EIG DCI - Economic Innovation Group Distressed Communities Index Data

General description

1. Database primary purpose - Provide a single, holistic, and comparative measure of economic well-being across communities throughout the United States.
2. Overall data type - Demographics, socioeconomic
3. Dataset type - Longitudinal
4. Data source - US Census
5. Data level - Person level
6. Geographic location of the data collection sites - United States
7. Sponsor, manager, or home institution - Economic Innovation Group
8. Date range - 2007 - 2019
9. Geolocation data - Zip codes (or Place ID/FIPS), city, county, state, census region
10. Clinical areas of interest - all
11. Number of records - In all, the zip code-level DCI captures 99 percent of the U.S. population and all 25,400- plus zip codes with at least 500 residents not in dormitories, group quarters, the armed forces, or other similar arrangements. At higher levels of geography, it captures 3,133 counties with at least 500 residents, 828 cities with at least 50,000 people, and 383 metro areas.
12. Variables that are uniquely present in this dataset - DCI combines seven complementary economic indicators into a single summary statistic that conveys each community’s standing relative to its peers. The index sorts the communities into five even quintiles, or tiers, of economic well-being: prosperous, comfortable, mid-tier, at risk, and distressed.

Applicable methods

1. Exploratory analysis (1, 2, 3)
2. Inferential test (4)
3. Association, such as multivariable regressions (5, 6), ANOVA (7), ANCOVA (8), hierarchical linear regression (9), (10), Generalized Linear Models (11)
4. Propensity score (12)
5. Machine learning, such as logistic regression, random forest, decision tree, gradient boosting, k-nearest-neighbor classification, and XGBoost tree models (13, 14)
6. Time to event (15, 16)

High-impact designs

• Enrichment of the DCI dataset through linkage to other datasets, such as ACS NSQIP1 (5) and Medicare claims (6)

Data dictionary

To access the data dictionary, click here

Variable categories

1. Patient demographics (e.g., race/ethnicity, level of education)
2. Geographic information (e.g., metropolitan statistical area, total population, town size)
3. DCI indicators (i.e., % of adults w/o a High School diploma, poverty rate, % of prime-age adults not in work, housing vacancy rate, median income ratio, % change in employment, % change in establishments, and employment suppression flag)
4. DCI Results (i.e., Distress Score, quintile, total and percentage of population in distressed zip codes, total and percentage of population in prosperous zip codes)

Linkage to other datasets

• EIG DCI can theoretically be linked to any dataset where the location (i.e., zip code, county) is mentioned, although that linkage might require the use of NLP and manual checking.