Job Loss, Human Capital Job Feature, and Work Condition Job Feature as Distinct Job Insecurity Constructs
Blau, Gary
The projected growth of new technologies, increasing use of automation, and continued consolidation of healthrelated services suggest that continued study of job insecurity is needed for health care professionals. Using a sample of 178 medical technologists over a 5-year period, this study’s findings extend earlier work by Blau and Sharp (2000) and suggest that job loss insecurity, human capital job feature insecurity, and work condition job feature insecurity are related but distinct types of job insecurity. A seven-item measure of job loss insecurity, a four-item measure of human capital job feature insecurity, and a four-item measure of work condition job feature insecurity were analyzed. Confirmatory factor analysis using a more heterogeneous sample of 447 working adults supported this threefactor structure. Using correlation and path analysis, different significant relationships of antecedent variables and subsequent organizational withdrawal cognitions to these three types of job insecurity were found. J Allied Health. 2004; 33:31-41.
IN A 1995 POLL,1 the top source of employee stress cited was job insecurity (i.e., fear of losing one’s job). To support that job security increasingly is disappearing for many, Pearce2 found that the growth rate of the temporary help industry was almost twice the growth rate of the U.S. gross national product from 1970 to 1984. Current worker concern about job insecurity can be found across many different industries in the United States, including health care.3 Despite the current identified employee shortages in such health care fields as nursing, radiologie technology, pharmacy, and medical technology,4-6 continued hospital and health care facility merger activity facilitates job insecurity perceptions.7,8 In addition, the projected growth of new technologies, use of robotics and automation, and continued consolidation of health care-related services6 will continue to make job insecurity an important issue for many health care employees. Job insecurity perceptions are not confined to the United States; they also are found in Great Britain, Israel, Holland, and Finland.9-11 The purpose of this study is to explore further the dimensionality of job insecurity beyond a paper by Blau and Sharp.12
Using a sample of medical technologists over a 3-year period (1995-1997), Blau and Sharp12 operationalized job loss insecurity versus job feature insecurity using two-item and one-item measures and found support for their distinctiveness. The present study uses the same general sample of medical technologists over a later time frame (1996-2000) and a new heterogeneous sample of working adults to test for distinguishing two types of job feature insecurity, human capital and work condition, from job loss insecurity, using an expanded job insecurity item domain. To distinguish the current study further from Blau and Sharp,12 there is minimal variable overlap.
Job Insecurity Research
DEFINING JOB INSECURITY
In a comprehensive treatment of the job insecurity construct, Greenhalgh and Rosenblattn presented a multidimension definition of job insecurity with a model summarizing antecedents, nature, and impact, including organizational consequences, of job insecurity. Greenhalgh and Rosenblatt13 defined job insecurity as “perceived powerlessness to maintain desired continuity in a threatened job situation.” Other definitions of job insecurity support the internalized perceptual nature of this variable.14,15 Greenhalgh and Rosenblatt13 defined threat to one’s job based on severity of the threat, which included considering whether job loss was permanent versus temporary and whether the whole job was lost (job loss) versus six identified job features (i.e., career progress, income stream, status/self-esteem, autonomy, resources, and community). Greenhalgh and Rosenblatt13 noted that loss of job features was an important but often overlooked aspect of job insecurity.
Consistent with Greenhalgh and Rosenblatt’s’13 distinction of job loss versus job feature insecurity, Johnson et al16 empirically found “high-source level” insecurity (job insecurity associated with reorganization and decline) to be related but distinct from “low-source level” insecurity (job insecurity associated with arbitrary supervision and technologic change). Closer inspection of the Ashford et al17 study results, comparing correlations of different job insecurity scales18 with other variables, indicates that it may be useful to separate job loss versus job feature insecurity. An overall job insecurity measure, combining job feature and job loss insecurity, partially may attenuate relationships with other variables. Given the limited operationalizations of job loss (two items) and job feature (one item) insecurity by Blau and Sharp,12 they did not compare variable relationships with overall job insecurity versus separate job loss and job feature insecurity measures.
FURTHER DISTINCTION WlTHIN JOB FEATURE INSECURITY
Careful scrutiny of the items within job feature insecurity scales”11,17 suggests at least two related but different types of job feature insecurity, human capital and work condition. Human capital job feature insecurity is defined as “a perceived personal threat that affects an individual’s monetary income and/or psychic income.” This definition is based on Becker’s’9 work in which human capital is concerned with activities (e.g., schooling, on-the-job training, benefits), that directly influence future employee monetary and psychic income. By contrast, work condition job feature insecurity is defined as “a perceived threat to nonpersonal, jobspecific situational characteristics.” As noted earlier, Greenhalgh and Rosenblatt13 identified six specific job features. Three of these identified features fit within human capital job feature insecurity (i.e., career progress, income stream, status/self-esteem), and the other three features fit within work condition job feature insecurity (i.e., autonomy, resources, community).
Perceived “vulnerability” is inherent in definitions of job insecurity.13,15 From a severity of threat perspective,13 job loss insecurity would be the most threatening or stressful because of its greater change intensity. ^° The Holmes and Rahe21 Social Readjustment Rating Scale supports this severity of threat “hierarchy” in that “job loss” (being fired) receives more points (47) than “human capital job feature change” (e.g., promotion, demotion) (29) versus “work condition job feature change” (change in working hours or conditions) (20). The first hypothesis is as follows:
1. Job loss insecurity, human capital job feature insecurity, and work condition job feature insecurity are related but distinct constructs.
TESTING THE DISCRIMINANT VALIDITY OF JOB LOSS, HUMAN CAPITAL, AND WORK CONDITION JOB FEATURE INSECURITY
As with prior studies,22,23 it can be argued that the five italicized antecedent variables proposed for this study fall within several of the general forms of powerlessness (in parentheses) suggested by Greenhalgh and Rosenblatt:13 downsizing (lack of protection); work ethic and career enrichment benefit satisfaction (unclear expectancies), and procedural justice and interactional justice (organizational culture). We can expect stronger links between certain antecedents and types of job insecurity perceptions and between job insecurity perceptions, and withdrawal cognitions.
Organizational decline is the dominant context in which job loss insecurity is likely to occur,24 and research shows that prior organizational downsizing has a strong impact on employee job loss insecurity.12,25,27 This downsizing may have less of an impact on job feature insecurity because of the employees’ focus on permanent loss of organizational membership as a priority over loss of job features.13
Unclear expectancies, as a category of powerlessness,13 involves employees not knowing what is needed to maintain current job status, including corrective action to take to avert a perceived threat. Two variables, work ethic and career enrichment benefit satisfaction, fall within this category and are negatively related to human capital and work condition job feature insecurity. Work ethic is an individual’s belief in the value of hard work.28 Career enrichment benefit satisfaction is an employee’s attitude toward organizational benefits focusing on skill development needs.29
Work ethic and career enrichment benefit satisfaction are individual, discretionary, “corrective action-oriented” variables and, as such, should have a stronger relationship to human capital job features, which are personal resource based. An employee may believe that working harder (i.e., having a stronger work ethic), can lessen one’s vulnerability concerning loss of monetary or psychic income.30 This stronger work ethic also may lead to lower work condition job feature insecurity. Employees who feel better about their skill development level via career enrichment benefits, such as continuing education, should feel more secure about various job features such as career progress, status/self-esteem, and autonomy, alleviating human capital and work condition job feature insecurity.13 Employees are less likely to believe that working harder or upgrading their skills would save their jobs.31
Greenhalgh and Rosenblatt13 noted that: “an employee’s sense of powerlessness would be exacerbated if the organization had no strong norms of fairness.” Procedural justice measures the perceived fairness of the organization32 (e.g., consistency of decision application, opportunity for voice). When employees perceive that the organization treats them fairly, this lessens their feelings of job loss insecurity.26,33 By being more of an “organizational level” variable,32,34 procedural justice does not have as strong of an impact on either human capital or work condition job feature insecurity. Human capital is more person based, and job feature is more job specific.19
Greenhalgh and Rosenblatt13 also noted that arbitrary supervision can exacerbate an employee’s sense of powerlessness. How employees perceive their immediate supervisor treats them (i.e., interactional justice)34,35 has a significant relationship to work condition job feature insecurity. Supervisors often control job-specific resources and opportunities (e.g., via delegation).36 Interactional justice should have a weaker influence on human capital job feature insecurity, which is based more on personal resources (e.g., job skills employees have). Supervisors typically are less able to influence job loss insecurity due to external factors, such as lower product demand.37
Based on stress theory,20 one might expect the stronger the perceived threat, the stronger the flight response. Applying this logic here, the strongest positive relationship should be between job loss insecurity to withdrawal cognitions, followed by human capital to job feature insecurity, and work condition to job feature insecurity.11,17 Cumulatively, these studies suggest the following hypothesis:
2. Job loss insecurity, human capital job feature insecurity, and work condition job feature insecurity will exhibit discriminant validity such that:
a. Organizational downsizing would have the strongest positive relationship to job loss insecurity.
b. Work ethic and career enrichment benefit satisfaction would have stronger negative relationships to human capital and work condition job feature insecurity than to job loss insecurity.
c. Procedural justice would have the strongest negative relationship to job loss insecurity.
d. Interactional justice would have the strongest negative relationship to work condition job feature insecurity.
e. Job loss insecurity would have the strongest positive relationship to organizational withdrawal cognitions, followed by human capital and work condition job feature insecurity.
Figure 1 summarizes the study hypotheses and shows the partial nomological network of variables tested.
Methods
SAMPLE ONE AND PROCEDURE
This sample is part of a longitudinal study12 on the career patterns of recently graduated medical technologists (MTs) by the Board of Registry of the American Society for Clinical Pathology (ASCP). MTs work in a laboratory in a variety of health-related settings (e.g., hospitals, independent laboratories). They are responsible for the accurate performance of tests that help to determine the presence of disease. Surveys were distributed to a sample of MTs across hundreds of different organizations in 1996, 1997, 1998, 1999, and 2000. Surveys were sent to individuals’ home addresses. Participants were invited to voluntarily participate in a longitudinal study on the career patterns of recently graduated MTs. In 1996, 652 of 1,156 (56%) surveys containing demographic (e.g., gender, age, marital status, education level), work ethic, and downsizing items were returned voluntarily. In 1997, 616 of 1,156 (53%) matched (by social security number) MTs responded, allowing demographic, downsizing, and career enrichment benefit satisfaction data to be collected. In 1998, 553 of 1,156 (48%) MTs returned their surveys containing demographic, downsizing, procedural justice, and interactional justice data. In 1999, 509 of 1,156 (44%) MTs returned their surveys containing demographic, downsizing, job loss, and job feature insecurity items. In 2000, 506 of 1,156 (44%) MTs returned their surveys containing demographic and organizational withdrawal cognitions data.
Although there were 506 repeat-respondents across the five surveys, based on the variables used in this study, complete data were available for only 178 MTs. Such a reduction in sample size over a 5-year time frame is common.38 This sample reduction was due to several factors, including missing data and eliminating respondents who indicated “not available” for career enrichment benefit satisfaction items, “did not know” for downsizing items, or that they were not employed in the laboratory that year. Respondents still were kept in the sample database, however, in case they became employed in the laboratory at a later date. A 1996 demographic comparison on gender, age, marital status, and education level of the 178 complete-data MT sample to the 474 (652 – 178) remaining MT sample indicated no significant demographic differences. A 1996 demographic breakdown of the sample of 178 MTs showed that their median age was 28 years (range 24-58 years), 82% were women, 93% had a baccalaureate degree, 7% had an advanced degree, and 65% were married. By 2000, these demographics either remained stable or increased as expected (e.g., age). Population demographics collected by the ASCP in 2000 on 73,471 MTs showed that 82% were female, and the median age was 43. This sample studied is representative for gender but is younger.
SAMPLE ONE SURVEY ITEMS
Job Loss and Job Feature Insecurity, 1999
Fifteen items were collected in 1999, 7 measuring job loss insecurity and 8 measuring job feature insecurity. Survey constraints limited the number of items that could be measured. The seven job loss items focused on being permanently laid off for different reasons, which is consistent with Greenhalgh and Rosenblatt.13 These items were based partially on previous work by Ashford et al17 and Kuhnert and Vance.39 The eight items measuring job feature insecurity were based on previous work by Ashford et al17 and Caplan et al.18 Given the current dynamic nature of the health care industry,7,8 respondents were asked alternatively to use “next year” and in the “next 3 years” time frames when answering 14 of the items. Previous research18 has used projected time ranges in assessing job insecurity perceptions. A 4-point response scale was used for all 15 items, where 1 = strongly disagree; 2 = disagree; 3 = agree; and 4 = strongly agree. Research has shown that the proportion of the scale used in a 4-point response format is no different than more common 5-, 6-, and 7-point formats.40 These 15 items were pilot tested among nine medical technology experts.41 Unless otherwise indicated, all other multi-item measures noted subsequently also used this 4-point response scale. For shortened scales, item retention was hased on strength of factor loadings in the cited reference.
Prior Organizational Downsizing, 1996 to 1999
This 4-item measure is based on aggregating respondent responses to yearly measures of overall organizational downsizing in 1996, 1997, 1998, and 1999. In each of these 4 years, MTs were asked, “did your institution downsize (i.e., reduce personnel),” where 1 = no and 2 = yes. Percentages by year are as follows: 1996, no (59%), yes (41%); 1997, no (66%), yes (34%); 1998, no (67%), yes (33%); 1999, no (70%), yes (30%). Precedent exists for using each time-based measure as an item and calculating a scale score based on adding items.42 If respondents indicated they “didn’t know” (a response option), this was coded as missing data.
Work Ethic, 1996
This 3-item scale is based on Blau and Ryan’s43 18-item scale. Survey constraints necessitated this, and a sample item is “hard work makes one a better person.”
Career Enrichment Benefit Satisfaction, 1997
This 6-item scale by Blau et al29 asked respondents to “indicate your level of satisfaction with each of the following benefits,” which included “education assistance/tuition reimbursement,” “tuition/grants for continuing education,” and “release time for continuing education/professional meetings.” These are examples of employee development activities.44 A 4-point response scale was used, 1 = dissatisfied; 2 = somewhat satisfied; 3 = satisfied; and 4 = very satisfied. If a benefit was not available in their organization, MTs were asked to respond 0 = not available. Following other benefit satisfaction research,45 these data were treated as missing.
Procedural Justice, 1998
This 4-item measure is based on Niehoff and Moorman’s46 6-item scale. Survey constraints necessitated using a shorter measure. A sample item is “job decisions are applied consistently by management across all affected employees.”
Interactional Justice, 1998
This 4-item measure also is based on Niehoff and Moorman’s46 6-item scale. Survey constraints necessitated the shorter measure. A sample item is “when making a decision affecting me, my supervisor treats me with dignity and respect.”
Organizational Withdrawal Cognitions, 2000
This 3-item measure is based on Blau.47 A sample item is “I intend to leave the organization where I am currently employed as soon as possible.”
SAMPLE TWO AND PROCEDURE
Complete data were collected in fall 2000 from a sample of 447 workmg adults working across many different organizations. Most respondents were working graduate or under’ graduate business students from a university in the northeastern United States. Occupationally, these respondents indicated they were represented as follows: 19% in the medical/health area (e.g., nursing, physical therapy, physician); 20% in the technical area (e.g., engineering, information technology, financial); 29% in the administrative area (e.g., management, advertising, government-related); 8% in education (e.g., teaching, religious, library services); 10% in service (e.g., hospitality, real estate, sales); and 14% in “other,” in which respondents were asked to fill in their current job title. Job titles reported included administrative assistant/secretary, human resources, clerical, coach, and nanny. Other demographic data on the 447 respondents were as follows: 60% were women, 51% were married, 84% were nonunion, 58% indicated that their income was the primary source of household income, 2% were younger than 21 years old, 52% were 21 to 35 years old, 31% were 36 to 50 years old, 14% were 51 to 65 years old, 1% were older than 65 years old, and 85% indicated that they worked at least 35 hr/wk. These demographics indicated that this sample was more heterogeneous than the MTs.
Class-based surveys were distributed and collected with the permission of the instructors, who returned the surveys to the investigator. Participation was voluntary, and respondents were told that the brief, anonymous survey was designed to measure their perceptions of job insecurity. A small percentage of surveys in this overall sample were collected on-the-job through personal contacts, in which contacts were asked to distribute to colleagues at work and collect the surveys to return to the investigator. There were no significant demographic differences between class versus on-the-job respondents.
SAMPLE TWO SURVEY ITEMS
Job Loss and Job Feature Insecurity, 2000
To test for the generalizability of the job insecurity measures, the same 15 items collected for sample one were collected for sample two. The same 4-point response scale ( 1 = strongly disagree, 4 = strongly agree) was used.
Results
PRELIMINARY ANALYSES
Confirmatory factor analysis was done using the work ethic, procedural justice, and interactional justice items48 to evaluate the discriminant validity of these shortened antecedent measures (i.e., do these items load [fit] on the appropriate factor?). Confirmatory factor analysis is appropriate when an investigator thinks he or she “knows” the number of expected factors.48 The following fit statistics were found for the three-factor solution: [chi]^sup 2^ = 151.36 (degrees of freedom = 41, p = 0.03); comparative fit index (CFI) = 0.953; adjusted goodness of fit index (AGFI) = .896; root mean square residual (RMSR) = 0.023; and root mean square error of approximation (RMSEA) = 0.074. Adequate fit is indicated when the fit indices are at least 0.90 and root mean square measures are less than 0.08.49,50 Overall, these reported statistics indicate an acceptable fit and support the discriminant validity of these three shortened measures.
Given the large amount of missing data, an analysis was done to determine if subject attrition was biasing the study results. Goodman and Blum51 recommended using logistic regression because it models the probability of being included in one of two response categories, remaining or leaving a sample. The response variable is staying versus leaving the sample, and all study variables constituted the explanatory variables in the logistic regression. Results indicated that none of the study variables was related significantly to remaining in versus leaving the sample.
JOB INSECURITY SCALE DEVELOPMENT-SAMPLE ONE
Because the 15 items collectively represent a unique set of job insecurity perceptions, exploratory factor analysis was used. Exploratory factor analysis is used when an investigator does not “know” the number of expected factors.48 The decision on how many factors to retain was based on two commonly accepted criteria, eigenvalues greater than 1 and a scree test.52 Given that the job insecurity factors were expected to be related to each other, an oblique rotation (promax method) was applied.52 A conservative cutoff of 0.50 for factor loadings was used to consider an item to be part of a factor.53 The results of the exploratory factor analysis are shown in Table 1. As the results show, a three-factor solution was found, with the first seven items loading on the first factor, the next four items loading on the second factor, and the final four items loading on the third factor. Three factors (scales) identified were named: job loss insecurity (items 1 to 7), human capital job feature insecurity (items 8 to 11), and work condition job feature insecurity (items 12 to 15). The factor correlations were 0.63, job loss-human capital job feature insecurity; 0.44, job loss-work condition job feature insecurity; and 0.52, human capital-work condition job feature insecurity. Scale scores were calculated by summing items and dividing by the number of items in that scale. Factor analyses also were done omitting “duplicate items” (i.e., using items with only “next year” versus “3 years” time frames). Results were consistent with those shown in Table 1.
JOB INSECURITY ITEM CONFIRMATORY FACTOR ANALYSIS-SAMPLE TWO
Having shown that a three-factor model was best in sample one, confirmatory factor analysis using Amos 4.0(48) was used on the same 15 job insecurity items in this new sample to see the degree to which these items “fit” the proposed three-factor structure (model). The following fit statistics were found for the full sample (N = 447): [chi]^sup 2^ = 408.97 (degrees of freedom = 87, p = 0.04); CFI = 0.952; AGFI = 0.889; RMSR = 0.062; and RMSEA = 0.083. Overall, these statistics indicate an adequate fit, although the AGFI is slightly less than 0.90, and the RMSEA is slightly greater than 0.08, the recommended threshold values.49,50 By contrast, the fit statistics were not as supportive for simpler (i.e., one- and two-factor) models of job insecurity. These model comparisons support using the three-factor model of job insecurity.
Table 2 shows the variance estimates for the 15 items for the three-factor solution. The variance estimate for each item shows how strongly an item “loads” on a particular factor. Composite reliabilities were 0.92 for job loss insecurity, 0.93 for human capital job feature insecurity, and 0.92 for work condition job feature insecurity.54 Factor correlations were 0.59 between the job loss insecurity and human capital job feature insecurity, 0.46 between job loss insecurity and work condition job feature insecurity, and 0.49 between human capital job feature insecurity and work condition job feature insecurity. Overall the exploratory and confirmatory factor analyses provide support for hypothesis 1 (i.e., job loss insecurity, human capital job feature insecurity and work condition job feature insecurity, are related but distinct).
DISCRIMINANT VALIDITY OF JOB LOSS, HUMAN CAPITAL, AND WORK CONDITION JOB FEATURE INSECURITY
Means, SDs, and correlations for variables using the MT sample are shown in Table 3. Means are based on the response scale used for that variable. All multi-item scales had internal consistencies greater than the general threshold of 0.70 except for the aggregated downsizing variable (0.62).55 The internal consistency of a measure represents its reliability (i.e., how consistent the items comprising a measure are with each other).55 The numbers given in the 1-to-10 matrix are correlations (r), and the criteria for judging their statistical significance is given in the footnote. A correlation indicates how the scores of two different measures covary with each other (e.g., a positive correlation indicates that as one measure increases, the other measure increases). Correlations can range in magnitude from 0 (no relationship) to 1.0 (perfect relationship).56 The positive, but not perfect, correlations between these job insecurity scales indicate that they are related but distinct. An overall job insecurity measure was created by summing all 15 job insecurity items. The results in Table 3 show that neither work ethic (r = -0.08) nor interactional justice (r = -0.14) was related significantly to overall job insecurity. Work ethic (r = -0.17) and interactional justice (r = -0.29) were related significantly negatively to work condition job feature insecurity, however. This suggests that an overall measure of job insecurity may “mask” variable relationships to components of job insecurity.17
Tests for significant differences between correlations (p
Results indicate partial support for hypothesis 2a (i.e., a stronger positive relationship of downsizing to job loss [r = 0.45] and human capital job feature insecurity [r = 0.34]). Partial support also was found for hypothesis 2b (i.e., work ethic shows a stronger negative relationship to only work condition job feature insecurity [r = -0.17], whereas career enrichment satisfaction shows a stronger negative relationship to only human capital job feature insecurity [r = -0.31]). Hypothesis 2c is supported (i.e., procedural justice is more negatively related to job loss insecurity [r = -0.34]). Hypothesis 2d also is supported (i.e., interactional justice is more negatively related to work condition job feature insecurity [r = -0.29]). Finally, hypothesis 2e is supported (i.e., the positive relationship between job loss insecurity and organizational withdrawal cognitions is strongest [r = 0.39], followed by human capital job feature insecurity [r = 0.27] and work condition job feature insecurity [r = 0.16]).
Figure 2 shows the path model results of testing the theoretical model (Figure 1) using Amos 4.0.48 This data analysis technique allows for examining the proposed relationships among all variables within a model. When analyzing the fit of the data to the proposed model, the Amos program automatically notes how this fit can be improved through the use of modification indices.48 Investigators are cautioned, however, to use modifications only if they make theoretical sense. 48 The only additional path added to Figure 2 (compared with Figure 1) is the significant path from prior downsizing to human capital job feature insecurity. The overall fit of the model is good.50 Hypotheses were tested further by looking at the significance of relevant path coefficients for the manifest (boxed) variables. The oval “error” terms for the outcome variables in Figures 1 and 2 are necessary to acknowledge that there are other variables that could affect these outcomes.48 These results are consistent with the results shown in Table 3. Overall, 63% of job loss insecurity, 44% of human capital job feature insecurity, and 19% of work condition job feature insecurity were explained. Overall, Figure 2 provides support for the theoretical model (Figure 1).
Discussion
Collectively the results of the study hypotheses provide initial support for job loss insecurity, human capital job feature insecurity, and work condition job feature insecurity being related but distinct job insecurity constructs. In addition, combining job loss and feature job insecurity into an overall measure can mask finer variable relationships. As such, this study extends initial work by Blau and Sharp,12 who used more limited job insecurity measures, did not distinguish between human capital versus work condition job feature insecurity, and did not test for the masking impact of overall job insecurity. In addition, this study found independent sample study support for this three-dimension measure of job insecurity.
As with previous research,22 it was argued that perceived powerlessness could be operationalized as different antecedent variables consistent with the general categories of powerlessness suggested by Greenhalgh and Rosenblatt.13 It would have been desirable, however, to have also measured powerlessness as an individual’s perceived ability to counteract threats,17 to test better its role within the job insecurity construct.2,15 The operationalization of the job loss insecurity construct emphasized permanent job loss due to downsizing and restructuring. This is consistent with other empirical research15 and represents a “purer” form than other job loss measures.11,17
It would have been desirable to use all 17 job feature items from Ashford et al17 to see if these two types of job feature insecurity could have been operationalized. Inspection of these 17 items suggests that some of the items (e.g., “potential to get ahead,” “potential to maintain current pay,” “status that comes with your position in the company,” and “your potential to attain pay increases”) seem to fall within human capital job feature insecurity,18 whereas other items (e.g., “your current freedom to schedule your own work,” “your current access to resources [people, material, information] in the organization,” “the physical demands the job places on you,” and “the opportunity to do a job from start to finish”) seem to be work condition job feature insecurity issues. This item distinction between types of job feature insecurity is consistent with the Greenhalgh and Rosenblatt13 framework.
Stress theory20,22,57 supports job loss insecurity having a stronger relationship to organizational withdrawal cognitions followed by human capital job feature insecurity and work condition job feature insecurity. When organizations suddenly downsize, surviving employees immediately can become concerned with permanent job loss and want to leave.17 Perhaps more insidious, however, is a potential progression within job insecurity constructs (i.e., from work condition job feature insecurity to human capital job feature insecurity to job loss insecurity). From an organizational standpoint, top management must recognize the importance of sending honest, timely messages to its employees regarding changes affecting their jobs. In the absence of such messages, unintended organizational clues and rumors are used with potentially more damaging effect.13
Looking at the results, work ethic had a stronger negative relationship to work condition job feature insecurity, whereas career enrichment benefit satisfaction had a stronger negative relationship to human capital job feature insecurity. Staw’s58 distinction between an individual’s attempt first to control his or her environment and, if not possible, then to make that environment more predictable seems applicable here. By working harder an employee may perceive that one can control at least to a degree one’s immediate working conditions (e.g., decision-making freedom, access to resources). Next, an employee can make his or her human capital worth more “predictable”19 by enhancing his or her credentials to the organization through career enrichment activities (e.g., getting a degree, involvement in continuing education).
Procedural justice was related significantly negatively to job loss insecurity, whereas interactional justice was related negatively to work condition job feature insecurity. These findings reinforce similar different “level” results for the relationships of these justice variables to other outcomes.34 It was not expected that such downsizing would impact as strongly on human capital job feature versus job loss insecurity. Perhaps there is an “accumulated effect,” in that respondents exposed to downsizing over a 4-year period also would feel less secure about their personal resources and begin to question the value of their job skills and job status in their organization.13
There are other study limitations to note. All data collected were self-report, and some of the variables, particularly downsizing, were measured crudely. Although the complete study sample of 178 MTs was shown to be representative for gender for the MT population of the ASCP and was not affected by attrition bias over the 5-year period,51 the “representativeness” of this sample for other demographics is questionable. Survey constraints not only hurt the domain coverage for job feature insecurity, but also necessitated using shortened measures for other variables and did not allow for assessing either perceived powerlessness or the severity of threat for each type of job insecurity. The degree to which job insecurity existed when the earlier variables were measured is not known. Time differences in the measurement period between variables may have affected study results (e.g., the relationships of 1996 work ethic versus 1998 procedural and interactional justice to 1999 job insecurity measures). These variables were asked only during specific waves of data collection. Repeated collection of antecedent measures could have controlled for time differences, affecting results.
Assuming that additional studies support generalizing the distinctiveness of these job insecurity constructs, future research is needed to test the potential progressive impact from work condition job feature insecurity to human capital job feature insecurity to job loss insecurity and the role of powerlessness in distinguishing among these job insecurity constructs.17 As noted at the beginning of this article, we believe that various trends, including the projected growth of new technologies, increasing use of automation, and continued consolidation of health-related services,6 will continue to support the need to study health care employee perceptions of job insecurity. We would encourage investigations of job insecurity using other health care samples, including dental hygienists, dietitians, speech-language pathologists, physical therapists, and occupational therapists. It is hoped that this study will stimulate this additional research.
We thank the Board of Registry for permission to use this data. We wish to thank Jim Arbuckle, John Deckop, and Frank Linnehan, for their assistance in the development of this article. We also wish to thank Lynne Andersson, Joe Bucci, Carol Carriere, Tom Daymont, George Eisele, Jo Anne Edwards, Deanna Geddes, Jesse Guiles, Mike Guglielmo, Ingo Kampa, Alison Konrad, David Malotsky, Mike Morris, Steve Oliver, Ron Roberts, Susan Sharp, Sumiko Sumida , Jerry Zeitz, and Ed Zuchelkowski for their help in collecting data on the second sample.
REFERENCES
1. Anonymous, job insecurity tops workplace stress. USA Today Mar 1995;123:8.
2. Pearce J. Job insecurity is really important, but not for the reasons you might think: The example of contingent workers. In: Cooper C, Rousseau D (eds). Trends in Organizational Behavior. London: Wiley, 1998, pp 31-46.
3. Daniel T. Between trapezes: The human side of making mergers and acquisitions work. Compens Benefits Manage 1999;15:19-37.
4. Aiken L, Clarke S, Sloane D, et al. Nurses’ reports on hospital care in five countries. Health Aff 2001;May/June:1-8.
5. Akroyd D. Survey reveals radiologic technologists affected by burnout. News Across the Profession. Dec 18, 2001. Available at: http://asahp.org.
6. Ward-Cook K. Medical laboratory workforce trends and projections: what is past is prologue. Clin Leadership Manage Rev 2002;Nov/Dec:364-369.
7. Armstrong-Stassen M, Cameron S, Mantler J, Horsburgh M. The impact of hospital amalgamation on the job attitudes of nutses. Can J Admin Sci 2001;18:149-162.
8. Kami K. Mergers and re-engineering-possible implications for allied health students (the Minnesota story). J Allied Health 1997; 26:35-40.
9. Hartley J, Jacobson D, Klandermans B, Van Vuuren T (eds). Job Insecurity: Coping with Jobs as Risk. London: Sage; 1991.
10. Kinnunen U, Mauno S, Natti J, Happonen M. Organizational antecedents and outcomes of job insecurity: a longitudinal study in three organizations in Finland. J Organ Behav 2000;21:443-459.
11. Rosenblatt Z, Ruvio A. A test of the multidimensional model of job insecurity: The case of Israeli teachers. J Organ Behav 1996;17: 587-605.
12. Blau G, Sharp S. Job-loss insecurity versus job-feature insecurity among medical technologists. J Allied Health 2000;29:86-91.
13. Greenhaugh L, Rosenblatt Z. Job insecurity: toward conceptual clarity. Acad Manage Rev1984;9:438-448.
14. Dekker S, Schaufeli W. The effects of job insecurity on psychological health and withdrawal: A longitudinal study. Aust Psychol 1995; 30:57-63.
15. Jacobson D. The conceptual approach to job insecurity. In: Hartley J, Jacobson D, Klandermans B, Van vuuren T (eds). Job Insecurity: Coping with Jobs at Risk. London: Sage; 1991, pp 23-39.
16. Johnson N, Bobko P, Hartenian L. Union influence on local union leaders’ perceptions of job insecurity: an empirical test. Br J Ind Relations 1992;30:45-60.
17. Ashford S, Lee C, Bobko P. Content, causes and consequences of job insecurity: a theory-based measure and substantive test. Acad Manage J 1989;32:803-829.
18. Caplan R, Cobb S, French J, et al. Job Demands and Worker Health. HEW pub no. (NIOSH). 75-160. Washington, DC: U.S. Department of Health, Education, and Welfare; 1975.
19. Becker G. Human Capital: A Theoretical and Empirical Analysis. New York: National Bureau of Economic Research; 1964.
20. Selye H. Stress in Health and Disease. Boston: Butterworths; 1976.
21. Holmes T, Rahe R. The social readjustment rating scale. J Psychosom Res 1967;11:213-218.
22. Klandermans B, Van Vuuren T, Jacobson D. Employees and job insecurity. In: Hartley J, Jacobson D, Klandermans B, Van vuuren T (eds). Job Insecurity: Coping with Jobs at Risk. London: Sage; 1991, pp 40-64.
23. Van Vuuren T, Klandermans B, Jacobson D, Hartley J. Predicting employees’ perceptions of job insecurity. In: Hartley J, Jacobson D, Klandermans B, Van vuuren T (eds). Job Insecurity: Coping with Jobs at Risk. London: Sage; 1991, pp 65-78.
24. Hartley J. Industrial relations and job insecurity: learning from a case study. In: Hartley J, Jacobson D, Klandermans B, Van vuuren T (eds). Job Insecurity: Coping with Jobs at Risk. London: Sage; 1991, pp 123-150.
25. Brockner J. The effects of work layoffs on survivors: research, theory and practice. Res Organ Behav 1988;10:213-255.
26. Brockner J, Grover S, Reed T, DeWitt R. Layoffs, job insecurity and survivors’ work effort: evidence of an inverted-U relationship. Acad Manage J 1992;35:413-425.
27. Ferrie J, Shipley M, Marmot M, et al. The health effects of a major organizational change and job insecurity. Soc Sci Med 1998;46″ 243-254.
28. Morrow P. The Theory and Measurement of Work Commitment. Greenwich, CT: JAI Press; 1993.
29. Blau G, Merriman K., Tatum D, Rudmann S. Antecedents and consequences of basic versus career enrichment benefit satisfaction. J Organ Behav 2001;22:669-688.
30. Furnham A. The Protestant Work Ethic: The Psychology of WorkRelated Beliefs and Behaviors. London: Routledge; 1990.
31. Greenhalgh L, Sutton R. Organizational effectiveness and job insecurity. In: Hartley J, Jacobson D, Klandermans B, Van vuuren T (eds). Job Insecurity: Coping with Jobs at Risk. London: Sage; 1991, pp 151-171.
32. Moorman R. Relationship between organizational justice and organizational citizenship behavior: do fairness perceptions influence employee citizenship? J Appl Psychol 1991;76:845-855.
33. Schweiger D, DeNisi A. Communication with employees following a merger: a longitudinal field experiment. Acad Manage J 1991;34: 110-135.
34. Masterson S, Lewis K, Goldman B, Taylor S. Integrating justice and social exchange: the differing effects of fair procedures and treatment on work relationships. Acad Manage J 2000;43:738-748.
35. Bies R, Moag J. Interactional justice: communication criteria of fairness. In: Lewicki R, Sheppard B, Bazerman M (eds). Research on Negotiations in Organizations, Vol. 1. Greenwich, CT: JAI Press; 1986, pp 43-55.
36. Blumberg M, Pringle C. The missing opportunity in organizational research: some Implications for a theory of work performance. Acad Manage Rev 1982;7:560-569.
37. Hartley J. Industrial relations and job insecurity: a social psychological framework. In: Hartley J, Jacobson D, Klandermans B, Van vuuren T (eds). Job Insecurity: Coping with Jobs at Risk. London: Sage; 1991, pp 104-122.
38. Winefield A, Tiggerman M. Employment status and psychological well-being: a longitudinal study. J Appl Psychol 1990;75:455-459.
39. Kuhnert K, Vance R. Job insecurity and moderators of the relation between job insecurity and employee adjustment. In: Quick J, Murphy L, Hurrell J (eds). Stress and Well-Being at Work: Assesssments and Interventions for Occupational Mental Health. Washington, DC: American Psychological Association; 1992, pp 48-63.
40. Matell M, Jacob, J. Is there an optimal number of alternatives for Likert-scale items? J Appl Psychol 1972;56:506-509.
41. American Society of Clinical Pathologists. Research and development committee meeting. New Orleans, Sept 24, 1999.
42. Adler S, Golan J. Lateness as withdrawal behavior. J Appl Psychol 1981; 66:544-554.
43. Blau G, Ryan J. On measuring work ethic: a neglected work commitment facet. J Voc Behav 1997;51:435-448.
44. Noe R, Wilk S, Mullen E, Wanek J. Employee development: issues in construct definition and investigation of antecedents. In: Ford J, Kozlowski S, Kraiger K, et al (eds). Improving Training Effectiveness at Work. Mahwah, NJ: Lawrence Erlbaum; 1997, pp 153-189.
45. Danehower C, Lust J. Understanding and measuring employee benefit satisfaction. Benefits Q 1995;11(1):69-75.
46. Niehoff B, Moorman R. Justice as a mediator of the relationship between methods of monitoring and organizational citizenship behavior. Acad Manage J 1993;36:527-556.
47. Blau G. Testing the generalizability of a career commitment measure and its impact on employee turnover. J Voc Behav 1989; 35:88-103.
48. Arbuckle J, Wothke W. Amos Users’ Guide (version 4.0). Chicago, IL: Smallwarers Corporation; 1999.
49. Bentler P. Comparative fit indexes in structural models. Psychol Bull 1990;107:238-246.
50. Browne M, Cudek R. Alternative ways of assessing model fit. In: Bollen K, Long J (eds). Testing structural equations models. Newbury Park, CA: Sage; 1993.
51. Goodman J, Blum T. Assessing the non-random sampling effects of subject attrition in longitudinal research. J Manage 1996;22:627-652.
52. Fabrigar L, Wegener D, MacCallum R, Strhan E. Evaluating the use of exploratory factor analysis in psychological research. Psychol Methods 1999;4:272-299.
53. Hair J, Anderson R, Tatham R, Black W. Multivariate Data Analysis. Englewood Cliffs, NJ: Prentice Hall; 1995.
54. Fornell C, Larcker D. Evaluating structural equation models with unobservable variables and measurement error. J Market Res 1981; 18:39-50.
55. Nunnally J. Psychometric Theory. New York: McOraw Hill; 1978.
56. McNemar Q. Psychological Statistics. New York: Wiley; 1969.
57. Lazarus R, Folkman S. Stress, Appraisal and Coping. New York: Springer; 1984.
58. Staw B. Rationality and justification in organizational life. In: Cummings L, Staw B (eds). Research in Organizational Behavior, Vol 2. Greenwich, CT: JAI Press; 1980, pp 45-80.
Gary Blau, PhD
Donna Surges Tatum, PhD
Keith McCoy, MS
Lidia Dobria, MS
Kory Ward-Cook, PhD
Dr. Blau is in the Human Resource Administration Department, Temple University, Philadelphia, Pennsylvania; and Dr. Surges Tatum, Mr. McCoy, Ms. Dobria, and Dr. Ward-Cook are on the ASCP Board of Registry, Chicago, Illinois.
Received July 16, 2002; accepted April 2, 2003.
The authors are members of the Research and Development Committee for the Board of Registry, American Society of Clinical Pathologists.
Address correspondence and reprint requests to: Gary Blau, Human Resource Administration Department, Temple University-Fox SBM, 384 Speakman Hall, Philadelphia, PA 19122; telephone: (215) 204-6906; email: gblau@sbm.temple.edu.
Copyright Association of Schools of Allied Health Professions Spring 2004
Provided by ProQuest Information and Learning Company. All rights Reserved
