Loading [MathJax]/jax/output/CommonHTML/jax.js
+ - 0:00:00
Notes for current slide
Notes for next slide

Exploring Socioeconomic Indices with R

Francis Boscoe, Ph.D.

Founder, Pumphandle, LLC

April 9, 2020

1 / 22

Where you live is a good indicator of socioeconomic position

2 / 22

Life expectancy and disability-free life expectancy at birth, by neighborhood income level, England, 1999-2003.

Fair Society, Healthy Lives: The Marmot Review (www.ucl.ac.uk/marmotreview)
3 / 22

Inequality in early cognitive development of children in the 1970 British cohort study, at ages 22 months to 10 years.

Fair Society, Healthy Lives: The Marmot Review (www.ucl.ac.uk/marmotreview)
4 / 22

Populations living in areas the least favorable environmental conditions across various environmental categories.

Fair Society, Healthy Lives: The Marmot Review (www.ucl.ac.uk/marmotreview)
5 / 22

Examples of two SES indexes currently in use:

Yost index – 7 variables from ACS

- Rent
- Income
- Home value
- Poverty rate
- Blue-collar employment
- Unemployment rate
- Years of education completed

Area Deprivation Index – 17 variables from ACS

- All of the Yost variables
- Education and poverty measured twice
- Income disparity
- Homeownership rate
- Mortgage payment
- Single-parent households
- Motor vehicle ownership
- Telephone ownership
- Functional plumbing
- Overcrowding
6 / 22

The Area Deprivation Index has its own web site (https://www.neighborhoodatlas.medicine.wisc.edu)

7 / 22

8 / 22

There are some problems with the ADI

  • Missing observations are treated as having a value of zero

    • Rent: 18% missing
    • Mortgage: 13% missing
    • Median household income: 6% missing
    • Other variables: 1-2% missing
  • Original factor loadings from 2000 are used

  • Income disparity is not adjusted for inflation
  • Variables are highly heteroscedastic
  • Access to a phone means something different in 2020 than in 2000
9 / 22

Yost index vs. ADI (using 2011-2015 ACS data)

10 / 22

Yost index vs. ADI (using 2011-2015 ACS data)

  • Yost index vs. ADI as published: R2=0.79

Where the two differ by >30 percentile points, Yost is higher over 90% of the time, because it used imputed rather than zero values missing data.

10 / 22

Yost index vs. ADI (using 2011-2015 ACS data)

  • Yost index vs. ADI as published: R2=0.79

Where the two differ by >30 percentile points, Yost is higher over 90% of the time, because it used imputed rather than zero values missing data.

  • Yost index vs. ADI calculated the same way: R2=0.92

Differences are normally distributed and typically involve outliers in the variables unique to the ADI

10 / 22

11 / 22

Yost index vs. ADI (continued)

  • Where ADI indicates a higher SES than Yost tend to be rural areas with low income but little income disparity, high rates of home ownership, few single-parent households, high vehicle ownership, and little crowding. (Examples in Arkansas, Alabama, Texas)
12 / 22

Yost index vs. ADI (continued)

  • Where ADI indicates a higher SES than Yost tend to be rural areas with low income but little income disparity, high rates of home ownership, few single-parent households, high vehicle ownership, and little crowding. (Examples in Arkansas, Alabama, Texas)

  • Where ADI indicates a lower SES than Yost tend to be urban and suburban areas with high cost of living but high income disparity, low rates of home ownership, many single-parent households, low vehicle ownership, and high crowding (Examples in San Francisco, Los Angeles, Washington DC)

12 / 22

13 / 22

14 / 22
15 / 22


Washington, DC
Yost Index
10%20%30%40%50%60%70%80%90%

NA
Leaflet | © OpenStreetMap contributors, CC-BY-SA
16 / 22

17 / 22

18 / 22

19 / 22

20 / 22

Conclusions

  • No single SES index will ever be perfect, but the Yost index has a lot of positive qualities

  • It is beyond time for socioeconomic measures to be a core element of population-based cancer surveillance

21 / 22
22 / 22

Where you live is a good indicator of socioeconomic position

2 / 22
Paused

Help

Keyboard shortcuts

, , Pg Up, k Go to previous slide
, , Pg Dn, Space, j Go to next slide
Home Go to first slide
End Go to last slide
Number + Return Go to specific slide
b / m / f Toggle blackout / mirrored / fullscreen mode
c Clone slideshow
p Toggle presenter mode
t Restart the presentation timer
?, h Toggle this help
Esc Back to slideshow