Journal of Health Care Communications Open Access

Asymmetric Accounting Performances and Labor Adjustments in the U.S. for-Profit Skilled Nursing Facilities, 1998-2014

Haizhen Lin¹ and Ayung Tseng^{2* 1}Department of Business Economics and Public Policy, Kelley School of Business, Indiana University, National Bureau of Economic Research, USA
²Department of Accounting, McDonough School of Business, Georgetown University, USA
^*Correspondence: Ayung Tseng, Department of Accounting, McDonough School of Business, Georgetown University, USA, Email:

Received: 29-Mar-2023, Manuscript No. IPJHCC-23-15994; Editor assigned: 31-Mar-2023, Pre QC No. IPJHCC-23-15994 (PQ); Reviewed: 14-Apr-2023, QC No. IPJHCC-23-15994; Revised: 19-Apr-2023, Manuscript No. IPJHCC-23-15994 (R); Published: 26-Apr-2023, DOI: 10.36846/2472-1654-8.2.8017

Abbreviations

(SNF) Skilled Nursing Facilities; (CMS) Centers for Medicare & Medicaid Services; (ROA) Return on Assets

Introduction

President Biden, in his State of the Union Address in October 2022, called for federal minimum staffing requirements for Skilled Nursing Facilities (SNF) because evidence suggests that understaffed SNFs have a significantly higher COVID-19 infection rate and SNFs with higher employee turnovers during the Great Recession of 2008 have a significantly lower quality of patient care [1,2]. Reflecting the aging U.S. population, SNFs serve about 1.3 million patients, with total spending approaching $196 billion [3,4]. Unlike hospitals, 70% of which are nonprofit entities, more than 55% of SNFs in the U.S. are for-profit entities, based on the Centers for Medicare & Medicaid Services (CMS) cost reports from 1998-2014 [5].

Existing studies have shown that for-profit entities across various industries rely on accounting performances to raise external capital, and capital providers continue using accounting performances to monitor their investments in these for-profit entities [6-8]. It is well-known in the accounting literature that accountants exercise conservative judgments. In particular, accountants tend to be conservative by reporting unfavorable business fluctuations earlier than favorable ones [9]. Therefore, SNFs’ reported accounting performances are expected to be more negative during adverse periods compared to positive accounting performances during favorable periods. Such asymmetric accounting performance reports could constrain SNFs’ ability to retain employees during downturns like COVID-19 or the Great Recession.

In this study, we propose an innovative approach to capture regional business fluctuations faced by individual SNFs to draw causal inferences on SNFs’ reported asymmetric accounting performances. More importantly, we further examine whether SNF’s labor adjustments present an asymmetry across positive and negative business fluctuations. We last discuss how our findings on SNF’s supply-side frictions add to the existing literature and ongoing policy debate.

Methodology

Proposed Instrument for Business Fluctuations and Data

Existing evidence on whether firms’ reported accounting performances or labor adjustments present asymmetry across business fluctuations is lacking mainly due to data availability and measurement issues. First, some firm-level instruments for business fundamentals are relevant but under the examined firms’ control. Hence, the same firms’ accounting reporting and operational responses simultaneously determine these firm-level instruments. Second, researchers are constrained by limited time-series data when using macroeconomic cycles to capture business fluctuations. We propose an innovative approach to capture regional business fundamentals specifically for skilled nursing facilities and avoid such problems.

In particular, we propose to use changes in the county-level occupancy rates (Δ Occupancy) as instruments for changes in the business fundamentals faced by SNFs operating in those counties. The county-level occupancy rate changes are remote from individual SNFs’ operational and accounting choices while still capturing key fluctuations in their relevant business conditions. The 3,000+ counties in the U.S. provide sufficient variations to infer asymmetric patterns. Our data source is state-audited annual cost reports that SNFs submitted to the CMS from 1998 to 2014. Our sample comprises 11,350 for-profit skilled nursing facilities, representing 108,345 facility years. Our sample starts in 1998 because the Balanced Budget Act of 1997 significantly altered industry dynamics (The Balanced Budget Act of 1997 changed how Medicare pays skilled nursing facilities. Prior to the Act, Medicare reimbursed incurred costs. Afterwards, Medicare pays a fixed rate per inpatient day). We include only for-profit entities in our sample to hold constant ownership types because studies have shown managerial incentives vary across SNFs’ ownership types [10].

Hypothesis Development

Our first hypothesis relies on the accounting theory, such that accountants tend to be conservative by reporting unfavorable business fluctuations earlier than favorable ones [9]. For example,accountants report loss events immediately but wait to report gain events until uncertainties are resolved [11]. Such asymmetric timeliness between gain and loss events leads to a greater correlation between accounting performances and adverse business fluctuations than the correlation between accounting performances and positive business fluctuations.

Our analysis starts with a commonly-used accounting profitability measure, return on assets (ROA), calculated as the annual net profit divided by the beginning-of-year total assets. ROA captures how much profits an SNF generates at the end of a year relative to the total asset invested (by external capital providers) at the beginning of the year. We expect that ROA becomes more negative during business downturns compared to increases in ROAs during upturns due to conservative accounting judgments.

To further verify that the underlying mechanism is due to accountants’ judgment rather than other asymmetric arrivals of arrivals of business events, we decompose an SNF’s ROA into two components: Accrual earnings that are produced by accountants and operating cash flows that do not involve accountants. If the conservative accounting theory is correct, we expect accrual earnings to respond more negatively to adverse business fluctuations than increases in accrual earnings during positive counterparts, but operating cash flows present symmetric responses across business fluctuations [12].

Our second hypothesis examines firms’ employee adjustments. Managerial accounting theories are split on how firms adjust their labor forces in response to business fluctuations. On the one hand, the cost curtailment argument postulates greater responses to adverse business fluctuations than positive counterparts due to the incurrence of employee severance payments related to layoffs [13]. On the other hand, the cost stickiness hypothesis argues that managers are more reluctant to reduce operating resources during economic downturns to avoid incurring current and future adjustment costs (e.g. to retrain skilled employees if business condition subsequently strengthens) than to increase resources during economic upturns, leading to a greater response to positive business fluctuations than to negative counterparts [14].

Skilled nursing facilities are labor intensive, employing higher- skilled registered and licensed practical nurses, lesser-skilled nurse assistants, and other personnel. We conjecture that cost curtailment likely applies to lesser-skilled employees, while cost stickiness applies to higher skilled employees. This is because higher-skilled nurse positions have higher adjustment costs, such as recruiting and training. We use the facility’s proportion of total hours worked that is attributable to higher-skilled nurses as the proxy for labor adjustment costs.

Empirical Model

To test our first hypothesis, such that accrual earnings respond more to adverse business fluctuations than positive counterparts, but operating cash flows present symmetric responses, we augment the piecewise linear regression model of Basu (1997) with our proposed instrument Δ Occupancy as follows:

The subscripts i and j denote skilled nursing facilities and counties, respectively. ROA is net income from services to patients deflated by beginning-of-year total assets. Cash flow is ROA minus Accruals (i.e., ROA=Accruals+Cash flow). Accruals are calculated based on the formula: Accruals=changes in current assets–changes in current liabilities–changes in the cash ending balance–depreciation and amortization expenses [12]. All accounting information comes from cost report worksheet G.

The explanatory variables include three sets of variables:

• A dichotomous variable for negative annual occupancy rate changes (D Δ Occupancy),

• A continuous variable of annual occupancy rate changes (Δ Occupancy), and

• The product of the corresponding values of (1) and (2).

Δ Occupancy is the annual growth rate in the room occupancy rate for the county where the facility is located. The occupancy rate is calculated as the sum of occupied rooms divided by the available rooms (cost report worksheets S-2 and S-3). Although we examine accounting reporting and labor adjustments for only for-profit SNFs, we include all SNFs across different ownership types in the county to construct the occupancy rate because such occupancy rates are closest to the business fluctuations monitored by SNFs.

We conduct Fama-MacBeth (1973) cross-sectional regressions annually and report the average value across annual coefficients or R-squared estimates. The Fama-MacBeth procedure allows us to remove macroeconomic fluctuations and focus on county-level business fluctuations [15]. The 11,000+ for profit skilled nursing facilities across 3,000+ counties provide sufficient variations to examine the asymmetric accounting reporting and labor adjustment responses during the 17 years from 1998 to 2014. Accounting conservative judgment is present in accrual earnings but not in operating cash flows. Hence, we expect γ³ to be significantly positive when ROA or Accruals is the dependent variable, but we expect γ³ to be insignificantly different from zero when cash flow is the dependent variable.

To test our second hypothesis of whether skilled nursing facilities present asymmetric labor adjustments, we replace the dependent variable in equation (1) with the percentage growth in the number of employees (Δ Employee). Next, we examine labor adjustment costs in the following regression model:

The subscripts i and j denote skilled nursing facilities and counties, respectively. Δ Employee is the annual growth rate in the number of employees (cost report worksheet S-3). Skill is the facility’s proportion of total hours worked that is attributable to higher-skilled nurses (cost report worksheet S-3 Part V). Higher- skilled nurses include registered nurses (Bachelor’s degree) and licensed practical nurses (a one or two year’s program). According to the Bureau of Labor Statistics, the median hourly wage is $ 31.6 for registered nurses, $ 24.1 for licensed practical nurses, and $ 13.4 for certified nursing assistants, and $ 14.2 for non-nurse employees, consistent with differential skill levels.

Results

Table 1 reports the number of observations, mean, quartiles, and standard deviation for each model variable. ΔOccupancy is the annual growth rate in the room occupancy rate for the county in which the facility is located (CMS cost report), where the occupancy rate is the sum of occupied rooms divided by the sum of available rooms, based on all facilities across different ownership types in the county (Cost report worksheets S-2 and S-3). ROA is net income from services to patients deflated by beginning-of-year total assets (cost report worksheet G). Cash flow is ROA minus Accruals (i.e., ROA=Accruals+Cash flow). Accruals= changes in current assets–changes in current liabilities– changes in the cash ending balance–depreciation and amortization expenses (cost report worksheet G). ΔEmployee is the annual growth rate in the number of employees (Cost report worksheet S-3). Skill is the facility’s proportion of total hours worked that are attributable to higher-skilled nurses (Cost report worksheet S-3 Part V). Higher-skilled nurses include registered nurses (Bachelor’s degree) and licensed practical nurses (a one- or two-year program). The sample includes 108,345 for-profit skilled nursing facilities-years from 1998–2014 for 11,350 distinct facilities. Skill is available only for 19,623 observations beginning in 2010, following the imposition of additional disclosure requirements in the CMS cost report. We require all variables to be not missing and all deflators to take positive values. All variables are winsorized at their 1% and 99% tails to avoid input errors.

Table 1: Variable Distributions.

The mean of ΔOccupancy is 0%, with a standard deviation of 5%. The mean of ΔEmployee is 23%, with a standard deviation of 114%. The mean of Skill is 22%, with a standard deviation of 5%. These standard deviations indicate sufficient variation, suggesting our tests should be reasonably powerful.

Table 2 reports the Pearson (upper right triangle) and Spearman (lower left triangle) correlations of the variables for the skilled nursing facility sample. *, **, and *** denote significance at the 10%, 5%, and 1%, respectively, levels in two-tailed tests.

Table 2: Correlation Matrix.

As expected, ΔOccupancy positively correlates with ROA, Accruals (Spearman only), cash flow, and ΔEmployee. Skill negatively correlates with ΔOccupancy and ΔEmployee, consistent with expanding (contracting) markets disproportionately hiring (firing) less-skilled employees.

Table 3 reports the time-series means and corresponding significance levels based on Fama-MacBeth (1973) standard errors of the estimated coefficients in annual OLS cross-sectional estimations of the equation (1) regression model. *, **, and *** denote significance at the 10%, 5%, and 1%, respectively, levels in two-tailed tests.

Table 3: Conservative Accounting Judgment in Skilled Nursing Facilities.

The first column of Table 3 reports the OLS estimation of equation (1) with the dependent variable ROA. The coefficient γ² on ΔOccupancy is insignificant, but the coefficient γ³ on DΔOccupancy × ΔOccupancy is significantly positive, consistent with scaled earnings reflecting conservative accounting judgment.

The second and third columns of Table 3 report the OLS estimation of equation (1) with the dependent variables Accruals and cash flow. For the Accruals model in the second column, the coefficient γ² on ΔOccupancy is significantly negative, perhaps because higher occupancy increases the usage of inventories or the accrual of payables for labor and other inputs. The coefficient γ³ on DΔOccupancy × ΔOccupancy is significantly positive, consistent with accrual earnings reflecting conservative accounting judgment. In contrast, for the cash flow model in the third column, the coefficient γ² on ΔOccupancy is significantly positive, and the coefficient γ³ on DΔOccupancy × ΔOccupancy is insignificant, suggesting symmetric responses across business fluctuations. Results in Table 3 are consistent with conservative accounting judgment being present in only accrual earnings but not in non-accountant-involved operating cash flows.

Table 4 reports the OLS estimation of equation (2) with the dependent variable ΔEmployee and the interactions of Skill with ΔOccupancy excluded in column (1) and included in column (2). The sample in column (1) includes all 108,345 observations with non-missing ΔEmployee, while the sample in column (2) includes 19,623 observations with non-missing ΔEmployee and Skill. *, **, and *** denote significance at the 10%, 5%, and 1%, respectively, levels in two-tailed tests.

Table 4: Changes in Employees and Labor Adjustment Costs in Skilled Nursing Facilities.

In the model that excludes the interactions of Skill with ΔOccupancy reported in column (1), the coefficient on DΔOccupancy × ΔOccupancy, is significantly negative, suggesting that the rate of layoffs during downturns is lower than the rate of recruiting during upturns. This evidence is consistent with the cost stickiness hypothesis that skilled nursing facilities are, on average, reluctant to curtail their employees during economic downturns.

In the model that includes the interactions of Skill with ΔOccupancy reported in column (2), the coefficient on DΔOccupancy × ΔOccupancy becomes slightly positive, and the coefficient on Skill × DΔOccupancy × ΔOccupancy is significantly negative. This additional finding reconciles the cost stickiness and the cost curtailment arguments. Since higher-skilled nurses have higher future adjustment costs (e.g., recruiting and training), skilled nursing facilities tend to retain higher-skilled nurses and lay off lesser-skilled employees during adverse business fluctuations.

Discussion and Conclusion

Overall, we find that SNFs’ accounting performances decline more rapidly during economic downturns than the corresponding increases during economic upturns due to conservative accounting judgment. In contrast, SNFs are reluctant to lay off employees during adverse periods, possibly due to state-level minimum staffing requirements. When SNFs have to curtail employees during challenging times, lower-skilled employees are more likely to be laid off than higher-skilled employees, consistent with higher future adjustment costs for higher-skilled labor [16,17].

Our evidence on SNFs’ supply-side frictions across business fluctuations complements existing studies focusing on SNFs’ staff level and (demand-side) patient outcomes or quality. For example, Lin (2014) exploits state-level legislation changes in minimum staffing requirements to present a causal relationship between higher-skilled employees and the quality of patient care but insignificant associations for lower-skilled employees. Lin’s evidence is consistent with our finding of SNFs retaining high-skilled nurses and laying off lesser-skilled employees during adverse business fluctuations.

Our study provides causal evidence of financial frictions faced by SNFs during economic downturns and identifies sources of these frictions as conservative accounting judgment and labor adjustment costs. In particular, we find that SNFs face three different sources of difficulties during economic downtowns:

1. A decrease in revenue due to lower demand, reflected in the declined occupancy rate,

2. A persistent level of (higher-skilled) employee costs due to expected higher future labor adjustment costs and state-level minimum staffing requirements, and

3. A potentially lower level of capital provided by external investors due to much more negative accounting performances related to conservative accounting reporting.

These results suggest that imposing minimum staffing requirements needs to consider the difficulties that SNFs face during economic hardship. Our results facilitate the debate over the federal-level minimum staffing requirements that are often based on demand-side evidence.

Acknowledgement

We thank Stephen Ryan for numerous insightful conversations that have improved the paper.

Conflict of Interest

The authors declare no conflicts of interest.

References

1. Harrington C, Ross L, Chapman S, Halifax E, Bakerjian D, et al. (2020) Nurse staffing and coronavirus infections in California nursing homes. Policy Polit Nurs Pract. 21(3):174-186.

   [Crossref] [Google Scholar]

2. Antwi YA, Bowblis JR (2018) The impact of nurse turnover on quality of care and mortality in nursing homes: Evidence from the great recession. Am J Health Econ. 4(2):131-163.

   [Google Scholar] [ResearchGate]

3. The Centers for Medicare and Medicaid Services (2022) CMS acts to improve the safety and quality of care of the nation’s nursing homes.
4. Skilled Nursing News (2021) National nursing home spending reaches $ 196.8 Billion in 2020.
5. Krishnan R, Yetman MH (2011) Institutional drivers of reporting decisions in nonprofit hospitals. J Account Res. 49(4):1001-1039.

   [Crossref] [Google Scholar]

6. Lang MH, Lundholm RJ (2000) Voluntary disclosure and equity offerings: Reducing information asymmetry or hyping the stock? Contemp Account Res. 17(4):623-662.

   [Crossref] [Google Scholar]

7. Allee KD, Yohn TL (2009) The demand for financial statements in an unregulated environment: An examination of the production and use of financial statements by privately held small businesses. Account Rev. 84(1):1-25.

   [Crossref] [Google Scholar]

8. Cassar G, Ittner CD, Cavalluzzo KS (2015) Alternative information sources and information asymmetry reduction: Evidence from small business debt. J Account Econ. 59(2-3):242-263.

   [Crossref] [Google Scholar]

9. Basu S (1997) The conservatism principle and the asymmetric timeliness of earnings. J Account Econ. 24(1):3-37.

   [Crossref] [Google Scholar]

10. Grabowski DC, Feng Z, Hirth R, Rahman M, Mor V (2013) Effect of nursing home ownership on the quality of post-acute care: An instrumental variables approach. J Health Econ. 32(1):12-21.

    [Crossref] [Google Scholar]

11. Penman SH, Zhang XJ (2020) A theoretical analysis connecting conservative accounting to the cost of capital. J Account Econ. 69(1):101236.

    [Crossref] [Google Scholar]

12. Collins DW, Hribar P, Tian XS (2014) Cash flow asymmetry: Causes and implications for conditional conservatism research. J Account Econ. 58(2-3):173-200.

    [Crossref] [Google Scholar]

13. Lawrence A, Sloan R, Sun E (2018) Why are losses less persistent than profits? Curtailments vs. Conservatism. Manage Sci. 64(2):673-694.

    [Crossref] [Google Scholar]

14. Banker RD, Basu S, Byzalov D, Chen JY (2016) The confounding effect of cost stickiness on conservatism estimates. J Account Econ. 61(1):203-220.

    [Crossref] [Google Scholar]  

15. Fama EF, MacBeth JD (1973) Risk, return and equilibrium: Empirical tests. J Polit Econ. 81(3):607-636.

    [Google Scholar] [ResearchGate]

16. Lin H (2014) Revisiting the relationship between nurse staffing and quality of care in nursing homes: An instrumental variables approach. J Health Econ. 37:13-24.

    [Crossref] [Google Scholar]

17. The White House Fact Sheet (2022) Biden-Harris administration announces new steps to improve quality of nursing homes.