SOC 1046 midterm

- A group of people
- Based on: location, interest, identities, age (biological), socioeconomic status

- Scatter diagram of relations between life expectancy at birth and national income per head for nations throughout the decades - Preston interpreted this shift as the effect of medical progress and health care over and above the effect of income. In many of the studies of this issue that followed Preston's lead, the assumption that income is the most important driver of mortality decline has been an unquestioned starting point.

Infant mortality rate (death per 1000 live births) VS infant deaths (per sq km)

- absolute measure of health disparity
o Retains the same unit of measurement as the health indicator
o Ex. Argentina has an infant mortality rate of 12 per 1000 births in 2013. Bolivia has an infant mortality rate of 31 per 1000births in 2013. The rate difference in 2013 is 31-12=19

- a relative measure of health disparity used to compare the incidence rates of events occurring at any given point in time.
o No unit of measurement required to interpret
o Ex. Argentina has an infant mortality rate of 12 per 1000 births in 2013. Bolivia has an infant mortality rate of 31 per 1000births in 2013. The rate ratio in 2013 is 31/12=2.58.

- Comparison of health outcomes for most and least advantaged group
- Can be an absolute or relative measure

- Differences and ratios
- Easy to calculate and interpret
- Only captures two groups at a time
- Don't account for group size

- The health concentration curve
- The slope index of inequality

- Regression based measure for determining the degree of inequality in a population
- The slope of the regression comparing the mean health outcome in a social group to the cumulative percent of the population ranked by social position
- Can be calculated as the difference between the fitted outcomes at the two extreme ends of the scale (absolute measure)
- Can also calculate the Relative Index of Inequality (RII) as the ratio of the two extreme outcomes (relative measure)

- Allows inequalities to be measured for groups with multiple categories
- Is sensitive to the distribution of population across categories

- Requires social groups to be ordered
- Assumes a linear relationship between social position and health

- "systemic differences in health of groups and communities occupying unequal positions in society" Graham 2004
- Difficult to measure; WHO GBD study favors ungrouped variation
- Begs the question of what the unequal positions are

- Health or health care differences between advantaged vs. disadvantaged groups; potentially amenable to policy (Braveman 2006)
- "racial or ethnic differences in the quality of health care that are not due to access-related factors or clinical needs, preferences, or appropriateness of intervention" (IOM 2002)
- Dominant US approach
- Begs the question of what defines racial/ethnic groups

- Avoidable, unfair and unjust differences in health/healthcare (Whitehead 1990)
- How are avoidable, unfair and unjust defined?
- How should health inequity be measured?

- Lots of options (think Krieger et al.)
- Consider the population
- Aim for comparability
- Understand etiologic period
- Use traditions to specify limitations

- Education, occupation, and income are robustly associated with wide range of health measures, esp. general measures
- New evidence suggests that gradients vary across social contexts
- New evidence suggests that gradients are steeper for health outcomes more amenable to medical care

A fundamental social cause of health inequalities:
1. Influences multiple health outcomes
2. Affects health outcomes through multiple risk factors
3. Involves access to resources that can be used to avoid risks
4. Reproduces its association with health/illness over time via the replacement of intervening mechanisms

- As new technologies, treatments, and programs are introduced, those with the most resources are best positioned to take an advantage of these advances
o Many health-promoting interventions will increase health inequalities
o Health inequalities are often greatest for most treatable health risks

- Not falsifiable?
- Not useful (because too broad?)
- Implies that declining mortality must increase inequality?
- Too focused on the purposive actions of individuals?

- Link and Phelan: social position provides access to an extensive array of resources that people can employ to improve their health
o More resources = more opportunities to improve/protect health
- Requires key assumptions about individual action
o Social position is linked to health through purposive action
o With sufficient resources/information, individuals will seek to reduce health risks
o Opportunities are shaped by individual resources

- Random chance
- Social position --> health
- Health-->social position
- Z -->health
--> social position

Z could be: place, family background, language, culture, politics

- Limiting ability to work or find work
- Increasing expenses
- Influencing the marginal utility of consumption
Reducing human capital accumulation (less schooling, credentials, work experience

- Control for confounding variables
o Hold z constant
- Longitudinal analysis
o Track trends over time
- Randomized experiment
o Treatment is unrelated to personal characteristics
- Quasi-experiments
o Find natural source of variation

- What about unmeasured confounders?
- Are health events or gains in resources independent?
- How feasible is randomized in health research?
o Ethical problems w double blind, health is different, group comparisons
- Difficulty in identifying true natural experiments

- Race is a variable system of social categorization involving variable phenotypic attributes
- Race as a categorization system is not necessarily skin color
- Race is a matter of life and death because it distributes power and resources
- Even if race is not objectively real, "the fact that [perceptions of race] are subjective does not, of course, mean that they do not have social consequences" (Villarreal 2012, cited in Monk 2015).

The probability of an outcome can be expressed in terms of the odds, or the probability that an outcome happens over the probability that an outcome doesn't happen:
- IMR in MA for Blacks/African Americans in 2018 = 8.4
- Odds = .0084 / .9916
- IMR in MA for Whites in 2018 = 3.2
- Odds = .0032 / .9968

The odds ratio quantifies the inequality in risks between groups:
- How much higher is the IMR for Blacks/African Americans vs. Whites?
- Rate difference = 8.4-3.2 = 5.2
- Rate ratio = 8.4/ 3.2 = 2.625
- Odds ratio = (.0084 / .9916) / (.0032 / .9968) = 2.639
- We can say the odds of infant mortality are 164% higher for B/AA
- We can also say the odds are 2.64 times greater for B/AA

- Socioeconomic position: is race really all about class?
o Education
o Employment
o Wealth
- Health behavior
o Consumption
o Health care
- Psychosocial stress: is America post-racial?
o Discrimination
o Hierarchy stress; experience of racism
- Institutional racism
o Segregation
o Housing discrimination
- Genetics
o -Clinal variation (gradational and geographical)
o More within-race than between-race variation

no, even in same household income level, whites have a higher life expectancy than blacks

Embodiment

bravemen: unstructured variation

bravemen: differences between advantaged/disadvantaged that can be addressed by policies

bravemen: hypothetical state where all have equal opportunities for health and resources (normative view)

Disability-adjusted life years: summary of population health. sum of years life lost due to premature mortality (YLLs) and years lived with disability (YLDs)

...