Union density and household income, nominal and cost-of-living adjusted.
The cross-state relationship between unionization and median household income is reported two ways.
The left chart uses Census ACS 2024 1-year median household income in nominal dollars.
The right chart divides that income by the BEA Regional Price Parity (all items, 2024; US = 100),
producing a rough cost-of-living-adjusted measure. The slope of the one-variable OLS fit
shrinks by about 57% once RPP is applied, but stays positive and statistically significant.
Regression slope · USD per +1 pp union
$1,256→$537
Shrinks by 57% once RPP-adjusted.
Correlation
0.540→0.352
Moderate, then weak-to-moderate after adjustment.
R-squared
0.292→0.124
About 29% → 12% of state-level variance explained.
Sample means
9.5% union
Nominal mean income $80,903, adjusted mean $83,068.
Side by side: union rate vs. income, before and after RPP
Both panels share the same y-axis so the compression is visible directly.
Bubble size reflects each state's employed wage and salary workforce. The dashed line in each panel
is that panel's own OLS fit. Hawaii sits furthest right at
26.5% union; North Carolina anchors the
low end at 2.4%.
California has the highest RPP at 110.7;
Arkansas the lowest at 86.9.
Regression comparison
One-variable OLS across the 50 states. Adjusting the outcome for BEA's
Regional Price Parity substantially compresses both the slope and the R-squared,
because high-union coastal states also have the highest cost of living.
Nominal income
RPP-adjusted income
Slope (USD per +1 pp union)
How much median income rises per one-point increase in union density.
$1,256
$537
Slope 95% CI
$688 – $1,825
$123 – $951
Slope p-value
5.172e-05
0.01214
Correlation
0.540
0.352
R-squared
0.292
0.124
Mean income (y)
$80,903
$83,068
Residual std. error
$10,530
$7,672
Who moves most when we adjust for cost of living?
State rankings by median household income shift once the RPP deflator is applied.
States that climb are affordable relative to their nominal income; states that fall
have high nominal income largely absorbed by high local prices.
Quadrant view (nominal)
Using the nominal scatter's sample means, states are grouped around the mean
union rate and the mean nominal household income. This keeps the quadrant
definitions comparable to the earlier version of the report.
Slope p-values: nominal 5.172e-05,
RPP-adjusted 0.01214. Even after
adjustment the slope remains positive and distinguishable from zero at
conventional levels, though with wider relative uncertainty.
This is still descriptive, not causal: the regression compares state-level averages and does not
control for industrial mix, education, demographics, or labor-force composition.