upstream and downstream approaches to inequalities in...

1

Upstream and Downstream Approaches to Inequalities in Health:

Challenges and Opportunities

George A. KaplanCenter for Social Epidemiology and Population Health

University of Michigan

©Ken Light CSEPH

Upstream and Downstream Approaches to Inequalities in Health

• The Context– The 800 lb. genome gorilla, health, and health inequalities– Social Epidemiology and Population Health– Socioeconomic inequalities in health

• Some current explorations (briefly)– The downstream/bloodstream side of inequality– The life course and cumulative disadvantage– Communities as crucibles for growing health inequality– Economic equity and health

CSEPH

2


• Challenges and Opportunities– Metaphors and Models– Need for Theory– Analytic and Methodologic Pitfalls– Perils and Promise of Interdisciplinarity– Personalizing Populations– Moderating Essentialism– Lost Opportunities

CSEPH

At the end of a century in which the average life expectancy in the United States has increased by nearly thirty years, victory over disease and disability has become an understandably popular and realistic goal. (emphasis added)

Harold Varmus, 1999

3

More than 150 years ago Rudolf Virchow, the great German pathologist, journeyed to Upper Silesia to investigate a great typhus epidemic. In his report to the Prussian government he focused not on a search for the critical bacilli, or the necessary animal vectors for transmission to humans. Instead, he called attention to the social, economic, and cultural factors responsible for the epidemic. In education and full employment, he argued, lay the cures for the prevention of future epidemics..

The primary determinants of disease are mainly economic and social, and therefore its remedies must also be economic and social. Medicine and politics cannot and should not be kept apart.

Geoffrey Rose, The Strategy of Preventive Medicine, 1992

4

Stearns Co, MN LE= 83

Baltimore Co, MD LE=73

41 year range in life expectancy between the shortest (AI males in 6 counties in SD) and the longest (A/PI females in Bergen County, NJ) is equal to 90% of the global range in life expectancy (males in Sierra Leone vs. females in Japan)(Murray et al.,1998)

5

1998

1991

1994

1996

<10% 10% to 15% >15% N/A

Prevalence of Obesity (+30 lbs) Among U.S. Adults: BRFSS, 1991-1998

Source: Mokdad et al. JAMA 1999; 282(16):1519-1522

Heart Disease Death Rates, Age 25-64 Average Annual 1979 - 1989, Men

Income B/W1980$ White Black Ratio

Less than $10,000 324.1 390.8 1.21

$10,000 - $14,999 255.4 292.8 1.15

$15,000 - $24,999 136.9 142.2 1.04

Source: Health, United States, 1998

6

Life

cour

se

Social and Economic Policies

Physical Environment

Institutions (including medical care)

Living Conditions

Social RelationshipsIndividual Risk Factors

Genetic/ConstitutionalFactors

Pathophysiologicpathways

Individual/PopulationHealth

Determinants of Population Health and Health Inequalities

l


• The Context– The 800 lb. genome gorilla, health, and health inequalities– Social Epidemiology and Population Health– Socioeconomic inequalities in health

• Some current explorations (briefly)– The downstream/bloodstream side of inequality– The life course and cumulative disadvantage– Communities as crucibles for growing health inequality– Economic equity and health

CSEPH

7

0

50

100

150

200

250

300

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

Pub

licat

ions

/yea

r

Publications with Social Epidemiology in title, abstract, or descriptors: MEDLINE, 1966-2002

What Do we Mean by Inequalities in Health?

Inequality

Disparity

Inequity

There are many potential types of inequality/disparity/inequity

In what follows I will focus on socioeconomic inequalities

8

Socioeconomic Position and Health

• Widespread• All age groups affected• Affects multiple organs and risk factors• Not fixed in time

CSEPH

Davey Smith et al.

9

Lord, 1955 17

Fatality Rates for Women Passengers on the Titanic

05

101520253035404550

1st Class 2nd Class 3rd Class

Cabin Class

Fata

lity

Rat

e (%

)

18

Number of publications per month with social class, socioeconomic factors, income or poverty listed as

descriptors: MEDLINE search, 1969 through mid-2000.

100

120

140

160

180

200

220

240

1969-72 1973-76 1977-80 1981-84 1985-88 1989-92 1993-96 1997-2000

Year

Publ

icat

ions

/Mon

th

Kaplan and Lynch, AJPH, 1997 + update 6/03/2000

10

Rogot et al., 1992 19

All-cause Mortality by IncomeNLMS, 25+ Years

020406080

100120140

White Male African-American Male

White Female African-American Female

SMR

Income Group < $5,000 $5-9,000 $10-14,999$15-19,999 $20-24,999 $25-49,999 > $50,000

0 50,000 100,000 150,000 200,000

Population Density

Relative Risk (RR) +/- 95% Confidence Interval for beta

RR = 1

Mean Income

0 50,000 100,000 150,000 200,000

1.0

2.0

1.5

2.5

0.5

Relative Risk of Death by Income & the Distribution of Income: USA, NLMS

Household IncomeMeanWolfson, Kaplan,Lynch, Ross, Backlund, BMJ, 1999

Income distribution

Relative risk of death compared to mean income

11

Deaths/10,000 in 300,685 Men: MRFIT, Age 35-57

Davey Smith et al., 1996

91.887.6 86.5

78.9 78.2

70.1 68.164.5 61.9 59 56.5 58.2

51.9 51.4

0

20

40

60

80

100

120Median Zip Code Family Income ($1,000s)

<10 10-12 12-14 14-16 16-1818-20 20-22 22-24 24-26 26-2828-30 30-32 32-34 34+



CSEPH

12

9.9

17.3

12.7

5.76.5

14.8

6.0

7.9

5.55.1

12.3

5.9 5.75.1

4.2

11.4

5.44.0

0

2

4

6

8

10

12

14

16

18

20

White, NH Black, NH Hispanic AmInd/AKNat Asian/PI

<12 12 13-15 >15

Infant Mortality Rates by Years of Education:United States, 1995, Mother 20+ years of age

Stunting* by Age and Poverty Status

USDHHS, 1986

12.9

6.87.4

5.23.9 3.7

2-5 6-11 12-17

Age

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

16.0Percent

Poverty Non-Poverty

*height<5th percentile

13

0

1

2

3

4

5

6

7

8

All White Af-Am

<$20K$20-29.9K$30-39.9K$40K+

Asthma Hospitalization Rate among children 1-4 years: U.S,.1989-91

Wolfson et al., 1993 26

Survival from Age 65 by Pre-Retirement Earnings Quintile546,759 Canadian Males in Canada Pension Plan

0.70

0.75

0.80

0.85

0.90

0.95

1.00

65 66 67 68 69 70 71 72 73 74Age

Surv

ival

Pro

babi

lity

Income Quintile (Earnings)1 (<$14,494)2 ($14,494-$22,278)3 ($22,279-$27,990)4 ($27,991-$35,986)5 (>$35,986)

1234

5

14



CSEPH

Chronic Conditions more Prevalent among those with < 12 Years of Education: NHIS, 1989, 65+ years

15

Lynch et al., 1997 30

4-year Progression of Plaque Height by Education and Income Level

0.2

0.22

0.24

0.26

0.28

0.3

0.32

<HighSchool

SomeHigh

School

HighSchool+

I(Lowest) I III IV4-Y

ear

Cha

nge

in M

ax IM

T (m

m)

Education Income

16


• Some current explorations (briefly)– The downstream/bloodstream side of inequality– The life course and cumulative disadvantage– Communities as crucibles for growing health

inequality– Economic equity and health

CSEPH

AnginaPectoris

MyocardialInfarction

Athero-sclerosis

Thrombosis

Lipidperoxidation

Othermechanisms

High LDL

Low HDL

High pro-oxidantsto antioxidants ratio

ObesityInsulinresistance

Hypertension

Low PA

Diet

SmokingPsychosocial

Factors

SES

17

Tissue Plasminogen Activator

0.10.20.20.30.3

1 (low) 2 3 4 (hi)

Income QuartileT

PA I

U/m

l

Fibrinogen and TPA by Income QuartilesKuopio Ischemic Heart Disease Risk Factor Study

Fibrinogen

2.72.82.93.03.13.2

1 (low) 2 3 4 (hi)

Income Quartile

Fibr

inog

en g

/l

LDL Cholesterol

3.5

3.8

4.0

4.3

1 (low) 2 3 4 (hi)

Income Quartile

LDL

mm

ol/ll

Total Serum Cholesterol and LDL by Income QuartilesKuopio Ischemic Heart Disease Risk Factor Study

Total Serum Cholsterol

5.55.65.75.85.96.06.1

1 (low) 2 3 4 (hi)

Income Quartile

Seru

m C

hole

ster

ol

mm

ol/l

18

Fasting Blood Glucose

4.5

4.8

5.0

1 (low) 2 3 4 (hi)

Income QuartileB

lood

Glu

cose

m

mol

/l

Systolic Blood Pressure and Fasting Glucose by IncomeQuartiles

Kuopio Ischemic Heart Disease Risk Factor Study

Systolic Blood Pressure

130.0

132.0

134.0

136.0

1 (low) 2 3 4 (hi)

Income Quartile

SBP

mm

Hg

130

135

140

145

150

155

160

Income Quintiles

WB

S (n

g/m

L)

Income and Whole Blood Serotonin in the KIHD Study

Everson-Rose et al., in progress

19

0102030405060708090

100

Pack-years(-100) % Drunk>2-3/mo % Sedentary

Primary or lessSome High SchoolHigh School or more

Behavioral Risk Factors by Level of EducationKuopio Ischemic Heart Disease Risk Factor Study

Lynch, Kaplan & Salonen, 1997

Sodium

2800290030003100320033003400

1 (low) 2 3 4 (hi)

Income Quartile

Na

(mg)

P/S Ratio

0.220.230.240.250.260.270.280.290.3

1 (low) 2 3 4 (hi)

Income Quartile

P/S

ratio

Vegetables

0

50

100

150

1 (low) 2 3 4 (hi)

Income Quartile

g/d

Fruits & Berries

0

50

100

150

200

2501 (low) 2 3 4 (hi)

Income Quartile

g/d

Dietary Consumption by Income QuartilesKuopio Ischemic Heart Disease Risk Factor Study

20

0

5

10

15

20

25

30

Depressio

n

Hopele

ssness

Hostilit

y

Sense of C

oherenc

e

Age

adj

uste

d Pr

eval

ence

(%)

Low Medium High

Lynch, Kaplan & Salonen, 1997

Prevalence (%) of Psychosocial Factors by EducationKuopio Ischemic Heart Disease Risk Factor Study: 42-60 year-old men

Prevalence of Depressive/Hopelessness Symptoms by Income Tertiles

0

5

10

15

20

25

30

ConsumersSurvey

Detroit Study Alameda Study Kuopio Study

Low Middle High

Everson et al., 2000

21

41

Prevalence of Hopelessness by Income Quintiles

0

5

10

15

20

25

1st (L

owest

)2n

d3r

d 4th

5th (H

ighest

)

Quintiles

Prev

alen

ce o

f Hop

eles

snes

s (%

)

4-Year Progression of Carotid Atherosclerosis by Hopelessness

Adjusted for Age and Baseline IMT

0.20

0.22

0.24

0.26

0.28

0.30

0.32

0.34

Low/Moderateat Baseline

Differing Levesat Baseline

High Levels atBaseline

4 Year Follow-Up of HopelessnessM

axim

um C

hang

e in

Intim

a-M

edia

Thi

ckne

ss (m

m

Kuopio Ischemic Heart Disease Study, Everson et al., 1997




CSEPH

22

Risk Factors &Environment

Adult Health Birth Outcomes

Child Health

“The childhood shows the man, As the morning shows the day”

John Milton. Paradise Lost (1667), Line 220-221.

0

0.2

0.4

0.6

0.8

1

1.2

Males Females

Hea

rt D

isea

se M

orta

lity

re: <

5.5

<5.5 6- 7- 8- 9- >10Birth Weight

Coronary Heart Disease Mortality in Hertfordshire Adults by Weight at Birth (lbs)

Osmond et al. , 1993

23

Adult Cognitive Function

Childhood Socioeconomic

Position

Mother’s Education

Mother’s Occupation

Education

Father’s Occupation

Father’s Education

Early Environmental Events and Later DevelopmentVarious studies indicate the following more common for poor children:HomelessnessPoor and unaffordable housingResidential mobilityInadequate heatingCrowdingCold, dampness, moldCockroaches, rats, micePoor quality child careDecreased verbal interactions with adultsInadequate schoolsFewer educational opportunities at homeFew stimulating activities at homeParental stress and depression

24

90.8

113.1

137.5

173.5

50

70

90

110

130

150

170

190

0 1 2 3

Tra

il-m

aker

Tes

t (se

cs)

Increasing Life Course Economic Disadvantage

Cognitive Function at Age 58 & 64 and Life Course Socioeconomic Disadvantage

worse

better

0

1

2

3

4

5

6

Disability Depression Pessimism Hostility CognitiveProblems

Incr

ease

re:

0 ti

mes

0 1 2 3

No.of Times <200% of Poverty

29-year Cumulative Impact of Economic Disadvantage on Five Health Outcomes

Lynch, Kaplan & Shema, NEJM, 1997

25

Washington, DC 1952New York City, 1952

Grandmothers of the Current Generation

Barrington, 2003

1

2.93

0.88

2.55

1.14

2.39

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

High Low

Adult SES

OR

: re

Hig

h/H

igh High Mid Low

Childhood SES

Childhood and Adult SES and Mortality from All CausesKuopio Ischemic Heart Disease Risk Factor Study

Lynch et al., 1994

26




CSEPH

Haan et al., 1987 52

0.80

0.85

0.90

0.95

1.00

0 1000 2000 3000 4000Days

Prop

ortio

n Su

rviv

ing

Non-Poverty Area

Poverty Area

Age-adjusted Survival by Poverty Area Residence: Alameda County Study, 1965-1974

Survival curves from Cox model, estimated at age=60

27

Those who lived in the poverty area over the next 9 years:

• had twice the decrease in physical activity

• were more likely to become depressed

• were twice as likely to become disabled

• were less likely to be “successfully” aging

Those in the Poverty neighborhoods

• Were 5.8-8.0 lbs (M and F) heavier and gained 2.4-3.0 lbs (M & F) more over the next 9 years

(after adjustment for age, race, education & income)

• They were 3-8 times (depending on gender, race, and income) to develop NIDDM over the next 34 years

28

1.00 1.00

2.63

1.43

0

0.5

1

1.5

2

2.5

3

Age Age+Income

Diabetes Mortality Differences between Whites and African-Americans before and after adjustment for Median Income of Area of Residence:

MRFIT: 320,909 men aged 35-47

Mor

talit

y R

isk

re: W

hite

75% reduction in risk with adjustment for median income of area


• The 800 lb. genome gorilla and health inequalities• Socioeconomic inequalities in health• Some current explorations (briefly)

– The downstream/bloodstream side of inequality– The life course and cumulative disadvantage– Communities as crucibles for growing health


CSEPH

29

U.S. Income Inequality

“The gap between rich and poor has grown into

an economic chasm so wide that this year the

richest 2.7 million Americans, the top 1 percent,

will have as many after-tax dollars to spend as

the bottom 100 million.”

NY Times, Sept 5, 1999

30

650

700

750

800

850

900

950

1000

17 18 19 20 21 22 23 24

Income Share Held by Poorest 50% of the Population

Age

-adj

uste

d T

otal

Mor

talit

y

Income Inequality and Mortality in US States, 1990(adjusted for State Median Income)

LA MS

NY

AL

KYWV

TX

TNAR

GA

OK

NM

FL

MO

CA

IL

AZ

SC

PAMI

MA

NCOH

VANJ

CT

MTKS

CO

RI

SD

OR

ND MNID

NEWA

NV

WY

IN

IA

ME

HI

DE

MD

AK

WI

VT

UT

NH

r = - 0.62

Kaplan et al. BMJ (1996)

Between 1980 and 1990, states that had the biggest increase in income inequality had the smallest decrease in mortality rates

r

% < High School Diploma -0.71

% High School Dropout -0.50

4th Grade Reading Scores -0.58

4th Grade Math Scores -0.64

Education / Total Spending -0.32

Library Books Per capita -0.42

Income Inequality and Selected Educational Indicators

31

4th (High)

3rd

2nd

1st (Low)

750

800

850

900

950

4th (High)

3rd

2nd

1st (Low)

Income Inequality Quartiles

4th (High) 3rd 2nd 1st (Low)

Per CapitaIncome

Quartiles

MortalityRate per100,000

High Inequality &Low Income

Low Inequality &High Income

895

786

926

812

Lynch, Kaplan,Lynch, Kaplan, Pamuk Pamuk, et al. AJPH (1998), et al. AJPH (1998)

How large are these mortality effects ?

High Income Inequality

Low Average IncomeLow Income Inequality

High Average Income

925.8 per 100,00 785.9 per 100,00

140 deaths per 100,000Equivalent to the combined loss of life from Equivalent to the combined loss of life from

lung cancer, diabetes, motor vehicle crashes, lung cancer, diabetes, motor vehicle crashes, HIV infection, suicide and homicide in 1995.HIV infection, suicide and homicide in 1995.

Lynch, Kaplan, Pamuck, et al, AJPH, 1998

32

0.15 0.20 0.25 0.30Median Share of Income

200

300

400

500

600W

orki

ng-A

ge (2

5-64

) Mor

talit

y R

ate

per 1

00,0

00

Income Inequality and Working Age MortalityUSA, Canada, Australia, Sweden Metropolitan Areas1990/91

US metro areas n=282CAN metro areas n=53AUS metro areas n=18SWE metro areas n=40

Source:Canada: Ross, Dunn, WolfsonUS: Lynch and KaplanAustralia: GloverSweden: Henricksson

Ross et al., in progress

DO NOT CITE



CSEPH

33

````


Neighborhoods and CommunitiesInstitutions (including medical care)

Living Conditions


Genetic/Constitutional Factors


Individual/Population Health

Pathophysiologic pathways

Life

cour

se


Political EconomyInstitutions

DiscriminationCultureHistory

Distal Social Connections

Family

StructuralMacrosocial

Factors

Proximal Social Connections

Neighborhood CommunityOrganizational Connections

Friends

Individual Characteristics

SocioeconomicBehavioralPsychosocial

Life Course

Conception Old Age

Human BiologyGenetics

Pathological Biomarkers

GeneticCharacteristics

Pathobiology

Health Status

Work

Lynch, 2000

34

Material, PsychosocialSocial Structure

PathophysiologyOrgan Impairment

Early Life

DMCHD

Wellbeing

Culture

Genes

Work

Psychological

CNS

Neuroendocrine

ImmuneSocial Environment Health

Behavior

Marmot, 2001

Race-Ethnicity

Gender

SocioeconomicStatus

Health Outcomes1. Mortality2. Institutionalization3. Morbidity (Chronic)4. Functional Limitations5. Self-rated Health6. Cognitive Function7. DepressionEXPLANATORY VARIABLES

Medical Care and Insurance

Psychosocial Risk Factors1. Health Behaviors2. Social Relationships

and Supports3. Chronic and Acute

Stress 4. Psychological

Dispositions5. Social Roles and

Productive Activities

Physical/Chemical and Social Environmental Hazards

Social, Political, and Economic Conditions

and Policy

House, 2002

35



CSEPH

36



CSEPH



CSEPH

37



CSEPH



CSEPH

38



CSEPH

In a world of science in which the proximal and molecular explanation is seen as the “best,” a complete, satisfying, and useful understanding of inequalities in health may require the methodologic lens of complex systems and systems biology (even), the sensitivity and insights of the historian, poet and ethnographer, the knowledge of the biologist and ecologist, and the tools of the policy analyst and public health practitioner.

39

While there are many challenges, the study and remediation of inequalities in health coiuld serve as a model for a new scientific paradigm– bridging the social and the biological.

A tall order, but an exciting one that holds promise for breaking the link between the social divides in society and the health divides, and in doing so perhaps (???) even helping to decrease the social divides……..

78

````


Neighborhoods and CommunitiesInstitutions (including medical care)

Living Conditions




Individual/Population Health

Pathophysiologic pathways

Life

cour

se


41




inequality– Economic equity and health– Globalization and health– How do we understand?

• 1.2 billion people (20% of world’s population) are estimated to live on less than $1 per day

• 3 billion (49% of world’s population) on less than $2 per day

• 110 million primary school age children are out of school (60% of them girls)

• Many more live without adequate food, shelter, safe water, and sanitation.

WHO, Global Poverty Report, July, 2000

42

Social Impacts of 1998 Asian Financial Crisis in Thailand

Increases in:

• poverty

• unemployment

• divorce rates

• child abandonment

• drug use

Shivakumar et al., 2000

Debt Service vs. Other Expenditures in Pakistan

Debt Service(350B)

Development

Health Budget

Bhutta, 2000

43

Into Out of Net

World Bank 250 514.2 -264.2

IMF 0 329.1 -329.1

Cash Flow in and out of Pakistan from World Bank and IMF, 1999-2000

Bhutta, 2000

Globalization of the EconomyCapital ConcentrationLeveraged buy-outs, etc.

Changes inOccupational Structure

Downsizing/Job Losses

Deforestation

Meat Production

FarmingMigration

SocialCohesion

Blood PressureSmokingObesity

Flue-curing oftobacco

Real WagesWork HoursJob InsecurityIncome and

Wealth Inequality

Social Resources

Stress &Time Pressure

Other Risk Factors

Convenience &Fast Food

Poor Nutrition

High-strain, minimum wage jobs

CardiovascularDisease

Epidemiologic Transitionin Developing Countries

44




inequality– Economic equity and health– Globalization and health– How do we understand?

46

New UrbanismSmall, self-contained neighborhoods with the center no more than a quarter of a mile from the edge -- a reasonable walking distance. Improved public transit systems and greater integration of different types of land uses at the neighborhood level.

vs

47

9-10 fodiffere

$137 vs $13,995

100-fold difference!!

48

LBW

Low Maternal Education

Maternal Age

Prenatal Care

Health Insurance

Individual Level

Unemployment

Per Capita Crime

Wealth

Low Per Capita

Income

Neighborhood Level

Interactions Between Neighborhood and Individual Risk Factors for Low Birth Weight

O'Campo et al 1997

49

9-10 fodiffere

Civic CultureSocial Values

IndividualIncome

MorbidityMortality by

Cause

SmokingDiet

ExerciseHostilityIsolation

PoliticalEconomy

Distributionof Income

Historical Factors

Political Institutions

Neo-Material Context

Education SystemWorking ConditionsIndustry BasePublic Health ServicesEnvironmental ControlsLand Use & ZoningRecreation FacilitiesTransportation SystemsAffordable Quality HousingAccess to Optimal NutritionChild CareAdvertising - cigarettes, etcDiscrimination

Perception of Placein the Social Hierarchy

DisrespectLack of Trust

Social CohesionSocial Capital

1

2

3

50

Trends in Inequality and Child Poverty 1967-1992

%Change in %Change in Inequality Child Poverty

UK + 30 + 30USA + 15 - 29 + 30 Sweden + 15 - 29 - 5Australia + 10 - 15 0Denmark + 10 - 15 - 5 Canada 0 - 5Finland 0 - 5Spain 0 0Israel 0 + 5 - 10W. Germany 0 + 5 - 10

How large are these mortality effects ?

High Income Inequality

Low Average IncomeLow Income Inequality

High Average Income

925.8 per 100,00 785.9 per 100,00

140 deaths per 100,000Equivalent to the combined loss of life from Equivalent to the combined loss of life from

lung cancer, diabetes, motor vehicle crashes, lung cancer, diabetes, motor vehicle crashes, HIV infection, suicide and homicide in 1995.HIV infection, suicide and homicide in 1995.

Lynch, Kaplan, Pamuck, et al, AJPH, 1998

51

Civic CultureSocial Values

IndividualIncome

MorbidityMortality by

Cause

SmokingDiet

ExerciseHostilityIsolation

PoliticalEconomy

Distributionof Income

Historical Factors

Political Institutions

Neo-Material Context

Education SystemWorking ConditionsIndustry BasePublic Health ServicesEnvironmental ControlsLand Use & ZoningRecreation FacilitiesTransportation SystemsAffordable Quality HousingAccess to Optimal NutritionChild CareAdvertising - cigarettes, etcDiscrimination

Perception of Placein the Social Hierarchy

DisrespectLack of Trust

Social CohesionSocial Capital

1

2

3

Trends in Inequality and Child Poverty 1967-1992

%Change in %Change in Inequality Child Poverty

UK + 30 + 30USA + 15 - 29 + 30 Sweden + 15 - 29 - 5Australia + 10 - 15 0Denmark + 10 - 15 - 5 Canada 0 - 5Finland 0 - 5Spain 0 0Israel 0 + 5 - 10W. Germany 0 + 5 - 10

upstream and downstream approaches to inequalities in...

Documents