Gender and pay equity in a global knowledge organisation

Abstract

The gender pay gap, the difference in earnings of men and women, has remained remarkably resilient over the past two decades in Australia and a number of other countries, despite a range of legislative and policy initiatives aimed at narrowing this difference. Drawing upon industry-wide data and an organisational case, this analysis of professional business services examines the nature of pay equity within and beyond organisational boundaries. We conclude that ‘success’ in gender and pay equity terms has been constrained by minimum government compliance requirements and limited attention to the role labour markets play in contributing to gendered patterns of pay.

Introduction

Although there is considerable common ground on the meaning and goals of gender pay equity, there is no standard international measure to assess its achievement. Despite extensive research and policy attention over 30 years, the gender pay ratio remains remarkably resilient in Australia and internationally. This indicates the complexity of assessing and achieving gender pay equity and the associated challenges for those framing relevant policies. Despite the complexity of defining and measuring pay equity, however, the Australian Commonwealth Government’s Equal Opportunity for Women in the Workplace Agency (EOWA, renamed, from November 2012, the Workplace Gender Equality Agency (WGEA) or ‘the Agency’) deploys several programmes to encourage organisations to enhance gender equity and promote pay equity.

The purpose of this article is to explore gender pay equity policies at an organisational level within a professional business services firm (‘ProServ’, a pseudonym) with a view to gaining insights for future policy and research on pay equity. In doing so, this article addresses three questions. First, what do gendered patterns of pay ‘look like’ in the professional services industry and how do these compare with those observed in a business professional services firm that has been recognised for its achievements in this area? Second, what insights can be gained from current processes about the challenges that exist for achieving gender pay equity in business professional services occupations? Third, what role did practices and policies promoted by EOWA play in encouraging gender pay equity within organisations that can be defined as proactive in their gender equity policies?

In addressing these questions, we first consider definitions of gender pay equity and summarise recent literature and policy relevant to pay equity in Australia. Next, we outline EOWA’s criteria for recognition in gender and pay equity. We then discuss the data and methods used to gain insights into pay equity in the organisational context of ProServ. The first part of the analysis considers the relevant labour market context in which ProServ operates by identifying existing Australian and international information on gendered patterns of pay in professional business services. We then explore detailed data provided by ProServ, used to monitor and evaluate their progress towards organisational gender pay equity. We draw tentative conclusions about the constraints on pay equity initiatives and the role played by accepted social constructs such as the ‘neutral’ role of markets. We conclude by considering a more proactive role for government agencies in their monitoring and support for gender pay equity policies.

While there is no standard international measure of gender pay equity, there is a degree of consensus about what it means as a policy objective within the Australian context. In general terms, gender pay equity is recognition that work of equal value should receive equal remuneration. This is the broad approach used by the International Labour Organization (ILO) in Convention 100 Concerning Equal Remuneration for Men and Women Workers for Work of Equal Value (1951), which is mirrored in Australia’s Fair Work Act 2009 (Cth), Section 2.7. Commonly reported measures of gender pay equity include gender wage ratios (GWRs), in which women’s average earnings are reported as a percentage of men’s average earnings, or gender pay gaps, which focus on the ‘gap’ between women’s and men’s average earnings. In Australia, the ratio of average weekly ordinary time earnings for women and men employed full-time is commonly used for this purpose and has usually been around 82% or 83% over the past two decades (Australian Bureau of Statistics (ABS), 2011). Our approach does not provide a better or alternative approach to defining ‘equal pay for work of equal value’ but provides insights into how an organisation that has been recognised as achieving ‘good practice’ in gender equity defines and implements this objective in their particular context.

This article contributes insights on pay equity in a context that is under-represented in existing research. First, by focusing on professional business services, it analyses a group of employees in which the levels of education and training among both men and women are relatively high and more homogenous than in the Australian population generally. Second, it focuses on a relatively well-paid section of the Australian labour market in which men and women are represented in roughly equal proportions. This provides a contrast with recent studies and policy debates that have predominantly focused on either low-paid feminised sections of the labour market or high-status positions such as corporate board membership, where male employees are ‘over’-represented.

Recent research and policy on gender and pay equity

The principle of equal remuneration for work of equal value poses practical difficulties for both research purposes and policy implementation. This arises because men and women often undertake different roles within the workforce, making it difficult to directly compare their work and its value. One framework commonly used to guide economic and policy analyses of gendered patterns of pay in occupations and industries is based on a ‘human capital’ approach to explaining wage rates. Using this approach, different levels of average wages between different groups of employees, such as men and women, are investigated on the basis of the different characteristics of the employees and the roles they take at work. Typically, different characteristics between men and women, such as years of work experience, education levels, working hours, industry and occupation of employment and trade union membership, account for some portion of the observed gender wage gap. The rest of the gender wage gap remains ‘unexplained’ and can be attributed to unobserved variables and/or discrimination in the labour market. This approach, usually based on Oaxaca’s (1973) and Blinder’s (1973) decomposition analyses, underlies a diverse range of economic analyses of national gendered patterns of pay in Australia (see, e.g. Cobb-Clark and Barón, 2010; Cobb-Clark and Tan, 2011; Eastough and Miller, 2004).

While useful for some purposes, this approach, like all specific research methods, has shortcomings, now well documented in the literature (Grimshaw and Rubery, 2007; Olsen and Walby, 2004). The first limitation arises from the philosophy and assumptions that underpin standard economic approaches to understanding human capital. In particular, using individual earnings and characteristics of workers and firms as key explanatory variables neglects important aspects of the institutional and cultural contexts in which wages and labour supply are determined. This is a major constraint because social institutions critically affect how different types of work and products/services are valued by both the individuals involved in a labour market transaction and the consumers involved in its purchase. Many relevant factors that influence the determination of wage rates remain outside the factors considered in decomposition analyses. ¹

The second area includes matters that cannot be readily included in quantitative data analysis, such as the lived experience of engaging in particular forms of work or the motivation for choosing relatively low-paid occupations (Austen et al., 2008; Elton et al., 2007; Masterman-Smith and Pocock, 2008; Probert et al., 2002). A third area of shortcoming includes neglect of issues that could potentially be included in quantitative data analysis but are omitted, usually due to lack of relevant data (e.g. how geographic or spatial dimensions of work affect labour supply and wage outcomes – Jefferson et al., 2011).

Gender segregation, commonly understood as occupational and industry segregation, has been identified as a contributor to pay inequity. As a factor in the undervaluation of women’s work, Grimshaw and Rubery (2007) discern multiple dimensions of segregation: sectoral, occupational, workplace and work group. Undervaluation, they explain, may be related to segregation at each, or combinations, of these different levels. They report that ‘job level studies of gender composition have more power in explaining gender pay differences than studies that only focus on the occupation’ (Grimshaw and Rubery, 2007: 58) and that work group gender segregation is important because separate pay hierarchies tend to be used for different workforce groups. Undervaluation arising from work being part-time has been documented by Grimshaw and Rubery (2007) and quantified (for the United Kingdom) by Olsen and Walby (2004).

The challenges associated with applying the equal remuneration provision under the Fair Work Act 2009 were apparent in the 2010–2011 case initiated on behalf of Social and Community Service workers (Fair Work Australia (FWA), 2012). The FWA hearings and the limitations documented above highlight the significance of other approaches to pay equity, such as the case analysis deployed in this article.

Policy discussions on gender equity were also provoked when, in 2010, a review proposed significant changes to EOWA, one of the main gender/pay equality ‘watchdogs’, and its underlying legislation (The Equal Opportunity for Women in the Workplace Act 1999; hereafter called the EOWW Act). A new Workplace Gender Equality Act 2012 (hereafter, ‘WGEA Act’) has been implemented in response to the review with reforms intended to strengthen the focus of the Agency. The recast Act includes acknowledgement of ‘pay equity and the caring responsibilities of both women and men as central to gender equality’ (Department of Families Housing Community Services and Indigenous Affairs (FAHCSIA), 2011). Along with augmented capacity to assist industry, and more streamlined reporting, compliance mechanisms are now strengthened to ensure that organisational reviews are accurate and industry benchmarks are deployed. The new WGEA Act retains measures for ensuring that the government deals only with compliant organisations (WGEA, 2012a). This directly affects the reporting by, and potentially the performance of, organisations regarding gender pay equity. Significantly, the changes have potential to allow the Agency more responsiveness to changing employment circumstances, and greater use of quantitative measures, in pursuing the goal of gender equality and in including gender pay equity as part of this pursuit.

Issues of gender equity in workplace contexts have also been prominent in discussions about board membership of Australia’s corporate organisations (Broderick, 2009, 2010; EOWA, 2010; Shave, 2011). In addition, the Australian Stock Exchange (ASX, 2010) introduced a new gender reporting requirement.

The conclusion from the above research and policy discussions is that gendered patterns of pay, together with gender equity in labour markets more generally, remain an area of ongoing policy concern. The resilience of gendered remuneration patterns, often unfavourable to women, is reflected in reviews of the labour markets in Australia and internationally (Borland, 1999; Hallock, 1999; Hyman, 2004; Preston and Crockett, 1999; Short, 2002).

EOWA’s gender and pay equity programme

Over the past decade, EOWA maintained several programmes to encourage and acknowledge business achievement with regard to gender equality and pay equity. First, a process for assessing organisations’ compliance with the Act has recently been amended through the new WGEA Act. In the past, organisations covered by the EOWW Act needed to develop a workplace programme and report annually on that programme retrospectively. Employers were required to demonstrate that they had identified the issues for women and responded by assigning responsibility for the programme to a sufficiently senior person. Those that had been compliant for a period of 3 years could be granted a waiver from reporting requirements for a fixed period (usually 1 or 2 years) (EOWA, 2012b). Following a transition period, from 2013 to 2014, under the new Act, organisations will be required to demonstrate progress over time in attaining gender equality goals (EOWA, 2012a; WGEA, 2012a). Equal remuneration between women and men is one of six standardised ‘gender equality indicators’ (GEIs) established by the new Act and has been specifically added to the revised list of ‘employment matters’ available to the Minister to draw on when determining GEIs (WGEA, 2012b). The Minister, in consultation with the Agency, may set (and change) minimum standards (WGEA, 2012a). There is no reporting waiver provision, meaning that organisations must continue to report annually (WGEA, 2012b).

Under the EOWW Act, a second programme assessed organisations for the status of an ‘employer of choice for women’ (EOCFW). The EOCFW citation was ‘awarded to non-government organisations with more than 100 employees that demonstrated policies and practices supporting women across the organisation, achieving positive outcomes for both women and the business’ (EOWA, 2012b). Unlike the compliance assessment, application for EOCFW was voluntary. In recent years, just fewer than 100 organisations of more than 2500 applicants annually (less than 4%) have met the requirements for an EOCFW citation, a very low level (EOWA, 2011). Table 1 summarises the criteria for recognition under the EOCFW scheme, highlighting subclauses relevant to pay equity. A third programme was EOWA’s Business Achievement Awards, which were presented in ‘recognition of excellence in equal employment opportunity’ (EEO) under six different categories (such as Leading CEO/Diversity manager, Outstanding EEO Initiative) (EOWA, 2012b).

Table 1. EOWA EOCFW criteria, 2011.

Source: EOWA Employer of Choice for Women Application Kit 2011.

Italics: items emphasised in original source; boldfaced: pay equity components; EOCFW: employer of choice for women; EOWA: Equal Opportunity for Women in the Workplace Agency.

Detailed analysis of the practices of ProServ, guided by these EOWA programmes, may help identify gaps that need to be addressed in implementing the new gender equality provisions. Documents and analyses provided by ProServ for monitoring and evaluating gender equity show that their organisation has drawn heavily on the reporting requirements developed by the former EOWA and available through that Agency’s website. The documents showed that the organisation also undertakes additional analysis beyond that suggested by the Agency. One outcome is that ProServ has demonstrated success in gender equity and the associated aspect of pay equity. ² Thus, ProServ can be viewed as an organisation that actively engages with the Agency’s approach of encouraging organisations to demonstrate self-improvement in gender and pay equity through enhancement of processes and increased depth and detail of analysis.

Method

Relevant industry and occupational classifications for comparative purposes

ProServ provides a range of business services to both corporate and private clients and, along with numerous other firms, has employees working in the broad areas of taxation, finance, accounting and human resources. ³ It has a range of characteristics that are consistent with the Professional, Scientific and Technical Services industry, which is the appropriate industry grouping for comparison with ABS data. For simplicity, we refer to ProServ as being a ‘professional business services’ firm. This reflects both its high proportion of professionally qualified employees, most of whom have undertaken university degrees in the business area, and its formal categorisation within the Professional, Scientific and Technical Services industry. This designation also assists to distinguish ProServ from other types of quite different professional service organisations such as engineering consultants, medical professionals and scientists who work within very different contexts from ProServ’s business professionals.

Professional business services can be considered as a subset of the ‘Professional, Scientific and Technical Services’ industry classification defined under the ABS Australian and New Zealand Standard Industry Classification (ANZSIC). The ANZSIC codes define the industry in terms of both its purpose (of supply services) and the type of labour utilised (specialised expertise) (ABS, 2006b: 311).

The relatively close alignment between the definition of the ABS Professional Services Industry Division and the type of labour it employs means that approximately 59% of employees in the industry division are categorised under the occupational code of ‘Professionals’ (ABS, 2006a). This is the second highest alignment between industry and occupational divisions in the Australian workforce (the highest is 60%) (ABS, 2010b). In terms of relevant Australian data with which to compare ProServ, the occupational subdivision of ‘Business, Human Resources and Marketing Professionals’ is closely aligned. Thus, contextual and comparative information below has been collated with reference to the following industry and occupational classifications: (1) the industry division of Professional, Scientific and Technical Services, (2) the occupational division of Professional and (3) the occupational subdivision of Business, Human Resources and Marketing Professionals.

Data

For industry and occupational comparisons, this article uses, where possible, average hourly ordinary time earnings (AHOTE) for full-time employees (ABS, 2010a). This measure of pay equity is preferred over ‘weekly ordinary time’ and full-time earnings commonly used to calculate gender pay ratios because it more accurately takes into account differences in hours worked by men and women, ⁴ despite slightly less regular data collection by the ABS. It does, however, have the key disadvantage of excluding part-time workers from the average earnings estimates used in this discussion, which is an ongoing issue for studies of gender and pay using ABS data (Jefferson and Preston, 2010). Publicly available national average full-time earnings data for comparable industry and occupational classifications have been compiled using the ABS Employee Earnings and Hours Survey data from May 2010, a period approximating that relevant to this analysis. Organisational data provided by ProServ is generally expressed as annualised salaries precluding comparison of hourly rates. Expressing gender wage patterns as a percentage of women’s earnings compared with men’s, known as a GWR, assists the comparison of the two forms of data.

Advice from pay equity and industry experts confirms that professionals perceive an organisational-level analysis as both timely and important, confirming wider insights that workplace analysis can be more illuminating of the gender pay gap than traditional analysis based on personal attributes (Grimshaw and Rubery, 2007). However, organisations are sensitive about divulging information about employees’ pay rates, particularly the differences between men’s and women’s pay. Obtaining organisational-level data was challenging, and a variety of avenues were deployed to gain research access. Initial inquiries were made to a range of professional, industry and consulting bodies. ⁵ The Diversity Council of Australia (DCA) was instrumental in facilitating contacts on behalf of the researchers. Four organisations have expressed interest in participating in aspects of our longer term study.

With ProServ’s consent, the key organisational data sources for this article include documents covering gender and pay equity compiled in relation to the EOWA’s programmes, additional documents containing the outcomes of a diversity analysis conducted by ProServ to inform its policies and exploratory face-to-face discussions. From the data available, a period of 3 years, 2009–2011, was suitable for analysis. The data documents contain annualised average grade-level salaries disaggregated by gender and full-time/part-time status, and summarise organisational policies and programmes. ProServ documents demonstrate that gender pay equity at the firm level and within each of the organisation’s main sections and pay grade levels were analysed in 2011. This was for the purpose of identifying areas that warrant further investigation (ProServ, 2011a: 3). The analysis, undertaken by the firm’s remuneration/human resources services, also encompassed performance appraisal data, promotions, total revenue and bonus data at different pay grades. Some of ProServ’s very senior employees are included in the overall, firm-wide calculation of the GWR but were not included in their detailed analysis. This approach is consistent with EOWA-recommended processes on which much of ProServ’s analysis is based.

Analysis

Our research method is based on an exploration and comparison of national and organisational data. The national-level data provide both the labour market context in which ProServ operates and a point of comparison between gendered patterns of average earnings in relevant industries and occupations nationally. Organisational-level data also provide insights into the extent to which pay equity is considered as a key component of gender equity within an organisation and the way in which pay equity is both measured and monitored.

In summary, the approach is largely inductive and undertaken to broaden existing insights into gender pay equity. The organisational focus of this article also provides an opportunity to consider the links between traditional, human capital approaches to investigating gender and pay and the institutional frameworks in which earnings are determined.

Industry and occupational context: Gender and pay in professional business services

Using the industry and occupations defined above, this section provides an essential profile with which to compare the data on ProServ. This analysis yielded data on a difficult-to-isolate ‘professional business services’ industry sector and provides an example of the insights that might be gained from changes to the Agency’s assessment methods which, under the new Act, will include comparison with industry norms (EOWA, 2012a; FAHCSIA, 2011). Table 2 contains industry-level AHOTE data. In the Australian context, the 78.9% GWR of the Professional, Scientific and Technical Services industry is relatively low compared with most other industries. The industry had the second lowest GWR of all Australian industries and lies 11.1 percentage points below the Australian average of 90% (see Table 2).

Table 2. AHOTE (full time) by industry and sex and GWR, Australia, May 2010.

Source: ABS (2010a, Table 11).

AHOTE: average hourly ordinary time earnings; F: female; GWR: gender wage ratio; M: male.

Table 3 shows similar data for major occupational divisions in Australia. This shows that Professionals have a GWR of 84.4%, also well below the Australian average. At a general level, the data in Tables 2 and 3 suggest that ProServ operates within an industry and occupational context of comparatively low GWRs.

Table 3. AHOTE (full time) by occupation and sex and GWR, Australia, May 2010.

Source: ABS (2010a, Table 10).

AHOTE: average hourly ordinary time earnings; F: female; GWR: gender wage ratio; M: male.

Tables 4 and 5 provide a more detailed account of the labour market context in which ProServ operates. Table 4 shows that the GWR for the industry subdivision of Business, Human Resources and Marketing Professions is 81.4%, again well below average, but within the subdivision, there is, however, considerable variation from 94% in the case of accountants, auditors and company secretaries to a low of 67.2% among financial brokers, dealers and investment advisers. Thus, even within a relatively specific occupational subdivision with relatively similar educational requirements, there are considerable differences in the relevant GWR.

Table 4. AHOTE – selected occupational minor groups by sex, Australia, May 2010.

Source: ABS (2010a).

AHOTE: average hourly ordinary time earnings; F: female; GWR: gender wage ratio; HR: human resources; M: male.

Table 5. Selected occupational minor groups, employment by full time, part time, total and sex, Australia, May 2010.

Source: ABS (2010b).

F: female; HR: human resources; M: male.

Table 5 provides information on the gender composition of the workforce in Business, Human Resources and Marketing. The proportion of male and female employees is roughly equal, being 52.9% and 47.1%, respectively. Approximately 15.7% of employees work on a part-time basis and about two-thirds of part-timers are women. Nationally, approximately 30% of Australian employees work part-time, indicating that this occupational group has a relatively high percentage of full-time workers.

GWRs in professional business services in Australia appear broadly consistent with available international data. International comparisons of wages and GWRs are difficult to make due to both differences in data collection and relevant definitions. However, some indicative information for two broadly relevant industry classifications, ‘Financial Intermediation’ and ‘Business Activity’, is included in Table 6. This information suggests that the Australian GWR of 81.4% for Business Professionals is ‘mid range’ compared with gendered patterns of pay in the 11 countries included in Table 6.

Table 6. Indicative gender wage gaps in Financial Intermediation and Business Activity Industries, selected countries, by comparison of mean and median wages, 2006–2007.

Source: International Trade Union Confederation (2008: 39–45).

In summary, investigating gendered patterns of work and pay in professional services offers potential insights into both industry and occupational sections of the Australian labour market. Professional, Scientific and Technical Services have the second lowest GWR of Australian industry divisions and employ approximately 7.5% of all female employees. Women classified as having professional occupations represent 11% of Australian employees, and despite having relatively high levels of education, they have below-average GWRs (ABS, 2010b).

ProServ therefore operates within an Australian industry context of very low GWRs and an occupational context of a below-average GWR. Internationally, the GWRs in ProServ’s area of operations vary considerably and Australia’s ratio appears to be in the mid range of indicative wage ratios. A tentative conclusion is that the industry’s gender wage gap in Australia does not reflect an international pattern of gendered patterns of work in similar industries. This suggests that national and subnational factors are likely to be important in determining observed patterns of work and pay by men and women in professional business services.

Gender and pay equity at ProServ

Workplace profile – Employment

As in September 2011, the vast majority of employees (90%) at ProServ were ‘professional’, categorised on their attainment of graduate-level qualifications and/or the type of work in which they are engaged. The remaining 10% of staff were categorised as either administrative staff or not yet fully qualified professionally. Women comprised 99% of administrative employees, 48% of total employees and 40% of professionally qualified employees (compared with 47% nationally).

Reflecting the data provided by ProServ, our primary focus in this article is the ‘professional’ occupations, which, to preserve anonymity, we have labelled Professional Level (PL) 1, 2 and so on. Male and female employees are employed in equal numbers at the entry level (PL 1). The number of women and men remains fairly similar at each of the three successively higher levels (PLs 2–4), but the proportion of women drops to one-third of total employees at PL 5. Women make up 17 % of senior management. ⁶ Together, PLs 1–4 comprise 72% of the company’s workforce and the proportion of female and male employees in this composite workforce group are quite equally represented (73% and 72% of the respective female and male employee numbers) (refer Table 7). The average age of employees at ProServ is 32 years.

Table 7. ProServ workplace profile summary, 2009, 2010 and 2011.

Source: Compiled from data provided by ProServ (2009, 2010, 2011c).

Percentages of M and F employees do not add to 100% due to the exclusion of some employees from the above analysis, including some contractors, ‘inactive’ employees (on long-term leave) and senior managers.

FT: full time; PT: part time; M: male; F: female.

Part-time workers comprise 7.9% of ProServ’s workforce, compared with 15.7% nationally for similar occupations (Table 5), and 90% of ProServ’s part-timers are women, compared with 66% nationally. The grades at which part-time work is most prevalent at ProServ are PLs 3–5, which coincide with cohorts of women who are in the main child-rearing years. Although women comprise a smaller proportion of senior management (not included in Table 7), information contained in ProServ’s analysis indicates that 39% of such women work part-time. This reflects one of ProServ’s recent goals to enhance flexibility by improving the availability of part-time work.

Salaries and gender pay ratios

To maintain the anonymity of ProServ, we have expressed the average salary for each occupational level within ProServ as a percentage of average earnings for all ProServ employees (Table 7, columns 8 and 9). This establishes a profile of earnings by occupational level and gender, which is sufficient for the purposes of examining organisational pay equity. In general terms, however, ProServ’s salaries are neither particularly high nor low within its industry context. Organisational entrants (PL 1) receive a salary similar to the average salaries paid to recent graduates who enter the workforce within comparable professional occupations (Bryant and Carroll, 2011). In 2011, there was a slight positive pay ratio at this level (102%). Professional salaries rise quite steeply by about 20% – 40% from level to level. The pay gaps evident in the ratios at each of Levels 2–5 at 99%, 97%, 97% and 96%, respectively, were fairly small (Table 7).

In comparison with professional occupations, the pay ratio for administrative staff is 108%, meaning that in this occupational group, women have higher average earnings than men. This reflects the high proportion of administrative women who have diverse ranges of experience and skills compared with the very small number of men in relatively few roles such as mailroom workers. The overall gender pay ratio for ProServ was 82% (Table 5). This is rather lower than for the specific levels and arises in part from occupational segregation – the influence of the female salaries of administrative staff, the remaining preponderance of men at the most senior, highly paid levels and the concentration of men in some higher paying sections or work groups within the organisation. The gender pay ratio for professional-only Levels 1–5 was calculated at 85%.

Gendered patterns of pay at ProServ 2009–2011

The ProServ data for 2009–2011 depict some interesting, fairly consistent trends over the 3 years. The relative proportion of employees at different levels remained fairly consistent over the 3 years, while the gender balance of new entrants (Level 1) reached parity (from 46/54 female/male in 2009 to 50/50 in 2011). The gender pay gap for administrative staff, which favoured women, has progressively narrowed from 130% in 2009 to 108% in 2011. Pay equity at the entry Level 1 fell slightly from 102% (2009) to 99% (2010) returning to 102% (2011). The pay ratios at each of Levels 2–4 have remained quite even and fairly small, that is, between 96% and 98% over the 3 years. The pay gap at Level 5 narrowed slightly from 93% in 2009 to 95% in 2010 and 96% in 2011.

Despite relatively small and consistent pay ratios at each occupational level within ProServ, the gender pay gap for the company as a whole deteriorated slightly over the 3 years from 84% to 83% then 82%. If administrative staff are omitted from these calculations, then we find that the gender pay ratio for PLs 1–5 likewise deteriorated from 88% to 87% then 85% from 2009 to 2011. This is slightly better than the equivalent national average of 81% in comparable occupations (in 2010) but is substantially below the Australian average of 90% for all occupations (refer Table 2).

Sectional patterns of pay, performance and promotion at ProServ 2011

Table 8 provides a summary of salaries, performance ratings and promotions for PLs 1–5 in 2011. These data were prepared for ProServ’s internal organisational purposes and demonstrate an extension of the type of analysis outlined in EOWA’s suggested reporting for purposes such as EOCFW applications. ⁷ However, mirroring EOWA recommendations, ProServ’s own analysis specifically identified differences of more than 5 percentage points for performance ratings, GWRs and promotions within particular sections, as specific areas for further investigation and possibly policy consideration within the organisation. For example, ProServ identified differences of greater than 5 percentage points in the proportion of females and males at PL 5 who received a ‘high’ performance rating in Sections B and C. There are 27 instances of such differences in Table 8 with 15 of these instances favouring females and 12 favouring males.

Table 8. Summary of salary, performance and promotions by section and sex, ProServ (2011).

Source: Compiled from data provided by ProServ (Internal Document, 2011b).

F: female; M: male; PL: professional level. TR: total revenue

ProServ applied a similar approach to analysing promotions and identified five cases of more than 5 percentage points difference in men and women receiving promotions to the next level. In all five instances, the difference indicated a greater rate of promotion among males than females. Finally, ProServ analysed GWRs, using annualised total earnings (which includes bonuses) for full-time staff, part-time staff and all staff for each PL within each section. Using this method of analysis, differences of more than 5% were identified in 20 instances, with 14 of these favouring males over females. Of the six instances of the GWR that favoured women, four occurred at the graduate entrant level (PL 1).

ProServ’s observations on pay data

ProServ’s documents show that the organisation investigated and provided explanations for the identified patterns of gender, pay, performance and promotion. The size of the overall gender pay gap is ascribed by the company to three main factors: First, there is a concentration of women in administrative roles at lower salaries and a concentration of men at more senior levels. That is, gendered patterns of pay are partly attributed to vertical occupational segregation within the company. Second, women’s relatively high participation in part-time work slows the rate at which they gain experience within the organisation, and this has implications for their performance evaluations and associated pay rate. Third, reliance on market remuneration indicators to assist with remuneration policies means that disparities in salaries between organisational sections are applied to both men and women. ProServ explained that it

reviews its remuneration ranges using external market data to ensure remuneration ranges we provide are competitive and comparable in the market … [T]hese ranges are shared with the business for their input, and adjustments can be made according to the role based on function[al] specialisation, market value, grade and location. The same salary range is offered regardless of gender. (ProServ, 2011c: 40)

There are some key strategies being developed by ProServ that might have an effect on its gender pay ratio. These strategies are based on achieving greater gender balance across occupational levels and, to a somewhat lesser extent, between sections within the organisation. Two key components of this approach are evident in ProServ documents. First, an objective stated in 2009 was to increase the number of women at Levels 4 and 5. There is reference to an instance of a female employee shifting from an administrative role to a professional role, though it is understood that this is not common. The objective is supported by a range of training and development programmes (gender-specific mentoring and leadership programmes) plus training for both men and women on ‘unconscious bias’ and its role in managerial decisions, including decisions where gender might be a focus. Despite these strategies, however, by 2011, there had not been an increase in the proportion of women at these levels, though at Level 4 an increase in part-time female workers indicated increased take-up of flexible work arrangements that may, in the longer term, support retention of senior women.

Second, throughout the 3 years, ProServ monitored vertical occupational differences by gender, reporting that it had undertaken job evaluation and (increasingly detailed) pay equity analyses in 2010 and 2011. Following the 2011 annual review and promotion round, the CEO requested that a detailed, firm-wide pay equity audit be undertaken with results being conveyed not only to the CEO but also to all employees. The intention was that actions arising from this audit would be integrated into pay-setting policies.

ProServ’s analysis identified the effects of part-time work on pay outcomes and, despite relatively small numbers of part-time employees, has identified this as an area requiring further, detailed investigation. In particular, ProServ specifically noted EOWA’s advice that part-time work may have equity implications if it affects access to training and career development and contributes to negative perceptions about its effect on a part time employee’s work performance, attitudes and loyalty to the organisation.

Discussion

The persistent and gradually widening gender wage gap evident in ProServ’s documents and analysis over 2009–2011 represents something of a paradox. The organisation has put considerable effort into collection and analysis of pay data during this period. The 2011 information on ProServ’s sectional analysis of pay, performance and promotions demonstrates interest in, and a comprehensive commitment to, pay equity and the analysis of its possible causes. Despite these programmes and clear endorsements of gender equity programmes from ProServ’s CEO, gendered patterns of pay have changed little among professional staff, and the apparently favourable gender ratio for administrative staff has declined (from 130% to 108% over 2009–2011). The gendered patterns of employment and pay at ProServ, together with the organisation’s analysis and explanations, provide insights into both the ongoing persistence of gender pay gaps and the potential role for tools of analysis such as those suggested by EOWA.

A key characteristic of ProServ’s documents is that despite ongoing and detailed attention to gender and pay, ProServ does not perceive that it has any specific gender pay ‘issues’ or ‘concerns’. Within the guidelines and processes suggested by EOWA via its website, these conclusions appear justified. Specifically, ProServ perceives that the gender pay ratios within each occupational level are acceptable. There appear to be two key reasons for this perception. First, ProServ’s pay gaps at each occupational level might be interpreted as being well within the guidelines and advice provided by EOWA for conducting a pay audit. EOWA comments in a footnote in its pay audit tool that ‘there is no legal guidance on what constitutes a significant difference but techniques of statistical analysis suggest that a 5% or greater difference can be regarded as significant, wherever it occurs’ (EOWA, n.d.: 5) ⁸ . In ProServ’s analysis, GWRs for each occupational level are at, or above, 95% and so are interpreted as being within the 5% gap that is deemed to not warrant further investigation (ProServ, 2010: 26–27, 2011c: 40–41).

EOWA’s approach has been that organisations should work towards improving their internal processes and depth of pay and pay equity analysis. From ProServ’s internal documents and depiction of its processes, it is evident that this has been occurring. This approach has been implemented by ProServ, and it has received recognition for its achievements in the area of gender equity. In the context of EOWA guidelines to date, there appears to be limited motivation or rationale for further investigating or initiating additional gender equity measures.

Nevertheless, we identified several reasons for the enduring pay inequities and potential areas of policy to which these relate. The first is the need to examine relative rates of change in remuneration at different occupational levels. For example, it is likely that the relatively slower growth in the remuneration of administrative staff has contributed to the declining GWR. Over the 3 years, average female administrative staff salaries have declined from 67.6% (as a percentage of average ProServ remuneration) in 2009 to 63.5% in 2011. In relative terms, this decline is larger than that experienced for professional occupations where more male employees are employed.

ProServ’s documents focus on positive areas of occupational performance and on building upon achievements. In this context, it might be questioned whether this approach provides an impetus for critical evaluation of data that show all indications of meeting EOWA recommendations. There is insufficient information in ProServ’s documentation to provide clear evidence of this possibility, but it is to be expected that critical information will be lacking if the organisation focuses mainly or exclusively on ‘positive’ stories. Some data suggest differences in the training and development provided for men and women. Women tend to be concentrated in gender-specific programmes involving mentoring and building women’s leadership capacity. However, they remain under-represented in executive-level programmes being undertaken by men either as a result of, or as a prelude to, promotion. While the various training and development programmes have, without doubt, been developed with intentions to promote equity, there would appear to be scope for a more critical investigation of why this is not occurring. One approach might be to survey employees about which sections they wish/do not wish to work in and what contributes to the perception of some sections as ‘desirable’ for male or female employees. Important differences in perceptions of male and female employees is demonstrated in a finance industry ⁹ survey, which showed marked disparities by gender in the perceived existence and effectiveness of policies and programmes aimed at improving gender equity in that sector (Financial Services Institute of Australasia (Finsia), 2010).

The sectional-level analysis undertaken for PLs 1–5 provides some further insights into the apparent paradox of ProServ’s GWR. There are at least five contributing factors that remain relatively neglected in this disaggregation. First, men and women appear to have relatively different rates of success in converting good performance ratings into promotion. The clearest example of this is within PL 4 in Section A. In 2011, 76.5% of females and 79.1% of males were rated as having either ‘excellent’ or ‘very good’ performance. Despite this, just 1% of females compared with 10.9% of males achieved promotion to PL 5. We calculated the percentage of females and males achieving ‘excellent’ performance ratings expressed as a ratio of the percentage promoted (Table 9). A ratio of 1.0 indicates that males and females are promoted in the same proportions that they achieve ‘excellent’ performance ratings. A ratio below 1.0 indicates that females with ‘excellent’ performance ratings are less likely to achieve promotion than males with ‘excellent’ performance ratings. In other words, exemplary performance by women does not get translated, to the extent that would seem justified by their performance, into promotion and access to higher salaries. A ratio above 1.0 indicates a situation where a disproportionately large number of females are being promoted compared with those who receive ‘excellent’ performance ratings. In some instances, the promotion rates favour females; however, these tend to be within lower occupational classifications. Excluding the ratios of 0.99 and 0.97 that are very close to 1.0, 11 ratios (>1.0) favour female employees compared with 17 (<1.0) that favour males. Research has revealed a variety of potential impacts of promotional systems on the achievement of pay equity including the need for women to prove stability, whereas for men to demonstrate (only) performance; significantly higher post-promotion wage growth for men than women and women receiving fewer career promotion benefits from training (Grimshaw and Rubery, 2007: 69).

Table 9. Proportion of female employees with ‘excellent’ rating as a ratio of females promoted (%).

Source: Compiled from data provided by ProServ (Internal Document, 2011b).

a No females were promoted and so a ratio is not possible for Level 5 in Section C.

Second, there are relatively large disparities in pay for the same occupational level across, and within, sections. Section A has higher average earnings at all occupational levels. This section also has the highest male-to-female employment ratio, which increases ProServ’s organisational pay gap. It is understood that there may be similar work group concentrations within sections. We are unable to assess whether changes in relative earnings or promotions have contributed to ProServ’s growing gender wage gap in recent years. However, either of these factors, or a combination of them, is a possible cause of a declining GWR and warrants investigation. The potential importance of this is highlighted by a UK study, which showed that work group segregation of men and women explained around one-quarter of the pay gap among full-timers and over one-tenth of the gap between female part-timers and male full-timers (Anderson et al., 2001, cited in Grimshaw and Rubery, 2007).

Third, the suggestion that a 5% difference in male and female earnings is a significant difference, a parameter consistent with the EOWA, appears to have been interpreted as differences of 5 percentage points being a key benchmark throughout ProServ’s sectional analysis. For example, ProServ’s analysis of promotions identifies all cases where there is a difference in excess of 5 percentage points between male and female rates of promotion. This, however, neglects the fact that 5 percentage points might be a much higher relative difference than 5%. If, for example, there is a work area employing 100 women and 100 men, the promotion of 5% of women and 10% of men means that 5 women and 10 men are promoted. That is, the promotion rate for men is 100% higher than for women, even though the difference is 5 percentage points.

There is no suggestion that ProServ’s use of a 5% difference is in any way a misinterpretation of the guidance provided via the EOWA website. Indeed, it might be interpreted as a straightforward application of the guidelines provided by EOWA. However, it is possible that a more appropriate form of analysis would be to compare directly the performance ratings of men and women and the ratios in which they are promoted, with a view to identifying differences and the possible reasons these occur.

Fourth, while ProServ has identified the position of part-time employees as an area warranting further investigation, there is little specific indication of how this will be done. It is known from research that undervaluation of part-time jobs is relatively common (Grimshaw and Rubery, 2007) and that while years spent working full-time are associated with increased wages, working part-time is not – not even prorata (Olsen and Walby, 2004). It is possible that ProServ’s part-time employees are acquiring valuable skills and experience that can contribute to their productivity, even if this is occurring outside the workplace. Current analysis appears to assume a close nexus between workplace experience and productivity and performance but warrants further investigation. For example, knowledge of changing policy or regulatory requirements is not proportional to the number of hours spent ‘on the job’ each week. A critical approach to examining links between part-time work and performance may seek to uncover possible assumptions in this area. A relatively uncritical influence of ‘billable hours’ and client facing time might underpin policy in this area (Ladva, 2010). Given ProServ’s success in providing time flexibility at very senior levels, a methodical review of pay, performance and part-time status is vital. It is important to ensure that various influences on pay are sufficiently disentangled. For example, despite increasing discussion of the importance of ‘soft skills’ (e.g. communication skills and emotional labour), pay and grading structures are commonly based on male-type skills with ‘soft skills’ not always being valued sufficiently, particularly in managerial and higher level jobs (Olsen and Walby, 2004: 60).

Finally, we note that ProServ, like many organisations, uses market remuneration surveys to assist with salary setting. In this way, the industry’s gendered patterns of pay may be conveyed into the organisation unnoticed. While one firm within an industry has limited capacity to affect pay structures, ProServ operates within an industry sector where new remuneration initiatives by particular organisations are likely to be influential. The data that companies derive from market remuneration surveys cannot be considered entirely independent of the market wage rates they report. A key theme of some submissions to FWA’s equal remuneration case was that markets are not ‘gender neutral’ in their operation (Austen, 2010; Junor, 2010; Meagher, 2010; Smith, 2010). In tshe United Kingdom, Adair Turner (2009), Chair of the Financial Services Authority, has noted that high market returns and associated earnings ‘can just as easily reflect market imperfections rather than be proof of social value’ (p. 5). Relying on market-based remuneration data can be highly problematic in terms of addressing gender wage differences as noted elsewhere:

Once wage scales are set up, the disparities are perpetuated by organizational inertia in the form of using past wages within the organization to set present wages or the use of market surveys of wages in other firms to set jobs’ pay levels. That is, wage scales get ‘institutionalized’. (England, 2005, cited in Grimshaw and Rubery, 2007: 60)

It is arguable that such structures could be altered through a proactive stance by large employers in this industry to actively establish equal pay for work of equal or comparable value within their own workplaces. Prior research on gendered patterns of remuneration has found that ‘employers often confuse previous salary with some notion of market forces … men are more likely to seek, or receive, outside offers to boost their internal pay … [W]omen are disadvantaged because of lower ability, or reduced willingness, to engage in individual bargaining’ (Grimshaw and Rubery, 2007: 67). Efforts are needed to ensure equity between salaries offered to new entrants (at all levels) and existing employees.

Conclusion

This article has provided an arm’s-length evaluation of gender pay equity within a large professional services firm via an external appraisal of the organisation’s own gender equity and pay analysis. This provided a springboard for considering the firm’s performance in this area, the underlying criteria of the government agency (EOWA) that have been used by organisations to shape their analysis of organisational gender equity and the influence of perceptions and assumptions of a ‘gender neutral’ labour market. The areas identified as contributing to the organisation’s ongoing pay gap within a context of continued monitoring and evaluation have wider resonance for organisations within and beyond professional business services.

This exploration of the data provided by ProServ shows that an employer can be proactive and conscientious on matters of gender and pay equity while simultaneously having highly resilient gendered patterns of pay. These patterns are strongly related to vertical and sectional occupational segregation and different patterns of career progression for men and women. Neither of these contributory causes to gender pay gaps is readily addressed by an individual organisation, and there is little guidance provided in current EOWA (now WGEA) guidelines for critically examining these factors. ProServ’s assessments demonstrate that it is very possible for an employer to extensively analyse and report on a wide variety of gender and pay equity measures and achieve well against the established criteria, yet have consistent, if not declining gender pay ratios. In an environment where remuneration and career progression are viewed as extensions of a ‘gender neutral’ market, measures to address gender pay gaps are constrained to areas that seek to have women’s patterns of employment more closely mirror those of men. Maintaining the relative earnings of feminised sectors of the organisation means that even this relatively constrained approach to pay equity has had limited success.

For organisations like ProServ, this raises the dilemma of apparent achievement but against criteria that may not be sufficiently robust to enable organisations to attain enhanced pay equity performance. Since ‘gender differentiation is a force that not only persists, but continually re-emerges’ (Grimshaw and Rubery, 2007: 131), such mechanisms need to create ongoing incentive for organisations to detect and overcome the many and varied barriers to gender pay equity. The research reported in this article suggests the benefits of making available detailed data to facilitate analysis that may further contribute to the achievement of pay equity in Australia.