Public Perceptions of Trade Unions in Countries of the European Union

Abstract

Given the ideological, political, and economic structural changes in the latter part of the twentieth century it might be expected that the demand for trade unions has significantly declined. Using a European-wide survey, this article addressed the extent to which European citizens perceive a need for trade unions. Our results indicate that contrary to expectations, a substantial majority of respondents perceived a need for strong trade unions to protect their pay and working conditions. Attitudinal formation appears to be more influenced by individual characteristics than either structural or institutional regimes, at least in a European context. Conversely the institutional measures of union presence and country of origin substantially account for the factors that determine why employees with favorable perceptions of trade unions become a union member. Among employees the extent to which positive attitudes converts into actual union membership appears to be critically dependent on a union-friendly institutional regime.

Keywords

Europe trade unions public perceptions joining institutional regimes

Introduction

Historically, unions have provided significant services to workers at enterprise and national level (Freeman and Medoff 1984). Trade unions not only emphasize the human aspect of the labor commodity, but in imposing a check on the exercise of power both express and foster democratic values and culture. Membership of a trade union also provides a stimulus to both electoral and political participation and even increases the life satisfaction of citizens, particularly the well-being of citizens with lower incomes (D’Art and Turner 2007; Flavin, Pacek, and Radcliff 2010). In Europe trade unions have been among the most important causal mechanisms in the development of welfare states (Radcliff 2005) and have played a key role in shaping social policy and the European social model (Hyman 2005). Arguably no viable alternative model of employee representation has emerged as a substitute for trade-union representation (D’Art and Turner 2003a). Trade unions in Europe experienced a massive surge of membership in the twentieth century, particularly after World War II (Ebbinghaus and Visser 2000). In most European countries trade unions enjoyed continuous growth up to 1980 (Ebbinghaus and Visser 2000). However, between 1980 and 2003 most countries, with the exception of Sweden, Denmark, and Finland, recorded often substantial declines in union density (Ebbinghaus and Visser 2000; Visser 2006). Trade unions now represent a smaller proportion of the employed labor force in Europe than at any other time since 1950 (Verma and Kochan 2004). Falling union membership and density levels may reflect more negative public attitudes toward trade unions in Europe, providing a less politically supportive institutional environment. In particular the dominance of neoliberal ideology and policies in many countries for some time is likely to have created a cultural framework that inculcated negative perceptions of unions and their functions in the wider public. A neoliberal agenda advocates a restoration of the primacy of the market mechanism that includes low social spending, low taxation, deregulation of financial and labor markets, and the privatization of state companies (Pollin 2003; Harvey 2005). In Europe the political balance of power generally swung against labor and even some nominally socialist governments pursued conservative economic strategies (Ferner and Hyman 1992).

In this union-hostile environment it might be expected that the demand for union membership has significantly declined. If declining union density is primarily due to falling employee demand then the outlook for unions is relatively bleak. Indeed, some have argued that the decline in union membership and density levels reflects a decline in the demand by workers for union representation (Troy 1999; 2004). An explicit aspect of the “decline in demand” thesis is that the functions unions provide are no longer necessary or relevant in a modern economy. Consequently, it might be expected that public attitudes and hence the demand for trade unions are likely to be relatively negative, that is, where less than a majority of the public favor trade unions. Yet the European survey used in this article indicates that public perceptions of trade unions in the fifteen EU countries are relatively favorable. A majority (72 percent) agree that workers need trade unions to protect their pay and working conditions while only 13 percent disagree and 15 percent have no opinion. Indeed, compared to the Euro-barometer findings for a similar question in 1984 and 1994, the proportion of respondents with a positive view of trade unions is significantly higher and those with negative attitudes significantly lower (Euro-barometer 44.3, 1996 and Euro-barometer 21, 1984).

In this article we examine the factors that are associated with positive perceptions of the need for strong trade unions in the contemporary workplace of the twenty-first century. Yet positive attitudes to trade unions do not necessarily convert into actual membership of a trade union. In the second part of the article the analysis is restricted to respondents who are employees and who agree with the need for strong trade unions. Specifically we test for the factors among this union-sympathetic group that determine why some employees become union members while others fail to join.

Determinants of Union Attitudes and Membership

An attitude is a hypothetical construct that represents an individual’s degree of like or dislike for something. Attitudes are judgments that express verbally an individual’s degree of preference for an entity, and can be indicative of a behavioral intention or a typical behavioral tendency (Eagly and Chaiken 1993). Consequently employees who believe in the need for strong trade unions are more likely to join a trade union. Generally the literature on union-joining identifies a number of broad, if sometimes competing, explanations. A worker’s decision to become a member of a trade union is best viewed as a function of both the presence of a union in the workplace and individual preferences for membership (Green 1990; Hartley 1992, 173-74). In the latter case membership is perceived to be a function of individual and instrumental characteristics. Alternatively union availability in the workplace is argued to be facilitated or constrained by certain structural factors such as firm size and occupational composition and/or the institutional regime and unique historical experience in a particular country.

Individual and Instrumental Characteristics

A number of individual characteristics have been advanced that influence workers’ demand and membership of trade unions. These include age, gender, education, occupational level, and the instrumentality of union membership. In a period of union decline in many European countries it can be argued that younger workers have less experience of trade unions and will display significantly lower levels of attachment to collective organizations. For instance, older workers occupying the lower reaches of the organizational hierarchy may over time develop a negative conception of management (D’Art and Turner 1999). Consequently they may be more likely to perceive a need for strong trade unions. Alternatively, given the marginalization of trade unions and their increasing absence in the workplace, younger workers may have little or no experience of collectivism. The attitudes of younger workers to unions may thus range from indifferent to hostile. However, in large-scale, cross-sectional surveys of national populations the positive responses of younger workers to unions outweigh the negative (Haynes, Vowles, and Boxall 2005; Hart Research Associates Inc. 1999). Thus the relationship between age and perceived need for a trade union remains an open question that is best settled empirically.

Historically the short-term and marginal experience of women in work relative to the male labor force, it has been suggested, has tended to discourage female workers from adopting a collectivist response to the issues of pay and conditions (Lockwood 1966; Hyman and Price 1983). Women were also perceived to be more passive and less confrontational than men, resigned to existing job conditions, and so less likely to engage in conflict and strikes (Fiorito, Gallagher, and Greer 1986; Wheeler and McClendon 1991; Schur and Kruse 1992). This “stereotype” has endured that women are uninterested in unions and identify mainly with family issues (Wajcman 2000, 187; Briskin and McDermott 1993, 7-8).Yet, women it seems are just as likely as men to have positive attitudes toward unions (Turner and D’Art 2003; Walters 2002), to vote for union recognition (Premack and Hunter 1988; Schur and Kruse 1992, 100) and be active union members (Klandermans 1992; Lawrence 1994, 94). We test for differences in attitudes to trade unions between women and men and specifically whether women are less likely than men to perceive a need for strong trade unions (hypothesis 1).

Attitudes toward trade unions may be associated with different educational levels. Educational level can plausibly be interpreted as a proxy measure of social class position. It can be expected that respondents with higher levels of education are less likely to perceive a need for strong trade unions (hypothesis 2). Similarly, being in a higher-level occupation may adversely affect attitudes to trade unions. Usually a concomitant of such occupations are high levels of discretion, job autonomy, and high trust relations (Fox 1974; Brown 1999). For employees in such occupations the relevance of trade unions may be low. Conversely, workers employed in lower-level occupations often perceive themselves as unable to influence their work environment as individuals and are more likely perceive a need for strong trade unions (Hartley 1992, 171). Thus respondents in lower-level occupations are more likely to perceive a need for strong trade unions than higher-level occupations (hypothesis 3).

Political orientation can be expected to have a significant impact on union attitudes and membership. Given the relatively strong association between social democratic parties and trade unions in many European countries it is likely that respondents with a left-wing political orientation are likely to be positively disposed toward trade unions (Hague, Harrop, and Breslin 1998). Consequently it can be expected that respondents with a left political orientation are more likely to hold positive union attitudes (hypothesis 4). Workers join unions to improve their pay and working conditions. Union membership is attractive to the extent that it is instrumental in achieving these goals (Crouch 1982; Waddington and Whitson 1997). Levels of satisfaction with income may thus be expected to affect attitudes to trade unions. Research in the United States has related dissatisfaction with wages to unionization or at least a propensity to regard unions more favorably. In Europe similar findings exist regarding dissatisfaction with terms and condition of employment (see Hartley [1992, 169-70]). For workers with high levels of income satisfaction, the instrumental basis for union membership maybe eroded and attitudes affected accordingly. This suggests that respondents who are satisfied with their income level are less likely to perceive a need for strong trade unions (hypothesis 5).

Structural Characteristics

The structural approach ascribes changes in union availability to a number of factors that impact on union availability (Kaufman 2001; Disney 1990). Long-term socioeconomic changes in industrial societies such as the shift from manufacturing to service employment, related occupational transformations from manual to mental work, declining size of firms, an increasingly integrated global economy, and the decline in left politics and associated working-class project are proposed as explanations for the decline in the fortunes of trade unions (Bain and Price 1983; Booth 1986; Deery and Cieri 1991; Ebbinghaus and Visser 1999). Thus predictions are that establishment size—large firms are more likely to be unionized than small firms, industrial sector—manufacturing firms are more likely to be unionized than service sector firms—and private-sector establishments are less likely to be unionized than public-sector organizations (Bain and Price 1983; Hartley 1992). In the UK, for example, the increase in nonmanual jobs, the growth of the service and private sectors, and an increasing number of part-time jobs performed by women may have adversely affected trade unions (Bryson and Gomez 2005). A growing proportion of employment shifted to workers who were supposedly less inclined to organize. Since the early 1980s there has been a substantial rise in the percentage of employees who have never been union members (Bryson and Gomez 2005). We test for two critical structural factors here: size of firm and employment sector. In the former case the larger the size of the firm the more likely employees are to hold positive attitudes toward unions (hypothesis 6) and in the latter case employees in the public sector are more likely to have positive attitudes toward unions than other sectors (hypothesis 7).

Institutional Regime

The institutional explanation focuses on the contextual historical development and the specific national institutions governing industrial relations. Changes in union availability have been attributed to a range of institutional factors and processes such as the nature, scope, and depth of collective bargaining, labor legislation, the Ghent system, and the effects of management strategies on union organization (Lind 2009; Freeman and Pelletier 1990; Cappelli and McKersie 1987). The core argument is that the supply of unions at the workplace and the supporting legislation for a union presence are the key elements accounting for union growth and decline (D’Art and Turner, 2003b). From this perspective it can be predicted that attitudes toward unions and membership levels will vary between European countries due to different and distinctive historical and institutional trajectories. Countries with institutional regimes that facilitate trade-union membership such as Sweden, Denmark, and Finland are likely to also have high levels of agreement with the need for trade unions. More significantly, such regimes are more likely to match the demand for unions with the supply through a greater union present across firms of all sizes, industrial sectors, and occupations. In the UK, for example, Bryson and Gomez (2005, 87) found that the single most important factor determining the probability of never being a union member is “whether or not an individual is employed in a workplace with a recognized union.” Thus even among those employees who agree with the need for strong trade unions, union membership is likely to be a function of union presence both in terms of availability and the coverage or density of members in the workplace. We expect that respondents who report a union present in their workplace are more likely to have favorable attitudes toward unions (hypothesis 8). Additionally, in countries with higher density levels and hence greater presence and visibility across firms and sectors of the economy employees are more likely to have positive attitudes toward unions (hypothesis 9). Crucially, the availability and ease of access of unions in the workplace reflects the nature of the specific institutional regime in each country. Thus it can be expected that employees who agree with the need for strong unions are more likely to be members of a union in countries with a favorable institutional regime (hypothesis 10).

To capture this institutional effect the fifteen EU countries in this study are clustered into four distinct groups according to the union-density level of only employees (that is employment density) in the population sample. This clustering facilitates multivariate analysis and is more parsimonious compared to reporting the results for fifteen separate countries. Countries in the sample with the lowest density levels (ranging for 9 to 13 percent) are France, Spain, Portugal, and Greece; the next group are Germany, UK, Italy, and Holland (with density ranging from 18 to 23 percent). A third group is comprised of Ireland, Austria, Luxemburg, and Belgium (with density ranging from 25 to 38 percent), and the group with the highest density levels are Finland, Sweden, and Denmark (ranging from 58 to 70 percent).

It may, however, be questionable whether the single measure of union density accurately reflects the nature of the industrial-relations regime and institutions in a particular country. Early studies on ranking countries occurred mainly in the literature on corporatism based on the degree of centralization of collective bargaining and peak organizations and levels of political exchange (Calmsfor and Driffill 1988; Henley and Tsakalotos 1993, 87). More recent ranking schemes have been developed using a wider array of indicators such as Kitschelt et al.’s (1999) institutional diversity of contemporary capitalism and Hall and Soskice’s (2001) firm-centered classification of varieties of capitalism. Kitschelt et al. (1999, 435) identify two forms of production regimes: coordinated market economies traditionally associated with medium to strong corporatist systems and liberal market economies typical of the Anglo-Saxon economies of the United States, UK, Canada, Australia, and New Zealand. CMEs are subdivided into national coordinated market economies (NCME) characterized by high levels of social equality and less egalitarian regimes labeled as sectoral coordinated market economies (SCME). Hall and Soskice’s (2001) essential distinction is between coordinated market economies and liberal market economies. In their work, France, Spain, Portugal, and Greece are identified as difficult to categorize and may “constitute another type of capitalism—marked by a large agrarian sector and recent histories of extensive state intervention” but retaining “more liberal arrangements in the sphere of labor relations” (Hall and Soskice 2001, 21). In the traditionally strong corporatist countries, trade-union movements have retained their influence albeit in a changed manner while in liberal market countries trade unions have been increasingly marginalized and weakened. Using the types of capitalism and varieties of capitalism categories to rank countries, Table 1 assesses the capacity of the single measure of union density to match up with these categories.

Table 1

Matching Types and Varieties of Capitalism and Union-Density Levels

Clustering by Union Density	Types of Capitalism Categories (Kitschelt 1999)	Varieties of Capitalism Categories (Hall and Soskice 2001)	Countries Correctly Predicted by Union-Density Level in the ESS 2002
Cluster 4
Sweden	NCME	CME	Yes
Finland	NCME	CME	Yes
Denmark	NCME	CME	Yes
Cluster 3
Ireland	LME	LME	No
Austria	SCME	CME	Yes
Luxemburg	—	—	—
Belgium	SCME	CME	Yes
Cluster 2
Germany	SCME	CME	No
UK	LME	LME	Yes
Italy	SCME	NC/C	Yes
Holland	SCME	CME	No
Cluster 1
France^b	SCME	NC/C	Yes
Spain	NC/C	NC/C	Yes
Portugal	NC/C	NC/C	Yes
Greece	NC/C	NC/C	Yes

Note: NCME = National Coordinated Market Economy; SCME = Sectoral Coordinated Market Economy; LME = Liberal Market Economy; NC/C = Not categorized or controversial; CME = Coordinated Market Economy.

Countries in the density clusters that do not conform to the types of capitalism and the varieties of capitalism categories are in cluster 3 (Ireland classed as an LME) and cluster 2 (Germany and Holland classed as CMEs—Luxemburg is not ranked in this literature). Yet in Ireland a relatively comprehensive form of corporatism has operated for over twenty years (Hardiman 2000; Turner 2002) and it may be more appropriate to rank Ireland as a medium-strong corporatist country located in the middle rankings of industrial relations regimes. Similarly, categorizing Germany as a CME is questionable in recent years. It has been argued that it is no longer possible to describe Germany as a “coordinated market economy” due to a distinct shift toward liberalization in collective bargaining, company-level governance, labor markets, and welfare state provision—caused among other factors by weakened trade unions (Streeck and Thelen 2005). Others perceive a partial liberalization in the labor market toward the British liberal market model (Hassel 2010; Seeleib-Kaiser and Fleckenstein 2007). Thus Germany’s position at the top of cluster 2 is, we believe, reasonably accurate. Alternatively, Holland is more appropriately defined as a CME and should be included in cluster 3 rather than cluster 2. However, given that clusters 1 and 4 are similar to the other rankings, we believe it best to remain consistent with the standard union-density measure rather than rearranging clusters 2 and 3. In any case the statistical effects on results are marginal even when types or varieties of capitalism rankings are used.¹ A further issue concerns the discrepancy between union density and collective- bargaining coverage. In many European countries collective agreements extend to nonunion firms guaranteeing very high union-coverage rates for collective agreements (Visser 2006, 46). However, not all employers may adhere to the collective agreement and there is a considerable empirical evidence showing that some employers consistently try to oppose unionization or union-backed collective agreements with a consequent expansion of nonunionism in some sectors (Doellgast and Greer 2007; Royle 2004). Thus there can be considerable variation within national systems regarding the effectiveness of extending collective-bargaining coverage to the nonunion sector. Although France has a bargaining coverage rate of over 80 percent, the union movement remains relatively weak with a union-density level of only 8 percent. Indeed, the severe decline of French unionism since the 1970s tends to reinforce the atypical character of the French system (Goetshcy and Rozenblatt 1992). According to Crouch (1996), French trade unions are probably, both in terms of low membership and internal conflict, now the weakest in the democratic world.

Data and Measures

The data used here comes from the 2002/2003 European Social Survey (ESS).² The European Social Survey is a biennial multicountry survey covering more than twenty states. The first round was fielded in 2002/2003. The dependent variable in our study is the extent to which respondents believe that trade unions are still necessary and is based on responses to the statement: Employees need strong trade unions to protect their working conditions and wages. Scored 1 = Strongly agree; 2 = Agree; 3 = Neither agree or disagree; 4 = Disagree; 5 = Strongly disagree. For ease of reporting, in many cases the categories of “strongly agree” and “agree” are combined, similarly “strongly disagree” and “disagree.” The meaning of “strong trade unions” is not elucidated in any detail in the survey questionnaire. Consequently, survey respondents take the statement at face value and attribute their own meaning to it. Broadly we interpret a positive answer to this question (strongly agree/agree) as favoring the need for trade unions.

Unfortunately, in later survey rounds in 2004/2005 and 2008/2009 this question is omitted. The survey is designed to chart and explain the interaction between Europe’s changing institutions and the attitudes, beliefs, and behavior patterns of its diverse populations. Sample population covered the eligible residential populations aged fifteen and over in each country. An essential element of the strategy is the achievement of high response rates in all participating countries. This is to ensure that the people interviewed in each country closely represent the country’s total population. The ESS then includes a high degree of representatives and reliability with well-designed and valid measures.

A unique aspect of the analysis in this article is that the ESS allows the various explanations: individual and instrumental characteristics, structural factors, and the nature of the institutional regime to be tested together. While previous studies have tested different aspects of these explanations, particularly individual and structural factors together, it is rare that such a comprehensive survey of individual respondents’ allows the addition of institutional measures in the same data set.

The fifteen EU member countries included in this article are Austria, Belgium, Germany, Denmark, Spain, Finland, UK, Greece, Ireland, Italy, Luxemburg, Netherlands, France, Portugal, and Sweden. We have excluded the new entrants such as Poland and Hungary as their experience of independent trade unions is relatively recent. Under communism nearly all workers joined official “transmission belt” unions that operated as an arm of the state rather than as independent representatives of workers. Consequently they were excluded as the legacy of this experience is likely to have been qualitatively different from workers in the fifteen member states (see Pollert [1999]; Martin and Cristesco-Martin [1999]; Blanchflower and Freeman [1997]).

A total of 29,110 surveys were returned from the fifteen countries. In total 24 percent (6,994) of the sample indicated membership of a trade union. Most respondents were active in the labor market: 21,362 of the sample were employees with employers and the self-employed accounting for 4,044 respondents. The proportion of respondents active in the labor market and members of a union (employment density) was 27 percent.

Results and Analysis

Table 2 focuses on the association between individual characteristics, structural factors, institutional regime, and perceptions of the need for trade unions. There is a statistical difference across most of the measures used, though not always in the direction predicted. Indeed women are significantly more likely to agree with the need for strong trade unions, 73 percent compared to 70 percent for men. Thus there is no support for hypothesis 1 that women are less likely than men to perceive a need for strong unions. However, there is statistically significant support for hypothesis 2 that respondents with higher levels of education are less likely to perceive a need for strong trade unions. A lower number of years in education is associated with increased agreement with the need for unions. Conversely a higher number of years in education is associated with lower agreement with the need for unions. Similarly the impact of occupational level on attitudes toward unions conforms to expectations as predicted by hypothesis 3, with manual workers significantly more likely to hold favorable attitudes toward unions compared to managers and professional employees. Not surprisingly, employees are significantly more likely to indicate agreement with the need for strong trade unions than employers and the self-employed.

Table 2

Individual Characteristics, Structural Factors, Institutional Regime and Perceptions of Union Need in the EU Fifteen Countries (Appropriate Weights Applied)

Gender	Agree %	Disagree %	Neither %	Mean	Sig.^a	N
Female	73	12	15	2.13	***	14,829
Male	70	15	15	2.21	13313
Age ^b
1. Under 18 years	81	6	12	2.01	*** (3 to 7)	1,073
2. 18 to 24 years	78	8	14	2.01	*** (3 to 7)	2,650
3. 25 to 34 years	73	13	15	2.14	** (4, 6, 7)	4,567
4. 35 to 44 years	70	15	15	2.22	5525
5. 45 to 54 years	71	14	15	2.17	4702
6. 55 to 65 years	69	15	16	2.23	4255
7. Over 65	69	15	16	2.22	5232
Years of Education
Up to 10 years	74	11	15	2.08	***	9,511
11 to 14 years	72	13	15	2.17	***	11,041
15 to 18 years	68	16	16	2.31	***	55,66
More than 18 years	66	15	19	2.32	***	1,426
Occupational Level
Managers	53	27	20	2.65	***	2,083
Professional	66	16	17	2.34	***	6,762
White collar	71	13	16	2.2	***	6,283
Manual workers	76	11	13	2.1	***	8,427
Active in the Labor Market
Self-employ/employer^c	60	22	18	2.48	***	3,907
Employee	74	12	14	2.12	21,125
Political Orientation
Left orientation	82	7	11	1.88	***	8,170
Centrist	68	15	17	2.26	***	10,941
Right orientation	58	23	19	2.52	***	5,816
Satisfaction Income Level
Living well	66	17	17	2.34	***	9,531
Coping	72	12	15	2.16	***	12,777
Difficult coping	79	10	11	1.95	***	4,277
Very difficult coping	77	10	13	1.91	***	1,257
Size
1. 1 to 10 employees	68	15	17	2.27	*** (2 to 4)	8,584
2. 11 to 24 employees	73	12	15	2.14	4,171	—
3. 25 to 99 employees	72	13	15	2.16	4,852	—
4. Greater than 100 employees	75	13	12	2.1	*** (3)	6,850
Industrial Sector
1. Industry	73	13	14	2.13	9620
2. Services	68	16	16	2.26	*** (1 and 3)	8,774
3. Public Sector	74	12	14	2.13	6,135	—
Union Presence
Union present	78	10	12	2.0	***	7,275
Union not present	66	16	18	2.28	4,810
Institutional Regime
Cluster 1 (Greece; Portugal; Spain; France)	78	9	13	1.94	***	6,945
Cluster 2 (Holland; Italy; UK; Germany)	68	16	17	2.3	***	8,392
Cluster 3 (Belgium; Ireland; Austria; Luxemburg)	71	13	16	2.18	***	7,459
Cluster 4 (Denmark; Sweden; Finland)	76	11	13	2.07	***	5,380

*p < 0.05; **p < 0.01; ***p < 0.001.

Significant differences reported using one way ANOVA, bonferroni post hoc test.

Independent samples T-Test used for dichotomous variables.

Source: European Social Survey (ESS), 2002/2003.

As predicted, those with a left political orientation are significantly more likely to agree with the need for unions than respondents in the center or on the right, supporting hypothesis 4 that respondents with a left orientation are more likely to hold positive union attitudes. There is also support for the instrumental attraction of trade unions that respondents who are satisfied with their income level are less likely to perceive a need for unions (hypothesis 5). A total of 66 percent of respondents who report “living well on their income” agree with the need for unions compared to 77 percent of those who find it “very difficult coping” with their income.

Structural factors also appear to be significant and in the direction predicted. Respondents employed in firms with ten or fewer employees were less likely to agree with the need for unions than those employed in companies with over one hundred employees, 68 percent compared to 75 percent. This provides some support for hypothesis 6 that the larger the size of the firm, the more likely employees are to hold positive attitudes toward unions. Respondents employed in the services sector are significantly likely to be less favorable to unions than those employed in industry and the public sector, providing some support for hypothesis 7 that employees in the public sector are more likely to have positive attitudes toward unions. However, there is no significant difference in attitudes between respondents employed in industry and the public sector.

The relationship between measures of the institutional regime, union presence, country clusters, and attitudes toward unions are all statistically significant. Union presence significantly affects union attitudes as expected with 78 percent of respondents who report a union presence in their workplace compared to 66 percent where no union is present, supporting hypothesis 8. In contrast there is no clear relationship between country clusters based on union density levels and attitudes to trade unions. Though there are statistically significant differences between the clusters it is not in the direction predicted. Cluster 1 with the lowest union-density levels has the highest proportion of respondents, 78 percent, agreeing with the need for strong trade unions. Consequently there is no support for hypothesis 9 that countries with a high proportion of respondents in trade unions are likely to have more positive attitudes toward unions.

Given the substantial proportion of respondents, 72 percent, agreeing with the need for strong trade unions there is limited variance in our dependent variable. As Table 2 indicates even such groups as employers, the self-employed, and managers have majorities agreeing with the need for unions. To test the combined effects of individual characteristics, structural factors, and institutional regime on union attitudes, Table 3 examines the characteristics that distinguish those respondents who disagree compared to those who agree with the need for unions. As the difference between “strongly dis(agree)” and “dis(agree)” is essentially marginal compared to the qualitative difference between respondents with positive perceptions of unions (strongly agree and agree) and negative perceptions (strongly disagree and disagree) it is more appropriate to use a dichotomous dependent measure. Excluding respondents with “no opinion,” there are 21,635 respondents who agree with the need for strong trade unions and approximately four thousand who disagree. The number of respondents included in the multivariate analysis varies considerably in each equation due to missing data in the measures.

Table 3.

Individual, Structural, and Institutional Factors Associated with a Negative View of Trade Unions (Respondents with “No Opinion” Omitted)

Multivariate Analysis: Binary Logistic Regression

Odds Ratios (Exp[B]) Reported, Standard Errors in Parentheses

	Disagree with Need for Unions^a
	1	2	3	4
Individual and Instrumental
Males	1.3*** (0.05)	1.3*** (0.05)	1.3*** (0.07)	1.3*** (0.05)
Reference group: females	—	—	—	—
Age over 55 years	2.3*** (0.1)	2.3*** (0.1)	2.0*** (0.1)	2.1*** (0.1)
Age 25 to 54 years	2.4*** (0.1)	2.3*** (0.1)	2.0*** (0.13)	2.3*** (0.1)
Reference group: under 25 years	—	—	—	—
Over 15 years education	1.4*** (0.07)	1.4*** (0.07)	1.5*** (0.11)	1.3*** (0.07)
11 to 14 years education	1.1* (0.05)	1.1* (0.06)	1.2** (0.08)	1.2*** (0.06)
Reference group: less than 11 years	—	—	—	—
Self-employed/employer	1.5*** (0.05)	1.4*** (0.07)	2.1*** (0.26)	1.4*** (0.07)
Reference group: employees	—	—	—	—
Nonunion respondents	3*** (0.06)	2.9*** (0.07)	2.7*** (0.09)	2.9*** (0.07)
Reference group: union members	—	—	—	—
Managers	2.5*** (0.08)	2.5*** (0.08)	3.7*** (0.13)	2.5*** (0.08)
Professional	1.6*** (0.06)	1.7*** (0.07)	1.8*** (0.1)	1.7*** (0.07)
White collar	1.3*** (0.06)	1.3*** (0.07)	1.5*** (0.1)	1.3*** (0.07)
Reference group: manual workers	—	—	—	—
Right political orientation	4.3*** (0.06)	4.3*** (0.06)	3.5*** (0.09)	4.3*** (0.07)
Centrist position	1.5*** (0.05)	1.5*** (0.05)	1.5*** (0.08)	1.5*** (0.05)
Reference group: left orientation	—	—	—	—
Satisfied/coping with income	1.4*** (0.07)	1.4*** (0.07)	1.3*** (0.12)	1.3*** (0.07)
Reference group: difficulties with income level	—	—	—	—
Structural Factors
1 to 10 employees	—	1.0	0.8	1.1
11 to 24 employees	—	0.9	0.9	1.0
25 to 99 employees	—	1.1	0.8	1.1
Reference group: greater 100 employees	—	—	—	—
Industry	—	1.2* (0.07)	1.0	1.2* (0.07)
Services	—	1.2*** (0.06)	1.1	1.2*** (0.06)
Reference group: public sector	—	—	—	—
Institutional Regime
Union not present	—	—	1.6*** (0.2)	—
Reference group: union present	—	—	—	—
Cluster 1 (Greece; Portugal; Spain; France)	—	—	—	−2.3*** (0.12)
Cluster 2 (Holland; Italy; UK; Germany)	—	—	—	1.2* (0.1)
Cluster 3 (Belgium; Ireland; Austria; Luxemburg)	—	—	—	1.1
Reference group: Cluster 4 (Denmark; Sweden; Finland)
Chi-sq.	1,575***	1,543***	734***	1,719***
Nagelkerke pseudo R²	0.15	0.16	0.15	0.17
N	17,593	16,941	8,469	16,941

p < 0.05; **p < 0.01; ***p < 0.001.

Dependent variable is perceived need for strong trade unions. Scored 1 = Disagree and 0 = Agree.

Respondents with “no opinion” are omitted from the analysis.

Source: European Social Survey (ESS), 2002/2003.

Equation 1 (Table 3) reports the impact of individual and instrumental characteristics on a respondent’s likelihood to disagree with the need for unions. All the measures included are statistically significant. Particularly notable are political orientation and union membership. Compared to respondents on the left, those with a right orientation are over four times more likely to disagree with the need for strong unions, and respondents who are not members of a union are three times more likely than union members. Managers are over two times more likely than manual workers to indicate disagreement with the need for unions. Age also has a significant impact, with respondents over the age of twenty-four being over two times as likely to disagree with unions compared to respondents under twenty-five. Though statistically significant, gender, employment status, and satisfaction with income have a relatively minor association with the dependent variable. While the overall model in equation 1 is statistically significant, the pseudo R2 is quite modest, explaining only 15 percent of the variance in the dependent variable. In equation 2 the structural measures of size and sector are added to the model. Firm size has no significant impact and sector has only a modest effect, with respondents working in the services sector 1.2 times more likely to disagree that unions are needed than those in the public sector. Also the inclusion of the structural measures adds little to the pseudo R2. Similarly the inclusion of union presence in equation 3 has a negligible influence. Though respondents who report no union presence in their workplace are 1.6 times more likely to disagree with the need for unions, it should be noted that a majority of respondents fail to answer this question. Along with individual characteristics and structural factors, the four country clusters are included in equation 4 (union presence is omitted due to small numbers). Respondents from the low-density countries in cluster 1 are over two times less likely to actually disagree with the need for unions than the high-density cluster 4 countries (note the minus odds ratio). Thus, contrary to expectations, the particular institutional regime appears to have no significant association with attitudes to trade unions for those surveyed in the fifteen EU countries. Overall, structural and institutional factors have little explanatory power in distinguishing between those who disagree or agree with the need for strong unions. Individual characteristics such as political orientation, union membership, and occupational level appear to have more purchase in accounting for differences in attitudes to unions.

Factors Determining Union-Joining among Employees Favorable to Trade Unions

A second and possibly more intriguing question in this article focuses on why some employees remain nonunion despite positive attitudes toward unions. In total 74 percent of employees agreed with the need for strong trade unions. Yet only 24 percent of this group of employees are members of a union and the majority 76 percent, though favorably disposed to trade unions, are not attached to a trade union. Table 4 attempts to identify those factors that account for this seeming representation gap where a substantial proportion of employees favorable to the need for unions remain nonunion. The dependent variable is structured in order to focus on the chances of respondents not being a union member. Using our measures of individual characteristics, structural factors, and institutional regime combined allows an evaluation of the relative efficiency of each explanation.

Table 4.

Individual, Structural and Institutional Factors Associated with Agreement with Strong Unions but Not Being a Union Member

Method: Binary Logistic Regression

Odds Ratios (Exp[B]) Reported, Standard Errors in Parentheses

	Dependent Variable: Respondents Not in a Union^a
	1	2	3	4	5
Individual Characteristics
Males	1.7*** (0.05)	1.8*** (0.04)	1.5*** (0.06)	1.9*** (0.05)	1.6*** (0.06)
Reference group: females	—	—	—	—	—
Age over 55 years	−2.0*** (0.09)	−1.9*** (0.09)	−3.3*** (0.1)	−1.8*** (0.09)	−3.2*** (0.13)
Age 25 to 54 years	−3.3*** (0.08)	−3.3*** (0.08)	−2.1*** (0.1)	−3.7*** (0.09)	−2.4*** (0.11)
Reference group: under 25 years	—	—	—	—	—
Over 15 years education	1.5*** (0.07)	1.3*** (0.06)	0.9	1.3*** (0.06)	−1.2* (0.08)
11 to 14 years education Reference group: less than 11 years	1.1	0.9	−1.2*** (0.06)	0.9	−1.1* (0.06)
Managers	1.1	—	—	—	—
Professional	1.2* (0.06)	—	—	—	—
White collar	1.1* (0.06)	—	—	—	—
Reference group: manual workers	—	—	—	—	—
Right political orientation	0.9	—	—	—	—
Centrist position	1.0	—	—	—	—
Reference group: left orientation	—	—	—	—	—
Satisfied/coping with income	1.7*** (0.07)	2.2*** (0.06)	2.1*** (0.07)	1.6*** (0.06)	1.4*** (0.08)
Reference group: difficulties with income level	—	—	—	—	—
Structural Features
1 to 10 employees	—	2.3*** (0.06)	1.1	2.5*** (0.06)	1.2* (0.09)
11 to 24 employees	—	1.6*** (.06)	0.9	1.8*** (.06)	1.1
25 to 99 employees	—	1.3*** (.05)	−1.2* (.06)	1.4*** (.05)	0.9
Reference group: greater 100 employees	—	—	—	—	—
Industry	—	1.5*** (0.06)	1.2** (0.07)	1.6*** (0.05)	1.3*** (0.07)
Services	—	2.1*** (0.05)	1.6*** (0.07)	2.3*** (0.06)	1.7*** (0.07)
Reference group: public sector	—	—	—	—	—
Union not present	—	—	8.5*** (0.07)	—	8.7*** (0.08)
Reference group: union present	—	—	—	—	—
Institutional Context
Cluster 1 (Greece; Portugal; Spain; France)	—	—	—	18.1*** (0.09)	35*** (0.15)
Cluster 2 (Holland; Italy; UK; Germany)	—	—	—	9.1*** (0.08)	15*** (0.14)
Cluster 3 (Belgium; Ireland; Austria; Luxemburg)	—	—	—	4.7*** (0.1)	7*** (0.17)
Reference group: Cluster 4 (Denmark; Sweden; Finland)	—	—	—	—	—
Chi-sq.	697***	1,555***	2,045***	2,837***	2,940***
Nagelkerke pseudo R²	0.08	0.14	0.29	0.25	0.4
N	13,031	14,957	8,473	14,957	8,473
Missing cases	3,059	1,113	7,617	1,133	7,617

p < 0.05; **p < 0.01; ***p < 0.001.

Dependent variable is union membership. Scored 1 = Not a union member; 0 = union member.

Source: European Social Survey (ESS), 2002/2003.

First, the dependent variable is regressed on the individual measures in equation 1 (table 4). Age has a moderate and significant impact on employee chances of not being a union member while gender, education, and satisfaction with income level have a minor effect. Employees over fifty-five are twice as likely to be union members (note the minus sign) than employees under twenty-five, and those aged twenty-five to fifty-four are three times more likely. Males are 1.7 times more likely not to be union member than females, and employees with over fifteen years of education are 1.5 times more likely not to be in a union compared to employees with less than eleven years of education. Employees who report being satisfied or coping with their incomes are 1.7 times more likely to be nonunion. Occupational level and political orientation have no significant impact on the likelihood of union membership and are dropped from the following equations to reduce the number of missing cases in the analysis.

Both structural measures of size and sector have a significant if moderate impact (equation 2). In firms employing ten or fewer employees, workers are 2.3 times more likely not to be in a union than those in firms with more than one hundred employees, and employees working in the services sector are 2.1 times less likely to be union members than employees in the public sector. The addition of the structural factors in equation 2 increases the pseudo R2 considerably and improves the overall model fit. The institutional measure of union presence is introduced in equation 3. Employees who report no union presence in their workplace are 8.5 times more likely not to be in a union. With this measure the pseudo R2 more than doubles, but this result must be interpreted cautiously given the large number of missing cases. In equation 4 union presence is omitted and only the country clusters are included. Employees from the low-density cluster 1 and cluster 2 are over eighteen times and nine times respectively more likely not to be a union member compared to cluster 4 countries. Equation 5 includes both institutional measures of union presence and country cluster. While the effect of union presence on the chances of being a union member remains stable, the effects of the country clusters are amplified. Employees in cluster 1 are thirty-five times more likely not to be a union member compared to respondents in cluster 4. In clusters 2 and 3 employees are fifteen and seven times more likely not to be in a union. Union availability is evidently strongly linked to country cluster with interaction effects such that unions are widely available to employees who have favorable union attitudes and wish to become a union member. These results support hypothesis 10 that employees who agree with the need for strong unions and reside in countries with a favorable institutional regime are more likely to be members of a union. In countries such as Sweden and Denmark, there appears to be a correspondence between attitudinal disposition, intention to join, and union availability.

Overall, equation 5 provides the best fit with the highest pseudo R2. There is also a clear pattern in the factors associated with union membership among employees with favorable attitudes toward trade unions. The institutional measures of union presence and country cluster substantially account for the reasons why many employees with favorable union attitudes nevertheless fail to join a union. Structural factors have a more moderate impact while individual characteristics have the smallest effect.

Discussion and Conclusion

Compared to previous Euro-barometer findings the proportion of respondents in the 2002 ESS with a positive view of trade unions appears to have significantly increased. A substantial majority (72 percent) of all respondents agreed that workers need trade unions to protect their pay and working conditions compared to only 13 percent who disagreed. Even employers, the self-employed, and managers have majorities agreeing with the need for unions. The nature of the ESS allowed a unique test in the same data set of three influential explanations for trends in the demand and membership of trade unions: individual characteristics, structural factors, and institutional regime. Aside from gender, the individual and instrumental characteristics were all significant and in the direction predicted. Respondents with lower levels of education, lower occupations, a left political orientation, and dissatisfaction with income level were more likely to hold favorable attitudes toward trade unions. Structural factors were partially supported, with employees employed in larger firms and the public sector more likely to have positive perceptions of trade unions. Institutional effects were more mixed, with union presence being significantly associated with more favorable attitudes but country clusters having no relationship with union attitudes.

Using multivariate analysis we focused on the small proportion of respondents who did not perceive a need for unions. Contrary to expectations, the particular institutional regime or structural features appeared to have little explanatory power in distinguishing between positive and negative perceptions of the need for strong trade unions in the fifteen EU countries. Individual characteristics such as political orientation, union membership, and occupational level had the most significant impact on attitudes to unions. Thus, perceptions of the need for trade unions appears to be more influenced by individual characteristics than either structural or institutional regimes, at least in a European context. Undoubtedly the overwhelming feature of the respondents surveyed in the ESS from those on the political right to employers and the self-employed are the number that agreed with the need for strong trade unions. Such positive perceptions of trade unions in Europe may be partly ascribed to a combination of social democracy on the political left and paradoxically social Catholicism, which provided a strong moral legitimacy for trade unions (Berman 2007; Misner 1991, 324). More fundamentally, the persistence of a belief in the necessity for independent collective representation can best be explained by workers’ experience of the employment relationship and its operation in a market system. Differing interests inherent in the employment relationship such as conflict round the wage/effort bargain, the commodity status of labor, and asymmetrical power relations creates the basis for trade-union demand. As these elements of the employment relationship are a constant in any market system, the demand for trade unions retains a continuing relevance for many workers (D’Art and Turner 2006).

A second question addressed was the relationship between the perceived need for trade unions and actual union membership. Employees who agreed with the need for strong unions were markedly more likely to be members of a union in those countries with a favorable institutional regime. The institutional measures of union presence and country cluster substantially account for the factors that determine why many employees with favorable union attitudes nevertheless fail to join a union. Structural factors had a more moderate impact on the likelihood of being in a union, with individual characteristics being the least significant. Consequently, among employees the extent to which a perceived need for strong trade unions converts into actual union membership appears to be critically dependent on a favorable and union-friendly institutional context. Historical trends in union membership numbers and density across the developed industrial societies reflect the main political and institutional developments (Western 1997). High union-density levels in countries such as Sweden and Denmark derive from political conditions that created a relatively benign institutional environment supportive of union membership and collective bargaining. A key feature of these regimes is the ready availability of trade unions across a broad range of workplaces in the public and private sectors of the economy.

The challenge for trade unions is to ensure a political, legal, and economic environment that facilitates employee choice of union membership. This entails a process of organizing, member activism, and union agitation that translates into a political capacity to influence or determine the institutional parameters and rules in which trade unions operate in a particular state.

Footnotes

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Bain

G. S.

Price

1983. “Union Growth: Dimensions, Determinants and Destiny.” In Industrial Relations in Britain, ed. Bain

G. S.

Oxford: Basil Blackwell.

Berman

2007. The Primacy of Politics: Social Democracy and the Making of the Twentieth Century. New York: Cambridge University Press.

Blanchflower

D. G.

Freeman

R. B.

1997. “The Attitudinal Legacy of Communist Labor Relations.” Industrial and Labor Relations Review 50 (3): 438-59.

Booth

1986. “Estimating the Probability of Trade Union Membership: A Study of Men and Women in Britain.” Economica 53 (209): 41-61.

Briskin

McDermott

1993. Feminism, Democracy, and Militancy. Toronto: University of Toronto Press.

Brown

1999. “Globalisation and the Political Economy of High Skills.” Journal of Education and Work, 12 (3): 233-51.

Bryson

Gomez

2005. “Why Have Workers Stopped Joining Unions? The Rise in Newer-Membership in Britain.” British Journal of Industrial Relations 41 (1): 67-92.

Calmsfors

Driffill

1988. “Centralisation of Wage Policy.” Economic Policy 6:13-61.

Cappelli

McKersie

1987. “Management Strategy in the Redesign of Work rules.” Journal of Management Studies 24 (5): 441-62.

10.

Crouch

1982. Trade Union: The Logic of Collective Action. London: Fontana.

11.

Crouch

1996. “Revised Diversity: From the Neo-Liberal Decade to beyond Maastricht.” In Industrial Relations in Europe: Traditions and Transitions, ed. Van Ruysseveldt

Visser

London: Sage.

12.

D’Art

Turner

1999. “An Attitudinal Revolution in Irish Industrial Relations: The End of Them and Us?” British Journal of Industrial Relations 37 (1): 101-16.

13.

D’Art

Turner

2003a. “Independent Collective Representation: Providing Effectiveness, Fairness, and Democracy in the Employment Relationship.” Employee Responsibilities and Rights Journal 15 (4): 169-81.

14.

D’Art

Turner

2003b. “Union Recognition in Ireland: One Step Forward or Two Steps Back?” Industrial Relations Journal 34 (3): 226-40.

15.

D’Art

Turner

2006. “New Working Arrangements: Changing the Nature of the Employment Relationship?” International Journal of Human Resource Management 17 (3): 523-38.

16.

D’Art

Turner

2007. “Trade Unions and Political Participation in the European Union: Still Providing a Democratic Dividend?” British Journal of Industrial Relations 45 (1): 103-36.

17.

Deery

De Cieri

1991. “Determinants of Trade Union Membership in Australia.” British Journal of Industrial Relations 29 (1): 51-73.

18.

Disney

1990. “Explanations of the Decline in Trade Union Density in Britain: An Appraisal.” British Journal of Industrial Relations 29 (2): 165-77.

19.

Doellgast

Greer

2007. “Vertical Disintegration and the Disorganization of German Industrial Relations.” British Journal of Industrial Relations 45 (1): 55-76.

20.

Eagly

A. H.

Chaiken

1993. The Psychology of Attitudes. Fort Worth, TX: Harcourt Brace Jovanovich College Publishers.

21.

Ebbinghaus

Visser

1999. “When Institutions Matter: Union Growth and Decline in Western Europe, 1950-1995.” European Sociological Review 15 (2): 135-58.

22.

Ebbinghaus

Visser

2000. Trade Unions in Western Europe since 1945. London: MacMillan.

23.

Ferner

Hyman

1992. Industrial Relations in the New Europe. Oxford: Blackwell.

24.

Fiorito

Gallagher

Greer

1986. “Determinants of Unionism: A Review of the Literature.” Research in Personnel and Human Resource Management 4, Greenwich, CT: JAI Press.

25.

Flavin

Pacek

Radcliff

2010. “Labor Unions and Life Satisfaction: Evidence from New Data.” Social Indicator Research 98 (3): 435-49.

26.

Fox

(1974) Beyond Contract: Work, Power, and Trust Relations. London: Faber and Faber.

27.

Freeman

Medoff

1984. What Do Unions Do? New York: Basic Books.

28.

Freeman

Pelletier

1990. “The Impact of Industrial Relations Legislation on British Union Density.” British Journal of Industrial Relations 28:141-64.

29.

Goetschy

Rozenblatt

1992. “France: The Industrial Relations System at a Turning Point?” In Industrial Relations in the New Europe, ed. Ferner

Hyman

Oxford: Blackwell.

30.

Green

1990. “Trade Union Availability and Trade Union Membership in Britain.” The Manchester School 58 (4): 378-94.

31.

Hague

Harrop

Breslin

1998. Comparative Government and Politics, 4th ed. London: Macmillan Press.

32.

Hall

Soskice

2001. Varieties of Capitalism: The Institutional Foundations of Comparative Advantage. Oxford: Oxford University Press.

33.

Hardiman

2000. “Social Partnership, Wage Bargaining, and Growth.” In Bust to Boom? The Irish Experience of Growth and Inequality, ed. Nolan

O’Connell

Whelan

Dublin: Institute of Public Administration.

34.

Hart Research Associates Inc. 1999. Americans’ Views on Key National Issues. Survey conducted on behalf of the AFL-CIO Washington DC. http://www.aflcio.org/labor99/am_attitude.htm.

35.

Hartley

1992. “Joining a Trade Union.” In Employment Relations, ed. Hartley

Stephenson

Oxford: Blackwell.

36.

Harvey

2005. A Brief History of Neoliberalism. Oxford: Oxford University Press.

37.

Hassel

2010. “Twenty Years after German Unification: The Restructuring of the German Welfare and Employment Regime.” German Politics & Society 28 (2): 102-15.

38.

Haynes

Vowles

Boxall

2005. “Explaining the Younger-Older Worker Union Density Gap: Evidence from New Zealand.” British Journal of Industrial Relations 43 (1): 93-116.

39.

Henley

Tsakalotos

1993. Corporatism and Economic Performance: A Comparative Analysis of Market Economies. Aldershot, UK: Edward Elgar.

40.

Hyman

2005. Trade Unions and the Politics of the European Social Model, Economic and Industrial Democracy 26 (1): 9-40.

41.

Hyman

Price

, eds. 1983. The New Working Class? White Collar Workers and their Organisations. London: Macmillan.

42.

Kaufman

2001. “The Future of US Private Sector Unionism: Did George Barnett Get it Right After All?” Journal of Labour Research 22 (3): 433-58.

43.

Kitschelt

Lange

Marks

Stephens

, eds. 1999. Continuity and Change in Contemporary Capitalism. Cambridge: Cambridge University Press.

44.

Klandermans

1992. “Trade Union Participation.” In Employment Relations—The Psychology of Influence and Control at Work, ed. Hartley

Stephenson

Oxford: Blackwell.

45.

Lawrence

1994. Gender and Trade Unions. London: Taylor and Francis.

46.

Lind

2009. “The End of the Ghent System as Trade Union Recruitment Machinery?” Industrial Relations Journal 40 (6): 510-23.

47.

Lockwood

1966. The Blackcoated Worker. London: Unwin University.

48.

Martin

Cristesco-Martin

1999. “Industrial Relations in Transformation: Central and Eastern Europe in 1998.” Industrial Relations Journal 30 (4): 387-404.

49.

Misner

1991. Social Catholicism in Europe: From the Onset of Industrialization to the First World War. London: Darton, Longman, and Tod.

50.

Pollert

1999. “Trade Unionism in Transition in Central and Eastern Europe.” European Journal of Industrial Relations 5 (2): 209-34.

51.

Pollin

2003. Contours of Descent: US Economic Fractures and the Landscape of Global Austerity. New York: Verso.

52.

Premack

Hunter

1988. “Individual Unionisation Decisions.” Psychological Bulletin 103 (2): 223-34.

53.

Radcliff

2005. “Class Organisation and Subjective Well-Being: A Cross-National Analysis.” Social Forces 84 (1): 513-30.

54.

Royle

2004. “Employment Practices of Multinationals in the Spanish and German Quick-Food Sectors: Low-Road Convergence?” European Journal of Industrial Relations 10 (1): 51-71.

55.

Schur

Kruse

1992. “Gender Differences in Attitudes towards Trade Unions.” Industrial and Labor Relations Review 46 (1): 89-02.

56.

Seeleib-Kaiser

Fleckenstein

2007. “Discourse, Learning, and Welfare State Change: The Case of German Labour Market Reforms.” Social Policy and Administration 41 (5): 427-48.

57.

Streeck

Thelen

2005. “Institutional Change in Advanced Political Economies.” In Beyond Continuity: Institutional Change in Advanced Political Economies, ed. Streeck

Thelen

Oxford: Oxford University Press.

58.

Toner

1989. “Women’s Views in Ireland on Unions.” Work and Occupations 16 (2): 214-24.

59.

Troy

1999. Beyond Unions and Collective Bargaining. New York: M.E. Sharpe.

60.

Troy

2004. The Twilight of the Old Unionism. New York: M.E. Sharpe.

61.

Turner

2002. “Corporatism in Ireland: A Comparative Perspective.” In Irish Employment Relations in the New Economy, ed. D’Art

Turner

Dublin: Blackhall Press.

62.

Turner

D’Art

2003. “The Feminization of Irish Trade Unions: Involvement, Solidarity, and the Relevance of Gender.” Women in Management Review 18 (5): 228-35.

63.

Verma

Kochan

2004. “Unions in the Twenty-First Century: Prospects for Renewal.” In Unions in the Twenty-First Century: An International Perspective, ed. Verma

Kochan

New York: Palgrave MacMillan.

64.

Visser

2006. “Union Membership Statistics in Twenty-Four Countries.” Monthly Labor Review, 129 (1): 38-49.

65.

Waddington

Whitson

1997. “Why Do People Join Unions in a Period of Membership Decline?” British Journal of Industrial Relations 35 (4): 515-46.

66.

Wajcman

2000. “Feminism Facing I.R. in Britain.” British Journal of Industrial Relations 38 (2): 183-01.

67.

Walters

2002. “Female Part-time Workers Attitudes to Trade Unions in Britain.” British Journal of Industrial Relations 40 (1): 49-48.

68.

Western

1997. Between Class and Market: Postwar Unionization in the Capitalist Democracies. Princeton, NJ: Princeton University Press.

69.

Wheeler

McClendon

1991. “The Individual Decision to Unionize.” In The State of the Unions, ed. Strauss

Gallagher

Fiorito

Madison, WI: Industrial Relations Research Association.