Table 1 Privatisation (psychosocial health impacts)
First author and interventionStudy designReported outcomes
Ferrie,2630 London, UKPrivatisation of a civil service department responsible for design and construction of public buildings (Property Services Agency) during the Whitehall II study of UK civil servants. Privatisation plans included large-scale redundancies. These plans were disseminated to PSA employees during the anticipation stage of the study. Change in public investment not reportedEvaluation covers 1985–94. Intervention occurred in 1992(A) Prospective controlled study of civil servants anticipating intervention. Three waves measuring a range of self-reported morbidity and physiological risk factors: 6 years (baseline), 3 years (anticipation), and 1 month (pretermination) before privatisationMethods appraisal*: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10(B) Cross-sectional study of the intervention’s effects on employment status and health, using the fourth wave 18–24 months after after privatisation and redundanciesPopulation: all employees aged 35–55 years. Total n = 14 121. Baseline response rate  = 73%. Follow-ups = 79–81% of baseline. No control for wave 4Methods appraisal*: 3, 4, 5, 6, 7, 8, 9, 10(A) Waves 2 and 3: some self-reported morbidity and physiological risk factors increased in the intervention group relative to control—for example, wave 3, body mass index (p<0.001); sleeping ⩾9 h (OR = 1.88; 95% CI = 1.3 to 2.8; p<0.01); ischaemia (OR = 1.45; 95% CI = 1.0 to 2.1). (B) For (1) insecurely employed (29%) and (2) job-seeking unemployed (19%), increased minor psychiatric morbidity (mean difference (1) 1.56; 95% CI = 1.0 to 2.2; (2) 1.25; 0.6 to 2.0) and >3 GP consultations in past year (OR (1) 2.04; 1.1 to 3.8; (2) 2.39; 1.2 to 4.7), relative to securely employed respondents (31%)
Nelson,31 32 UKPrivatisation of most of the UK’s water industry. Change in public investment not reportedEvaluation covers 1989–91 Intervention occurred in 1989Prospective cohort study of employees’ self-reported health 1 month (T1) prior, 8 months (T2) and 20 months (T3) afterTotal pop.  = 4501Initial sample = 1500. Baseline response rate  = 41%. Follow-ups = 79%, 64% of baselineMethods appraisal*: 1, 3, 4, 6, 7, 8, 10OSI mean scores suggest worsening mental health for clerical and administrative staff at T2 compared with baseline (p = 0.018). Otherwise, little change in mean OSI scores for mental or somatic symptoms for clerical and administrative staff, manual workers or managers at either T2 or T3. These figures are based on the final data reported from Nelson’s thesis of 1999,39 rather than from the preliminary findings of 1995.38
Cunha,33 PortugalPrivatisation of a paper factory and two cement companiesGovernment shares floated = 44% (paper factory), 20% and 59% (cement companies)Evaluation covers 1994–96 Interventions occurred in 1994 (cement companies) and 1995 (paper factory)(A) Repeat cross-sectional study with control group (paper factory employees, 1 year before and after). Total pop. = 2800. Initial sample  = 1244. Achieved sample  = 128. Response rate  = 10%Methods appraisal: 1, 2, 3, 4, 8, 10(B) Cross-sectional comparison of paper factory’s employees before privatisation and cement companies’ employees 18 months after privatisation. Total pop. = 5800. Initial sample = 2119. Achieved sample = 289. Overall response rate = 14% (paper factory = 10%, cement factories = 18% and 28%)Methods appraisal*: 2, 3, 4, 8, 10(A) Paper factory: self-reported mental and physical health scores using the OSI were similar for intervention and control groups before and after privatisation (F(1,123) = 0, p>0.05)(B) Analysis of variance for mental and physical OSI symptom scores suggest positive relationship between ill health and the extent to which a company was privatised: mental ill health (F(2,285) = 6.3, p<0.01), and physical ill health (F(2,285) = 6.0, p<0.01)
  • PSA, Professional Staff Association; OS1, occupational stress indicator.

  • *Methods appraisal: 1 = prospective; 2 = appropriate comparison group; 3 = representative sample; 4 = all intervention groups exposed (uncontaminated control group); 5 = baseline response ⩽60%; 6 = follow-up⩽60%; 7 = adjustment for non-response and dropout; 8 = appropriate statistical tests; 9 = exploration of and (if necessary) adjustment for confounding interventions; 10 = conclusions substantiated by data.