Does Public or Not-for-Profit Status Affect the Earnings of
Hospital Workers?
Edward J. Schumacher
Departments of Health Care Administration and Economics
Trinity University
San Antonio, TX 78232
Edward.Schumacher@Trinity.edu
210.999.8137
Abstract
This paper examines the earnings differentials among health industry workers in the
public, private nonprofit, and private for-profit sectors. Utilizing data from the 1995
through 2004 Current Population Surveys, unadjusted earnings are highest in the private
nonprofit sector and lowest in private for-profit firms. Once measurable characteristics
are accounted for health practitioners in for-profit and nonprofit firms earn similar wages
while public sector health care workers earn small but significant wage penalties. Wage
change analysis using pooled two year panels constructed from the CPS indicate no
significant differences in earnings is found between the three sectors of employment.
Whatever the role of the sector of employment on the overall earnings of hospital
workers, there is sufficient worker mobility within the industry to largely eliminate
systematic wage differences across type of hospital.
July 2006
I.
Introduction
The health care industry is a unique industry in the US economy and accounts for a large
and growing proportion of its resources. Spending in 2003 was 1.67 trillion dollars or $5,670 per
capita, and accounted for over 15 percent of GDP. In 1985 employment in the industry accounted
for 8.2 percent of all employment and by 2003 healthcare employment was 11.3 percent of all
employment. This industry is unique, in part, due to the fact that the public sector, the for-profit,
and private not-for-profit sectors all compete directly in the same markets. There has been
considerable research examining the theory and outcomes of the differing incentives of these
sectors on output markets (Chakravarty, et al. 2005; Horwitz, 2005; Sloan, 2000), but there has
been little research examining how labor market outcomes are different within these sectors of the
industry. Sloan and Steinwald (1980) use data from the 1970s and find no significant wage
differences in wage rates paid by hospitals of different ownership types. This paper looks at how
the earnings and employment of workers in the health care industry are different across these
three sectors of the industry.
Previous research on public labor market tends to find moderate to large wage premium
for federal government workers (on the order of 10-20 percent), with larger differentials for
females than for males and for nonwhite males than for white males.1 These differentials,
however, are reduced substantially once occupation is controlled for in detail (Moulton 1990).2
Thus once occupation is controlled for in sufficient detail the public wage differential disappears.
This assumes, however, that jobs are similar across sectors, which is generally not the case.
Thus, prior studies finding a non-profit or public wage effect are not comparing truly similar jobs.
Rhum and Borkowski (2003) examine compensation in the nonprofit sector and lay out
why wages might differ between private for-profit and non-profit industries. Since nonprofits are
barred from distributing net earnings, Hansmann (1980), Easly and O’Hara(1983) and Handy and
1
For a review of this literature, see Ehrenberg and Schwarz (1986) or Gregory and Borland (1999).
2 Postal workers are the clear exception (Hirsch, Wachter, and Gillula; 1999).
Katz (1998) argue that nonprofits will be prevalent in markets where consumers lack information
about the good (price, quality, quantity, etc.) since nonprofits will help alleviate the consumer
trust problem. Rhum and Borkowski suggest this characteristic results in two reasons for why
wages might be higher in the nonprofit sector. First, managers may have less incentive to hold
down wages since they do not gain from the resulting costs savings. Second, nonprofits have less
incentive to shirk on quality and so may choose to hire better quality workers. At the same time
though, workers may be willing to work for lower wages in nonprofit firms as a way of donating
to a “worthy cause.” Also Lakdawalla and Philipson (1998) suggest that nonprofits are likely to
be concentrated in more competitive and less profitable industries and this should put additional
downward pressure on wages. Rhum and Borkowski use data from the 1994-1998 Current
Population Surveys and find that weekly wages are about 11 percent lower in nonprofit jobs as
compared to for-profit jobs, but this is due almost entirely to differences in hours worked and the
concentration of nonprofit jobs in low paying industries.
Thus previous studies examining either the public vs. private wage differential or the forprofit vs. not-for-profit wage differential have faced the difficulty of comparing similar workers
doing similar jobs. This paper focuses specifically on occupations within the hospital industry
since it has the unique characteristic of containing large segments of public, private for-profit, and
private nonprofit employers and workers who are performing similar tasks across sectors. Data
from the Current Population Survey are utilized to analyze how wages and employment differ
across the three sectors of employment. Using both cross-sectional data as well as pooled twoyear panels, wage differentials between public, private for-profit, and private-non-profit workers
in the health care industry are estimated. The findings show that while wage differences are
small, earnings are lowest in the public sector and highest in the private not-for–profit sector(once
differences in measurable characteristics are accounted for). This is counter to the non-health
care market where large premiums are found for public sector employment. Wage change
analysis, however, fails to find a significant difference between the three sectors of employment.
2
II.
Data and Descriptive Statistics
Data
The primary data for this study are drawn from the monthly Current Population Survey
(CPS) Outgoing Rotation Group (ORG) earnings files for January 1994 through December 2004.
Beginning in 1994 the CPS began collecting information on the profit status of the workers in the
private sector. Previous surveys contain information about public sector workers (federal, state,
or local government), but 1994 was the first year they separated private for-profit from private
nonprofit employers. The focus is on hospital workers. In order to make more detailed
comparisons, this group is broken into four groups: RNs, other health practitioner occupations3,
those non-practitioners with at least a college degree, and those non-practitioners with less than a
college education. Comparisons are also made to workers outside the health care industry.
The samples include employed wage and salary workers ages 18 and over, with positive
weekly earnings and hours. The wage on the primary job is measured directly for workers
reporting hourly earnings who do not receive tips, commissions, or overtime. For others the wage
is calculated by dividing usual weekly earnings on the primary job (which includes usual tips,
commissions, and overtime) by usual hours worked per week.4 Usual weekly earnings are topcoded at $1,923 through 1997 and at $2,885 beginning in 1998. Those at the cap are assigned
mean earnings above the cap based on year and gender-specific estimates that assume a Pareto
distribution for earnings beyond the median.5 Workers with measured hourly earnings less than
$3 or greater than $150 (in 2004 dollars) are omitted.
Workers who have had their earnings, occupation, or industry allocated by the Census are
also excluded. Few workers have occupation or industry allocated. A large number (20-30%)
have earnings imputed by the Census based on a cell hot-deck procedure in which
3
Other practitioners include the following. For 1994-2002: Health Diagnosing Occupations (excluding RNs,
physicians, dentists, and veterinarians), and Health Assessment and Treating Occupations. For 2003-2004: Health
Practitioner and Technical Occupations (excluding RNs, physicians, dentists, and veterinarians).
4 Those reporting “variable” hours have their implicit wage calculated, if necessary, using hours worked last week.
5 Estimates of gender-specific means above the cap for 1973-2003 are posted by Barry Hirsch and David Macpherson
at http://www.unionstats.com. Values are about 1.5 times the cap, with somewhat smaller female than male means and
modest growth over time.
3
nonrespondents are allocated the earnings of a “donor” with an identical set of match
characteristics. CPS imputation rates have been particularly large in recent years. Imputed
earners are excluded from the sample since their inclusion may result in match bias (Hirsch and
Schumacher, 2004). Since non-responding workers are not in general matched to the earnings of
workers in similar settings (Industry and public/private sector are not match criteria, and
occupation is defined at only a broad level), wage differences between public, private for-profit,
and private nonprofit are biased toward zero. Following the above sample restrictions, the sample
sizes of RNs, other health practitioners, non-practitioners with at least a college degree, and nonpractitioners with less than a college degree are 15,212; 12,183; 7,143; and 23,247 respectively
Descriptive Statistics
Table 1 displays employment (overall and by sector of employment) for hospital workers
between the years 1994 through 2004. Employment increases for all the groups, but the increases
are mostly confined to the private sector. Public sector employment decreases for RNs, other
practitioners, and non-practitioners with less than a college degree. Practitioner employment
increases the most in the private nonprofit sector. Note that part of the jump in employment for
practitioners between 2002 and 2003 could be due to the occupational code changes that were
implemented in the 2003 CPS. In January 2003, the CPS adopted the 2002 Census industry and
occupational classification systems derived from the 2002 North American Industry
Classification System and the 2000 Standard Occupational Classification system. The industry
changes did not substantially affect the definition of the hospital industry, but there were
numerous occupational changes that affected the classification of practitioners. It is not possible
to construct time consistent occupational codes to cover the period. Overall the table shows an
increase in employment in the industry with a shift away from the public sector towards private
sector employment.
4
To illustrate relative wage movements during the period, an adjusted wage index is
constructed for each group. A separate log wage equation is estimated for each of the healthcare
groups and for non-healthcare workers with and without a college degree. The regressions include
controls for years of schooling, experience and its square, and dummy variables for gender, union
membership, part-time employment, race (3), marital status (2), region (8), and year. The
coefficients from the year dummies are converted to a percentage of the wage in 1994 (the
omitted year), which is set equal to 100. Figure 1 displays these indexes, revealing that early in
the period, healthcare wages were relatively flat or declining (especially for nonpractitioners with
a college degree). Beginning around 1997, however, wages began rising. Health industry wages
flattened out again in the 1999 to 2001 period and then rose rapidly in the 2001 to 2002 period. It
appears that wage growth has slowed in the most recent years. Practitioners had particularly
rapid wage growth in the later years. A hospital RN in 2004 is estimated to earn a little more than
12 percent more than a similar RN in 1994. For other hospital workers wages grew by 8-10
percent over the period.
Table 2 displays descriptive statistics for each of the groups. Overall, unadjusted wages
are about 8-9 percent higher in the private non-profit sector, although for RNs there is little wage
difference across sectors. Among other practitioners workers in the private non-profit sector earn
about 6.6 percent higher wages than those in the public sector and 7.4 percent higher than those in
the private for-profit sector. For nonpractitioners with a college degree, public sector wages are
about 6.0 percent lower and for-profit wages are about 10.0 percent lower than private non-profit
wages. Public sector workers tend to be older, are more likely to hold a graduate degree, are less
likely to be white, and more likely to be a union member. The next section examines how these
differentials change once measured characteristics are accounted for.
5
III.
Cross-Sectional Estimates of the Public/Nonprofit/For-Profit Wage
Differentials.
The estimation begins with a standard cross-sectional log-wage equation:
(1)
J
Y
j 1
y 2
ln Wit   jXijt  yYEARiy   1 Public it   2 Pr ivateNPit   it ,
where lnWi is the log of the real wage for worker i, at time t, X contains J-1 personal,
job, and labor market characteristics (e.g., education, potential experience, union status, region,
etc.) and  contains the corresponding coefficients (X1=1 and 1 is the intercept). Public is a
dummy variable equal to 1 if the worker is employed in the federal, state or local government,
and PrivateNP is equal to one if the worker is employed by a private nonprofit firm (with private
for-profit as the omitted group). YEAR includes dummy variables for the years 1995-2004. For
now ei is assumed to be a well behaved error term. The coefficients 1 and 2 are the adjusted log
earnings differences relative to the for-profit sector. That is, 2 is an estimate of the wage
differential between private for-profit and nonprofit workers, holding all other variables constant,
and 1 is the unbiased differential between public sector workers and private for-profit workers.
Table 3 shows estimates of these differentials, pooled over the years 1994 to 2004. Other
than the variables shown in the table, dummy variables for marital status (2), race/ethnicity (3),
year (10), and region (8) are included in the model. Unlike industries outside of health care, there
is a negative and significant differential associated with employment in the public sector.
Overall, health industry workers in public employment earn about 2.5 percent lower wages than
similar workers employed in the private for-profit sector, and 4.8 percent less than workers in
private non-profit hospitals. For RNs, the differential is only about 1.8 percent while for other
practitioners this differential is about 3.1 percent between public and private for-profit and about
6.4 percent between public and non-profit hospital workers. While these effects are not
particularly large, they are counter to other studies examining all industries, which tend to find
that wages are 10 to 20 percent higher in the public sector (Ehrenberg and Schwarz, 1986;
Gregory and Borland, 1999). For nonpractitioners with less than a college degree there is little
6
difference between public sector and private nonprofit earnings, while wages are about 3 percent
lower in the for-profit sector. Among nonpractitioners with a college degree there is about a 6
percent wage advantage for working in a nonprofit hospital, compared to a public or for-profit
hospital.
Among the other coefficients there are large returns to education and relative flat returns
to experience among practitioners. There is not a wage differential associated with gender for
RNS (where about 94 percent are female), but a relatively large penalty among those nonpractitioners with a college degree. The union wage differential is relatively small within the
industry6. Among RNs there is a higher wage for part-time employment. This is counter to the
other occupations within the industry and to results for the economy as a whole (Hirsch 2005).
This premium for part-time employment could be due to lower non-wage benefits for part-time
workers, older workers selecting into part-time jobs, or to differences in shift work. If part-time
jobs are more likely to be on weekends or late night shifts and there is a premium for these shifts,
then this will show up as a part-time wage premium.
The cross-sectional regressions find that once measurable characteristics are accounted
for, there are little differences in wages for hospital RNs employed across the three different
sectors. Among other hospital workers, however, private nonprofit hospital workers make 5-6
percent more than workers in public hospitals (with the exception of RNs and nonpractitioners
with less than a college degree), while wages in for-profit hospitals tend to be 3 to 6 percent
lower than private nonprofit hospitals. If the included right-hand-side variables control
completely for human capital and other determinants of earnings, or if those factors not included
are uncorrelated with sector employment, then the differentials could be interpreted as a true
wage premium for working in the private nonprofit sector. Since public sector hospitals tend to
have a much higher proportion of uncompensated care (both bad debt and charity care) than do
private sector hospitals, it could be that public hospitals attract workers who are also more willing
6
The negative differential for non-practitioners with a college degree is likely capturing management positions which
should be relatively high wage with little or no unionization.
7
to “donate” a portion of their wage to help fulfill the mission of the organization (though this
would not explain the for-profit/nonprofit differential).
Alternatively public sector employment could offer job amenities such as greater job
security, more regular hours, or other nonwage attributes that workers are willing to pay for in the
form of lower wages. On the other hand, it could be that private hospitals are able to attract a
higher quality worker into their facilities due to better facilities, more stable financial conditions,
more attractive clientele, etc. In this case the lower wage in the public sector would reflect ability
or quality differences between the workers and would not be a true wage penalty for working in
the public industry. For profit hospitals may face tighter financial pressure than private nonprofit
hospitals and this tends to push wages down relative to the nonprofit sector. That the private
nonprofit/for profit wage differential is largest for nonpractitioners with a college degree (who are
more likely to be managers with more influence over wages) is consistent with this wage
differential being a true rent. However, since the CPS data utilized here only contains
information on wages, it is possible that for profit hospitals offer larger non-wage compensation.
Such things as stock options and other types of compensation in the for-profit sector could well
offset the 6 percent premium for private nonprofits over for-profit hospitals.
IV.
Longitudinal Estimates of the Public/Nonprofit/For-profit Wage Differentials.
If there are omitted factors correlated with sector of employment and earnings, then the
cross-sectional differentials will be biased estimates. Suppose the error term , can be broken
down into two parts: it = i + eit, where i is a person-specific component fixed over time, and eit
is a random, well behaved, component. If the omitted fixed effect is negatively correlated with
public employment (e.g., less able workers are located in the public sector), then estimates of the
public differential above are biased downward.
Letting the symbol  represent changes between adjacent years, a wage change equation
will take the form (dropping the individual subscript i):
8
(2)
J
D
j 1
d 2
 ln Wd   jXjd  1 ' Public id   2 '  Pr ivateNPid  dPERIOD d  e' d ,
where d indexes the time periods over which values are differenced, and PERIODd are
dummies for the periods 1995/96 through 2001/02 (with 1994/95 the reference period). The
major distinction between equations 2 and 1 is that the effects owing to unmeasured skills fall
out, potentially allowing for unbiased estimates of the quality-constant premium between sectors
of employment. For equation 2 to provide an unbiased measure of the public/nonprofit/forprofit
wage differential, sectoral switching is assumed to be exogenous and ability must be equally
valued at the margin by employers in all sectors (Gibbons and Katz, 1992) and within a year’s
time.7
Panel data are constructed from multiple panels from the CPS ORG files for 1994/95 through
2001/02 by matching individuals in the same month in consecutive years. Because measurement
error is a particular concern with longitudinal analysis, those with industry, occupation or
earnings allocated by the Census are deleted from the sample. The resulting panel data set for
1994/95 through 2001/02 contains data on 13,447 hospital workers in the health care industry,
4,049 RNs, 2,408 other practitioners, 3,488 nonpractitioners with at least a college degree, and
5,432 nonpractitioners with less than a college degree. Table 4 displays movements by group as
well as the average change in wage for each category. While there are still a small number of
movers in absolute terms, one of the advantages of examining the health care industry is that
there are a relatively large number of movers into and out of the public sector as compared to
studies examining the economy as a whole. Of the 1,894 workers in the public sector in year 1,
78.2 percent stay in the public sector, 14.6 percent move to the for-profit sector, and 7.2 percent
move to the nonprofit sector. There are similar movements in the other sectors and for the other
groups of workers. There is much more movement between the for profit and not-for-profit
7
The above specification forces the effects of joining and leaving sector to be symmetric. That it, it
assumes that the gain (or loss) in wage to moving out of the public sector is equal to the loss (or gain) in
wage to moving into the public sector. This assumption can be relaxed by including eight separate dummy
variables for the nine categories of changes between years (public to for-profit, public to nonprofit, public
to public, etc.). Small sample sizes, however, make estimates from this specification unreliable.
9
sectors than between the private and public sectors. Part of this could be due to a higher degree
of measurement error on the private for-profit and not-for-profit distinction. Overall there is a 2.7
percent increase in wages for health care workers between years. The wage change for RNs is
3.8 percent and for non-practitioners with a college degree it is 5.3 percent. There does not
appear to be a clear pattern in the wage changes among those who changed sectors of
employment.
Table 5 shows wage level regressions using the first year from the CPS matched panel data as
well as wage change regressions. Here the differentials are similar to those estimated from the
full CPS data. Wage change regressions, however, find little to know wage differences across all
three sectors. The coefficient estimates are all very small and none of them are statistically
significant.
These results suggest that there is no wage difference across the three sectors of employment
in the health care industry. The small wage differentials found in the cross sectional results are
not found in the wage change results. This suggest that these small differentials are not true wage
premiums, but are due to omitted factors correlated with sector of employment. That is, the
negative wage differential associated with public employment found in Table 3 is not due to these
workers donating part of their earnings to the organization in support of its mission or to
differences in job amenities, but may be explained by lower quality workers being attracted to the
public sector.
Also shown in Table 5 are estimates of the part-time wage differential. Note that the positive
differential for RNs in the cross-sectional analysis of about 6 percent is quite similar to the wage
change result. This suggests that there is a true differential associated with part-time
employment, and is consistent with part-time jobs being associated with differences in shift work.
V.
Conclusion
This paper examines the role of the sector of employment in the earnings of workers in the
health care industry. The results find that before accounting for measurable characteristics
10
workers in the private non-profit sector earn higher wages than private for-profit and public sector
workers, with for-profit workers earning the lowest wages. But once personal characteristics are
accounted for, workers in the public sector earn significantly lower wages than those in the other
sectors of employment. Public sector workers are found to earn about 4.6 percent lower wages
than private nonprofit workers, and about 2.5 percent less than private for-profit workers. RNs
appear to be the exception to this as there are few wage differences across sectors. Wage change
analysis utilizing two-year panels constructed from the CPS finds no significant differences
between the groups. This suggests that the small public private wage differential is due to
unmeasured ability, suggesting that lower quality workers are attracted to the public sector.
Alternatively, measurement error in the longitudinal data could be biasing the wage change
results towards zero. But given that even the cross-sectional differentials are quite small, the
results here suggest that there are few differences across sector and is consistent with competitive
wage setting behavior.
There is a large IO literature examining differences in the behavior of for-profit and not-forprofit hospitals. The results are somewhat mixed, but for the most part few differences are found
in output, costs, efficiency, pricing, and provision of uncompensated Care (Chakravarty, et al.
2005). The results here are consistent with there being few differences in costs across hospital
sectors. While there are small wage differences between workers across the three sectors, it
appears that public, private for-profit, and private not-for-profit hospitals operate in the same
labor markets.
11
Figure 1
Wage Indices by Year by Group
116
114
112
110
Wage Index
108
106
104
102
100
98
96
94
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
Year
RNs
Other Practitioners
Non-Practitioner School >=16
Non-Practitioner School <16
Non Health Industry School >= 16
Non Health Industry School < 16
Data are from the 1994-2004 Monthly Outgoing Rotation groups of the Current Population Survey. Shown are
log wage regression coefficients on year dummies, with 1994=100 estimated separately for each group. The
indices are derived from a separate log wage equation for each group including controls for years of schooling,
potential experience (years since finishing school) and its square, and dummies for race (2), gender, part-time
status, marital status (2), region (8), large metropolitan area, and year (10). The coefficients on the year
dummies are converted to a percentage index by: exp()*100, where  is the coefficient.
12
2004
Table 1
Hospital Employment: 1994-2004
(in thousands)
RNs
All
Non-Practitioner School 16
Other Practitioners
Non-Practitioner School <16
Public
1994
1,324
Non- ForAll
Public Non- For- All
Public Non- ForAll
Public
Non- ForProfit profit
Profit profit
Profit profit
Profit profit
190
370
764
936
145
268
551
538
101
198
238 1,928
377
468 1,092
1995
1,290
176
436
678
951
147
274
530
538
91
209
237
1,962
355
542
1,065
1996
1,274
149
439
685
932
141
294
498
563
103
209
251
2,012
359
569
1,085
1997
1,322
156
480
686
986
142
355
489
557
99
217
240
2,006
348
603
1,054
1998
1,305
150
466
689
1,009
122
357
530
609
94
245
269
1,963
317
627
1,018
1999
1,363
150
501
713
1,004
107
359
539
558
96
232
230
1,933
307
644
982
2000
1,310
147
498
665
965
122
350
493
534
91
215
228
1,923
272
597
1,053
2001
1,290
137
491
662
928
105
314
508
611
90
245
276
2,013
281
631
1,102
2002
1,396
149
519
728
977
100
344
533
619
92
252
275
2,016
288
625
1,103
2003
1,545
137
551
857
1,371
159
486
726
666
101
268
297
2,044
259
702
1,083
2004
1,540
123
580
837
1,307
185
473
650
699
110
303
286
2,033
313
655
1,066
Data are from the 1994-2004 Monthly Outgoing Rotation groups of the Current Population Survey. Practitioners includes the following. For 1994-2002: Health
Diagnosing Occupations (excluding physicians, dentists, and veterinarians), and Health Assessment and Treating Occupations. For 2003-2004: Health
Practitioner and Technical Occupations (excluding physicians, dentists, and veterinarians). Non Practitioner School  includes those individuals in the health
care industry with at least a college degree but not in practitioner occupations. Non Practitioner School < 16 is defined similarly.
13
Table 2
Hospital Employee Descriptive Statistics by Group by Sector of Employment
Real
Wage
Log
Wage
Age
Bach.
Degree
Graduate
Degree
Female
PartTime
Black
Other
24.88
(9.25)
Non
Profit
25.16
(8.59)
Forprofit
24.09
(15.02)
Other Practitioners
Public
Non
Profit
20.87
21.39
(16.42)
(15.65)
Forprofit
20.02
(14.34)
Health Care School 16
Public
Non
ForProfit
profit
23.52
24.97
23.01
(13.57)
(15.24)
(16.07)
Health Care School <16
Public
Non
Profit
13.59
13.28
(9.31)
(6.08)
Forprofit
12.52
(9.26)
2.789
3.157
3.176
3.160
2.879
2.935
2.865
3.034
3.086
2.979
2.519
2.509
2.441
40.03
(11.17)
.252
(.434)
.082
(.274)
.818
(.386)
.213
(.410)
.126
(.332)
.060
(.238)
.623
(.485)
.183
(.386)
21,754
42.84
(9.81)
.464
(.499)
.094
(.292)
.882
(.322)
.151
(.358)
.120
(.326)
.088
(.284)
.655
(.475)
.328
(.470)
1,428
41.52
(9.83)
.442
(.497)
.088
(.283)
.929
(.257)
.330
(.470)
.028
(.164)
.038
(.192)
.715
(.452)
.165
(.371)
5,277
40.57
(9.82)
.474
(.499)
.060
(.238)
.932
(.252)
.279
(.449)
.066
(.248)
.081
(.272)
.701
(.458)
.168
(.374)
6,129
41.04
(10.73)
.243
(.429)
.190
(.392)
.670
(.470)
.116
(.321)
.146
(.354)
.093
(.290)
.574
(.495)
.250
(.433)
1,264
39.45
(10.29)
.299
(.458)
.149
(.356)
.762
(.426)
.221
(.415)
.065
(.246)
.043
(.203)
.628
(.483)
.077
(.267)
3,734
38.70
(10.47)
.252
(.434)
.130
(.326)
.757
(.429)
.213
(.409)
.102
(.302)
.070
(.256)
.640
(.480)
.096
(.295)
4,699
43.76
(10.33)
.545
(.498)
.455
(.498)
.609
(.488)
.077
(.267)
.097
(.296)
.087
(.282)
.631
(.482)
.168
(.374)
987
41.76
(10.26)
.615
(.487)
.385
(.487)
.659
(.474)
.124
(.329)
.050
(.217)
.046
(.218)
.655
(.475)
.041
(.199)
2,509
40.56
(10.46)
.636
(.481)
.364
(.481)
.682
(.466)
.132
(.338)
.082
(.275)
.078
(.269)
.641
(.480)
.058
(.234)
2,200
43.02
(11.57)
0.00
41.83
(12.23)
0.00
40.24
(12.48)
0.00
0.00
0.00
0.00
.689
(.463)
.107
(.309)
.212
(.409)
.067
(.250)
.555
(.497)
.303
(.460)
2,901
.784
(.411)
.177
(.382)
.111
(.314)
.030
(.172)
.582
(.493)
.090
(.286)
6,544
.804
(.397)
.188
(.390)
.193
(.394)
.036
(.186)
.555
(.497)
.104
(.306)
8,726
All Health Care
Public
Non
Profit
18.93
20.05
(12.66)
(12.34)
Forprofit
18.74
(14.05)
2.804
2.872
42.71
(10.90)
.229
(.420)
.125
(.331)
.715
(.451)
.114
(318)
.162
(.369)
.080
(.271)
.592
(.492)
.278
(.448)
6,580
41.24
(10.96)
.281
(.449)
.110
(.313)
.804
(.397)
.223
(.416)
.069
(.253)
.037
(.190)
.641
(.480)
.180
(.384)
18,064
RNs
Public
Currently
Married
Union
Member
Sample
Size
Data are from the 1994-2004 Monthly Outgoing Rotation groups of the Current Population Survey. Standard deviations are in parenthesis. See Table 1 note for
description of groups.
14
Table 3
Public-Private Nonprofit Hospital Log Wage Differentials
Public
Private Non
Profit
Some College
Bacc Degree
Grad Degree
Experience
Exp2/100
Female
Union Member
Part-Time
All Health
Care
-.022
(.006)
.021
(.004)
.345
(.005)
.643
(.005)
.811
(.007)
.027
(.001)
-.046
(.001)
-.042
(.005)
.104
(.006)
-.004
(.005)
RNs
Other
Practitioners
-.027
(.013)
027
(.009)
.214
(.013)
.486
(.013)
.711
(.017)
.024
(.001)
-.040
(.003)
-.092
(.010)
.024
(.014)
-.003
(.010)
-.010
(.009)
.003
(.006)
.131
(.022)
.216
(.022)
.323
(.024)
.018
(.001)
-.030
(.002)
-.011
(.010)
.078
(.007)
.063
(.006)
Health Care
School 16
Health Care
School <16
.015
(.017)
.066
(.013)
--
.031
(.008)
.025
(.006)
.182
(.005)
--
-.268
(.012)
.037
(.002)
-.069
(.005)
-.118
(.013)
-.111
(.013)
-.162
(.018)
-.021
(.001)
-.032
(.002)
-.105
(.006)
.071
(.008)
-.105
(.007)
Data are from the 1995:8-2004:12 Monthly Outgoing Rotation groups of the Current Population Survey.
Private for-profit is the omitted group. Also included in the model are dummy variables for marital status
(2), race/ethnicity (3), year (9), and region (8).
15
Table 4
Descriptive Statistics on Two-Year Group Changes
All Health Care
Sample
Log
Size
Wage
RNs
Sample
Size
Log
Wage
Other Practitioners
Sample
Log
Size
Wage
NP School 16
Sample
Log
Size
Wage
NP School < 16
Sample
Log
Size
Wage
ALL
7,986
.028
2,529
.015
1,476
.029
937
.057
3,044
.028
Public to
Public
888
(79.9%)
.014
218
(77.9%)
-.001
146
(78.5%)
.020
129
(78.2%)
.032
395
(82.3%)
.014
Public to
For-profit
140
(12.6%)
.028
42
(15.0%)
.023
26
(14.0%)
-.017
20
(12.1%)
.073
52
(10.8%)
.037
Public to
NFP
83
(7.5%)
.052
20
(7.1%)
.109
14
(7.5%)
-.066
16
(9.7%)
.061
33
(6.9%)
.062
For-profit to
Public
111
(3.2%)
.072
33
(3.0%)
-.005
20
(3.0%)
.098
11
(3.4%)
.319
47
(3.5%)
.057
For-profit to
For-profit
2,451
(70.9%)
.017
799
(71.5%)
-.001
471
(71.3%)
.034
222
(68.9%)
.055
959
(70.8%)
.015
For-profit to
NFP
893
(25.8%)
.013
285
(25.5%)
-.012
170
(25.7%)
.009
89
(27.6%)
.043
349
(25.8%)
.029
NFP to
Public
104
(3.0%)
.049
38
(3.4%)
.100
14
(2.2%)
.100
7
(1.6%)
.088
45
(3.7%)
-.017
NFP to Forprofit
1,028
(30.1%)
.036
338
(29.9%)
.017
202
(32.1%)
.021
127
(28.2%)
.082
361
(29.9%)
.047
NFP to NFP
2,288
.042
756
.041
413
.038
316
.052
803
.041
(66.9%)
(66.8%)
(65.7%)
(70.2%)
(66.4%)
Data are from the matched CPS ORG Monthly files for 1994/95 through 2001/02. Shown are sample sizes
and the average change in log wage for each group and by the nine two-year mover categories. The
numbers in parenthesis are the percentage of workers in the given sector in year 1 that are in that category
in year two. For example, 79.9 percent of workers in the public sector in year 1 are in the public sector in
year 2, 12.6 percent of workers in the public sector in year 1 are in the private for-profit sector in year 2,
and 7.5 percent are in private not for profit in year 2.
16
Table 5
Wage Change Regression Results
Public
Private Non
Profit
All Healthcare
Wage
Wage
Level
Change
RNs
Wage
Level
-.016
(.013)
.022
(.009)
.021
(.030)
.006
(.013)
Public
--
Private Non
Profit
--
Part-Time
Part-Time
---
.007
(.015)
-.010
(.007)
.049
(.011)
---
Wage
Change
---
NP School  16
Wage
Wage
Level
Change
NP School < 16
Wage
Wage
Level
Change
-.032
(.027)
.044
(.018)
.024
(.042)
.110
(.030)
.026
(.019)
.021
(.014)
-.005
(.032)
-.015
(.014)
.066
(.014)
.021
(.012)
Other Practitioners
Wage
Wage
Level
Change
---
---
.065
(.030)
-.003
(.013)
.027
(.021
.031
(.022)
---
---
.048
(.038)
-.014
(.018)
-.126
(.042)
.005
(.027)
---
---
.016
(.022)
-.005
(.010)
-.075
(.018)
.105
(.039)
-.003
(.020)
Data are from the matched CPS ORG Monthly files for 1995/94 through 2001/02. Also included in the
wage level model is years of schooling, potential experience and its square, dummy variables for gender,
race (3), marital stauts(2), region (8), and year(10). The wage change model also includes the change in
experience squared, the change in union status, and dummy variables for year 1997 through 2002.
17
References
Borjas, George J., H.E. Frech III, and Paul B. Ginsburg. 1993. “Property Rights and Wages: The
Case of Nursing Homes,” Journal of Human Resources, 17(2):103-110.
Chakravarty, Sujoy. Martin Gaynor, Steven Klepper, and William B. Vogt. 2005. “Does the
Profit Motive Make Jack Nimble? Ownership Form and the Evolution of the U.S.
Hospital Industry,” National Bureau of Economic Research, Working Paper 11705,
October.
Ehrenberg R. and J. Schwarz, "Public Sector Labor Markets," Handbook of Labor Economics,
Vol. 2, edited by O. Ashenfelter and R. Layard, North-Holland, 1986, Ch. 22.
Frank, Richard G., and David S. Salkever. 1994. “Nonprofit Organizations in the Health Sector.”
Journal of Economic Perspectives, 8(4):129-44.
Gibbons, Robert, and Lawrence F. Katz. 1992. "Does Unmeasured Ability Explain Inter-Industry
Wage Differentials?" Review of Economic Studies, 59(3): 515-35.
Gregory, Robert G., and Jeff Borland. 1999. “Recent Developments in Public Sector Labor
Markets,” in Handbook of Labor Economics, Vol. 3c. edited by O. Ashenfelter and David
Card, North Holland, Ch 53, 3573-3630.
Hatch, Julie. 2004. “Employment in the Public Sector: Two Recessions’ Impact on Jobs,”
Monthly Labor Review 127(10) 38-47.
Holtman, A.G., and Todd L. Idson. 1993. “Wage Determination of Registered Nurses in
Proprietary and Nonprofit Nursing Homes.” Journal of Human Resources 28(1):55-79.
Horwitz, Jill R. 2005. “Does Corporate Ownership Matter? Service Provision in the Hospital
Industry.” NBER Working Paper #11376, May.
Hirsch, Barry T. 2005. “Why Do Part-Time Workers Earn Less? The Role of Worker and Job
Skills,” Industrial and Labor Relations Review, 58(4): 525-51.
Hirsch, Barry T., and Edward J. Schumacher. 2004. “Match Bias in Wage Gap Estimates Due to
Earnings Imputation.” Journal of Labor Economics 22(3): 689-722.
Ippolito Richard. 1987. "Why Federal Workers Don't Quit," Journal of Human Resources,
22(2):281-99.
Katz L. and A. Krueger, "Changes in the Structure of Wages in the Public and Private Sectors,"
Research in Labor Economics, 12:137-72.
Krueger, Alan B. 1988. “Are Public Sector Workers Paid More Than Their Alternative Wage?
Evidence from Longitudinal Data and Job Queues.” In When Public Sector Workers
Unionize, ed. Richard B. Freeman and Casey Ichniowski, 217-40. Chicago Univeristy
Press.
Moulton Brent R. 1990. "A Reexamination of the Federal-Private Wage Differential in the United
States," Journal of Labor Economics, 8(2):270-93.
18
Roomkin, Myron J., and Burton A. Weisbrod, 1999. “Managerial Compensation and Icentives in
For-Profit and Nonprofit Hospitals.” Journal of Law, Economics, and Organization
15(3):750-28.
Ruhm, Christopher J., and Carey Borkoski. 2003. “Compensation in the Nonprofit Sector,” The
Journal of Human Resources, 38(5):992-1021.
Sloan, Frank A. 2000. “Not-for-Profit Ownership and Hospital Behavior,” in A.J. Culyer and J.P.
Newhouse, Handbook of Health Economics, Elsevier Science B.V., 2000. 1141-74.
Sloan, Frank A., and Bruce Steinwald. 1980. Hospital Labor Markets: Analysis of Wages and
Work Force Composition, D.C. Health-Lexington Books, Lexington, MA.
19