Development and validation of an instrument for assessing job demands arising from accelerated change: The intensification of job demands scale (IDS)

Abstract

Accelerated societal and organizational changes have placed new pressures on employees. Especially, service employees are exposed to intensified workloads, planning and decision-making, and learning demands. Despite the growing attention given to this intensification of job demands, a comprehensive measure is missing. In the present study, we developed the Intensification of Job Demands Scale (IDS) and validated it in four samples (N = 1363). Confirmatory factor analyses supported the differentiation into five subscales, namely work intensification, intensified job-related planning and decision-making demands, intensified career-related planning and decision-making demands, intensified knowledge-related learning demands, and intensified skill-related learning demands. This five-factor structure holds for both the German and the English versions of the instrument. Convergent and discriminant validity tests showed that the IDS subscales are moderately related to established measures of job demands, but at most have small correlations with negative affectivity. Providing support for the incremental validity, the IDS subscales were found to add to the prediction of burnout and job satisfaction beyond established job demands. Finally, the IDS subscales helped to identify employees who experienced changes in their work situation. In sum, the results indicate that the IDS is a valid and reliable measure to assess the intensification of job demands.

Keywords:

• Social acceleration
• Job demands
• Changing working conditions
• Intensification
• Validation
• Service industries

The world of work has changed considerably over the past few decades. There are continuing shifts away from manufacturing towards services (postindustrialization; Bell, 1973) and from previously closed national economies towards global ones (globalization; Giddens, 1990). Moreover, organizational structures have become increasingly flexible and new management practices have been implemented (Cascio, 2003). Besides these qualitative changes, there is also a quantitative one: With all of these changes, the speed at which transformations occur has accelerated (Rosa, 2003).

The accelerated economic, societal, and organizational changes inevitably alter what is expected of employees. Employees have found themselves working under a new dictum putting increasing emphasis on speed (e.g., Cascio, 2003), planning and decision-making (Pongratz & Voß, 2003), and knowledge (Loon & Casimir, 2008). For example, the percentage of employees working at high speed and under tight deadlines has increased since the 1990s (Green & McIntosh, 2001). Moreover, employees are confronted with more job autonomy (Wood, 2011) and higher learning demands (Loon & Casimir, 2008), since innovation and knowledge are recognized as sources of competitive advantage.

Despite the growing attention given to the intensification of job demands (e.g., Korunka & Kubicek, 2013), a comprehensive measure for assessing the extent to which employees experience such intensification is still missing. Although, there is a long tradition of assessing change with the use of longitudinal data, this approach may, despite its well-known advantages, have downsides, especially when it comes to the assessment of long-term transformations. Response shift (Schwartz & Sprangers, 1999) and regression effects (Popper, Spiel, & von Eye, 2012) may hamper the detection of changes using longitudinal data. Employees will, however, be able to retrospectively assess long-term changes in job demands if they are directly asked to do so (see e.g., Obschonka, Silbereisen, & Wasilewski, 2012). Therefore, an instrument that allows us to measure changes in job demands more directly will be developed.

The aim of the present article is twofold. In the first step, we shall integrate previous findings to identify and conceptualize job demands subject to intensification. In the second step, we shall develop an instrument for measuring these demands and validate the newly developed Intensification of Job Demands Scale (IDS) among four independent samples of service workers. Based on the construct validity assessment, we will show that the factor structure of the IDS subscales is the same for employees from different countries and with different languages. By analysing convergent and incremental validity indicators, we will show that the IDS subscales are associated with, but also add information to, existing scales when it comes to the predication of various outcomes. Moreover, we will show that the extent to which employees perceive intensified demands differs with regard to working conditions (use of information and communication technologies (ICTs) and new ways of working (NWW)). Finally, based on discriminant validity tests, we will distinguish the IDS subscales from negative affectivity.

The Intensification of Job Demands Due to Accelerated Change

As indicated by Rosa’s (2003) theory of social acceleration as well as by empirical evidence on changing working conditions (e.g., Green & McIntosh, 2001; Obschonka et al., 2012; Pongratz & Voß, 2003), there has been an intensification of job demands over the past few decades. In particular, workload, planning and decision-making demands (regarding one’s job and career), and learning demands (regarding one’s knowledge and skills) have intensified. We shall describe these demands and develop an instrument for assessing them.

Work intensification

To gain competitive advantage, companies try to shorten product cycles and speed up decision and production processes. Consequently, employers put more pressure on employees to work at higher speeds and under tighter deadlines (Green, 2004). It is therefore not surprising that cumulative evidence points towards an intensification of work over the past few decades (Burchell, Ladipo, & Wilkinson, 2002; Green & McIntosh, 2001; Paoli & Merllié, 2005). Using trend data, researchers have identified an intensification of work during the 1990s in Europe (Green & McIntosh, 2001; Paoli & Merllié, 2005) and in the United States (Handel, 2005). Likewise, Burchell et al. (2002), who asked their respondents directly whether they had experienced an increase in work effort, found an intensification of work for a majority of British employees. Combining previous research, we suggest that work intensification refers to the fact that the amount of effort an employee needs to invest during the working day increases. It is a multifaceted construct that is characterized by the need to work at increasing speed, perform different tasks simultaneously, or reduce idle time.

Intensified job-related planning and decision-making demands

Western societies are currently undergoing accelerated change, meaning that social norms and practices change more frequently (Rosa, 2003; Stone, 2010). Organizations and businesses react to this accelerated change by speeding up decision processes and implementing more flexible organizational structures (Cascio, 2003), leading not only to work intensification but also to a new logic of corporate labour control (Pongratz & Voß, 2003). The direct, “Tayloristic”, way of controlling employees’ labour seems to be outdated. Instead organizations reduce direct control and increase requirements for autonomous planning and decision-making. Consequently, employees’ job autonomy has increased (Wood, 2011). But as opposed to the traditionally positive view of autonomy (see, Karasek et al., 1998), employees do not just have the possibility of making decisions on their own; rather, they are forced to do so. They are increasingly expected to plan and structure their workday autonomously, to determine how to handle work tasks as well as to set and control work goals. Therefore, they experience intensified job-related planning and decision-making demands.

Intensified career-related planning and decision-making demands

Intensified planning and decision-making demands relate not only to one’s current job but also to career planning: employees are increasingly required to become strategic actors who continuously secure and prove their value for their current employer and increase their future employability outside the organization (Briscoe & Hall, 2006; Pongratz & Voß, 2003). This increasing need to autonomously plan one’s career is partly captured by concepts of the “boundaryless” and the “protean” career. While the concept of the boundaryless career draws attention to the fact that employees develop their careers across physical and psychological boundaries (Arthur & Rousseau, 1996), the concept of the protean career points out that careers are driven by the career actors and their values (Briscoe & Hall, 2006). Both career concepts take an affirmative view of current developments, highlighting the freedom, opportunities for self-directedness, and choices employees have in managing their careers. Conversely, Pongratz and Voß (2003) point out that self-directedness and individual responsibility for one’s career are also demands employees need to master successfully to remain attractive on the labour market. Following Pongratz and Voß’s notion of career development as a demand, we conceptualize intensified career-related planning and decision-making demands as the increasing requirement to autonomously plan and pursue one’s career inside and outside the current organization (Zeitz, Blau, & Fertig, 2009). This entails sustaining external networks as well as being aware of and open to career opportunities from current and possible future employers.

Intensified knowledge-related learning demands

Knowledge represents a significant economic resource and an integral feature of successful companies (Pyöriä, 2005), as it provides organizations with the opportunity to stay competitive in an increasingly complex world (Loon & Casimir, 2008). Although knowledge as a whole is growing in importance (Pyöriä, 2005), the half-life of knowledge is decreasing. For example, knowledge of technological devices needs to be updated at shorter time intervals (Obschonka et al., 2012) because rapid technological innovations render technical equipment outdated more quickly (Pitkäaho, Ryynänen, Partanen, & Vehviläinen-Julkunen, 2011). Consequently, developing one’s own knowledge as a means of sustaining employability represents a growing prerequisite for employees (Obschonka et al., 2012). Thus, knowledge-related learning demands are becoming more intense, meaning that the pressure to permanently refresh old and acquire new work-related knowledge increases.

Intensified skill-related learning demands

Job-related learning not only entails updating work-related knowledge but also means that new skills have to be acquired to perform one’s job effectively (Loon & Casimir, 2008). Because skill variety has increased over the past few decades (Wood, 2011), retaining one’s skill level has become more demanding. Employees are increasingly urged to continuously adjust their skills to new equipment, work practices, and regulations as well as to develop their competences in a self-directed and strategic manner (Korunka & Kubicek, 2013; Obschonka et al., 2012). They are therefore confronted with intensified skill-related learning demands.

Job- and Person-Related Factors Associated with Job Demands Arising from Accelerated Change

In the following, we shall build arguments on how traditional job characteristics and personality traits are related to the five job demands affected by intensification.

Relation to job-related factors

We propose that the perception of intensified workload, planning and decision-making demands, and learning demands should be moderately related to their corresponding traditional job characteristics.

First, because work intensification manifests itself in the need to work faster, reduce idle time, and multitask, it should be related to traditional measures of time pressure. Nevertheless, high levels of time pressure are not necessarily the result of an intensification process but may have stabilized at consistently high levels. Therefore, a medium-sized positive association is expected between the perception of work intensification and time pressure. In line with this assumption, Franke (in press) found a moderate intercorrelation between a traditional measure of work intensity and a single-item measure of work intensification.

Hypothesis 1:

The perception of work intensification is positively related to time pressure.

Second, intensified job-related planning and decision-making demands should be related to job control (also termed autonomy). Initially, job control was conceptualized as the extent to which individuals can freely choose how to carry out their work tasks (e.g., Hackman & Oldham, 1976). Recently, this broad conceptualization of job control has been split up into distinct aspects. Morgeson and Humphrey (2006) distinguish between discretion in scheduling work tasks, making task-related decisions, and selecting work methods as components of autonomy. We expect moderate positive associations between general measures as well as distinct aspects of job control and intensified job-related planning and decision-making demands based on two reasons: (1) In order to experience planning and decision-making as an increasing demand, employees need to have at least some control in their job. (2) The two concepts should not, however, be perfectly related, because job control only gives the possibility of taking over control, whereas employees facing intensified job-related planning and decision-making demands are forced to plan and make decisions.

Hypotheses 2a–d:

Perceived intensified job-related planning and decision-making demands are positively related to job control (a), work scheduling (b), decision-making (c), and work methods autonomy (d) as distinct aspects of job control.

Third, intensified career-related planning and decision-making demands reflect the extent to which employees perceive an increasing need to plan and pursue their careers independently of their current organization. As such, they include the increasing demand for securing one’s future employability. We therefore expect that the perception of intensified career-related planning and decision-making demands will be positively related to requirements for self-directed career development (Höge, 2011). These requirements are described as the need to take personal responsibility for one’s career development and to strategically develop one’s work capacities.

Hypothesis 3:

Perceived intensified career-related planning and decision-making demands are positively related to requirements for self-directed career development.

Fourth, we expect that intensified knowledge- and skill-related learning demands will be related to the extent to which jobs put pressure on employees to learn new things. This pressure may—among other aspects—be determined by the number of tasks associated with a job (Loon & Casimir, 2008). Task variety (e.g., topic-related variety), in contrast to specialization, is associated with increased learning rates, since it offers the possibility of developing a complex and abstract knowledge and probing one’s knowledge under diverse conditions (Schilling, Vidal, Ployhart, & Marangon, 2003). Thus, we propose that the greater the variety of tasks employees have to perform, the more job demands for learning they face (Loon & Casimir, 2008) and the higher the chance that they will need to update their knowledge and skills.

In addition to task variety, employers’ expectations regarding training participation may also determine the extent to which employees experience an intensification of learning demands (Höge, 2011). Employees whose employers expect them to attend training on and off the job should experience a stronger intensification of learning demands.

Hypotheses 4a–d:

Perceived intensified knowledge-related learning demands (a, c) and intensified skill-related learning demands (b, d) are positively related to task variety and requirements for self-directed learning.

Relation to person-related factors

According to studies on the role of negative affectivity in job stress research, employees’ personal dispositions have substantive effects on their perception of and susceptibility to job demands (e.g., Cavanaugh, Boswell, Roehling, & Boudreau, 2000; Spector, Zapf, Chen, & Frese, 2000). Despite these effects of individual characteristics, it is still important to distinguish the perception of intensified demands from employees’ dispositions, such as negative affectivity. Because the perception of intensified job demands and negative affectivity represent different constructs, they should not be, or should only marginally be, related (Hinkin, 1998). Such marginal relations are, according to Cohen’s (1988) classification of correlation magnitude, smaller than .30.

Hypotheses 5a–e:

Negative affectivity will at most have a small correlation with the perception of work intensification (a), intensified job-related decision-making demands (b), intensified career-related planning and decision-making demands (c), intensified knowledge-related learning demands (d), and intensified skill-related learning demands (e).

Differences in the Intensification of Demands Between Employees

There are not many studies that have assessed increasing demands and linked them with verifiable changes. However, one study (Green & McIntosh, 2001) showed that work intensity increased more strongly for employees who used computers frequently than for employees who used computers rarely. Moreover, NWW (Baarne, Houtkamp, & Knotter, 2010)—conceptualized as the combination of temporal and spatial flexibility and the use of ICTs—were shown to contribute to autonomy and work intensification. On the one hand, NWW are enablers of flexibility and therefore increase employees’ control over when, where, and for how long they work (Kelliher & Anderson, 2008). On the other hand, NWW also have a downside. Constant connection to work implies that work never stops and therefore increases work intensity (Kelliher & Anderson, 2008).

Hypotheses 6a–c:

The use of ICTs is positively related to perceived work intensification (a). The extent to which employees perform NWW is positively related to the perception of work intensification (b) and intensified job-related planning and decision-making demands (c).

Another verifiable indicator that may be linked to increasing demands is changes in employees’ work situation. Such changes may involve promotions, changes in the types of tasks employees need to perform on the job, or organizational restructuring. As changes in the work situation render acquired work-related knowledge and skills partly obsolete, they place pressure on employees to learn new things (Loon & Casimir, 2008). Moreover, changes in the work situation challenge employees’ routinized and learned ways of performing a task (Hacker, 2003). Consequently, employees need to actively decide how to approach and meet the demands associated with the new tasks and thus need to use conscious effort in order to attain their work goals. Such knowledge-based ways of action regulation (Hacker, 2003) may therefore entail increased planning and decision-making demands.

Hypotheses 7a–c:

Employees who experience changes in their work situation perceive higher levels of intensified job-related planning and decision-making demands (a), intensified knowledge-related learning demands (b), and intensified skill-related learning demands (c) compared to employees who do not experience changes in their work situation.

Consequences of Job Demands Arising from Accelerated Change

Burnout as a potential consequence

As job demands arising from accelerated change are characterized by the intensification of workload, planning and decision-making demands, and learning demands, meeting these demands requires increased effort. Sustaining such physical or cognitive effort is associated with certain costs (Van den Broeck, De Cuyper, De Witte, & Vansteenkiste, 2010). According to the health impairment process of the job demands-resources (JD-R) model (Bakker & Demerouti, 2007), high demands trigger an energetic process of “wearing out” in which demands deplete employees’ mental and physical resources, leading—at least in the long run—to burnout and health impairments. Following the predications of the JD-R model (Bakker & Demerouti, 2007), we hypothesize that perceptions of the intensification of job demands are associated with emotional exhaustion and cynicism. These positive associations with emotional exhaustion and cynicism should be present even when controlling for traditional job demands.

Hypotheses 8a–e:

The perception of work intensification (a), intensified job-related planning and decision-making demands (b), intensified career-related planning and decision-making demands (c), intensified knowledge-related learning demands (d), and intensified skill-related learning demands (e) add to the prediction of emotional exhaustion and cynicism, over and above traditional job demands.

Job satisfaction as a potential consequence

Despite their energy-depleting effect, some of the intensified demands may also affect positive work-related outcomes, such as job satisfaction. For instance, intensified job-related planning and decision-making and learning demands may fulfil basic human needs for autonomy and competence (Deci & Ryan, 2000) and may therefore support personal goals and foster personal development. According to self-determination theory (Deci & Ryan, 2000), enabling employees to feel autonomous and competent by offering jobs that include planning and decision-making as well as learning is an important prerequisite for intrinsic motivation. This in turn is associated with positive work attitudes such as job satisfaction (e.g., Cavanaugh et al., 2000; Gagné & Deci, 2005). Work intensification and intensified career-related planning and decision-making demands, on the other hand, do not offer such an opportunity for satisfying employees’ needs for autonomy and competence. Work intensification requires additional effort and may press employees to trade quality for quantity, which may frustrate employees’ need for competence and may consequently diminish feelings of intrinsic motivation (Amabile, DeJong, & Lepper, 1976). Intensified career-related planning and decision-making demands are associated with uncertainty regarding one’s future job prospects and the fear of becoming unattractive to the labour market (Höge, 2011). Therefore, this demand is also expected to impair employees’ job satisfaction.

Hypotheses 9a–e:

The perception of work intensification (a) and intensified career-related planning and decision-making demands (c) are negatively related to job satisfaction, whereas the perception of intensified job-related planning and decision-making demands (b), intensified knowledge-related learning demands (d), and intensified skill-related learning demands (e) are positively related to job satisfaction, after controlling for traditional job demands.

Developing the Ids

In developing the IDS, we first consulted existing instruments to find items measuring intensified demands. Given the novelty of the constructs, we had to develop new items, with the exception of one item measuring work concentration (adapted from the ERI Questionnaire; Siegrist, Wege, Pühlhofer, & Wahrendorf, 2009). We sought to create items that reflect the construct definitions as outlined earlier and capture the conceptual breadth of each demand.

In addition to these conceptual considerations, the following principles guided item formulation. First, we used a direct measure of change (Burchell et al., 2002). That is, the item wording reflects the intensification of the respective demands and respondents are asked to indicate whether they have experienced an increase in the respective demand. Following Obschonka et al. (2012), the time frame to be considered when assessing intensified demands was set at 5 years. Therefore, each statement started with the phrase “In the last five years…”. Second, we restrict the item content to changes observable in service-sector jobs. Third, we decided to formulate the items impersonally in order to measure perceived changes in job demands rather than generally negative personal dispositions. We therefore indicated in the introduction that employees should assess changes that have occurred in their work. In addition to framing the questions in this way, we formulated the items in a descriptive rather than in an evaluative manner (see e.g., Spector et al., 2000) and refrained from addressing employees personally in order to avoid questions being mistakenly interpreted as asking for the employee’s abilities to meet his/her job demands. Fourth, because the inclusion of negatively and positively worded items into one scale was shown to produce artificial factor solutions (Dalal & Carter, 2014), all items were positively worded, so that higher levels indicate a stronger intensification. The items were grouped by content and the response format was a 5-point scale ranging from 1 = not at all to 5 = completely.

After generating a pool of potential items (k = 44), we used various probing methods to assess the intelligibility and appropriateness of the item wording as well as the content validity. The think-aloud technique, which is respondent-driven (Collins, 2003), was used to determine the thoughts that led participants to their responses (Prüfer & Rexroth, 1996). This technique helped to assess whether participants indeed fully understood what was meant by the intensified demands and whether they understood all items in the intended way. Furthermore, an interviewer-driven technique was used to explore retrieval, judgement, comprehension, and response processes used by respondents (Collins, 2003). Focus groups and expert interviews with human resource managers (Presser & Blair, 1994) helped to assess the content validity. Based on the information gathered from the probing methods and statistical analysis of the item pool, we adapted the items linguistically and reduced them to a 19-item scale.

Using this final version of the IDS, we investigated the validity of the newly developed instrument in four independent samples. We aimed at assessing the instruments’ (1) factor structure using confirmatory factor analyses, (2) convergent validity using correlations with job-related factors (Hypotheses 1–4), (3) discriminant validity using correlations with negative affectivity (Hypothesis 5), (4) construct validity using mean differences between groups of employees (Hypotheses 6–7), and (5) incremental validity using relations with outcomes (Hypotheses 8–9). Because different samples were used for different validity tests, Table 1 provides an overview of the analyses conducted in each sample.

TABLE 1 Demographic characteristics of the participants and constructs measured and analyses conducted in each sample

	Sample 1	Sample 2	Sample 3	Sample 4
	Varied sample of service employees	Varied sample of service employees	Employees of the sales departments of an Austrian firm with international subsidiaries	Office workers of a governmental institution
Country/region	Germany	Austria	Austria, Germany, Russia, Commonwealth of Independent States, Western Europe, Asia Pacific, Middle East	Austria
Language	German	German	German, English	German
N (total)	491	185	193 (response rate: 38.1%)	628 (response rate: 55.4%)
N (after quality check)	406	163	193	601
Age	42.71 (SD = 13.44)	37.88 (SD = 9.87)	35.73 (SD = 10.31)	42.10 (SD = 8.84)
Organizational tenure	12.16 (SD = 11.08)	9.06 (SD = 9.53)	7.69 (SD = 8.07)	13.56 (SD = 10.10)
Weekly working hours	43.45 (SD = 8.98)	42.81 (SD = 7.47)	43.89 (SD = 9.39)	41.70 (SD = 7.42)
Male participants	47.5%	50.9%	44%	59.7%
Tertiary education	41.6%	38.7%	39.4%	61.7%
Factor structure	CFA, reliabilities	CFA, reliabilities
Invariance tests	Across samples 1 and 2	Across samples 1 and 2	Across language versions
Relation to job-related factors: convergent validity	Time pressure (H1), job control (H2a), RSC (H3), RSL (H4c–4d)	Time pressure (H1), work scheduling—(H2b), decision-making—(H2c), and work methods autonomy (H2d), RSC (H3), RSL (H4c–d), task variety (H4a–4b)
Relation to person-related factors: discriminant validity	Negative affectivity (H5)	Neuroticism (H5)
Differences in the intensification of demands between employees: construct validity			Use of ICTs (H6a), NWW (H6b–c)	Changes in work situation (H7a–7d)
Relation to outcomes: incremental validity	Emotional exhaustion, cynicism (H8a–8e)	Emotional exhaustion, job satisfaction (H9a–9e)

SD = standard deviation; CFA = confirmatory factor analysis; RSC = requirements for self-directed career development; RSL = requirements for self-directed learning; NWW = new ways of working; H = hypothesis; ICTs = information and communication technologies.

Method

Sample and procedure

Data were gathered from four independent samples. Two of the four samples consisted of service workers who were accessed via an ISO-certified (ISO 26362) German online panel (www.respondi.com). This panel recruits its participants via offline and online methods. It ensures high quality through minimizing participation frequency, focusing on intrinsic motivation instead of financial dependency, and conducting continuous controls. Sample 1 was intended to be similar to the general population of workers in the German service sector with regard to gender and working hours. The other two samples were drawn from two Austrian organizations: one from the sales departments of a production firm with headquarters in Austria and several international subsidiaries, the other from an Austrian governmental institution. For all four samples missing values were imputed using the expectation-maximization algorithm (EMA) as implemented in NORM 2.03 (Schafer, 2009). Demographic characteristics of the samples are presented in Table 1.

Measures

Time pressure was measured with a 5-item scale using the instrument for stress-related job analysis (ISTA, Semmer, Zapf, & Dunckel, 1998). The response format ranged from 1 = very rarely/never to 5 = very often.

Autonomy, in sample 1, was measured with a 5-item job control scale of the ISTA (Semmer et al., 1998). The response format ranged from 1 = very little to 5 = very much. In sample 2, we used the German version of the Work Design Questionnaire (WDQ; Stegmann et al., 2010) to differentiate between work scheduling, decision-making, and work methods autonomy. Each subscale of the WDQ consisted of three items that had to be answered on scales ranging from 1 = strongly disagree to 5 = strongly agree.

Requirements for self-directed career development (e.g., “In my work, my employer expects me to take personal responsibility for my career development”) and requirements for self-directed learning (e.g., “In my work, my employer expects me to attend in-service training courses”) were measured with two items each using the Flexibility Requirements Scale (Höge, 2011). The response format ranged from 1 = strongly disagree to 6 = strongly agree.

Task variety was measured with four items using the WDQ (Stegmann et al., 2010). The response format ranged from 1 = strongly disagree to 5 = strongly agree.

NWW were measured with their three key characteristics: temporal and spatial flexibility and use of ICTs (Baarne et al., 2010). As in previous research (Shockley & Allen, 2007), flexibility in terms of time was assessed with the following item: “I can schedule my working hours flexibly”. The response format ranged from 1 = not at all to 5 = completely. Based on previous research (Ten Brummelhuis, Bakker, Hetland, & Keulemans, 2012), flexibility in terms of place and the use of ICTs were also measured with single items (i.e., “How often do you telework?”; “How often do you use a smartphone, tablet PC or personal digital assistant for your work?”). The response format ranged from 1 = very rarely/never to 5 = very often. All three items were summed to form an index for NWW (1 = no use of NWW up to 15 = frequent use of NWW).

Change in work situation was measured with one item: “Did your work situation change during the last five years (promotion, new work activities, new employer etc.)?” The response format ranged from 1 = yes to 2 = no.

In sample 1, negative affectivity was measured with the German translation of the Positive and Negative Affect Scales (PANAS; Krohne, Egloff, Kohlmann, & Tausch, 1996). Answers were given on a scale ranging from 1 = not at all to 5 = very much. In sample 2, we used the neuroticism subscale of the German language translation of the Big Five Short Inventory (BFI-S; Gerlitz & Schupp, 2005). The response format for the three neuroticism-items ranged from 1 = does not apply at all to 7 = applies perfectly.

Burnout was measured with the German translation of the Maslach Burnout Inventory (MBI-D; Büssing & Perrar, 1992). We focused on the two core dimensions of burnout, emotional exhaustion and cynicism (Shirom, 2003), both consisting of five items with a response format ranging from 1 = never to 6 = very often.

Job satisfaction was measured with a single item (i.e., “All in all, how satisfied would you say you are with your job?”; Wanous, Reichers, & Hudy, 1997). Answers to the question were provided on a seven-step Kunin scale.

Results

Factor structure

First, we analysed the factor structure of the IDS by comparing different models using confirmatory factor analysis (CFA) as implemented in AMOS 19. The model tests were based on maximum likelihood estimation. To assess the quality of the models we used chi-square statistics, the comparative fit index (CFI), the Tucker–Lewis index (TLI), and the root mean square error of approximation (RMSEA) with its corresponding confidence interval. CFI and TLI values close to .95 and above and RMSEA values of .06 or lower indicate a good fit to the data (Hu & Bentler, 1999). Chi-square difference scores (Δχ²) and the Akaike information criterion (AIC) were used to compare the models. The hypothesized five-factor model (work intensification, intensified job-related planning and decision-making demands, intensified career-related planning and decision-making demands, intensified skill-related learning demands, and intensified knowledge-related learning demands) was compared with a four-factor model (which had one factor for intensified learning demands in general), a three-factor model (which collapsed the subscales of intensified learning as well as intensified planning and decision-making into two factors), a one-factor two higher-order-factor model (which had one factor for work intensification and two higher order factors, one for learning demands and one for planning and decision-making demands), and with a second-order model (a model with a single overarching latent factor with five first-order factors that were identical to those in the five-factor model). The model comparisons are presented in Table 2.

TABLE 2 Fit indices for measurement models in sample 1 and sample 2

Model	Sample	χ^2	df	CFI	TLI	RMSEA [90% CI]	AIC	Model comparison	Δχ^2	Δdf	p
Five-factor model (WI, IJP, ICP, IKL, ISL)	Sample 1	399.57	142	.96	.95	.067 [.059–.075]	495.57
	Sample 2	291.97	142	.92	.93	.081 [.068–.094]	387.97
Four-factor model (WI, IJP, ICP, IL)	Sample 1	524.53	146	.93	.92	.080 [.073–.087]	612.53	M2–M1	124.86	4	.001
	Sample 2	363.36	146	.89	.89	.096 [.084–.108]	451.36	M2–M1	71.39	4	.001
Three-factor model (WI, IP, IL)	Sample 1	764.91	149	.89	.88	.101 [.094–.108]	846.91	M3–M1	347.34	7	.001
	Sample 2	435.60	149	.86	.83	.109 [.097–.121]	517.60	M3–M1	143.63	7	.001
One-factor two-higher-order-factor model (WI, IP—IJP&ICP, IL—IKL&ISL)	Sample 1	420.43	145	.95	.94	.068 [.061–076]	510.43	M4–M1	20.86	3	.001
	Sample 2	293.59	145	.63	.91	.080 [.066–.093]	383.59	M4–M1	1.62	3	ns
Second-order model (with five first-order factors)	Sample 1	453.51	147	.95	.94	.072 [.064–.079]	539.51	M5-M1	53.94	5	.001
	Sample 2	314.02	147	.92	.90	.084 [.071–.097]	400.02	M5-M1	22.05	5	.001
Five-factor model: correlations between the latent factors		WI & IJP	WI & ICP	WI & IKL	WI & ISL	IJP & ICP	IJP & IKL	IJP & ISL	ICP & IKL	ICP & ISL	IKL & ISL
	Sample 1	.43	.52	.50	.57	.67	.66	.57	.70	.67	.89
	Sample 2	.52	.54	.52	.53	.66	.62	.57	.57	.56	.83

χ² = chi-square fit index; df = degrees of freedom; CFI = comparative fit index, TLI = Tucker–Lewis index; RMSEA = root mean square of approximation; CI = confidence interval; AIC = Akaike information criterion; Δχ² = chi-square difference; WI = work intensification; IP = intensified planning and decision-making demands; IJP = intensified job-related planning and decision-making demands; ICP = intensified career-related planning and decision-making demands; IL = intensified learning demands; IKL = intensified knowledge-related learning demands; ISL = intensified skill-related learning demands. Sample 1 (n = 406); Sample 2 (n = 163).

For sample 1, all models but the three-factor model yielded good fit indices. Thus, no further modifications, such as correlations between errors, were necessary. Our proposed five-factor model yielded the best fit indices. Only the RMSEA was slightly above the upper limit, but still less than .08 and therefore acceptable (Browne & Cudeck, 1993). Comparing the models, the chi-square difference test and the AIC value showed that the five-factor model was superior to all other models. The items, means, standard deviations (SD), corrected item-total correlations, and standardized loadings for the five-factor model are presented in Table 3.

TABLE 3 Items, means (M), standard deviations (SD), corrected item-total correlations (CITC), and standardized loadings based on the confirmatory factor analysis of the IDS in sample 1

				Latent dimensions
Item wording	M	SD	CITC	1	2	3	4	5
1. Work intensification
… it is increasingly rare to have enough time for work tasks	2.97	1.36	.86	.91
… it is increasingly harder to take time for breaks	2.66	1.39	.80	.84
… the time between the more intense work phases has decreased	2.94	1.42	.78	.85
… one has more often to do two or three things at once (such as eating lunch, writing emails, and talking on the phone)	2.98	1.44	.75	.80
… ever more work has to be completed by fewer and fewer employees	3.41	1.48	.63	.67
2. Intensified job-related planning and decision-making demands
… it is increasingly becoming necessary to plan the workflow (activities, appointments, breaks, etc.) by oneself	3.65	1.31	.79		.85
… one increasingly has to determine by oneself how to do the work	3.41	1.34	.81		.87
… one increasingly has to determine the sequence of activities by oneself	3.46	1.34	.78		.84
… more often decisions have to be made without consultation with supervisors	3.15	1.37	.74		.78
… one increasingly has to check independently whether the work goals have been reached	3.02	1.38	.67		.71
3. Intensified career-related planning and decision-making demands
… one is increasingly demanded to maintain one’s attractiveness for the job market (e.g., through advanced education, networking)	3.12	1.41	.69			.79
… one’s own professional development increasingly requires to keep other alternatives open	2.80	1.27	.68			.79
… one increasingly has to plan one’s professional career independently	3.24	1.45	.65			.75
4. Intensified knowledge-related learning demands
… one has to acquire new expertise for the job more often	3.21	1.25	.84				.88
… one increasingly has to acquire new knowledge to handle job tasks	3.23	1.20	.81				.93
… one has to update one’s knowledge level more frequently	3.48	1.24	.58				.78
5. Intensified skill-related learning demands
… one increasingly has to familiarize oneself with new work processes	3.06	1.22	.81					.91
… one increasingly has to get used to new workflows	3.00	1.23	.80					.92
… one has to use new work equipment (devices, programs, etc.) more often	3.02	1.23	.63					.67

N = 406.

To cross-validate the five-factor structure, we used data from sample 2 (see Table 2 for results). The chi-square difference tests revealed that the five-factor model was not significantly better than the one-factor two-higher-order-factor model. Although this higher-order model shows that job-related planning and decision-making demands and career-related planning and decision-making demands as well as knowledge-related learning demands and skill-related learning demands belong to overarching factors, namely intensified planning and decision-making demands and intensified learning demands, it still demonstrates that it is advisable to distinguish between the five IDS subscales. Thus, data from both samples lend support to the hypothesized factor structure of the IDS.

Invariance tests

In a final step we assessed the metric and the structural invariance (Byrne, Shavelson, & Muthén, 1989) of the German and the Austrian samples (samples 1 and 2) and of the German and English versions of the IDS (sample from the international firm). We started by comparing the Austrian and the German subsamples. In order to assess the metric invariance of the IDS, we compared a model with the constraint that factor loadings across the two subsamples had to be equal in size with a model that allowed the factor loadings to be freely estimated. The model comparison led to a non-significant chi-square difference test (Δχ² (14) = 11.45, p = .650). Then we assessed structural invariance by setting the factor variances and the factor covariances to be equal across the samples. These constraints did not result in a significant deterioration of the chi-square statistic (Δχ² (29) = 23.59, p = .749). We then repeated the analyses using data from employees of the international firm in order to compare the German and the English language versions of the IDS. Again, the test for metric invariance led to a non-significant chi-square difference test (Δχ² (14) = 16.30, p = .296). However, the test for structural invariance revealed a significant difference. As recommended (Byrne et al., 1989), we therefore proceeded with testing for partial structural invariance. After relaxing the constraints of all correlations between intensified skill-related learning demands and intensified knowledge-related learning demands, we found support for partial structural invariance (Δχ² (23) = 23.96, p = .406). Finally, we ran the metric invariance test for all three samples together. We did not conduct the structural invariance test using data from all three samples, since we only found support for partial structural invariance in sample 3. The model comparison for testing metric invariance across the three samples led to a non-significant chi-square difference test (Δχ² (14) = 14.95, p = .382). These analyses showed that factor loadings are invariant across the Austrian and the German subsamples, across the German and English versions of the IDS and also across all three samples together. Moreover, the IDS measure is structurally invariant across the German and the Austrian samples and partially structurally invariant across the English and German language versions.

Reliabilities

The internal consistencies of the five subscales of the IDS (see Table 4) were all satisfactory (Nunnally & Bernstein, 1994). The corrected item-total correlations were all above .58 and thus acceptable (see Table 3).

TABLE 4 Intercorrelations among and Cronbach’s alphas of the IDS subscales as well as variables assessed in samples 1 and 2

		2	3	4	5	6	7	8	9	10	11	12	13	14	15
2	WI	(.91/.87)	.44***	.46***	.47***	.49***	–	.58***	.21**	.15	.27***	−.15*	−.15	−.17*	.08
3	IJP	.41***	(.90/.87)	.59***	.57***	.48***	–	.20**	.15	.21**	−.01	.20*	.19*	.20*	.27**
4	ICP	.44***	.60***	(.82/.79)	.50***	.47***	–	.24**	.35**	.23**	.13	.04	−.07	−.04	.06
5	IKL	.52***	.54***	.58***	(.90/.90)	.76***	–	.19*	.19*	.47***	.09	.06	.10	.16*	.30***
6	ISL	.45***	.62***	.61***	.79***	(.87/.87)	–	.17*	.15	.28***	.05	−.01	−.02	.04	.21**
7	Job control	−.05	.38***	.17**	.10*	.16**	(.89/–)	–	–	–	–	–	–	–	–
8	Time pressure	.58***	.19***	.21***	.27***	.23***	−.07	(.88/.87)	.17*	.11	.22**	−.19*	−.17*	−.23**	.17*
9	RSC	.06	.29***	.31***	.22***	.34***	.44***	.08	(.92/.89)	.39***	.04	.12	.10	.09	.14
10	RSL	.12*	.32***	.31***	.31***	.42***	.38***	.12*	−54***	(.68/.81)	.04	.30***	.36***	.34***	.48***
11	Negative affectivity	.23***	−.05	.10*	.07	.03	−.29***	.24**	−.14**	−.11*	(.88/.69)	−.24**	−.27***	−.36***	−.20*
12	WS autonomy	–	–	–	–	–	–	–	–	–	–	(–/.88)	.75***	.76***	.42***
13	DM autonomy	–	–	–	–	–	–	–	–	–	–	–	(– /.94)	.87***	.57***
14	WM autonomy	–	–	–	–	–	–	–	–	–	–	–	–	(–/.92)	.58***
15	Task variety	–	–	–	–	–	–	–	–	–	–	–	–	–	(–/.90)

Intercorrelations for sample 1 (n = 406) are presented below the diagonal and intercorrelations for sample 2 (n = 163) are presented above the diagonal. Correlations that were included in hypotheses testing are in boldface. WI = work intensification; IJP = intensified job-related planning and decision-making demands; ICP = intensified career-related planning and decision-making demands; IKL = intensified knowledge-related learning demands; ISL = intensified skill-related learning demands; WS autonomy = work scheduling autonomy; DM autonomy = decision-making autonomy; WM autonomy = work methods autonomy; RSC = requirements for self-directed career development; RSL = requirements for self-directed learning.

*p < .05, **p < .01, ***p < .001.

Relation to job-related factors: convergent validity

To establish the construct validity of the IDS we correlated the subscales of the IDS with existing and published scales measuring similar constructs (for results see Table 4).

Perceived work intensification correlated positively with time pressure in the German and the Austrian service worker samples (samples 1 and 2), supporting Hypothesis 1. Perceived intensified job-related planning and decision-making demands had significant positive associations with job control and with work scheduling, decision-making, and work methods autonomy, supporting Hypotheses 2a–d. Perceived intensified career-related planning and decision-making demands had moderate positive associations with requirements for self-directed career development in the German and the Austrian service worker samples, supporting Hypothesis 3. Perceived intensified knowledge-related and perceived intensified skill-related learning demands showed significant positive associations with task variety as well as with requirements for self-directed learning, supporting Hypotheses 4a–d.

Relation to person-related factors: discriminant validity

Hypotheses 5a–e were tested using data from the German and the Austrian service worker samples (see Table 4 for results). Across all IDS subscales, only work intensification (in both samples) and intensified career-related autonomy demands (only in the German sample) were significantly related to negative affectivity. These correlations were, however, low in magnitude (Cohen, 1988). Based on the small amount of significant correlations, the low magnitude of the significant correlations, and the congruence of the non-significant correlations in both studies, we concluded that there were only limited and at most small relations between the IDS subscales and negative affectivity. Thus, hypotheses 5a–e were supported.

Differences in the intensification of demands between employees: construct validity

Hypotheses 6a–c were tested with the international sample (sample 3). The frequency of ICT use was positively related to the perception of work intensification (r = .24, p = .001). Moreover, perceived work intensification (r = .23, p = .001) and perceived intensified job-related planning and decision-making demands (r = .23, p = .001) were positively related to the extent to which employees perform NWW. Taken together, the results support hypotheses 6a–c.

Hypotheses 7a–c were tested using data from sample 4. Employees who experienced changes in their work situation (N = 292) perceived higher levels of intensified knowledge-related learning demands, t(599) = 2.70, p = .007 (M_{change group} = 3.87, SD = 0.87; M_{no change group} = 3.68, SD = 0.90) and job-related planning and decision-making demands, t(599) = 2.09, p = .037 (M_{change group} = 3.19, SD = 0.95; M_{no change group} = 3.03, SD = 0.98). But their perception of intensified skill-related learning demands, t(599) = 0.86, p = .390, did not differ from employees who experienced no such changes. Thus, hypotheses 7a and b were supported, whereas hypothesis 7c was not supported.

Relation to outcomes: incremental validity

The incremental validity of the IDS subscales was tested using multiple regression analyses (for results see Table 5). In accordance with Hypotheses 8a–d, work intensification, intensified job-related planning and decision-making demands, intensified career-related planning and decision-making demands, and intensified knowledge-related learning demands correlated positively with emotional exhaustion and cynicism even when taking traditional demands into account. Hypothesis 8e was only partially supported. After controlling for traditional demands, intensified skill-related learning demands were positively associated with emotional exhaustion but not with cynicism.

TABLE 5 Relations to outcomes in samples 1 and 2

	Dependent variables
	Sample 1 (German service workers)				Sample 2 (Austrian service workers)
	Emotional exhaustion		Cynicism		Emotional exhaustion		Job satisfaction
Independent variable	Step 1	Step 2	Step 1	Step 2	Step 1	Step 2	Step 1	Step 2
Time pressure	.483***	.325***	.276***	.112^†	.580***	.401***	−.320***	−.161^†
Work intensification		.274***		.282***		.310***		−.276**
ΔR^2		.050***		.053***		.064***		.051**
Job control	−.289***	−.336***	−.385***	−.433***
Work scheduling autonomy					.017	−.002	−.153	−.150
Decision-making autonomy					.031	.029	.188	.189
Work methods autonomy					−.449**	−.469**	.520***	.523***
Intensified job-related planning and decision-making demands		.123*		.126*		.181*		−.030
ΔR^2		.013*		.014*		.031*		.001
Requirements for self-directed career development	−.090^†	−.125*	−.209***	−.276***	.080	−.022	−.082	−.024
Intensified career-related planning and decision-making demands		.113*		.217***		.295***		−.165*
ΔR^2		.011*		.043***		.076***		.024*
Requirements for self-directed learning	−.057	−.115*	−.202***	−.239***	.069	.017	.033	.055
Task variety					−.189*	−.211*	.379**	.388**
Intensified skill-related learning demands		.140*		.089^†		.221**		−.095
ΔR^2		.016*		.007^†		.044**		.008
Requirements for self-directed learning	−.057	−.110*	−.202***	−.237***	.069	−.031	.033	.088
Task variety					−.189*	−.213*	.379**	.392**
Intensified knowledge-related learning demands		.172**		.113*		.239**		−.132
ΔR^2		.027**		.011*		.044**		.014

Sample 1: n = 406; Sample 2: n = 163.

^†p < .10, *p < .05, **p < .01, ***p < .001.

In support of Hypotheses 9a and c, work intensification and intensified career-related planning and decision-making demands showed significant negative associations with job satisfaction after controlling for time pressure and requirements for self-directed career development, respectively (see Table 5). However, no such associations were found for the other IDS subscales. Thus Hypotheses 9b, d, and e were not supported.

Discussion

In this research, we developed a measure capturing the intensification of job demands. Based on a comprehensive literature review, we identified five job demands that have been subject to significant intensification over recent decades and developed the IDS. The scale comprises five subscales, namely work intensification, intensified job-related planning and decision-making demands, intensified career-related planning and decision-making demands, intensified knowledge-related learning demands, and intensified skill-related learning demands. The validity of the IDS was assessed using data from four independent samples. CFA revealed that the hypothesized five-factor structure of the IDS was supported in a large sample of German service workers. In a sample of Austrian service workers, which was used to cross-validate this result, the model with two higher-order factors, one for intensified learning and one for intensified planning and decision-making demands, was not significantly worse than the five-factor model. Thus, overarching factors seem to represent the theoretically related subscales job- and career-related planning and decision-making demands on the one hand and skill- and knowledge-related learning demands on the other hand. Yet the existence of higher order factors does not render the five-factor structure of the IDS invalid. Rather, the results suggest that it is advisable to distinguish among the five subscales, even if these scales partially belong to overarching dimensions. Although the model with one general overarching (acceleration) factor had good fit indices, it was not superior to the five-factor model. This may be explained by the fact that employees are not necessarily affected by all intensified demands to a similar extent. For example, even if intensified learning occurs, this does not presuppose that work intensification also has to occur.

Moreover, we showed that the measurement structure and the latent structure are invariant across two national samples. Using a sample of employees from an international firm, we demonstrated that the German and the English versions of the IDS were invariant with regard to their factor loadings. However, the different language versions were only partially structurally invariant. This could be due to the fact that the cross-validation was based on a small sample size (Byrne et al., 1989). Taken together, the results suggest that the factor structure of the IDS is metric invariant across different countries (Germany and Austria) and languages (German and English), indicating the robustness of the scale. Moreover, the results show that the IDS is fully structurally invariant across countries, but only partially structurally invariant across the English and German versions. Finally, all items loaded high on their corresponding latent dimensions and all five dimensions of the IDS yielded satisfactory internal consistencies.

Another aim of this research was to develop a measure that—although capturing intensified demands—is related to existing measures of job demands. Using different samples of service workers, we showed that all five subscales of the IDS had moderate positive correlations with their respective traditional demands. These results provide evidence for good construct validity, since the IDS subscales are not identical to established measures of work characteristics, but still moderately related to them.

The fact that negative affectivity is at most to a small extent related to intensified demands also argues in favour of the construct validity of the IDS. This finding lends support to the assumption that the IDS indeed measures changes in job demands rather than employees’ tendencies to see the world in a negative way and that the items are formulated in a non-affective and descriptive way (Cavanaugh et al., 2000). Nevertheless, the IDS still measures subjective perceptions of intensified demands rather than objective changes. Considering the assumptions made by the transactional stress model (Lazarus & Folkman, 1984), assessing subjective perceptions of demands is important since these perceptions determine the stress process.

Despite the importance of employees’ perceptions, we aimed to verify the IDS measure by replicating recent research findings on differences in the intensification of demands among different subgroups of employees. In line with previous findings (Green & McIntosh, 2001), the IDS is able to detect differences between employees who use ICTs at work. Moreover, we demonstrated that employees in NWW conditions perceive a stronger intensification of work and of job-related planning and decision-making demands than their colleagues without NWW. These findings are in line with previous research showing that flexible work arrangements contribute to more discretion over the scheduling and location of work tasks as well as work intensification (Kelliher & Anderson, 2008). Since increases in demands are often linked to changes in work situations (Höge, 2011), we also showed that employees who experience changes in their work situation perceive higher levels of intensified knowledge-related learning and intensified planning and decision-making demands than their counterparts with stable jobs. However, the IDS measure was not able to distinguish between employees whose work situation changed and those whose work situation remained stable in terms of skill-related learning demands. This result may be due to the fact that employees who experience changes in their work situation do not necessarily need to intensify their skill acquisition. The finding that employees with a change in their work situation perceive a stronger intensification in demands does not question the assumption of an overall intensification in job demands due to the acceleration of societal, economic, and organizational changes. Rather this result shows that the amount of intensification in demands employees are confronted with may vary depending on individual factors (Korunka & Kubicek, 2013). Taken together, the study results indicate that the ratings employees provide in terms of intensified demands based on the IDS measure are in line with recent verifiable changes reported by existing research on increasing demands (e.g., Green & McIntosh, 2001).

Finally, supporting the incremental validity of the IDS, its subscales were found to add to the prediction of emotional exhaustion and cynicism beyond traditional job demands.

Contribution to traditional measurements of work characteristics and outlook for future studies

The IDS expands existing measures of job demands in at least four ways. First, existing literature already points to the importance of “new”/intensified demands, which are triggered by accelerated economic, societal, and organizational changes (Cascio, 2003; Green, 2004; Obschonka et al., 2012; Stone, 2010). Despite the growing attention directed towards the intensification of various job demands, no measure has existed until now which is able to capture the wide range of intensified demands. The IDS closes this gap by providing a comprehensive instrument measuring various aspects of current changes in job demands.

Second, the IDS offers the opportunity to measure change directly, instead of using longitudinal trend or panel data. Such a measure of intensified job demands is not prone to response shift (Schwartz & Sprangers, 2009) or regression effects (Popper et al., 2012). When change is determined through longitudinal data by comparing employees’ assessments at two points in time, it is assumed that study participants have an internalized standard for judging the level of job demands that remains stable over time (Visser, Smets, Sprangers, & Dehaes, 2000). It is, however, likely that employees will shift their metric of assessment when they are confronted with transformations in working conditions (Golembiewski, Billingsley, & Yeager, 1976; Schwartz & Sprangers, 1999), especially in case of long-term transformations. In addition, people with initially high levels of job demands have only limited possibilities for indicating that their demands have increased over time in a longitudinal assessment, because they reach the upper limits of the response format. Therefore, our approach for measuring intensified job demands seems, despite the advantages of longitudinal data, more effective for dealing with long-term changes in job demands. Moreover, assessing employees’ perceptions of such changes is important to examine the impact of current macro-level (societal and organizational) transformations on psychological outcomes such as individual well-being (e.g., Obschonka et al., 2012).

Third, the IDS extends existing measures by adding to the prediction of work outcomes. In fact, each subscale of the IDS explained additional variance of emotional exhaustion after controlling for traditional job demands. In contrast to traditional instruments, the IDS takes not only the actual amount of exposure, but also the increase in demands over a period of at least 5 years into account. This could be a particular advantage when it comes to the effects of demands on employee well-being in the long run. With its long-term perspective of 5 years, one could assume that the IDS is especially effective at explaining long-term impairments of well-being (e.g., emotional exhaustion).

Fourth, the IDS was able to identify expected differences between employees working in flexible and traditional work arrangements. This suggests that the IDS could be helpful in differentiating among groups of employees who are more or less affected by current technological, societal, and organizational changes.

Limitations

This study has some limitations that need to be considered. First, in all four samples, we relied solely on self-reported data. Using the same source of data may inflate the associations between the constructs (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). However, because we also included traditional job demands as control variables in the regression models, this limitation is unlikely to have affected all our hypotheses. Since variance that is shared by the predictors because of common methods of data collection gets attributed to neither predictor in multiple regression analysis (Cohen & Cohen, 1983), the observed relations are likely the result of true covariation between the constructs.

Second, the IDS focuses on the intensification of job demands in service industries. Thus, the instrument may not be appropriate for measuring changes in construction work or farming, for example. However, given differences in job demands and their transformations among industries, this restriction seems necessary. If the items of the IDS were designed to fit all industries, they might be unintelligible or might not capture relevant changes.

Third, we recruited two samples via online panels. We tried to ensure the quality of the study samples in three ways: (1) The first sample resembles the German population working in the service industries in terms of gender and working hours. (2) We checked the data from the German and the Austrian service worker samples for careless responding (response duration, unacceptable number of missing responses, etc.) and excluded persons so identified (Johnson, 2005). (3) We used data from an organizational sample to replicate the factor structure of the IDS. Considering all three points, we assume that our findings are generalizable and not prone to selection or response bias.

Fourth, the IDS measure was not structurally invariant across the German and English versions. However, small sample sizes, especially for those who answered the English version of the IDS, could explain this effect (Byrne et al., 1989). Moreover, only the correlation between the two intensified learning demands was not structurally invariant.

Fifth, although the IDS was designed to measure intensification in job demands over time, we were not yet able to compare responses on the IDS with longitudinal changes. This would, however, be an important further step in validating the IDS.

Implications

From a practical perspective, the IDS offers the opportunity for people in charge of (re-)designing jobs to identify demands subject to intensification as well as side effects of changes in work conditions. Work intensification may, for example, be a negative side effect of flexible work arrangements (Kelliher & Anderson, 2008). Moreover, information on which aspects of work have been intensified over the last few years may also lead to ideas about interventions that have the potential to buffer the negative effects of such changes or boost their positive effects. For example, offering job resources may help employees avoid the negative effects of intensified job demands (Bakker & Demerouti, 2007).

From a theoretical perspective, our findings have implications for stress research. With regard to work intensification, our research indicates that task-related stressors such as time pressure or work overload (Sonnentag & Frese, 2003) should be complemented by the intensification and compression of work. In fact, we were able to show that employees’ reactions to workload or work intensity are determined not only by the absolute amount of work that has to be completed within a working day but also by the increase in work intensity over the past 5 years. Thus, stress models and theories (such as the JD-R model; Bakker & Demerouti, 2007) should not only focus on job demands at a specific point in time, but may also want to consider the intensification of such demands.

With regard to intensified planning and decision-making and learning demands, our findings reveal difficulties in classifying job characteristics as either job demands or job resources, because planning and decision-making and learning—once unambiguously conceptualized as resources—can become demands. In his vitamin model, Warr (1994) suggests that specific work characteristics such as decision latitude may exert positive effects only up to a certain level and may become harmful once this threshold is exceeded. Our findings indicate that intensified planning and decision-making and learning demands may have such negative effects. For both demands we found positive associations with emotional exhaustion and cynicism. For intensified career-related planning and decision-making demands we even found a negative association with job satisfaction. These findings suggest that an ever-growing amount of planning and decision-making and learning has detrimental effects on employees’ well-being. Thus, future research on job demands and job resources should consider that having more and more of a good thing may become stressful in the end.

Conclusion

In the present research, an instrument for assessing job demands arising from accelerated change was developed and validated. Based on a literature review and previous empirical findings, five domains of job demands were identified that are subject to intensification: workload, job-related planning and decision-making demands, career-related planning and decision-making demands, knowledge-related learning demands, and skill-related learning demands. Using various cognitive probing techniques, a set of 19 items was developed. In four samples, the construct, convergent, and incremental validity of the newly developed IDS were supported. Thus, the instrument provides a reliable and valid measure to assess the extent to which employees experience an intensification of job demands.