Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Dec 2.
Published in final edited form as: Policy Polit Nurs Pract. 2014 Aug 13;15(0):72–80. doi: 10.1177/1527154414546868

Wage, Work Environment, and Staffing: Effects on Nurse Outcomes

Matthew D McHugh 1, Chenjuan Ma 2
PMCID: PMC4667784  NIHMSID: NIHMS739772  PMID: 25121923

Abstract

Research has shown that hospitals with better nurse staffing and work environments have better nurse outcomes—less burnout, job dissatisfaction, and intention to leave the job. Many studies, however, have not accounted for wage effects, which may confound findings. By using a secondary analysis with cross-sectional administrative data and a four-state survey of nurses, we investigated how wage, work environment, and staffing were associated with nurse outcomes. Logistic regression models, with and without wage, were used to estimate the effects of work environment and staffing on burnout, job dissatisfaction, and intent to leave. We discovered that wage was associated with job dissatisfaction and intent to leave but had little influence on burnout, while work environment and average patient-to-nurse ratio still have considerable effects on nurse outcomes. Wage is important for good nurse outcomes, but it does not diminish the significant influence of work environment and staffing on nurse outcomes.

Keywords: burnout, job satisfaction, outcomes, quality of work environment, staffing levels

Introduction

Burnout and job dissatisfaction are perennial problems resulting in costly employee turnover (Larrabee et al., 2003) and poor patient outcomes (McHugh, Kutney-Lee, Cimiotti, Sloane, & Aiken, 2011; Vahey, Aiken, Sloane, Clarke, & Vargas, 2004). Nurses working in hospitals with excessive patient workloads and poor work environments are more likely to be burned out and dissatisfied with their job (Aiken, Clarke, Sloane, Lake, & Cheney, 2008; Aiken, Clarke, Sloane, Sochalski, & Silber, 2002; Maslach, Schaufeli, & Leiter, 2001). These conditions—the level of nurse staffng and the quality of nurses’ work environment—can be changed through good management and organizational practices that value professional nursing.

A satisfactory wage is a significant factor in job-seeking behavior and is especially important in keeping workers in their current positions. Increasing wage to solve institutional workforce recruitment and retention problems is an easy-to-implement intervention in the short run (May, Bazzoli, & Gerland, 2006). Wage, however, is not the only factor; many nonwage job characteristics are important considerations for workers in selecting and staying at a workplace (Antonazzo, Scott, Skatun, & Elliott, 2003; Blau, 1991; Chiha & Link, 2003; García & Molina, 1999; Kovner, Brewer, Wu, Cheng, & Suzuki, 2006; Shields, 2004; Woodbury, 1983). Work takes place within a larger context of hierarchies, relationships, management environments, ethical climates, operating rules, resources, and space distribution (Maslach et al., 2001). Work environments that are more favorable to workers are associated with lower burnout, job dissatisfaction, and intent to leave across a number of sectors, including health care and nursing (Aiken et al., 2008; Kovner et al., 2006; Kovner, Brewer, Greene, & Fairchild, 2009; Maslach & Jackson, 1982; Shields, 2004; Shields & Ward, 2001; Ulrich et al., 2007). To nurses, the organizational climate may be as or even more important than wages as a reason for staying in their job (Hayes et al., 2006; Stone et al., 2007).

Wage increases may need to be combined with non-pecuniary factors, especially modifiable hospital factors such as work environment and workload, to recruit adequate numbers of nurses and prevent them from leaving the hospital setting to work in other jobs (Buerhaus, 1991; Ehrenberg & Smith, 2008; McHugh et al., 2011; Spetz & Given, 2003). Frequently missing from nurse outcomes studies, however, is the potentially confounding effect of wage. One might expect that better staffed hospitals with the best work environments also pay higher wages to their nurses. If true, work environment and staffing might be acting as a proxy when better wages is actually the determining factor for mitigating poor nurse outcomes like burnout and job dissatisfaction. Another possibility is that hospitals with less favorable work environments or staffing levels pay higher wages to compensate for poor work conditions. In either case, failing to account for the effects of both wage and nonwage factors on nurse outcomes could confound results. Our aim was to clarify how wage, work environment, and staffing are associated with burnout, job dissatisfaction, and intent to leave.

Methods

Data and Sample

Nurses

The parent study of our cross-sectional secondary analysis used a two-stage sampling design with surveys mailed to the homes of registered nurses in four states (California, Florida, New Jersey, and Pennsylvania) in 2006–2007. The response rate was 39%—more than 100,000 respondents. To address potential nonresponse bias, another random sample of 1,300 nonresponders generated a response rate of 91%. A comparison of the two samples indicated no response bias (Aiken et al., 2011; Smith, 2009). The sampling approach has been detailed elsewhere (Aiken et al., 2011). The survey collected information on individual nurses’ demographic characteristics, work status, setting, role, burnout, and job satisfaction.

Hospitals

The nurse survey detailed earlier was also the source of information about hospital work environment and nurse staffing levels. Hospital nurses provided their employers’ names, allowing us to aggregate nurses’ responses for work environment and staffing measures. We limited analysis to data from staff nurse respondents working in adult, nonfederal acute care hospitals with at least 10 respondents. We also used Magnet hospital status—designated through the American Nurses Credentialing Center (ANCC) Magnet Recognition program—as an alternative measure indicating the presence of a good work environment for nurses. Magnet-recognized hospitals have been demonstrated to exemplify good places for nurses to work, and outcomes for both nurses and patients are better in Magnet versus non-Magnet hospitals (Aiken et al., 2008; Lake, Shang, Klaus, & Dunton, 2010; Lake et al., 2012; McHugh et al., 2013).

Information on nonnursing structural hospital characteristics was obtained from the 2006 American Hospital Association (AHA) Annual Survey. These characteristics included number of beds, technological status, teaching status, and geographic location.

Wage

Average hourly wage data for registered nurses working as staff nurses in each hospital came from the 2006 Centers for Medicare & Medicaid Services’ (CMS) Medicare Wage Index Occupational Mix Survey. The survey collected average hourly wage data for nurses and other workers from Medicare-participating hospitals. The survey differentiated patient care nurses from those in administrative roles. The survey did not include data for nurses in skilled-nursing, psychiatric, or rehabilitation facilities, allowing us to focus on staff nurses in adult, nonfederal acute care hospitals.

Measures

Burnout

As in earlier works (Kelly, McHugh, & Aiken, 2011; McHugh et al., 2011), burnout was measured using the Emotional Exhaustion subscale of the Maslach Burnout Inventory. The intraclass correlation coefficient ICC(1,k) for the Emotional Exhaustion subscale was acceptable at .62 with 10 nurses per hospital (Glick, 1985). Nurses were classified as burned out if their score on this subscale was higher than or equal to 27, the published average for health care workers (Maslach & Jackson, 1982, 1986).

Job dissatisfaction

Job dissatisfaction was measured using nurses’ responses to the question, “How satisfied are you with your current job?” The 4-point Likert-type scale response options ranged from very satisfied to very dissatisfied. We dichotomized the measure such that nurses who reported being either very dissatisfied or a little dissatisfied were described as dissatisfied and nurses reporting being moderately satisfied or very satisfied were described as satisfied (Kelly et al., 2011).

Work environment

Work environment was measured using the Practice Environment Scale of the Nursing Work Index (PES-NWI; Lake, 2002). We used the four subscales of the PES-NWI, measuring nurse manager ability, leadership, and support; nursing foundations for quality care; collegial nurse–physician relations; and nurses’ participation in hospital affairs. The staffng and resource adequacy subscale was omitted due to high correlation with our staffng measure—patient-to-nurse ratio. An overall PES-NWI score for each hospital was calculated as the mean of the hospital-level subscales. The intraclass correlation coefficient ICC(1, k) for the hospital composite was acceptable at .85 with 10 nurses per hospital (Glick, 1985). For descriptive purposes, hospitals were classified into three categories based on the PES-NWI scores: hospitals in the top quartile were classified as good, hospitals in the low quartile as poor, and hospitals in between as mixed. In regression models estimating the effect of work environment on nurse outcomes, continuous standardized PES-NWI scores were used.

We also evaluated models substituting an alternative indicator of a good work environment—Magnet hospital status. Hospitals recognized as Magnet hospitals by the ANCC as of 2007 were included (53 Magnet and 481 non-Magnet hospitals in our sample).

Staffing

Staffing was measured as the hospital’s average patient-to-nurse ratio from the nurse survey data by dividing the average number of patients reported by nurses on their units on their last shift by the average number of nurses on that unit (Aiken et al., 2011). Hospitals were categorized into three groups for descriptive purposes: hospitals in the highest quartile as having poor staffing, hospitals in the lowest quartile as having good staffing, and hospitals in between as having mixed staffing. In regression models, a continuous standardized patient-to-nurse ratio was used.

Wages

The average hourly wage (total of paid wages and salaries/total paid hours) data for staff nurses, including charge nurses but excluding nurses in administrative or leadership positions, were obtained from the 2006 CMS Medicare Wage Index Occupational Mix Survey. Total paid wages and salaries included overtime, vacation, holiday, sick, lunch, severance, other paid time off, and bonuses, but not fringe benefits or wage-related costs. Paid hours included regular hours, overtime hours, paid holiday, vacation, sick, severance pay hours, and other paid-time-off hours.

Hospital characteristics

Our analytic models included a number of hospital structural characteristics as covariates that may affect nurse outcomes (Aiken et al., 2002). These variables were drawn from the AHA Annual Survey. Teaching status was a categorical variable where hospitals without any trainees (postgraduate medical residents or fellows) were nonteaching, hospitals with a 1:4 or smaller trainee-to-bed ratio were minor teaching, and those with higher than 1:4 trainee-to-bed ratios were major teaching. The number of beds was categorized as small (<100 beds), medium (101–250 beds), or large (>251 beds). High-technology hospitals were designated with a binary variable, where high-technology hospitals were those facilities that provided services for open heart surgery, organ transplantation, or both. We used the Herfindahl–Hirschman index—the sum of squared market shares for the hospital service area—as a market competition indicator. We used dummy variables for state as well as urban–rural location (urbanized areas=50,000 or more people; urban clusters=49,999–2,500 people; and rural areas encompass any area not included within an urban area or cluster).

Nurse characteristics

We included nurse-level demographic information that may have affected our nurse outcomes. These variables included gender, type of unit the nurse worked on (medical–surgical, intensive or critical care, or other), highest level of education attained (less than a Bachelor of Science in Nursing [BSN] degree vs. a BSN degree or higher), and years of experience as a registered nurse.

Data Analysis

The data sources were merged to create a final analytic sample of 26,005 registered nurses from 534 hospitals in the four states. First, we assessed the characteristics of nurses and the hospitals in which they worked. We then described the average wages by hospital characteristics. Logistic regression models were used to estimate the effects of wage, work environment, and nurse staffng on three separate nurse outcomes—burnout, job dissatisfaction, and intent to leave—while accounting for individual nurse and hospital characteristics. Our independent variables were standardized, which allowed us to interpret our results as the expected change in the outcome, corresponding with a one standard deviation (SD) change in the predictors. This permitted more relevant comparisons of the effects of different predictors on outcomes. We estimated robust standard errors and significance levels that accounted for the clustering of individual nurses within hospitals (White, 1980; Williams, 2000). The study protocol for the parent study was approved by the University of Pennsylvania institutional review board.

Results

Table 1 shows characteristics of the 534 study hospitals. Roughly half of the hospitals were nonteaching hospitals; among the teaching hospitals, most were minor teaching hospitals with a trainee-to-bed ratio of 0.25 or less. More than 90% of the hospitals in our study had more than 100 beds. Approximately 10% of the hospitals in our sample were Magnet hospitals. About 2% of the hospitals were located in rural areas. The mean score of nurse work environment was 2.73 (SD=0.22). On average, the patient-to-nurse ratio among these hospitals was approximately 5:1 (SD=1.1). Nurses’ average hourly wage was $37.20 (SD=$8.84).

Table 1.

Hospital Characteristics and Average Hourly Wage by Hospital Characteristics (n=534).

n (%) Mean wage (SD)
Nurse work environment
 Poor 134 (25) $33.82 (6.61)
 Mixed 267 (50) $37.49 (9.20)
 Good 133 (25) $40.49 (9.05)
Nurse staffing
 Poor 134 (25) $32.23 (6.48)
 Mixed 267 (50) $36.43 (7.75)
 Good 133 (25) $43.67 (9.11)
Magnet hospital
 Magnet 53 (10) $36.29 ($6.54)
 Non-Magnet 481 (90) $37.30 ($9.05)
High technology
 High technology 246 (46) $38.22 (8.86)
 Not high technology 288 (54) $36.54 (8.85)
Teaching status
 No 273 (51) $36.97 (9.22)
 Minor 217(41) $37.67 (8.65)
 Major 44 (8) $37.70 (8.00)
Bed size
 ≤100 51 (10) $34.80 (9.35)
 101–250 234 (44) $36.79 (9.22)
 >250 249 (47) $38.33 (8.34)
Ownership
 Not for profit 444 (83) $37.35 ($9.08)
 For profit 90 (17) $36.47 ($7.50)
State
 California 204 (38) $45.27 (7.72)
 Florida 132 (25) $32.15 (3.61)
 New Jersey 69 (13) $37.58 (3.69)
 Pennsylvania 129 (24) $29.72 (5.03)
Urban–rural status
 Urban area 482 (90) $37.91 ($8.64)
 Urban cluster 40 (7) $31.03 ($8.07)
 Rural 12 (2) $29.33 ($7.57)
Open in a new tab

Descriptive information on registered nurses’ average hourly wage by hospital characteristics is also presented in Table 1. The average hourly wage was higher in hospitals with good versus mixed and mixed versus poor work environments. There was little difference, however, in the average wage between Magnet and non-Magnet hospitals. Average hourly wage was higher in hospitals with lower patient-to-nurse ratios. The average hourly wage was also higher in teaching hospitals or hospitals with high technology or more beds. Similarly, wage was higher for hospitals located in urban areas compared with hospitals in rural areas. There was variation in the average wage across the states: Hospitals in California had the highest average wage ($45/hour), whereas hospitals in Pennsylvania had the lowest ($30/hour).

Table 2 describes the nurse characteristics and outcomes in the final sample. Of the 26,005 nurse respondents, approximately 25% expressed dissatisfaction with their current job and about 34% reported high burnout. Roughly 14% of the nurses expressed intent to leave their current position within the year.

Table 2.

Nurse Characteristics and Outcomes (n=26,005).

n
Characteristics
 Age, mean (SD) 44.7 (10.7)
 Work experience, mean (SD) 16.97 (11.2)
 Female 23,704 (93.2%)
 BSN degree or above 10,460 (41.0%)
Outcomes
 Job dissatisfaction 6,199 (24.8%)
 Burnout 8,489 (33.6%)
 Intent to leave 3,455 (13.7%)
Open in a new tab

Note. BSN=Bachelor of Science in Nursing.

We examined whether the differences seen between hospitals with good work environments and the other hospitals was due to their location in urban areas. Tabulating the work environment and staffing data by urban–rural location (Table 3) shows that the average wage was still generally higher in hospitals with better work environments or good staffing levels. A similar relationship was found between urban–rural locations and staffing: Average wage was higher in hospitals with lower patient-to-nurse ratios. All the Magnet hospitals were in urban areas, preventing such an evaluation.

Table 3.

Mean (SD) Hourly Wage by Work Environment, Staffing, and Urban–Rural Area Classification.

Urban area Urban cluster Rural
Work environment
 Poor $33.82 ($6.07) $30.15 ($7.21) $28.40 ($5.40)
 Mixed $38.43 ($8.86) $30.02 ($7.91) $30.77 ($9.00)
 Good $40.80 ($8.92) $35.58 ($9.41) $23.00 (–)
Nurse staffing
 Poor $33.52 ($6.01) $26.98 ($4.91) $27.63 ($7.92)
 Mixed $36.73 ($7.66) $32.86 ($8.60) $32.74 ($6.38)
 Good $43.71 ($9.24) $42.45 ($3.30) –
Open in a new tab

In addition to unadjusted models, we estimated two versions of our final models, both of which jointly consider the effects of work environment (measured alternatively with the PES-NWI and Magnet status) and staffing when controlling for nurse characteristics and structural hospital factors (Table 4). The difference between the models was that one was estimated without wage, while the other included wage so that we could evaluate how our estimates would change when wage was held constant.

Table 4.

Odds Ratios Indicating the Effect of Wage, Work Environment, and Staffing on Individual Nurse Outcomes.

Adjusted ORs [95% CI]
Unadjusted
ORs [95% CI]		 PES-NWI
 Magnet Hospital
Without wage With wage Without wage With wage
Job dissatisfaction
 Wage 0.81 [0.77, 0.86]*** – 0.91 [0.86, 0.96]*** – 0.91 [0.84, 0.99]*
 Work environment 0.66 [0.64, 0.69]*** 0.70 [0.67, 0.72]*** 0.70 [0.67, 0.72]*** – –
 Magnet status 0.82 [0.71, 0.94]** – – 0.82 [0.72, 0.94]** 0.82 [0.72, 0.94]**
 Staffinga 0.78 [0.75, 0.82]*** 0.90 [0.86, 0.95]*** 0.91 [0.86, 0.95]*** 0.84 [0.79, 0.90]*** 0.84 [0.79, 0.90]***
Burnout
 Wage 0.88 [0.84, 0.92]*** – 0.96 [0.91, 1.02] – 0.97 [0.90, 1.04]
 Work environment 0.77 [0.74, 0.79]*** 0.78 [0.75, 0.81]*** 0.77 [0.75, 0.81]*** – –
 Magnet status 0.90 [0.80, 1.00] – – 0.87 [0.78, 0.97]* 0.87 [0.78, 0.97]*
 Staffinga 0.86 [0.82, 0.89]*** 0.92 [0.87, 0.96]** 0.92 [0.88, 0.96]** 0.87 [0.83, 0.92]*** 0.87 [0.83, 0.92]***
Intent to leave
 Wage 0.89 [0.84, 0.94]*** – 0.89 [0.82, 0.97]** – 0.90 [0.81, 0.99]*
 Work environment 0.75 [0.71, 0.80]*** 0.72 [0.68, 0.76]*** 0.72 [0.68, 0.76]*** – –
 Magnet status 0.81 [0.70, 0.94]*** – – 0.84 [0.73, 0.97]* 0.84 [0.73, 0.97]*
 Staffinga 0.94 [0.89, 0.99]* 0.93 [0.86, 0.99]* 0.93 [0.87, 1.00] 0.87 [0.80, 0.94]*** 0.86 [0.80, 0.94]***
Open in a new tab

Note. PES-NWI=Practice Environment Scale of the Nursing Work Index; OR=odds ratio; CI=confidence interval. For interpretability, we report standardized coefficients. Unadjusted estimates are based on models evaluating the effect of a predictor variable, separately, without accounting for any other covariates. Adjusted models account for nurse gender, education level, unit type, years of experience, and hospital characteristics including market competition with the Herfindahl–Hirschman index, teaching status, number of beds, technology level, ownership, state, and urban–rural location.

a

To interpret coefficients in a similar direction, the standardized staffing variable is based on the reciprocal of the patient-to-nurse ratio.

*

p <.05.

**

p<.01.

***

p<.001.

Our results related to job dissatisfaction show that nurses working in hospitals with better work environments and better staffing have lower odds of job dissatisfaction. When we added wage to the models, the effect of wage was significant, but the effects of work environment and staffing remained essentially unchanged. Thus, an odds ratio (OR) of 0.70 (95% confidence interval [CI] [0.67, 0.72]) for work environment suggests that even when we account for wage and all other covariates, nurses working at hospitals at the 50 th versus the 16 th percentile or the 84 th versus the 50 th percentiles in terms of the work environment had 30% lower odds of job dissatisfaction. The OR for both staffing (OR=0.91; 95% CI [0.86, 0.95]) and wage (OR=0.91; 95% CI [0.86, 0.96]) suggests that a one standard deviation change in either average wage or average staffing was associated with 9% lower odds of nurses reporting job dissatisfaction. The same was true for our models that substituted Magnet status as an indicator of a good work environment. The odds of job dissatisfaction was 18% (OR=0.82; 95% CI [0.72, 0.94]) lower in Magnet compared with non-Magnet hospitals, accounting for wage and all other covariates.

Our results related to burnout were similar to job dissatisfaction. The notable exception was that there was no statistically significant relationship between wage and burnout in our adjusted models with either the PES-NWI (OR for wage=0.96; 95% CI [0.91, 1.02] in the PES-NWI model) or Magnet status (OR for wage=0.97; 95% CI [0.90, 1.04] in the Magnet status model).

Work environment and staffing were significantly associated with intent to leave even when we accounted for average wage. The one caveat was related to models measuring the work environment with the PES-NWI—when we go from a model not including wage to one that includes wage, the relationship between staffing and intent to leave becomes statistically insignificant at the p=.05 level (p=.058). The estimate, however, remained essentially unchanged (OR=0.93; 95% CI [0.86, 0.99] in the PES-NWI model without wage; OR=0.93; 95% CI [0.87, 1.00] in the PES-NWI model with wage). The estimates for staffing in models using Magnet status, both with and without the inclusion of wage, were statistically significant. Interactions between the work environment and wage, Magnet status and wage, and staffing and wage were not significant.

Discussion

The significant association between more favorable nurse work environments and nurse outcomes, net of wage effects, implies that wages are important, but they do not account for the better outcomes associated with the work environment and nurse staffing. This supports earlier findings that more favorable work environments are associated with lower burnout, job dissatisfaction, and intent to leave across a number of sectors, including nursing (Aiken et al., 2008; Kovner et al., 2006, 2009; Maslach & Jackson, 1982; Stone et al., 2007). Our findings are also consistent with the literature demonstrating that it requires more than good pay to attract nurses into hospital employment and keep them working there (McCloskey, 1974; Stone et al., 2007).

Transforming the organizational culture to support and integrate a model of professional nursing practice can be a valuable investment for hospitals (DeBaca, Jones, & Tornabeni, 1993; Mark, Lindley, & Jones, 2009; Needleman & Hassmiller, 2009; Zelauskas & Howes, 1992), but it requires organizational willingness and commitment. The Magnet Recognition Program is one approach offering a concrete model for creating a good work environment for nurses. When we substituted Magnet hospital status as an indicator of the work environment, our findings showed that—holding wage constant—Magnet recognition was associated with lower odds of burnout, job dissatisfaction, and intent to leave. The Magnet estimates are likely conservative given the low ratio of Magnet hospitals.

The finding that a higher patient-to-nurse ratio was associated with job dissatisfaction and burnout, regardless of average wage, supports previous research (Aiken et al., 2002). An excessive workload exhausts workers’ energy and makes recovery impossible. Effective and gratifying work—the satisfaction that comes from providing good quality care to patients—becomes less attainable. Although hiring more nurses can be costly, some of these costs would be offset by increased productivity, a reduction in turnover and retraining costs, and, more importantly, better patient outcomes (Dall, Chen, Seifert, Maddox, & Hogan, 2009; Jones, 2004; Needleman & Hassmiller, 2009; Rothberg, Abraham, Lindenauer, & Rose, 2005).

Our intent was not to discount the importance of wage; wages are an important tool for administrators to use to attract and expand their workforce (Buerhaus, 2008). Our goal, however, was to evaluate whether, as we found, the effects of the work environment and staffing persisted when we accounted for wage. The wage effect, as one might expect, was still important—except in the instance of burnout. Wage may do little to compensate for the conditions leading to burnout. Reforming underlying work conditions at the root of emotional exhaustion may be key to reducing and preventing burnout.

Although nurses are paid relatively well in the United States, wages are compressed (Greipp, 2003) and hospitals can set nurses’ wages below their value (Staiger, Spetz, & Phibbs, 2010). This has not gone unnoticed. Nurses, particularly those in direct patient care roles in hospitals and nursing homes, have reported dissatisfaction with wages, as well as nonwage benefits such as health care, tuition reimbursement, and retirement benefits (McHugh et al., 2011). Wage rates and distribution should match nurse skill level to encourage entry into the profession and retention within the institution. Competitive wages, combined with good benefits and nonpecuniary factors, may be necessary to recruit adequate numbers of nurses to meet the ongoing care demands of the upcoming decades and prevent cyclical shortages that have defined the past half century (Spetz & Adams, 2006; Spetz & Given, 2003).

Researchers have estimated that California increased wages 12% higher than other metropolitan hospitals outside of California between 2000 and 2006 (Mark, Harless, & Spetz, 2009). It has been suggested that this rise in wages was a consequence of California’s mandate that limited the number of patients that nurses could care for at a given time. We also find that California’s hospital wages were higher compared with hospitals in the other states in our analysis. However, our models estimating the effect of wage, work environment, and staffing on nurse outcomes account for differences by state.

There are limitations to our study. Foremost, the cross-sectional design limited our ability to draw causal inferences. The study, however, strengthens the basis for causal inference by accounting for an important potential confounder that had previously been omitted from most research.

Another limitation is that our survey and wage data predate the economic recession that began at the end of 2007. In an environment where any job is difficult to find, the financial downturn could have tempered perceptions of less desirable aspects of the work environment. A number of factors, however, limit our concern here. The first is that, over the many investigations examining the association between work environment and nurse outcomes, the relationships that we found have been consistently identified regardless of time (Aiken et al., 2002; Aiken & Patrician, 2000; Hare & Skinner, 1990; Kelly et al., 2011; Kutney-Lee et al., 2009; Larrabee et al., 2003; McHugh et al., 2011). What is important about our findings is that we show that these relationships persist when we account for wages as a potential confounder.

Furthermore, although work environments may improve over time, evidence suggests that where changes occur, they are associated with changes in nurse outcomes, that is, if work environments and staffing improve, rates of burnout, job dissatisfaction, and intent to leave decrease (Kutney-Lee, Wu, Sloane, & Aiken, 2013). Additionally, research showed that the percentage of direct care hospital nurses who were very satisfied with their present job in 2006 (prerecession) was 29% compared with 28% in 2008 (during the recession). The percentage of nurses who would rate the quality of the salary and benefits package in their current or most recent work setting as excellent or very good was not significantly different from 2006 (28%) to 2008 (30%). Researchers also suggested that the recession effects on the nursing workforce were temporary (Staiger, Auerbach, & Buerhaus, 2012). We expect that if we repeated our study today, we would find the same relationships. Nevertheless, additional study is warranted.

Another consideration is that our outcomes were nurse-specific, but our wage data were not. Although nurse-level wage data might have allowed a more refined analysis of the association between individual wage and our nurse outcomes, we expect that the gains would have been marginal given wage compression. Additionally, some rural hospitals are not represented in our data because there were an insufficient number of nurses from small hospitals for reliable estimates. Many of such hospitals are Critical Access Hospitals, which are not paid under the Inpatient Prospective Payment System (IPPS) and, therefore, were not required to complete the survey that was the source of our wage data.

We also do not measure union presence or membership, although evidence varies regarding how much of a premium unions confer (Ash & Seago, 2004). We do, however, include many covariates we expect would at least partially account for the variation associated with unionization, including state and urban–rural indicators, teaching status, ownership, and size.

Last, the relative impact of wage on nurse outcomes may vary in different countries. Some studies conducted outside the United States, for example, have shown that dissatisfaction with wage played the biggest role with nurse job dissatisfaction and turnover (Fochsen, Sjogren, Josephson, & Lagerstrom, 2005; Klopper, Coetzee, Pretorius, & Bester, 2012; Palmer, 2014). These studies, however, also found that characteristics of the work environment were important contributing factors. Multihospital studies outside the United States and cross-national studies would be valuable.

Conclusion

Our findings suggest that better wages do not explain the relationship between working in well-staffed hospitals with good practice environments and nurse outcomes such as nurse burnout, job dissatisfaction, and intent to leave. Although good wages are important, interventions that improve the work environment and maintain reasonable staffing levels may be more critical to attracting and retaining satisfied nurses in the hospital workforce.

Acknowledgments

The authors thank Ms. Mikaella Hill and Ms. Myra Eckenhoff for their assistance with the preparation of this manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Robert Wood Johnson Foundation Nurse Faculty Scholars program (to M. D. McHugh) and the National Institute of Nursing Research (R01-NR-004513 and P30-NR-005043; PI: L. Aiken).

Biographies

Matthew D. McHugh, PhD, JD, MPH, RN, FAAN is The Rosemarie Greco Term Endowed Associate Professor in Advocacy at the Center for Health Outcomes & Policy Research at the University of Pennsylvania School of Nursing.

Chenjuan Ma, PhD is a post-doctoral fellow with The National Database of Nursing Quality Indicators, University of Kansas Medical Center.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

  1. Aiken LH, Cimiotti JP, Sloane DM, Smith HL, Flynn L, Neff DF. Effects of nurse staffing and nurse education on patient deaths in hospitals with different nurse work environments. Medical Care. 2011;49(12):1047–1053. doi: 10.1097/MLR.0b013e3182330b6e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Aiken LH, Clarke SP, Sloane DM, Lake ET, Cheney T. Effects of hospital care environment on patient mortality and nurse outcomes. Journal of Nursing Administration. 2008;38(5):223–229. doi: 10.1097/01.NNA.0000312773.42352.d7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Aiken LH, Clarke SP, Sloane DM, Sochalski J, Silber JH. Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA. 2002;288(16):1987–1993. doi: 10.1001/jama.288.16.1987. [DOI] [PubMed] [Google Scholar]
  4. Aiken LH, Patrician PA. Measuring organizational traits of hospitals: The Revised Nursing Work Index. Nursing Research. 2000;49(3):146–153. doi: 10.1097/00006199-200005000-00006. [DOI] [PubMed] [Google Scholar]
  5. Antonazzo E, Scott A, Skatun D, Elliott R. The labour market for nursing: A review of the labour supply literature. Health Economics. 2003;12(6):465–478. doi: 10.1002/hec.737. [DOI] [PubMed] [Google Scholar]
  6. Ash M, Seago JA. The effect of registered nurses’ unions on heart-attack mortality. Industrial & Labor Relations Review. 2004;57(3):422–442. [Google Scholar]
  7. Blau DM. Search for nonwage job characteristics: A test of the reservation wage hypothesis. Journal of Labor Economics. 1991;9(2):186–205. [Google Scholar]
  8. Buerhaus PI. Economic determinants of annual hours worked by registered nurses. Medical Care. 1991;29(12):1181–1195. doi: 10.1097/00005650-199112000-00002. [DOI] [PubMed] [Google Scholar]
  9. Buerhaus PI. The potential imposition of wage controls on nurses: A threat to nurses, patients, and hospitals. Nursing Economics. 2008;26(4):276–279. [PubMed] [Google Scholar]
  10. Buerhaus PI, DesRoches C, Donelan K, Hess R. Still making progress to improve the hospital workplace environment? Results from the 2008 National Survey of Registered Nurses. Nursing Economics. 2009;27(5):289–301. [PubMed] [Google Scholar]
  11. Chiha YA, Link CR. The shortage of registered nurses and some new estimates of the effects of wages on registered nurses labor supply: A look at the past and a preview of the 21st century. Health Policy. 2003;64(3):349–375. doi: 10.1016/s0168-8510(02)00197-5. [DOI] [PubMed] [Google Scholar]
  12. Dall TM, Chen YJ, Seifert RF, Maddox PJ, Hogan PF. The economic value of professional nursing. Medical Care. 2009;47(1):97–104. doi: 10.1097/MLR.0b013e3181844da8. [DOI] [PubMed] [Google Scholar]
  13. DeBaca V, Jones K, Tornabeni J. A cost-benefit analysis of shared governance. Journal of Nursing Administration. 1993;23(7–8):50–57. doi: 10.1097/00005110-199307000-00014. [DOI] [PubMed] [Google Scholar]
  14. Ehrenberg R, Smith R. Modern labor economics: Theory and public policy. 10th Addison-Wesley; Reading, MA: 2008. [Google Scholar]
  15. Fochsen G, Sjogren K, Josephson M, Lagerstrom M. Factors contributing to the decision to leave nursing care: A study among Swedish nursing personnel. Journal of Nursing Management. 2005;13(4):338–344. doi: 10.1111/j.1365-2934.2005.00546.x. [DOI] [PubMed] [Google Scholar]
  16. García I, Molina JA. How do workers decide their jobs? The influence of income, wage and job characteristics. Managerial and Decision Economics. 1999;20(4):189–204. [Google Scholar]
  17. Glick WH. Conceptualizing and measuring organizational and psychological climate: Pitfalls in multilevel research. Academy of Management Review. 1985;10(3):601–616. [Google Scholar]
  18. Greipp ME. Salary compression: Its effect on nurse recruitment and retention. Journal of Nursing Administration. 2003;33(6):321–323. doi: 10.1097/00005110-200306000-00004. [DOI] [PubMed] [Google Scholar]
  19. Hare J, Skinner DA. The relationship between work environment and burnout in nursing home employees. Journal of Long Term Care Administration. 1990;18(3):9–12. [PubMed] [Google Scholar]
  20. Hayes LJ, O’Brien-Pallas L, Duffield C, Shamian J, Buchan J, Hughes F, Stone PW. Nurse turnover: A literature review. International Journal of Nursing Studies. 2006;43(2):237–263. doi: 10.1016/j.ijnurstu.2005.02.007. [DOI] [PubMed] [Google Scholar]
  21. Jones CB. The costs of nurse turnover: Part 1: An economic perspective. Journal of Nursing Administration. 2004;34(12):562–570. doi: 10.1097/00005110-200412000-00006. [DOI] [PubMed] [Google Scholar]
  22. Kelly LA, McHugh MD, Aiken LH. Nurse outcomes in Magnet® and non-Magnet hospitals. Journal of Nursing Administration. 2011;41(10):428–433. doi: 10.1097/NNA.0b013e31822eddbc. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Klopper HC, Coetzee SK, Pretorius R, Bester P. Practice environment, job satisfaction and burnout of critical care nurses in South Africa. Journal of Nursing Management. 2012;20(5):685–695. doi: 10.1111/j.1365-2834.2011.01350.x. [DOI] [PubMed] [Google Scholar]
  24. Kovner CT, Brewer C, Wu Y-W, Cheng Y, Suzuki M. Factors associated with work satisfaction of registered nurses. Journal of Nursing Scholarship. 2006;38(1):71–79. doi: 10.1111/j.1547-5069.2006.00080.x. [DOI] [PubMed] [Google Scholar]
  25. Kovner CT, Brewer CS, Greene W, Fairchild S. Understanding new registered nurses’ intent to stay at their jobs. Nursing Economics. 2009;27(2):81–98. [PubMed] [Google Scholar]
  26. Kutney-Lee A, McHugh MD, Sloane DM, Cimiotti JP, Flynn L, Neff DF, Aiken LH. Nursing: A key to patient satisfaction. Health Affairs (Millwood) 2009;28(4):w669–w677. doi: 10.1377/hlthaff.28.4.w669. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kutney-Lee A, Wu ES, Sloane DM, Aiken LH. Changes in hospital nurse work environments on nurse job outcomes: An analysis of panel data. International Journal of Nursing Studies. 2013;50(2):195–201. doi: 10.1016/j.ijnurstu.2012.07.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lake ET. Development of the practice environment scale of the Nursing Work Index. Research in Nursing & Health. 2002;25(3):176–188. doi: 10.1002/nur.10032. [DOI] [PubMed] [Google Scholar]
  29. Lake ET, Shang J, Klaus S, Dunton NE. Patient falls: Association with hospital Magnet status and nursing unit staffing. Research in Nursing and Health. 2010;33(5):413–425. doi: 10.1002/nur.20399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lake ET, Staiger D, Horbar J, Cheung R, Kenny MJ, Patrick T, Rogowski JA. Association between hospital recognition for nursing excellence and outcomes of very low-birth-weight infants. Journal of the American Medical Association. 2012;307(16):1709–1716. doi: 10.1001/jama.2012.504. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Larrabee JH, Janney MA, Ostrow CL, Withrow ML, Hobbs GR, Jr., Burant C. Predicting registered nurse job satisfaction and intent to leave. Journal of Nursing Administration. 2003;33(5):271–283. doi: 10.1097/00005110-200305000-00003. [DOI] [PubMed] [Google Scholar]
  32. Mark B, Harless DW, Spetz J. California’s minimum-nurse-staffing legislation and nurses’ wages. Health Affairs (Millwood) 2009;28(2):w326–w334. doi: 10.1377/hlthaff.28.2.w326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mark BA, Lindley L, Jones CB. Nurse working conditions and nursing unit costs. Policy, Politics, and Nursing Practice. 2009;10(2):120–128. doi: 10.1177/1527154409336200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Maslach C, Jackson SE. Burnout in health professions: A social psychologists analysis. In: Saunders G, Suls J, editors. Social psychology of health and illness. Lawrence Erlbaum Association; Hillsdale, NJ: 1982. pp. 227–251. [Google Scholar]
  35. Maslach C, Jackson SE. Maslach Burnout Inventory manual. 2nd Consulting Psychologists Press; Palo Alto, CA: 1986. [Google Scholar]
  36. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annual Review of Psychology. 2001;52(1):397–422. doi: 10.1146/annurev.psych.52.1.397. [DOI] [PubMed] [Google Scholar]
  37. May JH, Bazzoli GJ, Gerland AM. Hospitals’ responses to nurse staffing shortages. Health Affairs. 2006;25(4):W316–W323. doi: 10.1377/hlthaff.25.w316. [DOI] [PubMed] [Google Scholar]
  38. McCloskey J. Influence of rewards and incentives on staff nurse turnover rate. Nursing Research. 1974;23(3):239–247. [PubMed] [Google Scholar]
  39. McHugh MD, Kelly LA, Smith HL, Wu ES, Vanak J, Aiken LH. Lower mortality in Magnet hospitals. Medical Care. 2013;51(5):382–388. doi: 10.1097/MLR.0b013e3182726cc5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. McHugh MD, Kutney-Lee A, Cimiotti JP, Sloane DM, Aiken LH. Nurses’ widespread job dissatisfaction, burnout and frustration with health benefits signal problems for patient care. Health Affairs (Millwood) 2011;30(2):202–210. doi: 10.1377/hlthaff.2010.0100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Needleman J, Hassmiller S. The role of nurses in improving hospital quality and efficiency: Real-world results. Health Affairs. 2009;28(4):w625–w633. doi: 10.1377/hlthaff.28.4.w625. [DOI] [PubMed] [Google Scholar]
  42. Palmer SP. Nurse retention and satisfaction in Ecuador: Implications for nursing administration. Journal of Nursing Management. 2014;22(1):89–96. doi: 10.1111/jonm.12043. [DOI] [PubMed] [Google Scholar]
  43. Rothberg MB, Abraham I, Lindenauer PK, Rose DN. Improving nurse-to-patient staffing ratios as a cost-effective safety intervention. Medical Care. 2005;43(8):785–791. doi: 10.1097/01.mlr.0000170408.35854.fa. [DOI] [PubMed] [Google Scholar]
  44. Shields MA. Addressing nurse shortages: What can policy makers learn from the econometric evidence on nurse labour supply? The Economic Journal. 2004;114(499):F464–F498. [Google Scholar]
  45. Shields MA, Ward M. Improving nurse retention in the National Health Service in England: The impact of job satisfaction on intentions to quit. Journal of Health Economics. 2001;20(5):677–701. doi: 10.1016/s0167-6296(01)00092-3. [DOI] [PubMed] [Google Scholar]
  46. Smith HL. A double sample to minimize bias due to non-response in a mail survey (Population Studies Center Working Paper Series, No. 09-05) University of Pennsylvania, Population Studies Center; Philadelphia: 2009. [Google Scholar]
  47. Spetz J, Adams S. How can employment-based benefits help the nurse shortage? Health Affairs. 2006;25(1):212–218. doi: 10.1377/hlthaff.25.1.212. [DOI] [PubMed] [Google Scholar]
  48. Spetz J, Given R. The future of the nurse shortage: Will wage increases close the gap? Health Affairs. 2003;22(6):199–206. doi: 10.1377/hlthaff.22.6.199. [DOI] [PubMed] [Google Scholar]
  49. Staiger DO, Auerbach DI, Buerhaus PI. Registered nurse labor supply and the recession—Are we in a bubble? New England Journal of Medicine. 2012;366(16):1463–1465. doi: 10.1056/NEJMp1200641. [DOI] [PubMed] [Google Scholar]
  50. Staiger DO, Spetz J, Phibbs CS. Is there monopsony in the labor market? Evidence from a natural experiment. Journal of Labor Economics. 2010;28(2):211–236. [Google Scholar]
  51. Stone PW, Mooney-Kane C, Larson EL, Pastor DK, Zwanziger J, Dick AW. Nurse working conditions, organizational climate, and intent to leave in ICUs: An instrumental variable approach. Health Services Research. 2007;42(3):1085–1104. doi: 10.1111/j.1475-6773.2006.00651.x. Pt. 1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Ulrich C, O’Donnell P, Taylor C, Farrar A, Danis M, Grady C. Ethical climate, ethics stress, and the job satisfaction of nurses and social workers in the United States. Social Science & Medicine. 2007;65(8):1708–1719. doi: 10.1016/j.socscimed.2007.05.050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Vahey DC, Aiken LH, Sloane DM, Clarke SP, Vargas D. Nurse burnout and patient satisfaction. Medical Care. 2004;42(2 Suppl):II57–II66. doi: 10.1097/01.mlr.0000109126.50398.5a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. White H. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica. 1980;48(4):817–838. [Google Scholar]
  55. Williams RL. A note on robust variance estimation for cluster-correlated data. Biometrics. 2000;56(2):645–646. doi: 10.1111/j.0006-341x.2000.00645.x. [DOI] [PubMed] [Google Scholar]
  56. Woodbury SA. Substitution between wage and nonwage benefits. The American Economic Review. 1983;73(1):166–182. [Google Scholar]
  57. Zelauskas B, Howes DG. The effects of implementing a professional practice model. Journal of Nursing Administration. 1992;22(7/8):18–23. doi: 10.1097/00005110-199207000-00006. [DOI] [PubMed] [Google Scholar]

RESOURCES