Although talent management is an integral part of organisational success, the concept itself lacks theoretical frameworks. Even though Collings and Mellahi (2009) are of the opinion that there is a lack of consistent definitions, they propose that talent management is seen as activities and practices that identify critical positions and develop high performers as well as establish ‘differentiated human resource architecture to facilitate filling of these positions’. Typically, there are three mainstream interpretations of talent management. Talent management may firstly be described as a new term for common human resource practices; secondly, as an allusion of succession planning practices; and lastly, as referring to the general management of talented employees (Al Ariss, Cascio, & Paauwe, 2014).

There is no one-size-fits-all solution to talent management. Generalisations made in terms of talent management strategies are complicated by diversity considerations, including the range of generations employed. Talent management strategies must be adjusted to accommodate the different values, characteristics, attitudes towards work and general world view of the different generational cohorts working together (Bluen, 2013; Festing & Schäfer, 2014; Knowledge Resources, 2017b). Given the broad definitions of talent management, our approach is to view talent management strategies to include HR practices and policies pertaining to the work environment, financial security and growth and development. The operationalisation of these variables will follow.

An organisation is only as good as its people, and talent management should thus form the foundation on which organisations are built (Grobler et al., 2017). Figure 1 presents a conceptual model for this study. The model highlights talent management strategies relating to the working environment, financial stability, and growth and development as the independent variable, and retention as the dependent variable. Diversity considerations, including generational cohorts, gender and ethnicity, are presented as variables which were controlled for in this study, as they may or may not have affected the strength of the relationship between talent management and retention. The model further suggests that both organisational considerations (talent management strategies) and individual considerations (diversity issues) play a role in determining the nature of the psychological contract, which in itself may influence the dependent variable of retention.

The work environment is vital to employee satisfaction. For the purpose of this study, work environment includes the conditions wherein an employee operates, their roles and responsibilities, whether sufficient authority is given to employees to make and implement decisions about their work and the nature of working relationships. One such working relationship is that between management and employees. This relationship needs to be one of trust, mutual respect and belief in employee capabilities. Changing ways of working necessitated by the knowledge-intensive nature of work as well as the continuously developing possibilities provided by information and communication technology have also inevitably changed the working environment. Several authors such as Bosch-Sijtsema, Fruchter, Vartiainen and Ruohomäki (2011); Gorgievski, Bakker and Schaufeli (2010); Peters, Poutsma, Van der Heijden, Bakker and Bruijn (2014) and Van der Voordt (2004) have explored these changing working practices. Flexible and mobile work practices have increased in popularity as they provide employees with more autonomy in deciding when, where and how their work gets done. Flexibility in work practices is especially important for females and ultimately leads to employee retention (Du Plessis & Barkhuizen, 2015). There are also the preferences that the different generational cohorts bring to the work environment. Gursoy, Maier and Chi (2008) found that Baby Boomers have a high respect for authority and hierarchy in the workplace. Conversely, Generation Xers are seen as being more self-reliant, independent and sceptical of authority (Crumpacker & Crumpacker, 2007). Millennials prefer more flexibility and time off work that will suit their lifestyle (Roy, 2008). These characteristics will, in turn, affect the preferred work environment of the different generations.

Financial security as a variable is complex and comprises of financial and non-financial rewards (Bussin & Van Rooy, 2014). Financial security is the individual’s peace of mind that their income is sufficient to cover their expenses. Research in the South African context with regard to financial security and generational cohorts is inconclusive as the different generations place value on different aspects of financial security (Bussin & Van Rooy, 2014; Masibigiri & Nienaber, 2011; Moore & Bussin, 2012). For the purpose of this study, factors contributing to financial security include work–life balance, recognition and achievements which, in turn, lead to job security.

The characteristics of a generation do not apply unilaterally to each member of a generational cohort as each individual is unique. Consideration must be given to the changing needs of employees which include freedom and flexibility in their job, opportunities to grow personally and professionally, as well as to experience fun at work. In this study, gender as a variable was controlled for as Clarke (2011) argues that development programmes should be specific to gender as this allows for discussion of challenges pertaining to that gender. D’Amato and Herzfeldt (2008) found that employees who have a higher organisational commitment and greater orientation are more likely to stay for the next 3 years. This finding highlights the importance of providing learning opportunities within the organisation to retain talent across generations.

In his seminal work, Rousseau (1990) defines a psychological contract as the individual’s beliefs in terms of reciprocal obligations between employees and employers. The contract emerges when a contribution is made, and thus, an obligation is created to provide future benefits or fulfilment of needs. This expected reciprocal obligation has an impact on employee performance beyond the levels expected in their formal contract (Ye, Cardon, & Rivera, 2012). Differences in the interpretation of what constitutes a psychological contract between parties may be a challenge because of the contract being of an unwritten nature and may cause problems within the organisation (Ye et al., 2012). The employee evaluates these individual perceptions by comparing organisational experiences such as talent management practices to their own psychological contracts (Festing & Schäfer, 2014).

When employers live up to the expectations of employees in terms of the psychological contract, employees are more likely to experience job satisfaction, demonstrate organisational commitment and eagerness to develop and learn within the organisation, and display willingness to contribute to organisational goals (Festing & Schäfer, 2014). However, if they fail to meet employee expectations, they may be met with negative reactions such as reduced loyalty, reduced job performance and commitment, and the intention to quit (Festing & Schäfer, 2014).

Crampton and Hodge (2007) postulate that generations develop different characteristics and work-related values and attitudes based on shared experiences in the formative years which contribute to shaping their lives. Research by Schambach (2001) suggests that the psychological contract may be perceived stronger by younger employees, as they feel the obligation of the reciprocal relationship more so than older employees, who have paid their dues to the organisation. A study on basic human value differences between generational cohorts indicated differences between Millennials and Generation Xers on ‘openness to change’ and ‘self-enhancement’ (Lyons, Duxbury, & Higgins, 2007). Younger generations, early in their careers, may be prone to accept higher levels of risk (Johnson & Lopes, 2008). The requirements of the psychological contract differ between individuals (Farndale, Pai, Sparrow, & Scullion, 2014). This will result in different generations having varying needs and expectations in terms of the psychological contract (D’Amato & Herzfeldt, 2008; Festing & Schäfer, 2014).

The above discussion leads to the formulation of the following hypothesis:

Research results claim that 65% of black employees have changed jobs at least once within three years (Zulu & Parumasur, 2009). Therefore, it is critical that organisations explore strategies aimed at enhancing employee retention. These include maintaining a work-life balance, and ways to increase job satisfaction and organisational commitment (Ilies, Wilson, & Wagner, 2009; Qu & Zhao, 2012). These strategies, however, need to be flexible when dealing with different generational groups. Baby Boomers seek power, influence and authority over others, seek out status symbols, seniority and experience, and value opportunities to mentor others (Close, 2015). Generation Xers value control over their lives and place an immense value on work-life balance (Close, 2015). Millennials desire career progression and advancement, and want opportunities for on-the-job education (Close, 2015). Millennials are attracted to work that is challenging, meaningful and flexible and offers career mobility (Close, 2015).

The above leads to the formulation of the following hypothesis:

Generational differences may also influence the preferred methods of communication within the organisation. Baby Boomers prefer to communicate face to face during meetings rather than over the phone (Simons, 2009). Wagner (2007), however, found that Generation X and Millennials generally prefer email and text messages. This may cause intergenerational issues as using communications methods that are unfamiliar or undesirable to one group may cause discomfort and conflict.

Baby Boomers are seen as being willing to sacrifice their personal and professional lives to achieve success and financial security (Glass, 2007), whereas Generation Xers focus on balancing their personal and professional lives, resulting in the reputation of being the generation that ‘work to live not live to work’ (Simons, 2009). Atkinson (2008) found that although Millennials see money as important, they are more motivated to maintain a good work-life balance. Millennials want strong workplace relationships, a sense of purpose and a feeling that they are making a difference. Not meeting these needs may decrease job satisfaction which may negatively impact on job security, resulting in the departure of highly qualified personnel, and the costs associated with employee turnover and weaker growth prospects (Cheese, Thomas, & Craig, 2008 as quoted in Kilfman, 2009).

Baby Boomers tend to be more established in their careers than Millennials and will thus not place the same value on opportunities for learning and development (Close, 2015). This is largely because of Millennials generally being in the early stages of their career and desiring on-the-job training (Carlson, 2015). Even though it is generally accepted that Baby Boomers may have more knowledge and experience within the organisation, it will still be expected that they should be challenged to develop new skills and refresh old ones.

Three main generations are identified in today’s workforce, namely Baby Boomers (born between 1945 and 1965), Generation X (1966–1977) and Millennials (1978–1990) (Bussin & Van Rooy, 2014). Although the exact dates that must be met to fall into one of these categories vary and some overlap exists, these generational cohorts show distinct characteristics which play a role in terms of how they perform in the workforce.

Cohort theory argues that growing up around the same time and experiencing events such as natural disasters, politics, economic conditions and pop culture at about the same point in their development lead individuals within a particular generational cohort to develop similar values, attitudes, opinions, employment expectations and life experiences (CIPD, 2009; D’Amato & Herzfeldt, 2008; Festing & Schäfer, 2014; Leuty & Hansen, 2011; Macky, Gardner, & Forsyth, 2008; Mayhew, 2014).

The above leads to the formulation of the following hypothesis:

South Africa has a wide range of cultural and racial diversity, as well as a history of discrimination against individuals on this basis. Issues around marginalisation and exclusion are still prominent in the South African context. Typically, South African organisations comprise of a multitude of cultures and must therefore develop talent management strategies that take all of these cultural differences into account. This is confirmed by The Economist Intelligence Unit (2010) report and Knowledge Resources (2017b), which highlight some of these challenges such as cultural conflict between employees (50%), problems understanding the local culture (47%) and variations in working styles and norms of behaviour in the office (42%). Statistics of South Africa reveals that the blacks comprise of 75% of the South African total employed population (Knowledge Resources, 2018). The above leads to the question as to how the nature of the psychological contract is perceived by different ethnic groups.

The above leads to the formulation of the following hypothesis:

A structured, closed-ended, five-point Likert-type questionnaire comprising 33 items was used to collect the data. The independent variable, talent management (TM) strategies, varies vastly among organisations based on size, type of industry and geographical location (Lewis & Heckman, 2006). In order to derive a standardised measure of TM, employees’ perceptions of TM strategies were assessed based on the motivational aspects derived from the job characteristics approach first developed by Campion in 1988. Twenty items, derived from the motivational scale developed by Campion and McClelland (1991) were clustered into three subsections comprising work environment (WE), financial security (FS), and growth and development (GD). Coefficient alpha values for this study ranged from 0.793 (WE), 0.807 (FS) and 0.844 (GD) and compared favourably to Campion and McClelland’s (1991) alpha values ranging from 0.81 to 0.94 (see Table 3).

Meyer and Allen’s (1997) scale measuring employees’ emotional attachment to, identification with, and involvement in the organisation served as proxy for measuring employees’ psychological contract (PC). The moderating variable, psychological contract, comprised eight statements. The coefficient alpha for this study was 0.847 (PC) and it was compared favourably to the study of Meyer and Allen’s (1997) validated scale ranging from 0.77 to 0.88.

Retention, the dependent variable, was measured using Lee, Mitchell, Sablynski, Burton and Holtom’s (2004) five-item intention to quit scale. The coefficient alpha for this scale was 0.859 (see Table 2). Literature indicates that behavioural intention is generally a good predictor of future behaviour (Armitage & Connor, 2001). Furthermore, Griffeth, Hom and Gaertner (2000) have shown that turnover intention is a strong predictor of actual turnover. In addition, generational cohort, gender and ethnicity were controlled for in this study in order to determine their impact on the psychological contract and retention.

The target population of this study was employees across the various generational cohorts in South Africa within the private and/or public sector and at all organisational levels including managerial, non-managerial and support and/or administrative employees, and comprising different gender and ethnic groupings. For the purpose of this study, the generational cohorts referred to include Baby Boomers, Generation X and Millennials as represented by the age categories presented in the discussion (refer to Table 1).

The respondents were identified by means of the snowball convenience sampling technique. The questionnaire link was emailed to potential respondents, and a total of 711 useable questionnaires were returned.

Structural equation modelling (SEM) was used to analyse the conceptual framework presented in Figure 2. The variables referred to in this study, work environment (WE), financial security (FS), growth and development (GD), psychological contract (PC) and retention (R), have been theoretically explored and operationalised using validated questionnaires. These variables were confirmed using exploratory factor analysis (EFA). Prior to the analyses being performed, the appropriateness of the data was assessed using Bartlett’s Test of Sphericity and the Keiser–Meyer–Olkin measures of sampling adequacy (MSA), and ensuring an adequate sample size. Once deemed adequate, the analysis was performed. The SEM analysis was performed using a two-tiered model-building approach. The confirmatory factor models were analysed to ensure suitable model fit prior to testing the structural models (Hair, Anderson, Tatham, & Black, 1998). Each model was assessed using several goodness-of-fit indices. Once an appropriate model fit was obtained, the path estimates were interpreted and conclusions drawn (Byrne, 2010).

The data were captured, cleaned and the biographical information analysed (Table 1). Of the 711 responses obtained, 44.8% were female and 55.2% male. The respondents were fairly evenly spread among the three generational age profiles: Baby Boomers (21.8%), Generation X (39.5%) and Millennials (38.7%). The African ethnic group had the greatest frequency at 40.5%. This was followed by the white ethnic group at 30.4% and the mixed race and Asian ethnic group at 29.1%.

The Bartlett’s Test of Sphericity and Kaiser–Meyer–Olkin MSA were calculated. The MSA value calculated was 0.915, significantly higher than the minimum required value of 0.50 (Hair, Black, Babin, & Anderson, 2014). The Bartlett’s Test of Sphericity was found to be significant. This along with a sample size of 711 deemed the data appropriate for EFA.

The EFA was performed and the results were consistent with the theoretical variables. Within this study, the measurement model comprises of 33 items that explain the five constructs (WE, FS, GD, PC and R). Factor loadings of 0.45 were considered significant, which falls above the recommendation of 0.40 for sample sizes greater than 350 respondents (Hair, Black, Babin, & Anderson, 2006). Several items did not load adequately and were therefore removed from the analysis. The EFA results provided evidence of the construct and discriminant validity of the measurement instrument. The internal reliability was assessed using Cronbach alphas and factor loadings, as seen in Table 2. A value of 0.70 or higher is considered reliable (Nunnally & Bernstein, 1994).

As each construct had a Cronbach’s alpha value of greater than 0.70, they were all deemed reliable. These constructs were then used in the SEM analysis. Each confirmatory factor model was analysed and assessed. The structural model based on the conceptual model in Figure 1 was analysed. The variables gender, generational cohort and ethnicity were then added to the structural model. The constructs GD and FS were removed from the analysis as the relationships between these constructs and PC were insignificant (Byrne, 2010). The resulting model can be found in Figure 2.

The model fit was assessed for the model in Figure 2 and indicates an average fit upon review of the indices. The CMIN/df value, also called chi-square to df ratio, indicating an adequate model fit, of this sample size, is a value less than 5. The CMIN/df from the model indicates an appropriate model fit with a value of 4.935. The root mean square error of approximation (RMSEA) value of 0.074 indicates a good model fit falling below the recommended value of 0.08. The goodness-of-fit index (GFI) and comparative fit index (CFI) values were 0.863 and 0.841, respectively, which are both close to the recommended value of 0.90 (Iacobucci, 2009). Therefore, the overall model fit indicates an average fit.

The unstandardised regression weights were then reviewed (Table 3). From the results, the relationship between WE and PC was significant at a 10% level, indicating a positive relationship. The relationship between gender and PC was significant at a 5% level, representing a negative relationship. The relationships of the mixed race ethnic grouping with PC, and Baby Boomers with PC, were found to be insignificant. The relationship between the white ethnic grouping and PC indicated a significant positive relationship, whereas the relationship between Millennials and PC indicated a significant negative relationship. This indicates that the white ethnic grouping tends to favour a higher level of PC, whereas the Millennials respondents did not favour a higher level of PC. The relationship between PC and retention indicated a significant positive relationship. This indicates that as the level of PC increases this leads to an increase in the level of retention.

From the results, it can be concluded that WE, white ethnic grouping and retention had a significant positive relationship with PC, whereas gender and Millennials had a significant negative relationship with PC.

A limitation of the research was that the sample size was relatively small and that a larger sample would have allowed the authors to make more meaningful inferences on the data obtained.

The provision of development opportunities is likely to increase the organisational commitment of younger generations and to attract talented employees. Future research could investigate the developmental opportunities and corresponding perception of cohorts in emerging markets. Further enquiry is necessary into the cross-cultural dimensions of generational and ethnic cohorts that impact an enabling environment supporting creativity and engagement. Our research indicated that gender had a strong negative relationship between work environment and the psychological contract. As such, we believe further research is necessary to determine specific work environment requirements within a cross-cultural frame for female employees.

Retention of key employees is critical to the long-term success of an organisation. Retaining talented employees plays a vital role in customer satisfaction, product sales, satisfied co-workers and effective succession planning. To achieve the benefits of a competitive advantage and long-term success, talent management must become a critical strategic imperative. For this to be achieved, a clear and compelling strategy must be developed to attract, engage and retain staff across the different generations. By becoming more aware of the needs and preferences of their diverse workforce, organisations can develop suitable TMS. It is also important to determine whether strategies can be uniformly applied across the board or whether different strategies are required for various generations of employees comprising different gender and ethnic groupings.