Unionization and the Wage
Structure of Nursing
Joanne Spetz, Ph.D.
University of California, San Francisco
Michael Ash, Ph.D.
University of Massachusetts at Amherst
June 2009
1
Union influence in health care is
growing

Nearly 21% of RNs were in unions in 2006





Unionization rate is 5 percentage points higher
than for other US workers
Increasing rate of unionization among health
care workers
Aggressive bargaining for contracts
Efforts to change state and national policies
Conflicts and competition between unions
2
Unionization has risen over time
3
What do unions do?

Seek to divert employers’ net revenues
(profits) to workers




Wages
Benefits
Working conditions
Control employer behavior



Hiring
Firing
Layoffs
4
What do unions do?

Maintain their position




Create solidarity among workers
Wages, benefits, working conditions
Public relations, public service, legal actions,
advocacy
Reduced wage dispersion, increased sense of
“fairness”
5
Cross-industry research finds…


Unions are associated with higher wages and
benefits
Public sector unions have smaller wage
premium


Is health care in the US public or private sector?
Wage premium for unionization has declined
over past 2 decades

Decline is smaller in health industry
6
How do unions affect hospital RN
wages?



Research from 1970s and 1980s find union
wage premia for RNs
Some evidence that non-union RN wages are
affected by union wages
This study: Do unions affect the structure of
wages?
7
Data and methods




US Current Population Survey data, 20032006
RNs identified by their self-reported
occupation
Hospital nurses identified by employment
setting
Number of observations: average=326/yr
unionized, 1472/yr non-union RNs

No fewer than 233 union or 1133 non-union RNs
8
Data and methods


Restricted sample to RNs who worked at
least 20 hours per week
Hourly wage computed




Usual weekly earnings divided by usual weekly
hours
All wages adjusted for inflation to 2006 dollars
Deleted outliers (< minimum wage, >$100/hr)
Deleted cases with imputed wages
9
Data and methods

Education is self-reported, based on degrees
received





“Some college” assumed to be Diploma RN
Associate degree assumed to be AD RN
Baccalaureate degree assumed to be BSN
It is possible that the nurse’s highest degree is not
in a nursing field (e.g., BS in biology, AD in
nursing)
Experience is estimated from years since
education
10
Characteristics of hospital RNs,
CPS, 2000-2006
Union
Non-Union
Hourly earnings
$33.50
$28.20
Female
91.1%
92.4%
Immigrant
18.9%
10.1%
Urban residence
85.6%
80.0%
White
71.9%
82.5%
Assoc. Degree
37.1%
40.4%
Bachelor’s Deg.
50.4%
49.0%
42.9
41.2
Age
11
Estimating the effect of unions on
wage dispersion, part 1

Ordinary least squares estimates of wages


Dependent variable: log (hourly wages)
Explanatory variables:







Education
Potential experience (age - 6 - years of ed)
Citizenship, immigration
Race/ethnicity, gender
18 regional dummies (9 census x urban/rural)
Time dummies
Separate equations for union & non-union
12
Estimating the effect of unions on
wage dispersion, part 2

Examine residual wages from OLS
regressions

Compare residuals for unionized RNs to nonunion RNs
13
Results, part 1 (OLS regression)
Female
Immigrant
Black
Hispanic
Other race
Union
-0.030
0.020
-0.035
-0.108
0.024
Non-union
-0.005
-0.064*
-0.068*
-0.042
0.043
Diff
-0.026
0.084
0.033
-0.066
-0.019
Non-union immigrants and blacks face a significant
wage penalty.
The coefficients of the union and non-union equations
are not significantly different from each other.
14
Results, part 1 (OLS regression)
Diploma
Bachelor’s
Master’s
Doctorate
Experience
Experience-sq
Union
0.075
0.083*
0.237*
0.063
0.017*
-0.0003*
Non-union
-0.053
0.074*
0.197*
0.045
0.002*
-0.0003*
Diff
0.128*
0.008
0.040
0.018
-0.002
0.00002
Diploma wage gap vanishes, other
education gaps don’t change.
Experience gap?
15
Wage-experience profile
16
Results, part 2 (residual ln wages)
17
Conclusions

Modest or no effect on the wage structure:

No race-wage gaps among unionized RNs



Differences between the union & non-union
coefficients are not significant
No penalty for diploma among union RNs
Lower premium for experience among union RNs

Differences in coefficients not statistically significant
18