## wpiea2021180-print-pdf

## Source details

**Canonical URL:** [wpiea2021180-print-pdf](https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021180-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2021/english/wpiea2021180-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2021/english/wpiea2021180-print-pdf.pdf.json)

---

### I. Introduction — context and motivation
- 2019 U.S. healthcare spending: about $3.8 trillion; about $11,000 per person.
- Healthcare share of GDP: increased from 5 percent in 1960 to about 18 percent in 2019.
- Three macro-critical motivations:
  - Underperformance in lifespan, infant mortality, burden of chronic diseases relative to countries that spend less per capita.
  - Large and increasing healthcare costs with important fiscal consequences and contribution to unsustainable growth in U.S. fiscal entitlement spending.
  - Employer-funded health insurance predominance could crowd out wage growth and shift compensation to non-cash benefits.

### Explanations proposed in the literature for high U.S. healthcare cost
- Luxury good hypothesis: consumption of healthcare increases more than proportionally as income rises.
- Supplier driven demand: providers suggest tests/procedures with unproven benefits; patients comply due to asymmetric information and lack of user-pay incentives.
- Baumol’s cost disease: equalization of wages across sectors increases unit costs in less productive sectors; healthcare considered a less-productive sector.
- Market power of providers and insurers: restricted entry by medical doctors and hospital mergers increase prices and may reduce quality of care.

### Paper’s three-pronged empirical approach
- Estimating market power:
  - Market power measured as markups (ability to charge prices above marginal costs).
  - Estimates for overall healthcare sector (providers, pharmaceuticals, healthcare equipment).
  - Micro-data on U.S. hospitals to calculate markups across states; hospitals account for about one-third of healthcare spending.
  - Wage premia for healthcare practitioners versus other sectors and marginal loss ratios of insurers examined.
  - Correlations assessed between hospital markups, wage premia, and insurer markups.
- Impact of market power on healthcare spending:
  - Connection analyzed between market power and determinants including income and Baumol’s cost disease in cross-country and U.S. state-level settings.
- Event study on ACA impact on healthcare sector wages:
  - Exploits heterogeneity in states’ choices and timing of Medicaid expansion.

### Key findings (summary)
- Market power increases:
  - Markups for publicly listed firms in the U.S. healthcare sector have almost doubled since the early 1980s.
  - Hospital markups increased significantly (by more than 6 percent on average) since the late 1990s across U.S. states.
- Contribution of markups to spending growth:
  - Incorporating markups into OECD cross-country regressions: on average, markups contributed to up to a quarter of annual per capita U.S. healthcare cost growth.
  - U.S. state-level regressions: hospital markups explain about 15 percent of variation in healthcare spending across states.
- Distributional impacts:
  - Rising hospital markups not associated with lower labor costs or lower insurance markups.
  - Physicians’ salaries increasingly rose above salaries of non-physicians with similar education and experience.
- ACA and wages:
  - Wages for healthcare practitioners increased more in Medicaid expansion states versus non-expansion states.
  - Hourly wages for healthcare practitioners and technical occupations increased by 2 to 3 percent more in expansion states over a five-year period.
  - Robust across OEWS, Medicare labor cost index, and CPS data.
  - No significant evidence that wages for other professions increased more in expansion than non-expansion states.

### II. Key institutional arrangements of the U.S. healthcare system (highlights)
- Ownership and provision:
  - Government-owned, non-profit, and for-profit institutions provide healthcare.
  - For-profit hospitals account for about one fifth of the total number of hospitals.
- Insurance coverage mix:
  - More than half of the population covered by private insurance; another 40 percent by Medicare and Medicaid.
  - Medicare: entitlement for persons age 65 and older; Part A (hospital insurance), Part B (medical insurance); Medicare Advantage (Part C) since 1973; Part D added in 2003.
  - Medicaid: state-administered, means-tested; eligibility varies by state.
- Affordable Care Act (effective 2014 provisions highlighted):
  - Expanded Medicaid eligibility in participating states from 100 percent to 138 percent of the federal poverty line.
  - Prohibited most plans from excluding preexisting conditions and several discriminatory practices.
  - Result: share of uninsured approximately halved, from 16 percent before the Act to about 9 ½ percent in 2019.

### Box 1 — Flow of funds in U.S. healthcare (2019, National Health Expenditures = US$3.8 trillion)
- Funding shares (2019):
  - About 40 percent of funds come from public sources (Medicare/Medicaid).
  - About one third come from private health insurance.
  - Out of pocket expenses account for about 10 percent of total expenditures.
  - A relatively small proportion (3 percent) is part of the Marketplace.
  - About 10 percent remains uninsured.
- National Health Expenditures composition (2019 chart highlights):
  - Out of pocket 11%
  - Private Health Insurance 32%
  - Medicare 21%
  - Medicaid 16%
  - Other Health Insurance Programs 4%
  - Other Third Party Payers 11%
  - Investment 5%

### III. Observations and U.S. comparison to other countries
- Growth and outlier status:
  - U.S. per capita health expenditures (USD PPP) rose to about four times the median of the OECD by 2019.
  - Higher U.S. per capita income does not explain the large spending differences.
- Health outcomes:
  - Life expectancy significantly lower than comparator countries and has declined after 2012.
  - Chronic diseases and obesity both about twice as high as in other advanced countries.
- Spending composition (2018 per capita, Table 1):
  - Inpatient and outpatient: USA $6,624; Comparable Countries $2,718
  - Prescription drugs and medical goods: USA $1,397; Comparable Countries $884
  - Administrative: USA $937; Comparable Countries $201
  - Long-term: USA $516; Comparable Countries $1,111
  - Preventive: USA $309; Comparable Countries $175
  - Other: USA $854; Comparable Countries $439
  - Total: USA $10,637; Comparable Countries $5,528
- Administrative spending more than four times higher per capita in the U.S. than in comparable countries.
- Price differences are broad based across inpatient and outpatient procedures; U.S. procedure prices are significantly higher than other countries.

### Increased utilization versus price growth (2014–18 decomposition)
- Price growth contributed approximately three times as much as changes in utilization to the growth in total expenditures.
- Average length of stay in hospital below the OECD average, indicating U.S. is not an outlier in utilization.

### Density of healthcare providers and workforce projections
- In 2018: 2.5 physicians per 1000 in the U.S., well below other advanced economies.
- Forecasted shortage: 139,160 physician jobs by 2030 (forecast cited).
- Policy example: Resident Physician Shortage Reduction Act of 2021 would raise Medicare supported graduate medical education positions by 2,000 per year for seven years, total 14,000 new slots (policy example cited).

### Market concentration in U.S. healthcare
- HHI thresholds: 1,500 moderately concerning; 2,500 high concentration.
- Hospital market concentration: HHI of 5,790 in 2016.
- Empirical evidence: hospital price increases often exceed 20 percent when mergers occur in concentrated markets; Cooper and others (2019) find prices increased by over 6 percent when merging hospitals were geographically close (e.g., 5 miles or less).

### Overall healthcare sector markups
- Data and method:
  - Markups estimated following Diez and others (2018) using Worldscope/Datastream data.
  - Sample: Worldscope contains information from over 81,000 publicly listed companies in 74 economies; markups computed for 9 advanced countries.
- Finding: U.S. healthcare median markups have increased by more than 70 percent since 1980; median markup of other advanced economies risen by about 40 percent.

### U.S. hospital markups (HCRIS data, 1997–2018)
- Sample: more than 5000 hospitals after cleaning.
- Finding: Hospitals’ median markups have been increasing by about 6 percent since 1997.
- Distribution: Broadened and become more skewed between 1998 and 2018.

### Healthcare practitioners — wage premia (CPS, 1997–2019)
- Method: Compare wages of healthcare practitioners and technical occupations to workers in other occupations with same education, experience, demographics.
- Findings:
  - Wage premia of healthcare professionals close to 30 percent on average.
  - Wage premia for physicians and surgeons about 50 percent on average.
  - Physician premia largely stable over two decades with a small upward trend in the last decade.
  - Physician and surgeon premia more cyclical than general healthcare professional premia.

### Insurers — medical loss ratios (MLR) and markups
- Insurer markup defined as inverse of MLR. MLR = medical claims expenditures / premium revenues.
- ACA MLR regulation: insurers required to maintain MLR of at least 80 percent in individual and small group markets and 80 percent in large group market segments; rebates required if below threshold.
- Data: CMS regulatory database for 2,486 insurance companies (post-cleaning); self-insured plans excluded.
- Empirical patterns:
  - MLRs vary significantly in the individual market; less variation for group insurance.
  - Individual market distribution of MLRs increased between 2011 and 2014 and shifted left thereafter.
  - Median MLR across states increased substantially in 2014 among the individual market and subsequently decreased.
  - MLRs negatively correlated with market concentration (Herfindahl index) for individual and small group markets; not significant for large group market.

### Correlations across market power measures
- Pairwise correlations:
  - Hospital markup with Insurer MLR (individual): -0.0723
  - Wage premium with Insurer MLR (individual): 0.0952* (*p<0.05)
- Interpretation:
  - Small positive correlation between hospital markups and wage premia implies higher hospital market power is not associated with lower wage premia for doctors and nurses.
  - Negative correlation between hospital markups and insurance MLRs implies hospital market power not associated with lower insurer markups.
  - Positive correlation between wage premia and insurer MLRs implies states where insurers have more market power, physicians and nurses earn less.

### Market power and healthcare cost determinants — econometric strategy (OECD-level)
- Basic regression (levels) dependent variable: log(HCE_per_capita_{i,t}).
- Key regressors:
  - Adjusted Baumol variable: 1/L_{i,t} · (log W_{i,t} − log Y_{i,t}).
  - log(MPR_{i,t}) overall markups for the healthcare sector.
  - Interaction: log(MPR_{i,t}) · 1/L_{i,t} · (log W_{i,t} − log Y_{i,t}).
  - Controls: real GDP per capita and percentage of population 65+.
- Estimator choices:
  - Primary: cross-sectionally distributed lag (CS-DL) regression (Chudik and others, 2017) for long-run relationships.
  - Comparison: Pedroni’s (1996) FMOLS.
- Data:
  - OECD sample excluding markups: 20 countries, 1970–2019.
  - When markups included: sample reduces to 9 countries, 1980–2018.
- Unit root testing: IPS and Fisher tests; for markups the null of a panel unit root cannot be rejected at 5 percent.

### Results — OECD regressions
- Fixed effects regressions: income and Baumol’s cost disease significant; coefficients on markup and interaction positive and significant.
- CS-DL (long-run): long-run positive effects of estimated markups and their interaction with Baumol variable on per capita healthcare costs.
- Quantitative contributions (CS-DL specification 3):
  - Markups account for about a quarter of annual increases in U.S. real health spending per capita on average since the early 1980s.
  - U.S. markups and interactions with the Baumol variable account for about 1 percent of the 4 percent average annual increases of total per capita U.S. healthcare costs since 1980.
  - Counterfactual: If U.S. markups had risen in line with the average of other OECD countries, healthcare costs would have risen by 0.6 percent less per year in the U.S.; real per capita spending today could be about 10 percent lower.
  - Baumol variable accounts for about 0.006 (0.6 percent) of annual healthcare cost growth on average.
  - Alternative index: markups accounted for about 43 out of a total increase of 264 in the index since 1981 (about 16 percent).

### U.S. states regressions — specification and data
- State regression (first differences) sample: all 50 U.S. states, 1980–2019.
- Dependent variable: Δlog(HCE_{s,t}), HCE from BEA.
- Key regressors: 1/L_{s,t} · (Δlog W_{s,t} − Δlog Y_{s,t}); controls include real GSP per capita, percentage 65+, poverty rate, union coverage, unemployment rate.
- Estimation strategy: first differences with time and state fixed effects; extract state fixed effects and regress on average hospital markups across states.

### U.S. states regression results
- Determinants of growth (first differences):
  - Baumol variable coefficient ≈ 0.05.
  - Per capita income coefficient ≈ 0.5.
  - Greater share of elderly associated with faster healthcare cost growth.
  - Higher unemployment rate growth associated with faster healthcare cost growth.
  - Roughly two-thirds of variation in growth of healthcare costs explained by right-hand side variables in first differences.
- Hospital markups and fixed effects:
  - Cross-state regression explains about 15.0 percent of the variation (R^2 = 15.0%).
  - Coefficient: a 0.1 increase in average hospital markup raises the extracted fixed effect by 0.29 percent.
  - No significant relationship between fixed effects and measures of wage premia or insurance markups.
- Robustness: MSA-level regressions using HCCI Healthy Marketplace index show hospital concentration significant over time in driving healthcare costs; positive and significant correlation between hospital concentration and the Baumol variable.

### Section VI — Impact of the ACA on healthcare sector wages (Medicaid expansion)
- Research design:
  - Exploits heterogeneity in Medicaid expansion timing across states.
  - Hypothesis: expansion states see healthcare practitioner wages grow faster than non-expansion states due to demand shock with limited supply response.
- Contextual facts:
  - Healthcare labor ~ one-half of total healthcare spending.
  - 2014 provisions insured about 20 million low-to-middle income non-elderly adults.
  - Insurance coverage increased about 4 percent more in expansion states than non-expansion states.
- Key empirical findings:
  - Aggregate/state-level:
    - Between 2014–2019, wage growth for healthcare practitioners in expansion states about 0.5 to 0.6 percent higher per year than in non-expansion states.
    - Basic difference-in-difference: hourly wages in expansion states grew about 0.7 percent faster on average in each expansion year relative to non-expansion states.
    - Dynamic effects: differential impact turns significant in the third to fourth year; cumulative state-level effect about 2 percent over five years.
  - Metropolitan and microdata:
    - Medicare wage adjustment indexes consistent with state-level findings; significant starting in 2012 for some areas.
    - Individual-level CPS regressions: wages in expansion states in expansion years about 4 to 6 percent higher on average (CPS top-coding may underestimate effect).
  - Supply-side:
    - Employment-to-population for healthcare practitioners increased in expansion states but not enough to offset wage pressures.
    - Higher ratio of healthcare employment to non-healthcare employment corresponds with a lower ratio of healthcare to non-healthcare wages.
  - Labor mobility:
    - Census Job-to-Job flow data show no higher net inflows of healthcare and social service workers in expansion states as a ratio of population.
- Interpretation:
  - Medicaid expansion increased demand and produced statistically and economically significant wage increases for healthcare practitioners.
  - Evidence consistent with limited supply elasticity and potential increases in provider market power when demand rises.

### Policy responses and recommendations
- Reduce barriers to entry and encourage entry by new health providers and insurers where entry barriers drive market power.
- Recalibrate licensing requirements or limits on flow of new medical professionals when binding constraints to entry.
- Adopt a more assertive approach to antitrust policies (federal and state) to identify and counter restraints of trade.
- Consider actions to constrain pricing power by providers and insurers in highly concentrated local markets.
- Consider creating an agency responsible for monitoring and oversight of healthcare markets.
- Increase transparency on price, quality, costs, and provider networks to improve consumer choice and competitiveness.
- Policy context noted:
  - Executive Order on “Strengthening Medicaid and the Affordable Care Act” (January 28, 2021) provides a special enrollment window.
  - 2021 American Rescue Plan (ARP) features described: fully paid coverage for lowest income workers, increased premium subsidies up to 400 percent of the federal poverty level, capped benchmark plan costs for two years, incentives for Medicaid expansion.
  - Proposed American Families Plan contains provisions to make ARP premium reductions permanent (proposal cited).

### Empirical evidence and selected econometric findings (selected statistics)
- Panel cointegration tests (OECD, logs):
  - Westerlund Panel means statistic: -2.1149; p-value: 0.0172
  - Pedroni - Phillips-Perron Panel means statistic: -2.1783; p-value: 0.0147
  - Kao- Augmented Dickey-Fuller Panel means statistic: -1.9924; p-value: 0.0232
- Fixed effects regressions on changes in insurer MLRs (Observations 400; Number of cid 50):
  - ∆HHI (Ind.) coefficient = -0.0315* (0.0167); R-squared = 0.010
  - ∆HHI (small) coefficient = -0.0220*** (0.00773); R-squared = 0.023
  - ∆HHI (large) coefficient = -0.00431 (0.00749); R-squared = 0.001
- Basic OECD first-difference regressions (selected coefficients):
  - baumol coefficients: 0.440*** (0.0670), 0.443*** (0.0722), 0.449*** (0.0765) (varying specs)
  - dlrgdp_pc coefficients: 0.460*** (0.0852), 0.462*** (0.0828)
  - R-squared examples: 0.331, 0.339, 0.344
- CS-DL and FMOLS (9-country sample):
  - lmarkup point estimates: 4.106* (2.373), 5.603** (2.691)
  - int_level (interaction) estimates: 0.355* (0.201), 0.508** (0.246)
  - R-squared examples: 0.391, 0.352; FMOLS R-squared 0.996 (reported)
- State-level regressions:
  - dbaumol coefficients: 0.0441*** (0.00304), 0.0491*** (0.00280)
  - dlog_GSP_pc example: 0.400*** (0.0594)
  - Cross-state regression: R^2 = 15.0% for hospital markup explaining fixed effects.
- Medicaid expansion wage analyses (selected):
  - Cross-section hourly wage growth (2014–2019): exp_dum coefficients: 2.776*** (0.959), 2.308** (0.971) across columns; Observations: 51.
  - Difference-in-difference state-level log hourly wages: exp_dum = 0.007** (0.003); Observations: 510; R-squared ~0.893–0.898.
  - Area-level (MSA) med_exp_dum = 0.026*** (0.002); Observations: 3,561; Number of areas: 376.
  - Individual CPS log hourly wages: med_year_dum coefficients: 0.059** (0.025), 0.049* (0.025); Observations: 5,945.
  - Between-occupational wage ratios: Medicaid expansion coefficient = 0.007** (0.003); Observations: 51; R-squared ~0.216–0.222.

### Conclusions (select points)
- Market power contribution:
  - Market power in the U.S. healthcare sector has increased since the 1980s and contributed to rising healthcare costs.
  - Markups for publicly listed firms in the U.S. healthcare sector have almost doubled since the early 1980s.
  - Hospital markups increased by more than 6 percent on average since the late 1990s and vary across states.
  - Incorporating markups into OECD regressions: on average, markups account for up to a quarter of annual increases in per capita U.S. healthcare costs.
  - U.S. state-level regressions: hospital markups explain about 15 percent of variation in healthcare spending across states.
- Medicaid expansion wage effects:
  - Medicaid expansion associated with higher wages for healthcare practitioners in expansion states; dynamic analyses show effects start around third to fourth year and accumulate to about 2 percent over five years at the state-level, with individual-level estimates of about 4 to 6 percent higher wages.
- Policy implication:
  - Coverage expansions that increase demand without commensurate supply-side reforms or competition improvements risk raising provider market power and prices in addition to direct coverage costs.

*Source: wpiea2021180-print-pdf — content unit.*

### References .............................................................................................................

### wpiea2021180-print-pdf - References .............................................................................................................

### I. INTRODUCTION — context and motivation
- In 2019, the United States spent about $3.8 trillion on healthcare or about $11,000 per person.
- The share of healthcare costs in GDP increased from 5 percent in 1960 to about 18 percent in 2019.
- Three macro-critical reasons for focusing on healthcare:
  - Underperformance in effective delivery of services (lifespan, infant mortality, burden of chronic diseases) that is significantly below countries that spend less per capita, increasing inequality and weighing on economic growth.
  - Large and increasing healthcare costs have important fiscal consequences and are a driver of unsustainable growth in U.S. fiscal entitlement spending.
  - Predominance of employer-funded health insurance means increasing healthcare costs could crowd out wage growth and shift compensation to non-cash benefits.

### Explanations proposed in the literature for high U.S. healthcare cost
- Luxury good hypothesis: consumption of healthcare increases more than proportionally as income rises.
- Supplier driven demand: providers suggest tests/procedures with unproven benefits; patients comply due to asymmetric information and lack of user-pay incentives.
- Baumol’s cost disease: equalization of wages across sectors increases unit costs in less productive sectors; healthcare considered a less-productive sector.
- Market power of providers and insurers: restricted entry by medical doctors and hospital mergers increase prices and may reduce quality of care.

### Paper’s three-pronged approach
- Estimating market power:
  - Measure market power as markups (ability to charge prices above marginal costs).
  - Provide estimates for the overall healthcare sector (providers, pharmaceuticals, healthcare equipment).
  - Use micro-data on U.S. hospitals to calculate markups across states; hospitals account for about one-third of healthcare spending.
  - Explore wage premia for healthcare practitioners versus employees in other sectors, and assess marginal loss ratios of insurers.
  - Assess correlations between hospital markups, wage premia, and insurer markups.
- Impact of market power on healthcare spending:
  - Analyze connection between market power and determinants including income and Baumol’s cost disease in cross-country and U.S. state-level settings.
- Event study on ACA impact on healthcare sector wages:
  - Exploit heterogeneity in states’ choices and timing of Medicaid expansion.

### Key findings (summary)
- Market power increased significantly since the 1980s:
  - Markups for publicly listed firms in the U.S. healthcare sector have almost doubled since the early 1980s.
  - Hospital markups increased significantly (by more than 6 percent on average) since the late 1990s across U.S. states.
- Contribution of markups to spending growth:
  - Incorporating markups into OECD cross-country regressions shows that, on average, they contributed to up to a quarter of annual per capita U.S. healthcare cost growth.
  - U.S. state-level regressions show hospital markups explain about 15 percent of variation in healthcare spending across states.
- Distributional impacts within the sector:
  - Rising hospital markups are not associated with lower labor costs or lower insurance markups (providers appear to use market power to raise consumer prices rather than reducing upstream costs).
  - Physicians’ salaries have increasingly risen above salaries of non-physicians with similar education and experience.
- ACA and wages:
  - After the ACA, wages for healthcare practitioners increased more in Medicaid expansion states versus non-expansion states.
  - Hourly wages for healthcare practitioners and technical occupations increased by 2 to 3 percent more in expansion states over a five-year period.
  - Results confirmed using state-level Occupational Employment and Wage Statistics, the metropolitan-level Medicare labor cost index, and individual-level Current Population Survey (CPS) data.
  - No significant evidence that wages for other professions increased more in expansion than non-expansion states.

### Contribution and interpretation
- Threefold contribution:
  - Presents evidence on evolution of market power for the healthcare sector overall and vis-à-vis other OECD countries, with granular U.S. microdata for hospitals.
  - Investigates how this evolution relates to healthcare costs in advanced countries and across U.S. states.
  - Documents that the supply of medical services may be relatively inelastic to demand boosts from Medicaid expansion—potentially due to barriers to entry, limited mobility of providers across states, or oligopolistic providers responding to higher demand with higher prices rather than increased supply.
- Structure of the remainder of the paper:
  - Section II: key institutional features of the U.S. healthcare system.
  - Section III: benchmarks U.S. healthcare costs and compares prices and outcomes internationally.
  - Section IV: analyzes evolution of market power in the U.S. healthcare sector (overall sector, hospitals, insurers, medical providers).
  - Section V: examines connection between market power and other determinants including Baumol’s cost disease and quantifies their effect.
  - Section VI: studies the impact of the ACA on healthcare sector wages.
  - Section VII: concludes.

### II. Key institutional arrangements of the U.S. healthcare system (highlights)
- Ownership and provision:
  - Healthcare provided by government-owned, non-profit, and for-profit institutions.
  - For-profit hospitals account for about one fifth of the total number of hospitals.
- Insurance coverage mix:
  - More than half of the population is covered by private insurance; another 40 percent by Medicare and Medicaid.
  - Medicare: entitlement for persons age 65 and older; Part A (hospital insurance), Part B (medical insurance); options for Medicare Advantage (Part C) since 1973; Part D added in 2003.
  - Medicaid: state-administered, means-tested program; eligibility varies by state.
- Employer-based coverage has historical roots in WWII employer benefit practices.
- The 2010 Affordable Care Act (ACA):
  - Retained structure of Medicare, Medicaid, and employer-sponsored market.
  - Effective in 2014: prohibited most plans from excluding preexisting conditions, restricted discrimination based on health status, prohibited annual monetary caps, required guaranteed issue/renewal, premium rating rules, nondiscrimination in benefits, and mental health/substance abuse parity.
  - Expanded Medicaid eligibility in participating states from 100 percent to 138 percent of the federal poverty line (effective 2014, with federal subsidies).
  - Introduced individual mandate and Marketplace (effective 2014); individual mandate penalty effectively repealed starting in 2019.
  - Result: share of uninsured approximately halved, from 16 percent before the Act to about 9 ½ percent in 2019.

### Box 1 — Flow of funds in U.S. healthcare (highlights)
- Financing flows from U.S. households into health insurance funds (governmental and private), which disburse funds to providers.
- Funding shares (2019, National Health Expenditures = US$3.8 trillion):
  - About 40 percent of funds come from public sources (Medicare/Medicaid).
  - About one third come from private health insurance.
  - Out of pocket expenses account for about 10 percent of total expenditures.
  - More than half of the population is covered by employer-sponsored health insurance; about forty percent by Medicare and Medicaid.
  - A relatively small proportion (3 percent) is part of the Marketplace.
  - About 10 percent remains uninsured.
- National Health Expenditures composition (2019 chart highlights):
  - Out of pocket 11%
  - Private Health Insurance 32%
  - Medicare 21%
  - Medicaid 16%
  - Other Health Insurance Programs 4%
  - Other Third Party Payers 11%
  - Investment 5%

### III. Observations and U.S. comparison to other countries
- Growth and outlier status:
  - Share of healthcare in GDP grew in all OECD countries, but growth in the U.S. was much faster; by 2019 the U.S. was above all other OECD countries.
  - U.S. per capita health expenditures (USD PPP) rose to about four times the median of the OECD by 2019.
  - Higher U.S. per capita income does not explain the large spending differences.
- Health outcomes:
  - High healthcare expenditures in the U.S. do not correspond to better outcomes; life expectancy is significantly lower than comparator countries and has declined after 2012.
  - Potential determinants include inefficient utilization, high prevalence of chronic diseases and obesity (both about twice as high as in other advanced countries), opioid crisis, and gun violence.
- Spending composition and international comparisons (2018 per capita, Table 1):
  - Inpatient and outpatient: USA $6,624; Comparable Countries $2,718
  - Prescription drugs and medical goods: USA $1,397; Comparable Countries $884
  - Administrative: USA $937; Comparable Countries $201
  - Long-term: USA $516; Comparable Countries $1,111
  - Preventive: USA $309; Comparable Countries $175
  - Other: USA $854; Comparable Countries $439
  - Total: USA $10,637; Comparable Countries $5,528
  - Comparable countries include Austria, Belgium, Canada, France, Germany, Netherlands, Sweden, Switzerland, and the United Kingdom.
- Administrative spending is more than four times higher per capita in the U.S. than in comparable countries; complexity of the system is a likely driver, but administrative costs do not account for the bulk of differences.
- Largest percentage of spending goes to inpatient and outpatient care (about 60 percent of total spending); U.S. per capita inpatient and outpatient spending is about twice that of comparable countries.
- Malpractice payouts average about $5 billion per year — a small fraction of total healthcare cost — and utilization of healthcare resources in the U.S. is not higher than comparable countries (countering the claim that fear of litigation drives excessive procedures).
- Price differences are broad based across inpatient and outpatient procedures; with rare exception, U.S. procedure prices are significantly higher than other countries.

*Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021180-print-pdf.pdf*

### conclusions.

### conclusions.

### Increased utilization versus price growth
- Decomposition of changes in total U.S. healthcare expenditures for the period 2014–18 shows price growth contributed approximately three times as much as changes in utilization to the growth in total expenditures.
- Notes: Metrics like average length of stay in hospital are below the OECD average, suggesting that the U.S. is not an outlier in utilization.

### Density of healthcare providers
- In 2018, the U.S. had 2.5 physicians per 1000, well below other advanced economies, and the difference to other countries has been increasing over time.
- There is no evidence that the U.S. has a higher density of practicing nurses than other countries today; historical differences have narrowed as U.S. nursing growth has not kept pace with other countries.
- Forecasts cited predict a shortage of 139,160 physician jobs by 2030.
- Policy effort example: the Resident Physician Shortage Reduction Act of 2021 would gradually raise the number of Medicare supported graduate medical education positions by 2,000 per year for seven years, for a total of 14,000 new slots.
- Barriers for immigrating physicians include strict licensing requirements and often higher U.S. residency year requirements for foreign-trained doctors.

### Market concentration in U.S. healthcare
- Herfindahl-Hirschman Index (HHI) thresholds: 1,500 indicates a moderately concerning concentration level; 2,500 indicates high concentration.
- Insurers, specialist physicians, and hospitals markets are all highly concentrated; hospital market concentration reported with an HHI of 5,790 in 2016.
- Primary care physicians are between the moderate and high concentration levels, with a rapid increase in HHI attributed in part to private practice acquisitions by larger groups.
- Empirical findings: hospital price increases often exceed 20 percent when mergers occur in concentrated markets; Cooper and others (2019) find prices increased by over 6 percent when merging hospitals were geographically close (e.g., 5 miles or less), but not when geographically distant (e.g., over 25 miles). These price increases did not appear to improve quality.

### Overall healthcare sector markups
- Methodology: Markups estimated following Diez and others (2018), using production function estimation and firm financial data (Worldscope via Datastream).
- Sample: Worldscope contains information from over 81,000 publicly listed companies in 74 economies; markups computed for 9 advanced countries (Australia, Canada, Germany, France, UK, Japan, Korea, Sweden, United States).
- Finding: U.S. healthcare median markups have increased by more than 70 percent since 1980. The median markup of other advanced economies has risen by about 40 percent.
- Within the U.S., healthcare sector markups have risen rapidly compared to industrials and less so compared to all other industries.

### Hospitals (U.S. hospital markups)
- Data source: CMS Healthcare Provider Cost Reporting Information System (HCRIS) cost reports; sample after cleaning: more than 5000 hospitals between 1997 and 2018.
- Finding: Hospitals’ median markups have been increasing by about 6 percent since 1997.
- Distribution: The distribution of hospital markups has broadened and become more skewed between 1998 and 2018, consistent with markup increases being driven by the largest hospital groups.

### Healthcare practitioners (wage premia)
- Data source: Current Population Survey, 1997 to 2019.
- Method: Wage premia estimated by comparing wages of healthcare practitioners and technical occupations to workers in other occupations with same years of education, work experience, and demographic characteristics; for physicians, residency time counted as work experience.
- Findings:
  - Wage premia of healthcare sector professionals are close to 30 percent on average.
  - Wage premia for physicians and surgeons are about 50 percent on average over the estimation horizon.
  - Wage premia for general healthcare professionals and for physicians have been largely stable over the past two decades, with physician premia on a small upward trend in the last decade.
  - Physician and surgeon wage premia appear more cyclical than general healthcare professional premia, suggesting some healthcare services could resemble a “luxury good” more than a “necessity good.”

### Insurers (medical loss ratios and markups)
- Definition: Insurer markup defined as the inverse of the medical loss ratio (MLR). MLR = expenditures on medical claims / premium revenues.
- ACA regulation: Insurers required to maintain an MLR of at least 80 percent in the individual and small group market segments and 80 percent in the large group market segment; insurers falling below threshold must send rebates to policyholders.
- Data: CMS regulatory database used to extract MLRs for 2486 insurance companies (post-cleaning). Self-insured plans are not required to report MLR and are excluded.
- Empirical patterns:
  - MLRs vary significantly among insurers in the individual market with much less variation for group insurance.
  - The individual market distribution of MLRs increased between 2011 and 2014 and shifted left thereafter, with more movement among firms close to the 0.80 federal threshold; very little variation across large group insurers.
  - Median MLR across states first increased substantially in 2014 among the individual market and subsequently decreased; less movement for the large group insurers.
  - There is significant cross-state variation in MLRs and MLRs are negatively correlated with market concentration (Herfindahl index). Regressions of annual changes in MLRs across states on Herfindahl indices show a clear negative relationship significant for the individual market and small group but not for the large group market.

### Correlations across market power measures
- Pairwise correlations indicate:
  - The correlation between hospital markups and wage premia is very small and positive, implying higher hospital market power is not associated with lower wage premia for doctors and nurses.
  - The correlation between hospital markups and insurance MLRs is negative, implying more hospital market power is not associated with lower insurer markups.
  - There is a positive correlation between wage premia of practitioners and insurance MLRs, implying that in states where insurance companies have more market power, physicians and nurses earn less.
- Interpretation: Results suggest hospitals use market power to raise prices to consumers rather than exercising monopsony power to lower provider wages. Evidence is consistent with insurer concentration being associated with lower physician prices.

### Market power and healthcare cost determinants
- Context: Previous literature typically finds income as the most important factor explaining healthcare cost differences across countries, with a health care income elasticity between 0.4 and 0.9 and a smaller effect of Baumol’s cost disease; limited evidence that aging is a significant determinant.
- Contribution: This work integrates estimates of overall markups for OECD countries and hospital markups for U.S. states into standard cross-country and state-level regressions of per capita healthcare expenditures to assess whether market power increases healthcare costs and exacerbates Baumol’s cost disease.

*Source: wpiea2021180-print-pdf - conclusions.*

### 33.      We first estimate regressions at the OECD level. The basic regression takes the

### 33. We first estimate regressions at the OECD level.

### Methodology: OECD-level regressions
- Basic regression specification (levels):
  - Dependent variable: log(HCE_per_capita_{i,t}) where HCE are per capita real current health-care expenditures.
  - Key regressors:
    - Adjusted Baumol variable: 1/L_{i,t} · (log W_{i,t} − log Y_{i,t})
      - L_{i,t} = (L_H / L_T)_{i,t} denotes the share of total labor employed in the services sector.
      - W = overall real wage per worker (total real wages and salaries divided by total persons employed).
      - Y = economy-wide output per worker (real GDP divided by total persons employed).
    - log(MPR_{i,t}) overall markups for the healthcare sector.
    - Interaction: log(MPR_{i,t}) · 1/L_{i,t} · (log W_{i,t} − log Y_{i,t}).
    - Z controls: include real GDP per capita and percentage of the population 65 years of age and older.
  - Fixed effects: country (μ_s) and time (τ_t) fixed effects.
  - Residual: ε_{s,t}.
- Econometric concerns and choices:
  - Variables in equation 1 are integrated of order 1 and there is evidence of cointegration and cross-sectionally dependent error terms.
  - Cross-sectional dependence likely due to worldwide technological progress and common shocks.
  - Estimator choice:
    - Primary: cross-sectionally distributed lag (CS-DL) regression (Chudik and others, 2017) to estimate long-run relationships; robust to serial correlation and breaks; valid for regressors whether I(0) or I(1).
    - Comparison: Pedroni’s (1996) fully modified OLS estimator for cointegrated panel models.
  - CS-DL specification includes cross-sectional averages of dependent variable and up to three lags for cross-sectional averages of regressors.

### Data: OECD and markups coverage
- OECD sample (excluding markups):
  - 20 OECD countries, 1970 to 2019.
  - Countries: Australia, Austria, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Korea, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, UK, USA.
- Markups coverage:
  - Limited; when markups are included sample reduces to 9 countries, 1980 to 2018.
  - Countries with markups: Australia, Canada, Germany, France, Japan, Korea, Sweden, UK, USA.
- Data sources:
  - Healthcare and demographic data: OECD Health Database.
  - Macroeconomic variables (GDPs, wages): OECD Annual National Accounts.
  - Employment data: OECD STAN Database for Structural Analysis.
- Unit root testing:
  - Im-Pesaran-Shin and Fisher-type unit root tests performed on all variables.
  - For markups, the null of a panel unit root cannot be rejected at the 5 percent level for the IPS and Fisher (Dicker Fuller) specification.

### Results: OECD regressions (fixed effects and CS-DL)
- Fixed effects regressions (levels and first differences):
  - Income and Baumol’s cost disease are significant drivers of healthcare costs across countries.
  - Estimated coefficients on the markup and interaction with the Baumol variable are positive and significant.
  - Other explanatory variables generally not significant.
  - Evidence of cointegration (Table 4) and cross-sectional dependence rejected at the 5 percent level for regressions in levels (CD statistic, Table 5).
- CS-DL results (long-run):
  - Provide evidence of long-run positive effects of estimated markups and their interaction with the Baumol variable on healthcare costs per capita (Table 6).
  - Included variables: Baumol variable, per capita income, overall healthcare sector markups, and their interaction with Baumol.
  - Robustness: checked with different numbers of lags; coefficients have expected sign and magnitudes broadly in line with fixed effects regressions.
  - Difference from fixed effects: real per capita income is no longer significant in CS-DL.
  - Explained variation: Roughly one-third of the variation in healthcare costs can be explained by the right-hand side variables.

### Quantitative contributions to U.S. healthcare cost growth (CS-DL based)
- Based on CS-DL regressions (specification 3):
  - Markups account for about a quarter of annual increases in U.S. real health spending per capita on average since the early 1980s.
  - U.S. markups and interactions with the Baumol variable (and their lags) account for about 1 percent of the 4 percent average annual increases of total per capita U.S. healthcare costs since 1980.
  - Counterfactual: If U.S. markups had instead risen in line with the average of other OECD countries in the sample, healthcare costs would have risen by 0.6 percent less per year in the U.S. and real per capita spending today could be about 10 percent lower today.
  - Baumol variable accounts for about 0.006 (or 0.6 percent) of annual healthcare cost growth on average.
  - Alternative index construction (1981=100): Markups accounted for about 43 out of a total increase of 264 in the index since 1981, or about 16 percent.

### U.S. states regressions: specification and data
- State-level regression (first differences):
  - Specification for state s at time t:
    - Δlog(HCE_{s,t}) = β0 + β1 · 1/L_{s,t} · (Δlog W_{s,t} − Δlog Y_{s,t}) + α · Δlog Z_H_{s,t} + μ_s + τ_t + ε_{s,t}.
    - L_{s,t} = (L_H / L_T)_{s,t} share of total labor employed in the healthcare sector.
    - HCE: real healthcare costs per capita from the Bureau of Economic Analysis (BEA).
    - W: overall real wage per worker (total real wages and salaries divided by total persons employed, including self-employed), by state-year from BEA.
    - Y: real gross state product (GSP) per worker (real GSP divided by total persons employed, including self-employed), by state-year from BEA.
    - Z: controls include real GSP per capita, percentage of population 65+, poverty rate, union coverage rate, and unemployment rate (Bureau of Labor Statistics, Bureau of the Census, Unionstats.com).
  - Sample: all 50 U.S. states, 1980-2019.
  - Estimation strategy:
    - Use first differences with time and state fixed effects due to short span of state-level hospital markups (start in 1997) and markups variation mostly across states rather than over time.
    - Extract state fixed effects from equation and regress them on average estimated hospital markups across states.

### Results: U.S. states regressions
- Determinants of growth (first differences):
  - Baumol’s cost disease and real per capita income are significant determinants of growth of healthcare costs across U.S. states.
  - Estimated coefficients:
    - Baumol variable coefficient ≈ 0.05.
    - Per capita income coefficient ≈ 0.5 (consistent with prior literature).
  - Other significant influences:
    - Greater share of elderly (population 65+) associated with faster healthcare cost growth.
    - Higher unemployment rate growth associated with faster healthcare cost growth (may reflect increased spending on Medicare during recessions).
  - Non-significant variables:
    - Growth of union rate and poverty rate unrelated to growth of healthcare costs.
  - Explaining variation:
    - Roughly two-thirds of the variation in growth of healthcare costs can be explained by the right-hand side variables in first differences.
    - A robustness check with a CS-DL specification yields a smaller R^2 but similar coefficients.
- Hospital markups and fixed effects:
  - Higher hospital markups are associated with larger state fixed effects from the state-level regressions.
  - Cross-state regression explains about 15 percent of the variation (R^2 = 15.0%).
  - Coefficient interpretation: A 0.1 increase in average hospital markup raises the extracted fixed effect by 0.29 percent (thereby raising healthcare cost growth).
  - No significant relationship found between fixed effects and measures of wage premia for healthcare professionals or insurance markups (possible measurement issues and hospital-based care being the largest component of U.S. healthcare spending).
- Robustness and geographic issues:
  - Caveat: hospital markets can span across states; geographic markets for medical care are difficult to define.
  - Additional check: regressions at the U.S. metropolitan (MSA) level using HCCI Healthy Marketplace index on hospital concentration:
    - Hospital concentration significant over time in driving healthcare costs.
    - Positive and significant correlation between hospital concentration and the Baumol variable, consistent with OECD results.

### Section VI — Impact of the Patient Protection and Affordable Care Act (ACA) on healthcare sector wages
- Research question:
  - Use heterogeneity in Medicaid expansion timing across states to identify ACA impact on labor costs and infer supply elasticity.
  - Hypothesis: Medicaid expansion states would see healthcare practitioner wages grow faster than non-expansion states due to increased demand not matched by supply.
- Context and prior evidence:
  - Healthcare labor makes up about one-half of total healthcare spending (Glied and others, 2015).
  - Evidence of supply not keeping pace with demand post-ACA:
    - Increased emergency response times (Courtemanche and others, 2019).
    - Increased wait times for Medicaid appointments (Miller and Wherry, 2017).
    - No statistically significant response in local medical education market (Dillender and others, 2019).
  - Evidence on hospital financials:
    - Rhodes and others (2020): Medicaid expansion led to decrease in uncompensated care and increase in average operating margins.
    - Moghtaderi and others (2020): hospitals in expansion states showed no significant relative gains in total patient revenues or operating margins compared to non-expansion states; non-expansion states saw relative gains in commercial insurance revenues.
  - Coverage and utilization increases:
    - 2014 provisions insured about 20 million low-to-middle income non-elderly adults (Garrett and Gangopadhyaya, 2016).
    - Increased demand for healthcare labor via additional job postings (Dillender, 2021).
    - Insurance coverage increased by about 4 percent more in Medicaid expansion states than in non-expansion states (Moghtaderi and others, 2020).
  - Literature review (Gruber and Sommers, 2019): strong evidence that ACA increased insurance coverage; positive effects on access and consumption of care; limited evidence on health outcomes; no evidence of significant reductions in provider access or large budgetary pressures on states through 2018.
- Empirical evidence on wages:
  - Descriptive: Between 2016 and 2019, average wages in Medicaid expansion states grew by about ¾ percent per year faster than in non-expansion states.
    - Seven states expanded Medicaid between January 2014 and end-2016.
    - Wage differentials tend to emerge within about three years (or less) after official implementation (implementation started in 2014).
  - Formal cross-section regression (wage growth over 2014-2019):
    - Y_i = α + β · expansi on_i + κ · X_i + ε_i
      - Y_i is wage growth over the five-year period 2014 to 2019 for state i.
      - Five-year window motivated by delayed wage response due to wage stickiness and offsetting employment effects.
- Interpretation:
  - Delayed wage response consistent with gradual adjustment of wages to demand shocks and possible offsetting employment gains in expansion states.

*Italic: Source: wpiea2021180-print-pdf — content unit 33 and surrounding sections.*

### 55.      We include a dummy variable 풆풆풙풙풆풆풆풆풆풆풆풆

### wpiea2021180-print-pdf - 55.      We include a dummy variable 풆풆퉁풙풙풆풆풆풆풆풆풆풆

### Methodology
- Key treatment variable
  - A dummy variable 풆풆퉁푒푥푒푒푒푒푒푒푒푒 (denoted in text) equals one for states that expanded Medicaid and zero otherwise; this dummy is the primary variable of interest.
- Cross-section and panel specifications
  - Cross-section regressions include control variables 푿푿풊풊: five-year average state-specific unemployment and inflation rates, five-year population growth, five-year average ratio of employment in the healthcare sector to population, and five-year average ratio of employment in the healthcare sector to total employment.
  - Difference-in-difference baseline (equation [5]):
    - Dependent variable: log hourly wage 푌푌풊풊,풕풕.
    - Controls: state fixed effects 훼훼풊풊, time fixed effects 훾훾풕풕, treatment dummy 푒푒푥푥푒푒푒푒푒푒푒푒풊풊풕풕, and 푿푿풊풊,풕풕 including inflation rate, unemployment rate, unemployment gap, population growth, share of medical employment in total employment, and share of medical employment to total population.
  - Leads-and-lags (equation [6]):
    - Event-time dummies 퐷퐷푖푡푗 index years relative to each state’s expansion; 훽훽0 is first expansion year effect, 훽훽1 second year, etc.; up to five lags are considered.
  - Distributed-lag in event time (equation [7]):
    - Cumulative treatment indicators 푒푒푥푥푒푒푒푒푒푒푒푒푖𝑡𝑘 equal one for expansion states k-year relative to the expansion year and all later years, covering up to five years.
  - Micro-level analysis (equation [8]):
    - Augmented Mincer framework using CPS: log hourly earnings 푌푌풊풊,풕풕 varies linearly with education and quadratically with job experience; controls include individual characteristics (race, sex, full-time work status, marital status) and macro variables (state-specific unemployment gap and log price levels).
    - Treatment and year dummies interact; coefficient 훾훾 on the interaction captures the policy effect.

### Data
- Sources and scope
  - CPS data from IMPUS CPS database (repeated cross sections).
  - Occupational Employment and Wage Statistics (OEWS): annual, by occupation and by state, back to 1997 with 2019 the latest year; includes wages, total employment, number of employees in occupation per 1000 employment.
  - Medicare wage adjustment indexes for metropolitan and micropolitan statistical areas from Centers for Medicare & Medicaid surveys (CMS).
  - Inflation: SAIRPD Implicit Regional Price Deflators by state from the Bureau of Economic Analysis (available starting in 2008).
  - Unemployment rate: Bureau of Labor Statistics local area unemployment statistics.
  - Population: Annual Estimates of the Resident Population, Census Bureau.
- Medicaid expansion timing
  - Kaiser Family Foundation documents status and timing (Table 9).
  - There are 39 states that have adopted or are expected to expand Medicaid as allowed by the ACA.
  - Some states expanded prior to 2014; some in January 2014; 13 states expanded later.

### Key empirical findings
- Aggregate and state-level wage impacts
  - Between 2014 to 2019, wage growth for healthcare practitioners in expansion states (accounting for about 60 percent of total healthcare practitioners in the U.S.) is about 0.5 to 0.6 percent higher per year than in non-expansion states.
  - Basic difference-in-difference regressions: hourly wages in expansion states grew about 0.7 percent faster on average in each of the expansion years relative to non-expansion states.
- Dynamic effects (leads, lags, and distributed lag)
  - Differential impact on wages starts to turn significant in the fourth year following the Medicaid expansion when plotting cumulative impact; relative-to-훽훽−3 plots show the differential starts in the third year.
  - Over the course of five years, the cumulative wage difference is about 2 percent.
  - Distributed lag model confirms wage impact starts to show up in the third year following expansion.
- Metropolitan-area level and microdata results
  - Results using Medicare wage adjustment indexes at metropolitan and micropolitan levels are consistent with state-level findings; wage differences turned statistically significant starting in 2012 (reflecting some pre-2014 expansions).
  - Individual-level CPS regressions: wages in expansion states in expansion years are about 4 to 6 percent higher on average (Table 13), though CPS top-coding may underestimate the full impact.
- Supply-side and wage-ratio evidence
  - Employment-to-population for healthcare practitioners increased in expansion states relative to non-expansion states, but not enough to offset demand-driven wage pressures (Figure 19).
  - Ratio of healthcare to non-healthcare wages rose in expansion states post-expansion; panel regressions with the wage ratio as dependent variable show a positive and statistically significant coefficient on the Medicaid expansion dummy (Table 14).
  - A higher ratio of healthcare employment to non-healthcare employment corresponds with a lower ratio of healthcare to non-healthcare wages.
- Labor mobility and other checks
  - Using Census Job-to-Job flow data, Medicaid expansion states do not show higher net inflows of “healthcare and social service workers” as a ratio of population.
  - Regression of inflow ratio on Medicaid expansion dummy and unemployment rate aligns with Figure 21 results.
  - Frictions such as declines in interstate labor mobility, state medical licensing requirements, and house price differentials likely limit perfect labor mobility.

### Interpretation and policy implications
- Mechanism and magnitude
  - Medicaid expansion increased demand for healthcare, producing statistically and economically significant wage increases for healthcare practitioners in expansion states relative to non-expansion states.
  - The observed wage increases may reflect increased provider market power when demand rises and supply is not sufficiently elastic.
- Supply-side considerations
  - Increasing supply elasticity could mitigate wage and price pressures from coverage expansions.
  - Evidence suggests that increasing healthcare sector employment-to-population could help reduce wage growth (i.e., larger supply dampens wage pressures).
- Pricing and revenue considerations
  - Although Medicaid expansion increased Medicaid revenues for hospitals, studies indicate no net gain in overall hospital revenues due to relative commercial revenue losses; higher provider wages may nonetheless be passed through to consumer prices with an elasticity greater than 1 (hospital markup growth is higher in expansion states; results available from authors upon request).
- Policy trade-offs
  - Policies that expand coverage and boost demand without supply-side reforms that reduce entry barriers could raise provider market power and increase prices in addition to direct coverage costs.
  - The findings point to benefits of making supply more elastic alongside coverage expansions.

### Conclusion (select points)
- Market power and markups
  - Market power in the U.S. healthcare sector has increased since the 1980s and contributed to rising healthcare costs; markups for publicly listed firms in the U.S. healthcare sector have almost doubled since the early 1980s.
  - Hospital markups have increased significantly (by more than 6 percent on average) since the late 1990s and vary across states.
  - Incorporating markups into OECD regressions: on average, markups account for up to a quarter of annual increases in per capita U.S. healthcare costs.
  - U.S. state-level regressions: hospital markups explain about 15 percent of variation in healthcare spending across states.
- Medicaid expansion wage effect summary
  - Medicaid expansion is associated with higher wages for healthcare practitioners in expansion states versus non-expansion states; dynamic analyses indicate effects start to materialize around the third to fourth year and accumulate to about 2 percent over five years at the state-level, with individual-level estimates of about 4 to 6 percent higher wages.

*Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021180-print-pdf.pdf*

### 74.      The significant contribution of market power to healthcare costs suggests the

### 74. The significant contribution of market power to healthcare costs suggests the need for carefully considered policy responses.

### Policy responses and recommendations
- Dilute barriers to entry and encourage entry by new health providers and insurers where entry barriers drive increased market power.
- Recalibrate licensing requirements or limits on the flow of new medical professionals when those measures have become increasingly binding constraints to entry.
- Adopt a more assertive approach to antitrust policies (at both the federal and state level) to identify and counter restraints of trade that unreasonably restrict competition.
- Consider actions to constrain the exercise of pricing power by providers and insurers in highly concentrated local geographic markets (National Academy of Social Insurance, 2015).
- Consider creating an agency responsible for monitoring and oversight of healthcare markets (Gaynor, 2020).
- Increase transparency to provide consumers with accurate and timely information about price, quality, costs, and provider networks to improve consumer choice and market competitiveness.
- Note policy context: The Executive Order on “Strengthening Medicaid and the Affordable Care Act” (January 28, 2021) provides a special enrollment window to the Marketplace for uninsured/under-insured Americans. The 2021 American Rescue Plan (ARP) fully paid for coverage for the lowest income workers, increased premium subsidies for those earning up to 400 percent of the federal poverty level, and capped the costs for a benchmark plan for two years while also providing incentives to encourage Medicaid expansions by states that have not done so. The proposed American Families Plan contains provisions to make the premium reductions of the ARP permanent.

### Market conduct issues highlighted
- Hospital practices such as bundling services where hospitals have greater market power with services where they have relatively less market power (“tying”) can extend market power and limit entry by competitors; such practices can be challenged under the Sherman Act and Clayton Act (National Academy of Social Insurance, 2015).
- Many local geographic markets are already highly concentrated, and antitrust cases are time and resource intensive; therefore direct constraints on pricing power merit consideration.

### Empirical evidence and econometric findings (selected)
- Panel cointegration tests (OECD data, variables in logs: Health expenditure p.c., Baumol variable, population over 65, real GDP p.c., markups and their interaction with Baumol variable):
  - Westerlund Panel means statistic: -2.1149; p-value: 0.0172
  - Pedroni - Phillips-Perron Panel means statistic: -2.1783; p-value: 0.0147
  - Kao- Augmented Dickey-Fuller Panel means statistic: -1.9924; p-value: 0.0232
- Fixed effects regressions on changes in insurer MLRs (Observations 400; Number of cid 50; Standard errors in parentheses):
  - Column (1): ∆HHI (Ind.) coefficient = -0.0315* (0.0167); Constant = -0.00162 (0.00339); R-squared = 0.010
  - Column (2): ∆HHI (small) coefficient = -0.0220*** (0.00773); Constant = -8.45e-05 (0.00109); R-squared = 0.023
  - Column (3): ∆HHI (large) coefficient = -0.00431 (0.00749); Constant = -0.000128 (0.000601); R-squared = 0.001
  - Significance codes: *** p<0.01, ** p<0.05, * p<0.1
- Pairwise correlations (selected):
  - Hospital markup with Insurer MLR (individual): -0.0723
  - Wage premium with Insurer MLR (individual): 0.0952*
  - Insurer MLR (small group) with Insurer MLR (individual): 0.5074*
  - Notes: *p<0.05
- Basic OECD regressions (all variables in first differences; dependent variable: ∆log of p.c. real health costs; Observations and R-squared vary by specification):
  - baumol coefficients across columns: 0.440*** (0.0670), 0.443*** (0.0722), 0.449*** (0.0765), 0.522*** (0.0940), 0.559*** (0.0953), 0.501*** (0.110)
  - dlrgdp_pc coefficients: 0.460*** (0.0852), 0.462*** (0.0828), 0.472*** (0.0865), 0.580*** (0.122), 0.763*** (0.122), 0.280** (0.0964)
  - Observations reported: 866, 866, 836, 370, 341, 333; R-squared: 0.331, 0.339, 0.344, 0.403, 0.420, 0.268
  - Notes: dl stands for the first difference in logs. Baumol defined as (wage-rate growth – labor-productivity growth) × 1/(share of Baumol sector in total employment).
- Log-level regressions (dependent variable: log of p.c. real health costs):
  - baumol_level coefficients: 0.0420** (0.0161), 0.0335 (0.0274), 0.0240 (0.0304), 0.0127 (0.0299)
  - lrgdp_pc coefficients: 0.971*** (0.0939), 0.994*** (0.0945), 1.011*** (0.0984), 1.075*** (0.117)
  - lmarkup coefficient (column 3): -0.00763 (0.116); lmarkup in alternative spec: 1.136* (0.494)
  - int_level (interaction) coefficient (col 4): 0.110** (0.0412)
  - Observations: 885, 337, 319, 319; R-squared: 0.971, 0.970, 0.971, 0.974
- CS-DL and FMOLS results (dependent variable: log of p.c. real health costs; Number of groups: 9; Observations vary):
  - lmarkup estimates: 4.106* (2.373), 5.603** (2.691), 5.603** (2.691), 3.505*** (0.780)
  - baumol_level estimates: 0.0675 (0.0891), 0.170* (0.0949), 0.170* (0.0949), 0.378*** (0.0334)
  - int_level estimates: 0.355* (0.201), 0.508** (0.246), 0.508** (0.246), 0.380*** (0.0874)
  - R-squared: 0.391, 0.352, 0.352, 0.996 (FMOLS)
- State-level regressions (dependent variable: per capita real healthcare expenditure; sample size and specifications reported):
  - dbaumol coefficients: 0.0441*** (0.00304), 0.0491*** (0.00280), 0.0772** (0.0362)
  - dlog_GSP_pc coefficient example: 0.400*** (0.0594); in IV/regression: 0.524*** (0.0704); CS-DL: 0.765** (0.318)
  - baumol_level (level regressions): 0.0559*** (0.00639), 0.0671*** (0.00964)
  - lmark_hospital coefficient (level regression, column reported): 0.471* (0.250); interaction lmark_hospital*baumol = 0.0528* (0.0308)
  - Observations commonly reported: 1,950; Number of cid: 50; R-squared examples: 0.624, 0.648, 0.615, 0.325, 0.946
  - IV regression uses state housing price indices as instrumental variable (see Bates and Santere, 2013). Specification (2) used to extract fixed effects.
- MSA regressions (dependent variable: per capita real healthcare expenditure; all variables in log-levels):
  - lconcen coefficients: 0.0900*** (0.0213) in FE; 0.0427** (0.0184) in CS-DL
  - lRGDP_pc coefficients: 0.351*** (0.0540), 0.122*** (0.0416), 0.448*** (0.0479)
  - baumol coefficients: 0.0530*** (0.0133), 0.0216** (0.00843), 0.0719*** (0.00541)
  - int (baumol*lconcen): 0.0102*** (0.00291), 0.00483** (0.00195)
  - Observations listed: 831, 761, 3,930; R-squared examples: 0.865, 0.944, 0.226
- Medicaid expansion and wage analyses:
  - List of states with Medicaid expansion source: Kaiser Family Foundation (2020), “Status of State Medicaid Expansion Decisions: Interactive Map”.
  - Cross-section regressions on hourly wage growth (2014-2019): exp_dum coefficients across columns: 2.776*** (0.959), 2.308** (0.971), 2.797*** (1.003), 2.653*** (0.952), 2.721*** (0.953); Observations: 51; R-squared range: 0.125–0.211
  - Difference-in-difference regressions (state-level log hourly wages): exp_dum coefficient = 0.007** (0.003) across multiple columns; Observations: 510; R-squared ~0.893–0.898; Year and State fixed effects included.
  - Area-level (MSA/micro) regressions: med_exp_dum coefficients = 0.026*** (0.002), 0.027*** (0.002), 0.027*** (0.002); Observations: 3,561; R-squared ~0.823–0.824; Number of areas: 376
  - Individual CPS regressions on log hourly wages: med_year_dum coefficients: 0.059** (0.025), 0.049* (0.025), 0.047* (0.025); med_dum coefficients vary (example: 0.165** (0.068) in one column); Observations: 5,945 (some specs 5,945, one with 5,945 and one with 5,945 reported); R-squared 0.146–0.166
  - Between-occupational wage ratios: Medicaid expansion coefficient = 0.007** (0.003) in several specifications; relative employment ratio coefficients around -2.670** (1.354) to -3.017** (1.366); Observations: 51 states; R-squared ~0.216–0.222
  - Job mobility regressions (net inflow ratios of “healthcare and social assistance” jobs): exp_dum coefficients range from -0.001 to -0.015 and -0.020 (standard errors reported); Observations: 501; R-squared up to 0.159

### Key empirical interpretations (as presented)
- The Baumol effect (defined here as (wage-rate growth – labor-productivity growth) × 1/(share of Baumol sector in total employment)) consistently enters positively and significantly in many specifications explaining growth or levels of per capita real health costs.
- Markups and their interaction with the Baumol variable are estimated to have statistically significant associations with health spending in multiple specifications (levels, CS-DL, FMOLS, state and MSA regressions).
- Medicaid expansion is associated with positive and statistically significant effects on healthcare practitioner wages in multiple specifications (cross-section, difference-in-differences, area-level, and individual-level analyses), with exp_dum or med_exp_dum coefficients consistently positive and often significant at conventional levels.
- Evidence across state, MSA, and individual-level analyses suggests links between market concentration, Baumol-type wage pressures, markups, and per capita health expenditure growth.

*Source: wpiea2021180-print-pdf - 74.      The significant contribution of market power to healthcare costs suggests the*

### REFERENCES

### REFERENCES

### Market power, pricing, and consolidation in health care
- Akcigit, U., W. Chen., F. J. Diez., R. A. Duval., P. Engler., J. Fan., C. Maggi., M. Mendes Tavares., D. A. Schwarz., I. Shibata., C. Villegas-Sánchez., 2021. “Rising Corporate Market Power: Emerging Policy Issues,” IMF SDN 21/01 (Washington: International Monetary Fund).
- Cicala, S., E.M.J. Lieber, and V. Marone, 2019. “Regulating Markups in U.S . Health Insurance,” American Economic Journal: Applied Economics, 11(4):71-104.
- Cooper, Z., S.V. Craig, M. Gaynor, J.Van Reenen, 2019. “The Price Ain’t Right? Hospital Prices and Health Spending on the Privately Insured,” The Quarterly Journal of Economics, Volume 134, Issue 1, February 2019, Pages 51–107.
- Dafny, Leemore, Kate Ho, Robin Lee. 2019. “The price effects of cross-market mergers: theory and evidence from the hospital industry.” The RAND Journal of Economics, 50 (2): 289–325.
- Dafny, Leemore, Nancy Beaulieu, Bruce Landon, Jesse Dalton, Ifedayo Kuye, and J. Michael McWilliams. 2020. “Changes in Quality of Care after Hospital Mergers and Acquisitions.” New England Journal of Medicine 384, 51–59.
- Gowrisankaran, G., A. Nevo, and R. Town. 2015. “Mergers When Prices Are Negotiated: Evidence from the Hospital Industry.” American Economic Review 105 (1): 172–203.
- Thompson, A., 2011. “The Effect of Hospital Mergers on Inpatient Prices: A Case Study of the New Hanover–Cape Fear Transaction.” International Journal of the Economics of Business, 18 (1): 91–101.
- Roberts E.T ., Chernew M.E ., and McWilliams J.M., 2017. “Market share matters: evidence of insurer and provider bargaining over prices.” Health Affairs (Millwood); 36(1):141.
- National Academy of Social Insurance, 2015. “Addressing Pricing Power in Health Care Markets: Principles and Policy Options to Strengthen and Shape Markets.” Final Report of the Academy’s Panel on Pricing Power in Health Care Markets, April 2015.
- Fulton, B.D., 2017. “Healthcare Market Concentration Trends In The United States: Evidence And Policy Responses” Health Affairs, vol. 36, No. 9.

### Health spending drivers, Baumol’s cost disease, and expenditure dynamics
- Baumol, W.J., 1967. “Macroeconomics of unbalanced growth: the anatomy of urban crisis.” American Economic Review, 57 (3) (1967), pp. 415-426.
- Bates, L.J., Santerre, R.E., 2013. “Does the U.S. healthcare sector suffer from Baumol's cost disease? Evidence from the 50 states.” Journal of Health Economics, Volume 32, Issue 2, March 2013, Pages 386-391.
- Hartwig, J., 2011. “Can Baumol's model of unbalanced growth contribute to explaining the secular rise in healthcare expenditure? An alternative test,” Applied Economics, 43 (2) (2011), pp. 173-184.
- Hartwig, J., 2008. “What drives healthcare expenditure? Baumol's model of unbalanced growth revisited,” Journal of Health Economics, 27 (2008), pp. 603-623.
- Colombier, C., 2017. “Drivers of Health-Care Expenditure: What Role Does Baumol’s Cost Disease Play?” Social Science Quarterly, Volume 98, Number 5, November 2017.
- Smith, S. D., J. P. Newhouse, and M. S. Freeland, 2009. “Income, Insurance, and Technology: Why Does Health Spending Outpace Economic Growth?” Health Affairs 28(5):1276–84.
- Newhouse, J., 1992. Medical care costs: How much welfare loss. Journal of Economic Perspectives 6, 3-21.
- Papanicolas, I., Woskie, L.R., and Jha, A.K., 2018. “Health Care Spending in the United States and Other High-Income Countries”, JAMA, 319 (10), pp. 1024-1039.
- Sisko, A. M., S. P. Keehan, J. A. Poisal, G. A. Cuckler, S. D. Smith, A. J. Madison, K. E. Rennie, and J. C. Hardesty, 2019, “National Health Expenditure Projections, 2018–27: Economic and Demographic Trends Drive Spending and Enrollment Growth,” Health Affairs 38 (3).

### Effects of the Affordable Care Act (ACA), Medicaid expansion, and insurance coverage
- Gruber, J., & Sommers, B., 2019. “The affordable care act's effects on patients, providers and the economy: What we've learned so far,” NBER.
- Garrett, B., & Gangopadhyaya, A., 2016. “Who gained health insurance coverage under the ACA, and where do they live.” The Urban Institute.
- Mazurenko, O., Balio, C. P., Agarwal, R., Carroll, A. E., & Menachemi, N., 2018. “The effects of Medicaid expansion under the ACA: A systematic review,” Health Affairs, 37(6), 944–950.
- Miller, S., & Wherry, L. R., 2017. “Health and access to care during the first 2 years of the ACA Medicaid Expansions,” New England Journal of Medicine, 376(10), 947–956.
- Moghtaderi, A., Pines, J., Zocchi, M. and Black, B., 2020. “The effect of ACA Medicaid expansion on hospital revenue,” Health Economics, Volume29, Issue12, December 2020, Pages 1682-1704.
- Rhodes, J., Buchmueller, T.C., Levy, H.G. & Nikpay, S.S., 2020. “The heterogenous effects of the ACA Medicaid expansion on hospital financial outcomes.” Contemporary Economic Policy, Volume 38, Issue 1.
- Dillender, M. 2021. “How did the ACA's Medicaid expansion affect the demand for healthcare workers? Evidence from Vacancy Postings”, Journal of Human Resources, Forthcoming.
- Dillender, M., Friedson, A., Gian C. & Simon K., 2019. “Does the healthcare educational market respond to short-run local demand?” Economics and Education Review, 73.
- Coutemanche, C., Friedson, A., Koller, A. & Rees, D., 2019. “The affordable care act and ambulance response times,” Journal of Health Economics, 67: 102213

### Workforce, labor mobility, and earnings in health sector
- Azzopardi, Damien, Fozan Fareed, Mikkel Hermansen, Patrick Lenain and Douglas Sutherland, 2020. “The decline in labor mobility in the United States: Insights from new administrative data,” OECD Economics Department Working Papers No. 16/2020.
- Glied, S., Ma, S. and Pearlstein I., 2015. “Understanding pay differentials among health professionals, nonprofessionals, and their counterparts in other sectors,” Health Affairs 34, No. 6 (2015) 925-935.
- Zhang, X., D. Lin, H. Pforsich, and Vernon W., 2020. “Physician workforce in the United States of America: forecasting nationwide shortages”, Human Resources Health. 2020; 18: 8.
- Baker, L.C , Bundorf M.K. , Royalty A.B. , Levin Z., 2014. “Physician practice competition and prices paid by private insurers for office visits”. JAMA 2014; 312 (16): 1653- 62.
- Mullangi, S., M. Agrawal, and K. Schulman, 2021. “The COVID-19 Pandemic—An Opportune Time to Update Medical Licensing,” Jama Network, https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2775345.

### Health, human capital, and socioeconomic outcomes
- Grossman, M., 1972. “On the concept of health capital and the demand for Health,” Journal of Political Economy 80, 223-255.
- Haas, S. A, Glymour M.M., and L.F. Berkman, 2011. “Childhood Health and Labor Market Inequality over the Life Course,” Journal of Health and Social Behavior, 52(3):298–313.
- Kane, J . B., 2015. “An Integrative Model of Inter-and Intragenerational Preconception Processes Influencing Birthweight in the United States.” Journal of Health and Social Behavior. 56(2):246–61.
- O’Brien, R.L., and C.L. Robertson, 2018. “Early-life Medicaid Coverage and Intergenerational Economic Mobility,” Journal of Health and Social Behavior. 59(2): 300–315.

### Empirical methods, panel econometrics, and macro implications
- Baltagi, B. H., Feng, Q. and Kao, C., 2016. “Estimation of heterogeneous panels with structural breaks,” Journal of Econometrics, vol. 191(1), pages 176-195.
- Baltagi, B. H., and F. Moscone, 2010. “Healthcare Expenditure and Income in the OECD Reconsidered: Evidence from Panel Data” IZA DP No. 4851.
- Pedroni, P., 1996. “Fully Modified OLS for Heterogeneous Cointegrated Panels and the Case of Purchasing Power Parity.” Working paper No. 96–020, Department of Economics, Indiana University.
- Pesaran, M. H., 2006. “Estimation and inference in large heterogenous panels with multifactor error structure”. Econometrica 74, 967-1012.
- Pesaran, M. H., 2007. “A simple panel unit root test in the presence of cross section dependence.” Journal of Applied Econometrics 22, 265-312.
- Carrion-i-Silvestre, J. L., 2005. “Healthcare expenditure and GDP: Are they broken stationary?” Journal of Health Economics 24, 839-854.
- Chou, W. L., 2007. “Explaining China’s regional health expenditures using Im-type unit root tests,” Journal of Health Economics 26, 682-698.
- Moscone, F., and E. Tosetti, 2010. “Health expenditure and income in the United States,” Health Economics, Volume 19, Issue 12, December 2010, Pages 1385-1403.
- Chudik A., Mohaddes K., Pesaran, M.H., and M. Raissi, 2017. “Is There a Debt-Threshold Effect on Output Growth?” The Review of Economics and Statistics, vol. 99(1), pages 135-150, March.
- Diez, F.J., Leigh, D., and Tambunlertchai, S., 2018. “Global Market Power and its Macroeconomic Implications,” IMF Working Paper WP/18/137 (Washington: International Monetary Fund).
- De Loecker, J., and J. Eeckhout, 2017. “The Rise of Market Power and the Macroeconomic Implications,” NBER Working Paper No. 23687.
- Woodridge J. (2012) Difference-in-Differences Estimation.

### Additional empirical and policy-relevant studies
- Baker, L.C., Bundorf, M.K., Royalty, A.B., Levin, Z., 2014. “Physician practice competition and prices paid by private insurers for office visits”. JAMA 2014; 312 (16): 1653-62.
- Nunn, R., J. Parsons and J. Shambaugh, 2020. “A dozen facts about the economics of the U.S. health-care system”, https://www.brookings.edu/research/a-dozen-facts-about-the-economics-of-the-u-  s-health-care-system/.
- Garrett, B., & Gangopadhyaya, A., 2016. “Who gained health insurance coverage under the ACA, and where do they live.” The Urban Institute.
- Sisko, A. M., et al., 2019. “National Health Expenditure Projections, 2018–27: Economic and Demographic Trends Drive Spending and Enrollment Growth,” Health Affairs 38 (3).

*Source: wpiea2021180-print-pdf - REFERENCES*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021180-print-pdf.pdf_