09e4150d193158d4c3000000 for aminah

33
Age stereotypes and discriminatory attitudes towards older workers: An East–West comparison Warren C.K. Chiu, Andy W. Chan, Ed Snape and Tom Redman ABSTRACT This study compared age stereotypes among 567 respondents sampled in the UK and Hong Kong and examined how these stereotypes were related to discriminatory attitudes at work. Compared to the Hong Kong sample, UK respondents saw older workers as more effective at work, but less adaptable to change. As expected, respondents’ own age was predictive of positive age stereotypes, although for supervisors this relationship was moderated in the case of perceptions of work effectiveness. Stereotypical beliefs were found to significantly affect respon- dents’ attitudes towards the training, promotion and retention of older workers, their willingness to work with older workers, and their support for positive discrimination. Findings also suggest that anti-age discrimination policies in the respondent’s organization had a positive impact on beliefs about the adapta- bility of older workers and possibly also on attitudes towards providing them with training. Implications of the findings were discussed in light of the existing socio-political environment in the UK and Hong Kong. KEYWORDS age discrimination age stereotypes equal opportunities human resource management 629 Human Relations [0018-7267(200105)54:5] Volume 54(5): 629–661: 017086 Copyright © 2001 The Tavistock Institute ® SAGE Publications London, Thousand Oaks CA, New Delhi

description

Age and resistance to change

Transcript of 09e4150d193158d4c3000000 for aminah

  • Age stereotypes and discriminatoryattitudes towards older workers: AnEastWest comparisonWarren C.K. Chiu, Andy W. Chan, Ed Snape and Tom Redman

    A B S T R AC T This study compared age stereotypes among 567 respondentssampled in the UK and Hong Kong and examined how thesestereotypes were related to discriminatory attitudes at work.Compared to the Hong Kong sample, UK respondents saw olderworkers as more effective at work, but less adaptable to change.As expected, respondents own age was predictive of positiveage stereotypes, although for supervisors this relationship wasmoderated in the case of perceptions of work effectiveness.Stereotypical beliefs were found to significantly affect respon-dents attitudes towards the training, promotion and retentionof older workers, their willingness to work with older workers,and their support for positive discrimination. Findings alsosuggest that anti-age discrimination policies in the respondentsorganization had a positive impact on beliefs about the adapta-bility of older workers and possibly also on attitudes towardsproviding them with training. Implications of the findings werediscussed in light of the existing socio-political environment inthe UK and Hong Kong.

    K E Y W O R D S age discrimination age stereotypes equal opportunities

    human resource management

    6 2 9

    Human Relations

    [0018-7267(200105)54:5]

    Volume 54(5): 629661: 017086

    Copyright 2001

    The Tavistock Institute

    SAGE Publications

    London, Thousand Oaks CA,

    New Delhi

    04chiu (ds) 29/3/01 7:35 am Page 629

  • Introduction

    Older workers have often encountered age discrimination from employers,for example through explicit age bars on job advertisements and lessfavourable treatment in recruitment, promotion and training (see, forexample: Walker & Taylor, 1993). A possible reason for this discriminationis that older workers may be perceived as having certain undesirable charac-teristics, including poor health, an inflexible attitude, resistance to changeand low trainability. There is now a significant amount of research on suchstereotypical beliefs in Western societies (see, for example: Doering et al.,1983; Hassell & Perrewe, 1995; Institute of Personnel Management, 1993;Kite & Johnson, 1993; Metcalf & Thompson, 1990; Taylor & Walker, 1992;Tillsley, 1990). Such beliefs about older workers, held by work colleagues,employers and by customers, may influence employment decisions and thusgive rise to age discrimination.

    This article deals with three basic questions. First, to what extent arethe kinds of stereotypical beliefs about older workers found in Westernsocieties also characteristic of Hong Kong? Here, we are interested in the rela-tive strength of such beliefs in Western and Eastern societies. Second, whatother factors appear to influence such beliefs? Much of the existing literaturefocuses on individual-level variables, whilst in this article we also examinecountry and organizational-level variables. Third, are stereotypical beliefsrelated to discriminatory attitudes towards older workers, in particular con-cerning their treatment on employment matters, and does this relationshipdiffer between East and West?

    We attempt to answer these questions through an analysis of surveydata collected in the UK and Hong Kong. Whilst the choice of the UK andHong Kong reflect the location of the researchers involved, and in that senseare chance selections, nevertheless the two serve to represent some of the keydifferences between Western and Eastern societies, respectively. The UKhas been characterized as an individualistic society with a relatively low toler-ance for inequalities of power and income (Hofstede, 1980). In contrast,Hong Kong has been seen as having many of the key characteristics of aChinese society, in particular a high degree of collectivism and a strongrespect for hierarchy (Bond, 1996; Hofstede, 1980). Thus, the UKHongKong comparison promises to provide some useful insights into EastWestdifferences in attitudes to older workers. Of course, Hong Kong has recentlyemerged from 150 years of British colonial rule and it is likely that this willhave left its mark on Hong Kong society. Furthermore, as a now advancedeconomy with extensive contact with international business, Hong Kong isarguably more likely than some other Asian societies to be subject to forces

    Human Relations 54(5)6 3 0

    04chiu (ds) 29/3/01 7:35 am Page 630

  • of modernization and convergence (Sparrow et al., 1994). Thus, our HongKongUK comparison may understate the attitudinal differences betweenEast and West which might be found by observing other Asian societies. Tothe extent that this is true, the present comparison represents a conservativetest of the impact of culture on ageist attitudes.

    After a brief discussion of the policy debates in Hong Kong and theUK, this article examines the literature on stereotypical beliefs about olderworkers. We derive a set of hypotheses, which are then tested using our HongKong and UK survey data. The article concludes with a discussion of theimplications of our findings.

    Age discrimination in Hong Kong

    One of the principles of traditional Chinese culture is to respect ones elders(Mencius, 1, 1, 7, 12; see Chen, 1990: 211). However, in employmentrelationships, discrimination on the grounds of age is a long-neglected issue,both in the private and the public sectors. In the 1995 Governors employ-ment summit, comprising representatives from employers, unions andgovernment, union leaders quoted cases of employers preferring to recruityounger applicants and selectively retrenching older workers (South ChinaMorning Post, 8 November 1995: 8; South China Morning Post, 10 Novem-ber 1995: 4), and union officials and leaders of grassroots societies urgedthe Hong Kong government to take action to protect older workers dis-placed from their jobs due to economic restructuring (Christian Times, 1May 1994: 1).

    In keeping with its laissez-faire approach to the economy, Hong Konghas been slower than the UK to develop legislation on equal opportunities.The government had no concrete proposals until 1993, when it published aconsultative Green Paper on Equal Opportunities For Women and Men.Even then, the paper focused mainly on gender and disability, rather than onage. Whilst 1995 saw the enactment of Sex Discrimination and DisabilityDiscrimination Ordinances, the bill covering age discrimination was voteddown (South China Morning Post, 29 July 1995: 1). This reflected a viewamongst legislators, especially those representing an employer perspective,that such legislation would constitute an unnecessary interference withmanagement discretion in recruitment and selection. Such legislation thatdoes exist may even encourage age discrimination in recruitment. Under theEmployment Ordinance, employers are required to pay higher rates of long-service payment to those employees aged 45 or above when they are dis-missed (Labour Department, 1997: 34), thus further discouraging employersfrom employing older workers in the first place.

    Chiu et al. Age stereotypes and discrimination against older workers 6 3 1

    04chiu (ds) 29/3/01 7:35 am Page 631

  • In addition to legislation, there have also been developments of a volun-tary nature. To cope with the challenge of equal opportunities, personnel andhuman resource management specialists professional bodies have started toadvise members to consider the issue of equal employment opportunities andto try to develop policies to reduce discriminatory practices (Hong KongInstitute of Human Resource Management, 1995, Clauses III7, 8 andVIII4; Hong Kong Industrial Relations Association, 1995, Clause I4).Whilst such developments are still quite recent, a survey of job advertisementsconducted in October 1995 found that only 14 percent of advertisementscontained explicit age requirements and comparison with similar earliersurveys suggests that this percentage has been falling (Wong, 1995). It seemsthat at least some human resource managers are becoming more aware of theimportance of providing equal employment opportunities.

    Age discrimination in the UK

    There is evidence of persistent and widespread discrimination against olderworkers in both the private and public sectors in the UK (Thompson, 1991;Walker & Taylor, 1993). However, although the UK has well-established legis-lation on sex and race discrimination, and has recently passed legislation ondisability, it lags behind countries such as the USA, Canada, Spain, France, NewZealand and Australia in having no laws prohibiting discrimination on groundsof age. Those in favour of such legislation have argued that it is needed to high-light the problem of age discrimination and to provide redress for thoseaffected. Opponents, particularly employers organizations such as the Con-federation of British Industry and the Chartered Institute of Personnel andDevelopment, prefer voluntary self-regulation and claim that legislation wouldbe difficult to enforce, would do little to change attitudes about older workersand would pose an additional administrative burden on employers.1 Some UKemployers have argued that legislation is unnecessary, since demographic andlabour market changes have obliged them to reconsider any negative attitudesto older workers, whilst a growing number of the over 50s are in any casetaking early retirement (McKay, 1998). However, the extent to which suchearly retirements are voluntary is open to question, with suggestions thatemployers are targeting older employees for redundancy (Taylor & Walker,1997), which may itself point to a need for protective legislation.

    Conservative governments throughout the 1980s and 1990s werefirmly opposed to legislation to prohibit age discrimination, and severalprivate members bills on the issue failed to win the support of the Houseof Commons. The Government also resisted European legislation on agediscrimination. In 1996, a proposal was made to include a clause on age

    Human Relations 54(5)6 3 2

    04chiu (ds) 29/3/01 7:35 am Page 632

  • discrimination and disability in a European Treaty on discrimination butfollowing representations from the UK and Denmark the clause was with-drawn. The election of a Labour government in May 1997 initiallyappeared likely to lead to legislation but by 1998 it was favouring a volun-tary approach. In the report Action on Age, produced in consultation withthe then Institute of Personnel and Development, the Confederation ofBritish Industry, the Trades Union Congress and Age Concern, theGovernment sets out a strategy based on encouraging rather than enforc-ing good practice (Department for Education and Employment, 1998).Upper age limits are to be removed from vacancies advertised through JobCentres and the Governments New Deal, aimed to encourage the unem-ployed into work, contained incentives to encourage the employment ofolder workers. The Labour Government did not rule out legislation andEurope provides momentum in this direction as the EU Anti-DiscriminationDirective, agreed in 2000, includes an article paving the way for membersto introduce anti-ageism legislation by 2006.

    Literature and hypotheses

    Stereotypical beliefs about older workers

    Tillsley (1990), drawing on her own and other researchers interview andsurvey work, compiled a list of age stereotypes held by employers. Thus,whilst older people are seen as more loyal and reliable, they are also seenas being less flexible and motivated, with outdated skills and poor healthand fitness. These findings are corroborated in a study sponsored by thethen Institute of Personnel Management (Warr & Pennington, 1993),involving a postal survey of 1140 UK personnel managers. Workers overthe age of 40 were seen as having more experience, being more loyal, reli-able, interpersonally skilled, conscientious, confident and more effective intheir jobs, and thinking more before they act, not taking things so easy andworking better in teams. Negative stereotypes included their being lessready to accept new technology, less adaptable to change, less able to learnand grasp new ideas, and being less interested in training. A factor analy-sis revealed two underlying dimensions to personnel managers attitudestowards people over 40: a work effectiveness factor which consisted ofthe first set of positive views, and a more negative maladaptability factor.Strong support for the existence of age-related stereotypes amongst bothworkers and managers/supervisors has emerged from other studies in boththe UK (Lyon & Pollard, 1997; Metcalf & Thompson, 1990; Taylor &Walker, 1992), and in the USA (Doering et al., 1983; Rosen & Jerdee,

    Chiu et al. Age stereotypes and discrimination against older workers 6 3 3

    04chiu (ds) 29/3/01 7:35 am Page 633

  • 1976a), although the detailed stereotypes vary with the exact definition ofolder workers used.

    We would like to emphasize that our concern in this article is withstereotypical beliefs rather than necessarily with actual job performance. Infact, the research evidence on the relationship between age and job per-formance is by no means conclusive. The meta-analyses of Waldman andAvolio (1986) and McEvoy and Cascio (1989) suggest that on average thecorrelation between age and job performance is close to zero. However, thereis a great deal of variation across studies; McEvoy and Cascio (1989), forexample, find correlations ranging from +0.66 to 0.44, with evidence of asmall positive relationship between age and performance for very youngemployees. More recent studies have continued to produce mixed findingsand there has also been a suggestion that any positive relationship is likelyto be non-linear, with performance eventually reaching a plateau and thengently declining as retirement age is approached (Warr, 1994). An expla-nation for these mixed results, methodological difficulties aside, is that jobsand their constituent tasks vary in the extent to which their performance isimpaired by physical ageing or enhanced by the accumulation of experienceand skill (see Warr, 1994, for an interesting typology of jobs drawn up onthis basis). What is clear is that older workers are not necessarily any lesseffective than their younger counterparts. It seems unlikely that age dis-crimination can be largely attributed to poorer performance on the part ofolder workers; the age stereotypes that we have described may be factuallyinaccurate, in general. This does not, of course, deny their possible role inunderpinning the discriminatory treatment of older workers, a possibilitythat we explore in this article.

    National characteristics affecting age stereotypes

    There has been little research on age stereotyping in Chinese societies. Debrah(1996) argues that age discrimination in Singapore arises mainly fromemployers stereotypical views on older workers (p. 815), but this proposi-tion is not explored empirically and there is no reference to any such researchin Chinese societies. A rare Hong Kong study is a government-sponsoredsurvey of employer preferences on recruitment, which suggested that employ-ers see younger workers as offering value for money, high productivity, andthe ability to deal with clients and learn new skills, whilst older workers arefelt to offer greater maturity and loyalty, and lower levels of turnover andabsence (Education and Manpower Branch, 1996: 20). There are similaritieshere with Western findings, although there was no direct comparison of theextent of such stereotypical views relative to Western societies, and the report

    Human Relations 54(5)6 3 4

    04chiu (ds) 29/3/01 7:35 am Page 634

  • made no mention of the correlates of such beliefs in Hong Kong. To ourknowledge, then, no study has yet attempted to systematically gauge the rela-tive prevalence of age stereotypes in Chinese as opposed to Western societies.Our aim in this study is to begin to fill this gap in the literature, by compar-ing Eastern and Western employees in terms of the extent to which theyexhibit the kinds of age stereotypes reported in the Western literature andbroadly replicated in the Hong Kong Governments study (Education andManpower Branch, 1996).

    We might expect that Chinese societies would show more positive atti-tudes towards older workers than those in the West, in line with the view thatChinese culture shows a high degree of deference towards age (Bond & Hwang,1986: 215). Finkelstein et al. (1995) have recently suggested that employersmay believe that older workers cost more to employ, for example due to higherhealth insurance claims. These economically based age stereotypes awaitrigorous testing in the Western literature, although it is interesting that theHong Kong Governments study provides some evidence of such a belief. Ingeneral, however, indirect employment costs are lower in Hong Kong than inthe UK and other Western countries, so that to the extent that stereotypesreflect true cost differentials, we may expect Hong Kong employers to be lessprone to such beliefs. This may in turn influence the attitudes of their staff.However, Hong Kong has seen a particularly rapid shift towards the servicessector and to jobs demanding higher levels of education and service skills, allof which might result in higher levels of age discrimination in Hong Kong com-pared to the UK, since older workers in particular may be felt to lack the neces-sary skills. Thus, concerns have been expressed in the media and amongstlabour activists about age discrimination, especially against those olderworkers who have been displaced from manufacturing and who find it diffi-cult to get jobs in the services sector. Add to this the possibility that a collec-tive and high power distance society such as Hong Kong may place lessemphasis than the West on individual rights (Hofstede, 1980), and it seems thatwe are not in a position to make an a priori hypothesis about the extent towhich Hong Kong and UK respondents will display negative or positive agestereotypes. Hence, we formulate the following exploratory research question:

    Research question 1: How do Hong Kong employees compare with UKemployees in the extent to which they hold positive or negative beliefsabout older workers?

    Organizational characteristics affecting age stereotypes

    The nature of the organization in which an individual works may be expectedto influence the individuals attitudes to older workers. We consider three

    Chiu et al. Age stereotypes and discrimination against older workers 6 3 5

    04chiu (ds) 29/3/01 7:35 am Page 635

  • such organizational-level variables in our analysis. First, the individuals atti-tude towards older workers may be influenced by the nature of the industryto which that individuals organization belongs. The argument is that dis-crimination may be driven by customer tastes. If customers prefer to beserved by younger workers, then those industries which involve frequentdirect contact between employees and customers are likely to be character-ized by higher levels of age discrimination and by a relatively young work-force as employers respond to customer preferences. Such age discriminationmay be rationalized by managers and supervisors through the adoption ofadverse stereotypes of older workers. The outcome is that individualsworking in industries requiring greater direct contact with customers may besocialized into a culture of stereotypical beliefs about older workers and areless likely to have any such beliefs challenged.

    Second, we may expect attitudes towards older workers to be influ-enced by the size of the organization in which the individual works. Thus,Lucas (1995), in a study of the UK hotel and catering industry, found evi-dence of more negative age stereotypes amongst employers in small organiz-ations than amongst those in larger organizations. She attributes this findingto smaller organizations lack of professional personnel management exper-tise and to their less sophisticated management practices, including perhapsa lower awareness of equal opportunities issues. Again, employees may besocialized into such beliefs and are less likely than employees in largerorganizations to see stereotypes challenged.

    Finally, a more specific measure of management practice is the presenceor otherwise of an equal opportunities policy with specific reference to age. Theintroduction of such a policy may help to cultivate a more tolerant atmospheretowards older workers in the organization by raising awareness and counter-ing stereotypical beliefs.2 Anti-age discrimination policies may include a rangeof measures, including raising upper age limits for training, redesigning jobs totake into account the needs of older workers, particularly in relation to moreflexibility in work schedules and gradual retirement, matching benefit systemsto suit the specific needs of older workers, and providing career-long trainingand development (Metcalf & Thompson, 1990).

    Based on the above discussion, we hypothesize that organizationalcharacteristics will be related to respondents age stereotypes. Specifically thefollowing hypotheses are proposed:

    Hypothesis 1a: Those who work in industries which require more fre-quent contact with customers (i.e. service industries as opposed tomanufacturing/ production) will have less favourable beliefs aboutolder workers than those working in industries where there is less cus-tomer contact.

    Human Relations 54(5)6 3 6

    04chiu (ds) 29/3/01 7:35 am Page 636

  • Hypothesis 1b: Respondents working in smaller organizations willhold less favourable beliefs about older workers than those who workin larger organizations.

    Hypothesis 1c: Respondents working in an organization with an anti-ageism policy will hold more favourable beliefs about older workersthan those working in an organization that does not have such a policy.

    Individual characteristics affecting age stereotypes

    Several studies have examined the individual characteristics that correlatewith beliefs about older workers. In a recent study of workers in three south-eastern US organizations, Hassell and Perrewe (1995) found evidence ofstereotypical beliefs, although the negative views were balanced by positivebeliefs. Taking an older worker as someone over 50, they found that therespondents own age was positively correlated with more favourable beliefsabout older workers, a finding consistent with other research (see, forexample: Bird & Fisher, 1986; Institute of Personnel Management, 1993).Such findings may be explained in terms of the in-group bias hypothesis,whereby raters evaluate others in light of their similarity to them (Finkelsteinet al., 1995). Older people may thus rate older workers more favourably ina bid to maintain a positive social identity as a member of the older in-group.

    The contact hypothesis suggests that contact between members ofdifferent groups can reduce intergroup hostility and discrimination (Brownet al., 1986). Thus, Hassell and Perrewe (1995) suggested that the frequencyof interactions an individual has with older workers might be expected toreduce negative beliefs, to the extent that these represent an inaccurate stereo-type that can be countered by direct experience. In their study, frequency ofcontact appeared to have no direct impact on beliefs, although frequency ofcontact negatively moderated the relationship between age and beliefs. Theimplication here is that younger respondents in particular, seeing olderworkers as members of an out-group, may have their less favourable beliefsmoderated through more frequent contact with older workers.

    Hassell and Perrewes (1995) study also shows a negative directrelationship between supervisory status and beliefs, in line with earlier USresearch finding supervisors to have less favourable beliefs about olderworkers than other employees have (Bird & Fisher, 1986; Kirchner & Dun-nette, 1954). A possible reason for this is that supervisors, often themselvesrather older than the majority of those they supervise, may seek to distin-guish themselves from the older group in an attempt to maintain a positiveself image and in so doing develop negative stereotypes of older employeesgenerally (Hassell & Perrewe, 1995). It may be that supervisors then identify

    Chiu et al. Age stereotypes and discrimination against older workers 6 3 7

    04chiu (ds) 29/3/01 7:35 am Page 637

  • with a more positive supervisor in-group, rather than with older workers ingeneral. There is also evidence that supervisory status acts as a negative mod-erator in the relationship between age and beliefs: older workers tended tohave more positive beliefs, but this is less so for older supervisors (Bird &Fisher, 1986; Hassell & Perrewe, 1995). Again, this may be due to super-visors distancing themselves from the older group, something that may bepursued with greater urgency by older supervisors.

    Based on the above discussion, we hypothesize that individual charac-teristics will be related to respondents age stereotypes. Specifically thefollowing hypotheses are proposed:

    Hypothesis 2a: Respondents own age will be positively correlated withmore favourable beliefs about older workers.

    Hypothesis 2b: Respondents who have more frequent contact witholder workers will have more favourable beliefs about such workersthan those who have less contact.

    Hypothesis 2c: Respondents who are supervisors will have lessfavourable beliefs about older workers than those who are not super-visors.

    Hypothesis 2d: Frequency of contact with older workers will negativelymoderate the relationship between age and beliefs, i.e. older respon-dents will have more favourable beliefs, but this tendency will be weak-ened where there is frequent contact.

    Hypothesis 2e: Supervisory status will negatively moderate therelationship between age and beliefs, i.e. older respondents have morefavourable beliefs, but this tendency will be weaker for supervisors.

    Hassell and Perrewe (1995) found no significant relationship betweengender and beliefs about older workers. Although we have no specific priorhypotheses on the likely impact of gender on stereotypical beliefs, we includethis variable in our analysis.

    Age stereotypes and biased attitudes in personnel practices

    It is clear from Western research that age stereotypes are very common.Furthermore, there is empirical support for the view that such stereotypes maylead to discriminatory human resources (HR) decisions. In two simulationstudies, involving US undergraduates and subscribers to the Harvard BusinessReview, Rosen and Jerdee (1976b, 1977) compared the responses of subjectsto a series of scenarios, designed to test the comparative treatment of olderand younger workers. They found that older workers were seen as being moreresistant to change than were younger employees, so that respondents were

    Human Relations 54(5)6 3 8

    04chiu (ds) 29/3/01 7:35 am Page 638

  • more likely to consider reassignment rather than corrective action to deal withwork behaviour problems of older workers. Respondents were also less likelyto recommend older workers for promotion to positions requiring creativity,innovation and financial risk-taking, and were less likely to recommend themfor transfer to a physically demanding job. Subjects saw older workers asbeing less motivated to update their skills; thus they were less likely to rec-ommend them for training and were more likely to see them as untrainableand hence to consider them suitable for dismissal.

    On the basis of such findings, we expect that the age stereotypes indi-viduals have will influence their attitudes towards the treatment of olderworkers on a range of personnel decisions, such as selection for training, pro-motion and redundancy. We may also expect a similar impact on the willing-ness of employees to work with older colleagues and on their support for theexercise of positive discrimination in selecting an older worker to work with.Specifically, we hypothesize:

    Hypothesis 3: There will be a positive relationship between positivebeliefs about older workers and attitudes towards their training, pro-motion and retention, willingness to work with older workers andwillingness to consider positive discrimination in selection on theirbehalf.

    We also test the effects on discriminatory attitudes of the organiz-ational and individual-level variables referred to earlier. In this analysis, weanticipate similar effects to those hypothesized in the prediction of beliefs.In addition, we are interested in finding out whether the relationshipbetween respondents age stereotypes and their attitudes towards discrimi-natory practices differs between Hong Kong and UK respondents. The argu-ment here is that the social and regulatory framework differs between thetwo locations, so that it is possible that the way in which age stereotypesinfluence an individuals attitudes towards discrimination will differbetween the two locations. For example, well-established laws and/or asocial climate that strongly condemns discrimination may weaken thisrelationship. Whilst we know that Hong Kong has less well-establishedequal opportunities law in general, neither Hong Kong nor the UK has legis-lation on age discrimination and at this stage we are unsure about the rela-tive strength of anti-age discrimination feeling in each place. Hence, ratherthan specify a hypothesis, we propose the following exploratory researchquestion:

    Research question 2: Will the relationship between respondents agestereotypes and their discriminatory attitudes towards personnel prac-tices be stronger amongst Hong Kong or UK respondents?

    Chiu et al. Age stereotypes and discrimination against older workers 6 3 9

    04chiu (ds) 29/3/01 7:35 am Page 639

  • Method

    Participants

    Samples were drawn from Hong Kong and the UK. A special feature of thesamples is the high proportion working in personnel management. The atti-tudes of such people are likely to be particularly important in shaping theemployment prospects of older people, given their involvement in selectionand other HR decisions. Their familiarity with such decisions also means thattheir views may provide a good indicator of the state of HR practice.

    Hong Kong sample

    A total of 302 respondents who attended part-time diploma courses inmanagement organized by the Hong Kong Polytechnic University wereinvited to participate in the survey. Of these 302 respondents, 72 percentwere females and 28 percent were males. Most of them were relatively young(M = 31, SD = 4.0, minimum = 23, maximum = 48). Half of them had atleast 10 years working experience. These respondents mostly worked forlarge organizations (i.e. more than 500 employees). The types of industriesthey belonged to were quite diverse, including finance (28%); health/edu-cation/social welfare (17%); manufacturing (16%); transport/warehouse(14%); wholesale/retailing (13%); restaurants/hotels (7%); and construction(4%).

    UK sample

    Data were collected from 265 respondents in the UK, all of whom were part-time students of management at the University of Teesside. More than halfof the respondents were females (56%). Like the Hong Kong sample, theserespondents were relatively young. The mean age of the UK sample was 33,with a standard deviation of 7.5, a minimum of 22 and a maximum of 60.Ninety percent were diploma level students, the remainder being Mastersstudents. All of them had work experience, and more than half had beenworking for at least 10 years. Two-thirds of the respondents worked fororganizations with more than 500 staff. The types of industries to which therespondents belonged included health/education/social welfare (54%);manufacturing (21%); finance (7%); construction (6%); transport/ware-house (4%); wholesale/retailing (3%); and restaurants/hotels (2%).

    We cannot claim that the two samples are directly representative of theHong Kong and UK populations of employees or managers. Both sampleswere relatively young and, having enrolled in part-time management courses,they are a self-selecting group who may have above-average levels of career

    Human Relations 54(5)6 4 0

    04chiu (ds) 29/3/01 7:35 am Page 640

  • ambition and awareness of current management issues. This represents alimitation in terms of the direct generalizability of our findings, although thesamples are likely to yield a conservative test for cultural differences, sinceany cross-cultural convergence in attitudes is likely to be greater amongst theyoung and better educated. Whilst the respondents were drawn fromamongst part-time students, all were in full-time employment and most hadconsiderable management and/or professional experience.

    The two samples are not an exact match. The UK sample was onaverage a little older (M = 33 vs 31 years old; t = 4.37, p < .000), reportedmore frequent contact with older workers (2 = 73.83, p < .000; for example,74% had daily contact compared to 46% of the Hong Kong sample), andworked for larger organizations (2 = 22.04, p < .000; for example, 66%worked for organizations with over 500 employees, compared to 51% forthe Hong Kong sample). The samples differed in their industry composition(2 = 136.350, p < .000). For example, 12.8 percent of the UK sample workedin manufacturing and production, compared with 10.7 percent in the HongKong sample, and 56 percent of the UK sample working in health, educationand welfare, with only 17 percent in Hong Kong. There was no significantdifference in the proportion of respondents with supervisory status.

    Procedure

    Questionnaires were administered to some 27 batches of students, 12 inHong Kong and 15 in the UK. Before respondents completed the question-naires, we briefly introduced the purpose of the study and answered ques-tions raised by the respondents. They were asked not to put down their namesto ensure anonymity and were assured that their participation was on avoluntary basis. Respondents were given as much time as they needed towork on the questionnaires. On average, they took 15 to 25 minutes to com-plete the questionnaire. All returned completed questionnaires.

    Variables

    Dependent variables

    Two sets of dependent measures were assessed: (1) age stereotypes aboutolder workers, and (2) discriminatory attitudes on employment practices.The age stereotypes consisted of two dimensions: Adaptability and WorkEffectiveness, measured by six and four questionnaire items, respectively.These and the other measures were based on items drawn from the Hasselland Perrewe (1995) questionnaire. The two dimensions broadly reflect earlier

    Chiu et al. Age stereotypes and discrimination against older workers 6 4 1

    04chiu (ds) 29/3/01 7:35 am Page 641

  • UK research on age stereotypes (Warr & Pennington, 1993). To confirm theunderlying structure of the age stereotypes, a confirmatory factor analysiswas conducted to compare the proposed two-factor model with a one-factormodel of age stereotypes across the Hong Kong and the UK samples. As sug-gested by the five goodness-of-fit indices shown in Table 1, the two-factormodel was a better fit than was the one-factor model and the indices of theproposed model were found to be generally acceptable. In light of the find-ings, two composite variables were constructed by taking the mean of theindividual items. Specific items and the reliability coefficients for bothsamples are listed in Table 2.

    Respondents were also asked for their attitudes on a range of discrim-inatory practices, as shown in Table 3. These items reflect commonly citedexamples of discriminatory treatment, including giving older workers lowerpriority in terms of training and promotion opportunities, and selecting olderworkers for retrenchment. One item refers to the willingness of respondentsthemselves to work with older workers and the final item looks at the extentto which respondents would be prepared to practise positive discriminationby choosing an older worker as a colleague in preference to an equally skilledyounger candidate. A principal components analysis with varimax rotationwas conducted on the five discriminatory attitudes (i.e. promotion, retention,willingness to work as peers, and a willingness to consider positive discrimi-nation in selection), producing a first general factor (eigenvalue greater than1) which accounted for only 37 percent of the total variance. The reliabilitycoefficient of that factor was a mere .40. Based on these findings, the five dis-criminatory attitudes were treated separately, each representing a differentaspect of employment practice.

    Human Relations 54(5)6 4 2

    Table 1 Goodness-of-fit indices of the proposed two-factor model of age stereotypes inHong Kong and UK samples

    2 df CFIa GFIb RMSEAc

    Hong KongOne-factor 124.36** 35 .73 .92 .09Two-factor 60.07* 34 .92 .96 .05UKOne-factor 208.68** 35 .60 .84 .14Two-factor 103.35** 34 .84 .93 .08

    aThe comparative fit indexbThe goodness-of-fit indexcThe root mean square error of approximation

    04chiu (ds) 29/3/01 7:35 am Page 642

  • Chiu et al.

    Age ste

    reo

    typ

    es a

    nd

    discrim

    inatio

    n a

    gain

    st old

    er w

    ork

    ers

    64

    3

    Table 2 Dependent variable scales: Age stereotypesa

    UKb Hong Kongc

    M SD M SD

    Adaptability 3.09 .66 2.68 .67Older workers are not interested in learning new skills 3.41 .92 2.93 .99Older workers do not want jobs with increased responsibilities 3.27 .97 2.81 .96Most older workers cannot keep up with the speed of modern industry 3.00 .93 2.83 .98Older workers prefer less challenging jobs than those they held when they were younger 3.15 .89 2.64 .94Older workers can learn new skills as easily as other employees (R) 3.09 .94 3.47 .87Older workers in our department work just as hard as anyone else (R) 3.86 .92 3.36 .92

    Cronbachs alpha = .75 (Hong Kong sample), .77 (UK sample)

    Work Effectiveness 3.02 .61 3.11 .54Older workers are more dependable (R) 3.25 .96 3.33 .85Older workers are better employees (R) 2.68 .76 2.72 .72Older employees usually turn out work of higher quality (R) 2.96 .80 2.56 .89Older employees are most loyal to the company (R) 3.18 .96 3.83 .88

    Cronbachs alpha = .60 (Hong Kong sample), .65 (UK sample)

    a Respondents were asked to rate all items according to a five-point Likert-type scale, ranging from 1 (strongly agree) to 5 (strongly disagree). Items considered by the authors to befavourable to older workers were reverse scored (R), so that higher scores indicate a more positive belief or attitude towards older workers.b The size for the UK sample ranged from 253 to 259.c The size for the Hong Kong sample ranged from 299 to 302.

    04chiu (ds) 29/3/01 7:35 am Page 643

  • Independent variables

    As described in the previous section, there are three sets of independent vari-ables: national, organizational and individual. The UK was coded as 1 andHong Kong as zero. There are three organizational level variables. Companysize refers to the number of employees working in the company. Two dummyvariables (1= yes; 0 = no) were constructed to indicate company size withlarge meaning 500 employees or more, and medium 100 to 499 employ-ees. Company size less than 100 employees, considered to be small, servedas the reference category.3 Variables were also included for industry type (1= services; 0 = manufacturing/production),4 and whether the respondentsorganization had a stipulated policy not to allow age discrimination (1 = hasageism policy; 0 = does not have ageism policy). There are four individual-level variables: female (female = 1; male = 0), age (ranging from 22 to 60),whether or not the respondent directs or manages other staff (supervisor) (1= yes; 0 = no) and the respondents reported frequency of contact with olderworkers in their job. The latter was coded with none, representing nocontact with elderly workers, serving as the reference category; some,defined as once a week or less (1 = yes; 0 = no); frequent, operationalizedas several times a week (1 = yes; 0 = no); and daily, defined as contact everyday (1 = yes; 0 = no). Interaction terms were constructed by multiplying age

    Human Relations 54(5)6 4 4

    Table 3 Dependent variables: Discriminatory attitudesa

    TrainingIt is a better investment to train younger workers rather than older workers

    PromotionOlder employees should 'step aside' (take a less demanding job) to give younger employeesadvancement opportunities

    RetentionI think younger employees should be given priority to stay if a company contracts its business

    Co-workingGiven a choice, I would not work with an older worker on a daily basis

    Positive discriminationIf two workers had similar skills, Id pick the older worker to work with me (R)

    a Respondents were asked to rate all items according to a five-point Likert-type scale, ranging from 1 (stronglyagree) to 5 (strongly disagree). Items considered by the authors to be favourable towards older workers werereverse-scored (shown as (R)), so that higher scores indicate a more favourable (less discriminatory) attitudetowards older workers.

    04chiu (ds) 29/3/01 7:35 am Page 644

  • with the three dummy coded variables on frequency of contact, age withsupervisory status, and respondents country with the two stereotypes.

    Analyses

    Hypotheses were tested using two separate sets of multiple regression analy-ses. The first set involved regressing the two age stereotypes on the respon-dents country, organizational characteristics, individual characteristics andfour interaction terms (i.e. age x supervisory status and age x contact fre-quency). The variables were entered in four consecutive blocks. The secondset of analyses involved regressing the discriminatory attitudes at work oncountry, organizational characteristics, individual characteristics, the two agestereotypes, and two interaction terms (respondents country x Adaptabilityand respondents country x Work Effectiveness). This was repeated for eachof the five discriminatory attitudes.

    Results

    Table 4 presents the means, standard deviations and zero-order correlations.The correlations ranged from .39 to -33, of which 80 percent are within the+.20s. Respondents country, the presence of an age discrimination policy,and respondents age were found to be correlated with at least one of the agestereotypes and with most of the discriminatory attitudes. The intercorrela-tions among the five discriminatory attitudes ranged from low to moderate.Adaptability was related to all five discriminatory attitudes while Work Effec-tiveness was related only to one. All variables had adequate discriminantvalidity and these results suggest that there should be no serious multi-collinearity problems in the regression analysis.

    Determinants of positive or negative age stereotypes

    To assess whether respondents country, organizational and individualcharacteristics influence their beliefs about older workers, the first set of mul-tiple regression analyses was conducted by regressing the two age stereotypeson the independent variables discussed earlier and the interaction terms. Asregards our research question on the relative prevalence of age stereotypesacross the two locations, Table 5 suggests that, compared to Hong Kongrespondents, UK respondents had a more negative view of older workers interms of their ability to cope with change (Adaptability), but a more positiveview of the relative Work Effectiveness of older workers. Importantly, these

    Chiu et al. Age stereotypes and discrimination against older workers 6 4 5

    04chiu (ds) 29/3/01 7:35 am Page 645

  • Hu

    man

    Rela

    tion

    s 54

    (5)

    64

    6

    Table 4 Means, standard deviations and correlations among variablesa

    Variables M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

    1. Country UKb .46 .50 1.002. Mediumb .22 .42 .04 1.003. Largeb .58 .49 .15** .27** 1.004. Age policyb .27 .44 .33* .09* .27** 1.005. Service indb .76 .42 .10* .06 .09* .17** 1.006. Age 31.80 6.02 .19** .02 .06 .13** .01 1.007. Femaleb .64 .48 .17** .06 .00 .08* .11* .20** 1.008. Supervisorb .79 .46 .02 .10* .00 .01 .07 .15** .14** 1.009. Someb .10 .25 .25** .18** .05 .08 .04 .09* .12** .01 1.0010. Frequentb .14 .35 .21** .01 .08* .11** .09* .12** .11** .04 .15** 1.0011. Dailyb .59 .49 .04 .07 .04 .00 .03 .00 .00 .06 .15** .17** 1.0012. Adaptability 2.87 .70 .29** .01 .04 .19** .03 .17** .00 .01 .14** .00 .08 1.0013. WorkEffect 3.10 .57 .08 .04 .06 .02 .06 .16** .05 .03 .04 .06 .14** .13** 1.0014. PosDiscrim 3.20 .96 .33** .08 .06 .09* .02 .01 .09* .02 .08* .10* .02 .16** .08 1.0015. Retaining 3.28 1.02 .17** .05 .03 .14** .01 .17** .05 .09* .04 .06 .03 .25** .10* .10* 1.0016. Promoting 3.83 .94 .07 .02 .04 .07 .03 .12** .03 .02 .01 .03 .03 .30** .03 .07 .32** 1.0017. Training 2.58 1.07 .28** .00 .03 .21** .09* .20** .07 .04 .14** .06 .01 .39** .01 .14** .33** .26** 1.0018. Co-working 3.77 .98 .19** .03 .02 .10* .04 .15** .00 .04 .15** .07 .04 .31** .03 .12** .17** .24** .29**1.00

    a Ns for descriptive statistics range from 536 to 561.b Dichotomous or categorical variable.* p < .05, two-tailed test; ** p < .01, two-tailed test.

    04chiu (ds) 29/3/01 7:35 am Page 646

  • Chiu et al.

    Age ste

    reo

    typ

    es a

    nd

    discrim

    inatio

    n a

    gain

    st old

    er w

    ork

    ers

    64

    7

    Table 5 Results of multiple regression analysis: The effects of national, organizational and individual characteristics on age stereotypes

    Adaptability Work Effectiveness

    Variables 1 2 3a 4b 1 2 3a 4b

    Country UK .41** .38** .36** .09* .08 .12*Medium size (vs small) .01 .01 .10 .07Large size (vs small) .03 .02 .04 .02Have ageism policy .15* .15* .01 .02Service industry .07 .06 .07 .07Age .02** .02**Female .11 .05Supervisor .01 .06Some contact (vs none) .18 .20*Frequent contact (vs none) .16 .03Daily contact (vs none) .01 .08Age x supervisor .01 .02*Age x some contact .04 .00Age x frequent contact .00 .00Age x daily contact .02 .01R2 .09 .10 .12 .14 .01 .01 .07 .08Final F 49.90** 11.40** 6.64** 5.52** 3.50* 1.31 3.36** 3.04**df 1,527 5,523 11,517 15,513 1,527 5,523 11,517 15,517

    a Unstandardized regression coefficients of all the main effects are shown before entering the interaction variables.b Only the unstandardized regression coefficients of the interaction effects at the final step are shown. p < .1; *p < .05; **p < .01.

    04chiu (ds) 29/3/01 7:35 am Page 647

  • country differences persist even when organizational and individual charac-teristics are controlled for in the regression analysis.

    The results in Table 5 also suggest that organizational characteristics, ingeneral, have little effect on stereotypical beliefs about older workers. Whetheror not the organization has a policy on ageism was the only exception here,emerging as a positive predictor of Adaptability but not of Work Effectiveness,providing only partial support for hypothesis 1c. The finding of no significantrelationship between the respondents industry type and organizational size onthe one hand and beliefs about older workers on the other contradicts hypoth-eses 1a and 1b. The older the respondent, the more favourable their beliefsabout older workers, in terms of both Adaptability and Work Effectiveness,thus providing support for the in-group bias hypothesis 2a. Female respondentsappear to have more positive views about the Adaptability of older workersonly, although this finding was significant only at the 10 percent level. Fre-quency of contact with older workers appears to have no effect on the Adapta-bility dimension, but those who reported some contact had more positiveviews of older workers Work Effectiveness than did respondents who reportedno contact. Perhaps surprisingly, those reporting frequent or daily contacthad no more positive beliefs than those reporting no contact. This suggests thatthere is no simple linear relationship between contact and beliefs, providing atbest only partial support for the contact hypothesis 2b. There was no signifi-cant relationship between supervisory status and beliefs about older workers,thus contradicting hypothesis 2c.

    No significant interactions were found between respondents age andthe frequency of contact with older workers. Thus we fail to find support forhypothesis 2d. There was, however, a significant interaction between age andsupervisory status in the case of the Work Effectiveness belief. The positiverelationship between own age and Work Effectiveness was weaker for super-visors. Supervisory status thus had a negative moderating impact on theagebeliefs relationship in the case of Work Effectiveness but not for Adapta-bility, partially supporting hypothesis 2e.

    Determinants of discriminatory attitudes in employment

    Next, separate regression analyses were conducted, entering the two age stereo-types, organizational characteristics, individual characteristics and country,along with two interaction terms as independent variables to predict respon-dents views on the five discriminatory attitudes. Table 6 shows that once theage stereotypes are included in the analysis, country was found to relate to twoof the discriminatory attitudes, with UK respondents having more favourableattitudes towards providing training for older workers, but being less prone to

    Human Relations 54(5)6 4 8

    04chiu (ds) 29/3/01 7:35 am Page 648

  • support the hiring of older workers in preference to equally qualified youngerworkers. With the age stereotypes in the equations, no systematic relationshipswere found between organizational characteristics and discriminatory atti-tudes, except for three marginal effects, two on training (i.e. having an ageismpolicy and being in a customer-driven industry are associated with positive atti-tudes) and one on positive discrimination (i.e. weaker support for positive dis-crimination from respondents in large as opposed to small companies).However, given the marginal nature of these latter findings, we should inter-pret them with caution, especially since the estimation of five separate equa-tions increases the risk that our findings capitalize on chance.

    Older respondents themselves had more favourable attitudes towardstraining and retaining older workers, with only a marginally significantrelationship between age and attitudes towards promoting older workers.Female respondents had more favourable attitudes than did males on train-ing and retaining older workers. Frequency of contact with older workerswas positively related to attitudes to providing them with training and withthe willingness to work alongside them.

    Respondents stereotypical beliefs about older workers were found tobe predictive of their discriminatory attitudes, with Adaptability emerging asa positive predictor of all five attitudes. The more respondents perceived olderworkers as being adaptable to change, the more favourable their views ontraining, promoting, retaining and working with older workers, and on exer-cising positive discrimination in their favour. These results support hypothe-sis 3 in the case of Adaptability. However, the Work Effectiveness belief wasfound to predict only the attitude to retention.

    As regards research question 2, the majority of the country by stereotypeinteraction effects were not significant. The exceptions were a positive co-efficient on the country Adaptability interaction in the case of promotion, amarginally significant positive coefficient for country Work Effectiveness inthe retaining older workers equation, and a positive coefficient on both inter-actions in the case of positive discrimination. These coefficients suggest that forUK respondents, the relationship between positive age stereotypes andfavourable attitudes on the treatment of older workers is stronger than amongstHong Kong respondents.

    Discussion

    We explored whether Hong Kong respondents would have more adversestereotypical beliefs about older workers than their UK counterparts. In fact,our results were mixed. Our Hong Kong sample saw older workers as less

    Chiu et al. Age stereotypes and discrimination against older workers 6 4 9

    04chiu (ds) 29/3/01 7:35 am Page 649

  • Human Relations 54(5)6 5 0

    Table 6 The effects of national, organizational and individual characteristics, and agestereotypes on the five discriminatory attitudes at worka

    Variables 1 2 3 4a 5b

    TrainingAdaptability .60** .56** .50** .46**Work Effectiveness .08 .07 .05 .07Medium size (vs small) .05 .09 .05Large size (vs small) .07 .10 .10Have ageism policy .33** .28** .19

    Service industry .13 .15 .19

    Age .02** .02**Female .26** .29**Supervisor .03 .00Some contact (vs none) .45** .40*Frequent contact (vs none) .24 .15Daily contact (vs none) .43** .32**Country UK .32**Country x Adapability .06Country x Work Effectiveness .15R2 .15 .17 .22 .24 .24Final F 46.87** 18.25** 12.24** 12.30** 0.72**df 2,531 6,527 12,521 13,520 15,518

    PromotionAdaptability .40** .40** .38** .38*Work Effectiveness .01 .03 .02 .02Medium size (vs small) .12 .12 .12Large size (vs small) .01 .00 .00Have ageism policy .04 .04 .04Service industry .03 .03 .03Age .01 .02

    Female .09 .08Supervisor .08 .08Some contact (vs none) .64 .70Frequent contact (vs none) .04 .05Daily contact (vs none) .02 .03Country UK .03Country x Adapability .24*Country x Work Effectiveness .00R2 .09 .09 .10 .10 .11Final F 25.56** 8.85** 4.77** 4.40** 4.10**df 2,531 6,527 12,521 13,520 15,518

    04chiu (ds) 29/3/01 7:35 am Page 650

  • Chiu et al. Age stereotypes and discrimination against older workers 6 5 1

    Table 6 continued

    Variables 1 2 3 4a 5b

    Retaining Older WorkersAdaptability .35** .33** .27** .25**Work Effectiveness .12 .13 .16* .15*Medium size (vs small) .18 .16 .18Large size (vs small) .07 .06 .06Have ageism policy .22* .19 .15Service industry .01 .02 .00Age .03** .03**Female .19* .20*Supervisor .14 .13Some contact (vs none) .24 .21Frequent contact (vs none) .04 .01Daily contact (vs none) .18 .13Country UK .14Country x Adapability .14Country x Work Effectiveness .25

    R2 .07 .08 .11 .12 .12Final F 18.79** 7.39** 5.51** 5.23** 4.79**df 2,531 6,527 12,521 13,520 15,518

    Working as peersAdaptability .45** .44** .41** .39**Work Effectiveness .12 .12 .10 .10Medium size (vs small) .02 .04 .02Large size (vs small) .12 .15 .16Have ageism policy .13 .09 .04Service industry .14 .10 .08Age .01 .01Female .11 .12Supervisor .08 .10Some contact (vs none) .05 .02Frequent contact (vs none) .25 .21Daily contact (vs none) .31** .27*Country UK .14Country x Adapability .14Country x Work Effectiveness .17R2 .10 .11 .14 .14 .14Final F 30.30** 10.93** 6.78** 6.44** 5.76**df 2,531 6,527 12,521 13,520 15,518

    04chiu (ds) 29/3/01 7:35 am Page 651

  • effective at work but, interestingly, they gave them a more positive rating thandid the UK respondents in terms of Adaptability. To the extent that the latterfinding accurately reflects the relative flexibility of older workers in the twosocieties, this may be a result of Hong Kongs more flexible labour market.Limited retirement and social security provision may mean that older workersin Hong Kong are forced by financial circumstances to remain adaptable tothe demands of a changing labour market. In the UK, the trend towards earlyretirement and the more extensive social security and pensions provision mayprovide older workers with a weaker incentive to remain flexible. Our aimhere is not, of course, to decry the provision of social welfare protection forolder workers the welfare arguments are overriding here and Hong Konghas been seen as a laggard in this respect. Rather, we simply point out theimplications for employee behaviour and for attitudes towards older workers.

    Human Relations 54(5)6 5 2

    Table 6 continued

    Variables 1 2 3 4a 5b

    Positive discriminationAdaptability .21** .20** .20** .12*Work Effectiveness .09 .09 .08 .06Medium size (vs small) .07 .06 .14Large size (vs small) .24* .19 .19

    Have ageism policy .11 .10 .09Service industry .09 .05 .03Age .01 .01Female .15 .09Supervisor .12 .07Some contact (vs none) .13 .24Frequent contact (vs none) .05 .12Daily contact (vs none) .20 .02Country UK .60**Country x Adapability .27*Country x Work Effectiveness .76**R2 .03 .05 .07 .14 .19Final F 8.27** 4.25** 3.12** 6.28** 7.99**df 2,531 6,527 12,521 13,520 15,518

    a Unstandardized regression coefficients of all the main effects are shown before entering the interactionvariables.b Only the unstandardized regression coefficients of the interaction effects at the final step are shown.p < .1; *p < .05; **p < .01.

    04chiu (ds) 29/3/01 7:35 am Page 652

  • There is a suggestion that workers in Hong Kong may be more dis-criminatory against older workers than are UK workers. Hong Kong respon-dents were more likely to believe that younger workers should have priorityin terms of training opportunities. Furthermore, when asked the question Inyour opinion, at what age are employees likely to begin to experience age dis-crimination in the region where you work?, 47 percent of UK respondentsanswered not at all compared to only 10 percent of the Hong Kong sample.Amongst the remainder who quoted a specific age in answer to this question,the mean was 47 for the UK and only 43 for Hong Kong, suggesting that agediscrimination may become an issue at an earlier age in the latter location.In our regression results, UK respondents showed less support for positivediscrimination in selection, but rather than reflecting negative attitudestowards older workers this may simply represent a more sophisticated under-standing of equal opportunities policy than is found amongst the Hong Kongsample. Overall, our findings suggest that it is Hong Kong workers who aremore prone to negative age discrimination.

    With regard to the other determinants of respondents beliefs aboutolder workers, we hypothesized several relationships on the basis of the priorliterature. As expected, we found that the respondents own age was predic-tive of positive attitudes towards older workers on both the Adaptability andWork Effectiveness dimensions. It seems that closer age proximity leads to amore favourable and perhaps more sympathetic assessment of older workers,providing support for the in-group bias hypothesis (Finkelstein et al., 1995).The presence of a policy on age discrimination in the respondents organiz-ation was associated with more positive beliefs on the Adaptability factor,but no significant relationship was found for Work Effectiveness. Contraryto our expectations, the respondents industry sector had no significant effecton stereotypical beliefs. Similarly, organizational size and whether or not therespondent had supervisory responsibilities appeared to have no significantimpact on beliefs. The respondents frequency of contact with older workershad no significant effect on Adaptability. Those reporting some contact hadmore favourable views on the Work Effectiveness factor than did thosereporting no contact, but those reporting more frequent (weekly or daily)contact had no more favourable beliefs than those having no contact. This israther curious, implying a non-linear relationship, and at best offers onlyqualified support for the contact hypothesis (Brown et al., 1986).

    Contrary to our hypothesis, we found no evidence that frequency ofcontact with older workers moderates the relationship between age andstereotypical beliefs. However, we did find a negative moderating effect ofsupervisory status on the relationship between age and Work Effectiveness.This is line with our hypothesis and with prior literature (Hassell & Perrewe,

    Chiu et al. Age stereotypes and discrimination against older workers 6 5 3

    04chiu (ds) 29/3/01 7:35 am Page 653

  • 1995), and suggests that the positive impact of own age on beliefs is offsetin the case of supervisors. One explanation for such a finding is that theimpact of own age in fostering positive beliefs is inhibited amongst super-visors because, whilst they tend to be older than many of their subordinates,they see themselves as valued members of the organization and wish to dis-tance themselves from older colleagues, in effect opting out of the older in-group, even to the extent of adopting negative stereotypes (Hassell &Perrewe, 1995: 466).

    Our results show that the extent to which respondents see olderworkers as adaptable has a significant impact on their discriminatory atti-tudes towards older workers, in terms of decisions on training, promotionand retention, their willingness to work alongside older workers and theextent to which they would support positive discrimination in selecting olderwork colleagues. These findings are consistent with earlier research showinga link between negative stereotypical beliefs about older workers and anintent to treat them less favourably than their younger colleagues (Rosen &Jerdee, 1976b, 1977). However, the only significant relationship between theWork Effectiveness belief and discriminatory attitudes was a positiverelationship with the willingness to retain older workers during retrench-ment. Why the two belief dimensions should differ so much in terms of theirlink with discriminatory attitudes is in need of explanation. One possibilityis that the Work Effectiveness dimension, as we have measured it, asksrespondents about the extent to which older workers are more effective thantheir younger colleagues. A low score does not necessarily imply that that therespondent sees older workers as less effective, simply that they are not moreeffective, so that such a neutral belief would not necessarily be expected tolead to a discriminatory attitude. There is also the possibility that the find-ings may simply reflect the lower internal consistency of the Work Effective-ness measure.

    Our findings suggest that the relationship between stereotypical beliefsand discriminatory work attitudes differs between Hong Kong and the UKin the case of only a minority of the discriminatory attitudes. The clearestfinding here is that for UK respondents there is a stronger relationshipbetween stereotypical beliefs and the willingness to positively discriminate infavour of an older worker.

    We found that the presence of an age discrimination policy in theworkplace is associated with more favourable attitudes on the provision oftraining for older workers, although this was only marginally significant.Add to this the finding noted earlier that age discrimination policies areassociated with more positive beliefs on the Adaptability of older workers,and it appears that such policies have a role in countering ageist beliefs and

    Human Relations 54(5)6 5 4

    04chiu (ds) 29/3/01 7:35 am Page 654

  • discrimination in the workplace. However, in our Hong Kong sample, only13 percent of respondents reported that they worked in an organization withsuch a policy, compared to 43 percent for the UK sample.5

    Unlike many others, both the UK and Hong Kong governments havein recent years adopted an essentially voluntary approach to dealing with agediscrimination, preferring to rely on exhortation and education of employ-ers rather than on legislation to outlaw age discrimination. Why discrimi-nation on the ground of age is best dealt with by purely voluntary means,whilst sex and race discrimination are tackled by legislation, is not entirelyclear, and Taylor and Walker (1997) outline the limitations of relying purelyon education of employers. Our study does not directly address the effec-tiveness or otherwise of legislation and voluntary approaches to combatingdiscrimination, but our findings do suggest that employer policies may havean educative effect on employees. Employers should therefore be encouragedto develop such policies and to communicate them to their workforce. Theunresolved question as far as our study is concerned is whether most employ-ers will implement such policies in the absence of legislation.

    The finding that stereotypical beliefs and discriminatory attitudes areat least to some extent related to the frequency of contact with older workerssuggests that familiarity with older workers can reduce negative stereotypesand discrimination. The implication here is that those employers who decideto recruit older workers will not necessarily experience a permanently nega-tive response from their younger staff. The very act of recruiting largernumbers of older workers into an organization can counter discriminatoryattitudes amongst the workforce. There is a positive message here for thoseemployers who have traditionally recruited large numbers of younger peopleinto particular positions, but who, perhaps in the face of long-term shortagesand demographic trends, are considering recruiting older staff.

    On the basis of our findings it is difficult to escape the conclusion thatHong Kong employees are more prone to age discrimination than are theirUK counterparts, whilst Hong Kong employers appear to lag in terms of theimplementation of age discrimination policies. Whether these differences areattributable to Hong Kong and the UK being at different stages in theimplementation of equal opportunities policy generally or whether there aremore enduring cultural influences at work here we cannot be sure, but theprinciple of equal opportunities has only relatively recently become an issueof active public policy and debate in Hong Kong and general awareness ofequal opportunities may yet increase.

    Given the more recent development of widespread participation inhigher education in Hong Kong relative to the UK, we cannot rule out thepossibility that these findings are to some extent a reflection of the larger

    Chiu et al. Age stereotypes and discrimination against older workers 6 5 5

    04chiu (ds) 29/3/01 7:35 am Page 655

  • education gap between older and younger cohorts in Hong Kong. To theextent that this is so, we may expect a convergence in Hong Kong and UKattitudes in the longer term, as the educational level of future cohorts ofolder workers rises. In the meantime, there is a need to provide educationaland training opportunities for older workers and to ensure that their dis-advantages are not compounded by discrimination in access to such oppor-tunities.

    In both the UK and Hong Kong, older workers have in recent yearsbeen disproportionately affected by industrial and organizational restructur-ing. In the former, early retirement has become almost the norm and isstrongly associated with organizations attempting to avoid compulsoryredundancy among young employees (Taylor & Walker, 1998). In the latter,retirement provision is usually more limited and the restructuring has beenif anything rather more dramatic. A sharp decline in manufacturing employ-ment since the mid-1980s, as production moves into the Chinese mainland,has led to a marked rise in unemployment for older people in particular. Akey issue has been the limited transferability of their manufacturing-basedskills into the expanding services sector, which emphasizes customer serviceand English-language skills. Government retraining schemes have made someimpact, but with unemployment rising in the Asian economies, the prospectsfor older workers may deteriorate. The need for legislative action on age dis-crimination may become more urgent.

    This article has a number of limitations. First, the nature of our samplelimits the degree to which we can generalize to Hong Kong and UK employ-ees as a whole. Our samples were not nationally representative, although wehave argued that the nature of our samples, relatively young employeesenrolled in part-time management education courses, means that our analy-sis constitutes a conservative test for the existence of cultural differences inage stereotypes. Whilst our Hong Kong and UK samples were not matched,our use of multivariate analysis allows us some degree of control for organiz-ational and individual variables in testing for country differences in stereo-typical beliefs and discriminatory attitudes.

    Second, amongst the organization-level variables, we found little evidence of a significant impact on stereotypical beliefs and discriminatoryattitudes, the main exception being the presence of a policy on age discrimi-nation. However, our inclusion of the size of the respondents organizationmay fail to adequately reflect the true impact of organizational-level influ-ences. We suggested that organization size is a proxy for the degree ofsophistication in personnel practices, which may be expected to influencebeliefs about older workers. It may be, however, that a more direct measure-ment of personnel practices, for example the extent to which there is a policy

    Human Relations 54(5)6 5 6

    04chiu (ds) 29/3/01 7:35 am Page 656

  • of promotion from within, would have yielded significant results. Further-more, the degree of personnel sophistication may impact on beliefs to theextent that employees are made aware of equal opportunities issues, theimplication being that employee awareness mediates the relationship betweenpersonnel policy sophistication and employee beliefs about older workers.Our research did not measure employee awareness of equal opportunities,nor did we have detailed measures of personnel practices, apart from theexistence or otherwise of an equal opportunities policy. Future research mightusefully include measures of personnel practice and employee awareness ofequal opportunities and might address the mediated nature of the organiz-ational characteristics-attitude relationship.

    Third, we were concerned specifically with employee beliefs and theirimplications for discriminatory attitudes. Whilst we suggested that employ-ees beliefs and attitudes may be influenced by consumer preferences, wemade no attempt in this study to measure consumer preferences directly. Weused industry as a proxy for the effects of customer tastes in influencingstereotypical beliefs amongst employees, but this is at best a very crudemeasure and the lack of significant findings may reflect this. To the extentthat consumer preferences are an important cause of discrimination inemployment, this is a key area for future research.

    Fourth, our aim was to explore the extent to which the age stereotypesidentified in the Western literature are found in an Eastern context, usingquestionnaire items developed from Western research. There remains a possi-bility that we have failed to adequately reflect the nature of Eastern stereo-types, to the extent that these may be distinct from those found in the West.Further research could usefully explore the content of Eastern stereotypesand develop specifically Eastern measures.

    Fifth, our analysis of discriminatory attitudes uses single-itemmeasures, which increases the likelihood of measurement error. Futureresearch needs to use reliable multi-item measures of discriminatory atti-tudes. Also, this analysis involves the estimation of five separate regressionequations, thus increasing the risk of capitalizing on chance, although toreduce this risk we are cautious about making inferences on the basis of mar-ginal (i.e. greater than 0.05 but less than 0.10) levels of significance.

    Finally, our concern in this article has been with general rather than job-specific age stereotypes. In contrast, the job stereotype hypothesis suggeststhat evaluations of older workers are influenced by the nature of the job inquestion. In particular, older workers doing or being considered for jobs whichare generally considered as older peoples jobs are likely to be evaluatedfavourably relative to their younger counterparts, whilst they will be morenegatively evaluated when doing or being considered for stereotypical young

    Chiu et al. Age stereotypes and discrimination against older workers 6 5 7

    04chiu (ds) 29/3/01 7:35 am Page 657

  • peoples jobs (Cleveland & Landy, 1987). The job stereotype hypothesis thussees stereotyping occurring at the level of the job and of the individual. Thereis considerable evidence of job stereotyping by age (Cleveland & Landy, 1987;Gordon & Arvey, 1986; Institute of Personnel Management, 1993; Lawrence,1988). Furthermore, Finkelstein et al. (1995) found some evidence thatyounger people were rated more positively in young jobs, although olderworkers were not rated more highly in old jobs and younger people weremore highly rated in age-neutral jobs. Perry et al. (1996) also found evidenceof such an asymmetric interaction between age and job type, suggesting thatthe negative consequences of being in an age-incongruent job may be offset inthe case of younger people in old-type jobs due to their appearing as highfliers when associated with an older and often more senior or prestigiousjob. These issues remain unexplored in the Eastern context and this is an areawhere further research would be of value.

    Notes

    1 Employers have often been characterized as being opposed to legislation, althoughit is interesting that Taylor and Walkers (1994) UK survey found that 53 percent ofemployers were in favour of legislation on age discrimination, with only 36 percentagainst.

    2 There is, of course, likely to be a correlation between organizational size and thepresence of an equal opportunities policy. The former is used by us as a proxy forthe sophistication and formalization of human resources management and manage-ment practices generally, whilst the latter relates to a specific policy initiative aimedat combating age discrimination. We felt it was important to include both, since evenin the absence of a specific equal opportunities policy, management practice mayhave an impact on discrimination and employee attitudes.

    3 Employment size was responded to as a categorical variable. Because few respon-dents were in the three smaller size categories (under 20, 2049 and 5099 employ-ees), we combined these into a single category.

    4 Respondents categorized their organization into one of seven industry categories. Wecombined these into a single services/manufacturing variable.

    5 The relative prevalence of age discrimination policies did not surprise us, althoughthe figure of 43 percent of UK respondents organizations having a stipulated policynot to allow age discrimination appears rather high. Arrowsmith and McGoldrick(1996) reported that 63 percent of their Institute of Management respondents hadwritten equal opportunities policies, but that only a third of these specifically men-tioned age. Our higher figure may reflect the composition of our sample, with two-thirds of organizations employing over 500 people. Also, our question may havebeen interpreted by respondents in terms of an understanding of implicit organiz-ational practice.

    Human Relations 54(5)6 5 8

    04chiu (ds) 29/3/01 7:35 am Page 658

  • References

    Arrowsmith, J. & McGoldrick, A. Breaking the barriers: A survey of managers attitudesto age and employment. London: Institute of Management, 1996.

    Bird, C. & Fisher, T. Thirty years later: Attitudes towards the employment of older workers.Journal of Applied Psychology, 1986, 71(3), 51517.

    Bond, M.H. Chinese values. In M.H. Bond (Ed.), The handbook of Chinese psychology.Hong Kong: Oxford University Press, 1996.

    Bond, M.H. & Hwang, K. The social psychology of the Chinese people. In M.H. Bond(Ed.), The psychology of the Chinese people. Hong Kong: Oxford University Press,1986.

    Brown, R., Condor, S., Mathews, A., Wade, G. & Williams, J. Explaining intergroup dif-ferentiation in an industrial organization. Journal of Occupational Psychology, 1986,59, 27386.

    Chen, J. Confucius as a teacher. Beijing: Foreign Language Press, 1990.Cleveland, J.N. & Landy, F.J. Sage perceptions of jobs: Convergence of two questionnaires.

    Psychological Reports, 1987, 60, 107581.Debrah, Y.A. Tackling age discrimination in employment in Singapore. International

    Journal of Human Resource Management, 1996, 7(4), 81331.Department for Education and Employment. Action on age: Report of the consultation on

    age discrimination in employment. London: DfEE, 1998.Doering, M., Rhodes, S. & Schuster, M. The ageing worker. Beverley Hills, CA: Sage, 1983.Education and Manpower Branch. Equal opportunities: A study on discrimination in

    employment on the ground of age. Consultation paper. Hong Kong: GovernmentPrinter, 1996.

    Finkelstein, L.M., Burke, M.J. & Raju, N.S. Age discrimination in simulated employmentcontexts: An integrative analysis. Journal of Applied Psychology, 1995, 80(6), 65263.

    Gordon, R.A. & Arvey, R.D. Perceived and actual ages of workers. Journal of VocationalBehavior, 1986, 28, 218.

    Hassell, B.L. & Perrewe, P.L. An examination of beliefs about older workers: Do stereo-types still exist? Journal of Organizational Behavior, 1995, 16, 45768.

    Hofstede, G. Cultures consequences: International differences in work-related values. Beverley Hills, CA: Sage Publications, 1980.

    Hong Kong Institute of Human Resource Management. A code of professional conduct &good practice for human resource specialists. Hong Kong: HKIHRM, 1995.

    Hong Kong Industrial Relations Association. Human resource management: Guides to pro-fessional conduct [in Chinese]. Hong Kong: HKIRA, 1995.

    Institute of Personnel Management. Age and employment: Policies, attitudes and practice.London: IPM, 1993.

    Kirchner, W. & Dunnette, M. Attitudes towards older workers. Personnel Psychology,1954, 7, 25765.

    Kirchner, W., Lindbom, T. & Paterson, D. Attitudes towards the employment of olderpeople. Journal of Applied Psychology, 1952, 36, 1546.

    Kite, M.E. & Johnson, B.T. Attitudes towards older and younger adults: A meta-analysis.Psychology and Aging, 1993, 3, 23344.

    Labour Department. A concise guide to the Employment Ordinance. Hong Kong: Govern-ment Printer, 1997.

    Lawrence, B.S. New wrinkles in the theory of age: Demography, norms, and performanceratings. Academy of Management Journal, 1988, 31(2), 30937.

    Lucas, R. Some age-related issues in hotel and catering employment. Service IndustriesJournal, 1995, 15(2), 23450.

    Lyon, P. & Pollard, D. Perceptions of the older employee: Is anything really changing? Per-sonnel Review, 1997, 26(4), 24557.

    Chiu et al. Age stereotypes and discrimination against older workers 6 5 9

    04chiu (ds) 29/3/01 7:35 am Page 659

  • McEvoy, G.M. & Cascio, W.F. Cumulative evidence of the relationship between employeeage and job performance. Journal of Applied Psychology, 1989, 74(1), 1117.

    McGoldrick, A.E. & Arrowsmith, J. Recruitment advertising: Discrimination on the basisof age. Employee Relations, 1993, 15(5), 5465.

    McKay, S. Older workers in the labour market. Labour Market Trends, July 1998, 3659.Metcalf, H. & Thompson, M. Older workers: employers attitudes and practices. IMS

    Report No. 194, Institute of Manpower Studies, June 1990.Perry, E.L, Kulik, C.T. & Bourhis, A.C. Moderating effects of personal and contextual

    factors in age discrimination. Journal of Applied Psychology, 1996, 81(6), 62847.Rosen, B. & Jerdee, T. The nature of job-related age stereotypes. Journal of Applied Psy-

    chology, 1976a, 61, 1803.Rosen, B. & Jerdee, T. The influence of age stereotypes on managerial decisions. Journal of

    Applied Psychology, 1976b, 61, 42832.Rosen, B. & Jerdee, T. Too old or not too old? Harvard Business Review, 1977, 55, 97106.Sparrow, P., Schuler, R. & Jackson, S. Convergence or divergence: Human resource prac-

    tices and policies for competitive advantage worldwide. International Journal of HumanResource Management, 1994, 5(2), 26799.

    Taylor, P. & Walker, A. What are employers doing to defuse the demographic time-bomb?Skills and Enterprise Briefing, 1992, 21.

    Taylor, P. & Walker, A. The aging workforce: Employers attitudes towards older people.Work, Employment and Society, 1994, 8(4), 56991.

    Taylor, P. & Walker, A. Age discrimination and public policy. Personnel Review, 1997,26(4), 30718.

    Taylor, P. & Walker, A. Policies and practices towards older workers: A framework for com-parative research. Human Resource Management Journal, 1998, 8(3), 6176.

    Tillsley, C. The impact of age upon employment. Warwick Papers in Industrial Relations,33, 1990.

    Thompson, M. Last in the queue? Corporate employment policies and the older workers.IMS Report No. 209, Brighton: Institute of Manpower Studies, 1991.

    Waldman, D.A. & Avolio, B.J. A meta-analysis of age differences in job performance.Journal of Applied Psychology, 1986, 71(1), 338.

    Walker, A. & Taylor, P. Ageism vs productive ageing: The challenge of age discriminationin the labour market. In S. Bass, F. Caro & Y. Chen (Eds), Achieving a productive ageingsociety. Westport, CT: Greenwood, 1993.

    Warr, P. Age and employment. In H.C.Triandis, M.D. Dunnette & L.M. Hough (Eds),Handbook of industrial and organizational psychology, Vol. 4. Palo Alto, CA: Con-sulting Psychologists Press, 1994, pp. 485550.

    Warr, P. & Pennington, J. Views about age discrimination and older workers. In Age andemployment: Policies and practices. London: IPM, 1993.

    Wong, J.W.P. An artificial barrier to employment. Hong Kong Staff, 1995, December, 223.

    Human Relations 54(5)6 6 0

    04chiu (ds) 29/3/01 7:35 am Page 660

  • Chiu et al. Age stereotypes and discrimination against older workers 6 6 1

    Warren C.K. Chiu is an Associate Professor in the Department ofManagement, Hong Kong Polytechnic University. He is an industrial andorganizational psychologist whose current research interests include agediscrimination, women in management, stress management and enter-prise reform in China.[E-mail: [email protected]]

    Andy W. Chan is an Assistant Professor in the Department of Manage-ment, Hong Kong Polytechnic University. His current research includesorganizational communication, joint consultation, age discrimination,trade unionism development in Hong Kong and human resourcesmanagement practices in China. He has published articles in academicjournals such as British Journal of Industrial Relations and Industrial Relations.[E-mail: [email protected]]

    Ed Snape was Professor of Human Resource Management at the Uni-versity of Bradford Management Centre, UK, but has recently joined theDepartment of Management at the Hong Kong Polytechnic University.Hiscurrent research includes projects on employee commitment to companyand union, age discrimination and performance appraisal.[E-mail: [email protected]]

    Tom Redman was Reader in Human Resource Management at theSchool of Business and Management, University of Teesside, UK, but hasrecently joined the University of Durham. His current research interestsfocus on human resources management strategy, union commitment, agediscrimination, employee involvement, organizational downsizing andcomparative studies of performance appraisal practice. He is the Editorof Personnel Review.[E-mail: [email protected]]

    04chiu (ds) 29/3/01 7:35 am Page 661