Public, private or both? Analysing Factors Influencing the Labour Supply of Medical Specialists.

Transcription

1 Public, private or both? Analysing Factors Influencing the Labour Supply of Medical Specialists. Terence C Cheng, Guyonne Kalb, Anthony Scott Melbourne Institute of Applied Economics & Social Research Centre of Research Excellence in Medical Workforce Dynamics The University of Melbourne Singapore Health Economics Association Conference 17 April / 19

2 Background The balance between public and private financing and provision of health care is a policy issue in many countries. Fiscal pressures have led governments to consider expanding the private sector. Channel demand away from public sector into private sector; relieve cost and capacity pressures. Subsidising private health/private insurance is expensive, and not cost-effective (Cheng 2014, JHE) 2 / 19

3 Background Supply-side implications: Supply-side implications: Recruitment and retention of public sector doctors. Problem of cream-skimming when doctors allowed to combine public and private work, or dual practice. (Biglaiser and Ma 2007, RAND) - Private providers have incentives to select patients with less severe conditions; public hospitals to deal with complex patients. Monitoring and diversion of effort into private work. Lack of consensus on the effects of dual practice reflected in heterogeneity in regulations across countries (González and Macho-Stadler 2013, JHE). 3 / 19

4 Background Medical labour market an interesting case to study multiple job holding (Paxton and Sicherman 1996, JLE). Hours constraint as motivation for second job not likely to be relevant. Job portfolio motive: doctors derive different utilities from different jobs. Doctors frequently combine different work: 6 work settings 21 different combinations observed. Preferences for different jobs differ across gender, age and other individual characteristics (e.g. rural background). 4 / 19

5 Background Table 1: Distribution of practice modes of MABEL specialists (%) Males Females Age bands (years) Public Private Both Public Private Both N 2,757 1,884 4,748 1, ,421 Note: Pooled data from waves 1-4 of MABEL survey. 5 / 19

6 Paper s Aims Investigate how pecuniary (sector specific earnings) and non-pecuniary factors influence allocation of time between public and private sectors by medical specialists. Estimate discrete choice structural labour supply model using cross-sectional data from a longitudinal survey of Australian doctors. The model is used to simulate the impact of changes in sector-specific earnings on supply of labour in the public and private sectors. 6 / 19

7 Structural model of public and private labour supply Doctors choose public and private working hours to maximise utility given net income (net earnings + other household income). Labour supply a discrete choice (Van Soest 1995). Doctors choose from a set of combinations of income and working hours in public and private sector jobs. Hours bands. We allow doctors to choose 11 different choices of public and private sector work: public & private hours per week {0, 1-34, 35-49, 50+}. Choice model estimated using multinomial logit. Estimation of public/private sector hourly earnings for each doctor using four waves of data. Calculate net income using tax and transfer rules. 7 / 19

8 The MABEL survey Data from first wave (2008) of the Medicine in Australia: Balancing Employment and Life (MABEL) longitudinal survey. ( All 54,750 doctors invited to participate. 10,498 (response rate = 19.4%) doctors form the baseline cohort, with 4597 specialists. Representative of population on age, gender etc (Joyce et al 2010). Wave 6 in field. Waves 1 to 5 available for use by researchers. Analysis sample: 1,670 males, 634 females. Use reported weekly hours by work setting to define public (public hospitals) and private (private hospitals, private consulting room) sector work. 8 / 19

9 Sector and hours worked Table 2: Sample means Male Female Diff. a Sector and hours worked: Public sector 0.26 (0.44) 0.40 (0.49) *** Private sector 0.18 (0.39) 0.18 (0.39) Mixed sector 0.56 (0.50) 0.42 (0.49) *** Total hours for public sector (11.54) (13.83) *** Total hours for private sector (14.68) (14.84) *** Total hours for mixed sector (10.82) (12.48) *** 9 / 19

10 Marginal effect on hours Table 3: Effects of covariates on weekly hours worked Male Female Public Private Total Public Private Total Covariates Hours Hours Hours Hours Hours Hours Age (in years) -4.04*** *** -3.66*** 2.24** -1.42** (0.64) (0.60) (0.36) (1.03) (1.07) (0.69) Has child aged 0-4 years *** *** (1.53) (1.63) (0.75) (1.51) (1.72) (0.90) Has child aged 5-9 years *** *** (1.66) (1.53) (0.75) (1.51) (1.60) (1.16) Has child aged years ** (1.46) (1.48) (0.68) (1.91) (1.77) (1.34) Partner is employed (1.41) (1.31) (0.74) (1.88) (1.80) (1.31) Partner not employed (1.64) (1.48) (0.76) (2.30) (3.01) (2.12) Note: Bootstrap standard errors shown in parenthesis. Significance: *** 1%; ** 5%; * 10%. 10 / 19

11 Results - Sector elasticity Table 4: Sector elasticities with one-percent increase in hourly earnings Male Female Public Private Both Public Private Both Elasticities earnings earnings earnings earnings earnings earnings Panel A: Sector elasticity Public sector 0.71*** -0.91*** ** (0.18) (0.14) (0.15) (0.22) (0.18) (0.14) Private sector -0.77*** 0.93*** * 0.69*** 0.13 (0.16) (0.14) (0.20) (0.33) (0.25) (0.30) Mixed sector (0.02) (0.05) (0.03) (0.05) (0.10) (0.09) Note: Bootstrap standard errors shown in parenthesis. Significance: *** 1%; ** 5%; * 10%. 11 / 19

12 Results - Hours elasticity Table 5: Hours elasticities with one-percent increase in hourly earnings Male Female Public Private Both Public Private Both Elasticities earnings earnings earnings earnings earnings earnings Panel B: Hours elasticity Public hours 0.40*** -0.46*** ** (0.11) (0.07) (0.12) (0.19) (0.11) (0.18) Private hours -0.43*** 0.54*** *** 0.09 (0.10) (0.08) (0.11) (0.20) (0.17) (0.19) Total hours (0.01) (0.01) (0.02) (0.04) (0.02) (0.06) Note: Bootstrap standard errors shown in parenthesis. Significance: *** 1%; ** 5%; * 10%. 12 / 19

13 Results - Hours elasticity by age Table 6: Male hours elasticities with one-percent increase in hourly earnings by age Male Public Std Private Std Both Std Elasticities earnings Err earnings Err earnings Err Public hours Age years 0.39*** (0.13) -0.42*** (0.12) (0.09) Age years 0.56*** (0.12) -0.62*** (0.08) (0.12) Age years 0.48*** (0.12) -0.55*** (0.10) (0.11) Age 60+ years 0.12 (0.13) (0.12) (0.10) Private hours Age years -0.52*** (0.16) 0.63*** (0.18) 0.11 (0.12) Age years -0.60*** (0.12) 0.77*** (0.09) 0.17 (0.11) Age years -0.45*** (0.11) 0.59*** (0.10) 0.14 (0.09) Age 60+ years (0.09) 0.14 (0.12) (0.03) Total hours Age years 0.02 (0.02) (0.01) (0.03) Age years 0.02 (0.01) (0.01) (0.02) Age years 0.01 (0.01) (0.01) (0.02) Age 60+ years (0.01) (0.02) (0.03) Note: Bootstrap standard errors shown in parenthesis. Significance: *** 1%; ** 5%; * 10%. 13 / 19

14 Results - Hours elasticity by age Table 7: Female hours elasticities with one-percent increase in hourly earnings by age Female Public Std Private Std Both Std Elasticities earnings Err earnings Err earnings Err Public hours Age years 0.31 (0.22) -0.36** (0.16) (0.15) Age years 0.25 (0.20) -0.33*** (0.13) (0.20) Age years 0.12 (0.25) (0.18) (0.20) Age 60+ years 0.14 (0.41) 0.08 (0.36) (0.18) Private hours Age years -0.55* (0.32) 0.78** (0.39) 0.23 (0.23) Age years (0.23) 0.47** (0.19) 0.11 (0.20) Age years (0.25) 0.17 (0.24) (0.21) Age 60+ years 0.04 (0.35) (0.47) (0.22) Total hours Age years 0.05 (0.06) (0.02) 0.04 (0.08) Age years 0.01 (0.04) (0.03) 0.01 (0.06) Age years (0.04) (0.03) (0.06) Age 60+ years (0.05) (0.06) (0.10) Note: Bootstrap standard errors shown in parenthesis. Significance: *** 1%; ** 5%; * 10%. 14 / 19

15 Results - Hours elasticity by specialty Table 8: Male hours elasticities with one-percent increase in hourly earnings by specialty Male Public Std Private Std Both Std Elasticities earnings Err earnings Err earnings Err Public hours Internal Med. 0.44*** (0.10) -0.61*** (0.09) -0.17* (0.09) Pathology 0.47*** (0.12) -0.66*** (0.09) (0.12) Surgery 0.59*** (0.13) -0.49** (0.21) 0.10 (0.26) Anaesthetic 0.57*** (0.13) -0.53*** (0.17) 0.03 (0.21) Diag. radiology 0.40*** (0.15) -0.58*** (0.10) (0.15) Obstetrics 0.62*** (0.13) -0.61*** (0.18) (0.24) Psychiatry 0.38*** (0.09) -0.49*** (0.06) (0.08) Other 0.49*** (0.11) -0.68*** (0.12) -0.19* (0.11) Total hours Internal Med. 0.03*** (0.01) 0.02** (0.01) 0.05** (0.02) Pathology 0.02 (0.02) 0.01 (0.01) 0.03 (0.02) Surgery 0.01 (0.01) (0.03) (0.03) Anaesthetic 0.01 (0.01) (0.02) (0.03) Diag. radiology (0.02) (0.02) (0.03) Obstetrics 0.01 (0.01) (0.03) (0.03) Psychiatry 0.03* (0.02) 0.03** (0.01) 0.05** (0.02) Other 0.01 (0.01) 0.01 (0.01) 0.14 (0.02) Note: Bootstrap standard errors shown in parenthesis. Significance: *** 1%; ** 5%; * 10%. 15 / 19

16 Summary of results Changes in earnings do not have any effect on total hours, but rather doctors respond by reallocating working hours to the sector with relatively higher earnings. Male specialists under the age of 60 years are most responsive, whilst those over 60 years are not responsive to changes in earnings. Considerable variation in labour supply response across different specialities. For a small number of specialties, increasing earnings can lead to increase in total hours worked. Family circumstances such as the presence of young dependent children influence the hours worked by female specialists but not male specialists. 16 / 19

17 T C Cheng, G Kalb and A Scott (2013). Public, private or both? Analysing Factors Influencing the Labour Supply of Medical Specialists. Institute for the Study of Labor, DP techeng@unimelb.edu.au Web: 17 / 19

18 Earnings regression Table 9: Public and private log hourly earnings regression Public Private Variables Coeff Std Err. Coeff Std Err. Female -0.06** *** 0.06 Australian medical school Temporary visa No. of postgrad qualifications *** 0.04 Experience (Ref: < 20 years) years years years years years years -0.15** *** 0.09 Significance: *** 1%; ** 5%; * 10%. 18 / 19

19 Earnings regression Table 9 (cont.) Public Private Variables Coeff Std Err. Coeff Std Err. Specialty (Ref: Internal Medicine) Pathology 0.17*** * 0.13 Surgery *** 0.08 Anaesthetics 0.22*** *** 0.07 Diagnostic Radiology 0.46*** *** 0.09 Obstetrics/Gynaecology *** 0.09 Psychiatry 0.06* * 0.07 Specialty - Other * 0.13 State (Ref: New South Wales) Victoria Queensland 0.18*** South Australia ** 0.09 Western Australia 0.11** Northern Territory Tasmania Australian Capital Territory SEIFA index -0.02* * 0.03 Constant 4.55*** *** 0.11 σ u σ e No. of observations 2,088 1,347 No. of ids 1, Significance: *** 1%; ** 5%; * 10%. 19 / 19