or more college degrees.  Because this study is a cross-sectional survey of participants at one particular time without any follow-up, it may have common method variance.  Because the respondents in this study resided in a right-to-work state, generalization of these findings is expected to extend only to people residing in other right-to-work states.

REFERENCES

Atleson, James B., “Commentary: Law and Union Power: Thoughts on the United States and Canada,” Buffalo Law Review, 42 Buffalo L. Rev. 463, Spring 1994.

Bureau of Labor Statistics, Union Membership (Annual), U.S. Department of Labor, URL:
http://www.bls.gov/news.release/History/union2_01252008.txt, January 2008.

Bureau of Labor Statistics, Union Membership (Annual), U.S. Department of Labor, URL:
http://www.bls.gov/news.release/History/union2_01212004.txt, January 2004.

Denholm, David, “Do Americans Support Labor Unions?” Capital Research Center, June 2004. 1-6.

Jarley, Paul, and Sarosh Kuruvilla, “American Trade Unions and Public Approval: Can Unions Please All of the People All of the Time?” Journal of Labor Research, XV, Spring 1994, 97-116.

Karier, Thomas, “Accounting for the Decline in Private Sector Unionization Representation Elections, Structural Changes and Restructuring,” Jerome Levy Institute Working Paper No. 44, February 1991.

Leap, Terry L., William H. Hendrix, R. Stephen Cantrell, and G. Stephen Taylor, “Discrimination against Prounion Job Applicants,” Industrial Relations: A Journal of Economy and Society, 29, 1990, 469-478.

Margalioth, Sharon Rabin, “The Significance of Worker Attitudes: Individualism as a Cause for Labor’s Decline,” Hofstra Labor and Employment Law Journal, 16, Fall 1998, 133 166.

Saporta, Ishak, and Bryan Lincoln, “Managers’ and Workers’ Attitudes towards Unions in the U.S. and Canada,” Industrial Relations, 50, Summer 1995, 550-567.