Transcript Do Political or Economic Factors Drive Healthcare

Do Political or Economic Factors Drive
Healthcare Financing Privatisations?
Empirical Evidence from OECD
Rasmus Wiese
[email protected]
Tammo H.A.
Bijmolt
R
[email protected]
2
Contribution
1) A novel methodology to detect healthcare privatisations
2) Empirical test of political economy theories and
conventional wisdom. Main results:
- Economic crisis trigger healthcare financing privatisations
- Political factors do not
3
Background
› Why/When do countries (not) undertake economic reform?
- Sovereign debt crisis imply a lack of reform
- Public healthcare expenditure 8% of GDP
- Profit and healthcare!
4
Motivation
› Lack of studies testing conventional wisdom and
theory
- Economic crisis, ideology, war of attrition, elections
(Drazen & Easterly 2001; Hibbs 1977; Alesina & Drazen 1991)
› Poor measurement of privatisations and reform
(Megginson & Netter 2001)
› No quantitative studies of healthcare financing
privatisations?
5
Do Political or Economic Factors Drive
Healthcare Financing Privatisations?
Presentation plan:
1) Detecting privatisations
2) Which factors triggered these privatisations?
3) Conclusion
6
Detecting privatisations
› General definitions of economic reform and privatisation
- Should have a significant economic impact and be a result of planned
policy
- A shift in ownership (e.g. Saltman 2003)
› Problems:
- Policy input or economic outcome data? (Rodrik 1996; Campos & Horváth
2012)
- Not necessarily any shift in ownership
› Specific definition of a healthcare financing privatisation:
A statistically significant policy induced shift from public to
private funding of healthcare
7
Detecting privatisations. 2 parts:
1)Detect statistically significant shifts in which “pockets” incur the
costs to healthcare. Private vs. public.
2)Validate these shifts using de jure evidence
Case study: Healthcare financing privatisations in Austria
Part 1: Data and statistical filter
›Data: Public and Private funds incurred to healthcare (OECD)
›Healthcare financing source:
publict
yt =
publict + privatet
8
Austria, h=6
0.74
0.72
0.70
0.66
0.68
privatisation
0.64
public relative to total expenditure
0.76
Healthcare financing source: Austria
1960
(Bai & Perron 1998, 2003)
1970
1980
1990
year
2000
2010
9
In Austria the filter detects:
•Privatisation in 1968
•Privatisation in 1988
Part 2: Validate the detected significant shifts
•Civil Servants’ Health and Work Accident Insurance Act of
1967
•Employment and Social Security Tribunal Act of 1987
Healthcare systems in transition, country reports (WHO)
10
Findings in a panel of 23 OECD countries
Findings:
• 33 privatisations detected by the filter
• 21 can be validated as the result of planned policy
• Evidence of ‘rigid institutions’ (Acemoglu et al. 2006)
• Evidence of ‘stroke-of-the-pen’ policies
›Privatisations in 14 out of 23 OECD countries (1960-2010)
11
Which factors triggered these privatisations?
› Panel of 23 OECD countries (unbalanced 1975-2006)
› Dependent variable privatisation
- Privatisation yit=1, no privatisations yit=0
› Binary outcome random effects logit model
12
Variables of interest
› Political factors
- Ideology (Potrafke 2009)
- War of attrition (Political fractionalisation, World Bank DPI)
- Elections (executive and legislative, World Bank DPI)
› Economic Crisis
- Job crisis (unemployment rate, mean + st.d., OECD)
- Debt crisis (interest rate on gov. debt, mean + st.d., OECD)
- Recession (year with negative growth, OECD)
13
Control variables
› Economic factors: Demographics, technological change,
inflation (Oxley & MacFarlan 1994, OECD)
- percentage of population over 65
- potential years of life lost
- inflation rate
› Duration dependence (Beck et al. 1998)
- probability of a reform is affected by earlier reforms
14
Main empirical results
Baseline specification, random-effects estimates
Dependent variable: Detected and validated privatisations
VARIABLES
(1)
(2)
(3)
(4)
(5)
Job crisis
1.575***
(0.004)
1.851**
(0.032)
1.727***
(0.005)
1.566***
(0.004)
1.778**
(0.038)
1.796***
(0.004)
0.229
(0.432)
1.511***
(0.006)
1.790**
(0.040)
1.731***
(0.005)
1.542***
(0.005)
1.785**
(0.041)
1.710***
(0.006)
0.085
(0.878)
1.523***
(0.006)
1.709*
(0.051)
1.785***
(0.004)
0.222
(0.474)
-0.122
(0.914)
0.067
(0.903)
23
550
23
539
Debt crisis
Recession
Ideologyt-1
Fractionalisationt-1
-0.398
(0.709)
Electiont-1
Number of countries
Observations
!
23
634
23
615
23
539
15
Conclusion
› The filter picks-up privatisations
- maybe too strict
› Economic crisis trigger privatisations robustly
1) Perception of the need for significant reform (Drazen 2000)
Uncertainty of reform outcome (Fernadez & Rodrik 1991)
2)
3) …
› Other political factors do not impact healthcare financing
privatisations: Ideology?
- Other measures of ideology
16
The end
Thank you for your attention
[email protected]