Certificate of need (CON) laws remain one of the most contentious public policy debates. Despite their inception nearly fifty years ago, ongoing questions about cost, access and the role of community healthcare remain. The purpose of this report is to provide a summary of the existing research on the topic, offer new data as states modernize their existing laws and provide a pathway for Georgia policymakers by dispelling misconceptions often cited in defense of CON.
CON laws were established as a means to control the rising costs of healthcare for consumers. Regulators were given the responsibility of determining healthcare needs. Notably, the enactment of these laws occurred during a period in which the government reimbursed providers based on how much they spent. As such, hospitals were incentivized to keep spending money by adding additional beds and new equipment.
Only a decade after passage, however, Congress realized the fallacy of this shift in policy and repealed the federal CON law. By requiring government approval, CON had effectively granted existing hospitals a competitor’s veto. While advancements in healthcare led to additional procedures being safely done in outpatient-based settings, competition and innovation were essentially restricted to what the incumbent provider chose to offer in most communities.
Elective surgery and imaging remain the most profitable service lines for hospitals, with health systems exerting their nonprofit status and the mandate to treat every patient regardless of ability to pay as justification for the preservation of CON. A Georgia without CON, its defenders claim, would result in the proliferation of ambulatory surgery centers (ASCs) and imaging facilities that would force the widespread closure of community hospitals.
CON laws were never intended to subsidize healthcare for the uninsured. There are local, state and federal reimbursement funds directly dedicated to offsetting these costs and ensuring that nonprofit hospitals remain financially viable. Despite the COVID-19 pandemic, many Georgia health systems retain profits in the hundreds of millions, especially in the wake of continued consolidation.
Most of the empirical literature we reviewed – well over 100 tests – found that CON laws do not contain cost, do not offer adequate and equitable access and do not provide quality improvement. Additionally, we did not find any correlation in the reduction of CON laws and an increase in rural hospital closures in the states we examined.
On October 4, 1988, U.S. Representative Roy Rowland took to the floor of Congress. His speech that day focused on the government regulation known as certificate of need (CON) and its “harmful impact” on a rural hospital in his congressional district. The regulation requires providers who wish to open or expand their facilities to first prove to a regulator that their community “needs” the service in question.
Rowland, a native of Wrightsville in eastern Georgia, was a family practice physician in nearby Dublin and a three-term member of the Georgia House of Representatives prior to his election to Congress in 1982. The congressman also had a personal connection to the issue, as the bill that established the federal CON law was what first inspired him to run for office. CON laws allow regulators to determine whether healthcare providers are allowed to open or expand facilities. As Rowland would later recall, “I felt the government was getting more involved in telling physicians and other medical people how to do their practices, how to take care of patients. So I decided I wanted to try and do something about it.”
However, by the time of his speech Congress had already eliminated the federal mandate requiring states to enforce CON laws in 1986. Rowland was now focused on ending the remaining state-based CON laws, specifically the one regulating healthcare providers in his home state. The congressman shared with his colleagues the story of how Putnam General Hospital in Eatonton sought to renovate its twenty-year old facility. The project was going to be financed by a local 1-cent sales tax that had been voted on and approved by the community. However, since there were capital improvements being made to the facility, the renovation required approval by Georgia’s health planning agency under the state’s CON law. After reviewing the proposed project, the state “looked over the request for the locally funded hospital improvements and decided to deny it — unless the hospital eliminated 10 beds.” The health planning agency determined there were too many hospital beds already in the area, despite the fact that Putnam General was not seeking to add additional beds as part of its renovation. Rowland went on to explain that not only would eliminating those ten beds not generate any significant healthcare cost savings, but the mandatory bed reduction would necessitate reducing the number of nurses being trained in the hospital’s licensed practical nurse (LPN) program. In addition, it would be much more costly if the hospital ever sought to regain those lost beds through the state’s CON program in the future.
“At first glance, [certificate of need] may have looked pretty good,” said Rowland. “In practice, however, the effect of certificate-of-need on healthcare costs has been dubious, at best. And the program has certainly been insensitive in many instances to the true needs of our communities… In my view, it’s a classic case of a bureaucracy paying more attention to numbers on a piece of paper than to reality. And the reality is the harmful impact this would have on the community without doing anything significant to cut costs.”
Unfortunately, the sentiment behind that speech from Rowland remains true 35 years later. For many Georgians, the state’s existing CON laws have limited their access to lower cost, higher quality healthcare services by providing entrenched incumbents with monopolistic control, often over entire counties and regions. Since then, Georgia communities spanning a wide range of socioeconomic statuses have suffered from the consequences of this regulation. In this study, we review the academic research on the efficacy of CON regulations and examine the arguments regarding access, cost containment and quality of care. We also assess the CON application process under the state’s existing structure and compare Georgia to other states with similar profiles that have reduced or repealed their CON laws. We focus, in particular, on the impact of the regulation on safety-net providers and their communities.
The Origin of and Rationale for CON Regulation
New York was the first state to establish a certificate of need program in healthcare in 1964. The biggest impetus for CON laws, however, came a decade later. In 1975, Congress passed, and President Ford signed the National Health Planning and Resources Development Act (NHPRDA). The NHPRDA threatened to withdraw federal healthcare funds from any state that refused to enact a CON program. Due to repeated postponement, that threat never actually materialized. Nevertheless, by the early 1980s, almost every state in the nation had adopted a CON program.
Then, as now, policy makers were worried about skyrocketing healthcare costs. In the run-on-sentences that characterize federal legislation, Congress lamented the “massive infusion of Federal funds into the existing healthcare system [that] has contributed to inflationary increases in the cost of healthcare and failed to produce an adequate supply or distribution of health resources, and consequently has not made possible equal access for everyone to such resources.”
Federal lawmakers held the common belief that healthcare was different. Medical services and technologies can be confusing. Patients are typically not experts and are often making once-in-a-lifetime decisions. And—as lawmakers knew well—someone else usually picks up the tab. So, they reasoned, patients might get suckered into agreeing to expensive hospital stays and unneeded procedures.
And there was some evidence for this. In 1959, UCLA health researcher Milton Roemer co-authored a study reporting a positive correlation between the number of hospital beds available per capita and the number of used hospital days per capita. The finding became known as “Roemer’s Law” and was shortened to the pithy characterization that “in an insured population, a hospital bed built is a hospital bed filled.”
In encouraging CON, lawmakers hoped hospitals would build fewer beds, fill them with fewer patients and spend less money. The main purpose of CON, therefore, was to reduce healthcare expenditures by rationing care. The authors of the NHPRDA also sought to reduce healthcare costs by encouraging “the use of appropriate alternative levels of healthcare, and for the substitution of ambulatory and intermediate care.” Beyond costs and expenditures, the authors of the NHPRDA hoped to ensure an adequate supply of care, especially for “underserved populations” including “those which are located in rural or economically depressed areas.” Finally, they hoped to “achieve needed improvements in the quality of health services.”
These goals—cost containment, adequate and equitable access and quality improvement—remain widely shared aims of health policy. And, as we will see, most of the empirical literature on CON tests whether the regulation serves these goals. Most of the literature finds it does not.
Why Assessing Healthcare Need is Difficult
Unlike other varieties of regulation, the CON process does not typically include an assessment of a provider’s qualifications. Nor do regulators appraise the adequacy of a provider’s facility or its safety record. Instead, regulators are charged with determining whether the community “needs” the service the provider hopes to offer. This is an unusual remit for a regulator. In the vast majority of markets, need is assessed by the service providers themselves, based on their expectation of profitability.
Several factors complicate the regulator’s task:
- First, compared to providers, regulators typically have less local knowledge about the community, its tastes, its culture and its economic situation.
- Second, as economic theory teaches, value is subjective. Consumers don’t just care about the technical attributes of care. They also care about the convenience of care, the modality of care and the cultural sensitivity of care. Different consumers will have subjective preferences around these factors.
Health care consumers may feel, for example, that they need a provider who understands their particular cultural, linguistic or religious needs. But providers who cater to a specific cultural or religious community have often been denied certificates of need because regulatory formulas make no room for such considerations.
- Third, while providers risking their own capital or borrowed capital have a strong incentive to accurately assess the viability of a project, public regulators have no such skin in the game.
- Fourth, the formulas on which regulators rely create perverse incentives. For example, if an existing provider knows that his potential competitor is likely to be denied a certificate of need if his facility is under-utilized, he then has an incentive to make sure his facility is under-utilized. So, ironically, the formula encourages providers to acquire equipment and then not use it, undermining efficiency rather than enhancing it.
- By far the most significant problem with regulatory needs assessment is the fact that it can so easily be used for anti-competitive purposes. In most CON states–including Georgia–a certificate can be denied if the regulator finds the new service will “duplicate” an existing service. This virtually guarantees a local monopoly, especially considering that in most CON states –including Georgia– incumbent providers are allowed to sit on the board that makes the decision. This is why the regulations are sometimes called “competitor’s veto laws.” And in most CON states–including Georgia–incumbent providers are allowed to be a part of the process, challenging competitors’ applications and in some cases even appealing decisions that they do not like. As a remarkable indication of the anticompetitive nature of CON regulations, competitors’ objections are often dropped and CONs are subsequently granted once applicants agree not to directly compete with incumbent providers.
These anticompetitive features help explain why antitrust authorities at the Federal Trade Commission and at the Department of Justice have taken the position for decades that CON laws are anticompetitive.
The States as Laboratories
In his dissent in New State Ice Co. v. Liebmann, Justice Louis D. Brandeis famously extolled what he called “one of the happy incidents of the federal system.” Under federalism, he declared, “a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.” New State Ice was a certificate of need case.
And due to its unique history, CON has proven to be an ideal topic for empirical study. By the mid-1980s the evidence was beginning to mount that certificate of need laws in healthcare did not achieve their stated goals. At the same time, observers on the left and the right were coming to appreciate that regulations often protect politically powerful incumbents from competition and harm consumers. It was in this context that Congress repealed the CON mandate in 1986.
Within three years, a total of twelve states had repealed their CON programs. Over the decades that followed, a handful of others would follow suit, and still more significantly pared their programs back. The most-recent full repeal was in New Hampshire in 2016. In 2019, Florida eliminated its CON requirement for hospitals, ASCs and most other services. And in 2021, Montana eliminated all of its CON requirements except that for long-term care facilities. Today, 3-in-10 Americans live in a state with no CON regulation in healthcare and 4-in-10 Americans live in a state with either no or only one CON requirement.
CON and non-CON states can be found in all regions of the country, they include both high- and low-income states and they include a wide variety of demographic and cultural populations. Thus, researchers can compare outcomes in CON and non-CON states to see what happens in states that have repealed or relaxed these regulations. Using modern econometric methods, researchers can control for factors such as economic and demographic differences that might also affect outcomes of interest.
In total, we have identified 94 peer-review studies assessing the effect of CON regulation. Since many studies include more than one test, these studies encompass well over 100 tests of certificate of need in healthcare. We summarize this literature in the final section.
CON Regulation in Georgia
The Legislative History of CON in Georgia
The framework for Georgia’s current CON system has been in place since 1975, when the state began reviewing new healthcare projects in accordance with the National Health Planning and Resources Development Act of 1974. Georgia’s CON program was established as state law by the General Assembly in 1979. Notably, competitors were not allowed to challenge certificate of need applications at the state level until the law was amended in 1983. Since then, attempts to reform, modernize and even repeal the program have been met with varying degrees of success.
In 1991, as healthcare technology and quality evolved, Georgia’s CON law was amended to allow for the emergence of outpatient surgeries in standalone ambulatory surgery centers, or ASCs. A less strenuous path that did not require a CON was established for single-specialty, physician owned ASCs to obtain a Letter of Non-Reviewability (LNR) from the state if their capital expenditures did not exceed $1 million.
In 2005, Governor Sonny Perdue established the Georgia Commission on the Efficacy of the CON Program. Georgia State University published a report for the commission that analyzed the effect that CON regulations had on the quality and cost of healthcare in Georgia and 10 other states.
The authors used a cross-border design to control for unobservable factors. They also used interviews and public information to develop an index measuring CON rigor based on fees, administrative requirements, reviewability, appeals and administrative complexity. They assess the effects of CON on acute care, long term care and home health markets, finding:
- CON is associated with higher private inpatient acute care costs.
- Acute care costs rise with the rigor of the CON program for the most resource-intensive acute care diagnoses.
- Some evidence that CON is associated with higher Medicaid costs for home health services.
- There is weak evidence that CON is associated with higher private long term care costs.
- There is weak evidence that CON is associated with higher Medicaid long term care costs.
- Some evidence that CON is associated with higher per-capita costs for home health services.
- CON is associated with fewer hospitals.
- CON is associated with fewer hospital beds.
- CON is associated with fewer home health agencies per 1,000 residents.
- CON is associated with fewer Medicare beneficiaries receiving home health services.
- There is no significant relationship between the percent of hospital admissions that are self-pay, though when controlling for the number of uninsured and family income, CON is positively related to self-pay admission per uninsured.
- There is no apparent difference in acute care quality in CON and non-CON markets.
- In long term care, CON is associated with better quality on two measures but worse quality on six measures.
- In home health markets, they find no evidence that CON affects any of 10 outcome measures of quality.
- Acute care markets are less competitive when CON is rigorous.
- CON is associated with lower levels of competition in home health agency markets.
As the Commission completed its study of CON, the Georgia House also established the Special Committee on Certificate of Need. This focus on CON led to the passage of significant reform in 2008. The bill was notable for its creation of “destination cancer hospitals” in the state code to allow for the opening of the Cancer Treatment Centers of America facility in Newnan, and the subsequent 35 percent cap on Georgia residents as patients as part of the definition of destination cancer hospital. The legislation also exempted certain non-medical expenditures such as parking decks and medical office buildings from CON requirements, and hospitals and physician practices were exempted from acquiring a CON for MRI and CT investments under $1 million. The LNR process for new physician-owned single-specialty ASCs was amended to require commitments for indigent, charity care, and Medicaid patients served, while the cap on capital expenditures under the LNR process was increased from $1.7 million to $2.5 million.
In 2018, concurrent efforts by the Georgia House and Senate studied the issue of CON and its impact on Georgia’s healthcare and economic landscape. The Georgia Senate Study Committee on Certificate of Need Reform recommended exempting all imaging and diagnostic equipment (except PET scans) from the state’s CON requirement, along with all mental health, psychiatric and substance abuse services, and allowing for multispecialty ASCs to open under the LNR process. The Georgia House Rural Development Council recommended replacing CON with a “rigorous accreditation and licensing requirement for new providers.” New healthcare providers in metro Atlanta would be exempted from the licensure requirement once accredited, and non-metro providers would require a state license if located within a 20-mile radius of an existing provider. The establishment of indigent and charity care requirements for non-profit, for-profit and specialty hospitals was also recommended, based on a rolling two-to three-year average of the state’s hospitals.
This led to renewed legislative efforts in 2019, including a bill that would have repealed CON requirements with the exception of long-term care facilities that was voted out of the House Special Committee on Access to Quality Healthcare. A special healthcare license and exemption process for certain facilities and services would have replaced the CON process, and diagnostic imaging would have been exempted from this process entirely. A pared back version of this bill made it to the House floor, but was defeated by a vote of 72-94. A bill ultimately passed that session and was signed into law that increased the threshold amounts for capital expenditures and diagnostic equipment, introduced a limitation on facilities that can oppose a CON application to within a 35-mile radius of the proposed project, established CON requirements for freestanding emergency departments, allowed CTCA to pursue a “general cancer hospital” designation through the CON process and revised the LNR process for the addition of imaging equipment, requiring that the physician be on-site at least 75 percent of the time the equipment is in use. The legislature also passed companion legislation in 2019 to increase hospital disclosures given the increase of nonprofit entities amassing cash and assets in their communities. The law requires hospitals to publicly provide audited financial statements, real estate holdings, ownership in any subsidiaries or captive insurance companies, patient debt collection practices, community benefit reports and salaries of the 10 highest paid administrators.
During the 2022 legislative session, the House Special Committee on Access to Quality Healthcare advanced a bill that would have repealed CON within two years and replaced it with a licensing process with oversight of healthcare facilities’ indigent care requirements. It would have also created a path for multispecialty physician-owned surgery centers to open immediately upon signing, along with stricter oversight of hospital authorities. The bill, however, was never brought up for a floor vote in the Georgia House, and subsequent attempts to attach it as an amendment to healthcare legislation in the Senate did not pass out of committee. A bill which would have repealed CON for all rural counties in Georgia also failed to make it out of committee in 2022, and similarly failed as an amendment to Senate legislation.
The Current CON Process
Under current law, the application fee for a CON in Georgia ranges from $1,000 – $50,000. Once the office of Health Planning deems an application complete, the agency must complete its review and issue a decision within 120 days. While most CON applications can be submitted at any time, applications for skilled nursing facilities, intermediate care facilities and home health agencies can only be submitted when the Department of Community Health determines there is an unmet need. This process is known as batching and healthcare providers are dependent upon the state determining there is a need for new facilities before the process to move forward can even begin.
In Georgia, a CON is required if hospitals or destination cancer hospitals wish to increase beds. In the non-hospital setting, a CON is required for increasing beds in skilled nursing facilities, intermediate care facilities, personal care homes, ASC’s, obstetrical facilities, freestanding emergency departments, and diagnostic treatment or rehabilitation centers. A CON is also required for the construction, development, expansion, or relocation of hospitals, special care units, skilled nursing facilities, intermediate care facilities, personal care homes, ASC’s, obstetrical facilities, freestanding emergency departments, health maintenance organizations, and diagnostic, treatment, or rehabilitation centers. Healthcare services that require a CON include imaging, biliary, lithotripsy, surgery, intensive care, coronary care, pediatrics, gynecology, obstetrics, general, medical care, medical surgical care, and patient, nursing, cardiac catheterization, open heart, surgery, inpatient rehabilitation, alcohol or drug abuse, services, and mental health services.
As of July 1, 2022, CON spending thresholds are in effect for all capital expenditures above $11.5 million (even if no other specific CON applies); single-specialty physician-owned ASCs above $3.7 million and joint venture ASCs above $7.4 million; equipment acquisition, excluding PET services, above $2.9 million; and equipment repair or replacement above $860,000.
Providers who wish to challenge the state’s CON decisions, whether in support of or in opposition to the proposed project, have the option to pursue administrative and legal challenges. As our review of Georgia’s CON applications from 2017 – 2022 reflects, this has often meant years added to the process for projects to move forward. While the legality of specific CON decisions has been challenged in the judicial system, the constitutionality of CON itself has not been overturned in the state, despite one case making it as far as the Georgia Supreme Court in 2017.
However, in a recent court decision contesting the opening of a Level II NICU at Cartersville Medical Center, Judge Stephen Dilliard of the Georgia Court of Appeals questioned the constitutionality of CON given how it infringes upon the due process and equal protection rights of healthcare providers attempting to enter a market. “I strongly encourage the General Assembly to revisit and carefully reexamine the efficacy and constitutionality of the State Planning and Development Act.” Judge Dilliard then closes his legal argument by making an economic observation: “One thing is for certain: Georgians don’t benefit from a system that props up health care monopolies.”
Georgia CON Applications
For this study, we analyzed each CON application the Department of Community Health (DCH) received from 2017 – 2022. Applications that were submitted under Governor Kemp’s suspension of CON via executive order during the COVID-19 pandemic – which required a different submission process – are addressed in the following section.
During the six-year period that we reviewed, 379 CON applications were filed with the state and cataloged in the department’s online repository. As of publication, 43 of the applications were still waiting for the initial decision on their CON by the state.
Our analysis showed the following:
- When a competitor objects to an application, the odds of denial more than double from 20 percent to about 50 percent..
- Every additional party opposed to the application increases the odds of denial by about 11 percent.
- Any opposition to an application adds 234 days to the wait time for a decision, but competitor opposition adds about 520 days to the wait time
- Each additional party opposed to the application adds another 129 days.
- The cost of the project is not statistically significantly related to either approval or wait times.
- There don’t appear to be any statistically significant trends over time in approval or wait time.
However, even this analysis does not present a comprehensive picture of the state of denial rate in Georgia’s CON program. While many proponents of the current system will offer that it is working as intended and that providers should just “file for a CON”, this often does not represent a realistic route for most applicants that are not health systems or hospitals. Many smaller physician offices choose to forgo equipment purchases or upgrades out of an acceptance they will not be able to compete with larger competitors willing to spend time and money on exacting regulatory barriers and legal appeals.
In 2020, Georgia was one of 24 states that suspended or reduced their CON laws to expand healthcare services in the midst of the COVID-19 pandemic. Governor Kemp’s Executive Order “authorized and directed” the Department of Community Health “to implement the suspension of [certificate of need] where such suspension would permit capable facilities to expand capacity, offer services or make expenditures necessary to assist with the needs of this Public Health State of Emergency.” Healthcare providers were still required to submit a CON application despite this suspension. In the first two months, 32 applications under the COVID exemption were submitted to DCH, including hospitals, ambulatory surgery centers, home health and rehab centers.
As nearly every hospital discontinued elective surgeries during those early months, 14 ASCs applied for a special CON exemption that would have allowed them to convert to a multispecialty ASC during the public health emergency. This would have provided physicians with the opportunity to continually serve their patients with non-COVID related issues and offer surgery during the initial peak of the pandemic. Despite the governor’s order, DCH did not approve a single application by an ASC. In each denial, the department’s pro forma response cited a subsequent executive order calling for healthcare providers to “begin treating patients as soon as practicable.” The department also noted that the hospitals in proximity had not requested the ASCs to treat the patients they were unable to serve. The delay and judgment in processing these applications during a public health emergency does not reflect positively on the ability of regulators to determine a community’s need.
CON Regulations and Safety-Net Hospitals
In 2007, Mark Botti, Chief of the Litigation I Section of the Antitrust Division with the U.S Department of Justice, testified before a Joint Session featuring members of the Georgia Senate Health and Human Services Committee and the Georgia House Special Committee on CON. Botti testified that his work and that of his colleagues not only reinforced the importance of competition in the healthcare industry, but demonstrated how regulatory barriers to entry harm consumers. The heart of Botti’s testimony focused on four critical factors undermining the rationale behind CON laws:
1. The original cost-control reasons for CON no longer apply since the federal government no longer reimburses on a “cost-plus basis” that incentivized capital expenditures by hospitals.
2. Protecting the revenues of incumbent providers does not justify CON laws. CON laws were never intended as a means of cross-subsidizing care for the indigent by protecting profitable service lines from competition, such as surgery and imaging. We further examine this argument and effect of CON repeal in other states on safety-net providers in the following section.
3. CON laws facilitate anti-competitive behavior by allowing incumbent providers the opportunity to delay new facilities and service lines into the marketplace by allowing them to utilize the appeals process even when a need has been determined by the state.
4. CON laws lead to less competition and higher prices.
Perhaps the most common argument in defense of CON is the protection of financially struggling hospitals, which are predominantly located in rural areas. Rural hospitals are sustained – outside of government subsidies – by elective surgeries and imaging, the profitable areas of a hospital. As such, the theory is that CON is needed in order to prevent ambulatory surgery centers and standalone imaging centers from cannibalizing the profit centers of rural hospitals, resulting in significant closures and lack of access to rural health care. Fortunately, we can examine how states with similar geographic profiles have fared after repealing or significantly reducing CON laws.
Perhaps the most informative state is Florida, given their significant reduction of CON laws in recent years. In 2019, Florida repealed the CON requirement for hospitals, complex medical rehabilitation beds and tertiary hospital services, including neonatal intensive care units and organ transplant centers. The repeal of CON for specialty hospitals (which focus on specific services for defined age ranges) was included in the same legislation, but did not go into effect until 2021. Florida still has CON requirements for nursing homes, skilled nursing facilities, hospice programs and intermediate care facilities.
Utilizing data provided by Becker’s ASC Review we can examine the location of new ambulatory surgery centers in Florida since the elimination of that CON requirement. From July 2019 – November 2022, 65 new or expanded ASCs were announced in Florida. Of those 65 announced ASCs, 64 are located in counties designated “Urban” by the Florida State Office of Rural Health. Notably, the one ASC development announced in a rural county was a facility proposed by the existing community hospital. A subsequent announcement by a physicians group planning to open a competing ASC led to pronouncements that it would need to be a joint venture, otherwise it would result in the closure of the local hospital completely. Notably absent from the public debate was the increased access and choice for this particular community.
Since Florida’s CON repeal in 2019, the Sheps Center for Health Services Research at the University of North Carolina at Chapel Hill reports three rural hospitals have closed in the state. However, of those three closures, two were converted to freestanding emergency departments, preserving emergency access for those communities.
The chart below compares Georgia to other states with approximate populations and rural proportions, along with each state’s rural hospital closures since January 2005 and an overview of their CON restrictions for hospitals and ambulatory surgery centers. Notably, the two states with the most rural hospital closures, North Carolina and Georgia, ranked second and sixth respectively among the strictest CON regulations in a national study by the Mercatus Center in 2020. To date, no researchers have found any correlation between rural hospital closures and reduced CON regulation. In fact, as the academic literature further details below, researchers find that, controlling for other possibly-confounding factors, there are 30 percent fewer rural hospitals in states with CON restrictions relative to non-CON states, and there are 13 percent fewer rural ASCs in CON states relative to non-CON states.
|U.S. Pop. Rank||State||2020 Pop.||Rural Pop.||% Rural||Rural Hospital Closures since Jan. 2005||CON for Hospitals & ASCs|
Summary of the Academic Research on CON
As discussed above, the original rationales for CON in the NHPRDA were to contain cost or spending, to ensure adequate and equitable access to care, to improve the quality of care and to ensure care for underserved populations. These remain the most commonly-stated goals for healthcare regulation and most of the research revolves around them. We therefore summarize the research in terms of these questions.
Let’s begin with spending. To date, we have reviewed 42 peer-reviewed papers assessing the effect of CON on spending and these papers contain a total of 77 separate tests.
There are three different ways that the literature has tackled the issue of CON and spending: spending per service ($/Q), spending per capita ($/capita) and efficiency (output/input). We will discuss each in turn.
a) Spending Per Service ($/Q)
In our judgment, spending per service ($/Q) is the most intuitive way to think about spending. In these tests, researchers assess the effect of CON on charges, reimbursements, prices or per-unit costs. The key is that these tests are looking at spending per service rendered. This is an intuitive way to think about spending because it is analogous to a market price, which is always expressed in per-unit or per-service terms (think of the price of a gallon of gas or of one dental cleaning). Putting spending in per-unit terms is helpful because we typically want to know how much we spend relative to some service rendered (think of the cost of a well-child checkup or of a knee replacement).
Standard economic theory offers two reasons to suppose that CON regulation might increase spending per service and no reasons to suppose that it will decrease it. First, CON is a supply restriction. As economists Jon Ford and David Kasserman explained nearly three decades ago, “the economic effect [of a CON] is to shift the supply curve of the affected service back to the left,” and “the effect of such supply shifts is to raise… [the] equilibrium price.” Second, because of its anti-competitive properties, CON seems likely to lead to local monopoly pricing.
The empirical literature on CON and spending per service, summarized in Figure 1, supports the standard economic theory. Among 37 tests assessing the effect of CON on spending per service, 26 find that CON is associated with more spending per service, nine find insignificant or negligible effects and just two find CON is associated with lower spending per service. For every one test that finds CON is associated with lower spending per service, there are more than ten that find it is associated with higher spending per service.
In one study, researchers found that reimbursements for coronary artery bypass grafts fell by about 9 percent in Pennsylvania and by about 3 percent in Ohio following CON repeal. A different study found hospital charges in non-CON states were 5.5 percent lower five years after repeal, and another found CON is associated with higher prices across 11 different procedures. Medicare reimbursements for total knee arthroplasty are 5 to 10 percent lower in non-CON states than in CON states, and spinal surgery reimbursements fall faster in non-CON states than in CON states (about 11 percent per year).
Among the two tests that found CON was associated with lower spending per service, one did not control for any possibly confounding factors and the other reported mixed results, finding CON to be associated with higher reimbursements for cervical discectomy in the inpatient setting but lower reimbursements in the outpatient setting.
In summary, both standard economic theory and the balance of available empirical evidence suggest that CON laws, by restricting supply, tend to raise the cost per service rendered.
b) Spending Per Capita ($/Capita)
Another way to think about spending is in per capita terms ($/capita). These studies assess the effect of CON on spending per patient or per person. The problem with this measure is that, unlike spending per service, it is not obvious that less spending per capita is a good thing. After all, an extremely stringent CON that ensured that there were no healthcare resources at all would result in $0 spending capita. But given that we tend to think of healthcare as a “good” and not a “bad” this would not improve wellbeing.
Theoretically, CON might increase or decrease spending per capita because it has two offsetting effects on a market. On one hand, it tends to increase spending per service rendered (see the previous section), and on the other, it tends to decrease the amount of services rendered (which we will discuss in the next section). Total spending per person might therefore increase or decrease, depending on which effect dominates. Given that consumers of healthcare are relatively price-insensitive, however, the most likely effect of CON is to increase spending per capita.
Once again, the empirical evidence supports standard economic theory. We have evaluated 19 papers which together contain 31 separate tests of the effect of CON on spending per capita. This literature is summarized in Figure 2. Among these 31 tests, 17 find that CON is associated with more spending per capita, 10 find insignificant or negligible results, and four tests find that CON is associated with less spending per capita. For every one test finding CON is associated with lower spending per capita, there are more than four that find it is associated with higher spending per capita.
In one study, researchers found that CON was associated with 20.6 percent higher hospital expenditures per capita. Others found hospital expenditures per admission are higher in CON than in non-CON states. And states that eliminate CON experience 5 percent reductions in real per capita healthcare spending.
State policy makers often worry that eliminating CON will cause Medicaid expenditures to skyrocket. If anything, it appears that CON causes states to spend more, not less, on Medicaid. One study, for example, found that CON is associated with higher per capita Medicaid community-based care expenditures. Another study found mixed results, but to our knowledge, no one has found clear evidence that CON increases Medicaid per capita spending.
In summary, while CON might reduce spending per person by rationing care, standard economic theory predicts that it is more likely to increase spending per person through its effect on spending per service. The balance of evidence supports this hypothesis.
c) Output per Input (Output/Input)
The final way that researchers have assessed the effect of CON on spending is by examining output per input. These studies look at whether inputs such as labor or capital are more intensely used in CON than in non-CON states. Like spending per capita, this measure is not an especially helpful gauge of human wellbeing, but it does give us a sense of how CON affects technical efficiency.
In contrast with the other two measures of spending, economic theory offers no clear predictions about how CON might affect output per input. On the one hand, it might increase output per input if it results in more services rendered by fewer providers. Since this will cause a more intense use of labor and capital by these providers, it will tend to increase output per input. On the other hand, CON might decrease output per input if the anticompetitive effects of the regulation make providers less attentive to efficiency.
In all, eight studies have assessed the effects of CON on output per input and together these studies contain 9 tests. Figure 3 summarizes the results. Four tests find that CON increases output per input, two find insignificant or negligible results, and three find that CON reduces output per input.
To give the reader a better sense of these results, let us consider one study that found CON to be associated with higher output per input. This study happens to have been co-authored by one of the authors of the present study. In this paper, Mitchell and Stratmann found that states that require a CON for hospital beds had 12 percent higher bed-utilization rates during the COVID-19 pandemic. One might interpret this as a “good” result since it means that each bed is able to serve more patients, yielding higher output per input. Unfortunately, the authors also found that hospitals in these states were 27 percent more likely to run out of beds during the pandemic. This underscores the fact that output per input is not an especially relevant measure of patient wellbeing. Given its frequent use as a metric, we thought it only fair to include the result here.
d) Summarizing the Spending and CON Literature
In all, 42 papers with 77 empirical tests assess the effect of CON on spending, either by looking at spending per service, spending per capita or output per input. Figure 4 summarizes this literature. In total, 46 tests find that CON is associated with “bad” spending outcomes, 21 tests find negligible or inconclusive results and 10 tests find CON is associated with “good” spending outcomes.
Thus, for every one test that finds CON is associated with a “good” spending outcome, there are more than four that find it is associated with a “bad” spending outcome. Once again, we emphasize that the spending tests that are most relevant for human wellbeing–those that assess the effect of CON on spending per service–are especially lopsided with more than ten “bad” results for every one “good” result.
With 58 papers and 132 separate tests, access is the most-studied aspect of CON laws. Broadly speaking, the literature has assessed the effects of CON on patient access to healthcare in two ways. Some tests look to see if CON regulation has any relationship with the availability of services while others see if it has any relationship to the utilization of these services. We take each in turn.
a) Availability of Care
By design, CON regulations limit the supply of technology and investment. It seems intuitive, then, that they are likely to reduce the availability of services. And that is what the bulk of the literature finds. In total, 35 papers containing 72 tests have assessed the effects of CON on the availability of services. These tests measure availability in different ways. One technique is to count the number of service providers per capita. Another is to count units of medical technology per capita. Some papers measure how far patients must travel to obtain care or how long patients must wait until they can be served.
Figure 5 summarizes this literature. Of the 72 tests assessing the effect of CON on the availability of services, 59 tests find that CON is associated with diminished availability, nine find negligible or insignificant results, and four find that CON is associated with greater access to certain services.
Controlling for other possibly-confounding factors, researchers find that the average patient in a CON state has access to:
- 30 to 48 percent fewer hospitals;
- 14 percent fewer ambulatory surgery centers (ASCs);
- 3 percent fewer rural hospitals;
- 13 percent fewer rural ASCs;
- 25 percent fewer open-heart surgery programs;
- 46 percent fewer facilities offering coronary artery bypass graft (CABG);
- 20 percent fewer psychiatric care facilities;
- 50 percent fewer home health agencies;
- fewer hospitals offering revascularization;
- fewer dialysis clinics;
- fewer hospitals per cancer incident;
- fewer neonatal intensive care units (NICU); and
- fewer alcohol and drug abuse facilities.
Patients in CON states have access to fewer medical imaging devices and fewer hospital beds. They face longer wait times, must typically travel farther to obtain care and are more likely to leave their states for care. As mentioned in the previous section, hospitals in states with bed CONs were 27 percent more likely to run out of beds during COVID.
Among the four positive results, one was interpreted by its author as a negative result. In one of the earliest CON studies, Fred Hellinger found that hospitals anticipated the introduction of CON and undertook investments before it went into effect. Thus, in his interpretation, it backfired because it encouraged the supply of capital that it was supposed to discourage. The other tests that found a positive association between CON and availability of services also found negative associations in other circumstances. One study found that in states that had expanded Medicaid, CON laws were associated with more non-profit substance abuse facilities, while in non-expansion states CON laws were associated with fewer non-profit substance abuse facilities. Another study found that CON was associated with shorter travel to radiation oncology in some parts of the country but prolonged travel to radiation oncology in other parts.
In summary, one of the most common ways to evaluate CON is to compare its status to the availability of services. That is, does it live up to its promise to ensure an adequate supply of healthcare services? Not only is this one of the most-studied aspects of CON but it is also an area with some of the most lopsided results. CON was associated with diminished availability of services in 82 percent of all tests. For every test that finds CON associated with greater availability, there are nearly 15 that find it to be associated with less availability.
b) Utilization of Service
Though there is abundant evidence that CON makes it more difficult to obtain care, this doesn’t necessarily mean that it will lead to diminished utilization of services. Most healthcare services are inelastically demanded, meaning patients will still seek care even if it is difficult or inconvenient to obtain. Moreover, CON laws might actually increase the utilization of certain services by suppressing the utilization of substitute services. For example, if two procedures can be used to treat an ailment and if CON applies to procedure A but not to procedure B, or if CON is more stringently applied to A than B, then we can expect CON to suppress the utilization of A while possibly enhancing the utilization of B.
Twenty-three papers, which together include 60 tests, assess the effect of CON on utilization of healthcare services. Of these 60 tests, 40 find no significant relationship between CON and utilization of services, 14 find CON is associated with less utilization of services, and 6 find that CON is associated with greater utilization of services.
Among the tests finding CON to be associated with diminished utilization, one study found that CON states have 13.7 percent fewer home health admissions from hospitals than non-CON states. Another found that in CON states, hospitals were 5.35 percent less likely to accept psychiatric patients on Medicare and that there were about 56 percent fewer psychiatric clients per capita. In CON states, patients are less likely to obtain medical imaging, less likely to obtain total hip arthroplasty,
Among the tests finding CON to be associated with increased utilization, one study found that CON is associated with greater growth in intensity modulated radiation therapy, an expensive and no-more effective treatment than alternatives, so the authors interpreted this as a negative result. Similarly, another study found that CON made radiation therapy more likely to be used on elderly patients who didn’t need it. They, too, interpreted this as a negative result, though for consistency we coded it as increasing utilization. Another found that following the removal of CON, there was a substitution from coronary artery bypass grafts (CABG) to an alternative treatment: percutaneous coronary interventions (PCI).
In summary, the bulk of evidence suggests that CON does not have a significant effect on utilization of services, though among the tests that do find an effect, more than twice as many find a negative effect on utilization as find a positive effect. Finally, among those tests that do find a positive effect, CON seems to encourage some, sometimes inferior, procedures over alternative procedures.
c) Summarizing the CON and Access Literature
Figure 7 combines the data from Figures 5 and 6 to summarize the access literature. Among 132 tests in 58 papers, 73 find CON is associated with diminished access to care, 49 find negligible results, and 10 find CON to be associated with increased access. For every test that finds CON is associated with increased access, more than seven find it is associated with diminished access.
As we have discussed, needs-assessment does not typically involve quality assessment. Nevertheless, the authors of the NHPRDA hoped that CON would “achieve needed improvements in the quality of health services,” and this continues to be a common rationale for the regulation. Assessing quality does not simply account for the outcome of the medical procedure on the direct patient area, but also mortality rates for hospital-acquired pneumonia and patient deaths from serious complications after surgery. There is an estimated 5.7 percent decrease in deaths from post-surgical complications without CON laws; in a state like Georgia that averages roughly 300,000 inpatient surgeries, this represents over 17,000 lives.
In theory, CON might either enhance or undermine quality. On the one hand, the regulation might enhance quality if it permits greater proficiency through volume. By reducing the number of providers, CON is likely to cause each provider to perform more procedures. And if providers improve these procedures the more they practice and provide them, CON might lead to better outcomes for those who receive care (and, of course, worse outcomes for those whose care is rationed, but this is likely to go unmeasured). On the other hand, competition tends to enhance quality so CON might undermine quality by undermining competition.
We identify 31 papers that together contain 78 tests assessing the relationship between CON and quality of care. Among these, 43 find that CON is associated with lower-quality care, 29 find negligible or insignificant results, and 6 find CON is associated with higher quality.
Among the papers that associate CON with lower quality, researchers find that, other factors being equal, CON states have:
- higher mortality rates among surgical inpatients with serious treatable complications;
- higher mortality rates for heart attack, heart failure, and pneumonia;
- higher mortality from natural death, septicemia, diabetes, chronic lower respiratory disease, influence or pneumonia, Alzheimer’s and COVID during the pandemic;
- higher readmission rates following heart attack, heart failure, and pneumonia;
- lower levels of functional improvement among home health patients for bathing, ambulating, transferring to beds, managing oral medication and managing pain;
- higher ER and acute care admissions among home health patients;
- more ER visits within 30 days and more infections within 6 months of knee arthroscopy;
- more surgeries performed by lower-quality surgeons;
- lower RN staff ratios and grater use of physical force in nursing homes;
- fewer patients giving their hospitals a 9 or 10 on a 10-point scale; and
- lower home health agency ratings.
Among the papers that associate CON with higher quality, one paper found CON was associated with better quality on two measures of home healthcare but worse quality on six other measures. Another found that CON was associated with better outcomes for postoperative pulmonary embolism but worse outcomes among eight other dimensions of quality. One study found that CON was associated with lower NICU mortality in states with large metropolitan areas, and another study found mortality was higher following CABG in non-CON states, though subsequent analyses found CABG mortality declined following CON repeal.
In summary, the balance of evidence suggests that CON does not enhance quality. For every test that associates CON with higher quality there are more than seven that associate the regulation with lower quality outcomes.
4. Underserved Populations
The final aim of the NHPRDA was to ensure care for “underserved populations” including those “located in rural or economically depressed areas.” It isn’t entirely clear how the authors of the legislation envisioned this happening. Supply restrictions tend to restrict supply, especially to communities for whom care is marginally profitable.
It is possible that they hoped regulators would be more restrictive in evaluating projects for well-served communities and that this might then cause providers to shift more resources to underserved communities. More recently, in public testimonies, hospital associations have offered another theory. By increasing the profitability of safety net hospitals, they contend, CON laws might permit these hospitals to cross-subsidize more care to underserved populations.
We have identified 9 papers which together contain 10 tests that assess the effect of CON on underserved populations. These papers often look at access to care but some also look at the financing of care for underserved populations (due to overlap, some but not all of these tests are included in the earlier figures). Figure 9 summarizes this portion of the literature. Among these 10 tests, 8 find that CON is associated with diminished care for underserved populations, two find neutral or insignificant effects, and no tests associate CON with better or more care for underserved populations.
Among the two negligible results, one found that, by itself, CON had no statistically significant relationship to uninsured admissions, and the other found that CON had no statistically significant relationship to uncompensated care. The first of these tests, however, found that uninsured admissions were lower in CON states that also had uncompensated care pools and community benefit requirement laws.
One study found substance abuse centers less likely to accept Medicaid patients in CON states. Another (commissioned by the state of Georgia) found that the uninsured were more likely to pay out of pocket in CON states than in non-CON states. Safety net hospitals in CON states also have lower margins than those in non-CON states.
One pair of studies found that a large black-white disparity in the provision of coronary angiographies disappeared when the procedure was exempted from the CON process. And two studies found that rural populations have less access to care in CON states compared with non-CON states.
Taken together, the literature offers no support for the hypothesis that CON encourages care for underserved, rural or economically depressed communities. If anything, it seems to make these communities worse off.
5. The Political Economy of CON
There is little evidence that CON achieves its stated goals. If anything, it seems to undermine them. Why, then, does it persist? In this section we review a subset of papers that examine the political economy of CON.
One possible reason why states retain CON is that it is profitable for the hospitals that lobby for it. There is abundant evidence that CON leads to more business. Figure 10 summarizes the research on CON provider volume. Of 15 tests assessing the effect of CON on provider volume, 13 find that it is associated with higher volume, one test finds negligible results and one test finds it to be associated with reduced volume. CON, it seems, steers more patients to incumbent providers.
This extra volume, however, may not turn into extra profit. Compared with other effects, the effect of CON on hospital profits has been relatively understudied. Only two papers have looked at this question and they reach somewhat surprising results. One paper found that following the 1996 repeal of CON in Pennsylvania, hospital margins initially fell but eventually recovered. In fact, over the long run, Pennsylvania hospitals were more profitable than hospitals in CON states. The other test, which was mentioned in the previous section, found that safety net margins were higher in non-CON states than in CON states. Together, these results suggest that if CON positively affects hospital margins, it only does so in the short run.
Even if hospital owners fail to benefit from CON over the long run, their employees may still benefit. Indeed, one study found that urban hospital CEOs earn more than $90,000 more in CON states than in non-CON states. It is also possible that CON laws benefit certain types of providers such as those that are effective in political markets. One study looked at the relationship between PAC contributions and CON approval in three states. In Georgia, the authors found that a 1 percent increase in contributions by an applicant firm increases the odds of approval by 6.7 percent.
Finally, one study examined several political factors to determine the likelihood of a state retaining its CON regulation. The authors find that CON laws are correlated with: 1) Democrats in upper and lower houses, 2) higher hospital costs, 3) more affluent and better-educated citizens, 4) fewer physicians and 5) a variable measuring hospital interests. This last variable includes the number of hospital industry-related interest groups active in a particular state multiplied by their average political action committee spending. While this factor was found to be significantly associated with retention of CON, legislative party makeup was found to be more important.
Neighboring states have taken action in recent years to modernize or repeal their CON programs. In 2021, Tennessee scaled back their program and eliminated CON requirements for imaging in the most populous counties by using a population-based threshold, targeting the Nashville, Memphis, Chattanooga and Knoxville markets. Tennessee also repealed CON in its entirety for all rural counties that are categorized as economically disadvantaged by the Appalachian Regional Commission and do not currently have a hospital. In 2022, the South Carolina Senate passed a bill by a 35-6 vote that would have repealed the state’s CON law for every healthcare facility with the exception of nursing homes. It was ultimately never brought up for a vote in the House. Also in 2022, the North Carolina State Senate passed a bill by a 44-2 vote that would have repealed CON for ambulatory surgery centers, psychiatric beds, MRIs, chemical dependency treatment facilities, home health agencies and dialysis. This was part of a legislative package that would have also expanded Medicaid under the Affordable Care Act and allowed APRNs, including nurse practitioners, certified nurse midwives and certified registered nurse anesthetists, to practice without physician oversight. Ultimately, the North Carolina House never voted on the bill.
The politics of CON repeal are not easy. In nearly every community across the state, hospitals provide access to emergency care and are often the largest employer, providing high-paying, steady jobs. Not just in rural Georgia, but in metro Atlanta where three of the four largest employers are health systems. However, the narrative that a reduction in CON laws would result in the widespread closure of hospitals has not been validated by what we have witnessed in other states. In fact, academic research has shown a reduction in the access to care.
The use of CON laws to protect profitable service lines has harmed not only patients, but new providers often wishing to practice medicine more closely aligned with the community’s need, such as imaging centers or birthing centers. The counterargument that new providers can simply acquire a CON does not represent a viable path forward given the legal, financial and political ability of nearly all of these health systems. Even the process of allowing new hospitals – subject to the same indigent care and emergency access requirements that health systems utilize to decry ASCs and imaging centers – is not immune from the political reality and protectionism of CON. As one county commision chairman recalled of his conversation with a competing hospital’s CEO at a community rally in 2022 to support their new hospital that was tied up in legal challenges, “I know I can’t stop that hospital from being built. But I know I can delay it for many years.”
If state policymakers wish to lower healthcare costs and increase options for patients, reducing CON laws is one path forward.