Abstract
This article provides an empirical assessment of the relationship between unionization and the adoption of equal opportunities (EO) policies and practices in British workplaces, using data from the 2004 Workplace Employment Relations Survey. The results demonstrate an association between union recognition and the adoption of a range of EO practices. However, this association largely only holds where unions are able to influence EO decision-making via negotiation or consultation. There is little evidence that the equality practices adopted in unionized workplaces where negotiation and consultation does not occur are any different from those adopted in non-unionized workplaces.
Introduction
The impact of trade unions on discrimination and labour market inequality has been a matter of public concern for many years. It is often argued that, historically, unions did little to support the interests of women and minority groups, and if anything reinforced rather than challenged inequalities (Colling and Dickens, 1998: 406; Dex and Forth, 2009: 237; Dickens, 1997: 287). More recently, however, observers have suggested that unions have increasingly sought to prioritize equality matters (Dex and Forth, 2009: 246; Dickens, 1999: 15; Hunt and Rayside, 2000). There is a growing expectation, for example, that they will actively bargain over equality and encourage employers to adopt equality practices in order to protect workers from discrimination. Some commentators have suggested that unions now have a central role to play in advancing the equality agenda and ultimately helping promote a more inclusive society (Dickens, 2007; Heery, 2006).
This article seeks to address these debates by providing an empirical assessment of the relationship between unionization and the adoption of equal opportunities (EO) policies and practices in British workplaces, using data from the 2004 Workplace Employment Relations Survey (Department of Trade and Industry, 2005). Considerable previous research has sought to address this issue, with several studies reporting that unions in Britain have had a positive effect in advancing the equality agenda in recent years (Dickens, 2007: 484; Metcalf et al., 2001). This is demonstrated by the greater prevalence of formal written EO policies in unionized workplaces than elsewhere (Dex and Forth, 2009: 243, 250; Hoque and Noon, 2004: 488; Kersley et al., 2006: 238; Walsh, 2007: 305), and also by the relationship that has been identified between unionization and the adoption of specific EO practices. For example, Kersley et al. (2006: 248) and Walsh (2007: 307) point to the greater prevalence of recruitment and selection monitoring by gender in unionized workplaces, while Bewley and Fernie (2003: 102) point to the greater prevalence of promotion monitoring and reviews of selection procedures by gender to identify indirect discrimination. Several recent studies also report that family-friendly practices in Britain are more widespread in unionized than non-unionized workplaces (Bewley and Fernie, 2003; Budd and Mumford, 2004; Dex and Smith, 2002).
This article seeks to expand on this research by exploring the proposition that while there may be a positive association between unionization and workplace equality practice overall, this relationship might be more pronounced in certain circumstances. In particular, it addresses the argument that the union effect might be particularly strong in instances where equality decision-making is subject to negotiation. In addition, whereas earlier studies focus mainly on the impact of unions and equality bargaining on gender equality practices, this article expands on these studies by considering practices covering a range of equality strands (ethnicity, disability and age as well as gender), thereby providing much needed research evidence on these other aspects of discrimination (Briskin, 2006: 12; Dex and Forth, 2009: 232).
Unions and negotiation over equality
The 2004 Workplace Employment Relations Survey (WERS) suggests that negotiation over equal opportunities issues occurs in 15 per cent of British workplaces with a recognized union (Kersley et al., 2006: 194). There is, however, some debate over what such negotiations are likely to achieve. On the one hand, the collective voice/institutional response model of unionism suggests that where unions voice female and minority employee demands for equality, employers may respond by providing improved equality practices (Freeman and Medoff, 1984). Collective voice effects of this nature are likely to be stronger in instances where unions can use their monopoly power within negotiations to extract improvements from management (Budd and Mumford, 2004: 206; Dickens et al., 1988). In addition, Dickens (1997: 288–9) argues that negotiation over equality can mainstream equality initiatives that might otherwise remain contingent and partial. It can also provide a mechanism by which legal rights might be converted into substantive outcomes (Dickens, 1999: 15; Heery, 2006: 522). Beyond this, it can affect employers’ cost–benefit analysis of taking equality action by increasing the cost of failing to do so (Colling and Dickens, 1998: 405).
On the other hand, there are several factors that might limit the impact of negotiation over equality. First, while getting equality onto the bargaining agenda might be an important step forward, the impact this subsequently has on employer equality practices depends on whether equality matters are prioritized within negotiations (Blackett and Sheppard, 2003: 437). This is unlikely to happen if women and/or minority groups are under-represented within union negotiating teams (Baird et al., 2009: 674–5; Dickens 1997: 288), as is frequently the case (Hart, 2002: 614).
Second, decentralized bargaining structures may have deleterious effects. As Colling and Dickens (1998: 400) argue, both union and employer commitment to equality, and also the knowledge and expertise necessary to facilitate meaningful negotiations over equality, are generally greater at corporate or national level than at local level. Given this, it is perhaps unsurprising that decentralized collective bargaining systems, according to Blackett and Sheppard (2003: 442), tend to accentuate enterprise-specific productivity or efficiency concerns rather than broader social concerns.
Third, it is widely argued that negotiation over equality will be more effective where it is underpinned by legal rights for equality, as these provide a ‘useful lever’ (Dickens, 1997: 288) for unions within negotiations (Blackett and Sheppard, 2003: 451; Gilbert and Secker, 1995: 204; Heery, 2006: 540). However, compliance with EU equality law in Britain has been somewhat minimalist (Colling and Dickens, 1998: 392), the result being a weak, individual enforcement model that while attempting to balance efficiency and justice concerns has tended to favour the former over the latter (Dickens, 2007).
A final argument relates to general union weakness, particularly in the private sector. As Colling and Dickens (2001) argue, the strength of unions to extract concessions from management within negotiations over equality (and within collective bargaining more broadly) is becoming increasingly questionable given the hostility of the environment in which unions operate and also given that employers may increasingly perceive less of a credible strike threat from unions.
The above arguments suggest, therefore, that there may be limits to what negotiation over equality is able to deliver. That said, much of the empirical evidence to date has suggested that where such negotiation occurs, successful outcomes generally emerge (see, for example, Baird et al., 2009: 687; Dickens et al., 1988: 65; Heery, 2006: 539). However, as Heery (2006: 522) argues, much of the evidence is case-based, and given this, there is a need for quantitative analyses of representative data to generalize the findings to the wider population of organizations. Given this, the first two aims of this article are to draw on data from WERS 2004 in order to identify: first, the extent to which there is an overall positive relationship between union recognition and employer equality practice (as identified within earlier studies); and second, the extent to which the relationship between unionization and employer equality practice is stronger in instances where equality decision-making is subject to negotiation than in instances where the union can exercise voice via consultation but cannot negotiate, or where the union is not involved in employer decision-making on equality whatsoever.
In considering this second aim, however, one must bear in mind that employer equality practice in many British workplaces is poor, with EO policies often representing little more than an ‘empty shell’ (Hoque and Noon, 2004). As highlighted by WERS 2004, less than one in four British workplaces monitor or review their procedures to identify indirect discrimination with respect to gender, ethnicity, disability or age. In addition, only 1 in 10 workplaces monitor and review procedures with respect to promotions, while relative pay rates are reviewed by ethnicity in only 5 per cent of workplaces and by gender in 7 per cent of workplaces (Walsh, 2007: 307). Hence, even if the analysis finds that equality policies are more prevalent in workplaces where negotiation over equality occurs, equality practices within such workplaces may still be notable by their absence. Given this, the third aim of the article is to evaluate the extent of adoption of equality practices in instances where negotiation over equality takes place, and whether the approach to equality in such instances can be considered to be one of substance.
The final aim of the article is to assess the relative influence of negotiation over equality in the public and private sectors. One might anticipate that its influence will be stronger in the public sector, where a higher union density and a higher percentage of female employees (Kersley et al., 2006: 112) alongside more positive employer attitudes (Kumar, 1993: 223) may provide public sector unions with greater bargaining power to extract concessions from management (see also Kirton and Greene, 2006: 432–3). In addition, the introduction of statutory duties to promote equality, the Single Status Agreement in local government and Agenda for Change in the NHS may have provided public sector unions with additional leverage not available to their private sector counterparts. In considering this issue, however, it is worth keeping in mind that the introduction of a statutory duty for public sector employers to promote race equality in 2001 did not prevent a reduction in the proportion of public sector employers using special procedures to attract job applicants from ethnic minority groups (Dex and Forth, 2009: 251; Kersley et al., 2006: 244). One should not automatically assume, therefore, that the apparently more favourable institutional environment in the public sector will increase the effectiveness of negotiation over equality.
Data and methods of analysis
The analysis uses data from the WERS 2004 management survey. This comprises 2295 observations with a response rate of 64 per cent (Kersley et al., 2006) and is designed to be nationally representative of workplaces with 5 or more employees within Standard Industrial Classification major groups D to O (agriculture, hunting, forestry and fishing and mining and quarrying are excluded), when probability weighted to take into account the complex nature of the WERS survey design. The respondent to the survey is the manager at the workplace who has primary responsibility for employment relations matters. The representativeness of WERS makes it an ideal data set by which to estimate macro-level union effects on workplace equality practice.
The first aim of the article is to estimate the relationship between union recognition and the adoption of workplace EO policy and practice. This is assessed in a series of equations within which the dependent variables relate to the EO policies and practices in operation at the workplace. Appendix 1 describes the dependent variables used and gives their means. Given the dichotomous nature of these variables, maximum likelihood survey probit is used. In order to identify whether the policies and practices asked about are more prevalent in unionized workplaces than elsewhere, a dichotomous independent variable is included in the equations where 1 = recognized union (22% of workplaces) and 0 = otherwise. Several observable characteristics that might affect the association between union recognition and the adoption of EO policies and practices are controlled for. These include workplace and organization size, Standard Industrial Classification major group, whether the workplace is in the public or private sector, whether the workplace is a single independent establishment, financial performance, national ownership, workplace age and the proportion of workforce that is female, ethnic minority, disabled or 50 years old or older. The means of the control variables are given in Appendix 1.
The article’s second key aim is to assess whether the association between union recognition and employer equality policy and practice is more pronounced in instances where negotiation over equality occurs than elsewhere. The survey asks respondents whether managers normally negotiate, consult with or inform trade unions when deciding EO issues. This allows for the creation of a five-part categorical variable 1 as follows:
Managers negotiate with unions when deciding EO issues (3.7% of workplaces)
Managers consult unions when deciding EO issues (9.6% of workplaces)
Managers provide unions with information on EO (4.1% of workplaces)
Managers do not involve unions when making EO decisions (4.3% of workplaces)
Non-union workplaces (78.3% of workplaces) 2
Replacing the dichotomous union recognition independent variable in the probit equations described above with this categorical variable enables an evaluation of whether the relationship between unionization and the adoption of equality policies and practices is stronger in instances where negotiation over equality occurs than in instances where unions can exercise voice effects via consultation but are unable to negotiate, or are unable to influence employer decision-making whatsoever.
The article’s third aim is to evaluate whether the approach to equality in instances where negotiation over equality can be considered to be one of substance. To address this, a 12-point scale 3 is constructed based on the dependent variables described in Appendix 1. This scale is then used to assess the number of equality practices adopted where negotiation over equality occurs relative to the number adopted elsewhere. The same scale is used to evaluate the article’s fourth aim of addressing whether the influence of negotiation over equality is stronger in the public than the private sector, by evaluating whether negotiation raises the number of practices used to a greater extent in the former than the latter sector.
The data are weighted throughout the analysis by the inverse of each workplace’s probability of selection into the sample. This is essential if unbiased population estimates are to be obtained.
Results
The presence of equal opportunity policies
Table 1 presents the analysis of the relationship between unionization and the presence of a formal written EO policy that makes explicit reference to gender, race, disability and age. There is no evidence that such policies are any more prevalent in unionized workplaces than elsewhere. The overall union recognition variable does, however, mask variation within the union sector, with gender, race, disability and age EO policies all being more widespread in workplaces where the union is consulted over equality than in non-union workplaces. It is perhaps notable that it is only consultation that makes a difference – there is no evidence that formal written EO policies are more prevalent in unionized workplaces where negotiation over equality occurs than in non-union workplaces.
Association between unionization and the presence of an equal opportunities (EO) policy.
Notes: Survey probit analysis. Coefficients given, standard errors in brackets.
Controls include: workplace size, standard industrial classification (SIC) major group, organization size, public sector, national ownership, workplace age, single independent workplace, financial performance. The equations also control for the following workforce characteristics: proportion of workforce female (column 1); proportion of workforce ethnic minority (column 2); proportion of workplace disabled (column 3); proportion of workforce 50 years old or older (column 4).
Significant at 1 per cent; **significant at 5 per cent.
Although it would appear that formal written EO policies are no more prevalent in unionized workplaces than in non-unionized workplaces, it is nevertheless possible that EO policies in unionized workplaces are more likely to be policies of substance. As discussed earlier, previous research suggests that many EO policies in Britain are ‘empty shells’, being little more than statements of good intent that contain few supporting EO practices (Hoque and Noon, 2004). Given this, the following sections seek to identify whether specific EO practices are more likely to have been adopted in unionized workplaces, and whether the association between unionization and the adoption of equality practices is greater in instances where negotiation over equality occurs.
Monitoring/reviewing of recruitment and selection and promotion
Table 2 explores whether recruitment and selection and promotions are monitored and reviewed to identify indirect discrimination by gender, race, disability and age. The results suggest that, overall, monitoring and reviewing of this nature is more common in workplaces with union recognition than elsewhere. In terms of whether the association is stronger where negotiation over equality occurs, the results suggest that the prevalence of recruitment and selection monitoring and reviewing, and the prevalence of promotion monitoring (but not promotion reviewing) is higher in instances where such negotiation occurs than in non-union workplaces. By contrast, there is no evidence that either recruitment and selection or promotion monitoring and reviewing is more widespread where the union is not involved in EO decision-making, and there is only very weak evidence to suggest that it is more widespread where the union is informed over EO decisions than in non-union workplaces.
Association between unionization and recruitment and selection monitoring/reviewing.
Notes: N = 1702 for all equations. Survey probit analysis. Coefficients given, standard errors in brackets. Controls as described in Table 1.
Significant at 1 per cent; **significant at 5 per cent; *significant at 10 per cent.
EO = equal opportunities.
However, the results also demonstrate that, as for negotiation, recruitment and selection monitoring and reviewing and promotion monitoring are more prevalent where consultation over equality occurs. In addition to this, promotion reviews are also more prevalent. Hence, the union effect is arguably stronger where unions are able to exercise voice effects via consultation than in instances where negotiation over equality occurs.
Special procedures to encourage applications from disadvantaged groups
Table 3 suggests that special procedures to encourage applications from ethnic minorities and disabled people are more prevalent among workplaces with union recognition than elsewhere, and that this association holds irrespective of whether the union is involved in equality decision-making. However, although special procedures to encourage applications from older workers and from women returning to work after having children are, overall, no more prevalent in unionized workplaces, they are nevertheless more prevalent among workplaces where negotiation over equality occurs than in non-union workplaces. This points to the importance of negotiation where these elements of employer equality practice are concerned.
Association between unionization and special procedures to encourage applications from disadvantaged groups.
Notes: Survey probit analysis. Coefficients given, standard errors in brackets. Controls as described in Table 1.
Significant at 1 per cent; **significant at 5 per cent.
EO = equal opportunities.
Reviews of relative pay rates
Table 4 explores reviews of relative pay rates by gender, race, disability and age. The results suggest that reviews by race are more prevalent, and reviews by disability are slightly more prevalent (at the 10% significance level), in workplaces with union recognition than elsewhere.
Association between unionization and reviews of relative pay rates.
Notes: Survey probit analysis. Coefficients given, standard errors in brackets. Controls as described in Table 1.
Significant at 1 per cent; **significant at 5 per cent; *significant at 10 per cent.
EO = equal opportunities.
In terms of whether the association is stronger in instances where negotiation over equality occurs, the results suggest that reviews of relative pay rates by race are more prevalent where such negotiation occurs than in non-union workplaces, and reviews by disability and age are slightly more prevalent (at the 10% significance level). By contrast, reviews of relative pay rates are no more prevalent in workplaces where the union is not involved in EO decision-making, while in workplaces where the union is informed about EO decisions, reviews of relative pay rates are less prevalent than in non-union workplaces.
However, the results also suggest that reviews of relative pay rates by race, disability and age are all more prevalent where consultation over equality occurs than in non-union workplaces, with the relationships for reviews by disability and age being stronger than where negotiation over equality occurs. Again, therefore, the results suggest that the union effect is stronger where unions are able to exercise voice effects via consultation than where negotiation over equality occurs.
Flexible working and family-friendly practices
In terms of the relationship between unionization and flexible working and family-friendly practices, the evidence presented in Table 5 suggests that four of the eight practices asked about (workplace nursery, financial help with childcare, leave for carers of older adults and maternity leave at full rate of pay) are more widespread, and job sharing is slightly more widespread (at the 10% significance level), within unionized workplaces than elsewhere.
Association between unionization and flexible working/family-friendly practices.
Notes: N = 1702 for all equations. Survey probit analysis. Coefficients given, standard errors in brackets.Controls as described in Table 1. All equations control for the proportion of workforce female with the exception of the ‘leave for carers of older adults’ equation.
Significant at 1 per cent; **significant at 5 per cent; *significant at 10 per cent.
EO = equal opportunities.
There is also some evidence pointing to the importance of negotiation, with workplace nurseries being more prevalent, and flexitime, financial help with childcare and leave for carers of older adults being slightly more prevalent (at the 10% significance level) in workplaces where negotiation over equality occurs than in non-union workplaces. In workplaces where the union is informed about equality decisions, job sharing and financial help with childcare are more prevalent, and workplace nurseries are slightly more prevalent than in non-union workplaces. However, flexitime is less prevalent. Where the union is not involved at all in equality decision-making, working at or from home in normal working hours, term-time only contracts are less widespread, and workplace nurseries are slightly less widespread than in non-union workplaces.
However, the association between unionization and the adoption of flexible working and family-friendly practices is particularly strong where consultation with the union over equality occurs, with six of the eight flexible working/family-friendly practices being more widely adopted in such workplaces than in non-union workplaces. Hence, the union effect on the adoption of flexible working and family-friendly practices would appear to be greater in instances where unions are able to exercise voice via consultation than in instances where negotiation over equality occurs.
The extent of adoption of equality practices and public/private sector differences
Overall, the results reported thus far demonstrate that the association between unionization and the adoption of a range of equality practices largely holds only in instances where negotiation or consultation over equality occurs. However, what the analysis to this point does not demonstrate is how extensively these practices are used within such workplaces. As argued earlier, while the uptake of employer equality practices may be higher than elsewhere it still might not be particularly high given that the usage of these practices is very low within British workplaces as a whole (Walsh, 2007). An exploration of this issue using the 12-point scale described earlier, however, shows that only 2 per cent of non-union workplaces and 14 per cent of unionized workplaces where the union is not involved in EO decision-making have adopted 6 or more of the 12 practices asked about. By contrast, the figure for unionized workplaces where consultation occurs is 35 per cent and where negotiation occurs it is 39 per cent. Thus, the approach taken to equality would indeed seem more likely to be of substance where either negotiation or consultation takes place.
That said, a significant proportion of workplaces in which negotiation or consultation occurs have adopted very few equality practices. In particular, 26 per cent of unionized workplaces where consultation occurs and 27 per cent of unionized workplaces where negotiation occurs have adopted only two or fewer practices. Although this compares favourably with figures of 51 per cent for workplaces where unions are only informed on equality decisions or are not involved in equality decision-making, and 70 per cent for non-union workplaces, it nevertheless suggests that there is an absence of EO practices in a significant minority of workplaces where negotiation or consultation over equality takes place.
The final aim of the article is to identify whether the influence of negotiation over equality is stronger in the public than the private sector. The results using the 12-point scale described earlier point to considerable evidence for negotiation and consultation effects in both sectors. In the public sector, 9 per cent of non-union workplaces, 20 per cent of unionized workplaces where the union is not involved in EO decision-making, 44 per cent of unionized workplaces in which consultation occurs and 40 per cent of unionized workplaces in which negotiation occurs have 6 or more of the 12 practices in place. In the private sector, the corresponding figures are 2 per cent, 12 per cent, 23 per cent and 39 per cent. Therefore, not only is there significant evidence of positive negotiation and consultation effects in both sectors, but if anything, negotiation effects appear to be more pronounced in the private than the public sector.
Discussion and conclusions
This article provides an empirical assessment of the relationship between unionization and the adoption of EO policies and practices in British workplaces. Where formal written EO policies are concerned, the analysis reported here suggests that these are no more likely to have been adopted in unionized workplaces than elsewhere. This is contrary to the findings of earlier research based on WERS 1998 (Hoque and Noon, 2004). It would seem, therefore, that the gap has closed between the union and non-union sectors with regard to this issue. Notably, the results appear to contradict Dex and Forth (2009: 250), Kersley et al. (2006: 238) and Walsh (2007), who report, also using WERS 2004 data, that EO policies are more prevalent in the union sector. While these authors are correct to point to these differences within the analyses they present, the differences disappear within the multivariate analysis conducted here, which controls for a range of observable workplace characteristics that might bias estimates of this relationship.
A different picture emerges with regard to EO practices. Here, the analysis finds that a range of EO practices (recruitment and selection and promotion monitoring and reviewing, special procedures to encourage applications from (some) disadvantaged groups, and reviews of relative pay rates by race and disability) are more prevalent in unionized than non-unionized workplaces. A number of family-friendly and flexible working practices (job sharing, workplace nurseries, financial help with childcare, leave for carers of older adults and maternity leave at the full rate of pay) are also more prevalent in unionized workplaces. The analysis therefore supports earlier research pointing to the greater uptake of such practices in unionized workplaces (Bewley and Fernie, 2003; Budd and Mumford, 2004; Dex and Smith, 2002). That said, there is no consistent evidence of a union effect in the 4 in 10 unionized workplaces where negotiation or consultation over equality does not occur – indeed, some equality practices are less prevalent in these workplaces than in non-union workplaces. This suggests that the positive union influence on employer equality practice might be benefitting a narrower segment of the female and minority workforce than has been assumed on the basis of previous research.
The second aim of the article was to identify whether the union effect on equality is particularly pronounced in instances where negotiation over equality occurs. The results demonstrate that the union effect is indeed greater in such instances than where unions are only informed about equality decisions or are not involved in equality decision-making. However, the union effect is equally if not more pronounced where unions are able to exercise voice effects via consultation. There are a number of ways this finding might be interpreted. One interpretation would be that consultation provides an equally if not more effective route than negotiation by which the joint regulation of equality might occur. Hence, ‘soft’ forms of joint regulation via consultation may provide greater scope for unions to influence employer equality practice than ‘hard’ joint regulation via collective bargaining. The results could also be viewed as supportive of the arguments made by several commentators discussed earlier that the potency of equality bargaining may be limited in Britain given: a lack of women in senior union negotiating positions; the decentralized nature of bargaining; the limited nature of equality legislation; and general union weakness.
Against this, however, the strength of the consultation effect could also be interpreted as reflecting the possibility that employers who are prepared to consult with unions on equality are perhaps more positively disposed in the first instance to the introduction of equality practices than are employers that need to be pressured into discussing equality with unions via negotiation. Hence, the apparently stronger consultation effect might not be an indication that consultation is an inherently more effective route by which unions might bring about change, but instead is an indication of employers’ predispositions to equality. The fact that consultation effects appear stronger in the public than the private sector may be testimony to this. Either way, further research addressing the differential impact of negotiation and consultation over equality would provide useful insights into the effectiveness of the different routes by which unions might be able to influence employer equality practice.
The third aim of the article was to identify whether the approach taken to equality in workplaces where negotiation over equality occurs can be considered to be one of substance. The analysis found that although employer equality practice is better in instances where negotiation (and consultation) occurs than in non-union workplaces, it remains the case that in over a quarter of such workplaces, equality practices are notable by their absence. This further suggests the caution urged in the literature in terms of what negotiation over equality might achieve may be entirely justified. It also suggests that the factors that limit the potency of negotiation might also limit the ability of unions to extract concessions from management through consultation.
The final aim of the article was to explore whether negotiation effects are more pronounced in the public than the private sector. In the event, the analysis found the opposite. While this may be testimony (as suggested above) to the extent to which private sector employers need to be pressured via collective bargaining into making improvements in equality practice, it is nevertheless perhaps ironic, as noted by Dickens (1997: 289), that while negotiation over equality would appear to be leading to positive gains in the private sector, the decline in private sector union coverage over the past three decades means that such gains will benefit relatively few women and minority groups.
It is worth raising a caveat with regard to the results presented here. Although equality practices are more widespread where negotiation and consultation over equality occurs, there is no guarantee that the practices will apply to the whole workforce or lead to better equality outcomes. As argued by Hoque and Noon (2004), equality practices on occasion apply only to certain segments of the workforce, and hence for large sections of the (particularly non-managerial) workforce, they are little more than ‘empty shells’. As such, while EO practices may be more widespread where negotiation and consultation occurs, there is no guarantee that these practices will actually apply to or result in equal treatment for all sectors of the female and minority workforce. Research on this issue would therefore provide useful insights into the precise nature and extent of the influence of unions on workplace equality.
Finally, the results have implications for both unions and government policy. Where unions are concerned, in recent times many unions have made commitments to ensuring equality of opportunity for their members (Dickens, 2007; Heery, 2006). The results here suggest that if they are serious in these commitments, there are two key routes they might pursue. First, it is essential (notwithstanding the barriers they might face regarding employer reluctance for equality to be determined via negotiation) that they seek to ensure that equality is included on the bargaining agenda, and that they encourage negotiating officers to prioritize equality matters. Second, it is essential that they engage with employers on equality via consultation processes, and where consultation over equality does not occur, this is something they press for. Should they do so, the results presented here suggest that this may enable them to have a significant influence on the equality practices employers subsequently adopt.
Where government policy is concerned, during its time in office the Labour government made efforts to encourage employers to adopt flexible and family-friendly practices, given the potential for such practices to influence workforce morale and productivity positively (Bloom and Van Reenan, 2006). The Conservative-led Coalition government has also made much of its commitment to fairness and the family. The Conservative manifesto, for example, made a commitment for Britain to become ‘the most family-friendly country in Europe’ (Conservative Party, 2010: 42), and it viewed employer provision of family-friendly practices as central in achieving this. The results presented in this article demonstrate that such practices are particularly prevalent in instances where negotiation and consultation with unions on equality occurs. Hence, one route by which the Coalition government might achieve its goals would be to encourage employers to engage in genuine joint regulation of equality with unions via either meaningful consultation or negotiation. The record of previous Conservative administrations in terms of union engagement suggests, however, that this is unlikely to happen. It is probable, therefore, that important avenues by which employer equality practices might be improved, and by which the government might achieve its commitment to fairness, will remain overlooked for the foreseeable future.
Footnotes
Appendix
Means of dependent and control variables
| Dependent variables | |
| Workplace has a formal written policy on equal opportunities (EO) or managing diversity that explicitly mentions equality of treatment or discrimination on the grounds of: | |
| Gender | 0.561 |
| Race | 0.551 |
| Disability | 0.534 |
| Age | 0.419 |
| Recruitment and selection monitored by: | |
| Gender | 0.186 |
| Ethnicity | 0.185 |
| Disability | 0.183 |
| Age | 0.175 |
| Recruitment and selection procedures reviewed to identify indirect discrimination by: | |
| Gender | 0.164 |
| Ethnicity | 0.173 |
| Disability | 0.164 |
| Age | 0.142 |
| Special procedures to encourage applications from: | |
| Women returning to work after having children | 0.107 |
| Members of ethnic minority groups | 0.083 |
| Disabled people | 0.091 |
| Older workers | 0.062 |
| Promotions monitored by: | |
| Gender | 0.088 |
| Ethnicity | 0.086 |
| Disability | 0.079 |
| Age | 0.066 |
| Promotion procedures reviewed to identify indirect discrimination by: | |
| Gender | 0.095 |
| Ethnicity | 0.095 |
| Disability | 0.092 |
| Age | 0.084 |
| Relative pay rates reviewed by: | |
| Gender | 0.056 |
| Ethnicity | 0.037 |
| Disability | 0.032 |
| Age | 0.051 |
| Family-friendly/flexible working practices in place for any employees at the workplace: | |
| Working at or from home in normal working hours | 0.256 |
| Job sharing schemes (sharing a full time job with another employee) | 0.245 |
| Flexitime (no set start or finish times but an agreement to work a set number of hours per week or per month) | 0.343 |
| Working only during school or term-time | 0.121 |
| Workplace nursery or nursery linked with workplace | 0.021 |
| Financial help with childcare (e.g. loans, repayable contributions to fees for childcare outside of the workplace, subsidized places not located at the establishment) | 0.056 |
| A specific period of leave for carers of older adults | 0.061 |
| Maternity leave at normal, full rate of pay | 0.556 |
| Control variables | |
| Workplace size: | |
| 5–9 employees | 0.426 |
| 10–24 employees | 0.338 |
| 25–49 employees | 0.123 |
| 50–99 employees | 0.063 |
| 100–199 employees | 0.030 |
| 200–499 employees | 0.016 |
| 500–999 employees | 0.003 |
| 1000+ employees | 0.002 |
| Organization size: | |
| 5–99 employees | 0.404 |
| 100–999 employees | 0.233 |
| 1000–9999 employees | 0.193 |
| 10,000+ employees | 0.169 |
| Standard industrial classification (SIC) major group: | |
| Manufacturing | 0.119 |
| Gas, electricity and water supply | 0.001 |
| Construction | 0.043 |
| Wholesale and retail trade | 0.234 |
| Hotels and restaurants | 0.093 |
| Transport, storage and communication | 0.046 |
| Financial intermediation | 0.053 |
| Real estate, renting | 0.162 |
| Public administration, defence, social security | 0.018 |
| Education | 0.035 |
| Health and social work | 0.119 |
| Other community, social, personal | 0.078 |
| Public sector | 0.104 |
| Single independent workplace | 0.381 |
| Financial performance: | |
| No competition/data | 0.080 |
| Above average financial performance | 0.474 |
| Average financial performance | 0.362 |
| Below average financial performance | 0.084 |
| National ownership: | |
| UK owned | 0.930 |
| North American owned | 0.034 |
| Non-UK EU owned | 0.023 |
| Rest of World | 0.013 |
| Workplace age: | |
| 0–4 years old | 0.102 |
| 5–9 years old | 0.152 |
| 10–19 years old | 0.244 |
| 20+ years old | 0.502 |
| Proportion of workforce: | |
| Female | 0.535 |
| Ethnic minority | 0.064 |
| Disabled | 0.009 |
| 50 years old or older | 0.217 |
N = 1702.
Acknowledgements
We would like to thank the WERS Sponsors – the Department of Trade and Industry, the Economic and Social Research Council, the Advisory, Conciliation and Arbitration Service, and the Policy Studies Institute – for allowing access to the data. They are not responsible for any of the findings or claims made in the article.
Funding
We would like to thank the Trades Union Congress for funding the analysis conducted here.