Week-5

Discrimination in hiring against immigrants and ethnic minorities: the effect of unionization

M. Harcourt a *, H. Lam

b , S. Harcourt

c and M. Flynn

d

a Department of Strategy and Human Resource Management, Waikato University, Hamilton, New Zealand;

b Centre for Innovative Management, Athabasca University, Alberta, Canada;

c Independent Researcher, Hamilton, New Zealand;

d Department of Human Resource

Management, Middlesex University, London, UK

There has been a long debate concerning whether unions are exclusive or inclusive with respect to immigrants and ethnic minorities. In the exclusive view of unions, unionization is expected to increase the likelihood of employers asking questions that discriminate against immigrants and ethnic minorities and decrease the likelihood of employers askingEqualEmployment Opportunity (EEO) questions related toimmigrant or ethnic minority status. The contrary is expected for the inclusive view. Our analysis, using New Zealand data for job applications, provides some support for the inclusive view of unions, as the higher the unionization rate, the more likely EEO information is sought, but no relationship is found between unionization rate and discriminatory questioning. This suggests that unions are probably helpful in promoting diversity but not yet in combating discrimination against immigrants and ethnic minorities in hiring.

Keywords: discrimination; immigrants; ethnic

Introduction

Worldwide, the workforce has become increasingly pluralistic and ethnically diverse as

more and more people migrate across nations. Although race- or ethnicity-based

discriminatory employment practices are prohibited in most developed countries, actual

discrimination is still widespread (Jain and Al-Waqfi 2000; Le and Kleiner 2000). Unions

have long been found to enhance equality and reduce discrimination within the workplace

by, for example, reducing the wage gap between visible minority and majority workers

(e.g. Ashenfelter 1972; Freeman and Medoff 1984; Metcalf, Hansen and Charlwood

2001). However, they have mostly focused on ‘treatment discrimination’ at the post-

employment stage (Jain and Al-Waqfi 2000). The unions’ role in ‘access discrimination’ at

the pre-employment stage is still controversial, as, in principle, two opposed views could

explain the union position.

Do unions want to embrace ethnic minorities and immigrant workers and help them

gain access to employment? This debate relates fundamentally to the arguments for the

monopoly and voice faces of unionism (Freeman and Medoff 1984) as well as the rationale

for craft versus industrial unionism (Ashenfelter 1972). According to the monopoly view,

a union’s main goal is to raise the wages of its members. If this is true, exclusionary

practices to constrain the number of alternative workers available for hire are likely to be

favoured. This is also in line with the craft union strategy of restricting entry into the trade

so as to push up wage levels. Accordingly, unions may lobby for limits to immigration,

ISSN 0958-5192 print/ISSN 1466-4399 online

q 2008 Taylor & Francis

DOI: 10.1080/09585190701763958

http://www.tandf.co.uk/journals

*Corresponding author. Email: [email protected]

The International Journal of Human Resource Management,

Vol. 19, No. 1, January 2008, 98–115

and be reluctant to offer help to ethnic minorities in obtaining employment. Ethnic

minorities are seen as a competitive threat to majority workers, in taking away the latter’s

jobs and driving down their wages. In contrast, if unions embrace a social democratic

role in trying to promote the welfare of all workers rather than only their members, then

inclusive practices are likely to be adopted (Lynk 2000; Godard 2000, p. 523). Such

practices are common to industrial unions, which are concerned with organizing across

occupational groups within an industry, or even across industry boundaries (Chaison and

Dhavale 1990). By organizing a wide range of workers, the union is able to ‘take wages

out of competition’, thereby simultaneously improving members’ terms and conditions

and the union’s chance of growth and survival.

Although there is no lack of studies on discrimination against minorities, systematic

empirical work specifically on unions’ stance and effect on ‘access discrimination’ against

ethnic minorities is scant. The current study further explores this relationship, using data

from New Zealand.

The diverse workforce

The workforce in most developed nations has become more diverse with rising global

migration. For example, in 2001, Hispanics represented about 13%, blacks 12.7%, and

Asians 4% of the US population (CBS News 2003). In a Census Bureau study

commissioned by the US Department of Commerce, 90% of projected total population

growth for 1995 to 2050 is expected to come from expansion of the minority population (He

and Hobbs 1999, p. 1). In Canada, according to the 2001 Census, total visible minorities

comprised about four million (13.4%) of the population, with Chinese and South Asians

being the largest minority categories (Statistics Canada 2003a). The same census indicated

that 18.4% of the population was foreign-born (Statistics Canada 2003b). In Europe, Lahav

(1997, p. 1) reported that there were approximately 15 million ‘foreigners, making

immigration a challenging issue for the European countries’. With the recent enlargement

of the European Union (EU), following the joining of ten formerly communist east

European nations, the UK prime minister admitted that immigration had ‘suddenly become

very high on the agenda’ because citizens were concerned about immigrants taking away

their jobs and/or claiming their social welfare (BBC News 2004). Some older EU members

have, as a result, secured the right to refuse work permits to people in the new member

nations for a transition period, which lasts up to seven years for countries like Germany and

Austria (EU Business 2004). In the South Pacific region, both Australia and New Zealand

have received large numbers of immigrants since the colonial days. Starting from the

1970s, more immigrants have been coming from the Asian areas and for New Zealand,

the Pacific Islands as well (Australian Bureau of Statistics 2001). More recently, New

Zealand reported almost 93,000 new permanent or long-term migrants (about 2.3% of the

population) for 2003, with net inflows from China, India, Japan, Fiji, South Africa and

the United States (Statistics New Zealand 2004). The population is expected to be more

ethnically diverse in the two decades ending 2021, with the Asian population projected to

have the highest growth rate of up to 120% (with a corresponding rise in the population

share from 7% to 13%) while the Pacific and Maori populations experience increases of

58% and 18%, respectively (Statistics New Zealand 2003c). Apparently, the demographic

trends and projections of developed nations point to further ethnic diversity in the future,

making this a critical issue for these nations, given the economic, social, and political

implications, and, in particular, the significant potential impact on the various parties of the

workforce – employers, unions and individual workers.

The International Journal of Human Resource Management 99

Employer discrimination theories

Despite the widespread enactment of anti-discrimination laws, discrimination against

minorities is ‘very much alive’ and is still ‘very obvious in the hiring and promotional

practices of United States companies’ (Le and Kleiner 2000, p. 98). If such discrimination

is so much alive in a nation which has been proactively promoting equal opportunities, one

would expect the problem to be comparable or even more serious in other ethnically

diverse nations.

According to Arrow (1998) and Jain and Al-Waqfi (2000), discrimination against

minorities can be explained by a number of theories. For example, Becker’s (1957) taste

theory posits that some employers have a prejudice or ‘taste for discrimination’ and are

even willing to pay a wage premium simply to avoid hiring members of so-called

undesirable groups. Workers may also have a taste for discrimination. Discriminating

members of the majority group may demand a wage premium if they have to work

alongside, at the same level as, or in a subordinate role to a minority worker. Employers

then avoid paying the wage premium by not hiring minority workers or by segregating

them into lower-paying, subordinate roles. Finally, customers may also have a taste for

discrimination. Discriminating members of the majority group may demand that they be

served by staff from the majority group, especially in prestigious, authoritative or skilled

roles. Employers then hire people accordingly, assigning the more public and important

positions to people from the majority group. Alternatively, they employ minority workers

in the skilled or responsible roles, but at lower wages that enable them to compete on

price in selling what prejudiced customers perceive to be an inferior product.

The statistical discrimination theory, on the other hand, is not based on preferences,

but on the employer’s lack of low-cost information about the true productivity

characteristics of individual job applicants. Rather than assess individuals one-at-a-time,

employers may find it easier and more cost-effective to screen applicants based on, for

example, the average productivity characteristics of their racial, educational, or gender

group (Phelps 1972; Fischer 1987; Arrow 1998). Hence, discrimination is the ‘rational

reactions of employers to different average productivity levels between easily identifiable

classes of job applicants’ (Tomaskovic-Devey and Skaggs 1999, p. 423). If employers

expect minority groups to have lower average productivity levels, they may prefer to hire

workers from the majority group, irrespective of an individual applicant’s productivity.

A second variant of statistical discrimination theory stresses the differences in

performance variance, rather than average performance levels, across groups.

If employers want to avoid hiring the particularly bad performers, they may prefer to

select from the group with a narrow performance distribution even when average

performance levels across groups are the same (Phelps 1972; Aigner and Cain 1977).

A third variant of statistical discrimination theory focuses on differences in the accuracy of

various selection methods across groups. If one group’s future performance can be more

accurately gauged through selection methods than another’s, employers may again prefer

to hire workers from the more predictable group (Phelps 1972; Aigner and Cain 1977;

Borjas and Goldberg 1978). In any form, statistical discrimination involves negative,

performance-related stereotyping. Such stereotypes may become a self-fulfilling prophecy

for minority workers, since discrimination lowers their rates of return to further education,

and so discourages them from pursuing additional qualifications (Coate and Loury 1993).

Error discrimination is similar to statistical discrimination: employers make critical

human resource decisions on the basis of perceived group characteristics. However, these

perceptions of group differences are erroneous. Either they do not exist at all or they are

M. Harcourt et al.100

greatly exaggerated. Even if real, intergroup differences are still dwarfed by intragroup

differences (England 1992).

Social capital theory explains discrimination in terms of social segregation and

deprivation (see, for example, Hayes James 2000; Metz and Tharenou 2001). In a world of

scarce employment opportunities, who you know and what they can do for you is very

important (Hayes James 2000). Minority groups have their own social networks, but these

are often not linked to management, the professions, or any positions of power and

influence in society. Minorities lacking in such connections may find it more difficult to

obtain references or win insider support for being hired or promoted. Even after hiring,

their outsider status may also make it harder to form a relationship with a mentor willing to

share knowledge and teach skills necessary for future promotion (Athey, Avery and

Zemsky 2000).

Marxian theory argues that discrimination is deliberately used to create divisions

within the working class, preventing workers from coalescing to take collective action

against employer exploitation. Discrimination means different treatment with respect to,

for example, hiring, firing, promotion, and pay for different religious, gender, and racial

groups. The privileged few, the labour aristocracy in the trades, professions, and

management, jealously guard their superior position, siding with owners against more

exploited workers. Minorities are particularly easy to exploit, given their often smaller

numbers and sometimes highly visible physical differences from other workers (Gordon

1972; Bowles and Gintis 1976; Borjas and Goldberg 1978; Roemer 1979).

Union role

What roles do unions play with respect to employer discrimination? Zeitlin and Weyher

(2001, p. 456) summarize the union influence this way: ‘Who fills the job vacancies is

decided by the employers, but within the kinds of constraints that unions are ready,

willing, and able to impose: they can try either to exclude black workers or to ensure that

employers treat them no differently than white workers in filling job vacancies’. Do unions

allay or accentuate employer discrimination? There are two opposing views on the union

position: the exclusive view and the inclusive view.

Exclusive practices

Why would unions want to exclude minorities? If unions are primarily concerned with

keeping members’ wages up, it is likely that they will want to exercise some form of

‘monopoly control’ over the supply of labour, where possible (Ashenfelter 1972).

Immigrants are seen to take away jobs from existing union members, and by providing

a ‘reserve army of low-wage workers’, decrease the unions’ bargaining power and

undermine the effectiveness of strike actions (Olzak 1989, p. 307; Daniels and Von Der

Ruhr 2003). An increase in labour supply inevitably drives down wages, assuming demand

remains the same. Immigrant workers are also regarded as less skilful or productive.

As such, including them as members affects the unions’ ability to obtain a productivity-

related wage premium. In addition, some unionists believe that immigrant workers are

easily manipulated by employers, unsupportive of unions and thus not easy to organize

(Defreitas 1993). According to Hunt and Rayside (2000, p. 405), race and multiculturalism

issues such as ‘occupational ‘ghettoization’, job discrimination, refugee and immigrant

status, sharp cultural and linguistic contrasts, and differences in religious practices’ are all

potentially divisive and so can pose an outright threat to union solidarity.

The International Journal of Human Resource Management 101

Unions also have few resources with which to serve the many and complex needs of a

diverse membership (Bertone and Griffin 1995). Reshef and Lam (1999, p. 125) outline

three types of security goals for the union. These are organizational, institutional, and

political security. Organizational security concerns the survival of the union; institutional

security relates to a union’s ability to withstand pressures that may undermine its

representational institutions (such as collective bargaining), whereas political security

involves how positively union leaders are seen by their members in terms of their ability

and integrity. Union leaders must prioritize their scarce resources to maximize their

security – to prevent losing members, to advance their collective bargaining capabilities,

and to promote a positive union image to their members. According to the median voter

theory, union leaders are likely to promote the needs of the median member who would

typically be a worker in the majority ethnic group. Hence, naturally, the concerns of

minority groups are likely to be neglected or given a lower priority than those of members

representing a larger proportion of the membership. Bertone and Griffin (1995) attribute

the lack of services for immigrant women in Australian unions to the attitude of senior,

full-time officials, most of whom are not members of minority groups and so not

sympathetic to minority needs. In sum, Bertone and Griffin (1995, p. 135) suggest that ‘the

provision of services is a political decision influenced by the needs of the members,

the resource available and the decisions of the officials’. Accordingly, the decision to focus

union services on the majority rather than minorities is politically-favoured, especially in

the short-term.

Jain and Sloane (1981) review the literature and identify five ways that unions have

discriminated against minority workers in the past. First, unions have excluded minority

workers from union membership, effectively denying them employment in what are often

high-paying jobs in workplaces with closed or union shops. Means of exclusion have

included: requiring new members to be sponsored by existing members, requiring new

members to sit an exam or test not relevant to job tasks, and failing to process minority

workers’ union membership application forms. Second, where they control access to

apprenticeship programmes, unions have often allocated places to trainees from the

majority rather than minority group, effectively denying the latter the opportunity to

advance to more highly skilled, higher-paying positions. Third, unions have often

segregated minority workers into separate locals, which are too small and poorly funded to

effectively press for improved terms and conditions. Fourth, unions have more

aggressively defended the contractual rights of majority than minority members, leaving

the latter more vulnerable to arbitrary and/or discriminatory mistreatment by management.

Fifth, unions have negotiated seniority rules for redundancy and promotion that have

tended to favour majority over minority workers, because discriminatory employers prefer

to hire majority workers ‘first’, giving them more seniority, and minority workers ‘last’.

Most of the evidence shows that, throughout history, unions have taken an antagonistic

approach towards immigration and the inclusion of minority groups. Hunt and Rayside

(2000, p. 401) conclude that ‘Canadian and American research finds that organized

labour’s engagement with race, ethnicity, gender and sexual orientation until recently has

been largely exclusionist’. They cite a number of studies to support this position. For

example, in the US, Foner (1974) found that trade unions were generally exclusionary

from the 1790s to the mid 1930s, at least as far as black workers were concerned. In Asher

and Stephenson’s (1990) edited book, which contains a collection of historical articles on

the topic, the 1940s and 1950s were described as a period when no forward movement was

made in unions’ stance on ethnic diversity. Hill (1985), looking at developments since the

1964 Civil Rights Act in the US, highlights union resistance to the legislation and general

M. Harcourt et al.102

defence of white male workers through support for discriminatory employer practices.

More recently, Riccucci (1990) lamented that, while some unions embrace equal treatment

of workers, there were many more which did not. In particular, the seniority system is seen

as perpetuating systemic discrimination against minority groups who are likely to be

newer entrants to the workforce. Studies have reported similar union positions for Canada,

especially earlier in the twentieth century when anti-Asian sentiment was high (Ward

1978). More recent work indicates that outright exclusionary practices, such as denying

union membership to minorities, still existed in the 1970s and that, until recently, unions

generally had not actively addressed minorities’ concerns (White 1993; Leah 1993).

A similar historical anti-immigration stance has been documented for Australia.

Quinlan (1989) contends that, for much of the twentieth century, trade unions in Australia

were strongly supportive of the White Australia Policy. In neighbouring New Zealand,

many citizens were said to ‘harbour suspicions about Asian migrants’ (James 1999, p. 22)

and Maori, the aboriginal people of New Zealand, have protested against the importation

of foreign workers (Newman 1995). Unfortunately, specific systemic research in New

Zealand on union policies and practices concerning immigrant workers is not available,

and so the general union position on ethnic diversity is far from clear.

Other authors have also found support for anti-immigration policies and union

exclusionism. Daniels and Von Der Ruhr (2003, p. 146), in their survey of ten advanced

economies, found that members of trade unions were more likely than nonmembers to

prefer restricting immigration. Moreover, Ashenfelter (1972, p. 435) found that craft

unions had the effect of lowering the ratio of black to white male wages by 5% as

compared with nonunion settings. This is not surprising as craft unions emphasize

organization along strict craft lines, adopt a more ‘job conscious’ philosophy, and tend to

be exclusionary (Godard 2000, p. 88). The tighter the unions’ control over the number of

skilled workers available, the higher the wages and benefits that can be negotiated. Such a

philosophy generally reflected the position of the American Federation of Labor (AFL)

prior to its merger with the Congress of Industrial Organizations (CIO) in 1955 (Godard

2000, p. 101). Gould (1977) argues that unions sometimes constitute a roadblock to equal

employment opportunity, even when they officially adopt an antidiscrimination stance.

Moreover, unions are commonly known to pay little attention to the plight of part-time

workers, a group in which minority workers are overly represented. As such, relative union

inaction in catering to the needs of part-timers is thus to some extent tantamount to the

exclusion of minorities.

In sum, the arguments for monopoly control, membership protection, leadership and

membership preference, as well as the historical empirical evidence described above,

all suggest that unions are ethnically exclusionary in their philosophy, policies and

practices. If this is true, we would expect unions to not object to, or even to covertly

support, employer discrimination against minorities. This would have some effect on

the hiring practices of employers, with the effect being greater the higher the level of

unionization.

Hypothesis 1: The higher the unionization rate of the workplace, the more likely it is

for the employer to discriminate against ethnic and immigrant minorities

in the hiring decision.

An exclusionary union position may also create a climate that hinders the promotion

of proactive human resource practices related to diversity. Most developed nations have

agencies or commissions that oversee fair treatment and employment representation

The International Journal of Human Resource Management 103

of minority workers under equal employment opportunity (EEO) principles. In New

Zealand, for instance, the EEO Trust is a nonprofit organization that promotes workplace

diversity through partnership, research, education, and information. While only the public

sector is required by legislation to abide by specific EEO requirements, private

organizations can become voluntary members if they want to be seen as actively involved

in EEO. As a result, these organizations may need to obtain EEO-related information in

their hiring process for reporting and planning purposes. However, in line with the

exclusionary view of unionism, we would expect the more unionized settings to be less

likely to embrace EEO and so less concerned about obtaining information for EEO

purposes.

Hypothesis 2: The higher the unionization rate of the workplace, the less likely it is for

the employer to seek EEO information in the hiring process.

Inclusive practices

Although there are many arguments put forth in support of exclusionism, most of them

have been contradicted by researchers and unionists alike. One such argument involves the

effect of immigration on unemployment. Contrary to traditional belief, Shan, Morris and

Sun (1999), using Australian and New Zealand data from 1983 to 1995, found no evidence

that immigration increases unemployment. Similar findings have been reported by Withers

and Pope (1985) for Australia. Given that immigrants are traditionally confined to low-

paying periphery jobs that do not appeal to natives, the immigrant threat to unemployment

is likely to be negligible. Indeed, Daniels and Von der Ruhr (2003, p. 146) note that ‘most

economic studies find small net gains in GDP per capita to host countries from increased

immigration and that past immigration has no obvious impact on unemployment in the host

country’.

As for the monopoly argument, limiting membership is not always, or even usually, an

effective way to boost wages. With advances in technology, many jobs have been

automated or deskilled and so employers are no longer as dependent on skilled

craftsmanship. More than perhaps ever before, employers can also make creative use of

nonunion people by, for example, hiring replacement workers in a strike, using temporary

workers from a recruitment agency, or by contracting out to nonunion firms. Employers

can also rely on market competition and the supply of cheaper nonunion labour to pressure

unions into accepting lower wages or moderating their demands for wage increases.

In contrast to exclusionism, it makes sense for unions to organize everyone, thereby

removing nonunion alternatives and ‘taking wages out of competition’.

Another recent concern to unions is the drastic decline in union membership. For

example, Farber and Western (2001, pp. 460–461) report that the union density rate

for nonagricultural employees in the US dropped from a high of 33.5% in 1954 to 13.3%

in 1998. Similarly, Crawford, Harbridge and Walsh (2000, p. 294) report that the

unionization rate in New Zealand declined from 44.7% in 1989 to 17% in 1999. The union

movement’s survival and growth (organizational security) thus depends on its ability to

expand the membership base. One way to do so is to include so-called nontraditional

workers who are largely comprised of minority groups. With the number of immigrants

fast increasing in many developed countries, minorities seem to be a logical target for

organizing efforts. As Hunt and Rayside (2000, p. 402) point out, ‘a growing union

recognition of diversity may be one of the ways the union movement is attempting to

reinvent and reposition itself for the twenty-first century’.

M. Harcourt et al.104

As for the view that ethnic and immigrant minorities are unsupportive of unions or

difficult to organize, evidence has often been to the contrary. Cohen and Kirchmeyer

(1994), comparing immigrants to the native-born in Israel, found that immigrants

participated more actively in unions and were also more committed. Defreitas (1993),

using data from a survey of youths in the US, found that immigrants had higher average

rates of unionization, with blacks exhibiting a noticeably stronger preference for union

representation. Funkhouser (1993, p. 259), analysing a 1983 Current Population Survey,

concluded that ‘recent immigrants to the US did not act differently about unionization

than . . . natives and, therefore, are as likely as young natives to be organized’. Similarly,

Kim and Kim (1997), doing a probit regression on a 1996 Current Population Survey,

found Asians were more likely to join unions that whites. Heywood and Mohtad

(1995) found, from their analysis of 748 unemployed workers in the US, that the union

applicant pool was comprised disproportionately of female and/or black workers, who

were apparently attracted to union settings. The rationale for this may well lie in the

instrumentality and self-protection concepts – that union jobs tend to pay a premium,

that ‘disadvantaged’ workers with weak bargaining power tend to have better protection

in union environments, and that they may benefit disproportionately from union attempts

to equalize wages across different employee classes (Kim and Kim 1997; Newton and

Shore 1992).

Furthermore, the social democratic political philosophy, widely supported by unions,

provides an ideological justification for inclusive recruitment practices. Lynk (2000), in

his Canadian study on union democracy, concludes that many unions venture beyond the

narrow concerns and activities of collective bargaining to embrace a broader, social

democratic focus on political and social causes. Social democracy ‘promotes reforms

within both the economy and society, in general, such as an improved workers

compensation system, reduced wage inequality within and across firms, and support for

human rights concerns . . . ’ (Godard 2000, p. 523). As such, one of the union movement’s

most important missions is the reduction of economic and social inequality and

corresponding alleviation of disadvantage. There is no room for exclusionary practices in

such a union role.

Plenty of empirical research work indicates that inclusive policies have been gaining

acceptance over time. Even as far back as the early twentieth century, prominent events

reveal just how much progress had already been achieved. For example, the 1919

Winnipeg General Strike in Canada was a classic example of workers coalescing against

employers, despite their ethnic divisions. The formation of the CIO in the US in 1937

signalled the emergence of a union philosophy radically different from the craft union

philosophy held by the AFL (Godard 2000, p. 97). The CIO called for the formation of

industrial unions which adopt inclusive rather than exclusive practices. Ashenfelter (1972)

suggests that industrial unions are less discriminatory (more egalitarian) towards

minorities. His study indicates that the ratio of black to white male wages was, in the

1960s, 4% higher in the industrial union sector than in nonunion settings. Leonard (1985)

found that, for 1974 to 1980, the employment share of black males increased more rapidly

in the union sector. This lends support to the argument that, even though union seniority

provisions disadvantage new entrants, unions have an overall positive effect on the

employment prospects of minority workers. There are also other positive indications that

US unions have been reaching out to immigrant and minority workers. For example, the

‘Justice for Janitors’ campaign in the US involved a broad union organizing drive of

custodian staff, comprised largely of minority members, to redress the power imbalance

between affluent building owners and struggling entry-level workers. As a result of this

The International Journal of Human Resource Management 105

major campaign, significant improvements in janitors’ contractual terms were obtained

across a number of cities (Wright 2001). More recently, the ‘Battle of Seattle’ in 1999

united some 30,000 unionists from many backgrounds to protest against the World Trade

Organization’s actions, which unions saw as infringing on human rights and threatening

the working class (Zeltzer 1999).

The union attitude towards minority issues has also changed in many other developed

nations during the twentieth century. In 1976, the Canadian Labour Congress issued

a National Anti-Racism Task Force report highlighting the racism problem, after which a

number of unions took more active measures to combat such discrimination (Hunt and

Rayside 2000). The Australian Council of Trade Unions, in the aftermath of World War II,

also became more receptive to immigrants. Unions started favouring the inclusion of

immigrants, having them ‘conform to existing union norms’ rather than excluding them

(Bertone and Griffin 1995, p. 120). Similarly, unions in Britain are now more proactively

addressing ethnic diversity issues by creating equal opportunity policies and making

antiracism statements (Virdee and Grint 1994).

In sum, with global migration an inevitable issue, many unions have begun to see

the inclusion of minorities as an opportunity to reinvent and re-position unions so as to

increase their membership base and bargaining power. Combating discrimination

against minorities is also the right thing to do from a social democratic perspective.

A quote from an interview with the Organizing Director of the South Florida Regional

Council of the United Brotherhood of Carpenters and Joiners of America neatly

summarizes the union position in the exclusion versus inclusion dilemma (Grenier and

Nissen 2000, p. 89):

Immigrants are positive and negative for our unions. It’s a double edge sword . . . we should exclude them . . . because our trade is being diluted. But if we keep them out, like the union has been doing, they become a double negative . . . If we go after them, it can revitalize us . . . give us purpose . . . the attitude [of exclusion] has changed.

With the shift in attitude towards inclusion, one could posit that unions would have a

positive effect in combating hiring discrimination against immigrant and ethnic minorities

by playing a watchdog role and by incorporating nondiscrimination employment

provisions in collective agreements. As a result, the antidiscrimination effect would,

contrary to that stated in hypothesis 1, be likely to be greater the more unionized the

workplace.

Hypothesis 3: The higher the unionization rate of the workplace, the less likely it is for

the employer to discriminate against ethnic and immigrant minorities

in the hiring decision.

The inclusive stance of unions may not be limited to the deterrence of discriminatory

practices. By raising workplace awareness of the value of inclusion and diversity, and

helping foster a culture receptive to it, unions could have an important effect in the

employers’ proactive embracement of EEO principles and practices. In line with this view,

the more unionized employers, contrary to hypothesis 2, are more likely to gather EEO-

related information for better diversity planning and any necessary reporting.

Hypothesis 4: The higher the unionization rate of the workplace, the more likely it is for

the employer to seek EEO information in the hiring process.

It will be interesting to see which of the two sets of competing hypotheses above is

supported by our empirical data.

M. Harcourt et al.106

Methods

Data for our dependent variables were obtained by perusing job application forms for

questions regarding nationality and ethnicity. The forms were acquired from organizations

that had advertised a vacancy in September, October or November 1999, on the internet or

in one of three major New Zealand newspapers: the New Zealand Herald, the Wellington

Dominion, or the Wellington Post. Only the 174 organizations that had instructed

interested job-seekers to complete a job application form were contacted. All 174 provided

their form. Comparable data were further obtained from 55 organizations participating in a

large-scale survey of human resource practices in small and medium-sized New Zealand

enterprises. Together, these sources yielded a total of 229 employers, of which two did not

reveal their identity to enable further data collection, and so were omitted from the sample.

These 227 employers collectively employ approximately 200,000 staff, in excess of 10%

of New Zealand’s workforce.

Organizations in our sample were contacted directly to ascertain whether they had a

human resource department.

The State Services Commission supplied a complete list of all agencies and

departments in the New Zealand civil service and covered by the 1988 State Sector Act.

The Registrar of Unions at the Department of Labour provided a complete list of New

Zealand trade unions and their contact details. All unions in existence at the end of 1999

were contacted for their membership numbers at each employer in the sample. All agreed

to provide the required membership data.

There are three dichotomous, dependent variables. The first is equal opportunities, the

second is immigration, and the third is discrimination. Each of these variables indicates

whether an employer asks a particular category of nationality/ethnicity question on its job

application form. Table 1 outlines the questions associated with each of these categories of

questions.

Equal opportunities assumes the value of ‘1’ if the employer asked a question about

the job applicant’s nationality or ethnicity for equal employment opportunities (EEO)

purposes and assumes the value of ‘0’ otherwise. Fifty-five (24.2%) employers in the

Table 1. Questions associated with the dependent variables.

Equal opportunities 1. What is your ethnicity (for EEO purposes)?* 2. What is your nationality (for EEO purposes)?* 3. What was your birthplace (for EEO purposes)?* 4. Which country are you a citizen of (for EEO purposes)?* Immigration 1. Are you a New Zealand citizen? 2. Are you a New Zealand resident? 3. Are you a New Zealand citizen or New Zealand resident? 4. Are you a holder of a valid New Zealand work visa/permit?

Discrimination What is your race? What is your ethnicity? What is your nationality? What was your birthplace? Which country are you a citizen of?

Note: * EEO questions were always identified as such and were often accompanied by detailed assurances that the information provided was for EEO purposes only, that it would not be used in the selection process.

The International Journal of Human Resource Management 107

sample asked one or more such questions. In line with hypotheses 2 and 4, we expect to see

some relationship between the level of unionization and the probability of an employer

asking one or more of these types of questions.

Immigration assumes the value of ‘1’ if the employer asked a question about whether

the job applicant had New Zealand citizenship, residency, or a valid work visa and

assumes the value ‘0’ otherwise. One hundred and nine (48%) employers in the sample

asked one or more such questions. It seems likely that most employers would ask questions

about an applicant’s immigration status to determine whether he or she has a legal right to

work in New Zealand. Although the answers to these questions can unnecessarily betray

someone’s nationality or ethnicity, discrimination is unlikely to be their primary or sole

underlying purpose. As a result, we do not expect the employer’s level of unionization to

have any effect on the likelihood of an immigration-oriented question.

Discrimination assumes the value of ‘1’ if an employer asks a question about the

applicant’s nationality or ethnicity and assumes the value of ‘0’ otherwise. Thirty-four

(15%) employers in our sample asked one or more such questions. None of these involves

immigration issues, such as New Zealand citizenship or residency, or an express EEO

purpose. As a result, we would expect employers to obtain information from such

questions for more invidious purposes related to discrimination. In line with hypotheses 1

and 3, we expect to see some relationship between the degree of unionization and the

likelihood of an employer asking at least one of these sorts of questions.

We use the variable union density to measure the level of unionization, as referred to in

hypotheses 1 to 4. It is expressed as the percentage of each employer’s workforce that is

unionized. In New Zealand, union membership is completely voluntary and unions may

only represent union members for collective bargaining purposes. As a result, we regard

union density as a good measure of union presence and influence in the workplace.

Two other variables, HR department and civil service, are included as controls. HR

department is a dichotomous variable. A ‘1’ value indicates that an employer had a human

resources department, whereas a value of ‘0’ indicates that it did not. Of the employers in our

sample, 151 (66.5%) had a human resources department. Human resource departments

ordinarily have professionally qualified and experienced staff. Their staff members are

therefore more likely than most managers to know about the legal entitlements to work, as set

out in the Immigration Act, and about the antidiscrimination provisions in the Human Rights

Act. For these reasons, we would expect employers with human resource departments to be

more likely to ask questions about an applicant’s immigration status (immigration) and less

likelytoask non-EEO questions about anapplicant’s nationality orethnicity(discrimination).

Civil service is also a dichotomous variable. A ‘1’ value indicates that an employer is

an agency or department of the New Zealand civil service, whereas a ‘0’ value suggests

otherwise. Twenty-nine (12.8%) of the employers in our sample belong to the New

Zealand civil service. Under the 1988 State Sector Act, all employers in the New Zealand

civil service are obliged to provide equal employment opportunities (EEO) as so-called

‘good employers’. The State Services Commission appraises each department’s progress

towards EEO goals, and uses this as one of several criteria in determining whether to

renew the chief executive’s fixed-term contract or provide him/her with an annual bonus.

Accordingly, we would expect civil service employers to be more likely to ask EEO-

related questions about an applicant’s nationality and ethnicity, but less likely to ask

comparable questions for discriminatory purposes.

The questions that comprise equal opportunities, immigration, and discrimination are

also used as dichotomous control variables in those models where each is not part of the

dependent variable. It is possible that the discrimination questions are merely clumsily

M. Harcourt et al.108

worded questions about immigration status. A seemingly discriminatory question about

nationality, for instance, could be a poorly worded attempt to assess whether the applicant

is a New Zealand citizen. Just as possibly, a seemingly discriminatory question about

nationality could be asked alongside a more neutral immigration question about whether

the applicant has a work visa, both for the purposes of assessing the applicant’s legal right

to work in New Zealand. In the former case, a discrimination question would be a

substitute for an immigration question and therefore negatively related. In the latter case,

a discrimination question would be a complement to an immigration question and

therefore positively related. However, the absence of any relationship would suggest that

discrimination questions were not being asked for immigration-related reasons.

Much the same holds for the relationship between discrimination and equal

opportunities. A discrimination question could be a crude attempt to collect information

for EEO purposes, in the absence of any explanation that this is being done. If so,

discrimination would be expected to be negatively (substitute) or positively (complement)

related to equal opportunities. However, the absence of such a relationship would suggest

that discrimination questions were being asked for reasons unrelated to EEO monitoring.

Legal right 1 is also used as a control variable. Legal right assumes the value of ‘1’ if an

employer asks whether an applicant has a legal right to work in New Zealand and ‘0’

otherwise. Although legal right involves no query about the applicant’s nationality, we

would nevertheless expect it to be negatively (substitute) or positively (complement)

related to equal opportunities, immigration, and discrimination to the extent that these

questions are about determining the applicant’s immigration status.

Summary statistics for our dependent and independent variables are provided in Table 2.

We use binomial logistic regression to analyse the data. Logistic regression estimates a

coefficient for each independent variable that best predicts the odds of an employer asking

the applicant at least one of the three types of question (equal opportunities, immigration,

and discrimination) on a job application form. The model logistic regression estimates has

the following form:

In (probability (question)/probability (no question)) ¼ b0 þ b1X1 þ b2X2 þ . . . . þ

bnXn. ‘Question’ in the model refers to an employer asking at least one of the questions

associated with equal opportunities, immigration, or discrimination, depending on which

of these is the dependent variable. In the other extreme, ‘no question’ in the model

indicates that the employer asked no questions related to the relevant dependent variable.

The probability of ‘question’ divided by the probability of ‘no question’ is an odds ratio,

and indicates the odds of a question being asked rather than not. The log (to the base e), or

ln, of this odds ratio is predicted by the Xs, the independent variables in the model.

The Betas (bs) are the coefficients and indicate the effect of a one-unit change in the

relevant independent variable on the log odds ratio.

Table 2. Descriptive statistics for the variables.

Frequency % of sample

Equal opportunities 55.0 24.2 Immigration 109.0 48.0 Discrimination 34.0 15.0 Union density 32.6* HR department 151.0 66.5 Civil service 29.0 12.8

Note: * mean.

The International Journal of Human Resource Management 109

Collinearity diagnostics revealed no obvious signs of multicollinearity among the

independent variables for the various models in the study. The variance inflation factors

(VIFs) for the Table 4 models were lower than 2.0 for all the independent variables except

resident (2.38) and work permit (2.06). The VIFs for the Tables 5 and 6 models were lower

than 2.0 for all the independent variables. In general, VIFs of 10.0 or higher indicate

serious multicollinearity problems (Neter, Kutner, Nachtsheim, and Wasserman 1996,

p. 387). Kendall tau-b correlation coefficients also show few signs of collinearity between

any two independent variables. The only correlations above 0.50 were for civil service and

EEO ethnicity (0.55), NZ citizen and resident (0.60), and resident and work permit (0.66).

Results

Results for the binomial regressions without controls for the equal opportunities,

immigration and discrimination questions are provided in Table 3. The parameter

estimates for the equal opportunities, immigration and discrimination binomial regression

models are outlined in columns 1, 2 and 3, respectively. The standard errors are provided

in parentheses. A positive parameter estimate suggests that a rise in the value of an

independent variable brings about a corresponding rise in the odds of an equal

opportunities, immigration, or discrimination question. A negative parameter estimate

suggests that a rise in the value of an independent variable brings about a fall in the odds of

such a question. However, only statistically significant effects are reported here.

Goodness of fit statistics for all three regression models are reported at the bottom of

each column. The Cox–Snell R square and the Nagelkerke R square both indicate that the

independent variables account for a lot of the uncertainty in the equal opportunities data

but not in the immigration and discrimination data.

The parameter estimates are all statistically significant in the equal opportunities model.

The intercept parameter indicates that the odds of a non-union employer, which has no

human resources department and is outside the civil service, asking a nationality- or

ethnicity-related EEO question on its job application form, to not asking any such question,

are 1:50. Consistent with expectation, the union density parameter estimate indicates

that the odds of an employer asking such a question rise 3% for each 1% rise in union density.

The HR department parameter suggests that the odds of such a question rise nearly three

times, if an employer has a human resources department. Similarly, the civil service

parameter suggests that the odds of a civil service employer asking a nationality- or ethnicity-

related EEO question are 26 times higher than for an employer outside the civil service.

In the immigration model, the union density parameter estimate is statistically

insignificant. Only the parameter estimate for civil service is significant. It indicates that

Table 3. Binomial logistic regression results without controls for equal opportunities, immigration, or discrimination.

Equal opportunities Immigration Discrimination

Intercept 23.876** (0.656) 20.383 (0.260) 21.299** (0.318) Union density 0.033** (0.007) 0.000 (0.004) 0.002 (0.006) HR department 1.069* (0.525) 0.272 (0.297) 20.895* (0.404) Civil service 3.256** (0.613) 0.891* (0.445) 0.265 (0.618) Cox–snell R square 0.328 0.030 0.022 Nagelkerke R square 0.490 0.039 0.038

Notes: Standard errors are shown in parentheses. * Statistically significant at the .05 level; ** Statistically significant at the .01 level (two-tailed tests).

M. Harcourt et al.110

the odds of a civil service employer asking an immigration question are more than twice as

high as for an employer outside the civil service.

In the discrimination model, the union density parameter estimate is, contrary to

expectation, statistically insignificant. However, the parameter estimates for the intercept

and HR department are both significant. The intercept parameter indicates that there are

4:1 odds against a non-union employer with no human resource department and outside

the civil service asking a potentially discriminatory nationality or ethnicity question on the

job application form. Conversely, the HR department parameter suggests that having a

human resource department more than halves the odds of asking such a question.

Table 4 outlines the parameter estimates for the equal opportunities (column 1) and

discrimination (column 2) models with individual immigration questions used as controls.

The first column shows that, of the new variables, only work permit appears to have a

statistically significant effect on the odds of an equal opportunities question. Specifically,

the odds of an employer asking an EEO question fall an estimated 80%, if an employer

asks whether or not the applicant has a work permit on the job application form.

The second column shows that, of the new variables, none has a statistically significant

effect on the odds of a potentially discriminatory question.

Table 5 contains the parameter estimates for the immigration (column 1) and

discrimination (column 2) models with the equal employment opportunity questions as

Table 4. Binomial regression results for job applications with immigration controls.

Equal opportunities Discrimination

Intercept 24.100** (0.753) 21.495** (0.390) Union density 0.038** (0.008) 0.002 (0.006) HR department 1.751** (0.665) 20.891* (0.435) Civil service 3.585** (0.686) 20.082 (0.679) NZ citizen 0.889 (0.695) 0.268 (0.552) Resident 0.364 (0.699) 1.016 (0.606) Resident or citizen 21.303 (0.943) 1.104 (0.699) Work permit 21.713* (0.716) 20.361 (0.551) Legal right 20.741 (0.572) 20.997 (0.676) Cox–Snell R square 0.359 0.071 Nagelkerke R square 0.536 0.124

Notes: Standard errors are shown in parentheses. * Statistically significant at the .05 level; ** Statistically significant at the .01 level (two-tailed tests)

Table 5. Binomial regression results for job applications with equal opportunity controls.

Immigration Discrimination

Intercept 20.392 (0.260) 21.296** (0.319) Union density 0.002 (0.005) 0.004 (0.006) HR department 0.299 (0.300) 20.868* (0.408) Civil service 1.114* (0.530) 0.657 (0.748) EEO ethnicity 20.334 (0.448) 20.671 (0.682) EEO nationality 0.393 (1.484) 23.436 (15.736) EEO birthplace 20.293 (0.852) 0.613 (1.200) Cox–Snell R square 0.034 0.028 Nagelkerke R square 0.045 0.050

Notes: Standard errors are shown in parentheses. * Statistically significant at the .05 level; ** Statistically significant at the .01 level (two-tailed tests).

The International Journal of Human Resource Management 111

controls. None of the parameter estimates for any of these new variables is statistically

significant in either model.

Table 6 contains the parameter estimates for the equal opportunities (column 1) and

immigration (column 2) models with the potentially discriminatory questions used as

controls. Of the new variables, only ethnicity appears to be statistically significantly related

to equal opportunities and only legal right appears to be statistically significantly related to

immigration. The parameter estimate for ethnicity suggests that the odds of an employer

asking an EEO question decline 95% if that employer asks a question about the applicant’s

ethnicity. The parameter estimate for legal right indicates that the odds of an employer

asking an immigration-related question are more than 95% lower, if that employer asks

whether the applicant has a legal right to work in New Zealand.

Discussion

We found some support for the view that unions are primarily inclusive in their treatment of

immigrant and ethnic minorities. More specifically, we found that the more unionized

employers were more likely, on their job application forms, to ask nationality and ethnicity-

related questions for equal employment opportunity purposes, thereby supporting

hypothesis 4 instead of hypothesis 2. This suggests that the more unionized employers are

more aware of EEO issues, more concerned that their hiring practices might discriminate

against minorities, and more prepared to take action to ameliorate such practices.

Unions may encourage employers to be more conscious of EEO issues for two

principal reasons. First, with their increasing concerns about declining membership

numbers, unions may be keen to recruit from among traditionally neglected minority

groups, who together represent an increasing proportion of the workforce. Second, many

unions adhere to a social democratic political philosophy which stresses the economic and

social equality of all individuals, regardless of nationality or ethnicity.

Not all of our results, however, support the inclusive view of unions. In particular, we

found no evidence of any link between union density and the likelihood of discriminatory

nationality- or ethnicity-related questions on job application forms. Hence, neither

hypothesis 1 nor hypothesis 3 is supported. Likewise, we uncovered no connection

between union density and the likelihood of what would normally be regarded as

immigration-related questions about nationality. Our additional statistical models, where

nationality, ethnicity and EEO questions are used as control variables, may suggest why.

Table 6. Binomial regression results for job applications with discrimination controls.

Equal opportunities Immigration

Intercept 23.870** (0.683) 20.394 (0.308) Union density 0.035** (0.007) 0.001 (0.005) HR department 1.159* (0.566) 0.997** (0.352) Civil service 3.558** (0.691) 1.116* (0.546) Ethnicity 23.084* (1.446) 0.926 (0.874) Nationality 0.075 (1.048) 0.424 (0.741) Birthplace 20.549 (1.204) 1.375 (0.727) Citizenship 20.057 (1.203) 20.259 (0.844) Legal right 20.305 (0.471) 22.589** (0.435) Cox–Snell R square 0.347 0.243 Nagelkerke R square 0.518 0.32

Notes: Standard errors are shown in parentheses. * Statistically significant at the .05 level; ** Statistically significant at the .01 level (two-tailed tests).

M. Harcourt et al.112

Some of the ostensibly discriminatory questions may be complements to (positively

associated), or substitutes (negative associated) for, more obviously legitimate questions

about either EEO or immigration issues. As a result, what looks like an attempt to

discriminate, on the surface, could instead be a poorly worded attempt to seek information

for EEO or immigration reasons. For instance, ethnicity-related questions are usually

asked when EEO questions are not, suggesting that the former maybe being used as a crude

substitute for the latter. Alternatively, we must acknowledge that highly discriminating

employers are unlikely to ask EEO questions, and it may be this that accounts for the

negative association between ethnicity-related questions and EEO questions.

Implications and conclusions

Our findings suggest that unions are not exclusionary, in the sense that they do not seem

to foster discriminatory hiring practices. Indeed, they even have a positive effect on

employers’ adoption of EEO practices, which tends to support an inclusion stance. Yet,

they do not seem to have a significant impact on reducing discriminatory questioning in

hiring. As raising awareness on the value of diversity is probably the first step in the

antidiscrimination crusade, the results revealed in this study are promising. It may mean

that unions are doing a good job in promoting inclusion, but their efforts have yet to extend

to combating discrimination against minorities in hiring. If this is the case, unions should

be legislatively empowered and resourced to take a more active role in enforcing

antidiscrimination legislation and in making employment decisions so that they have

fewer discriminatory consequences.

As acknowledged above, some seemingly discriminatory questions may be clumsily

worded attempts at obtaining information for EEO or immigration purposes. If true, this

would suggest that many more managers and human resource practitioners need to be

educated about Human Rights Act requirements and EEO principles. It would also suggest

that union officials might also benefit from similar educational efforts, especially if they

are to play a watchdog role in encouraging diversity and driving out discrimination.

Note

1. Legal right is not subsumed within immigration because it is always a lawful question. It cannot be used to unlawfully identify and then discriminate against people born in other countries.

References

Aigner, D., and Cain, G. (1977), “Statistical Theories of Discrimination in Labor Markets,” Industrial and Labor Relations Review, 30, 1975–1987.

Arrow, K.J. (1998), “What has Economics to say about Racial Discrimination?,” The Journal of Economic Perspectives, 12, 2, 91–100.

Ashenfelter, O. (1972), “Racial Discrimination and Trade Unionism,” Journal of Political Economy, 80, 3, 435–464.

Asher, R., and Stephenson, C. (Eds.) (1990), Labor Divided: Race and Ethnicity in United States and Labor Struggles 1935–1960, Albany, NY: State University of New York Press.

Athey, S., Avery, C., and Zemsky, P. (2000), “Mentoring and Diversity,” American Economic Review, 90, 4, 765–786.

Australian Bureau of Statistics (2001), “Populations of Australia and New Zealand: A Comparison,” Australian Demographic Statistics, March Quarter 2001, (Catalogue. No. 3101.0).

BBC News (2004, 27 April), “Blair Pledges Immigration Review,” retrieved on 4 May 2004 from http://www. news.bbc.co.uk/1/hi/uk_politics/3661261.stm

Becker, G.S. (1957), The Economics of Discrimination, Chicago, IL: The University of Chicago Press. Bertone, S., and Griffin, G. (1995), “Immigrant Female Workers and Australian Trade Unions,” Relations

Industrielles/Industrial Relations, 50, 1, 117–145.

The International Journal of Human Resource Management 113

Borjas, G., and Goldberg, M. (1978), “Biased Screening and Discrimination in the Labor Market,” American Economic Review, 68, 5, 918–922.

Bowles, S., and Gintis, H. (1976), Schooling in Capitalist America: Educational Reform and the Contradictions of Modern Life, Princeton, NJ: Princeton University Press.

CBS News (2003, 21 January), “Hispanics now Largest US Minority,” CBSNEWS, retrieved on 19 April 2004 from http;//www.cbsnews.com/stories/2002/01/21/national/main537369/shtml

Chaison, G., and Dhavale, D.G. (1990), “The Changing Scope of Union Organizing,” Journal of Labor Research, 11, 3, 307–321.

Coate, S., and Loury, G. (1993), “Will Affirmative-Action Policies Eliminate Negative Stereotypes?”, American Economic Review, 83, 5, 1220–1240.

Cohen, A., and Kirchmeyer, C. (1994), “Unions and Ethnic Diversity: The Israeli Case of East European Immigrants,” Journal of Applied Behaviorial Science, 30, 2, 141–158.

Crawford, A., Harbridge, R., and Walsh, P. (2000), “Union and Union Membership in New Zealand: Annual Review for 1999,” New Zealand Journal of Industrial Relations, 25, 3, 291–302.

Daniels, J.P., and Von Der Ruhr, M. (2003), “The Determinants of Immigrant Policy Preferences in Advanced Economies: A Cross-country Study,” Atlantic Economic Journal, 31, 2, 146–158.

Defreitas, G. (1993), “Unionization Among Racial and Ethnic Minorities,” Industrial and Labor Relations Review, 46, 2, 284–301.

England, P. (1992), Comparable Worth: Theories and Evidence, Hawthorne, NY: Aldine de Gruyter. EU Business (2004, 4 May), “EU Newcomers Urge End to Worker Restrictions,” retrieved on 4 May 2004 from

http://www.eubusiness.com/afp/040504161955.bu573r44 Farber, H.S., and Western, B. (2001), “Accounting for the Decline of Unions in the Private Sector, 1973–1998,”

Journal of Labor Research, 22, 3, 459–485. Fischer, C.C. (1987), “Toward a More Complete Understanding of Occupational Sex Discrimination,” Journal of

Economic Issues, 21, 1, 113–138. Foner, P.S. (1974), Organized Labor and the Black Worker 1619–1973, New York: Praeger. Freeman, R.B., and Medoff, J.L. (1984), What Do Unions Do?, New York: Basic Books. Funkhouser, E. (1993), “Do Immigrants have Lower Unionization Propensities than Natives?,” Industrial

Relations, 32, 2, 248–261. Godard, J. (2000), Industrial Relations, the Economy, and Society, North York, Ontario: Captus Press. Gordon, D. (1972), Theories of Poverty and Unemployment, Lexington, MA: Heath. Gould, W. (1977), Black Workers in White Unions: Job Discrimination in the United States, New York: Cornell

University Press. Grenier, G., and Nissen, B. (2000), “Comparative Union Response to Mass Immigration: Evidence from an

Immigrant City,” Critical Sociology, 26, 1/2, 82–91. Hayes James, E. (2000), “Race-related Differences in Promotions and Support: Underlying Effects of Human and

Social Capital,” Organization Science, 11, 5, 493–508. He, W., and Hobbs, F. (1999), “The Emerging Minority Marketplace–Minority Population Growth: 1995 to

2050,” US Department of Commerce. www.mbda.gov/documents/mbdacolor.pdf (Retrieved 16 April 2004). Heywood, J.S., and Mohtad, H. (1995), “The Racial Composition of Unemployment: The Role of Unions,”

The Manchester School, 63, 2, 175–195. Hill, H. (1985), Black Labor and the American Legal System: Race, Work, and the Law, Madison, WS: University

of Wisconsin Press. Hunt, G., and Rayside, D. (2000), “Labor Union Response to Diversity in Canada and the United States,”

Industrial Relations, 39, 3, 401–444. Jain, H., and Al-Waqfi, M. (2000), “Racial Discrimination in Employment in Canada,” Proceedings of the

XXXVIIth Conference of the Canadian Industrial Relations Association, 47–63. Jain, H., and Sloane, P. (1981), Equal Employment Issues: Race and Sex Discrimination in the United States,

Canada, and Britain, New York: Praeger. James, C. (1999), “Migrant Repellent,” Far Eastern Economic Review, 162, 27, 22. Kim, D., and Kim, S. (1997), “The Effects on Union Membership of Race and Immigrant Status,” Journal of

Applied Behaviorial Science, 30, 3, 378–396. Lahav, G. (1997), “Ideological and Party Constraints on Immigration Attitudes in Europe,” Journal of Common

Market Studies, 35, 3, 377–406. Le, P.M., and Kleiner, B.H. (2000), “A Review of Current Empirical Research Concerning Race Discrimination

at Work,” Equal Opportunities International, 19, 6/7, 98–100. Leah, R. (1993), “Black Women Speak Out,” in Women Challenging Unions: Feminism, Democracy, and

Militancy, eds. L. Briskin and P. McDermott, Toronto: University of Toronto Press, pp. 157–171. Leonard, J.S. (1985), “The Effect of Unions on the Employment of Blacks, Hispanics, and Women,” Industrial

and Labor Relations Review, 39, 1, 115–132. Lynk, M. (2000), “Union Democracy and the Law in Canada,” Journal of Labor Research, 21, 1, 37–63. Metcalf, D., Hansen, K., and Charlwood, A. (2001), “Unions and the Sword of Justice: Unions and Pay Systems,

Pay Inequality, Pay Discrimination and Low Pay,” National Institute Economic Review, 17, 6, 61–75. Metz, I., and Tharenou, P. (2001), “Women’s Career Development: The Relative Contribution of Human and

Social Capital,” Group and Organization Management, 26, 3, 312–342.

M. Harcourt et al.114

Neter, J., Kutner, M., Nachtsheim, C., and Wasserman, W. (1996), Applied Linear Models (4th ed.), Chicago, IL: Irwin.

Newman, S. (1995), “Population growth in the Pacific,” Viewpoints, 95, 2, 6–8. Newton, L.A., and Shore, L.M. (1992), “A Model of Union Membership: Instrumentality, Commitment and

Opposition,” Academy of Management Review, 17, 2, 275–298. Olzak, S. (1989), “Labor Unrest, Immigration, and Ethnic Conflict in Urban America, 1880–1914,” The American

Journal of Sociology, 97, 6, 1303–1333. Phelps, E. (1972), “The Statistic Theory of Racism and Sexism,” American Economic Review, 64, 59–61. Quinlan, M. (1989), “Unions and Immigrants: The Post Second World War Experience,” in Australian Unions.

(2nd ed.), eds. B. Ford and D. Plowman, Melbourne: Macmillan, pp. 203–224. Reshef, Y., and Lam, H. (1999), “Union Responses to Quality Improvement Initiatives: Factors Shaping Support

and Resistance,” Journal of Labor Research, 20, 1, 111–131. Riccucci, N.M. (1990), Women, Minorities, and Unions in the Public Sector, New York: Greenwood Press. Roemer, J. (1979), “Divide and Conquer: Microfoundations of a Marxian Theory of Wage Discrimination,”

Bell Journal of Economics, 10, 695–705. Shan, J., Morris, A., and Sun, F. (1999), “Immigration and Unemployment: New Evidence from Australia and

New Zealand,” International Review of Applied Economics, 13, 2, 253–260. Statistics Canada (2003a, 23 April), “Visible Minority Population, Provinces and Territories: 2001 Census,”

retrieved 16 April 2004 from http://www.statcan.ca/english/Pgdb/demo40a.htm Statistics Canada (2003b, 12 May), “Proportion of Foreign-born Population, Provinces and Territories,” retrieved

16 April 2004 from http://www.statcan.ca/english/Pgdb/demo46a.htm Statistics New Zealand (2003c, 12 December), “Projections Overview: Ethnic Population Projections,” retrieved

23 April 2004 from www.stats.govt.nz/domino/external/web/prod_serv.nsf/htmldocs/projectionsþ overview#EthnicPopulationProjections

Statistics New Zealand (2004, 4 February), “Visitors Arrivals up 12 Percent. External Migration December 2003 – Media Release,” retrieved 16 April 2004 from http://www.stats.govt.nz/domino/external/pasfull/ pasfull.nsf/web/MediaþReleaseþExternalþMigrationþDecemberþ2003?open

Tomaskovic-Devey, D., and Skaggs, S. (1999), “An Establishment-level Test of the Statistical Discrimination Hypothesis,” Work and Occupations, 26, 4, 422–445.

Virdee, S., and Grint, K. (1994), “Black Self-organization in Trade Unions,” The Sociological Review, 42, 2, 202–225.

Ward, W.P. (1978), White Canada Forever: Popular Attitudes and Public Policy towards Orientals in British Columbia, Montreal: McGill-Queen’s University Press.

White, J. (1993), Sisters and Solidarity: Women and Unions in Canada, Toronto: Thomson Educational Publications.

Withers, G., and Pope, D. (1985), “Immigration and Unemployment,” Economic Record, 61, 554–563. Wright, C. (2001), “A Clean Sweep: Justice for Janitors,” Multinational Monitor, 22, 1/2, 12–14. Zeitlin, M., and Weyher, L.F. (2001), “‘Black and White, Unite and Fight:’ Interracial Working-class Solidarity

and Racial Employment Equality,” The American Journal of Sociology, 107, 2, 430–467. Zeltzer, S. (1999), “Global Victory for Workers in Seattle: Workers, Students, Activists Rout WTO,” Labournet,

retrieved 10 February 2004 from http://www.labournet.net/docks2/9912/wto37.htm

The International Journal of Human Resource Management 115