Skip to main content
PMC home page
PLOS ONE logoLink to PLOS ONE
. 2024 Jul 24;19(7):e0306571. doi: 10.1371/journal.pone.0306571

The determinants of nonprofit hospital CEO compensation

Derek Jenkins 1,*, Marah N Short 1, Vivian Ho 1,2,3
Editor: Gabriel A Picone4
PMCID: PMC11268666  PMID: 39046937

Abstract

Hospital CEO salaries have grown quickly over the past two decades. We investigate correlates of rising nonprofit hospital CEO pay between 2012 and 2019 by merging compensation data from Candid’s IRS 990 forms with hospital data from the National Academy for State Health Policy Hospital Cost Tool. Almost half of the measured increase in CEO compensation (44.5%) accrued to a “base case” CEO, who was leading a non-teaching hospital system or independent hospital with fewer than 100 beds that earned 0 profits and provided no charity care. Another 28.5% of the measured salary increase resulted from changes in the generosity with which observable metrics were rewarded, particularly the reward for heading a system with 500 or more beds. The remaining 27% resulted mostly from hospital systems or single hospitals that increased their profits or bed size over time. The increase in CEO compensation associated with leading larger healthcare systems and earning greater profits may explain the increase in healthcare system consolidation which has occurred over the last several years.

Introduction

Hospital CEO salary has been growing quickly in the past two decades. Mean CEO compensation increased 92% from $1.6 million to $3.1 million among leaders of 22 major nonprofit medical centers between 2005 and 2015 [1]. While mean CEO pay by industry across the entire U.S. nonprofit sector ranged between only $100,000 and $200,000 in 2020, the mean pay of nonprofit hospital CEOs was around $620,000 [2]. In North Carolina, the nine largest health systems paid CEOs, chief financial officers, vice presidents and other highly paid administrative positions more than $1.75 billion from 2001–2021, with about 20 percent of this sum paid only to the CEOs of these health systems [3]. Not only is CEO compensation growing, but it is growing disproportionately relative to other healthcare workers; the wage gap between CEOs and mean salary of a registered nurse grew from 23:1 in 2005 to 44:1 in 2015 [1].

High levels of nonprofit CEO compensation may be warranted if the compensation is linked to better quality of care and greater community benefit. Joseph Newhouse first posited in 1970 that hospital trustees would reward nonprofit hospital CEOs for maximizing trustee utility instead of profits, by providing the optimal quantity and quality of care [4]. However, previous literature finds no association between better patient outcomes (i.e. risk adjusted mortality and readmissions) and executive compensation at nonprofit hospitals [5–7], although higher patient satisfaction has been linked to better pay. There have been mixed results on the association of CEO compensation with financial performance and a strong association with organization size [6,8,9].

Literature on agency theory suggests CEOs are incentivized to behave on behalf of the board of directors [10]. Previous empirical studies in a broader literature of CEOs in all industries have shown CEO compensation is associated with financial performance of the firm [11]. Hospital CEOs often have generous incentive packages encouraging them to meet various financial, managerial, and strategic goals. One news article series published in 2013 identified multiple large nonprofit health systems that linked bonuses to specified targets for profits, revenues, and hospital system growth [12]. In a 2017 survey of hospital CEOs, 69 percent reported that financial performance was a determinant of their incentive pay, while only 18 percent reported that serving the community, including activities to improve community or population health, affected their compensation [13]. Additionally, a 2010 study of nonprofit hospitals in Connecticut found that CEOs were driven to increase the number of privately insured patients as opposed to publicly insured ones [14]. These findings are consistent with the general economics literature, which suggests that CEO pay is closely linked to firm size and the size of the average firm in the economy [15,16].

As nonprofit hospitals consolidate, becoming more profitable, and amassing significant cash reserves, the share of expenses going to generating community benefit has not been increasing [17–19]. Federal laws obligate nonprofit hospitals to provide charity care and community benefits in order to preserve their tax-exempt status [17]. Our objective is to measure the determinants of executive compensation and estimate how the relationship between CEO compensation and operating profit, charity care, teaching status and hospital size has changed from 2012 to 2019.

Methods

Data sets

We obtained data on executive compensation from Candid’s compilation of Internal Revenue Service (IRS) 990 tax forms filed by hospitals in 2012 and 2019 [20]. We focused on schedule H and the base 990 tax filings, which report executive compensation at the employer identification number (EIN) level. Some health systems file a group return using only one EIN, while others submit their tax filings at the hospital level and have multiple EINs.

Hospital financial data and characteristics were obtained from Medicare Cost Reports, which are maintained by the Center for Medicaid and Medicare Services. These cost reports are submitted annually by all hospitals treating Medicare patients. CMS publishes the information from hospitals’ cost reports in its Healthcare Cost Report Information Systems (HCRIS) [21] database. We applied the methodology from the National Academy for State Health Policy’s (NASHP’s) Hospital Cost Tool to the HCRIS data to construct measures of hospital profits, spending on charity care, and cash reserve balances. Hospital profit is defined as net patient revenue minus total hospital operating costs. Charity care is defined as hospital operating costs for providing patient care under the hospital’s charity care policy, less charity care patient payments and restricted grant funds received. We do not include bad debt in this amount. Charity care and profits are aggregated to the system level for health systems.

Following previous research by Joynt et. al. on nonprofit CEO pay, our sample includes CEOs of hospital systems and CEOs of single hospitals not affiliated with other hospitals [5]. We use the American Hospital Association annual survey to identify hospitals belonging to the same system, as well as teaching hospitals [22]. The AHA defines a teaching hospital as a hospital that has an American Medical Association residency program, is a member of the Council of Teaching Hospitals, or has a ratio of full-time equivalent interns and residents to beds of 0.25 or higher.

To merge the compensation data with our other data sources, we develop a CMS certification number (CCN) to EIN crosswalk (See S1.1 Fig in S1 Appendix) [23]. Since executive compensation is reported at the EIN level in the Candid personnel file, multiple hospitals belonging to the same system were in some cases merged with one EIN in the Candid data. In other cases, health systems reported to the IRS using multiple (hospital-level) EINs. There were 190 hospitals (93 in 2012 and 97 in 2019) in our data that were associated with EINs that included hospitals from different systems. These hospitals were dropped from our analysis, so that we could collapse the compensation data to the system level. In all cases we aggregated data (e.g. profits, number of beds, and charity care spending) to the system level. Our final sample consists of both hospital systems and independently operated hospitals that are not part of a larger hospital system. A detailed explanation of how the sample is constructed is available in S1.1 Fig in S1 Appendix [23].

Identifying the CEO and CEO compensation

We identify a CEO for each hospital or health system using titles and compensation reported in the 990 forms. Health systems consisting of many hospitals may have multiple executives with titles such as, “CEO”, “President”, or “Executive Director”. To tackle this issue, we begin with the methods of Song et al and identify executives with titles likely to indicate a CEO [24] and create a flag for executives likely to be the leader of the health system. After the possible leaders have been flagged, we designate the “CEO” of each system or independently operated hospital in our sample as the highest paid executive who is flagged as a leader for each organization. A deeper explanation of how we identified CEOs is available in S1.2 and S1.3 Figs in S1 Appendix. [23].

The measure of compensation used in this analysis is total compensation, which consists of W-2 and or 1099-Misc compensation including base compensation, bonus and incentive compensation, and other reportable compensation. Total compensation also includes retirement and other deferred compensation as well as nontaxable benefits.

Empirical strategy

We first divided CEO compensation into deciles and compared hospital characteristics across these deciles. To understand the association between CEO compensation and hospital profit, charity care costs, teaching status and hospital size, we estimated a linear regression of the log of CEO wages on these independent variables. CEO wage, profit, and charity care costs are defined as total compensation, operating profits, and charity care spending for fiscal years 2012 and 2019. We constructed three categorical measures of hospital or system size. For systems in our sample, these measures reflect the total number of beds, discharges, or hospitals in the system. For our main specification, we created the following categories for number of beds, <100, 100–299, 300–499, and 500 or more beds. As a robustness check we also created categorical variables for number of adjusted discharges (<5,000, 5,000–19,999 20,000–99,999, >100,000), and number of hospitals in the system (single hospital, systems with 2–3 hospitals, systems with 4–9 hospitals, and systems with 10+ hospitals). Adjusted discharges scales up inpatient discharges to include outpatient care by multiplying inpatient discharges by the ratio of total hospital charges to inpatient charges. The coefficients represent the percentage change in CEO wages associated with a one-unit change in the independent variable. We present cash reserve levels in the descriptive statistics presented in Table 1, but are not able to include them in the regression analysis due to multicollinearity with the categorical variables for number of beds. S1.6 Table in S1 Appendix in the appendix presents a cross tabulation of the categorical variables for number of beds and number of hospitals [23].

Table 1. Descriptive statistics by CEO compensation deciles.

Bottom Decile Deciles 2–9 Top Decile
Variable 2012 2019 2012 2019 2012 2019
Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD)
CEO Wage 132,932 (34,118) 148,546 (37,438) 733,774 (481,157) 883,027 (618,073) 3,945,610 (1,926,787) 5,624,209 (4,390,476)
Profit ($ mill) 2.91 (8.37) 1.79 (8.66) 65.59 (129.19) 88.23 (196.97) 417.34 (505.57) 788.27 (726.22)
Charity Care ($ mill) 0.53 (1.39) 0.39 (0.81) 10.67 (26.70) 10.84 (34.80) 65.17 (95.50) 84.35 (91.90)
Cash Reserves ($ mill) 10.63 (22.42) 9.25 (14.25) 206.9 (383.9) 338.5 (622.7) 1528.9 (3053) 2924.9 (3428)
Teaching Status 0 (0) 0 (0) 0.06 (0.22) 0.06 (0.21) 0.36 (0.39) 0.30 (0.30)
Number of Beds 39.3(71.2) 26.9 (25.9) 274.3 (433.4) 309.7 (638.2) 1435.7 (1875.3) 2258 (2319.6)
Number of Hospitals 1.1 (0.4) 1.0 (0.2) 1.8 (2.7) 2.1 (4.1) 6.4 (10.8) 10.4 (13.5)
Adjusted Discharges 2,706 (4942) 1,960 (3035) 27,929 (41,448) 34,890 (68,006) 141,043 (179,758) 247,444 (242,071)
Open in a new tab

The distribution of hospital profits is skewed to the right, because our sample contains both single hospitals and large health systems. While we are controlling for hospital size as a covariate, this skewness could still bias our regression estimates. Therefore, we estimated additional specifications where we transform profit using the cube root transformation that allows for negative values of profit.

We use a Oaxaca decomposition to quantify the portions of changes in CEO pay between 2012 and 2019 that are attributable to differences in hospital characteristics such as hospital size, versus increases in the generosity of compensation in 2019 versus 2012 for any given characteristic [25]. This decomposition has been used to decompose differences in wages between males and females that are attributable to differences in educational attainment or years of work experience by sex, versus differences in the rate at which these characteristics are rewarded for men versus women. The method has been applied to male-female wage differentials across time, industries, and countries [26]. [23] S1.5 Fig in S1 Appendix provides our regression specification, a summary of the Oaxaca method, and the regression coefficients and descriptive statistics used to construct the decomposition.

Results

Our final sample consists of 1,113 independent hospitals and health systems in 2012 and 868 in 2019. The largest health system contained 101 hospitals in 2019. In 2012, CEOs of independent hospitals or health systems in our sample made an average of $996,000 (in 2019 dollars). By 2019, CEOs earned over 30 percent more, an average of $1.3 million. Over the same period, registered nurses’ mean wages grew from $75,652 (in 2019 dollars) to $77,460 [27,28], only a 2.3% increase. Fig 1 graphs the levels of CEO compensation by year. As CEOs of independent and smaller hospital systems in 2012 consolidated into larger health systems by 2019, the distribution of CEO compensation shifted to the right.

Fig 1. Distribution of CEO wages in 2012 and 2019.

Fig 1

Open in a new tab

Fig 1 graphs the distributions of CEO compensation in 2012 and 2019. Compensation is presented in 2019 dollars and truncated at $5+ million.

Table 1 presents descriptive statistics on CEO compensation and hospital characteristics for 2012 and 2019. All descriptive statistics are reported in 2019 dollars. Total compensation grew for all decile groups, with the top decile experiencing a 42.5% increase, from $3.95 million in 2012 to $5.62 million in 2019. The middle eight, and bottom decile groups also had an increase in average CEO compensation of 20.3% and 11.7% respectively.

Table 1 presents descriptive statistics for health systems and independent hospitals in our sample. CEO wage, profit, charity care, and cash reserves are all presented in 2019 dollars. The table is stratified by year and deciles of CEO compensation, with the middle 8 deciles presented together. CEO wage is the compensation of the highest paid executive, likely to be the organization leader. Number of beds, hospitals, and discharges are presented at the system level. Adjusted discharges are defined as the calculated inpatient and outpatient hospital discharges. This is computed by multiplying inpatient volume by an outpatient factor. The outpatient factor is defined as hospital charges divided by inpatient hospital charges. Teaching status is defined as the percentage of adjusted discharges that occurred at a teaching hospital within the hospital or health system.

Mean operating profits increased from $417.3 million to $788.3 million among health systems in the top decile of CEO compensation, while the profits of hospitals and health systems not in the top or bottom decile grew from $65.6 million to $88.2 million. Profits for hospitals in the bottom decile of CEO compensation fell from $2.91 to $1.79 million over the same time period.

Charity care increased among health systems in the top decile of CEO compensation, growing from $65.2 to $84.4 million, while the group containing the middle eight deciles of CEO compensation increased charity care from $10.7 to only $10.8 million. Charity care spending decreased from $0.53 to $0.39 million for the bottom decile.

CEO compensation and hospital characteristics

A one million dollar increase in hospital profits was associated with a 0.06 percent (95% CI: 0.03, 0.09) increase in CEO pay in 2012, and a 0.05 percent (95% CI: 0.03,0.06) increase in CEO pay in 2019. A one million dollar increase in charity care had no significant association with CEO compensation.

Fig 2 presents the results from the regression analysis estimating the association between CEO compensation and profit, charity care spending and hospital size. Hospital size had a large, statistically significant association in both 2012 and 2019 with CEO compensation. In 2012, after controlling for profit, teaching status, and charity care, we estimate hospitals with 100–299 beds paid CEOs 89 percent more (95% CI: 80, 98) than the reference group of hospitals with fewer than 100 beds. Hospitals and health systems with 300–499 paid CEOs 121 (95% CI: 108, 135) percent more than the reference group, and hospitals with 500 or more paid their CEOs 144 (95% CI: 130, 159) percent more. By 2019 CEOs were paid 170 (95% CI: 156, 185) percent more than the reference group for leading a hospital with more than 500 beds, 124 (95% CI: 110, 138) percent more for hospitals with 300–499 beds, and 99 (95% CI: 89, 109) percent more for hospitals with 100–299 beds. There was no increase in relative pay from 2012 to 2019 for the 300–499 beds category.

Fig 2. Association between CEO compensation and profit, charity care, teaching status and Bedsize.

Fig 2

Open in a new tab

Fig 2 presents the results of our regression analyses for 2012 and 2019 where we regress profit, charity care, and categorical measures of teaching status and bed size on log CEO compensation. CEO compensation, profit (in $millions), and charity care (in $millions) are defined using 2019 dollars. Teaching status is defined as the percentage of adjusted discharges that occurred at a teaching hospital within the system or hospital. We multiply the regression coefficients by 100 to approximate the associated percentage change in CEO compensation. The estimates for profit and charity care can be interpreted as the estimated percentage change in CEO compensation associated with a one million dollar increase in profit or charity care. While the estimates for the bed size and teaching categorical variables should be interpreted as the increase in CEO compensation associated with each category of hospital size or teaching status relative to the reference group (hospitals with less than 100 beds, or hospitals/health systems with no adjusted discharges at teaching hospitals). Confidence intervals are presented at 95 percent confidence levels. The base case log CEO compensation (or the intercept term in our regression analysis, defined as estimated compensation for CEOs at hospitals with zero profits or charity care, less than 100 beds, and no adjusted discharges at teaching hospitals) is 12.56 in 2012 and 12.64 in 2019 which corresponds to 286,338 and 307,807 in 2019 dollars. The difference in base case CEO compensation from 2012 to 2019 is not statistically significant. There are 1,113 hospital or health systems in the 2012 sample and 868 in 2019.

Although all of the coefficients (except on charity care) are significantly different from zero at the 95 percent confidence level, the individual coefficient estimates for 2019 versus 2012 are not significantly different from each other. However, Exhibit S1.7 in S1 Appendix presents the regression results of a sensitivity analysis including the interactions between a 2019 year dummy and the explanatory variables in our main regression. A Wald test shows the 2019 coefficients are jointly significantly different from the 2012 coefficients (F (8, 1965) = 3.63; P = 0.0012) [23]. Additionally, the magnitude of the returns to managing a hospital or system with more beds was substantially larger in 2019 versus 2012. As a robustness check, we used discharges as the variable for hospital size, and the effects are similar. The results of the specification using a cube root transformation on profit are again similar and can be found in [23] Fig S1.4 in S1 Appendix.

The Oaxaca decomposition (See S.1.5 in S1 Appendix) of the change in log wages from 2012 to 2019 is presented in waterfall form in Fig 3 [23]. The results are similar when using the untransformed data, but we discuss the results in log form, which accounts for the skewness in CEO pay. The total difference in log wages is 0.162 from 2012 to 2019. The left-hand side of Fig 3 depicts the change in log wages resulting from differences in the magnitude of the coefficients for each characteristic in 2019 versus 2012 (i.e., the difference in compensation reward/penalty associated with a unit change in each hospital characteristic). By reporting the coefficient differentials sequentially, the waterfall format enables one to gauge the individual importance of each hospital/system characteristic in changing CEO compensation between 2012 and 2019. If the solid line is below (above) the bar attached to a hospital characteristic, then the difference in regression coefficients between 2012 and 2019 represents an increase (decline) in CEO compensation associated with that characteristic in 2019 versus 2012. Correspondingly, the right-hand side of the waterfall presents the sequential change in log wages that is accounted for by the differences in the mean of each hospital characteristic across time periods (e.g. how much the proportion of health systems with over 500 beds increased over time). The leftmost bar represents the difference in the base case across years, which is the difference in log wage for a hospital with fewer than 100 beds, no adjusted discharges at teaching hospitals, and zero profits and charity care in 2019 versus 2012.

Fig 3. Oaxaca decomposition.

Fig 3

Open in a new tab

Fig 3 presents the Oaxaca decomposition of the change in log wages from 2012 to 2019. The left-hand side of the chart presents the difference in log wages that is accounted for by the differences in coefficients from 2012 to 2019 and the right-hand side of presents the difference in log wages that is accounted for by the differences in the means of covariates. The leftmost bar represents the difference in regression intercept terms across years (the base case), which is the difference in log wage when the values of all explanatory variables equal 0. The right-most bar is the total difference in log wages, comprising the difference due to coefficient changes (including the base case), and the total of the difference due to covariate changes. Red (blue) columns indicate a negative (positive) contribution to the difference in log wages. Teaching status is defined as the percentage of adjusted discharges that occurred at a teaching hospital within the system or hospital.

On the left-hand side of Fig 3, the base case bar is by far the largest in magnitude, amounting to 0.072 relative to the overall log wage differential of 0.162. Compensation associated with being CEO of a larger hospital or health system was the next most important factor that was more generously rewarded in 2019 versus 2012. The magnitude of the bar for heading a hospital or system with 100 to 299 beds was 0.027, and the bar for heading a system with over 500 beds was even larger (0.048). In contrast, payments for greater profits, teaching status, or charity care did not differ markedly between 2012 and 2019.

On the right side of Fig 3, the larger share of CEOs managing systems with over 500 beds in 2019 versus 2012 was associated with 0.069 greater log wages. Because hospitals consolidated between 2012 and 2019, much of this differential was offset by a 0.048 reduction in log wages associated with fewer CEOs heading systems with 100 to 299 beds. Mean profits adjusted for inflation also rose between 2012 and 2019, amounting to 0.026 greater log wages. [23] S1.5 Fig in S1 Appendix presents the coefficients and means used to construct these estimates.

Overall, 28.5% of the difference in log wages is explained by coefficient differences from 2019 and 2012 and 27% of the difference is explained by changes in hospital characteristics across time. The remaining 44.5% reflects the difference in log wages for the base case hospital in 2019 versus 2012.

Discussion

Policy implications

In an analysis of 1,113 nonprofit health systems and independent hospitals in 2012 and 868 in 2019, we found that mean CEO compensation rose 30% between 2012 and 2019, but mean salary for CEOs in the top decile grew even more, by 42%. Similar to other studies that apply the Oaxaca decomposition to determine the relative importance of differences in characteristics versus differences in payment generosity for those characteristics, we perform our analysis using log wages to control for the observed skewness in earnings. While changes in log wages may be difficult to conceptualize, the method illustrates that almost half (44.5%) of the increase in log CEO compensation between 2019 and 2012 was attributable to an increase in pay for the base case hospital. This result suggests that a hospital CEO that did not improve her performance over time still enjoyed a significant (inflation-adjusted) boost in pay. Part of this pay increase might be explained by the increased complexity of operating a hospital or system, including a wider range of affiliated outpatient facilities such as ambulatory surgery centers, and greater vertical integration with physicians, which we did not account for in our analysis.

Health systems or hospitals that were larger in size or earned higher profits than the base case hospital accounted for most of the additional 27% increase in CEO pay observed over the time period. Of the remaining 28.5% increase in log CEO compensation, most was attributable to an increase in pay generosity for CEOs managing systems with over 500 beds.

The substantial increase in compensation that occurred for all nonprofit hospital CEOs regardless of performance of their institution is consistent with reports from other sources of a widening gap between healthcare executive pay versus nurses’ compensation [1]. The increase in CEO compensation associated with leading larger healthcare systems and earning greater profits may explain the massive increase in healthcare system consolidation which has occurred over the last several years.

A previous study of nonprofit hospital CEO compensation by Joynt et. al. [5] and another by Mulligan et. al. [6] did not report an association between either profits or system size and compensation, although both studies found that greater patient satisfaction was linked to higher CEO pay. Both studies specified profits in terms of margins (the ratio of net income to revenues), while we specified profits in millions of dollars. We specified profits in dollar terms, because this criterion has been used by both for-profit and nonprofit hospital boards to determine incentive pay [12,29]. Both studies included number of beds as an explanatory variable in their regressions, although neither study reported coefficient estimates for this hospital characteristic. Mulligan et. al. chose to exclude CEOs of hospital systems from their sample and instead included CEOs of individual hospitals within systems. The range of profits and facility size in their sample would be more limited relative to our analysis or that in Joynt et. al, which may partially explain a lack of association between CEO pay and hospital profits or size.

Our analysis advances the literature by quantifying the significant role that higher profits and increasing system size play in explaining rising CEO pay. Unlike the two studies just mentioned, we did not measure the association between hospital quality and CEO pay. However, both past studies found no association between CEO pay and a range of process of care and patient outcome measures. Both past studies found a link between greater patient satisfaction and higher CEO pay. However, the mean patient recommendation of hospitals in Mulligan et. al’s analysis rose from 68.8 in 2009 to only 71.2 (out of 100) in 2014, which, when applied to the study’s coefficient estimates, implies an almost negligible increase in CEO pay [6].

Recently, there have been news articles highlighting state and local efforts to either cap CEO compensation or remove tax exempt status for hospitals. In Pennsylvania, public school districts pushed back against the nonprofit health system in the area “Tower Health,” and four hospitals had their tax-exempt status revoked. The executive salaries at these four locations were deemed to be excessively large, and none of these four hospitals were spending even one percent of their expenses on charity care [7]. This case is not the only example of state and local government pushing back against nonprofit hospital leaders prioritizing executive compensation over community benefit. Recently, in California, Oregon, and Arizona the Service Employees International Union (SEIU) have proposed capping hospital executive salaries considering reports showing how large these salaries have grown [30]. Affordable healthcare continues to be a problem for many Americans. As many as 41% of adults currently have debt caused by medical or dental bills [31]. Misaligned incentives for executives at nonprofit hospitals contributes to the rising burden of healthcare costs to American families and suggests their financial priorities are not aligned with the obligations to the communities they serve. If there is little distinction between nonprofit and for-profit hospital behavior [17,18], additional policies to ensure nonprofit hospitals deliver community benefits should be considered, such as reporting of performance criteria that determine CEO bonuses and required reporting of hospital CEO to employee pay ratios.

Limitations

CEOs are incentivized by the board of directors to meet various organizational goals. Thus, decisions on CEO compensation are likely endogenous, and our estimates of the association between CEO compensation and hospital performance cannot be interpreted as causal effects. However, our decomposition provides a useful descriptive analysis of what financial and organizational characteristics are associated with rising CEO pay.

Some hospitals did not include a CEO in their 990 forms. Many 990 forms used other titles for the CEO (i.e. President/COO, Executive Officer, Chairman/President), while some others had missing titles. In addition, some executives are listed as a CEO at one hospital and hold a different position at a related hospital. These issues are discussed in greater detail in [23] S1.2 Fig in S1 Appendix. In some cases, these challenges could lead us to selecting a higher paid executive within the health system who is not the current CEO. Additionally, individual characteristics such as age, sex, and tenure may be important factors in determining CEO pay, but we do not observe these in the data.

Another limitation of our analysis is that data for the IRS 990 forms can be reported for either the calendar year or the fiscal year. Hospital cost reports were analyzed for the fiscal years 2012 and 2019. For this reason, we do not conduct a sensitivity analysis using lagged values of profits and charity care, because it is likely that many of our observations already represent lagged values in our sample.

We may have underestimated profits for some healthcare systems, which are reported at the hospital level in the HCT dataset. When we aggregated hospitals from the NASHP HCT to the health system level, some hospitals were dropped from the analysis, because we could not find a matching facility in the Candid dataset. However, 91.9% of the nonprofit hospitals in the HCT data were matched to the final data set in 2012 and in 2019 94.3% were matched, so this shortfall in profits is likely small. Additionally, profits can be volatile, and because we only estimate cross-section regressions of observables associated with CEO compensation in 2012 and 2019, our coefficient estimates on profit may be biased toward zero.

Conclusion

Executive compensation grew substantially for all CEOs of nonprofit hospitals and healthcare systems from 2012 to 2019. Compensation increases for CEOs of healthcare systems with over 500 beds were even larger, and the proportion of CEOs leading such large systems grew over time. Higher profits were the next most important determinant of higher CEO pay, while changes in charity care and their association with CEO compensation were negligible. Rising executive compensation is contributing to the affordability crisis in American healthcare and should remain in the forefront of the minds of policy makers.

Supporting information

S1 Appendix. Supplement to the determinants of nonprofit hospital CEO compensation.

(DOCX)

pone.0306571.s001.docx (70.3KB, docx)

Acknowledgments

The authors thank Marilyn Bartlett for helping them understand the National Academy for State Health Policy Hospital Cost Tool; and Mireille Jacobson for her insightful comments.

Data Availability

The compensation data used in this study were originally collected by Candid by compiling information stored in IRS 990 tax forms that individually, are publicly available at (https://projects.propublica.org/nonprofits). Candid owns these data, and we cannot share these data publicly per our data use agreement. Parties interested in purchasing these data should contact Candid at (https://Candid.org/use-our-data). Raw hospital financial data is available publicly for download at (https://tool.nashp.org). Our dataset is supplemented with the American Hospital Association survey data for system identification and teaching status. These data may not be shared publicly, but interested parties may inquire about purchases at (https://www.ahadata.com/aha-annual-survey-database). The authors had no special access privileges to the data in this study. Others are able to purchase and access the data in the same manner as the authors via the links provided in the data availability statement.

Funding Statement

The authors declare funding support from Arnold Ventures grants No. 22-06846. The funders had no role in study design, data collection and analysis, publication decisions, or manuscript preparation.

References

  • 1.Du Jerry Y., Rascoe Alexander S., Marcus Randall E. The Growing Executive-Physician Wage Gap in Major US Nonprofit Hospitals and Burden of Nonclinical Workers on the US Healthcare System. Clin Orthop. 2018;476:1910–9. doi: 10.1097/CORR.0000000000000394 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Economic Research Institute. 2021 CEO Pay Trends in the Nonprofit World. 2021. [Google Scholar]
  • 3.North Carolina State Health Plan for Teachers and State Employees, Rice University’s Baker Institute for Public Policy, Johns Hopkins University of Bloomberg School of Public Health, State Treasurer of North Carolina. Hospital Executive Compensation: A Decade of Growing Wage Inequity Across Nonprofit Hospitals. 2022 Oct.
  • 4.Newhouse JP. Toward a Theory of Nonprofit Institutions: An Economic Model of a Hospital. Am Econ Rev. 1970;60(1):64–74. [Google Scholar]
  • 5.Joynt KE, Le ST, Orav EJ, Jha AK. Compensation of Chief Executive Officers at Nonprofit US Hospitals. 2014;174(1). doi: 10.1001/jamainternmed.2013.11537 [DOI] [PubMed] [Google Scholar]
  • 6.Mulligan K, Choksy S, Ishitani C, Romley JA. What do nonprofit hospitals reward? An examination of CEO compensation in nonprofit hospitals. PLOS ONE. 2022. Mar 21;17(3):e0264712. doi: 10.1371/journal.pone.0264712 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Garber J. “Eye-popping” executive salaries led these hospitals to lose their property tax benefits [Internet]. Lown Institute. 2023. [cited 2023 Sep 19]. Available from: https://lowninstitute.org/eye-popping-executive-salaries-led-these-hospitals-to-lose-their-property-tax-benefits/. [Google Scholar]
  • 8.Brickley JA, Van Horn RL. Managerial Incentives in Nonprofit Organizations: Evidence from Hospitals. J Law Econ. 2002. Apr;45(1):227–49. [Google Scholar]
  • 9.Shay PD, White KR. Executive compensation in health care: A systematic review. Health Care Manage Rev. 2014. Jul;39(3):255–67. doi: 10.1097/HMR.0b013e3182915e44 [DOI] [PubMed] [Google Scholar]
  • 10.Mirrlees JA. The Optimal Structure of Incentives and Authority within an Organization. Bell J Econ. 1976;7(1):105–31. [Google Scholar]
  • 11.Murphy KJ. Chapter 4—Executive Compensation: Where We Are, and How We Got There. In: Handbook of the Economics of Finance [Internet]. Elsevier; 2013. [cited 2024 May 28]. p. 211–356. Available from: https://www.sciencedirect.com/science/article/pii/B9780444535948000045. [Google Scholar]
  • 12.Hancock J. Hospital CEO Bonuses Reward Volume and Growth. Kaiser Health News [Internet]. 2013. Jun 16 [cited 2023 Sep 19]; Available from: https://kffhealthnews.org/news/hospital-ceo-compensation-mainbar/. [Google Scholar]
  • 13.Athey LA. Chief Executive Employment Contracts and Performance Evaluations: Current Practices. Foundation of the America College of Healthcare Executives; 2017. [Google Scholar]
  • 14.Kramer J, Santerre RE. Not-for-Profit Hospital CEO Performance and Pay: Some Evidence from Connecticut. Inquiry. 2010;47(3):242–51. doi: 10.5034/inquiryjrnl_47.03.242 [DOI] [PubMed] [Google Scholar]
  • 15.Gabaix X, Landier A. Why has CEO Pay Increased So Much? Q J Econ. 2008. Feb 1;123(1):49–100. [Google Scholar]
  • 16.Zhang H. Are CEOs rewarded for managerial ability? Evidence from nonprofit hospitals. J Public Budg Account Financ Manag. 2023. Mar 23;35(2):219–43. [Google Scholar]
  • 17.Bai G, Demirkan S, Zare H, Anderson GF. Trustee Compensation And Charity Care Provision In US Nonprofit Hospitals: Study examines the relationships between trustee compensation and charity care at US nonprofit hospitals. Health Aff (Millwood). 2023. Apr 1;42(4):526–30. [DOI] [PubMed] [Google Scholar]
  • 18.Jenkins D, Ho V. Nonprofit Hospitals: Profits And Cash Reserves Grow, Charity Care Does Not: Study compares changes in hospital profits with changes in hospitals’ charity care and cash reserves. Health Aff (Millwood). 2023. Jun 1;42(6):866–9. [DOI] [PubMed] [Google Scholar]
  • 19.Young GJ, Flaherty S, Zepeda ED, Singh SR, Rosen Cramer G. Community Benefit Spending By Tax-Exempt Hospitals Changed Little After ACA. Health Aff (Millwood). 2018. Jan;37(1):121–4. doi: 10.1377/hlthaff.2017.1028 [DOI] [PubMed] [Google Scholar]
  • 20.Candid. Candid Custom Data Set [Internet]. Candid. [cited 2024 Feb 23]. Available from: https://candid.org/. [Google Scholar]
  • 21.Cost Reports | CMS [Internet]. [cited 2023 Dec 19]. Available from: https://www.cms.gov/Research-Statistics-Data-and-Systems/Downloadable-Public-Use-Files/Cost-Reports.
  • 22.American Hospital Association. AHA Annual Survey Database [Internet]. [cited 2024 Feb 23]. Available from: https://www.ahadata.com/aha-data-resources. [Google Scholar]
  • 23.Supporting information. To access the Supporting information file, click on the Supporting Information link in the box to the left of the article online. [Google Scholar]
  • 24.Song PH, Lee SYD, Toth M, Singh SR, Young GJ. Gender Differences in Hospital CEO Compensation: A National Investigation of Not-for-Profit Hospitals. Med Care Res Rev. 2019. Dec 1;76(6):830–46. doi: 10.1177/1077558718754573 [DOI] [PubMed] [Google Scholar]
  • 25.Oaxaca R. Male-Female Wage Differentials in Urban Labor Markets. Int Econ Rev. 1973;14(3):693–709. [Google Scholar]
  • 26.Blau FD, Kahn LM. The Gender Wage Gap: Extent, Trends, and Explanations. J Econ Lit. 2017. Sep;55(3):789–865. [Google Scholar]
  • 27.U.S. Bureau of Labor Statistics. Occupational Employment and Wage Statistics [Internet]. May 2019. Registered Nurses. 2020 [cited 2023 Dec 19]. Available from: https://www.bls.gov/oes/2019/may/oes291141.htm. [Google Scholar]
  • 28.U.S. Bureau of Labor Statistics. Occupational Employment and Wages—May 2012 [Internet]. 2013. [cited 2023 Dec 19]. Available from: https://www.bls.gov/news.release/archives/ocwage_03292013.pdf. [Google Scholar]
  • 29.HCA Healthcare. Notice of Annual Meeting of Stockholders to be held April 21, 2022 [Internet]. 2022. [cited 2023 Dec 22]. Available from: https://d18rn0p25nwr6d.cloudfront.net/CIK-0000860730/44bb77ec-3b84-402f-bbcd-4783968b69e8.pdf. [Google Scholar]
  • 30.Work MC. As a Union Pushes to Cap Hospital CEO Pay, It’s Accused of Playing Politics [Internet]. KFF Health News. 2023. [cited 2023 Sep 19]. Available from: https://kffhealthnews.org/news/article/union-hospital-ceo-pay-cap-seiu/?utm_campaign=KHN%3A. [Google Scholar]
  • 31.Lopes L, Montero A, 2022. Health Care Debt In The U.S.: The Broad Consequences Of Medical And Dental Bills—Main Findings [Internet]. KFF. 2022 [cited 2022 Nov 20]. Available from: https://www.kff.org/report-section/kff-health-care-debt-survey-main-findings/. [Google Scholar]

Decision Letter 0

Gabriel A Picone

8 Apr 2024

PONE-D-24-07430The determinants of nonprofit hospital CEO compensationPLOS ONE

Dear Dr. Jenkins,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. In particular, we would like you to provide a more precise description of the empirical analysis, including more precise definitions and an equation for the specification estimated. Also, endogeneity is a significant issue. The authors carefully use the term "association" but do not explore how these associations arise. 

Please submit your revised manuscript by May 23 2024 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Gabriel A. Picone

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at 

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and 

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please note that funding information should not appear in any section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form. Please remove any funding-related text from the manuscript.

3. Thank you for stating the following in the Competing Interests section: 

   "I have read the journal's policy and the authors of this manuscript have the following competing interests: the authors receive funding from Arnold Ventures. VH and MNS receive funding from the Heath Care Service Corporation and National Academy of State Health Policy.  The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

VH is a member of the Blue Cross Blue Shield of Texas Community Advisory Board; Community Health Choice: Houston Business Coalition on Healthcare Advisory Board; Texas Employers for Affordable Health Care Advisory Board and the Blue Cross Blue Shield Research Alliance. VH receives support for attending membership committee meetings from the National Academy of Medicine. VH owned stock in IBM. 

"

We note that you received funding from a commercial source:  Arnold Ventures, Heath Care Service Corporation and National Academy of State Health Policy

Please provide an amended Competing Interests Statement that explicitly states this commercial funder, along with any other relevant declarations relating to employment, consultancy, patents, products in development, marketed products, etc. 

Within this Competing Interests Statement, please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared. 

Please include your amended Competing Interests Statement within your cover letter. We will change the online submission form on your behalf.

4. We note that you have indicated that there are restrictions to data sharing for this study. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. 

Before we proceed with your manuscript, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., a Research Ethics Committee or Institutional Review Board, etc.). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of recommended repositories, please see

https://journals.plos.org/plosone/s/recommended-repositories. You also have the option of uploading the data as Supporting Information files, but we would recommend depositing data directly to a data repository if possible.

We will update your Data Availability statement on your behalf to reflect the information you provide.

5. When completing the data availability statement of the submission form, you indicated that you will make your data available on acceptance. We strongly recommend all authors decide on a data sharing plan before acceptance, as the process can be lengthy and hold up publication timelines. Please note that, though access restrictions are acceptable now, your entire data will need to be made freely accessible if your manuscript is accepted for publication. This policy applies to all data except where public deposition would breach compliance with the protocol approved by your research ethics board. If you are unable to adhere to our open data policy, please kindly revise your statement to explain your reasoning and we will seek the editor's input on an exemption. Please be assured that, once you have provided your new statement, the assessment of your exemption will not hold up the peer review process.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: General Comments

1. The topic of the paper is an interesting and an important one. The paper potentially contributes to the literature on determinants of CEO compensation (although the existing economic literature on the topic is not reviewed. There is one citation (13) from this literature.

2. On page 3, lines 69-70, the authors indicate that hospital CEOs are offered generous incentive pages to meet financial, managerial, and strategic goals. If so, would not such covariates as profit, cash reserves and hospital and and system characteristics be endogenous? A survey of hospital CEOs is cited on lines 72 ff, the results indicating that achieving financial goals is a major determinant of incentive pay. Then profit and CEO relationship is certainly bidirectional. Although decompensation studies do not typically deal with endogeneity, the case for doing so seems strong here.

3. Teaching status is never defined. There are alternative definitions, including presence of an accredited residency program, the number of residents per bed, membership in the American Association of Medical Colleges.

4. There are other cases for which details about empirical specification is lacking. I assume the authors picked the highest paid CEO when there are several CEOs in the hospital system, but this is not stated. Is the observational unit the system when the hospital is part of a system and a hospital otherwise? Are there systems in which the total number of beds is say 250? Or do most systems have 500+ beds. Some systems, e.g., HCA have many multiples of 500 beds. Is profit defined for one year, the current year or a lagged year, or the year the CEO was appointed? It would be good to see the main equation specification that was estimated.

5. There is no theoretical framework. Changes in coefficients between the two years could reflect structural changes in markets for hospital executives. Demand for such executives could have shifted outward relative to a stable inelastic supply curve. How would one know whether ‘’executive compensation is excessive.”? (line 396). This could explain the base case change in coefficients. Journalists will be looking for a statement that compensation is excessive, but the paper does not present empirical evidence for this. The price of eggs may rise rapidly because of a supply shock. This does not mean that the price of eggs is “excessive.”

6. Will readers understand the purpose of and how an Oaxaca decomposition is performed? I am among potential interested readers who does not know what a waterfall form is.

7. What happened to the asset covariate?

Specific Comments (the numbers are line numbers)

52. Use mean rather than average.

53. The $1.75 billion figure is hard to understand. Divide by 9?

55. What is a “handful?”

61 ff. More precisely, in the Newhouse model, the trustees have a utility function which depends on quantity and quality. Optimal quantity and quality are determined by maximizing trustee utility subject to a zero- profit constraint.

64. Which patient outcomes have been studied. Hospital quality is hard to measure from a distance. And there are many quality attributes.

78. One study I have read finds that payer mix does not change after hospital mergers.

86. What do you mean by charity care?. There is uncompensated care which is the sum of charity care (care the hospital expects to subsidize at patient admission) and bad debt care (revenue the hospitals expect to collect but does not). Uncompensated care is a more comprehensive definition.

98. How do you deal with the EIN problem?

129. Then hospitals in systems have multiple CEOs. Chose the highest paid CEO? I checked a chain I am familiar with. The head of a hospital that is part of the chain but not the leading hospital has a “President.”

144. Be more specific about how hospital bed size is defined in the case of systems.

156. You know from your data that profits are skewed to the right and that there are negative values. This is more than “likely.”

160. What does “back transform” mean?

165 ff. More explanation needed.

Table 1. You mention cash reserves in the table. But they disappear elsewhere. Cash reserves can be highly endogenous in that a highly paid entrepreneur may use such funds for hospital/system improvement/expansion. Why use cash reserves and not other items from the balance sheet?

223. A 0.06% change is trivial. Also, the 95%Cis overlap for all results in Fig. 2 (see bars and whiskers), implying no statistically significant changes in structure between 2012 and 2019.

247. The definition of teaching status must include a definition of teaching hospital.

286.larger hospital or health system. Again, we need a precise definition of how size is constructed. Does system trump hospital when a hospital belongs to a system?

295. Are there systems with 100-299 beds? If so, what is the N for this category?

394. This statement may reflect equity considerations. But relative wages can change over time for many reasons. There is a literature on monopsony in a hospital context. Is this being implied here?

378. Some of the points I make under General Comments apply here. In particular, it is likely that some hospital executives receive generous pay offers because boards think the execs can help solve the hospital’s financial problems.

388. Profits are volatile. For this reason, the parameter estimate may be biased toward zero (consider it an error in variable).

Fig. 1. How do you handle the open-ended category at the far right in the regression analysis?

Fig. 2. I prefer regression results for the preferred specifications. But perhaps this is a matter of taste. I have a stronger preference for discussing methods in the text. None of the parameter estimates are statistically different between 2012 and 2019. What is the sample size? I do not see it here or elsewhere in the paper. Hospitals dropped?

Fig. 3. Can there be better titles than unexplained coefficient effects? Effects implies causation.

They are really changes in how CEO time is priced. The right side is changes in the mean values of pay determinants. Again, methods should be explained in the text.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2024 Jul 24;19(7):e0306571. doi: 10.1371/journal.pone.0306571.r002

Author response to Decision Letter 0

30 May 2024

All line number references below refer to the clean version of the manuscript.

Reviewer #1: General Comments

1. The topic of the paper is an interesting and an important one. The paper potentially contributes to the literature on determinants of CEO compensation (although the existing economic literature on the topic is not reviewed. There is one citation (13) from this literature.

Thank you for the suggestion that we include more sources from the broader economics literature on CEO compensation. On line 71 we have added the text below, along with citations of (Mirrlees Bell J Econ 1976 [10] and Murphy Handbook of the Economics of Finance 2013 [11]),

“Literature on agency theory suggests CEOs are incentivized to behave on behalf of the board of directors [10]. Previous empirical studies in a broader literature of CEOs in all industries have shown CEO compensation is associated with financial performance of the firm. [11]”

2. On page 3, lines 69-70, the authors indicate that hospital CEOs are offered generous incentive pages to meet financial, managerial, and strategic goals. If so, would not such covariates as profit, cash reserves and hospital and system characteristics be endogenous? A survey of hospital CEOs is cited on lines 72 ff, the results indicating that achieving financial goals is a major determinant of incentive pay. Then profit and CEO relationship is certainly bidirectional. Although decompensation studies do not typically deal with endogeneity, the case for doing so seems strong here.

We agree that CEO compensation is likely endogenous and we do not attempt to estimate causal effects in this paper. Thank you for letting us know that this was not clear to the reader. We have added the following sentences on line 402 at the beginning of the limitations section to acknowledge that we are not estimating a causal relationship.

“CEOs are incentivized by the board of directors to meet various organizational goals. Thus, decisions on CEO compensation are likely endogenous, and our estimates of the association between CEO compensation and hospital performance cannot be interpreted as causal effects. However, our decomposition provides a useful descriptive analysis of what financial and organizational characteristics are associated with rising CEO pay.

3. Teaching status is never defined. There are alternative definitions, including presence of an accredited residency program, the number of residents per bed, membership in the American Association of Medical Colleges.

We apologize for the oversight. The following definition of teaching status has been added on line 117.

“The AHA defines a teaching hospital as a hospital that has an American Medical Association residency program, is a member of the Council of Teaching Hospitals, or has a ratio of full-time equivalent interns and residents to beds of 0.25 or higher.”

4. There are other cases for which details about empirical specification is lacking. I assume the authors picked the highest paid CEO when there are several CEOs in the hospital system, but this is not stated. Is the observational unit the system when the hospital is part of a system and a hospital otherwise? Are there systems in which the total number of beds is say 250? Or do most systems have 500+ beds. Some systems, e.g., HCA have many multiples of 500 beds. Is profit defined for one year, the current year or a lagged year, or the year the CEO was appointed? It would be good to see the main equation specification that was estimated.

• Thank you for letting us know that these details were not clear in the main text. All hospitals within systems are aggregated to the system level. Our final sample includes both hospital systems, and independently operated hospitals. The following sentence has been added to line 128.

“In all cases we aggregated data (e.g. profits, number of beds, and charity care spending) to the system level. Our final sample consists of both hospital systems and independently operated hospitals that are not part of a larger hospital system.”

• We have also updated the explanation of CEO identification in the main text to be more specific. The sentence beginning on line 137 now reads,

“After the possible leaders have been flagged, we designate the “CEO” of each system or independently operated hospital in our sample as the highest paid executive who is flagged as a leader for each organization. A deeper explanation of how we identified CEOs is available in appendix Figs S1.2 and S1.3.[23].”

In the appendix, the details of CEO identification are explained at length. We felt it was cumbersome to include the full discussion of these methods in the main text.

• Thank you for this question. Yes, there are systems where the total numbers of beds is lower than 250. We have added a cross tabulation to the appendix for our categorical variables of bed size and number of hospitals. We include a mention to this table on line 164 with the sentence,

•

“Table S1.6 in the appendix presents a cross tabulation of the categorical variables for number of beds and number of hospitals. [23]”

• Profit is defined as total operating profit in the current fiscal year. We have added a sentence in the text on line 151 that reads,

“CEO wage, profit, and charity care costs are defined as total compensation, operating profits, and charity care spending for fiscal years 2012 and 2019.”

Some other studies use lagged values of financial performance, which we could not do with confidence with our merged dataset. We have added an explanation on line 417 in the Limitations section:

“Another limitation of our analysis is that data for the IRS 990 forms can be reported for either the calendar year or the fiscal year. Hospital cost reports were analyzed for the fiscal years 2012 and 2019. For this reason, we do not conduct a sensitivity analysis using lagged values of profits and charity care, because it is likely that many of our observations already represent lagged values in our sample.”

• In the last sentence before the Results section, we have added a mention that our regression specification is included in the Appendix Fig S.1.5.

5. There is no theoretical framework. Changes in coefficients between the two years could reflect structural changes in markets for hospital executives. Demand for such executives could have shifted outward relative to a stable inelastic supply curve. How would one know whether ‘’executive compensation is excessive.”? (line 396). This could explain the base case change in coefficients. Journalists will be looking for a statement that compensation is excessive, but the paper does not present empirical evidence for this. The price of eggs may rise rapidly because of a supply shock. This does not mean that the price of eggs is “excessive.”

You are correct that we do not present a theoretical framework and cannot claim from our analysis that executive compensation is excessive. We agree the statement you refer to was misleading. The paragraph on page 18 has been revised to remove this sentence. We also revised the sentence beginning on line 393 to remove the phrase “excessive compensation” and now reads,

“Misaligned incentives for executives at nonprofit hospitals contributes to the rising burden of healthcare costs to American families and suggests their financial priorities are not aligned with the obligations to the communities they serve”

6. Will readers understand the purpose of and how an Oaxaca decomposition is performed? I am among potential interested readers who does not know what a waterfall form is.

Thank you for pointing out this shortcoming. We have expanded the explanation of the Oaxaca decomposition on page 8 line 174 to read:

“We use a Oaxaca decomposition to quantify the portions of changes in CEO pay between 2012 and 2019 that are attributable to differences in hospital characteristics such as hospital size, versus increases in the generosity of compensation in 2019 versus 2012 for any given characteristic.[25] This decomposition has been used to decompose differences in wages between males and females that are attributable to differences in educational attainment or years of work experience by sex, versus differences in the rate at which these characteristics are rewarded for men versus women. The method has been applied to male-female wage differentials across time, industries, and countries.[26] Appendix[23] Fig S1.5 provides a summary of the Oaxaca method and the regression coefficients and descriptive statistics used to construct the decomposition.”

The concept of graphing results in waterfall format is more general than a Oaxaca decomposition. We have modified the text on page 13 line 281 to clarify this point. Moreover, we modified Fig 3, so that it is easier to see the bar associated with each hospital characteristic, so that one can readily tell whether each factor contributes to a positive or negative change in CEO pay in 2019 vs 2012:

The Oaxaca decomposition (See Appendix S.1.5) of the change in log wages from 2012 to 2019 is presented in waterfall form in Fig 3… The left-hand side of Fig 3 depicts the change in log wages resulting from differences in the magnitude of the coefficients for each characteristic in 2019 versus 2012 (i.e., the difference in compensation reward/penalty associated with a unit change in each hospital characteristic). By reporting the coefficient differentials sequentially, the waterfall format enables one to gauge the individual importance of each hospital/system characteristic in changing CEO compensation between 2012 and 2019. If the solid line is below (above) the bar attached to a hospital characteristic, then the difference in regression coefficients between 2012 and 2019 represents an increase (decline) in CEO compensation associated with that characteristic in 2019 versus 2012. Correspondingly, the right-hand side of the waterfall presents the sequential change in log wages that is accounted for by the differences in the mean of each hospital characteristic across time periods (e.g. how much the proportion of health systems with over 500 beds increased over time). The leftmost bar represents the difference in the base case across years, which is the difference in log wage for a hospital with fewer than 100 beds, no adjusted discharges at teaching hospitals, and zero profits and charity care in 2019 versus 2012.

7. What happened to the asset covariate?

The cash reserves variable was highly correlated with bed size and could not be included in the regression. We felt it still provided useful context, so we kept it in Table 1 as a descriptive statistic. A sentence has been added to the methods section on line 162 stating,

“We present cash reserve levels in the descriptive statistics presented in Table 1, but are not able to include them in the regression analysis due to multicollinearity with the categorical variables for number of beds.”

Specific Comments (the numbers are line numbers)

52. Use mean rather than average.

Thank you for this correction. On line 50 “average” has been changed to “mean” in the text.

53. The $1.75 billion figure is hard to understand. Divide by 9?

This figure refers to the total amount paid to highly paid executives, which includes chief financial officers, vice presidents, chief legal officers and management positions over clinical care staff like chief medical officers. It should be interpreted as spending on highly paid administrative staff, not involved with clinical care. We have rephrased the sentence on line 54 to read:

“In North Carolina, the nine largest health systems paid CEOs, chief financial officers, vice presidents and other highly paid administrative positions more than $1.75 billion from 2001-2021…”

55. What is a “handful?”

We are referring to the nine hospital CEOs at the health systems mentioned earlier in the sentence. We have revised the text on line 54 to read,

“In North Carolina, the nine largest health systems paid CEOs, chief financial officers, vice presidents and other highly paid administrative positions more than $1.75 billion from 2001-2021, with about 20 percent of this sum paid only to the CEOs of these health systems.”

61 ff. More precisely, in the Newhouse model, the trustees have a utility function which depends on quantity and quality. Optimal quantity and quality are determined by maximizing trustee utility subject to a zero- profit constraint.

Thank you for this correction. We have revised the sentence on line 62 to read, “Joseph Newhouse first posited in 1970 that hospital trustees would reward nonprofit hospital CEOs for maximizing trustee utility instead of profits, by providing the optimal quantity and quality of care.”

64. Which patient outcomes have been studied. Hospital quality is hard to measure from a distance. And there are many quality attributes.

The literature we reference includes risk adjusted mortality and readmissions as patient outcomes. In the main text we have updated the sentence on line 64 to read, “However, previous literature finds no association between better patient outcomes (i.e. risk adjusted mortality and readmissions)”.

78. One study I have read finds that payer mix does not change after hospital mergers.

Thank you for this comment. We have updated the sentence on line 80 to read,

“Additionally, a 2010 study of nonprofit hospitals in Connecticut found that CEOs were driven to increase the number of privately insured patients as opposed to publicly insured ones.[12]

86. What do you mean by charity care?. There is uncompensated care which is the sum of charity care (care the hospital expects to subsidize at patient admission) and bad debt care (revenue the hospitals expect to collect but does not). Uncompensated care is a more comprehensive definition.

We apologize for the oversight. We have revised the text on line 109 to provide a better definition.

“Charity care is defined as hospital operating costs for providing patient care under the hospital's charity care policy, less charity care patient payments and restricted grant funds received. We do not include bad debt in this amount. Charity care and profits are aggregated to the system level for health systems.”

98. How do you deal with the EIN problem?

Thank you for letting us know this wasn’t clear. Our treatment of EINs attached to a set of hospitals belonging to two or more systems is explained in detail in the appendix exhibit S1.1, but we have added the following two sentences on line 125,

“There were 190 hospitals (93 in 2012 and 97 in 2019) in our data that were associated with EINs that included hospitals from different systems. These hospitals were dropped from our analysis, so that we could collapse the compensation data to the system level.”

129. Then hospitals in systems have multiple CEOs. Chose the highest paid CEO? I checked a chain I am familiar with. The head of a hospital that is part of the chain but not the leading hospital has a “President.”

Thank you for letting us know there is not enough detail in the main text for the reader to understand our methodology. Appendix S1.2 explains our methods for identifying the CEO of each system in detail, but we felt outlining the entire process was too cumbersome for the main text. We have revised the text in the manuscript on line 137 that now read,

“After the possible leaders have been flagged, we designate the “CEO” of each system or independently operated hospital in our sample as the highest paid executive who is flagged as a leader for each organization. A deeper explanation of how we identified CEOs is available in appendix Figs S1.2 and S1.3.[23].”

144. Be more specific about how hospital bed size is defined in the case of systems.

We have corrected the text in our methods section to define how bed size is defined for systems. On line 153, we added the text:

“We constructed three categorical measures of hospital or system size. For systems in our sample, these measures reflect the total number of beds, adjusted discharges, or hospitals in the system”.

156. You know from your data that profits are skewed to the right and that there are negative valu

Attachment

Submitted filename: Reviewer_Comments.docx

pone.0306571.s002.docx (37.4KB, docx)

Decision Letter 1

Gabriel A Picone

20 Jun 2024

The determinants of nonprofit hospital CEO compensation

PONE-D-24-07430R1

Dear Dr. Jenkins,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. If you have any questions relating to publication charges, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Gabriel A. Picone

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I have read all the comments I had on the previous draft, and my concerns have been addressed. I also read the revised paper.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Frank A. Sloan, Ph.D.

**********

Acceptance letter

Gabriel A Picone

2 Jul 2024

PONE-D-24-07430R1

PLOS ONE

Dear Dr. Jenkins,

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now being handed over to our production team.

At this stage, our production department will prepare your paper for publication. This includes ensuring the following:

* All references, tables, and figures are properly cited

* All relevant supporting information is included in the manuscript submission,

* There are no issues that prevent the paper from being properly typeset

If revisions are needed, the production department will contact you directly to resolve them. If no revisions are needed, you will receive an email when the publication date has been set. At this time, we do not offer pre-publication proofs to authors during production of the accepted work. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few weeks to review your paper and let you know the next and final steps.

Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

If we can help with anything else, please email us at customercare@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Gabriel A. Picone

Academic Editor

PLOS ONE

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    S1 Appendix. Supplement to the determinants of nonprofit hospital CEO compensation.

    (DOCX)

    pone.0306571.s001.docx (70.3KB, docx)
    Attachment

    Submitted filename: Reviewer_Comments.docx

    pone.0306571.s002.docx (37.4KB, docx)

    Data Availability Statement

    The compensation data used in this study were originally collected by Candid by compiling information stored in IRS 990 tax forms that individually, are publicly available at (https://projects.propublica.org/nonprofits). Candid owns these data, and we cannot share these data publicly per our data use agreement. Parties interested in purchasing these data should contact Candid at (https://Candid.org/use-our-data). Raw hospital financial data is available publicly for download at (https://tool.nashp.org). Our dataset is supplemented with the American Hospital Association survey data for system identification and teaching status. These data may not be shared publicly, but interested parties may inquire about purchases at (https://www.ahadata.com/aha-annual-survey-database). The authors had no special access privileges to the data in this study. Others are able to purchase and access the data in the same manner as the authors via the links provided in the data availability statement.

    Articles from PLOS ONE are provided here courtesy of PLOS

    RESOURCES