e-Recruitment, gender discrimination, and organizational results of listed companies on the Spanish Stock Exchange

Nowadays, the development of new technologies has turned the Internet into the prevalent recruitment source ( Pfieffelmann, Wagner, & Libkuman, 2010 ). For example in Spain, according to the Observatorio Nacional de las Telecomunicaciones y la Sociedad de la Información (ONTSI, 2015) [the Spanish Observatory for Telecommunications and the Information Society] during 2014, 75.8% of companies with Internet access had their own website and 21.1% of companies used it to advertize the jobs or receive work requests. This means an increase of 18.3% and 3.4% respectively from 2008, which is the first year with available data on this issue. This new form of recruitment, named e -recruitment, implies important changes with regard to the traditional recruitment. According to Cober, Brown, Keeping, and Levy (2004) , the differences are detectable when we refer to the experiences which become available with Internet web pages (e.g., images, sound, animations) as well as the possibility of providing a dynamic experience the job-seeker is involved in the recruitment process instead of being a passive informational receiver. Moreover, e -recruitment offers many practical benefits in comparison with traditional methods such as the use of companies’ own corporate websites as a source of recruitment ( Lievens & Harris, 2003 ), which serve to attract future workers through web content ( Cober, Brown, Blumenthal, & Levy, 2001 ) and appearance ( Thompson, Braddy, & Wuensch, 2008 ), and which can result in substantial savings in terms of financial costs and time invested ( Sylva & Mol, 2009). Although e -recruitment offers many possibilities, it has several limitations. To start with, there are more applicants which can prove detrimental to the quality of the e-recruitment process (Boehle, 2000 ), and some potential applicants may become excluded due to demographic differences in Internet access ( McManus & Ferguson, 2003).

During the recruitment process, organizations to obtain biographical data ( biodata ) from candidates. Biodata measures come from standardized methods of measuring past behaviors (e.g., education, job experience) that are relevant to future job performance ( Becton, Mathews, Hartley, & Whitaker, 2009 ). The use of biodata has certain advantages such as their low collection cost and acceptable predictive validity (between .30 and .40) across a wide range of criteria and situations ( Allworth & Hesketh, 2000 ). In addition, applicants do not usually fake their answers ( Schmitt & Kunce, 2002 ). However, biodata cannot be considered trouble-free: biodata are perceived as somewhat lacking in terms of their legality ( Furnham, 2008 ) and recent research has shown that it entails a much higher adverse impact than was previously supposed ( Bobko & Roth, 2013 ). Nowadays, the concerns related with legal compliance and discrimination in the use of biodata are actually very relevant issues for organizations. For instance, the inclusion of content suggesting a preference for a specific gender which is unrelated to job performance could be a first step towards discrimination ( Stone-Romero & Stone, 2005 ). Similarly, the inclusion of inappropriate or invasive content could promote the self-exclusion of potentially valuable applicants ( Truxillo et al., 2004 ) and generating negative reactions from applicants ( Bauer et al., 2006).

Following Guion (2011) , the information gathered during recruitment should be job-related in accordance with a prior job analysis. In consequence, any question unrelated with the job content could be considered as invasive and may be conducive to the generation of negative reactions and perceptions of unfairness in applicants ( Stone-Romero, Stone, & Hyatt, 2003 ). This is because it bestows upon decision makers the ability to make discriminatory choices ( Stone-Romero & Stone, 2005).

Given its negative consequences, one expects that companies will not invasive content in their ion processes. Surprisingly however, empirical evidence exists which indicates that public ( Wallace, Tye, & Vodanovich, 2000 ) and private ( García-Izquierdo et al., 2010 ) companies use potentially discriminatory items in e -recruitment. This could be explained by the Rational Bias Theory ( Larwood, Gutek, & Gattiker, 1984 ). Rational Bias Theory has shown that employees might discriminate if they believe that both their supervisors and the organizational climate promote discrimination (e.g., Larwood, Szwajkowski, & Rose, 1988; Powell & Butterfield, 1994; Szwajkowski & Larwood, 1991 ), and this even in the absence of their own prejudices ( Trentham & Larwood, 1998). According to Trentham and Larwood (1998) the Rational Bias Theory s two preconditions which serve to produce gender discrimination: the preference norm and the compliance instrumentality. On the one hand, the preference norm refers to the perception of decision makers that other significant parties (mainly supervisors, but also clients) expect gender-discriminative choices from them. On the other hand, compliance instrumentality suggests that decision makers feel that their own careers could be affected if they do not support the existing bias. Although the authorship of the recruitment content is usually unknown, there is a broad agreement that executives are the persons involved in their preparation ( Barr, Stimpert, & Huff, 1992 ). Consequently, and in accordance with the Rational Bias Theory explained above, the presence of potentially discriminatory content promotes discrimination. That is, if cues of discrimination are found in the e -recruitment sections it could signify that discrimination is expected by the managers of the organization according to the preference norm. In order to study in-depth what we could consider as discrimination, the legal framework must be taken into ac.

Since its inception, the European Union has protected equality in the workplace, developing extensive legislation to prevent discrimination between women and men ( García-Izquierdo & Ramos-Villagrasa, 2012 ). This equal-opportunity legislation commenced in the seventies with the enactment of some relevant Directives such as the Council Directives 75/117/EC and 76/207/EEC regarding equal pay, equal treatment in employment, and vocational training. At the ning of the present century there was a renewed attempt to deal with this matter with the Council Directive 2000/78/EC to prevent all types of discrimination (gender, age, etc.). Some years later, the Council Directive 76/207/EEC gave member states an active role in achieving the objective of equality between men and women for example when formulating and implementing laws, and the Council Directive 2004/113/EC implemented the principle of equal treatment between men and women in the access to and supply of goods and services. More recently a new Directive has been provided, the EU Directive 2006/54/EC, which contains what we could consider as discrimination.

According to the EU Directive 2006/54/EC, discrimination can be performed directly if one person is treated less favorably on grounds of gender, or indirectly if an apparently neutral provision, criterion, or practice would put persons of one gender at a particular disadvantage compared with persons of the opposite gender. It should be noted that indirect discrimination alone is only acceptable based on the three following assumptions: (a) for professional activities gender is a factor because of the nature or conditions of the activity to be performed, (b) to protect women, particularly in pregnancy and childbirth, and (c) to promote equal opportunities between men and women (art. 2.6 EU Directive 76/207 on Equal Treatment).

All in all, these Directives 2002/73/EC, 2004/113/EC and 2006/54/EC admonished member states to incorporate their content into ries’ respective legal systems between 2005 and 2008. Spain for example developed its own legislation in 2007 by implementing the Organic Act 3/2007, relating to the effective equality between women and men (LOIMH). In accordance with García-Izquierdo and García-Izquierdo (2007) , art.5 of the LOIMH establishes that information gathered during personnel ion must focus on data related with job performance and must seek to avoid unrelated and invasive information as well as unfair procedures or unequal decisions. However, one of the major concerns of the LOIMH is what one understands to be discriminatory content, at least in absence of jurisprudence. In line with legislation, discrimination is only acceptable when dealing with professional activities gender is a factor because of the nature or conditions of the activity to be performed. Consequently, we may consider that the inclusion in e -recruitment sections of any gender-related content which fails to justify a clear job-deion (role) is potentially a discriminatory practice.

Equally important, as a well-built legal framework, is the impact of the legal dispositions on real practices, given that the true drivers of change are the practitioners, not the policies themselves ( Yanow, 1999 ). This places the focus on the legal fulfillment. Research has shown that legal dispositions need some time to produce changes (e.g., Jain, Lawler, Bai, & Lee, 2010; Weichselbaumer & Winter-Ebmer, 2007 ). However, the time needed remains unclear. In fact, the number of articles addressing this problem is very scarce. Weichselbaumer and Winter-Ebmer (2007) performed a meta-analytic study regarding the gender-wage gap in 62 ries and found a small but positive effect during the first three years after the passing of the legislation. In a wider perspective, Jain et al. (2010) investigated a set of equal employment opportunity indicators in Canada during seven years, finding an upward trend over time. These studies that it take at least a year to see changes in the practices of organizations.

The present study focuses on searching for potentially discriminatory content against women in the websites of listed companies on the Spanish Stock Exchange. As we stated before, we developed a preliminary investigation regarding this question in the past ( García-Izquierdo et al., 2010 ). In that study, we found that companies did not change their practices according to law enforcement. However, we want to complete our approach to Spanish e -recruitment by: (1) focusing on gender equality, because Spanish legislation s specific guidelines in that sense; (2) developing and testing hypotheses based on the aforementioned arguments; (3) increasing our scope of the e -recruitment practices, including recruitment statements; and (4) analyzing the impact of the Equal Employment Opportunity (EEO) law fulfillment in terms of its impact on organizational results. Therefore, we reach a better perspective of the Spanish e-recruitment.

Since the LOIMH has been in force from 2007, it is easy to be expected that companies would follow legal dispositions, but empirical studies have shown discrimination cues after the law enforcement ( Alonso, Táuriz, & Choragwicka, 2009; García-Izquierdo et al., 2010; Mateos, Gimeno, & Escot, 2011 ). Therefore, here we offer an analysis of changes in the practices implemented by companies in Spain as a result of the introduction of LOIMH. If some improvement indicators are identified, then we can expect a progressive trend over time (e.g., Jain et al., 2010; Weichselbaumer & Winter-Ebmer, 2007 ), but if no changes appear to exist, then some kind of intervention is recommended (e.g., developing specific guidelines). Moreover, potentially discriminatory content could be detected when asking questions regarding gender, but also regarding marital status and children, as women are stereotyped for taking care of their family, especially when children are involved (e.g., Bobbit-Zeher, 2011; López-Zafra & García-Retamero, 2011; Sarrió, Barberá, Ramos, & Candela, 2002 ). In relation to the foregoing ideas, we believe that content regarding gender, marital status, or children could potentially be considered as discriminatory against women.

An important and closely related issue is the timespan of the study. The transposition of European Council directives by member states should have been finished between 2005 and 2008. Given that in 2005 the Spanish Government had not even published any draft legislation, we decided to commence our data gathering by aiming to capture “pre” enforcement practices in e-recruitment. Additionally, as Amor (2010) states, since 2005 Spain has moved from a period of economic growth to a recession and higher unemployment rates, the latter have been more pronounced for women. Job scarcity and concerns regarding unemployment have ed obtaining a job in one of the most desired objectives of the Spanish population, making equality considerations in personnel ion even more important. With respect to “post” condition measures, the LOIMH was finally enforced in 2007. Given that prior research has shown that legal dispositions between one and three years to produce an observable impact (e.g., Jain et al., 2010; Weichselbaumer & Winter-Ebmer, 2007 ), we decided to gather data two years before and after the LOIMH enforcement (i.e., 2005 and 2009).

As stated previously, nowadays discrimination is subtler than in the past ( Pager & Western, 2012 ). Thus, we indirect methods for investigating discrimination. In this study, we have relied on narrative policy analysis, which helps to conciliate the policy narrative and the er narrative (i.e., opposed position, in our case company practices) by comparison of both the former and the latter in order to generate a metanarrative acceptable to both sides ( Hampton, 2004 ). From our point of view, this metanarrative should demonstrate that compliance with legal dispositions can in fact produce better organizational results than discriminative practices.

Regarding organizational results, Huselid (1995) stated that Human Resources practices contributed to the profitability of organizations. Generally speaking, research shows a weak but existing relationship ( Wright & Gardner, 2003 ). This is because Human Resources practices are a somewhat remote indicator of results which are additionally affected by a pool of other different variables ( Paauwe, 2009 ). In recent years, some authors have explored the relationship between equality practices and organizational results (e.g., Armstrong et al., 2010; Bloom, Kretschmer, & Van Reenen, 2011; Lee & DeVoe, 2012 ). However, to the best of our knowledge nobody has actually d prior to this study the impact of e -recruitment discrimination or equality practices on organizational results, in Spain or else.

Based on our review of the literature in connection with the legal dispositions regarding equality at work, the e -recruitment in personnel ion, and their impact on organizational results, we propose the three following hypotheses. First of all, in accordance with the foregoing, if companies fulfill the LOIMH, one expects that they will only job-related content, otherwise, the inclusion of potentially discriminatory content could be seen as a preference norm (Rational Bias Theory; Trentham & Larwood, 1998 ), serving to promote discrimination to job entry ( Stone-Romero & Stone, 2005 ). Given that prior research in Spain has shown discrimination cues in the ion process ( Alonso et al., 2009; García-Izquierdo et al., 2010; Mateos et al., 2011 ), we expect to find potentially discriminatory content in e-recruitment.

H 1: Companies’ websites potentially discriminatory content in their e-recruitment sections.

Secondly, and according with this idea, our study mentioned above shows that at the category level there is no change over time in the application forms (e.g., the number of companies asking for gender remains stable). However, research has shown an improvement in equality practices after the enactment of the law (e.g., Jain et al., 2010; Weichselbaumer & Winter-Ebmer, 2007 ). Given that Spanish EEO law is focused on gender, we expect that some positive changes could be seen when gender-content is d as a whole.

H 2: There is significant improvement in e-recruitment practices regarding gender equality (i.e., a reduction of potentially discriminatory content) before the LOIMH and after its enforcement.

Lastly, given that previous studies have shown a positive relationship between equality practices and organizational results ( Armstrong et al., 2010, Bloom et al. , 2011; Lee & DeVoe, 2012 ), we expect to find that companies with better equality practices (i.e., less potentially discriminatory content in e -recruitment) tend to obtain better organizational results than other companies.

H 3: A negative association exists between the potentially discriminatory content of websites and the organizational outcomes of the companies.

Our first hypothesis proposed that companies’ websites d potentially discriminatory content (comments involving gender, marital status, number of children, and military service) in their recruitment sections. After analyzing recruitment statements ( n = 75), we found that some of them could be related with gender discrimination in a direct or indirect way. These categories were “women” ( n = 8, 10.64%) from the personnel set, “reconciliation of work and home-life” ( n = 3, 4.00%) from the activities or processes in the companies set, and “discrimination” ( n = 3, 4.00%) and “equal opportunities” ( n = 3, 4.00%) from the activities or processes related to the companies’ set. Although we find categories that could be related with gender discrimination, the context in which they were used does not necessarily signify a discriminatory intention. After that, we did not consider appropriate to the recruitment statements for 2009 (for further detail, see García-Izquierdo et al., 2010 ).

Talking about application forms, some companies have not been evaluated in both years (2005 and 2009) due to mergers and acquisitions between companies, changes in the e -recruitment sections (e.g., removal of application forms and substitution requesting that résumés should be sent by email), and technical problems of the websites. Taking all of these factors into ac, our sample size suffers shrinkage from 76 in 2005 to 66 in 2009 (86.84% of the original sample). With respect to results and focusing on the four potentially discriminatory items (gender, marital status, number of children, and military service), we can see in Table 5 that most companies maintain their policy as regards e-recruitment (n = 46, 69.70%). With reference to the rest, surprisingly, some companies have in fact increased the potentially discriminatory content ( n = 11, 16.67%) whilst, on the other hand, some companies have reduced their discriminatory content ( n = 9, 13.63%). It is worth highlighting that those companies which performed worse tend to ask for gender, which is precisely the item most directly related to gender discrimination. As we found cues of preference norm against women ( Trentham & Larwood, 1998 ) in application forms in 2005 and 2007, we consider H1 supported. Hypothesis 2

Improvement according with the Passage of the LOIMH