The purpose of my thesis is to investigate whether the workplace construct, generational membership, predicts employee preferences and attitudes towards compensation and benefits (C&B). My thesis builds on very recent developments in this emerging research program. I will test four hypotheses based on central constructs of presumed increased generational divergence; individually-oriented rewards, workforce mobility, motivation and work-life balance. A cross-sectional design is used to inquire into this problem, and my sample consist of working economics/administration post-graduates. The hypotheses of workforce mobility and motivation were supported, and there is a statistically significant relationship with age under the hypothesis of work-life balance as well, albeit in a different direction than initially hypothesized. Overall, this support means my research question was answered in the affirmative, there were differences.
Generational differences has been a literary topic throughout history, from Antiquity (like Plato’s Republic) up to our times. However, lower life expectancy meant that fewer generations coexisted previously (Denney, McNown, Rogers, & Doubilet, 2013), with life expectancy only being 28 years in Antiquity (Encyclopædia Britannica, 2013). Increased life expectancy is making the topic more pressing, since the near future will witness an unprecedented generational make-up, in that up to five different generations will work side-by-side; Matures (1925-42), Baby Boomers (1943-60) and Generation-X (1961-81), -Y (1982-2000), and -Z (2000-20) (Kulik, Ryan, Harper, & George, 2014). Of all things this demographic shift might have implications for, C&B is hitherto one of the least studied (Xavier, 2014).
Lyons and Kuron (2014) recently reviewed the research on generational differences, building on the inconclusive review by Parry and Urwin (2011), and thence determined that generations was ready to enter the scene as a new workplace variable. This new variable is still controversial, and rightly so, which is why it has to be studied further in order to determine the boundaries of its workplace validity. Hence, if this is a valid workplace variable, then it presumably has to manifest itself in important workplace issues such as C&B.
The purpose of my thesis is to answer a hitherto neglected research question (Xavier, 2014), and that is what impact generational membership has on C&B preferences and attitudes. An organizations C&B program plays a pinnacle role in its ability to attract, retain and motivate employees (Martocchio, 2014). C&B programs can be “one size fits all”, individualized or group-ased (WorldatWork, 2007). That latter option would be interesting if generational differences is a deep fault-line for C&B preferences, which some suggest (F. Hansen, 2010; Thomsen, 2012).
I accommodate the recommendations for further inquiry put forth by Lyons and Kuron (2014), in that I will have a strong link to theory, consider the context, study a topic of probable generational tension, and heed their methodological admonitions. My theoretical contribution is to test the scope of the generation construct, i.e. whether it is relevant for C&B. The practical contribution is to better help HR professional understand how generations effect C&B.
There are contending directions both within the field of employee rewards and generational differences. Within these, I pick directions based on their theoretical and empirical merits. I will henceforth give an in-depth presentation of the theoretical framework from which I delineate the independent and dependent variables, and argue why I selected the four generational issues (incl. hypotheses) introduced below.
Generational research. There are five theoretical ways to explain generational differences; cohort, aging, social forces, life/career-stage and contextual. The latest literature reviews highlight that the social forces direction is the most promising for future research (Lyons & Kuron, 2014; Parry & Urwin, 2011). Hence, my focus throughout the thesis will be on social forces and life/career-stage, given how the former is holistic and also encapsulate key aspects of the other directions.
Cohort effects. A cohort is a group of individuals that are differentiated from other groups, by shared experiences happening within a given time-frame (Laufer & Bengtson, 1974). These are thought to create a shared identity (attitudes, perceptions, preferences and values), e.g. among the Baby boomers, that differentiates them from other generations. As such, cohort research promotes the use of clear generational categories, which is problematic both theoretically and methodologically. A big challenge for cohort studies is to distinguish between true cohort effects, and the confounding influence of aging (maturation) and the contextual period effect (Lawler III, 2011). Attitudes stabilize in early adulthood (Stockard, Carpenter, & Kahle, 2014), and sampling for generational research should take that into account.
Aging effects. There is support for the case that biological and psychological developments across one’s lifetime affects attitudes and behavior, which in turn can explain differences we see between generations. Changes in mental plasticity, personality, priorities, and fluid mental ability are directly related to aging (Goldstein, 2010). As an example, we know that young adults are much more risk seeking than their older peers, as per the higher incidence of automobile accidents (Karaca-Mandic & Ridgeway, 2010). Which in part is explained by endocrinal and neurological developments (Goldstein, 2010). Further, there is an increase in self-confidence, neuroticism, and extraversion across generations (Lyons & Kuron, 2014). While some think such psychometric properties are contingent on context, psychological developments have also demonstrated stability over time, after crystalizing in early adulthood (Stockard et al., 2014). Further reinforcing the case for careful sampling.
Social forces effects. Karl Mannheim started the generational research program back in 1923 (Pilcher, 1994), but his holistic approach has not been pursued until recently (Lyons & Kuron, 2014). Mannheim’s “social forces“ explanation is that socio-cultural context, complemented by aging and cohort effects, has a formative influence on generational attitudes and behavior (Joshi, Dencker, & Franz, 2011). As such, combining four of the five. His logic is that people growing up during a certain period, integrate historical stereotypes into their social identity, and form an implicit psychological contract with their cohort (Dencker, Joshi, & Martocchio, 2008). The key difference between this and the cohort explanation, is the focus on how contextual factors cause generational differences. Cohort studies think of generations as more chronological groups than sociologically differentiated groups.
Historians generally agree on some key changes rising forth in Western societies, which also falls on generational fault lines: anti-authoritarianism (political/familial/religious), equality, interconnectivity (globalization and ICT), education boom, mobility and short-termism courtesy of higher frequency of change (Cook, 2001; Davis, 2012). A possible repercussion of interconnectivity is that a “global generation” might have emerged (Edmunds & Turner, 2005). Contrarians argue that these developments are heavily moderated by cultural factors (Vincent, 2005). Nonetheless, these developments might have influenced, indirectly, different generation’s thoughts about C&B. The logic being that as the framework of society changes, different generations enter the workforce with age-contingent attitudes and preferences. It follows that my respondents presumably have developed an implicit psychological contract with their respective cohorts, as a function of micro and macro social pressures (Dencker et al., 2008; Joshi et al., 2011). Still, discrepancies between nations historical developments, makes some argue that generational group demarcations likely differ between them (Deal et al., 2012), which suggests a liability of using strict generational groups like GenX and GenY in research.
Life/career-stage effects. Some of the differences we see are most likely the effect of their current life/career stage, which in turn can be relatively homogeneous within a given generational group (Parry & Urwin, 2011). Granted, this is a contrarian argument often used to question and critique the ontology of generations. While there is overlap between life- and career-stage, they can also develop somewhat independently, in that e.g. life-stage developments can precede career-stage. An example would be that someone begets children while still an unemployed student. New graduates usually have student depth, little experience in the workforce, no children, nor are owners in the real estate market. Hence, they could be more interested in base pay, overtime work, and other extrinsic motivators. On the other hand, people in later adulthood are in the opposite situation, and likely to report the inverse of the former. However, multiple contrarian studies find that youth actually do not put as much emphasis on extrinsic motivators when choosing workplace, and rather focus on more intrinsic ones like e.g. TAD (training and development) (Lindquist, 2008; Yeaton, 2008).
Contextual period effects. Workplace culture and national characteristics appear to influence how individuals rate themselves on various workplace variables (Lyons & Kuron, 2014; Parry & Urwin, 2011). There is a known relationship between reward and job preferences (Cable & Judge, 1994), and various socio-cultural dimensions create other differences as well (Chiang & Birtch, 2012). Hence, it is practical to have some level of homogeneity in one’s research sample. Geert Hofstede’s (1984) magnum opus on international differences in work-related values, explores cultural dimensions that explain underlying differences. According to Hofstede’s masculinity/femininity spectrum, Norway is one of the most feminine countries in the world (Myers, 2012). That means that Norwegians rate competition, achievement and success lower than many other countries, and this might naturally affect how they rate various workplace variables. A Norwegian proverb reflecting its deferential culture is “janteloven”, which admonishes that “you should not think you are better than other people”.
On another dimension, Norway is also somewhat more collectivistic than many other western countries (Hofstede, 1984; Myers, 2012). More, strong labor movements, protectionism, and a general skepticism of free-market capitalism has influenced Norway’s political history (Cook, 2001). American culture in particular, has nevertheless made many inroads into the Norwegian psyche through multinational companies, and its powerful entertainment industry. This might have persuaded the younger generations in Norway, who have gown-up with it, to accept different ideas on C&B. Still, the managers who determine workplace policies are usually GenX or older (Joshi et al., 2011), so support for these policies (incl. C&B) might differ between generations.
Empirical support and critics. Hitherto the empirical research on generational differences fractionally support the theories, howbeit suffering from methodological inconsistencies, lack of coordination, and often minimal theoretical grounding (Lyons & Kuron, 2014; Parry & Urwin, 2011). Nevertheless, researchers have unearthed consistent generational differences on various constructs (Beutell, 2013; Cogin, 2012; Gursoy, Chi, & Karadag, 2013; Krahn & Galambos, 2014; Lester, Standifer, Schultz, & Windsor, 2012). The aim of Lyons and Kuron’s (2014) recent literature review was to settle the score between those who have, and those who did not find support for generational differences. They concluded that present research provides sufficient reason to establish generations as a workplace variable, and that the next stage is to map-out the role of mediators and moderators.
Critics argue that generations are not homogeneous enough to be placed into generational groups, and that extraneous variables explain most of the differences found (Giancola, 2008). Yet, a contemporary meta-analysis demonstrated that work attitudes exhibit stability over time (Douglas Low, Mijung, Roberts, & Rounds, 2005), and high school work interest has been used to predict occupational membership 12 years post-graduation (J.-I. C. Hansen & Dik, 2005). I do however meet this criticism, since I do not primarily use a rigid generational membership group variable in this paper, only secondarily.
Compensation and benefits. Compared with other fields within HR, rewards has not received research attention commensurate with its critical role in organizations (Gupta & Shaw, 2014). Yet, reward packages have historically become increasingly broad and intricate, and are imperative for firm strategy (WorldatWork, 2007). C&B has also been front-center in some of the most heated socio-political debates, as exemplified in it being most common cause of strikes (Barber, 2009).
Theories on rewards. Some relevant theories are self-determination theory and social exchange theory. Two central tenants of self-determination theory is that people are motivated by a need to gain fulfillment and psychological growth (Olafsen, Halvari, Forest, & Deci, 2015). As such, this theory leans towards a focus on intrinsic motivation. Competence, autonomy, connection and a sense of belonging are necessary to achieve psychological growth. Social support is important to sustain this. Extrinsic rewards can undermine this if e.g. they are so strong that they constrain autonomy. On the other hand, social exchange theory has a different perspective more focused on social behavior as an exchange process (Emerson, 1976). The goal is to maximize benefits and minimize costs of social relationships, and we tend to discontinue those were the sum of the two is negative. In an employment relationship, we are more likely to quit if the arrangement is not mutually beneficial.
Choosing a reward taxonomy. The literature distinguish between two types of rewards: financial (C&B), where compensation is directly financial and benefits are indirectly so, and non-financial rewards like recognitions. “Total rewards” is a recent and very broad concept in vogue that mixes these two types, and includes; C&B, work-life, performance and recognition, development and career opportunities (WorldatWork, 2007). The total rewards concept is rarely practiced by companies (Brown, 2014), hence I will stick to C&B. C&B is arguably the most exacting reward concept for studying generational differences anyway, since it is well established and widely practiced in organizations, it serves as a metric of extrinsic motivation, and preferences towards it has a clearer theoretical and empirical rationale.
Components of C&B for measurement. Age is not a foreign element of contemporary pay systems, in that it e.g. is common to reward tenure and regulate employee C&B according to previous work experience, career- and life-stage (WorldatWork, 2007). The compensation itself might be fixed monthly, hourly, or piecemeal. And many companies use additional variable incentive systems like commissions, profit sharing and bonuses: these can in-turn be directed towards individual-, group- or organizational-level efforts (WorldatWork, 2007).
Benefits come in many shapes and sizes, and the common denominator is that their financial elements are indirect. While benefits packages are more extensive under well-fare capitalism, they can add up to substantial value in well-fare states as well (Dulebohn, Molloy, Pichler, & Murray, 2009). Examples of common benefits in Norway are electronics like phone and PC, gym membership, travel expenses, and financed education/training (Skattebetalerforeningen, 2014).
Research Question and Conceptual Model
The biggest challenge of generational research is the inherent high level of complexity, given how very different explanations could be given to the same observations. It will be interesting to see whether my research strengthens or weakens the validity of generational membership as a workplace variable. As will become apparent, there is much speaking in favor of both outcomes, and both are equally desirable. While confirmatory research seems unduly overrepresented in journals, disconfirmatory research is of equal importance (Boyd, Gasper, & Trout, 1991).
Research question. My contribution to generational research is to explore the relationship between age and generational differences on C&B attitudes and preferences. I investigate a selection of attitudes likely to affect a generation’s relationship with C&B, and I give respondents an opportunity to allocate preferences within C&B categories. Collectively, the attitudinal Likert and preference rank-order data will answer the research question and four underlying hypotheses.
The research question for this thesis paper is formulated as follows: Can we use generational membership to predict C&B attitudes and preferences?
Conceptual model. The figure on page 8 shows that age is the independent variable used to predict the following dependent constructs; individually-oriented rewards, motivation, workforce mobility and work-life balance. The latter three are measured with multiple dependent variables (both Likert and rank-order) in order to fully tap into the underlying construct. Relationships in the model are tested as linear. While one could hypothesize various mediating, moderating and interactive effects, the theoretical framework in generational research is still too young and conceptually impoverished to explore such nuances. Case in point, most research papers in this field do not even have a theoretical framework, being essentially descriptive in nature (Lyons & Kuron, 2014). More, some of the measurement instruments used herein are not necessarily strong enough to handle such investigations, i.e. were not pilot tested.
There is little consensus in the literature on how to conceptualize and measure my independent and dependent variables (Lyons & Kuron, 2014; Parry & Urwin, 2011). I naturally had to choose those most likely to accurately test the age-C&B relationship. I choose multiple variables to tap into the more complicated constructs in the model (right-hand side), and divided the dependent variables used into primary (strongest theoretically/empirically) and secondary (weaker, but relevant). Those classified as primary variables were those in the literature most frequently used to test these constructs, while the secondary ones were less used. However, due to strong disagreements among generational researchers, I choose to rather err on commission than omission. Primary variables are hence more critical to determine support for hypotheses, while secondary variables serve to corroborate any primary findings.
Independent variable (IV). Age is the continuous independent variable from which I gather information about generational membership. As such, each successive year represents a new generation. That is, I will not subscribe to the use of generational categories to study this (i.e. Boomers, X and Y), since too much information is lost in the process, and the theoretical rationale for these demarcations is very weak (Giancola, 2008). Still, since researchers differ on whether they study generations with a group or continuous age variable (Lyons & Kuron, 2014), I will make a group variable (Baby Boom (1943-60), GenX (1961-81) and GenY (1982-2000)) to check whether there are any discrepancies between the two.
Dependent variables (DV). The common denominators of both generational and C&B relevance identified in the latest reviews, concerned presumed underlying differences on individualism, workforce mobility, extrinsic vis-à-vis intrinsic rewards, and work-life balance (Lyons & Kuron, 2014; Parry & Urwin, 2011).
Instead of measuring general individualism, I measure individually-oriented rewards (primary DV). Which is understood as the extent to which someone wants an individual reward allocation of C&B. I also measured the opposite dimension, i.e. collectivistic reward allocation. As will be explained later, that one was excluded due to low internal consistency.
Workforce mobility is herein understood as the physical and psychological attachment someone has to a company (Lyons et al., 2012), as manifested in ratings of turnover intention (primary DV) and long-term incentive C&B preferences (tenure based-pay, stock options and pension plan: all secondary DV’s).
The construct motivation consists of the intrinsic and extrinsic dimensions of it (primary DV’s): Extrinsic motivation is driven by the instrumental value of an act, opposite intrinsic motivation. I also use C&B preference items to tap into these dimensions (base-/variable-pay, C&B distribution items, and TAD: secondary DV’s).
Work-life balance is understood as ones need for balance (or not) between demands of personal- and work-life (WorldatWork, 2007). Measured by willingness to work inconvenient work hours (overtime/evenings) and weekends/holidays (primary DV’s), and preference for financed work-related expenses and gifts (secondary DV’s).
If there are generational differences, then they ought to manifest themselves as specified in my conceptual model, given the following reasons. While the effect sizes in the sources below varies from moderate to strong, they were of sufficient magnitude to warrant inclusion.
H1: Age and individually-oriented rewards. Increased individualism across generations (Blok, 1998) is likely to affect attitudes towards group- and organizational-level rewards. Some call this development the rise of generation me (Twenge, 2013a), also called millennials or GenY, though critics say this is myth making (Arnett, Trzesniewski, & Donnellan, 2013). Leadership preferences reflect this tendency, in that successive generations seem to prefer relationship- over task-oriented leaders, where the former is less interested in organizational outcomes (Gentry, Deal, Griggs, Mondore, & Cox, 2011). Leadership positions are however contingent on generational power-dynamics (Joshi et al., 2011), though average leader age varies between industries.
Individualism is also a cultural dimension (Hofstede, 1984) , and it is a long-standing debate over the historical teleology of increased individualism, like whether it’s a side-effect of modernization or other forces (Hamamura, 2012). As the aforementioned social forces theory would presuppose, this generational development appears to overlap with the historical development of management practices. In that organizational structure has become decreasingly hierarchical and departmentalized (Scott & Davis, 2007), while downsizing and outsourcing have become more common (Witzel, 2012). It is possible that the observed change in behavior across generations, simply is a natural reaction to these new management practices, but that is a chicken and egg paradox. With this presumed increasing individualism as a backdrop, my contribution is to study whether this also means that we can identify an increase in individual reward orientation across generations. I hence propose the following hypothesis:
H1: There will be a negative relationship between age and preference for individually-oriented rewards.
H2: Age and workforce mobility. This pertains both to one’s physical (geographic) and psychological attachment. Succeeding generations seem to exhibit a greater degree of workforce mobility than the preceding (Dries, Pepermans, & De Kerpel, 2008; Lyons, Schweitzer, Ng, & Kuron, 2012), concomitant with decreasing employee loyalty (Cennamo & Gardner, 2008). Exemplifying increasing physical mobility, Norwegian youths exhibit an increased propensity to leave place of birth to advance their careers (Barlindhaug et al., 2004). Such decisions might have a long-term psychological effect, in engendering a general sense of rootlessness.
A recent meta-analysis (Costanza, Badger, Fraser, Severt, & Gade, 2012) also found an overall decrease in organizational commitment and job satisfaction across generations, and an increase in intention to quit (ITQ). Still, extraneous effects can explain this, since sorting effects in part explain the relationship between age and tenure (Meyer, Stanley, Herscovitch, & Topolnytsky, 2002). Nonetheless, this seems to fall into a general pattern, from which I draw the following implications. The findings on increased ITQ across generation might be replicated through a measure of turnover intention, given their conceptual similarities.
There might also be corresponding differences in how generations perceive short- vis-à-vis long-term rewards in C&B packages, in that newer generations might be more inclined to avoid the latter in order to maintain their mobility. Tenure-based pay, discounted stock options and pension are some common rewards used to retain employees, and more generally incite long-term commitment (Martocchio, 2014; WorldatWork, 2007). Granted, this assumes rational behavior, and that people actually care about C&B package composition. I propose the following hypothesis:
H2: There will be a negative relationship between age and turnover intention, but a positive relationship between age and preference for long-term rewards (tenure-based pay, stock options, pension plan).
H3: Age and motivation. A 30 year time-lagged study in U.S. found that there has been an increased desire for extrinsic values, like status and money across generations (Twenge, Campbell, Hoffman, & Lance, 2010). A later corroborative meta-analysis found a positive relationship between age and intrinsic motivation, and a negative between age and extrinsic motivation (Kooij, De Lange, Jansen, Kanfer, & Dikkers, 2011). Therefore, there seems to be a pattern here as well. However, sorting effects and job type could effect this as well.
These findings are interesting in regards to C&B, since C&B is primarily a financial extrinsic reward, even though benefits are less extrinsic than compensation. Hence, I expect that my younger respondents are more favorably inclined towards compensation than benefits, on the distribution question in my questionnaire: respondents must choose a percentage distribution between C&B. Von Bonsdorff (2011) found that younger nurses tended to prefer extrinsic rewards, though her findings are hard to generalize since e.g. 97% where women. Others (Dulebohn et al., 2009) have emphasized that benefits is a likely issue of generational divergence, in that it becomes a deeper concern as you age. Taken into account the rise in extrinsic motivation across generations, we would then expect to see that attitudinal difference replicated in my paper. In regards to C&B preference items, we would expect younger respondents to rate extrinsic rewards higher than the older ones. That is, we expect them to rate compensation as more important than benefits, and also rate the most extrinsic compensation and benefits items highest. Hence, I propose:
H3: There will be a negative relationship between age and extrinsic motivation and preference for extrinsic rewards (compensation, base-/variable pay), but a positive relationship between age and intrinsic motivation and preference for more intrinsic rewards (benefits, TAD).
H4: Age and work-life balance. While the total rewards movement positions this concept as something separate from benefits, it is more correctly framed as a sub-component of benefits. Both time-lagged (Wray-Lake, Syvertsen, Briddell, Osgood, & Flanagan, 2011) and cross-sectional studies (Gursoy et al., 2013) show that there has been an increased appreciation for work-life balance across generations. However, this is likely affected by the work hours per week increasing in US over the last 30 years. This increasing preference for work-life balance is alternatively explained as an exhibition of lower work-morale among newer entrant to the workforce. Often this is seen in connection with the rise of “generation Me”, i.e. that there is a sense of entitlement and/or self-centeredness among youth. A life-cycle explanation from a recent study affirms that GenX are especially concerned with work-life balance (Becton, Walker, & Jones‐Farmer, 2014). While this hypothesis is the least original one, it could be of empirical interest due to my sample, and the national context situating my research. Thence, I propose the following hypothesis:
H4: There will be a positive relationship between age and preference for work on overtime/evenings and work on weekends/holidays, but a negative relationship between age and preference for gifts and financed work-related expenses.
My method is cross-sectional, which is weaker than various alternatives (Lyons & Kuron, 2014), but was the best option given my time available and access to data. While I would have liked to use time lag, that became impractical for my partner organization.
Sample and procedure. Econa, the biggest trade union for business and economics postgraduates in Norway, assisted in data access and collection. My contact at Econa sent an email survey in early March to a randomized sample of 1970 of their 20,000 members (9,86%). The contact also sent a reminder to those who had not responded within one week of the initial requisition. In order to maximize response rate, Econa provided an incentive in the form of 5 gift cards valued at 1000 NOK each. I was at first somewhat concerned that the presence of a carrot could have a distorting effect on the answers, given the theme of the survey. However, further inquiry on the actual effect of such incentives (Ryu, Couper, & Marans, 2006), and modification of phraseology in the email, neutralized that risk. The participants did not know that the thesis was about generational differences, only knowing the topic was C&B, hence generational stereotypes were not primed when answering.
The survey was sent through a web-based tool called SurveyXact, which is the one Econa subscribes to. I had the option of using Qualtrics, but that would be impractical for Econa, and limit my sample to 700. Seeing as going from 700 to 1970 is an increase of 181%, I choose SurveyXact after doing some due diligence on the program. More, there were additional benefits of using SurveyXact, in that I got access to allot of prerecorded background data on my sample (control variables), such as age, gender, work sector, and county of residence. While I initially where to gain intel on the specifics of respondents educational background and work, that had to be omitted for privacy reasons. Anyway, that omission matter little considering the members by default have similar backgrounds.
We assured the participants of confidentiality. Even though I did not collect sensitive data, I had to send a submission to Norsk Samfunnsvitenskapelig Datatjeneste (NAD), since NAD considers known membership in a trade union sensitive. They replied that they did not find anything problematic about the survey.
Demographics. A risk with this kind of cross-industry study is that the respondent’s preference data lose some of its meaning, since I do not know their current C&B package. There are nevertheless benefits with a broader sample that counterweight weaknesses, to also collect data on their C&B packages would create great privacy issues and make the survey so long as to hurt response rate. I wrote an intro in the survey where I made it clear for the respondents that I wanted to know how they would want their C&B package configured, and admonished them to think independently. Still, I have data on their line of work, so I will control for that in the analysis and discussion.
Our first prompt yielded 507 responses, and the second one increased it to 653, we did not send a third one due to the likely depreciation of data quality (Rao & Pennington, 2013). The response rate of 653 (33%), i.e. whom completed the entire survey, is slightly above the number Econa usually gets. Average age was 42 (SD = 11), with the youngest being 25 and the oldest 67. Regarding categorical generational membership, 25% where GenY (1982-2000), 60% GenX (1961-81), and 15% Baby Boomers (1943-60). This discrepancy in distribution, further argues in favor of measuring generation as a continuous age variable were each year of birth represents a new generation. Predominance of GenX makes sense given how membership in a trade union is likely affected by life-stage. Both genders where adequately represented, considering that males are overrepresented in these postgraduate programs: 62% where male and 38% female. Geographically speaking, my sample was highly representative of Norway, with all counties adequately represented. Following is the industry sector representation: private 36%, NHO 14%, unregistered 14%, state government 13%, FA 8%, county/municipality 6%, specter 3%, specter/health 2%, Oslo municipality 1%, independent 1%, KS-business 1%, Virke 1%. While the industry distribution is not as nationally representative, the deviation is not troubling given the samples backgrounds.
I studied the demographic characteristics of the non-respondents (n = 1259, 67%), and found no notable differences between them and the respondents. I also had data on those who started, but did not complete, the survey (n = 58, 2,94%). There were some notable characteristics there, in that 45% where male and 55% female, meaning that women appear to be less interested. Regarding work sector, those working in NGOs and municipality where least likely to complete the survey, indicating some degree of self-selection. This self-selection might relate to the rigid nature of rewards in those sectors, yet a sample of 58 is not highly generalizable.
Measures. I am venturing into relatively unexplored territory, and there are thence no ready-to-go surveys for my problem. I have built on previous research to the greatest extent possible. My dependent variables where the most challenging part, while the independent one (age) was naturally easy to measure. Measurement of individually/collectivistically-oriented rewards, motivation and turnover intention used a 5-point Likert scale (normative), ranging from 1 (strongly disagree) to 5 (strongly agree). Measures of C&B preferences used an ipsative ranking design, ranging from most (1) to least preferred (highest value). Using both normative and ipsative measures enables me to circumnavigate their respective weaknesses (Field, 2009; Hair, Black, Babin, & Anderson, 2014). Response biases is a risk with normative, and psychometric issues with ipsative (Chieh-Chen Bowen, Martin, & Hunt, 2002). Likert items were turned into scales of the underlying construct(s), like e.g. extrinsic motivation. Rank-order items, while belonging to underlying themes like e.g. compensation, could not be transformed into scales, since they did not measure any underlying psychometric construct (Hair et al., 2014).
Independent variable. I measure it by collecting info about respondent’s date of birth, from which I can delineate generational differences. I also made a generational category variable: GenY (25-33 yo), GenX (34-54 yo), and Boomers (55+ yo). Given the conflicting demarcations of these categories, further complicated by national/cultural differences, it might be wise to focus on age (scale). Another issue is that information is lost when continuous data is grouped (Hair et al., 2014).
Dependent variables: Likert. I use three psychometric instruments developed by my supervisor, for intrinsic (a 0.89) and extrinsic motivation (a 0.74), and for turnover intention (a 0.90) (Kuvaas & Dysvik, 2009, 2011). My supervisor and I also developed the measure of individual/collectivist in relation to C&B. An EFA identified both factors, though the Cronbach’s Alpha was moderate 0.58 for individually-oriented reward allocation and 0.49 for collectivistically-oriented allocation. Cutoff level for internal consistency differs depending on the nature of the study, with lower levels between 0.5 and 0.6 considered sufficient in exploratory research (Cho & Kim, 2015). Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was well-above the minimum of 0.6 for all Likert measures.
Dependent variables: Rank-order. Alas, absence of standardized questionnaires for C&B preferences was not unexpected, considering that strategic C&B planning and design is a relatively new phenomenon (Martocchio, 2014). I could have alternatively measured a related, but different, concept like pay satisfaction. The big disadvantage with that concept is that it is harder to interpret such results, vis-à-vis a preference survey, when you do not know how their actual C&B packages are structured. The C&B items are divided into four groups: compensation (a 0,9), benefits (a 0,85), base pay vs. variable pay distribution (a 0,9), compensation vs. benefits distribution (a 0,9). Some of these variables do not have any assumed relationship with age, and not of direct relevance to my hypotheses. Items like “health and recreation”, “educational background” and the “base-/variable-pay distribution” items were included for the interest of Econa. Their absence would also be very strange in the questionnaire groups they appear.
Values of internal consistency were calculated through optimal scaling, as recommended by SPSS’s e-manual (IBM, 2015), though some argue that its not meaningful to calculate this for ipsative. Primary sources of C&B preference items are (Chonko, Tanner Jr, & Weeks, 1992; Lopez, Hopkins, & Raymond, 2006), and secondary ones are (Guthrie, 2000; Hackman & Oldham, 1980; R. J. Long, 2001; McClelland, 1961). These were adapted to the Norwegian context, and represent the most common reward items in Norway such; i.e. high level of content validity.
Other measures. Econa had set aside 5000 NOK to buy gift cards, and they wanted to give the respondents the opportunity to choose between alternative ways of allocating this sum: (a) give 10 NOK to the rainforestfund, (b) give their potential share back to Econa, and (c) be in the poll for a gift card. Beyond these questions, Econa asked the participants whether they would like to comment the survey, participate in a follow-up survey, and whether they would like to participate in surveys on other topics in the near future.
For ease of use, means are listed from low to high in the table. The acronyms in this and following tables refer to the hypotheses: H1 (individually-oriented rewards), H2 (workforce mobility), H3 (motivation) and H4 (work-life balance), denoting which hypothesis a given variable tests. ““Health and recreation”, “educational background” and the “base-/variable-pay distribution” items were not were not relevant for any of the hypotheses, but included for the interest of Econa.
The measures of spread, are naturally affected by the forced rank order nature of these four groups. High kurtosis and positive skewness in base pay and pension was expected, and anything else could signal that respondents were very a-typical, misunderstood the survey, or not take it seriously. I similarly expected top-scores to heavily congregate in the upper portion of the distribution split options, given how uncommon the rest are in real-life. Someone wanting e.g. 90% or 70% pay to be variable would be abnormally risk-prone. We see a higher variety of preferences regarding benefits. Overall, the data show that respondents are rather unanimous in their rating of “typically” highly and lowly preferred items. In other words, their preference allocation seems normal based on what we know from previous research (WorldatWork, 2007), excepting their rather high preference for variable pay. Working weekends/holidays (compensation) and gifts (benefits) were lowest ranked.
Measures of central tendency and spread show a more normal distribution in all these items. Skewness indicates that there are somewhat more negative high scores pulling the tail in that direction in terms of e.g. intrinsic motivation, while the opposite is true for turnover intention. Kurtosis values are much more moderate in these Likert scores, further indicating higher degree of spread. Looking at the means, my sample is quite intrinsically motivated, individualist, but their score on turnover intention indicates that they are somewhat discontent with their job. Arguably, their high intrinsic motivation could compensate for their stronger turnover intention.
Outliers. Given how the scales are limited, outlier responses was a reduced issue. I also did not notice any troubling degree of outliers in the form of abnormal responses, like e.g. someone rating gifts as 1 and pension as 6. Virtually all responses seemed reasonable, so respondents appeared to take the survey seriously.
I found that researchers differed greatly on which inferential technique(s) they used to analyze datasets with both ranking and Likert data. Due to this uncertainty in the research papers and researchers I personally consulted with, I will compare methods as advised. I will primarily use MANOVA, but triple check by comparing results with Kruskal-Wallis and multiple regression: i.e. with age as dependent variable in the latter case. Most common analyses in cases with multiple dependent and one independent variable, where Kruskal-Wallis and MANOVA. A big limit with the former is that it treats the variables individually, and offers very little data on effect sizes, but it is designed to handle ranking data. I especially compared MANOVA and Kruskal-Wallis results when investigating ranking items. MANOVA is quite robust even when assumptions are broken (Almli, Øvrum, Hersleth, Almøy, & Næs, 2015; Bristow, Amyx, Castleberry, & Cochran, 2011; Field, 2009), and variables included in the final model were, as aforementioned, crosschecked.
Through plot analyses, I found substantive improvement of a quadratic equation over a linear one regarding inconvenient work hours (R2 0,037-0,044), pension (R2 0,046-0,066), and intrinsic motivation (R2 0,026-0,032), which suggests curvilinear relationship. However, my sample only consist of people between 25-67 years, with a 60% predominance of GenX (34-54 yo). This represent an overrepresentation of GenX, considering that GenY recently became the largest group in the US workforce (Fry, 2015). An explanation for this is that, GenX’ers are more likely to be members of a trade union, given their life-stage. This does however mean that one should be careful not to extrapolate too much from these curvilinearities.
MANOVA test statistics. The MANOVA procedure provides both univariate and multivariate output. I used Levene’s test to check homogeneity of variance assumption on each variable, and it was broken on gender, pension, tenure and base pay. Box’s test of equality of covariance matrices was significant, which affirms that the assumption was broken. However, MANOVA is robust to violations of homogeneity, and these values are not concerning given how the SD’s were not too big between the variables. Stevens (1980) recommends that number of dependent variables be kept below 10 to avoid distortions in MANOVA, unless the sample size is 1000+. There are 9 in this model, so I don’t expect any issues regarding power.
All tests of multivariate significance, i.e. Pillai’s Trace, Wilks’ Lamba, Hotellin’g, Trace and Roy’s Largest Root were p < .01. Of these multivariate tests, values of Pillai’s Trace are least influenced by violations of assumptions (Field, 2009). Pillai’s value is V = 0.92 F(378, 5454) = 1.65, p < .01. Partial Eta squared for Pillai’s was 0,10, so 10% of the variability across the dependent variables is accounted for by age. Yet, the Eta value was much higher in Roy’s Largest Root test, 0,278 or 27,8% of variability, but Roy is far too sensitive to assumption violations.
The model. Only statistically significant variables are included in the model on page 21, and were found through an iterative process driven by study of the structural matrix and theory. Directions of the effects was also found in the matrix. MANOVA, multiple regression, and Kruskal-Wallis agreed on variables for inclusion in the model. However, a stepwise multiple regression fallaciously retained multiple variables the other two found highly insignificant p = .9, like e.g variable pay and gifts. When comparing the retained variables of the scale age variable (in the model), with the generational category variable, they agreed on all excepting that the latter chose to retain gifts. Kruskal-Wallis also found that age predicted preference for gifts (benefits) p < .01. However, preference for gifts is not very theoretically meaningful, nor crucial to test my fourth hypothesis, and is p = .9 in MANOVA. For post-hoc MANOVA tests, I choose Fisher’s LSD.
Revisiting the hypotheses. Of all the variables in the dataset, only 9 were significant enough to support the hypotheses. The Likert and rank-order variables alone numbered 28 (counting each rank-order item as a separate variable), and I had access to 8 demographic variables. Granted, multiple of these were unlikely to have any relationship with age, but were included to the benefit of Econa.
Regarding H1, there was no support for any relationship between age and individually-oriented rewards (H1) p = .3. The measure of collectively-oriented rewards, its inverse, was also not significant at p > .05.
Regarding H2, both the primary (turnover intention p < .001) and two of the secondary variables (turnover-based pay p < .01 and pension p < .001) were significant: however, preference for stock options was not significant p > .05. Turnover intention decreased with age, while preference for tenure-based pay and pension increased with age.
Regarding H3, both the primary (extrinsic and intrinsic motivation p < .01) and two of the secondary variables (base pay p < .001 and financed TAD p < .01) were significant: the secondary variables variable pay and distribution items were all insignificant p > .05. Extrinsic motivation, preference for base pay and TAD decreased with age, while intrinsic motivation increased with age.
Regarding H4, we see that only one primary variable (work more overtime/evenings p < .01) was significant, and preference for such work decreased with age. However, no other variables was significant for H4 p > .05.
Effect size and structural matrix. Eta2 is a measure of effect size, anything around .02 is considered small, .13 is medium, and at or above .26 is large (Field, 2009; Pierce, Block, & Aguinis, 2004). Univariate between subjects effects of significance are in model above. All variables above have medium effect size, and are quite decent given that large effect sizes are very rare in generational research (Lyons & Kuron, 2014; Parry & Urwin, 2011). Considering that MANOVA is like multiple ANOVA’s being run, one could argue that an alpha correction is needed, and that we lower the cut off to p < .025. In that case, gender is not significant. The gender variable tell us that the older respondents are less likely to be female than the younger ones, which sounds very reasonable given how females gravitation towards my samples degrees is a recent phenomenon. Hence, I retain gender in the model.
Discriminant analysis. To investigate further relationships between dependent variables, I conducted this analysis, which revealed 8-9 discriminant functions. Initial statistics showed the eigenvalues, converted into percentage of variance accounted for, found that the 1st variate accounted for 34,7% of variance (canonical R2 = 0,27), 2nd for 16,6% (canonical R2 = 0,15), 3rd for 14,2% and 4th at 11,1% being a cumulative 76.7%. Following, I got Wilks’s Lambda and the significance of all 9 variates (1 through 9) at (Ʌ = 0,37, x2(378) = 626,754, p < .01), and successive significances after variates are removed. I found that when the 9 variates are tested in combination, they do significantly discriminate on age, but Lambda becomes insignificant (p = .053) when the 2nd variate is removed (Ʌ = 0,59, x2(280) = 319,267, p < .05). This means that 8-9 underlying dimensions in combination, explain the age differences found in MANOVA.
Sector differences. It was a big concern how this could affect how they allocate preferences on C&B, and cause systematic error in those items. I made dummy variables of my sector variable to test this, with private sector as baseline. In the following, I only report significant differences (p < .05). Regarding compensation items: employees in the State prefer variable pay less, while FA and Virke prefer it more than those in the private sector; Spekter and state prefer reward for educational background more. Regarding base pay vs. variable pay distribution: State and Spekter prefer 90/10 allocation more, while entrepreneurs prefer 50/50 more. Regarding benefits items: State, SpekterHealth and Spekter prefer stock options less; State prefer financed TAD more. TAD is the only variable also in MANOVA model.
These findings are very reasonable, and not so large as to spell trouble for my data. Natural differences between the state and private sector are reflected here, in that e.g. variable pay and stock option programs are very uncommon in the former. It is interesting to note that there were no differences in other non-private organizations, like e.g. the municipality. Bear in mind that only 13% of respondents worked in the State, and I did not find anything suggesting this distorts my hypothesis testing.
Gender differences. There were some significant differences (p < .05), like women being less favorable towards variable pay, but more favoring of base pay increases. Women also rated educational background, financed TAD and work-related expenses as more important. However, they were less favorably inclined towards discounted stock options. No significant differences were found on the Likert variables. Relating to work sector, women were more likely to work in the state, specter or spekterhealth than males (p < .05).
Giftcard allocation. I found that 38% choose option (a: 10 NOK to rainforestfund), with 4% choosing (b: give their potential share back to Econa), and 58% opting for (c: giftcard lottery). Chances of winning the giftcard was 0.15%, while option (a) and (b) were 100% guaranteed monetary allocations.
The purpose of this thesis was to investigate whether generational membership predicts employee preferences and attitudes towards C&B. I had a neutral relationship to my hypotheses, since there was allot speaking in favor of both outcomes. Indeed, it is incredibly important not to be biased when doing research, since biases can make you consciously or unconsciously do things that increase the likelihood that they are confirmed (Boyd et al., 1991; Bryman & Bell, 2011).
While I did find significant relationships between age and various variables, all effect sizes were medium. I will follow the recommendations in the latest literature review (Lyons & Kuron, 2014), and holistically discuss my results. Given how researchers differ on conceptualization and measurement of the constructs in my model, I will compare and discuss their differences throughout.
H1: Age and Individually-oriented Rewards
There will be a negative relationship between age and preference for individually-oriented rewards
It is interesting that I did not find any generational differences, given how other researchers have support for the individualism of millennials, with “generation Me” being a common characterization (Arnett et al., 2013; Twenge, 2013a, 2013b). I checked both with the age (scale) and generational (group) variables, both p >.05.
There are various possible explanations for this. Psychometrically speaking, my instrument had low reliability, 0.58 for individualist- and 0.49 for collectivist-orientation. As such, it was never a general individualism measure, yet conceptually closely related. A general individualism measure might have garnered different results. Still, I chose against that, since that would simply be retesting what many have done before, and would not tell us anything about the C&B-age link.
How individualistic behavior is measured differs between researchers, with some opting for measures of peripheral concepts like narcissism and self-confidence (Arnett et al., 2013). Arnett (2013) argues that while he did not find any increase in narcissism, he does contend that younger people are more confident and have higher expectations than what’s reasonable. However, he thinks this is a psychological reaction to youth’s uncertain life-stage. If we take into account the financial situation of my younger respondents life- and career-stage, one could expect to find a significantly higher preference for individually-oriented rewards. This battle between Arnett and Twenge is illustrative of many others in the generational research program, in that they do not really have diametrically opposed views, and differ more on nuances and explanatory frameworks.
Interestingly, that my sample consisted of economics/administration post-graduates might actually have canceled-out generational differences. Economic psychologists have documented a strong relationship between what economists are taught and their ensuing behavior (Barber, 2009; Henrich et al., 2005; W. Long, Malhotra, & Murnighan, 2011). Those who were taught an individualistic line of economic theory behaves accordingly, i.e. self-fulfilling prophecy. Further still, economic theories were much more individualistic in the past than now (Ferraro, Pfeffer, & Sutton, 2005; Ghoshal, 2005). Which means that the older respondents in my sample might respond more individualistic as a function of their prior training, which in turn could cancel out the younger respondent’s inherently higher level of individualism.
Of course, the easiest explanation for the lack of significant differences in my data, is that there never were any in the first place, and that critics are correct in their supposition that generational individualism only manifest itself under certain preconditions (Arnett, 2013; Arnett et al., 2013). A study of cohort effects from 1976-2006 with a GenY(Me) sample of 477,380 found little evidence for changes in either individualism, or similar constructs like egotism, self-esteem and self-enhancement (Trzesniewski & Donnellan, 2010).
Giftcard allocation. As just stated, there is support for the notion that the correlation between economics education and individualistic behavior might have some causal properties (W. Long et al., 2011). While the giftcard allocation question does not directly measure individualism, it is an indicator of self-interest. Of my sample, 58% choose to participate in the lottery for a gift card (option c), 38% choose to give 10 NOK to the rainforestfund (option a), while 4% choose to give their potential share back to Econa (option b). Females where more disposed to give to the rainforestfund than males, but there were no other demographic differences on this item. While I do not have a comparison group with a different educational background, the statistics speak for themselves. Comparative studies of people with different degrees, find that economics graduates consistently stick out by acting more out of self-interest (W. Long et al., 2011). Such findings are also rather consistent across cultures (Henrich et al., 2005).
Conclusion. Overall, this hypothesis is not confirmed in this research context. There are numerous explanations for the lack of significant differences, and of these, the arguments from economic psychology do have substantial empirical substance. However, Norway is much less capitalistic than USA, so the older people in this sample might not been exposed to the same teaching as those.
H2: Age and Workforce Mobility
There will be a negative relationship between age and turnover intention, but a positive relationship between age and preference for long-term rewards (tenure-based pay, stock options, pension plan).
An increasing degree of workforce mobility (Dries et al., 2008; Lyons et al., 2012), and decreasing employer loyalty (Cennamo & Gardner, 2008) is observed in successive generations. Seeing as how there were no standardized measure of this construct, I used multiple items to measure aspects of workforce mobility; turnover intention p < .001, tenure-based pay p < .01, preference for discounted stock options p > .05 and pension plan p < .001.
My measure of turnover intention showed that younger employees were more likely to consider quitting their jobs, corroborating previous research (Arnold & Feldman, 1982; Costanza et al., 2012). That variable was the primary, and most important, dependent variable for this hypothesis. Other researchers have also found similar differences on related constructs, such as overall commitment (D’Amato & Herzfeldt, 2008), affective commitment (Brunetto, Farr-Wharton, & Shacklock, 2012), continuance commitment (Lub, Bijvank, Bal, Blomme, & Schalk, 2012) and job satisfaction (Parry & Urwin, 2012). However, Benson & Brown (2011) did not find any differences, and a meta-analysis of turnover predictors among U.S. child welfare workers, found only small to negligible relationships with age (Kim & Kao, 2014). Nonetheless, general stress, job satisfaction and work-life balance are better predictors of turnover intention than any demographic variable (Katsikea, Theodosiou, & Morgan, 2015; Kim & Kao, 2014; Nohe & Sonntag, 2014).
We saw that preference for tenure-based pay increased with age, and tenure is a very common compensation element globally, though how formalized the process is differs (WorldatWork, 2007). There might naturally be a career-stage effect behind this, essentially driven by self-interest among older employees. Pay-for-performance systems (PFP) (Gerhart & Fang, 2014) is a more contrary movement, with a focus on meritocratic metrics, sometime even in place of tenure (Martocchio, 2014). As such, one could expect that tenure does not factor into pay as much in companies with strong PFP. Beyond career-/life-stage, there also might be a social forces effect here (Johnson, 2010), in which the historical development of C&B systems partly explain the differences. Maybe the younger respondents doubt how much tenure will matter for them as they age with their company?
On the other hand, there was no significant relationship between age and preference for discounted stock options, though few companies use this in C&B below top management level (WorldatWork, 2007). Research on how employees perceive stock options, also show that employees often do not understand underlying mechanisms sufficiently for it to enhance motivation or retention (Convery, Farrell, Krische, & Sedatole, 2013). They can nonetheless have a very powerful effect on income, and increased grant of stock options to executives largely account for the increased discrepancy between staff and executive pay (Bruvik & Gibson, 2011).
The positive relationship between age and preference for pension is unsurprising, given how its relevance increase with age (Dulebohn et al., 2009). Which corroborate findings by others on benefits (Chonko et al., 1992; Lopez et al., 2006). The relationship was partly curvilinear, in that pension peaked in importance at 50 years.
There are various ways to explain the higher level of mobility among youths in my data. Firstly, the social forces explanation is that the younger respondents have grown up with a sense of rootlessness that inhibits them from developing deeper attachment to their company. Historically, labor markets have become increasingly mobile, courtesy of advances in transportation, ICT and globalization (Davis, 2012). This is also reflected in management practices (Scott & Davis, 2007; Witzel, 2012). Hence, it is likely that younger generations have simply adapted to this change (Johnson, 2010). Secondly, the first job post-graduation is probably the hardest one to get, and that would increase the risk that youth will say yes to a job out of necessity rather than passion. As they later build their professional network, they start to shop around for jobs more congruent with their inner passions. The younger respondents in my dataset might be in this situation now, as per e.g. their turnover intention.
Another alternative explanation for my data is that they in some way are erroneous, though that is unlikely given that the turnover instrument used has very high reliability, and given that the other measures used are theoretically and methodologically reasonable. The increase in workforce mobility is one of the strongest findings in generational research (Lyons & Kuron, 2014), so chances are high that my findings are valid.
Conclusion. The lower turnover intention among older workers corroborate their preference for these long-term reward items. As such, this hypothesis is confirmed, and I have offered some explanations for why this is the case. Of the explanations offered, it would appear that this could be a combination of career/life- and some social-forces effects.
H3: Age and Motivation
There will be a negative relationship between age and extrinsic motivation and preference for extrinsic rewards (compensation, base-/variable pay), but a positive relationship between age and intrinsic motivation and preference for more intrinsic rewards (benefits, TAD).
I found significant relationships with age in both motivational directions, in that younger employees were more extrinsically p < .01, but less intrinsically p < .01 motivated. As such, this finding corroborates those by other researchers (Kooij et al., 2011; Twenge et al., 2010), even though the measuring instrument is a bit different. We also saw in the rank order items that preference for base pay p < .001 and TAD p < .01 decreased with age. However, the age-variable pay relationship was p >.05. It is also of interest that none of the distribution items were significant either.
A couple of longitudinal studies have found that the importance of extrinsic motivators is curvilinear across time-periods, peaking in 1990s (Twenge et al., 2010; Wray-Lake et al., 2011). I only found a substantive improvement of a quadratic equation over a linear one regarding intrinsic motivation, and the curve peaked at 51 years, before slightly dropping off. Which might indicate a career-cycle effect, in which advancement opportunities drop-off around that age. It could also pertain to life-cycle, in that retirement becomes salient when you approach 55+. Further, decrease in extrinsic and increase in intrinsic motivation would presumably develop as an employee gradually commits to a given organization, and their task/skill variety and autonomy increase (Fried & Ferris, 1987; Ng & Feldman, 2012).
Some of those who are skeptical of these differences regarding motivation, draw a distinction between age and generational membership, and concedes differences in the former (Wong, Gardiner, Lang, & Coulon, 2008). As indicated above, it is very hard to demarcate age from generations, since there is not enough consensus on how to group generations in the first place. One of the most pinnacle parts of science it to have a theoretical basis for constructs (Boyd et al., 1991; Bryman & Bell, 2011). If we take the most common generational groups, then there is little historical rationale for the demarcations; matures (1925-42), Boomers (1943-60) and GenX (1961-81), GenY (1982-2000). One could make the wartime and post-wartime case for matures and early Boomers, respectively, but the rest are very arbitrary, and is not reflective of the average age at which people have children either.
Krah and Galambos (2014) found interactions between age and cohort effects, in that extrinsic rewards was more important to the younger cohort, but only increased in importance for the younger cohort. A problem with some of these studies is that they also included adolescent respondents, and work values are not that stable before late 20s (Jin & Rounds, 2012). Fortunately, the youngest respondents in my sample is 25 years, with the average being 42, meaning it is expected that their responses are more stable (Stockard et al., 2014). My samples educational background also increases the likelihood that they have thought long and hard about how they stand on rewards prior to the survey.
Base pay is the most fundamental piece of a company’s C&B program (WorldatWork, 2007), and we saw that preference for it was negatively related to age. Granted, 92% of my sample choose base pay as their number one preference, so one should be careful not to read too much into this age finding. Still, this is corroborated by the negative relationship between age and extrinsic motivation. Others have found contrarian evidence that older employees tend to prefer extrinsic rewards, like base pay and variable pay, more than their younger peers (Chonko et al., 1992). A later study found that younger employees preferred pay increases over fringe benefits, and that promotions consistently ranked higher for younger employees, suggesting a career-stage effect (Lopez et al., 2006).
The negative relationship I found between age and preference for TAD is interesting, given how TAD is just as important for older workers (Blume, Ford, Baldwin, & Huang, 2010). One could even argue that TAD is especially important for older workers, given how younger employees are recent graduates with allot of fresh knowledge in mind. Maybe the older respondents are too complacent? Interestingly, this negative relationship between TAD and age conflict with the idea that younger employees are less intrinsically motivated.
We can explain my result in various ways. Firstly, the social forces explanation would tie this to the findings on various attitudinal differences, and that these correlate with historical developments affecting each cohort. Hence, its expected that younger people have internalized more of these sentiments than older ones. Secondly, the life-stage argument made about hypothesis 2 also hold in this case, in that their intrinsic motivation likely affected by them still not having found their way. There is also the risk that the results were erroneous to begin with, but there are too many corroborating findings in this case to lend that much credence.
These findings on motivation, intrinsic in particular, gives more substance to the finding of higher turnover intention. Other significant demographic characteristics found regarding the motivation variables, was men being more extrinsically motivated than women were. Significantly stronger extrinsic motivation was also found among people working at Virke, but one should not read too much into that given how they only represent 1% of respondents, i.e. only 7 people.
Conclusion. This hypothesis is supported through multiple items collectively rallying against the null hypothesis. Firstly, younger respondents did prefer the most extrinsic reward components, such as base pay. Though decreasing TAD preference with age weight in the opposite direction. Secondly, the intrinsic and extrinsic motivation measures corroborate each other, and show a clear age difference.
H4: Age and Work-life Balance
There will be a positive relationship between age and preference for work on overtime/evenings and work on weekends/holidays, but a negative relationship between age and preference for gifts and financed work-related expenses.
The underlying rationale for this hypothesis, was the abovementioned research suggesting that successive generations increasingly desire work-life balance. This is thought to reflect a change in work values and life-style (Lyons & Kuron, 2014). Researchers differ in how they measure work-life balance, and I thence choose to include multiple items likely to best encapsulate it. There are standardized measures of some sub-components of work-life balance, but these are not as suited for generational research. The problem is that they are too narrow or presume a certain life-stage, like e.g. family commitments. Most sources cited in the two latest reviews on this topic (Lyons & Kuron, 2014; Parry & Urwin, 2011), used work-life balance measures that arguably were too life-stage constrained. The items I used to tap into this was preference for inconvenient work hours (p < .01), work on weekends/holidays (p > .05), gifts (p > .05) and financed work expenses (p > .05).
It was at first surprising to find a negative relationship between age and preference for inconvenient work-hours, still that can be explained as a function of life- and career-stage. This relationship was slightly curvilinear, in that it was least preferred by those between 35-55, which for many is the most intense child-rearing phase. Younger employees are less likely to have kids, and might also feel compelled to work more to accelerate their careers. Interestingly, this preference somewhat conflicts with their lower intrinsic motivation from H3, since these work-hours arguably require a combination of extrinsic and intrinsic motivation. More, my finding also conflicts with time-lag studies who found that leisure values decreased with age, though they also found a curvilinear relationship suggesting life-stage effects (Twenge et al., 2010; Wray-Lake et al., 2011). A cross-sectional study of hospitality workers also found differences (Gursoy et al., 2013). Life-stage differences is the most likely explanation of this curvilinear effect, in that particularly “work interfering with family” (WIF) peaks with GenX, whom are in the process of raising a family, before decreasing with GenY (Becton et al., 2014; Beutell, 2013).
Interestingly, there was no significant relationship between age and presumably relevant variables such as preference for work on weekends/holidays, gifts and financed work-related expenses. There is however no substantive reason why there would be generational differences on the latter three variables, and they were included in the preference sets according to how common they were in C&B packages (WorldatWork, 2007). Still, one might expect to see differing attitudes regarding work on weekends/holidays, under a socio-cultural argument that younger people have a weaker bond to the deep religious or political background of them. My younger respondents appear desirous of sacrificing their leisure for pay, which sharply conflicts with the stereotype that they lack initiative and work morale (Ferri-Reed, 2014). Which is a strong case against the hypothesis.
Conclusion. While I had six items tapping into work-life balance, only two of these was p < .05. As such, my hypothesis is disconfirmed, since inconvenient work hours was significant in a direction opposite the one hypothesized, and pension does not sufficiently counterweight the overall picture. The lack of significant difference in these other items questions the validity of previous research (Lyons & Kuron, 2014), which proposed a more general age-contingent difference in work-life balance. Life/career-stage and social forces effects accounted for the lion’s share of explanations, which is natural in this case considering the nature of the hypothesis.
There are strengths and weaknesses with the theoretical and methodological framework of my thesis. A basic issue with all research is that you have to circumscribe its breath and scope for practical reasons. Collecting too much data creates privacy issues, and a long survey is likely to hurt response rate. A general challenge in generational research is that this field is still young, and there are thence various chinks not worked out jet. Like how to define and measure the underlying constructs. Researchers differ greatly in theory and methods used, and I have used an integrative approach. In the following, I will present limitations identified by my respondents, and present steps taken to compensate for those and other limitations.
They could give some comments about the survey upon completion. Most did not comment, and the majority of those who did wrote supportive or neutral comments. Most of the neutral and positive comments were short: “great job”, “liked it” etc.. Given that this is the limitation section, I will focus on the comments that mentioned limitations. Those comments fall into three broad categories; those who found the survey too short, some who thought items did not apply to their current situation, and some who took offense of the option to allocate their possible reward.
Too short? I certainly would have benefited from knowing their salary, tenure, department position, their marital situation, household income, costs of living, specific of their C&B programs etc., but that would lengthen my survey and increase the risk of people not answering. Some of these questions are also very private in nature. My survey was also significantly longer than those usually sent out by Econa, and 3% of those who started quit halfway through the survey, so more questions would be very risky for response rate. However, increased adoption of Big Data research in social science has the potential to alleviate such informational limitations in the future, while maintaining anonymity (White & Breckenridge, 2014).
Did not apply to me. Some commented that the items either did not fully relate to their line of work, or that their life/career stage affected their answers: e.g., one said he was about to retire, and another person was in the process of changing jobs. However, on the surveys first page, I clearly admonished that they try to answer independent of past/current C&B package. I stated that I wanted to know their opinions, as if they could choose themselves. There was no practical way to gain access to the myriad of C&B packages my respondents actually have. This is just an inherent limitation of throwing a wide net vis-à-vis focusing on a single organization.
Giftcard allocation. This last group was smallest in number, but certainly the harshest, being aggravated that they had to choose between; (a) giving 10 NOK to the rainforestfund, (b) give their potential share back to Econa, and (c) be in the poll for gift cards. Some stated that they felt deceived, while others extrapolated that they found it moralizing to have to choose between these alternatives. Still, there was nothing stopping anyone from choosing (c), and 58% choose that option anyway.
Finding differences is one thing, but explaining why they manifest themselves is a much bigger challenge. A key issue pointed out in later literature reviews (Lyons & Kuron, 2014; Parry & Urwin, 2011), is the absence of explanatory frameworks. Having a thick explanation for statistical findings is particularly important, and can compensate for that deficiency. This research is also in a different cultural context, in that most others are conducted within USA, which I initially thought would reduce likelihood of any differences manifesting themselves.
Research design. Cross-sectional designs like mine, are considered weak because attitudes are not stable over time. Attitudes do stabilize in early adulthood (Stockard, Carpenter, & Kahle, 2014); youngest respondents in my sample were 25, with 42 being the average. My design primarily pick-up the period effect, and it is thence difficult to clearly demarcate it from other effects (Twenge et al., 2010). The decision to use a cross-sectional design was driven by practical considerations, since my partner Econa did not want to ask too much of their members, and my survey was one of the longest they have sent out in the first place. Glenn (1976) has written much on how to distinguish the effects introduced in the theory section, but there still is no single method which can disentangle them. Cohort effects are easiest to find through longitudinal sequential data, and time-lag also manages to offer some compelling evidence. Other methods of interest are retrospective accounts, various qualitative methods, cross-temporal meta-analyses, and most recently Big Data analytics.
Sampling. A problem with many generational studies is that their samples often are constrained to a given organization, profession or workplace sector (Lyons & Kuron, 2014). This desire for control over extraneous variables makes their findings harder to generalize, due to respective idiosyncrasies. Making matters worse, a disproportionate number of studies are on people in nursing, hospitality and accounting, were you have a-typical demographic compositions; e.g., there are very few males in nursing or women in accounting. My sample avoids many of these issues in that it consists of people across organizations, professions and work sectors. Even though all members of Econa must be postgraduates in economics/
administration, their members range from graduates in finance to those in event management (Econa, 2015). The demographics in my sample are very favorable both regarding age, geographic representation and gender (62% male and 38% female). The higher number of males is expected from educational background, but both are adequately represented. All my respondent were in the workforce, unlike other generational researchers who mix with students, or even consist only of students.
Did non-academic sources affect the thesis? As aforementioned, generational differences and C&B are very hot topics in non-academic circles; i.e. the popular press, professional organizations, and consultancies. Of these, I only use professional organizations in this thesis, since their access to workplace data is unparalleled. Still, neither of these hold themselves to the same standards as peer-reviewed journals. The problem with their work pertains to various, methodological, economic, and political conflicts of interests that might unduly influence their results.
There is a chance that such organizations have had an undue influence on the academic development and treatment of this topic, in either creating an issue that does not really exist, caricaturing it, or otherwise affecting how it is studied. On the other hand, they have conducted various surveys and analyses, which contain insights. Based on the literature, I suspect that the generational categories (Boom, GenX and GenY) are courtesy of these organizations, rather than academia.
These organizations have found various findings corroborating mine, but their reports lack transparency, references, the scientific approach, and only serve as soft support. The biggest professional organizations are, SHRM (2012), CIPD (CIPD, 2012, 2013) and WorldatWork (2007). SHRM found significant differences between generations on the importance of the firms’ financial situation, and aspects of the benefits program. CIPD (2012) found that newer generations where significantly more positive towards pay for individual performance than older generations. The biggest HR consultancies on these topics are Towers Watson (2012), Hay Group (2013), Aon Hewitt (2012), Deloitte (Flynn & Valenzuela, 2014), and Mercer (2012).
Suggestions for Future Research
While the later literature reviews affirm that generations is a valid workplace variable, and this thesis suggest its impact on C&B, we do not really know its full scope and replicability. My suggestions for future research fall into three camps; quantitative, qualitative and triangulated.
Quantitative. I initially confronted various challenges by having a mix of Likert and rank-order variables. It seems like Likert is overused in the field, and testing more with alternative measuring instruments can unearth new perspectives, and mitigate common respective weaknesses. Further, generational researchers often use very limited samples in regards to demographics, organizations, professions and work sectors. Exploring more with broader samples, as in my case, is of interest given how my macro-level sample corroborated various findings from the meso-level.
Research questions. Survey: What is the relationship between career-stage and C&B preferences? Survey: Does career- and life-stage develop congruently or not? Big Data social-science mining (White & Breckenridge, 2014): Which variables best predicts job-seeking search-terms?
Qualitative. Within this line of research, we want to use methods such as deep interviews, text analysis, and focus groups to collect thick data on this topic. Even though such data is less generalizable, one can reduce that liability by properly screening potential participants. The primary benefit is the volume of information gathered. One of the biggest issue in generational research, is that we do not have enough information to clearly distinguish between explanatory effects.
Research questions. Focus group: How do different generations perceive theirs, and each other’s, C&B attitudes and preferences? Text analysis: Compare the narratives behind generational differences in the media vis-à-vis academia.
Triangulated. My thesis is an example of a triangulated approach, were a quantitative method is used, but the analysis also applies qualitative perspectives. While one stands the risk of overcomplicating the research design and interpretation, this approach is recommended (Lyons & Kuron, 2014; Parry & Urwin, 2011).
Research question. Survey and focus group: What is the difference between pay satisfaction and preferences between generations, and how is this understood?
Implications for Practice
Many of the topics above are hot among practitioners, yet while there were some significant generational differences, ones should also heed all those who were not so. The practical benefit of this thesis is that measures used are much more objective than those in the popular press, and interpretations are based on a neutral stance were all outcomes are equally desirous. As such this can help in bridging the research-practice gap on this topic (DeNisi, Wilson, & Biteman, 2014).
The finding that age predicts preferences for tenure-pay and pension is probably not very revelatory, and maybe of reduced practical implication. We might have a causal relationship between tenure-pay and increased intrinsic motivation, and if so, companies could benefit by promoting tenure-pay for younger employees as well. However, that younger employees are open to work inconvenient work hours conflicts with the idea that they are not as committed as older peers, and further conflicting with the idea that they are more concerned with work-life balance. It might relate to their financial life-stage, and reflected in their preference for extrinsic motivators. Preference for TAD decreased with age, which conflicts with how older employees appear to be more intrinsically motivated. If anything, the need for TAD increases with age, with the sudden emergence of computers as an example, hence employers can consider whether that need should be communicated better.
Undoubtedly, turnover is a big issue for companies, since it is very costly to recruit and train new employees (Cascio & Boudreau, 2011). That turnover intention appears to be higher among younger employees is concerning, and this finding was corroborated by their lower intrinsic motivation. It could then behoove employers to emphasize intrinsic motivation more in recruitment, in order to avoid staffing with people who do not fit well with organizational culture, or who essentially intend to use the employer as a temporary pit stop. Employers can take steps to enhance intrinsic motivation and as such reduce risk of turnover, seeing as how the matrix analysis showed a relationship.
Overall, when taking the pulse on workforce demographics, some often-stated generational differences manifest themselves. The effect sizes are not large enough to warrant allot of resources from a given HR department, since there are other factors with bigger impact on the workplace. Still, findings could have value for how HR communicates and relate to different age groups (Hicks & Block, 2014).
This thesis explored the intersection between two fields in HR of increasing importance, C&B and generational research. It also has clear practical implications for companies and researchers that may peruse it. Interestingly, two out of four hypotheses were confirmed. The biggest challenges with this research paper, was how this field is young with various chinks not worked out yet. As such, “nothing came for free”, in that the ontological, epistemological and methodological framework for inquiry required more active decision-making by the researcher. On various facets, my approach improved upon prior research, in being much more transparent and avoided various common theoretical and methodological traps.
Almli, V. L., Øvrum, A., Hersleth, M., Almøy, T., & Næs, T. (2015). Investigating individual preferences in rating and ranking conjoint experiments: A case study on semi-hard cheese. Food Quality and Preference, 39, 28-39. doi: 10.1016/j.foodqual.2014.06.011
Aon Hewitt. (2012). Total rewards survey: Transforming potential into value. Retrieved from: http://www.aon.com/human-capital-consulting/thought-leadership/talent_mgmt/2012_aonhewitt_total_rewards_survey.pdf
Arnett, J. J. (2013). The evidence for Generation We and against Generation Me. Emerging Adulthood, 1(1), 5-10. doi: 10.1177/2167696812466842
Arnett, J. J., Trzesniewski, K. H., & Donnellan, M. B. (2013). The dangers of generational myth-making: Rejoinder to Twenge. Emerging Adulthood, 1(1), 17-20.
Arnold, H. J., & Feldman, D. C. (1982). A multivariate analysis of the determinants of job turnover. Journal of Applied Psychology, 67(3), 350-360.
Barber, W. J. (2009). History of Economic Thought. Middletown, CT, USA: Wesleyan University Press.
Barlindhaug, J. P., Angell, E., Landstad, B., Sandberg, P., Grande, S. K., Malonæs, K., . . . Sund, Å. (2004). NOU, Norges offentlige utredninger: Livskraftige distrikter og regioner. Retrieved from: https://www.regjeringen.no/contentassets/468c7a31567345bc9499b6499b87a829/no/pdfs/nou200420040019000dddpdfs.pdf
Becton, J. B., Walker, H. J., & Jones‐Farmer, A. (2014). Generational differences in workplace behavior. Journal of Applied Social Psychology, 44(3), 175-189.
Benson, J., & Brown, M. (2011). Generations at work: Are there differences and do they matter? International Journal of Human Resource Management, 22(9), 1843-1865. doi: 10.1080/09585192.2011.573966
Beutell, N. (2013). Generational differences in work-family conflict and synergy. International Journal of Environmental Research and Public Health, 10(6), 2544-2559.
Blok, A. (1998). The narcissism of minor differences. European Journal of Social Theory, 1(1), 33-56.
Blume, B. D., Ford, J. K., Baldwin, T. T., & Huang, J. L. (2010). Transfer of training: A meta-analytic review. Journal of Management, 36(4), 1065-1105.
Boyd, R., Gasper, P., & Trout, J. D. (1991). The Philosophy of Science. USA: MIT Press.
Bristow, D., Amyx, D., Castleberry, S. B., & Cochran, J. J. (2011). A cross-generational comparison of motivational factors in a sales career among Gen-X and Gen-Y college students. Journal of Personal Selling & Sales Management, 31(1), 77-85. doi: 10.2753/PSS0885-3134310105
Brown, D. (2014). The future of reward management: From total reward strategies to smart rewards. Compensation & Benefits Review, 46(3), 147-151. doi: 10.1177/0886368714549303
Brunetto, Y., Farr-Wharton, R., & Shacklock, K. (2012). Communication, training, well-being, and commitment across nurse generations. Nursing Outlook, 60(1), 7-15. doi: 10.1016/j.outlook.2011.04.004
Bruvik, K., & Gibson, J. W. (2011). The past, present and future of executive compensation. Business Studies Journal, 3(1), 69-83.
Bryman, A., & Bell, E. (2011). Business Research Methods (3 ed.). NY, USA: OUP Oxford.
Cable, D. M., & Judge, T. A. (1994). Pay preferences and job search decisions: A person-organization fit perspective. Personnel Psychology, 47(2), 317-348.
Cascio, W., & Boudreau, J. (2011). Investing in People: Financial Impact of Human Resource Initiatives (2 ed.). New Jersey: Financial Times Press.
Cennamo, L., & Gardner, D. (2008). Generational differences in work values, outcomes and person-organisation values fit. Journal of Managerial Psychology, 23(8), 891-906.
Chiang, F. F. T., & Birtch, T. A. (2012). The performance implications of financial and non-financial rewards: An Asian Nordic comparison. Journal of Management Studies, 49(3), 538-570. doi: 10.1111/j.1467-6486.2011.01018.x
Chieh-Chen Bowen, K. D., Martin, B. A., & Hunt, S. T. (2002). A comparison of ipsative and normative approaches for ability to control faking in personality questionnaires. International Journal of Organizational Analysis (1993 – 2002), 10(3), 240.
Cho, E., & Kim, S. (2015). Cronbach’s coefficient alpha: Well known but poorly understood. Organizational Research Methods, 18(2), 207-230. doi: 10.1177/1094428114555994
Chonko, L. B., Tanner Jr, J. F., & Weeks, W. A. (1992). Selling and sales management in action: Reward preferences of salespeople. Journal of Personal Selling & Sales Management, 12(3), 67.
CIPD. (2012). Employee attitudes to pay 2012. Retrieved from: https://www.cipd.co.uk/hr-resources/survey-reports/employee-attitudes-pay-2012.aspx
CIPD. (2013). Annual survey report 2013. Reward Management. Retrieved from: https://www.cipd.co.uk/binaries/reward-management_2013.pdf
Cogin, J. (2012). Are generational differences in work values fact or fiction? Multi-country evidence and implications. International Journal of Human Resource Management, 23(11), 2268-2294. doi: 10.1080/09585192.2011.610967
Convery, S. P., Farrell, A. M., Krische, S. D., & Sedatole, K. L. (2013). How to help employees better value stock options as compensation. Journal of Financial Planning, 26(1), 34-41.
Cook, B. A. (2001). Europe Since 1945: An Encyclopedia. New York, USA: Garland Publishing.
Costanza, D., Badger, J., Fraser, R., Severt, J., & Gade, P. (2012). Generational differences in work-related attitudes: A meta-analysis. Journal of Business & Psychology, 27(4), 375-394. doi: 10.1007/s10869-012-9259-4
D’Amato, A., & Herzfeldt, R. (2008). Learning orientation, organizational commitment and talent retention across generations A study of European managers. Journal of Managerial Psychology, 23(8), 929-953. doi: 10.1108/02683940810904402
Davis, A. H. (2012). History: From the Dawn of Civilization. New York: USA: DK Publishing.
Deal, J. J., Stawiski, S., Graves, L. M., Gentry, W. A., Ruderman, M., & Weber, T. J. (2012). Perceptions of Authority and Leadership: A Cross-National, Cross-Generational Investigation. In E. S. Ng, S. T. Lyons, & L. Schweitzer (Eds.), Managing the New Workforce: International Perspectives on the Millennial Generation (pp. 281-306): Cheltenham, U.K. and Northampton, Mass.: Elgar.
Dencker, J. C., Joshi, A., & Martocchio, J. J. (2008). Towards a theoretical framework linking generational memories to workplace attitudes and behaviors. Human Resource Management Review, 18(3), 180-187. doi: 10.1016/j.hrmr.2008.07.007
DeNisi, A. S., Wilson, M. S., & Biteman, J. (2014). Research and practice in HRM: A historical perspective. Human Resource Management Review, 24(3), 219-231. doi: 10.1016/j.hrmr.2014.03.004
Denney, J. T., McNown, R., Rogers, R. G., & Doubilet, S. (2013). Stagnating life expectancies and future prospects in an age of uncertainty. Social Science Quarterly (Wiley-Blackwell), 94(2), 445-461. doi: 10.1111/j.1540-6237.2012.00930.x
Douglas Low, K. S., Mijung, Y., Roberts, B. W., & Rounds, J. (2005). The stability of vocational interests from early adolescence to middle adulthood: A quantitative review of longitudinal studies. Psychological Bulletin, 131(5), 713-737. doi: 10.1037/0033-2909.131.5.713
Dries, N., Pepermans, R., & De Kerpel, E. (2008). Exploring four generations’ beliefs about career: Is “satisfied” the new “successful”? Journal of Managerial Psychology, 23(8), 907-928. doi: 10.1108/02683940810904394
Dulebohn, J. H., Molloy, J. C., Pichler, S. M., & Murray, B. (2009). Employee benefits: Literature review and emerging issues. Human Resource Management Review, 19(2), 86-103. doi: http://dx.doi.org/10.1016/j.hrmr.2008.10.001
Econa. (2015). Hvem kan bli medlem? Retrieved 1.2., 2015, from https://www.econa.no/hvem-kan-bli-medlem
Edmunds, J., & Turner, B. S. (2005). Global generations: social change in the twentieth century. British Journal of Sociology, 56(4), 559-577. doi: 10.1111/j.1468-4446.2005.00083.x
Emerson, R. M. (1976). Social exchange theory. Annual Review of Sociology, 2, 335-362. doi: 10.1146/annurev.so.02.080176.002003
Encyclopædia Britannica. (2013). Mortality. Retrieved 21.10, 2014, from http://www.britannica.com/EBchecked/topic/393100/mortality
Ferraro, F., Pfeffer, J., & Sutton, R. I. (2005). Economics language and assumptions: How theories can become self-fulfilling. Academy of Management Review, 30(1), 8-24. doi: 10.5465/AMR.2005.15281412
Ferri-Reed, J. (2014). Are millennial employees changing how managers manage? Journal for Quality & Participation, 37(2), 15-35.
Field, A. (2009). Discovering Statistics Using SPSS (3 ed.). London: SAGE Publications Ltd.
Flynn, J., & Valenzuela, J. (2014). 2014 global top five total rewards priorities survey. Retrieved from: http://www2.deloitte.com/content/dam/Deloitte/global/Documents/HumanCapital/dttl-2014-top-five-global-employer-rewards-priority-survey-report-20140423.pdf
Fried, Y., & Ferris, G. R. (1987). The validity of the job characteristics model: A review and meta-analysis. Personnel Psychology, 40(2), 287-322.
Fry, R. (2015). Millennials surpass Gen Xers as the largest generation in U.S. labor force. Retrieved 17.5, 2015, from http://www.pewresearch.org/fact-tank/2015/05/11/millennials-surpass-gen-xers-as-the-largest-generation-in-u-s-labor-force/
Gentry, W. A., Deal, J. J., Griggs, T. L., Mondore, S. P., & Cox, B. D. (2011). A comparison of generational differences in endorsement of leadership practices with actual leadership skill level. Consulting Psychology Journal: Practice & Research, 63(1), 39-49. doi: 10.1037/a0023015
Gerhart, B., & Fang, M. (2014). Pay for (individual) performance: Issues, claims, evidence and the role of sorting effects. Human Resource Management Review, 24(1), 41-52. doi: 10.1016/j.hrmr.2013.08.010
Ghoshal, S. (2005). Bad management theories are destroying good management practices. Academy of Management Learning & Education, 4(1), 75-91. doi: 10.5465/AMLE.2005.16132558
Giancola, F. (2008). Should generation profiles influence rewards strategy? Employee Relations Law Journal, 34(1), 56-68.
Glenn, N. D. (1976). Cohort analysts’ futile quest: Statistical attempts to separate age, period and cohort effects. American Sociological Review, 41(5), 900-904.
Goldstein, E. (2010). Cognitive Psychology: Connecting Mind, Research and Everyday Experience. CA, USA: Cengage Learning.
Gupta, N., & Shaw, J. D. (2014). Employee compensation: The neglected area of HRM research. Human Resource Management Review, 24(1), 1-4. doi: 10.1016/j.hrmr.2013.08.007
Gursoy, D., Chi, C. G.-Q., & Karadag, E. (2013). Generational differences in work values and attitudes among frontline and service contact employees. International Journal of Hospitality Management, 32, 40-48.
Guthrie, J. P. (2000). Alternative pay practices and employee turnover: An organization economics perspective. Group & Organization Management, 25(4), 419.
Hackman, J. R., & Oldham, G. R. (1980). Work Redesign. Reading, Massachusetts: Addison- Wesley.
Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2014). Multivariate Data Analysis: Pearson New International Edition (7 ed.). London: Pearson Education Ltd.
Hamamura, T. (2012). Are cultures becoming individualistic? A cross-temporal comparison of individualism–collectivism in the United States and Japan. Personality and Social Psychology Review, 16(1), 3-24.
Hansen, F. (2010). Currents in compensation and benefits. Compensation & Benefits Review, 42(6), 435-449. doi: 10.1177/0886368710387345
Hansen, J.-I. C., & Dik, B. J. (2005). Evidence of 12-year predictive and concurrent validity for SII Occupational Scale scores. Journal of Vocational Behavior, 67(3), 365-378. doi: http://dx.doi.org/10.1016/j.jvb.2004.08.001
Hay Group. (2013). The changing face of reward. Retrieved from: http://www.haygroup.com/downloads/uae/The_Changing_Face_of_Reward_Website.pdf
Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., . . . Tracer, D. (2005). ‘Economic man’ in cross-cultural perspective: Behavioral experiments in 15 small-scale societies. Behavioral and Brain Sciences, 28(6), 795-855.
Hicks, M., & Block, L. (2014). “What’s in It for Me?” Targeting Rewards Messaging to All Generations. Benefits Quarterly, 30(2), 47-50.
Hofstede, G. (1984). Culture’s Consequences: International Differences in Work-Related Values. California, USA: SAGE Publications.
IBM. (2015). What Is Optimal Scaling? Retrieved 2.3, 2015, from http://www-01.ibm.com/support/knowledgecenter/SSLVMB_20.0.0/com.ibm.spss.statistics.help/optimal_scaling_whatis.htm
Jin, J., & Rounds, J. (2012). Stability and change in work values: A meta-analysis of longitudinal studies. Journal of Vocational Behavior, 80(2), 326-339. doi: 10.1016/j.jvb.2011.10.007
Johnson, M. J., L. (2010). Generations, Inc. New York: AMACOM.
Joshi, A., Dencker, J. C., & Franz, G. (2011). Generations in organizations. Research in Organizational Behavior, 31, 177-205. doi: 10.1016/j.riob.2011.10.002
Karaca-Mandic, P., & Ridgeway, G. (2010). Behavioral impact of graduated driver licensing on teenage driving risk and exposure. Journal of Health Economics, 29(1), 48-61. doi: 10.1016/j.jhealeco.2009.10.002
Katsikea, E., Theodosiou, M., & Morgan, R. E. (2015). Why people quit: Explaining employee turnover intentions among export sales managers. International Business Review, 24(3), 367-379. doi: 10.1016/j.ibusrev.2014.08.009
Kim, H., & Kao, D. (2014). A meta-analysis of turnover intention predictors among U.S. child welfare workers. Children and Youth Services Review, 47(Part 3), 214-223. doi: 10.1016/j.childyouth.2014.09.015
Kooij, D. T. A. M., De Lange, A. H., Jansen, P. G. W., Kanfer, R., & Dikkers, J. S. E. (2011). Age and work-related motives: Results of a meta-analysis. Journal of Organizational Behavior, 32(2), 197-225. doi: 10.1002/job.665
Krahn, H. J., & Galambos, N. L. (2014). Work values and beliefs of ‘Generation X’ and ‘Generation Y’. Journal of Youth Studies, 17(1), 92-112.
Kulik, C. T., Ryan, S., Harper, S., & George, G. (2014, 08//). Aging populations and management, Editorial. Academy of Management Journal, pp. 929-935. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=97333431&site=ehost-live
Kuvaas, B., & Dysvik, A. (2009). Perceived investment in employee development, intrinsic motivation and work performance. Human Resource Management Journal, 19(3), 217-236. doi: 10.1111/j.1748-8583.2009.00103.x
Kuvaas, B., & Dysvik, A. (2011). Permanent employee investment and social exchange and psychological cooperative climate among temporary employees. Economic and Industrial Democracy, 32(2), 261-283. doi: http://eid.sagepub.com/archive/
Laufer, R. S., & Bengtson, V. L. (1974). Generations, aging, and social stratification: On the development of generational units. Journal of Social Issues, 30(3), 181-205. doi: 10.1111/j.1540-4560.1974.tb00733.x
Lawler III, E. E. (2011). Creating a new employment deal: Total rewards and the new workforce. Organizational Dynamics, 40(4), 302-309. doi: http://dx.doi.org/10.1016/j.orgdyn.2011.07.007
Lester, S. W., Standifer, R. L., Schultz, N. J., & Windsor, J. M. (2012). Actual versus perceived generational differences at work: An empirical examination. Journal of Leadership & Organizational Studies (Sage Publications Inc.), 19(3), 341-354. doi: 10.1177/1548051812442747
Lindquist, T. M. (2008). Recruiting the millennium generation: The new CPA. CPA Journal, 78(8), 56-59.
Long, R. J. (2001). Pay systems and organizational flexibility. Canadian Journal of Administrative Sciences (Canadian Journal of Administrative Sciences), 18(1), 25.
Long, W., Malhotra, D., & Murnighan, J. K. (2011). Economics education and greed. Academy of Management Learning & Education, 10(4), 643-660. doi: 10.5465/amle.2009.0185
Lopez, T. B., Hopkins, C. D., & Raymond, M. A. (2006). Reward preferences of salespeople: How do commissions rate? Journal of Personal Selling & Sales Management, 26(4), 381-390.
Lub, X., Bijvank, M. N., Bal, P. M., Blomme, R., & Schalk, R. (2012). Different or alike?: Exploring the psychological contract and commitment of different generations of hospitality workers. International Journal of Contemporary Hospitality Management, 24(4), 553-573. doi: doi:10.1108/09596111211226824
Lyons, S., & Kuron, L. (2014). Generational differences in the workplace: A review of the evidence and directions for future research. Journal of Organizational Behavior, 35, S139-S157. doi: 10.1002/job.1913
Lyons, S., Schweitzer, L., Ng, E., & Kuron, L. (2012). Comparing apples to apples: A qualitative investigation of career mobility patterns across four generations. Career Development International, 17(4), 333-357.
Martocchio, J. J. (2014). Strategic Compensation: A Human Resource Management Approach (7 ed.). London: Pearson Education Limited.
McClelland, D. (1961). The Achieving Society. Princeton, New Jersey: Van Nostrand.
Mercer. (2012). Generation Y: Born to bee wild? Retrieved from: http://www.mercer.de/content/dam/mercer/attachments/europe/Germany/Gen_Y_Born_to_bee_wild.pdf
Meyer, J. P., Stanley, D. J., Herscovitch, L., & Topolnytsky, L. (2002). Affective, continuance, and normative commitment to the organization: A meta-analysis of antecedents, correlates, and consequences. Journal of Vocational Behavior, 61(1), 20-52. doi: 10.1006/jvbe.2001.1842
Myers, D. (2012). Social Psychology (11 ed.). NY, USA: McGraw-Hill Higher Education.
Ng, T., & Feldman, D. (2012). Evaluating six common stereotypes about older workers with meta-analytical data. Personnel Psychology, 65(4), 821-858. doi: 10.1111/peps.12003
Nohe, C., & Sonntag, K. (2014). Work–family conflict, social support, and turnover intentions: A longitudinal study. Journal of Vocational Behavior, 85(1), 1-12. doi: 10.1016/j.jvb.2014.03.007
Olafsen, A. H., Halvari, H., Forest, J., & Deci, E. L. (2015). Show them the money? The role of pay, managerial need support, and justice in a self‐determination theory model of intrinsic work motivation. Scandinavian Journal of Psychology, 56(4), 447-457. doi: 10.1111/sjop.12211
Parry, E., & Urwin, P. (2011). Generational differences in work values: A review of theory and evidence. International Journal of Management Reviews, 13(1), 79-96. doi: 10.1111/j.1468-2370.2010.00285.x
Pierce, C. A., Block, R. A., & Aguinis, H. (2004). Cautionary note on reporting Eta-squared values from Multifactor ANOVA designs. Educational and Psychological Measurement, 64(6), 916-924. doi: 10.1177/0013164404264848
Pilcher, J. (1994). Mannheim’s sociology of generations: An undervalued legacy. British Journal of Sociology, 45(3), 481-495. doi: 10.2307/591659
Rao, K., & Pennington, J. (2013). Should the third reminder be sent? International Journal of Market Research, 55(5), 651-674. doi: 10.2501/IJMR-2013-056
Ryu, E., Couper, M. P., & Marans, R. W. (2006). Survey incentives: Cash vs. in-kind; face-to-face vs. mail; response rate vs. nonresponse error. International Journal of Public Opinion Research, 18(1), 89-106. doi: 10.1093/ijpor/edh089
Scott, W. R., & Davis, G. F. (2007). Organizations and Organizing: Rational, Natural, and Open System Perspectives. New Jersey: Pearson Education.
SHRM. (2012). 2012 Employee job satisfaction and engagement: How employees are dealing with uncertainty. Retrieved from: http://www.shrm.org/Research/SurveyFindings/Articles/Documents/SHRM-Employee-Job-Satisfaction-Engagement.pdf
Skattebetalerforeningen. (2014). Lønn og frynsegoder. Retrieved 21.10, 2014, from http://www.skatt.no/skatt/tema/frynsegoder/
Stevens, J. P. (1980). Power of the multivariate analysis of variance tests. Psychological Bulletin, 88(3), 728-737. doi: 10.1037/0033-2909.88.3.728
Stockard, J., Carpenter, G., & Kahle, L. R. (2014). Continuity and change in values in midlife: Testing the age stability hypothesis. Experimental Aging Research, 40(2), 224-244. doi: 10.1080/0361073x.2014.882215
Thomsen, D. J. (2012). From the trenches: Projecting the future reality of compensation and benefits. Compensation & Benefits Review, 44(2), 66-72. doi: 10.1177/0886368712450543
Towers Watson. (2012). Global workforce study. Retrieved from: http://www.towerswatson.com/assets/pdf/2012-Towers-Watson-Global-Workforce-Study.pdf
Trzesniewski, K. H., & Donnellan, M. B. (2010). Rethinking ‘Generation Me’: A study of cohort effects from 1976-2006. Perspectives on Psychological Science, 5(1), 58-75. doi: 10.1177/1745691609356789
Twenge, J. M. (2013a). The evidence for Generation Me and against Generation We. Emerging Adulthood, 1(1), 11-16.
Twenge, J. M. (2013b). Overwhelming evidence for Generation Me: A reply to Arnett. Emerging Adulthood, 1(1), 21-26. doi: 10.1177/2167696812468112
Twenge, J. M., Campbell, S. M., Hoffman, B. J., & Lance, C. E. (2010). Generational differences in work values: Leisure and extrinsic values increasing, social and intrinsic values decreasing. Journal of Management, 36(5), 1117-1142. doi: 10.1177/0149206309352246
Vincent, J. A. (2005). Understanding generations: Political economy and culture in an ageing society. British Journal of Sociology, 56(4), 579-599. doi: 10.1111/j.1468-4446.2005.00084.x
Von Bonsdorff, M. E. (2011). Age-related differences in reward preferences. International Journal of Human Resource Management, 22(6), 1262-1276. doi: 10.1080/09585192.2011.559098
White, P., & Breckenridge, R. S. (2014). Trade-offs, limitations, and promises of Big Data in social science research. Review of Policy Research, 31(4), 331-338. doi: 10.1111/ropr.12078
Witzel, M. (2012). History of Management Thought. Florence, KY, USA: Routledge.
Wong, M., Gardiner, E., Lang, W., & Coulon, L. (2008). Generational differences in personality and motivation. Journal of Managerial Psychology, 23(8), 878-890. doi: doi:10.1108/02683940810904376
WorldatWork. (2007). The WorldatWork Handbook of Compensation, Benefits and Total Rewards. New Jersey: John Wiley and Sons.
Wray-Lake, L., Syvertsen, A. K., Briddell, L., Osgood, D. W., & Flanagan, C. A. (2011). Exploring the changing meaning of work for American high school seniors from 1976 to 2005. Youth & Society, 43(3), 1110-1135.
Xavier, B. (2014). Shaping the future research agenda for compensation and benefits management: Some thoughts based on a stakeholder inquiry. Human Resource Management Review, 24(1), 31-40. doi: 10.1016/j.hrmr.2013.08.011
Yeaton, K. (2008). Recruiting and managing the ‘why?’, Generation: Gen Y. CPA Journal, 78(4), 68-72.