Objective To investigate short- and long-term effects of major organisational change on minor psychiatric disorder and self-rated health for women and men in different employment grades.
Methods Minor psychiatric disorder and self-rated health among 6710 British civil servants (1993 women and 4717 men) in three employment grades from the Whitehall II study were examined from 1985 to 1988 under stable employment conditions. The short-term effects of organisational change were investigated in 1991–1993 after a time of major restructuring aiming at increasing the influence of market forces in the civil service and the long-term effects were investigated in 1997–1999.
Results Those who had experienced organisational change and those who anticipated organisational change reported more negative short-term health effects (minor psychiatric disorder and poor self-rated health) compared with those who reported no change. No major differences were found depending on employment grade or gender. The negative health effects had diminished during 1997–1999 for those who reported that a major change had happened before 1991–1993. Those who anticipated an organisational change in 1991–1993 still reported more ill-health in 1997–1999 (both minor psychiatric disorder and self-reported health) than those in the comparison group.
Conclusions The results indicate that organisational change affects employees’ health negatively in the short term but also that it is possible to recover from such negative effects. As it was not possible to discern any definite difference between the gender and grades, the results point at the importance of working proactively to implement organisational change for women and men at all levels.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
What this paper adds
The long-term health consequences of major organisational change are unclear as well as to what extent the health effects of change differ depending on employment grade and gender.
Analyses of the effects of major organisational change indicated that the short-term negative health effects of changes implemented before 1991–1993 had diminished at 1997–1999, but the negative health effects were still present in 1997–1999 for changes anticipated during 1991–1993.
It was not possible to discern any definite difference between gender or employment grades.
This study indicates that it is possible to recover from short-term negative health effects of a major organisational change and emphasises the importance of working proactivelyto reduce negative health consequences during a major change for women and men in all organisational levels.
Organisational change is a common feature of today's working life.1 Previous research has indicated that organisational change often results in negative health consequences for the employees.2–7 Less is known about the permanency of such negative effects and the possibility of recovering, especially when controlling for confounders and investigating mental health.8 One of the major trends that has been affecting employees all over the world since the late 1980s is the increased introduction of market forces in the public sector.9 The consequences of such changes have been extensively investigated from an economic perspective, but less is known about how such changes affect the health of the employees.10
The observed negative health consequences of organisational change have been attributed to the inherent nature of uncertainty that lies within the concept of change.11 ,12 Uncertainty during organisational change has been associated with different stress reactions such as increased blood pressure,13 increased levels of total cholesterol14 and a decrease of the hormone DHEA-S that is involved in recovery.3 The uncertainty associated with a change usually differs over time (see Zapf et al15). For example, the anticipation phase of an organisational change has been shown to be associated with high levels of uncertainty and negative health consequences for employees.13 ,16 ,17 Short-term negative health consequences have also frequently been reported for different kinds of major changes such as downsizing,18 expansion6 and mergers.19 While it seems that the negative health effects diminish over time,2 the effects could still be present several years after a change.6 Major downsizing was, for example, associated with a lower self-rated health 4 years after the downsizing20 and the risk for cardiovascular death has been shown to be elevated 4–7.5 years after a major downsizing.5
It has also been suggested that the perceived uncertainty during organisational change differs depending on employees’ social status in the organisation.21 Two fundamental categories of social status are employment grade22 and gender23 and, hence, the effects of organisational change may differ between these groups.24 The health of employees in lower positions could be more negatively affected by organisational change25 due to lower levels of personal control.10 ,12 However, managers have reported higher level of change-related stress than non-managers which has been attributed to extensive responsibilities for carrying out organisational changes.26 Employees in higher grades also have more of important attributes such as power and prestige to lose during times of turbulence.2 In respect of gender, higher risk of long-term sickness absence6 and musculoskeletal problems25 have been observed in women compared with men following major organisational change. However, similar patterns in health as well as sickness absence for women and men during downsizing have also been observed.4 ,18
A problem with studying the consequences of organisational change is the difficulty to measure a baseline that is unaffected by the change. The present study uses data from the Whitehall II study that was set up during the relatively stable employment conditions which characterised the British civil service during the 1980s.27 The first data collection (1985–1988) was accomplished before the major restructuring that aimed to increase the importance of market forces in the Civil Service through the ‘Next Steps’ programme and that was launched in 1988.28 This programme was part of a major deregulation in Britain that was followed by many other countries in Europe and worldwide.9 The Next Steps programme separated the executive functions of government from policy advice and created ‘agencies’ of the executive functions. Each Next Steps agency became responsible for reaching its targets in terms of finance and service and to publish an annual report and account. The Next Steps programme was also associated with other aspects of restructuring, such as a reduction of middle management,29 and decentralisation of personnel and finance to all civil service units, whether agencies or not, which meant massive reorganisation of departments.30 The increased focus on effectiveness and competition has also been argued to impact on the resources that were set aside to equal opportunities initiatives and to erode gains already made in this area, such as access to managerial positions for women, part-time and flexible working patterns, child care and career break initiatives.29 Studies on Whitehall II data, focusing on the change into agencies, have shown that employees who had been transferred or anticipated transfer to an agency reported more self-reported psychological and physiological morbidity as well as increase in some clinical measures such as body mass index and increased blood pressure31 as well as higher level of work disability.27 The present study covers a broader range of organisational changes and adds the long-term perspective of health consequences of organisational change as well as a focus on social status. The specific aim of this study was to investigate how major organisational change was associated with short- and long-term effects on minor psychiatric disorder and self-rated health for women and men at different employment grades.
All non-industrial civil servants aged 35–55 years who were working in the London offices of 20 departments between 1985 and 1988 were invited to participate in the Whitehall II study. With a response rate of 73%, the final cohort consisted of 10 308 participants (3413 women and 6895 men).32 Although the respondents were mostly white collar (office) workers, they covered a wide range of grades. Twelve non-industrial grades of employment were identified based on salary and grouped into three employment grades (administrative, professional/executive, clerical/support) in this study. Phase 1 (1985–1988) involved a clinical examination and a self-administrated questionnaire. In 1989–1990, all of the original study participants were sent a postal questionnaire (phase 2) and all participants were invited to take part in a further screening examination and questionnaires in 1991–1993 (phase 3). Data were then collected at every other occasion with postal questionnaires only (phase 4: 1995–1996; phase 6: 2001; phase 8: 2006) and every other occasion both with questionnaires and screening (phase 5: 1997–1999; phase 7: 2003–2004; phase 9: 2008–2009).32 This study analyses data from phases 1, 3 and 5.
The analyses of short-term effects of major organisational change as well as the stratification by gender and employment grade were based on those 6710 (1993 women and 4717 men) who had complete data for all study variables at phase 1 (with imputation from phase 2 where data were missing for longstanding illness and negative affectivity) and phase 3. The analyses of long-term effects of major organisational change were based on those with complete data of all phase 1 variables (with imputation for the same variables as above), the rating of organisational change at phase 3 and the phase 5 dependent variable of interest. This meant that 5494 participants (1565 women and 3929 men) were included in the phase 5 analysis of minor psychiatric disorder and 5560 participants (1603 women and 3957 men) were included in the phase 5 analysis of poor self-rated health. The participants in this study were working in the Civil Service during the major restructuring that was implemented between phases 1 and 3 and were then followed irrespective of whether or not they were still working in the Civil Service at phase 5. Table 1 shows the characteristics of the participants. The 3598 participants not included in the analysis of the short-term effects tended to be slightly older, more often men, single, divorced, separated or widowed and from the low employment grade. Those not included in the analyses, to some extent, also had worse health at baseline (longstanding illness, minor psychiatric disorder and poor self-rated health) and more negative affectivity compared with the sample included in the analyses.
Ethical approval for the Whitehall II study was obtained from the University College London Medical School committee on the ethics of human research. All study participants gave informed consent to participate in the research.
Since the Civil Service was going through a major change, starting with the implementation of the Next Steps programme in 1988, a question was included at phase 3 (1991–1993) that read: ‘Major changes in the organisation and location of civil service departments have been made and/or are planned. Which of these changes affect you? Major changes in the organisation or management of your department.’ The question was followed by the options: (1) has happened, (2) is planned, (3) not certain what will happen and (4) is not planned. In the analyses, three groups were constructed, two exposure groups of change—those who had been through major changes (option 1 above) and those who anticipated major changes (a combination of those who had answered option 2 and 3)—and a comparison group consisting of those who answered that no major changes were planned (option 4). The question about major changes only applied to those who were still working at phase 3.
Self-rated health measures
Minor psychiatric disorder was measured at phases 1, 3 and 5 with the 30-item General Health Questionnaire (GHQ), a well-established screening questionnaire for psychiatric disorder suitable for use in general population samples.33 The variable was dichotomised so that all those scoring 0–4 were considered ‘non-cases’ and those scoring 5+ were considered ‘GHQ cases’, based on results from a validation study carried out at phase 1.34
At phases 1 and 3, health over the past year was self-rated as very good, good, average, poor or very poor. For the purpose of the analysis, this measure was dichotomised and used as an indicator of poor self-rated health (very poor, poor or average) compared with good (good or very good). At phase 5, this measure was assessed using the self-rated general health question from the 36-item Short Form Health Survey (the SF-36 scale).35 General health was self-rated as excellent, very good, good, fair or poor. This variable was dichotomised to poor (fair or poor) and good (excellent, very good, good).
The following covariates were drawn from the phase 1 questionnaire: age (divided in 5-year age groups), gender, marital status (married/cohabiting, single/divorced/separated/widowed) and current civil service grade of employment (administrative, professional/executive, clerical/support). Ill-health at phase 1 was measured by self-reported presence of longstanding illness, disability or infirmity (yes, no). Negative affectivity was measured using the five negative affect items from the 10-item Bradburn Affect Balance scale.36 As these last two measures were only administered to a part of the sample at phase 1, data were imputed from phase 2 when phase 1 data on longstanding illness (n=1486) and negative affectivity (n=1519) were missing. Phase 1 measures of minor psychiatric disorder (GHQ) and poor self-rated health were also included as covariates.
Logistic regression was used to estimate ORs for GHQ cases and poor self-rated health at phase 3 (1991–1993) and phase 5 (1997–1999) in the two exposure groups of change compared with the reference group (no planned change) adjusting for gender and age (Model 1) and additionally adjusting for employment grade (Model 2). In order to adjust for possible initial group differences in health, further adjustment was made for phase 1 values of the dependent variable of interest (Model 3), with GHQ at phase 1 included as a score (ranging from 0 to 30) and poor self-rated health included as a binary variable (very poor, poor or average vs good or very good). Additional adjustments were then made for possible confounders such as marital status, longstanding illness and negative affectivity (Model 4).
In order to investigate whether the association of major change with health varied according to gender and employment grade, the analyses for phase 3 were repeated stratified by employment grade and gender. These stratified analyses were run adjusted for baseline values of the dependent variable of interest as well as age. A formal test of effect modification was carried out by extending Model 3 to include two- and three-way interaction terms among change, employment grade and gender (ie, the interactions terms: change by employment grade, change by gender, gender by employment grade and change by employment grade by gender). Since employees may have changed employment grade between phases 1 and 3, additional analyses were carried out where phase 3 employment grade was used instead of phase 1. All analyses were performed using the statistical package IBM SPSS Statistics V.19.
Health status (both GHQ cases and poor self-rated health) at phase 3 was worse in the two change exposure groups compared with the comparison group where no changes were planned (table 2). These differences in health between the exposure groups and the comparison group remained after adjusting for phase 1 values of age, gender, employment grade, marital status, longstanding illness and negative affectivity, as well as the baseline value of the dependent variable (GHQ or self-rated health). The worse health among those who reported that major changes had happened before phase 3 (1991–1993) had diminished at phase 5 (1997–1999). However, those reporting that they anticipated changes at phase 3 (1991–1993) still reported worse health (both GHQ cases and poor self-rated health) at phase 5 follow-up (1997–1999) after adjustments for all covariates (table 3).
When the results for short-term effects (1991–1993) were stratified by employment grade and gender (table 4), the pattern of results was similar for minor psychiatric disorder (GHQ cases) and poor self-rated health. The general trend was that all participants, irrespective of gender and employment grade, reported worse health in the two exposure groups of change, with ORs above 1 in all groups. However, the association between organisational change and health reached significance in only some of the groups. For women, those in the highest grade who anticipated changes reported more psychiatric disorder as well as poorer self-rated health than women in the comparison group with no planned changes. Also women in the lowest grade reported more psychiatric disorder where changes had happened than women in the comparison group in the same grade. Men in the two change exposure groups reported more psychiatric disorder and poorer self-rated health compared with men where no changes were planned, both in the high and middle employment grades. When formally tested, the three-way interaction terms among change, employment grade and gender were not significant (GHQ cases p=0.284; poor self-rated health p=0.441) and the comparisons between the different strata of gender and grade should therefore be interpreted with caution.
In this study, we found that the risk for minor psychiatric disorder and poor self-rated health increased in the short term among employees who had experienced or were anticipating a major organisational change intended to increase the influence of market forces in the British Civil Service. This is in line with earlier research that indicates that organisational change may have negative consequences for employees’ health.2 ,4 ,5 We observed that both changes that had happened and changes that were anticipated were associated with similar negative short-term health outcomes (at phase 3), which is consistent with other studies from Whitehall II16 ,31 and elsewhere.17 In respect of long-term effects, the results indicated no long-term (1997–1999) effects of changes reported to have happened before phase 3 (1991–1993) suggesting that it is possible to recover from negative health effects of change. This is in line with studies indicating that organisational instability per se can have negative health effects7 that diminish as the stress associated with the change decreases.2 However, the changes that were anticipated at phase 3 seemed to have affected the employees’ health several years later (1997–1999), both in sense of minor psychiatric disorder and poor self-rated health. This could either be an effect of a long time of insecurity associated with anticipation or, in fact, short-term effects of changes that were planned at phase 3 (1991–1993) but implemented shortly before 1997–1999. Irrespective of which of these plausible scenarios may actually have happened these employees seemed to have experienced increased ill-health for many years.
The knowledge about how groups with different social status are affected by organisational change is limited. The few studies that have investigated how organisational change affect health for employees in different occupational grades10 ,12 ,26 ,37 or gender4 ,6 ,18 ,25 ,38 have shown inconclusive results. This study suggests that the major organisational change that was implemented in the British Civil Service from 1988 affected both women and men in all grades negatively to some extent. However, one group that may be worth mentioning is women in the highest grade who were anticipating change. For these women, anticipation of change was associated with a high risk of self-reported minor psychiatric disorder and poor self-rated health. Since women have lagged behind their male colleagues in gaining access to higher posts in the Civil Service and since the introduction of the Next Steps programme has been criticised to narrow the opportunities for women and erode earlier gains in equal opportunities29 it may be that the uncertainty that is associated with anticipation of change 16 ,17 was especially burdensome for women in the highest grade. Other studies have also noted that a restructuring process may exacerbate gender bias and that gender bias is likely to be more pronounced at higher levels in a corporate hierarchy.38
The major organisational change investigated here was a result of large scale introduction of market forces in the British Civil Service that was launched in 1988. Great Britain was one of the first countries to introduce those kinds of changes but similar changes were introduced worldwide especially during the 1990s.9 While it is impossible to generalise widely from one study of a particular restructuring effort, the present study adds to the evidence that organisational change, such as an increase of market forces in the public sector, can have negative effects for the health of the employees in the short run. These temporary effects are likely to be more generalisable than long-term outcomes of change since the short-term effects probably are due to instability and insecurity during the change process per se rather than to the organisational structure that the changes eventually lead to. The amount of services in a society that are governed by the state or by the market are something that has been fluctuating through recent history9 and that is constantly under debate. The risk of increased ill-health for employees, at least in the short term, is worth considering in that discussion.
This study followed a large number of participants over three phases of data collection which means that participants were followed over 9–14 years, thus adding to the research of long-term consequences of organisational changes that has been called for.8 However, even if the baseline was before the major restructuring started, it was not possible to ascertain when the changes ended. We therefore treated our two exposure groups (those reporting that major change ‘had happened’ or ‘was anticipated’) as ‘intention-to-treat groups’ as in clinical trials. This means that those who reported exposure at phase 3 (1991–1993) remained in their respective treatment group regardless of what happened to them after they were assigned to their group.
Few, if any studies, have analysed how organisational change affects health for women and men in different employment grades with a baseline measured before the implementation of change which make it possible to adjust for several confounders such as ill-health before change. One reason for the lack of research is the difficulty to get a baseline measure of health before a major change is announced and, second, since the labour market is gender segregated, it is hard to obtain a sufficiently large sample to study how gender and employment grade interact during organisational change. Even the large sample in the present study suffers from lack of power due to the low number of women in the highest grade as well as few men in the lowest grade.
The fact that we used employment grade at phase 1 to determine the gradient means that our measure of social status is not contemporaneous with the measure of self-reported psychological and physiological health at phases 3 and 5. Although there was a general uplift in grade until phase 3, the uplift was similar between the three groups that were compared (exposure groups and comparison group of organisational change) and analyses that were repeated with employment grade at phase 3 produced findings very similar to those presented (data not shown).
Another issue worth mentioning is that both exposure and outcome measures were based on self-reports, which theoretically could produce overestimation of ORs and differences in the analysis if the exposure groups contained more respondents with a tendency to over-report health problems. To minimise this risk we included negative affectivity as a covariate which did not change the estimated effects in any substantial way.
There were differences at phase 1 between the analytic sample and those who were excluded from the analyses due to missing data. The clerical/support grade was under-represented in the sample as were women, the group of single, separated, divorced and widowed persons and those with more morbidity and negative affectivity, thus indicating a ‘healthy worker’ effect. If employees with already poor health at baseline would be more negatively affected by organisational change, the estimates reported in this study are likely to be an underestimation of the negative health effects of organisational change.
In this study, self-rated health was measured with the same question in phases 1 and 3, but the wording both of the question and the response options had been changed until phase 5 (when SF-36 was used). Self-rated health is a measure that has been described both to provide summative information about various domains of health and to capture aspects of health that are difficult to measure with other methods. It also has a strong association with mortality even after adjustments for comorbidity, depression and functional status.39 Since the advantages of using self-rated health are many, we wanted to include it also when analysing long-term effect of organisational change. Even if different self-rated health measures may not be directly comparable they measure the same underlying construct of health.40 The fact that the observed long-term effects of organisational change both for minor psychiatric disorder and self-rated health were similar indicates that the measure may be proper to use.
This study was limited to investigate the health effects of organisational change. In future studies it would be interesting to investigate the mechanisms behind the health consequences. For example, to investigate how the amount of insecurity and work characteristics, such as demand and control, varies during different phases of change for women and men in different employment grades.
The major organisational change that was a result of the large scale introduction of market forces in the British Civil Service, launched in 1988, resulted in poorer self-rated psychological and physiological health status when measured during 1991–1993 both for those who anticipated organisational change and for those who reported that organisational change had already happened, while controlling for health as baseline as well as other possible confounders. These results support earlier findings that major organisational change may carry a risk of increased ill-health for employees. The worse health observed for those who had been exposed to change before 1991–1993 had, however, declined when measured again during 1997–1999, thus pointing at the possibility of recovering from the negative health effects of a major change. Those anticipating change in 1991–1993 still reported poorer health status in 1997–1999. Based on the result from this study, we could not make any definite conclusion that there was a gender or grade difference, thus pointing at the importance of working proactively with supporting women and men in all employment grades when implementing major organisational changes.
We thank all participating men and women in the Whitehall II Study; all participating Civil Service departments and their welfare, personnel and establishment officers; the Occupational Health and Safety Agency; and the Council of Civil Service Unions. The Whitehall II Study team comprises research scientists, statisticians, study coordinators, nurses, data managers, administrative assistants and data entry staff who make the study possible.
Contributors All authors participated in designing the analysis, interpreting the data and critically reviewing the paper. HF analysed the data and wrote the first draft.
Funding The Whitehall II study has been supported by grants from the Medical Research Council; Economic and Social Research Council; British Heart Foundation; Health and Safety Executive; Department of Health; National Heart Lung and Blood Institute (HL36310), US, NIH; National Institute on Aging (AG13196), US, NIH; Agency for Health Care Policy Research (HS06516); and the John D and Catherine T MacArthur Foundation Research Networks on Successful Midlife Development and Socio-economic Status and Health. HF was supported by a fellowship from Riksbankens Jubileumsfond as well as a grant from the Swedish council for working life and social research (2007-0933) and FAS Marie Curie International Postdoc Fellowship Programme (2012-0091). EIF was supported by a fellowship from the Swedish council for working life and social research (2010-1823). JH is partially supported by the National Institute on Aging, NIH (R01AG013196).
Competing interests None.
Ethics approval Ethical approval for the Whitehall II study was obtained from the University College London Medical School committee on the ethics of human research.
Provenance and peer review Not commissioned; externally peer reviewed.