Transcript Stuckler_EUPHA_11_28.. - Centre for History and Economics

What is the evidence of the impact
of the economic crisis on public
health?
David Stuckler, PhD MPH
Oxford and LSHTM
With
Marc Suhrcke, Sanjay Basu, Adam Coutts and Martin McKee
Disaster?
“Econocide to surge as
recession bites”
– BBC Mar 2009
Blessing?
“Recession is a lifestyle
blessing in disguise”
– Times Oct 2008
Great Depression:
not so bad?
Source: Hanley 1931 Life
Sources: NYT Oct 1930; AP 1932; WP 1930
Evidence
- Income ↓ 30%
- All-Cause Mortality
Rates ↓ 10%
- Suicide ↑40%
New Deal
Prohibition
Post-Soviet Depression: devastating
Alcohol and Surrogates
Prisons and Tuberculosis
Rising Unemployment
Overview
- Income↓ 30%
- All-Cause Mortality
Rates ↑ 20%
- Suicide ↑about 40%
But important variations:
Worse in countries with:
Russia & Belarus
- rapid economic change,
- weak social protections,
- low social capital
Cuba, Finland:
- Incomes fell but health
stable
Source: Stuckler et al 2009 Mass privatisation and the post-communist mortality crisis Lancet
IMF Programs and TB Control
- Gov Spending↓ 8%
- Doctors ↓ 8%
- Directly Observed
Therapy avg. ↓ 40%
- TB incidence ↑ 17%
Source: Stuckler et al 2008 IMF Programs and Tuberculosis outcomes PLoS Med
Lessons?
Social
Protection
Social
Cohesion
Drugs and
Alcohol
Unemployment-Mortality
1% rise in
unemployment
associated with:
- 0.8% ↑Suicide
- 0.8% ↑Homicide
- 1.4% ↓Traffic
death
No effect on allcause mortality
Source: Stuckler et al 2009 Lancet
Some populations more vulnerable
Average Labour Market Protection: $88 per head
1985
1990
2000
15
1980
2005
1985
1990
8
6
2000
2005
Year
Year
Suicide Male
1995
4
25
20
15
10
1995
2
30
25
20
Suicide Rates per 100,000 (under 64)
12
11
10
9
8
1980
10
Sweden
Average Labour Market Protection: $362 per head
Unemployment Rate (Labour Force Surveys)
Spain
Suicide Male
Unemployment Rate
Unemployment Rate
United Kingdom
10
12
8.5
8
8
6
7.5
4
7
2
6.5
1970
1980
Suicide
1990
Year
2000
Unemployment Rate
2010
Unemployment Rate (Official)
Suicide Rates per 100,000 (under 64)
Average Labour Market Protection: $95 per head
Social Protections Help…
Each 100 USD
greater social
spending reduced the
effect on suicides by:
- 0.38%, active labour
market programmes
- 0.23%, family support
- 0.07%, healthcare
- 0.09%,unemployment
benefits
Source: Stuckler et al 2009 Lancet
Spending> 190 USD no effect of
unemployment on suicide
Recommendations
• Active Labour Market Programmes
• Health Stimulus Funding
• Evidence-Based Crisis Response Plan
An opportunity for whom?
Appendix
East Asian Financial Crisis:
it depends
Sources: Hopkins 2006 Health Policy; Chang 2009 Soc Sci Med
Implications for Europe?
1) Social spending to re-integrate and prevent
job loss appears effective
2) Effects not as bad as some suggest
3) Potential long-term health effects (nutrition)
4) Vulnerable groups at greater risk (withincountry inequality)
5) Vulnerable populations, budget cuts (e.g. IMF
programs)
Rhetoric or reality?
-“Grave debt crisis”;
“public finances out of
control”
BUT
-Debt/GDP lower than
US, France, Germany,
EU average
-About same as 10
years ago
UK Debt
AND
- Temporary deficits OK
Dow and Deficit
Future Research?
1) Evaluate Prohibition & New Deal
2) Investigate East Asian Crisis
3) Iceland, nutrition & dietary change
4) Health co-benefits to social protections
5) Study IMF programs effects
6) Speculative finance and commodity bubbles
Some Underlying Publications
Stuckler, D., S. Basu, M. Suhrcke, A. Coutts, M. McKee. 2009. The public health effect of economic crises and
alternative policy responses in Europe: an empirical analysis. The Lancet. 374(9686):315-23
Stuckler, D., L. King, and M. McKee. 2009. Mass privatization and the postcommunist mortality crisis: a crossnational analysis. The Lancet. 373(9661): 399-407
Stuckler, D, S. Basu, M. Suhrcke, A. Coutts, M. McKee. 2009. The health implications of financial crisis: A review
of the evidence. Ulster Medical Journal. 78(3): 142-5
Lock, K., D. Stuckler, K. Charlesworth, M. McKee. 2009. Potential causes and health effects of rising global food
prices. British Medical Journal. 339:b2403
Stuckler, D and S. Basu. 2009. The International Monetary Fund's effects on global health: before and after the
2008 financial crisis. International Journal of Health Services. 39(4): 771-81.
Stuckler, D., S. Basu, and L. King. 2008. International Monetary Fund programs and tuberculosis outcomes in
post-communist countries. Public Library of Science Medicine 5(7): e143
Stuckler, D., L. King and A. Coutts. 2008. Understanding privatisation's impacts on health: Lessons from the
Soviet Experience. Journal of Epidemiology and Community Health 62(7): 664.
Stuckler, D., S. Basu, M. McKee, L. King. 2008. Mass incarceration can explain population increases in TB and
multi-drug resistant TB in European and central Asian countries. Proceedings of the National Academy of
Sciences. v105(36):13280-5.
Stuckler, D., C. Meissner and L. King. 2008. Can a bank crisis break your heart? Globalization and Health 4(1):112.
Stuckler, D. 2008. Population causes and consequences of leading chronic diseases: A comparative analysis of
prevailing explanations. The Milbank Quarterly 86(2):273-326.
I. Great Depression
When the full record of the year 1931 is written, no chapter will
be more striking than the one dealing with public health. The
country was in the grip of the most serious industrial
depression in a generation...By all the signs and all the
precedents, hard times so seriously prolonged should have
brought in their train disease and death.
Actually, 1931 was one of the healthiest years in the history of
the country. The evidence is overwhelming.
– New York Times, Jan 5th, 1932
Empirical Setup
Sample and Period: 114 U.S. Cities, 1929-1937
‘Shock’: Bank Suspensions: Federal Deposit Insurance Corporation
– percentage of banks suspended in calendar year
Mortality: U.S. Bureau of the Census
– Crude death rates for >30 causes
– Epidemiology: city-level, weighted U.S. urban populations
– Econometric: state-level, weighted U.S. state means
Economic and Demographic Controls:
–Urbanisation Rates, Literacy Rates, Percentage Black Ethnicity, Age-structure
Fixed Effects + Hodrick Prescott Filter
Thanks to Price Fishback for collecting data
Immediate Causes of Death
Effects of the Great Depression
Potential Role of Prohibition
II. Postcommunist Mortality Crisis
Three million excess deaths
Privatisation & Mortality
Unemployment: A mediating factor
Adjustment for unemployment reduces privatisation coefficient by :
-10% (mass privatisation program)
- 25% (1 unit on EBRD scale)
Social Capital: A protective factor
Empirical Setup
Sample and Period: 26 Postcommunist CEE and FSU Countries, 19912002
‘Shock’: Mass Privatization: EBRD and WB
– transferring at least 25% of large-state owned enterprises to the private
sector within 2 years
– jump from 1 to 3 on EBRD large-scale privatisation index
Mortality: WHO and UNICEF
– Age-standardised mortality rates in working-age men (15-59)
– Life expectancy, alcohol-related mortality rates, suicide rates
Economic and Demographic Controls: WB, EBRD, ILO
– Price liberalisation, Trade liberalisation, GDP per capita, Democratisation
Index, Dependency Ratio, Urbanisation Rates, Education Levels
Two Way Fixed Effects + Control Function
AMRit    PRIVit  Xit   it  i  t  it
Clue #1: Working-Ages
Clue #2: Men
Russian Life Expectancy
Clue #3: Immediate Causes: Heart Attacks,
Accidents, Suicides
Clue #4: Healthcare Collapse?
700
600
Russia
deaths/100,000
• Death rate from avoidable
mortality in UK and Russia
similar in 1965, when little
could be done
• Gap began to widen in
1970s, and has continued to
do so since
• It rose rapidly in the early
1990s before falling back
500
400
300
200
UK
100
0
1965
1975
1985
Year
1995
Shock Therapy versus Gradualism
“Despite economists’ reputation for never being able to agree
on anything, there is a striking degree of unanimity in the
advice that has been provided to the nations of Eastern
Europe and the former Soviet Union (FSU). The legions of
economists who have descended on the formerly Communist
economies have provided advice very similar … The three “ations”—privatization, stabilization, and liberalization —must all
be completed as soon as possible.”
Lawrence Summers (1994: 252-253)
Potential Health Effects of Mass Privatization
Mass Unemployment
Mass Privatization
Reduced Social
Provision
Firm Failure
↑ Adult mortality
↑ mortality
↑ mortality
“… I was the person sent to Russia to live in the factories and feed back
information to the bank. I spent months in the factories. In one, I slept in an
empty ward in the sanatorium. I realised very quickly that the master plan of
privatising Russian industry overnight was going to impose huge costs on
hundreds of thousands of people. These factories were producing goods that
once they were launched, no one would want in a too competitive market. They
would have to slash tens of thousands of jobs. But also, these factories provided
schools, hospitals, health care and retirement - cradle to grave. I raised these
concerns in Washington, to say there weren't any safety nets in place. It became
clear to me that it was really a political play, that they wanted to take assets out
of the state's hands, so the Communist Party wouldn't come back.”
-- WB administrator overseeing Mass Privatization in Russia
Russia & Belarus
Mass Privatization and Life Expectancy
Postcommunist Countries
Mass Privatizers
Non Mass Privatizers
Life Expectancy at Birth (years)
71
1991
70
69
1994
1997
68
2000
Year
Data Source: WHO European Health for All Database 2007
Table 2. Effect of Privatization on Male Life Expectancy in
Transition Countries, 1991-2002
Covariates
Mass Privatization
EBRD Average
Privatization Index
EBRD Cumulative
Privatization Index
Log(GDP per capita)
EBRD Price
Liberalization Index
Heritage Foundation
Democracy Index
Military Conflict
Percentage of
Population Urban
Population
Dependency Ratio
Percentage of
Population with
Tertiary Education
(1)
(2)
(3)
-1.58
(0.23)***
–
–
–
-0.49
(0.18)**
–
–
–
1.61
(0.30)***
-0.22
(0.15)
0.24
(0.05)***
-0.74
(0.29)**
-0.63
(0.14)***
-0.08
(0.08)
1.81
(0.30)***
-0.26
(0.14)
0.25
(0.06)***
-0.64
(0.30)*
-0.45
(0.13)***
-0.15
(0.06)
-0.03
(0.03)
1.79
(0.35)***
-0.28
(0.15)
0.22
(0.06)***
-0.63
(0.31)*
-0.39
(0.14)**
-0.16
(0.07)*
-0.01
(0.01)
-0.02
(0.02)
-0.01
(0.02)
Working-Age Men
Table 3. Effects of Mass Privatization and Log GDP per Capita on Log Male Death Rates
in 25 Transition Countries by Age, 1991-2002
Covariates
Mass Privatization
Log GDP per
capita
Infanta
Under-5a
5-14
15-59
60+
2.92%*
(1.44)
-0.10%***
(0.01)
2.00%
(1.49)
-0.11%***
(0.01)
0.20%
(0.95)
-0.05%***
(0.01)
13.51%***
(2.31)
-0.14%***
(0.03)
1.32%***
(0.38)
-0.02%**
(0.01)
Path Analysis: Alcohol and Unemployment
Web Annex A2. Effects of Mass Privatization and Log GDP per Capita on Pathways of Population Health Impacts
Determinant
Mass Privatization
Log GDP per Capita
Alcohol
Consumption
(Liters per Capita)
Log Male
Alcohol Death
Rates
Log Male
Heart Disease
Death Rates
Log Male
Suicide Rates
Homicide
Rates
Log Male
Unemployment
Rates
Log Crime
Rates
0.81
(0.24)***
14.64%
(4.21)**
5.98%
(1.85)***
16.86%
(4.51)***
5.60
(1.85)**
44.83%
(13.62)***
4.19%
(4.13)
-0.04
(0.30)
-0.11%
(0.06)
-0.10%
(0.02)***
0.04%
(0.08)
-0.88
(2.45)
-0.66%
(0.15)***
-0.23%
(0.05)***
Note: Robust panel-corrected standard errors in parentheses. Results presented from seven separate regression models. Two-way fixed effects models, using PraisWinsten transformation to adjust for country-specific serial correlation, control for the effects of EBRD price liberalization index, occurrence of military conflict,
percentage of population urban, age-dependency ratio, and percentage population with tertiary education.
* = p<0.05, ** = p<0.01, *** = p<0.001 (two-tailed tests).
Job Loss & Health: Evidence from RLMS
- Hazard Dying: ↑~3-4-fold
- Odds Excess Vodka: ↑40%
Who loses jobs?
Chance job loss next
year:
10.4% if privatized
9.1% if Russian
private
7.9% if foreign
6.2% if state
- 40% higher risk at
private firm
Failing Tuberculosis Control
Inaccurate and Misleading Data
Life expectancy at birth, in years
71
70
69
68
67
66
65
64
63
62
1970
1980
1990
2000
2010
Russia’s regions: pace of economic change and LE
Table 3. Vodka consumption per capita
Year
Belarus
Russia
1990
3.6
3.6
1991
4.4
4.1
1992
4.8
3.8
1993
5.9
4.9
1994
5.6
5.5
1995
5.5
7.0
1996
6.4
5.5
1997
6.2
5.5
1998
5.8
6.0
1999
5.1
6.5
2000
4.2
6.5
2001
3.8
6.3
2002
3.7
6.2
2003
3.2
6.2
Table WB7. Effect of Male Adult Mortality and Life
Expectancy on Log GDP per Capita
Health Measure
Adult Male Mortality Rates
Male Life Expectancy
Effect on Log GDP per Capita
-0.45%***
(0.12)
0.04***
(0.01)
Note: Coefficients calculated as semi-elasticities presented; Robust standard errors in
parentheses. Models also control for EBRD price-liberalization index, EBRD trade
liberalization index, democratization, population dependency, the percentage of
population living in urban settings, education rate, and country-specific fixed effects.
Significance at * p-<0.05, ** - p<0.01, *** - p<0.001
Probability of surviving from 20 to 60
W Germany
England & Wales
Lithuania
Latvia
Women
Estonia
Men
Ukraine
Russia
Belarus
0
0.2
0.4
0.6
0.8
1
Health Resources
Covariates
Model 1:
Physicians
Model 2:
Nurses
Model 3:
Dentists
-3.92**
(1.37)
Model 4:
Hospital
Beds
-0.72**
(0.20)
Model 5:
Inpatient
Admissions
0.08
(0.36)
Mass Privatization
-0.15*
(0.07)
-30.31
(16.32)
GDP
0.10
(0.07)
-44.31*
(18.61)
4.00**
(1.47)
0.22
(0.22)
2.05**
(0.41)
Urbanization
0.07**
(0.02)
18.21**
(5.23)
0.41
(0.42)
0.32**
(0.06)
0.74**
(0.11)
Dependency
-0.01
(0.01)
6.21*
(2.66)
0.70**
(0.22)
0.16**
(0.04)
0.11
(0.06)
Fertility
0.11
(0.11)
37.99
(23.81)
-1.89
(2.01)
1.53**
(0.33)
2.98**
(0.58)
Price Liberalization
-0.06
(0.03)
-18.40
(8.69)
1.24
(0.70)
-0.31**
(0.10)
-0.34
(0.19)
Education
0.00
(0.00)
0.81
(0.72)
0.16**
(0.06)
-0.01**
(0.00)
0.11**
(0.02)
Number of
Observations (NxT)
286
296
299
278
303
Number of
Countries (N)
26
25
25
26
26
R2
0.93
0.94
0.96
0.94
0.96
Note: Models 1 scaled per 1,000 population, Models 2-4 per 100,000 population,
Model 5 per 100 admission; Constant not reported; Fixed Effects with country effects
presented, period effects do not alter results; Prais-Winsten transformation used to
accommodate AR(1) error structure; * = p<0.05; ** = p<0.01 (two-tailed tests)
In writing about Stalinism, John Maynard Keynes
observes that “We have a fearful example in
Russia today of the evils of insane and
unnecessary haste. The sacrifices and losses of
transition will be vastly greater if the pace is
forced…”
Yale Review, 1933
IV. Recessions in Europe
Men
Women
Other Measures
Interaction Term