Health Insurance Options for Small Businesses

Summary

Ensuring that affordable options for health insurance are available to small businesses has long been a focus of Congressional interest. Yet the relatively low rate at which small businesses offer coverage suggests they often face challenges in finding affordable and attractive options for their workers. In this report, the Congressional Budget Office examines the factors that influence small businesses’ decisions to offer coverage, describes the options available to them, and analyzes trends in enrollment in those options from 2014 to 2024, the most recent year for which data were available for CBO’s analysis.

Why Does Employment-Based Coverage Matter?

Employment-based health insurance remains the predominant source of coverage for Americans, with more than 164 million people obtaining insurance through an employer in 2024. When workers are not offered coverage by their employer, they and their families may obtain coverage from another family member’s employment-based plan, subsidized nongroup coverage in the marketplace established by the Affordable Care Act (ACA), Medicaid, or other options. Some remain uninsured.

Small businesses—defined in most states as those with 50 or fewer employees—employed about 38 million of the 139 million private-sector workers in 2024. Just over half of workers at small businesses (53 percent) worked for businesses that offered health coverage, compared with nearly all workers employed by large businesses. As a result, only 36 percent of small business workers were enrolled in a plan offered by their own employer.

Coverage challenges affect small business owners, as well. Among owners of small businesses with at least one employee, 20 percent were uninsured in 2023 (the most recent year for which relevant data existed for CBO’s analysis).

Among uninsured people in households with workers at small businesses but not large businesses, CBO estimates, 60 percent were eligible for some form of subsidized coverage, such as tax credits that apply toward premiums for nongroup coverage in the ACA marketplace, employment-based coverage through another family member, or coverage by Medicaid or the Children’s Health Insurance Program (CHIP). The remaining 40 percent were not eligible for any form of subsidized coverage.

Small businesses that offer health insurance do so to help attract and retain well-qualified workers. Increasing access to employment-based coverage would likely reduce the number of uninsured people in households with only small business workers. Doing so has important implications for the federal deficit, as well. Providing subsidized coverage through marketplace coverage or Medicaid costs the federal government, on average, more than twice as much as the forgone federal revenues from the tax exclusion for employment-based coverage. The relative cost to the federal government depends on the income of the affected workers, with employment-based coverage costing relatively more for higher-income workers and Medicaid or marketplace coverage relatively more for lower-income workers.

How Does Health Insurance Coverage Differ for Workers at Large and Small Businesses?

Compared with people in households that include a worker at a large business, people in households with only small business workers are three times as likely to be uninsured, four times as likely to obtain coverage through the nongroup market, and two and a half times as likely to enroll in public programs such as Medicaid.

From 2013 to 2023 (the most recent year for which relevant data for this part of the analysis were available), the difference in uninsured rates between what this report calls “small business households” and “large business households” narrowed. The uninsured rate for people in small business households fell from 25.4 percent to 13.6 percent, whereas the rate for large business households declined from 9.7 percent to 4.4 percent. Coverage gains among small business households were driven by expanded eligibility for Medicaid and the availability of premium tax credits for nongroup coverage in the ACA marketplace. Between 2018 and 2023, employment-based coverage in small business households also increased.

What Health Insurance Options Are Available to Small Businesses?

Small businesses can support health insurance coverage through several types of arrangements that differ in their legal requirements for benefits and for the rating method used to calculate premiums.

Fully Insured Small-Group Health Plans. Small businesses can have one of two types of plans purchased from licensed insurers in the small-group market.

Community-Rated Plans. The ACA introduced community rating rules, which limit the factors that insurers are allowed to use when setting premiums for plans sold in the small-group market. Premiums for such plans are based principally on the community as a whole and can vary only by age, geographic location, family size, and tobacco use and cannot reflect enrollees’ health history. These plans must comply with other rules as well, including a requirement that they accept all businesses and their workers regardless of the health status or claims experience of employees and their dependents (which is to say that “guaranteed issue” applies) and that they provide what are termed essential health benefits.

Grandfathered and Transitional Plans. Some businesses maintain pre-ACA plans that are either grandfathered or transitional. These plans were purchased before the passage of the ACA in 2010 (grandfathered plans) or after that date but before 2014, when the ACA’s main provisions took effect (transitional plans). Both types of plans are exempt from some of the ACA’s rules but remain available to certain businesses that continuously offered them since before the law’s implementation.

Self-Insured Plans. Some small businesses self-insure, assuming financial responsibility for their employees’ health care costs while purchasing stop-loss insurance and administrative services from a third-party administrator. Many of these arrangements are known as “level-funded plans” because the plan’s costs are spread evenly across 12 months, reducing variability for the employer. Self-insured plans are generally exempt from community rating and requirements to include essential health benefits, though they still must comply with many of the ACA’s other provisions, such as providing coverage for preexisting conditions.

Multiple-Employer Arrangements. Some small businesses join association health plans (AHPs) or similar arrangements that allow them to obtain coverage through larger groups, potentially offering more choices of plans or lower premiums.

Health Reimbursement Arrangements. Employers that do not offer traditional group health plans can contribute to health reimbursement arrangements (HRAs) to help employees purchase nongroup coverage with tax-excluded funds.

No Offer of a Plan. Alternatively, small businesses can choose not to offer health insurance, as they are not generally mandated to do so. A business can offer higher wages instead. When employers do not offer insurance, workers may obtain coverage from Medicaid, from a spouse's or a parent’s employer, or through the nongroup market—or they may remain uninsured.

What Factors Might Influence a Small Business’s Decisions About Health Insurance?

A small business must decide whether to offer health insurance and which type of coverage to provide. Those decisions affect compensation costs and the business’s ability to recruit and retain workers. The choice to offer coverage depends largely on financial capacity, the demand by employees, regulatory requirements, and government subsidies.

Businesses with tighter profit margins might find premiums difficult to incorporate into compensation costs. Workers who have access to Medicaid, a spouse’s or a parent’s plan, or marketplace subsidies sometimes prefer higher wages instead of employment-based coverage. State and federal rules also affect the availability and cost of plans. Among businesses that offer coverage, choices of plans reflect the workforce’s characteristics and preferences: Businesses with younger or healthier workers often pay less in self-insured arrangements, whereas those with older or higher-risk workers often obtain a better deal in community-rated plans. Businesses that have employees with differing health care needs, costs, or preferences can find it advantageous to offer HRAs to some or all employees to help them purchase nongroup coverage.

How Has Enrollment in Health Insurance Offered by Small Businesses Changed Over Time?

From 2014 to 2024, enrollment by workers and their families in health insurance plans offered by small businesses increased from 16.9 million to 18.6 million, a 10 percent rise. That growth was slower than the growth in total employment among small businesses, which increased 21 percent. Enrollment in grandfathered and transitional plans declined sharply, from 60 percent of enrollment in 2014 to 4 percent in 2024. Enrollment in ACA community-rated plans grew early in the period, reaching 60 percent in 2018, but declined to 50 percent by 2024. That decline in community-rated plans occurred as enrollment in self-insured and other non-community-rated arrangements rose to 46 percent, driven largely by increasing use of level-funded plans. AHPs and union plans accounted for relatively small shares of enrollment. The number of small businesses offering HRA contributions toward nongroup coverage increased but remained low.

How Much Are Premiums, and How Do They Differ in Different Market Segments?

From 2014 to 2024, enrollment-weighted average premiums for small business health plans rose 43 percent overall—46 percent for ACA community-rated plans and 35 percent for experience-rated plans. (By comparison, the average growth in premiums for plans offered by large employers was 47 percent over that same period.) In 2024, average annual premiums per enrollee for a small business health plan were $7,300—higher for community-rated plans at $7,750 and lower for experience-rated plans at $6,800. Premiums for plans offered by small businesses are generally somewhat lower than those for plans offered by larger businesses, perhaps in part because plans chosen by small businesses often have higher deductibles.

In different states, growth in premiums varied significantly, especially for community-rated plans and especially among states with a smaller share of enrollment in such plans. Those differences are probably partly related to departures by small businesses with healthier employees from community-rated plans for experience-rated options that offered them lower premiums, although many other factors affect premiums as well.

The Role of Small Businesses in Health Insurance Coverage

Of the 38 million people who worked at small businesses in 2024, including employees and working owners, about half were offered health insurance by their employer, on average. Offers of health insurance tend to be least likely for the smallest firms. Offering health insurance can help businesses attract and retain workers, but the cost of health insurance plans can be high enough that some businesses might not be able to cover those expenses. Some businesses predominantly hire employees who prefer cash compensation instead.

People in small business households were three times as likely to be uninsured in 2023 as people in large business households—that is, households in which someone is employed by a large business.1 After the implementation of the ACA, the share of people in small business households who were uninsured fell from 25 percent in 2013 to 13 percent in 2023, whereas the shares with public, nongroup, and employment-based coverage grew.

Most states define a “small-group” health plan as one offered by a business with 50 or fewer full-time employees. In 2016, four states (California, Colorado, New York, and Vermont) expanded the definition to include businesses with 100 or fewer employees.2 Survey data sometimes rely on different definitions in grouping businesses by size. In those cases, CBO uses the closest equivalent grouping in the survey data.

People Affected by Small Business Health Insurance

When a small business offers health insurance, the offer typically is extended to both employees of the business (along with working owners) and their spouses and children, who can enroll as dependents in employment-based plans.

Workers at Small Businesses. In 2024, 38.3 million people worked in small businesses (see Figure 1). The figure includes owners who worked at their own small businesses as well as the people they employed on a full-time or part-time basis. From 2011 to 2024, private-sector small businesses added 8.6 million workers, growing at about the same rate as the total workforce.3 The COVID-19 pandemic caused a notable dip in employment by small businesses in 2020 and 2021.

As of 2024, 14.3 million people (or 37 percent of employees of small businesses) worked in businesses with fewer than 10 employees; 12.3 million (32 percent), in businesses with 10 to 24 employees; 9.9 million (26 percent), in businesses with 25 to 49 employees; and 1.8 million (5 percent), in businesses with 50 to 99 employees (in the four states that expanded their definition of the small-group market). The shares of private-sector employees in each size category have remained mostly stable over time.

Family Members of Small Business Workers. More than a third of covered workers have, in addition to their own coverage, their spouse or children covered by health insurance offered by their employer. According to CBO’s analysis using the household component of the Medical Expenditure Panel Survey (MEPS-HC), private-sector small-group family health insurance plans covered about two dependents for each policyholder, on average, from 2014 to 2021.

Small Businesses’ Offers of Health Insurance

Only some businesses offer health insurance to their workers, and the smallest businesses are especially likely not to do so. In 2024, about 53 percent of small business workers were in firms that offered a plan (see Figure 2). The share of employees at businesses that offered a plan was lowest among the smallest firms and higher at somewhat larger firms: 32 percent of workers in businesses with fewer than 10 employees, 55 percent in businesses with 10 to 24 employees, 72 percent in businesses with 25 to 49 employees, and 95 percent in businesses with 50 to 99 employees (in the four states that include businesses of that size in the small-group market). Offer rates decreased somewhat from 2011 to 2015 but generally increased after that year and remained over 50 percent from 2019 to 2024.

Part-time workers are less likely to work for businesses that offer health insurance, and even if their employer offers a health plan, part-time workers are less likely to be eligible for it, generally because their employer restricts the offer to full-time employees (see Figure 3). In addition, some employers impose a waiting period of up to 90 days after hiring.

Even when offered a health plan at work, not every eligible employee enrolls. Some employees have coverage through a spouse, parent, Medicaid, or Medicare. Some get coverage through the nongroup market. If their employer’s offer is deemed unaffordable, they may be eligible for subsidized coverage in the ACA marketplace.4 Finally, some remain uninsured. Small employers can vary the features of health plans they offer, affecting the affordability and attractiveness of the plans to workers. (For a description of common features, see Box 1.)

Incentives for Small Businesses to Offer a Health Plan

Small businesses vary in many ways, including size, the average wage of employees, the industry they are in, the mix of full-time and part-time workers, and workers’ access to alternative sources of coverage. Those variations can affect the demand for an offer of health insurance. Health insurance comes at a cost, and employees generally expect the employer to pay most of the cost of premiums. However, providing the benefit to workers can have important advantages for some types of businesses.

Benefits of Offering a Health Insurance Plan. Small businesses offer health insurance for the same reasons that larger businesses do: to attract and retain high-quality employees and to allow employees to benefit from tax incentives. Most people under age 65 in the United States get health insurance through either their own job or through a spouse’s or a parent’s coverage.5 As a result, many workers have an expectation that a job offer will include the option of signing up for a health plan to cover themselves and their family. Given that nearly all large businesses provide health insurance, some small businesses offer similar benefits to attract and retain top talent and remain competitive.6 Expecting employees to rely on alternatives such as nongroup coverage may not be attractive to some potential employees. Although most plans in the nongroup market are subject to many of the same rules as small-group plans, including community rating and coverage of essential health benefits, nongroup plans tend to have narrower provider networks, which can lead to higher out-of-pocket costs.7

The federal tax system provides tax advantages for health insurance purchased through an employer. Employers’ payments for health insurance premiums are excluded from federal income and payroll taxes.8 For about two-thirds of workers at small businesses, the employees’ premium contributions are also tax-excluded.9 Self-employed individuals, including working owners of small businesses, are often eligible for a deduction on their federal income taxes for their own health insurance premiums.10

The tax subsidy benefits workers by allowing them to receive employment-based health insurance without paying income or payroll taxes on their employer’s contributions (and, in most cases, the employee’s contributions as well), effectively lowering the cost of coverage and encouraging employers to offer health insurance. Unlike Medicaid or income-based subsidies for nongroup insurance, the tax subsidy is universally available, covering workers who do not qualify for other assistance—although the benefit of the tax subsidy for employment-based coverage is smaller for low-income workers because their marginal tax rates are generally lower than those for high-income workers.

Drawbacks of Offering a Health Insurance Plan. The cost of offering a health plan can present a challenge for many small businesses. Small businesses with low total revenues sometimes find it difficult to afford the cost of health insurance premiums for employees.11 For some businesses, especially those with lower-wage workers, the businesses’ contributions for health insurance can account for a large share of total compensation. Average health insurance premiums for group coverage in the United States in 2024 were about $8,500 for single coverage and $24,500 for family coverage, much of which workers would expect their employer to pay.12 Meanwhile, average total wages and salaries for small businesses were about $54,500 for a full-time worker in 2024.13 Moreover, there is considerable variation in average wages across small businesses. Small businesses that employ more low-wage workers are much less likely to offer a health insurance plan, and eligible employees at those firms are less likely to enroll in a plan if one is offered.

CBO examined MEPS-IC data for private-sector businesses with fewer than 50 employees, comparing small businesses with half or more of their employees qualifying as “low-wage workers” (low-wage businesses) with small businesses with fewer than half of their employees falling into that category (higher-wage businesses). In 2024, 58.6 percent of workers at higher-wage small businesses were in firms that offered a health plan, compared with 26.5 percent of employees at lower-wage small businesses. In that same year, among eligible workers at higher-wage small businesses that offered a plan, 67.7 percent enrolled in the plan, whereas at lower-wage small businesses, only 55.1 percent of eligible employees enrolled in the offered plan.14 Some research has shown that lower-wage workers sometimes value health insurance coverage at less than its cost, which helps explain the differences in the rates at which plans are offered and employees enroll.15 For some small businesses hiring low-wage workers, the minimum wage can present a barrier to reducing cash wages to offer health insurance or other benefits, lessening the likelihood that such businesses will offer a health plan.16

Some factors that influence premiums are things that small businesses can directly control, such as the extent of coverage and the choice of a provider network. However, other factors, such as local providers’ prices, market competition, and the age and health mix of employees, they cannot directly control.

Some businesses employ people who, for various reasons, prefer higher wages rather than an offer of employment-based health insurance. Some workers have an alternative source of employment-based coverage through a family member.17 Low-wage workers sometimes qualify for Medicaid, and some others are eligible for premium tax credits for nongroup coverage (though the credits are unavailable if their employer’s plan meets the ACA’s affordability standards).18 Still other workers, who can tolerate the risk, prefer higher cash wages even if that choice means forgoing coverage.19

In some industries, the norm is not to offer health insurance, and small businesses can hire and retain workers without offering a health plan. Some such businesses engaged in services (including food service and accommodation) and retail trade employ many part-time and seasonal workers.20 Full-year health insurance contracts often do not align with seasonal employment, and health insurance premiums could constitute a very high share of total compensation for part-time employees.

Administrative costs of offering benefits can also pose a challenge for small businesses.21 Small businesses sometimes also lack sufficient knowledge and time to select and set up a health plan. For the latter reason, most small businesses that do offer a health plan rely on agents, brokers, insurers, and third-party administrators to assist in selecting and setting up the plan.22

Differences in Health Insurance Coverage for Workers at Large and Small Businesses

Many households with at least one worker have employment-based health insurance, but that situation is less common for workers at a small business and their families.

Overall, small business households have experienced a significant drop in uninsured rates.23 Since 2013, the uninsured rate gap has narrowed between working households where someone works for a large business and those with only a small business worker, whereas enrollment in Medicaid and subsidized nongroup coverage has increased for small business households.

Coverage in 2023. In 2023, people under age 65 in small business households were less likely to have employment-based insurance and more likely to be uninsured than were those in large business households. Analyzing data from the MEPS-HC, CBO found that in that year people in small business households were three times as likely to be uninsured as those in large business households (13.6 percent versus 4.4 percent, respectively) (see Figure 4). People in small business households were significantly less likely to have employment-based insurance (48.5 percent versus 80.4 percent) and more likely to rely on public insurance (24.9 percent versus 10.4 percent) or nongroup insurance (8.0 percent versus 2.0 percent) than were those in large business households. Those differences in coverage are likely driven not just by differences in offers of health insurance by employers, but also by average wages and access to alternative sources of coverage.

Notably, 20.3 percent of the owners of small businesses (with at least one employee) were uninsured in 2023. Only 16.6 percent of small business owners were policyholders for an employment-based insurance plan.24

Coverage Over Time. From 2013 to 2023, the gap in uninsured rates between large and small business households narrowed (see Figure 5). The uninsured rate in small business households, which was 25.4 percent in 2013, dropped by 11.8 percentage points, to 13.6 percent. In comparison, there was a 5.2 percentage-point decline in the uninsured rate for large business households, or a decline from 9.7 percent to 4.4 percent.

Most of the coverage gains in small business households came from public insurance, driven by expanded access to Medicaid, and increased enrollment in nongroup coverage, driven by the introduction of subsidies for such coverage in the ACA marketplace. Additionally, employment-based coverage in small business households rose after 2018.

In 2020, the federal government increased federal funding for Medicaid and required states to provide continuous eligibility to enrollees through the end of the coronavirus public health emergency. Those provisions contributed to large increases in enrollment in Medicaid from 2020 through 2023.25 In addition, from 2021 to 2025, the federal government provided greater subsidies for marketplace coverage by reducing the maximum share of income that eligible enrollees must contribute toward premiums and by extending eligibility to people whose income is above 400 percent of the federal poverty level.26

Uninsured People in Small Business Households and Their Access to Subsidized Coverage

Although the number of people in small business households without health insurance coverage has declined in the past decade, millions of small business workers and their families remain uninsured. CBO has previously analyzed the entire uninsured population under age 65 in 2019 and its access to subsidized coverage.27 For this report on small businesses, the agency performed the same analysis for the same year specifically for people without health insurance in small business households. Besides having the benefit of extending that earlier analysis, CBO’s choice of further analyzing 2019 for the current report avoids presenting data for years when the federal government enacted many temporary policy changes to mitigate the effects of the economic disruption caused by the COVID-19 pandemic. Given that most of those policies have now expired, the agency decided that circumstances in 2019 were likely to be more relevant for policymakers in considering options for the future.28

In 2019, CBO estimates, about 3.8 million, or 32 percent, of the 11.9 million uninsured people in small business households had access to premium tax credits for nongroup marketplace coverage, and about 1.8 million, or 15 percent, were eligible for coverage by Medicaid or CHIP (see Figure 6). Another 1.6 million, or 13 percent, had access to employment-based coverage but declined that coverage. The remaining 40 percent were not eligible for any type of subsidized coverage.

If policymakers sought to reduce the number of people in small business households who lack health insurance, increasing the share of small businesses that offer a health plan could cost the federal government less than expanding access to other subsidized options such as marketplace subsidies or Medicaid or CHIP. In general, the cost per person to the federal government of providing marketplace subsidies or coverage by Medicaid or CHIP is about twice as high as the estimated loss in federal revenues from the tax exclusion for employment-based coverage.29 Because the federal cost of different types of subsidized coverage depends on factors such as household income, geographic location, age, and health status and because people’s willingness to enroll in different types of coverage also varies depending on most of those same factors, the cost of expanding health insurance coverage would depend crucially on the target population and the incentives the federal government might provide to encourage more enrollment.

Health Plan Options for Small Businesses

Small businesses have various options for providing health insurance for their employees.

Key Characteristics That Small Businesses Consider

Characteristics that small business owners may consider when deciding which option to offer to their employees include the following:

What Premiums Are. For many small businesses, the most important factor to be considered is the premiums for different options, including both the portion of the premiums paid by the employer and the share paid by employees who enroll themselves (and often their family members) in the plans. Premiums will be affected by many factors, some that the employer can control (such as whether to offer a plan with low out-of-pocket costs or a high-deductible plan and how broad the provider network is) and some that the employer has less control over (such as the age and health risk of workers and the number of dependents they have).

Whether Plans Are Fully Insured or Self-Insured. Plans purchased from a licensed insurer, referred to as fully insured plans, are regulated differently than plans in which the employer self-insures. In general, fully insured plans are regulated at both the state and federal levels, whereas self-insured plans are largely regulated at the federal level.

Whether Plans Are Guaranteed Issue. “Guaranteed issue” means plans must accept all businesses and their workers, regardless of the health status or claims experience of employees and their dependents. The requirement ensures access but can lead to higher premiums because insurers are not allowed to reject groups with high-cost enrollees.

What the Rating Method Is. Community-rated plans base premiums on a larger pool of enrollees, such as the entire small-group market in a state, whereas experience-rated plans use the specific claims history of enrollees at that business. Experience-rated plans can cost less as long as enrollees’ medical costs remain low, whereas community-rated plans can provide more predictable premium costs from year to year even if one or more enrollees have unexpectedly high medical costs.

Whether Costs Vary Within a Year. Most types of plans offer a flat premium or premium-equivalent each month. Employers sponsoring some types of self-insured plans can find that their claims costs vary from month to month, increasing their financial risk.

How Much Flexibility Is Available in Covered Benefits. Some plan types are required to cover a wide range of benefits such as the full set of essential health benefits defined in the ACA and certain benefits mandated by the state, though others are exempt from that requirement. Some types of plans allow flexibility in tailoring benefits, for example, by excluding benefits of less interest to workers, offering wellness programs to encourage healthy practices, or adding coverage for other services that workers may value receiving as tax-advantaged benefits. Excluding benefits will generally reduce premiums, while adding benefits will generally increase premiums.

Whether Plans Are Subject to State Premium Taxes. Plans regulated at the state level must pay state and local premium taxes, which average 1.4 percent of premiums nationwide but vary among states, from nearly zero to more than 6 percent of premiums.30

What Responsibility for Compliance Involves. Plans differ in reporting and privacy requirements and in who manages compliance, which can affect the administrative burden on businesses.

Options Available to Small Businesses

Health plan options available to small businesses fall into several broad categories: fully insured health insurance plans sold within the small-group market, self-insured plans, multiple-employer arrangements, and support for purchasing plans in the nongroup market (see Figure 7). Small businesses can also choose not to sponsor a plan or offer other tax-subsidized support for health coverage and instead provide compensation in the form of cash wages. Health plan options are subject to different state and federal regulations. Plans’ features and costs are likely to differ across all of the options. Which option a business selects can have implications for premiums, covered benefits, risk, the ability of the business to attract and retain qualified employees, and other factors. The decisions of businesses to offer or not offer health insurance coverage, and how much those plans cost, also affect federal tax revenues.

Options in the Fully Insured Small-Group Market. Within the fully insured small-group health insurance market, since the implementation of the ACA, only two main options exist: community-rated plans that comply with all the rules established under the ACA and grandfathered and transitional plans that the business has continuously offered since before the implementation of the ACA.

An ACA Community-Rated Plan. One option available to all small businesses is to pay premiums to a commercial insurance carrier to cover medical claims—for a fully insured health plan. Such plans are subject to state regulations and to federal requirements enacted as part of the ACA. The ACA created a system that pooled together the health risk of enrollees in all small-group plans in each state—the state’s “single risk pool.” Requirements under the ACA include these:

• Premiums are community-rated in that a given plan’s premiums can vary among individuals and businesses only on the basis of age, geographic area, family size, and tobacco use.
• Plans must participate in a risk adjustment system in their state that redistributes funds from plans with healthier enrollees to plans with less healthy enrollees to mitigate adverse selection within the small-group market, whereby less healthy people enroll in certain types of plans (and thus alter plans’ costs).31
• Plans are guaranteed issue and guaranteed renewable.32
• Plans must cover a set of essential health benefits.33

States can also impose additional requirements, which can include additional mandated benefits, further limits on rating, and required payment of premium taxes. Small businesses with fewer than 25 relatively low wage employees may qualify for a small business tax credit by offering an eligible health plan.34 According to preliminary tax data provided to CBO by the Joint Committee on Taxation, relatively few small businesses have claimed the credit, particularly in 2014 and later years, perhaps for several reasons: Small businesses can claim the credit for only two years; and starting in 2014, the requirement to select plans certified by the Small Business Health Options Program to qualify for the credit would exclude grandfathered and transitional plans, self-insured plans, and multiple-employer plans such as AHPs.

Grandfathered and Transitional Plans. Some small businesses can keep their pre-ACA plans if the plans have been in place since before the ACA’s implementation.35 Certain plans from 2010 are grandfathered and thus exempt from most of the ACA’s regulations.36 Plans in place since 2013, known as transitional plans, are exempt from fewer ACA regulations, and only 27 states allowed such plans in 2022.37 Both types of small-group plans are exempt from federal requirements like community rating, guaranteed issue, risk adjustment, and the provision of essential health benefits. Grandfathered plans are also exempt from rate reviews, from the requirement to cover preventive services without cost sharing, and from a prohibition on salary-based discrimination.38 Neither grandfathered nor transitional plans are available to new businesses or to businesses that have not been offering them continuously since before the implementation of the ACA.

Self-Insured Plans. Some small employers choose to self-insure, assuming financial responsibility for employees’ health insurance claims rather than paying premiums to an insurer. (Self-insured plans are sometimes referred to as self-funded plans.) Stop-loss insurance is essential for managing financial risk, as the variability of claims can pose a greater risk for small groups than for large ones. Such policies are regulated at the state level and cover claims above a set amount, with employers covering costs below that threshold.39

Employment-based health plans that are self-insured are subject to fewer of the ACA’s regulations than fully insured plans. Self-insured plans are largely regulated under the Employee Retirement Income Security Act (ERISA) and are required to comply with only some of the provisions that the ACA imposed on fully insured plans. For example, self-insured plans are not required to provide the full set of essential health benefits. They are exempt from state regulations requiring specific benefits.40 Self-insuring also allows businesses to avoid community rating and risk adjustment and instead pay premiums that are based on the age and health of their own workforce. Such plans are not necessarily available or advantageous to businesses with workers who are older or less healthy than average. Self-insured plans also allow employers more flexibility to tailor the benefits to the interests and needs of their employees—that is, they can omit benefits that are not of interest to their employees and add benefits that are not normally included in fully insured community-rated plans offered under ACA’s rules. Self-insuring imposes added responsibilities, including fiduciary duties and compliance with privacy rules. Third-party administrators often assist with compliance, but the ultimate responsibility lies with the employer.

Level-Funded Plans. Many small businesses that self-insure choose level-funded plans, which bundle a self-insured health plan with stop-loss insurance, a reserve account, and administrative services such as claims processing and network management. Employers make fixed monthly payments that cover the stop-loss premiums, service fees, and a portion of expected claims below the stop-loss threshold. That structure helps manage both high-cost claims and unpredictable timing of claims. If actual claims fall below expectations, employers sometimes receive a partial refund at year’s end.

Similar to other self-insured plans, level-funded plans are experience-rated, meaning costs depend on the claims history of enrollees from the specific small business rather than being pooled with the claims history of enrollees in other small businesses in the state. Although stop-loss coverage limits risk, high claims in one year can lead to sharp increases in stop-loss premiums, potentially making level-funded plans more expensive than ACA-compliant alternatives after a year with high medical claims.

Other Self-Insured Plans. Small businesses that have sufficient financial resources sometimes opt not to use the full range of services available in a level-funded plan. For example, a business might pay claims directly up to the stop-loss threshold rather than having the costs spread out evenly over 12 months. However, an employer opting for that arrangement typically would still contract with a third-party administrator to process claims and manage the provider network.

Multiple-Employer Plans. In addition to the above types of plans, there are several options that allow small businesses or their employees to band together to obtain coverage as a larger group. The most prominent are AHPs, union plans, and professional employer organizations (PEOs).41

Association Health Plans. AHPs are business associations that offer health insurance to employees of member employers.42 Regulatory treatment of these plans depends on the association’s structure. If an AHP qualifies as a “single employer” under ERISA section 3(5), it is regulated as a large-group plan, exempt from the ACA’s rules for small-group plans such as community rating.43 Such AHPs can vary premiums by member firms’ claims experience, allowing small businesses with healthier workforces to obtain lower premiums than in the community-rated market.

In 2018, the Department of Labor eased the rules for AHPs and established a new, less restrictive pathway for groups to form associations that offer plans.44 The rule retained the original pathway for groups to form associations and offer AHPs but created a new pathway that had less stringent requirements for associations to assert a “commonality of interest.” However, the rule was struck down in 2019.45 After the appeal of this decision was dismissed, the rule was formally rescinded in 2024, and the Department of Labor noted that it was not aware of any remaining AHPs that were established under the new pathway created by the 2018 rule.46

AHPs that meet the single employer standard are not required to provide all essential health benefits, potentially lowering costs. However, in a 2019 analysis, CBO found that benefits in AHPs were similar to those in community-rated small-group plans and that cost differences for businesses switching to AHPs mostly resulted from the exemption from community-rating requirements rather than from eliminating benefits.47 States can regulate AHPs and create their own rating restrictions and requirements for essential health benefits.48 Most AHPs purchase fully insured plans. If the association is fully insured and does not meet the definition of an employer, small-group employers that are members are subject to all of the ACA’s and the state’s regulations that apply to the small-group market.

Union Plans. Union plans, also known as Taft–Hartley plans or multiemployer health plans, are administered jointly by labor organizations and employers and are funded mainly by employers' contributions as specified in collective bargaining agreements. These plans are relatively more prominent in the construction industry.49 The plans are collectively bargained between employers and a union. Such plans can allow workers to transfer benefits or earn credits toward benefits while moving from one employer to another, as long as both employers participate in the same plan.50

Most union plans are self-insured and employ an insurance company to act as a third-party administrator. When self-insured, such plans are experience-rated. Union plans are exempt from state regulation if 85 percent of participants are employed under a collective bargaining agreement.51 Although such plans are not required to provide the full set of essential health benefits, union plans often have comprehensive benefits.

Professional Employer Organizations. PEOs offer a source of coverage like AHPs for small businesses. Businesses pay PEOs to administer their payroll, taxes, and employee benefits, sometimes including health insurance. For health insurance, a PEO signs as a coemployer with the small business and then enrolls the employees in large-group insurance plans sponsored by the PEO. Such plans can be fully insured or self-insured. Just like AHPs, PEOs can generally charge different groups of employers different rates based on claims experience, making these plans experience-rated.52 According to the National Association of Professional Employer Organizations, their PEO membership included more than 230,000 small and mid-sized businesses with about 4.5 million employees in 2025, with 85 percent of member businesses having fewer than 50 employees.53 Approximately 53 percent of PEO member businesses are in Florida, California, New York, and Texas.54 Data on the number of enrollees in health plans sponsored by PEOs who work for small businesses are unavailable.

Support for Coverage in the Nongroup Market. Small businesses can provide tax-advantaged assistance to workers who buy their own coverage in the nongroup market by contributing to certain types of HRAs.

Two main types of HRAs can be used to help employees purchase a health plan in the nongroup market: a qualified small employer health reimbursement arrangement (QSEHRA) or an individual coverage health reimbursement arrangement (ICHRA). The 2016 21st Century Cures Act allows small businesses (with fewer than 50 full-time employees) without a group plan to offer tax-free funds through a QSEHRA, which employees can use for health care or insurance premiums.55 In 2019, the Departments of Treasury, Labor, and Health and Human Services published a final rule that permitted employers of any size to provide funds through an ICHRA to employees to purchase health insurance through the nongroup market.56

Employers’ contributions to QSEHRA and ICHRA accounts are excluded from income and payroll taxes. If an employee uses the HRA funds to purchase a health plan in the nongroup market outside the ACA marketplace and the employer offers a Section 125 cafeteria plan, the net amount the employee pays toward premiums can also be excluded from federal income and payroll taxes.57 (A cafeteria plan allows workers to choose between a taxable benefit, such as cash wages, and nontaxable fringe benefits.)

Some employees are eligible for premium tax credits in the ACA marketplace. Eligibility depends on income, family size, and other factors—including whether the employee and their family have an affordable offer of coverage from the employer.58 An offer of an ICHRA or QSEHRA contribution from the employer is considered an affordable offer if the contribution is enough to bring the net premium of the lowest-cost silver plan for the employee in that area below the affordability threshold for that year.59 Businesses with 50 or fewer employees are not subject to penalties for not complying with the requirement to offer insurance that generally applies to larger businesses (known as the employer mandate), so they can offer a QSEHRA or ICHRA contribution even if such an offer might be considered unaffordable for some employees.60

If an employer’s offer of a QSEHRA or ICHRA contribution constitutes an affordable offer for some employees, those employees are not eligible for premium tax credits during the same time period. If the offer is un­­­affordable for an employee and they are eligible for premium tax credits that would be higher than the employer’s ICHRA contribution, the employee can decline the ICHRA offer and take the premium tax credits instead.61 However, if a QSEHRA contribution constitutes an unaffordable offer for a given employee, the employee can accept both the employer’s QSEHRA contribution and the premium tax credits, but the amount of credits is reduced by the amount of the QSEHRA contribution.62

Either type of HRA option allows the employer to set the amount of its contribution and allows the employee to select from many plans in the nongroup market.63 Thus, employees have more choice but also the responsibility to shop for and select a plan. A nongroup plan may provide greater flexibility and may allow an employee who changes jobs to keep the same health plan when moving to another employer—if that new employer also offers an HRA contribution to help purchase a nongroup plan.

No Health Insurance Plan or Support. In 2024, about 47 percent of small business employees worked at firms that did not offer a health insurance plan. Because there is no employer mandate for businesses with 50 or fewer employees, most small employers do not face employer mandate penalties from the federal government if they choose to not offer health insurance to their employees.64

Although many workers value jobs that offer health insurance, others have alternative sources of coverage or less interest in employment-based plans. Some workers are covered through a spouse or parent, are enrolled in subsidized marketplace coverage, or are enrolled in public programs such as Medicaid or Medicare. Others simply value health insurance less than wages. When a small business does not offer health benefits, it may be able to offer higher pay or other forms of compensation to appeal to workers who have other sources of coverage or do not place a high value on health insurance.

Choosing Among Available Options for Supporting Health Coverage

Small employers choose among the options discussed above, often relying heavily on agents and brokers who present them with choices that may work best for them and their employees. The key characteristics of each option in deciding what type of plan to offer, if any, include whether the plan is fully insured or self-insured, whether the plan is guaranteed issue, whether it is community-rated or experience-rated, whether it features level monthly premiums or variable monthly claims payments, what flexibility there is in covering benefits, and who is responsible for compliance with fiduciary, privacy, and other rules (see Table 1).

Enrollment by Segment in the Small-Group Market

CBO analyzed data from several data sources to estimate enrollment in each segment of the small-group market from 2014 to 2024 and found notable changes over time in the composition of enrollment (see Figure 8). The key data sources include administrative data on ACA community-rated plans and on all fully insured plans in the small-group market as well as survey data on the number of small business workers enrolled in plans offered by their employers. CBO relied on multiple data sources because no single source provides comprehensive information.65

Total enrollment in health plans offered by small businesses grew by 10 percent from 2014 to 2024. That growth was slower than the growth in employment at small businesses, which increased by 21 percent over that same period. As noted earlier, an increasing share of people in small business households enrolled in public and nongroup coverage (see Figure 5). The share of people in such households enrolled in employment-based coverage also grew modestly. Survey data collected in 2020 are problematic; hence, total enrollment in the small-group market and the number of people enrolled in self-insured and other plans in that year are both unclear.66

Enrollment in grandfathered and transitional plans declined substantially, as fewer small businesses remained in older plans set up before the implementation of the ACA. Enrollment in community-rated plans rose through 2018, as more small businesses chose those plans but then fell as more businesses switched to other types of plans. Enrollment in self-insured and other types of plans grew in 2019 and later years, in part because of the increasing availability of level-funded plans.

In 2024, about half of enrollees in group health plans sponsored by small businesses were in an ACA community-rated plan. The count of enrollees includes the policyholders (small business employees and working owners) as well as dependents on those policies (generally spouses and children).

Fully Insured Plans in the Small-Group Market. Among fully insured plans, enrollment in grandfathered and transitional plans predominated at the beginning of the period examined, whereas enrollment in community-rated plans did so at the end.

ACA Community-Rated Plans. Enrollment in community-rated small-group plans subject to all of the ACA’s rules grew substantially in the first few years after these new plans were introduced by the legislation in 2014. By 2018, the community-rated market segment (at 10.7 million people) made up 60 percent of enrollment in small-group market plans. After that year, the share enrolled in the segment fell, dropping to 50 percent (at 9.3 million people) in 2024.

Grandfathered and Transitional Plans. In 2014, by CBO’s estimates, when the ACA’s rules for the small-group market were first implemented, enrollment in fully insured grandfathered and transitional plans (at 10.1 million people) made up 60 percent of enrollment in plans sponsored by small businesses. However, after that year, enrollment in such plans declined as small businesses switched away from the plans, states stopped allowing transitional plans, and insurers stopped offering the plans.67 In addition, newly formed businesses did not have the option of choosing such plans if they were not offering coverage before 2014. By 2024, grandfathered and transitional plans (covering 0.7 million people) made up only 4 percent of enrollment in the small-group market.

Self-Insured and Other Plans. Data on enrollment in plans outside the fully insured small-group market are not available through administrative sources. Consequently, CBO used survey data on enrollment by small business workers through plans offered by their employer and then subtracted enrollment in fully insured plans to estimate the number of small business workers and their dependents enrolled in self-insured and other types of plans. Using that approach, CBO estimates that from 2014 to 2024, enrollment in self-insured plans sponsored by small businesses as well as some enrollment through multiple-employer plans grew from 18 percent (at 3.0 million) to 46 percent (at 8.6 million).

Self-Insured Plans. CBO’s conversations with insurance industry experts and other data sources suggest that self-insured plans, particularly those known as level-funded plans, are the primary driver of the increase in enrollment in experience-rated plans—though estimating the share of enrollment accounted for by level-funded plans (as opposed to more traditional self-insured arrangements) is difficult.

Multiple-Employer Plans. Many AHPs and union plans are self-insured or regulated in the large-group market. Enrollment in these plans is included in CBO’s estimates of enrollment in the “self-insured and other plans” segment. The MEPS-IC includes data on enrollment in AHPs and union plans that CBO aggregated to the national level.68

Enrollment in AHPs was less than 650,000 from 2014 through 2018 and then rose rapidly in 2019 after the rule establishing a new pathway for their creation went into effect (see Figure 9). Enrollment then dropped as the AHP rule was contested in court. Enrollment rose again, to over 1.6 million in 2022, before falling, to 1.3 million in 2024, when the rule was rescinded; in 2024, enrollment in AHPs constituted about 7 percent of enrollment in the small-group market. Although the higher enrollment in AHPs in recent years is perhaps surprising given the court action and eventual rescission of the rule, news stories may have drawn attention to the availability of AHPs. Such plans have gained popularity in some states as an experience-rated alternative for businesses that might otherwise have considered setting up a level-funded plan.

Union plans accounted for about 4 to 5 percent of enrollment in the small-group market from 2014 to 2024. Enrollment in such plans is likely driven by union membership and employment in unionized industries that make use of these multiple-employer arrangements for union workers who often move from one employer to another.

Small businesses that enter into coemployment agreements with PEOs are most likely not captured in MEPS-IC data as small business workers. Instead, they would be reported as employees of the PEOs, which are large employers. As a result, enrollment by small business workers via such an arrangement cannot be determined.

Coverage in the Nongroup Market Funded by QSEHRAs and ICHRAs. HRAs that can be used to purchase nongroup coverage are options that were introduced relatively recently. Survey data indicate that some small employers are offering HRA options to some or all of their employees, but how many employees use such options to buy nongroup coverage is unclear.69

Industry reports and KFF survey data show small but increasing numbers of employers offering QSEHRAs and ICHRAs since they became available in 2017 and 2020, respectively.70 According to preliminary tax data from the Joint Committee on Taxation, use of the arrangements to enroll in nongroup coverage appears to be slight. By CBO’s estimates, in 2024 fewer than 50,000 small business employees (along with some of their dependents, presumably) enrolled in nongroup plans using QSEHRAs. ICHRAs appear to be somewhat more popular, perhaps because they have fewer restrictions and can be offered by businesses of any size. However, accurately estimating offers of ICHRAs by small businesses and use of ICHRA funds by their employees to enroll in nongroup coverage is difficult because small businesses are not subject to the reporting requirements for ICHRAs that apply to larger businesses. Consequently, CBO’s analysis of enrollment by market segment does not include people whose employers provide them with HRA funds through QSEHRAs and ICHRAs and who use those funds to purchase coverage in the nongroup market.

Premiums

Nationally, average annual premiums in the small-group market have been rising by about 3.6 percent per year—with growth for community-rated plans somewhat faster and for experience-rated plans somewhat slower. At the state level, growth in premiums for community-rated plans has varied widely. Premium growth has tended to be higher in states where a relatively small share of the small-group market is enrolled in community-rated plans.

National Level

From 2014 to 2024, at the national level, premiums for ACA community-rated plans grew faster than the average for all plans offered by small businesses (see Figure 10). Average premiums per enrollee in community-rated plans grew by 46 percent in nominal terms, but for experience-rated plans, that figure was 35 percent. In the overall small-group market, including all market segments, average premiums grew by 43 percent.71

Most recently, from 2023 to 2024, average premiums for community-rated plans grew 6.1 percent, CBO estimates. For experience-rated plans, average premiums grew 2.6 percent. Although the difference is consistent with adverse selection, empirical research has not confirmed it was the cause. The average for all market segments was 4.3 percent.

State Experiences

Premium growth for community-rated plans differs among states. For the 12 states with the most growth, the increase from 2014 to 2024 was 89 percent, whereas for the 12 with the least, the average was 31 percent. The differences among states reflect variation in both differences in adverse selection and other factors, such as varying market concentration among health care providers and insurers. Adverse selection is of particular concern in states with low shares of enrollees in community-rated plans.

The two states with the slowest growth in premiums for ACA community-rated plans—Mississippi and Michigan—had quite different shares of small business enrollees in such plans in 2024: 38 percent in Mississippi and 83 percent in Michigan. The three states with the fastest such growth—Ohio, Kentucky, and Georgia—faced more than a doubling of premiums for community-rated plans; each of those states had 10 percent of enrollment or less in those plans, with 90 percent or more of small business enrollees in experience-rated plans.

If many healthier enrollees opt out of the community-rated market segment, that segment can become smaller. When that phenomenon of adverse selection occurs and the share of the small-group market that remains in the community-rated pool becomes small, that pool can become higher risk on average, producing higher premiums, CBO’s analysis suggests. Variation in premium growth and in the size of the community-rated market segment among states suggests that premium increases may accelerate once enrollment in community-rated plans becomes sufficiently small.

By CBO’s estimates, the number of states with less than 25 percent of small business enrollees in community-rated plans grew from two in 2020—Ohio and Wyoming—to seven in 2024—with the addition of Georgia, Indiana, Kentucky, Missouri, and South Carolina. In all of those states, premiums for community-rated plans grew at rates well above the national average. When small businesses see very high premiums in community-rated plans, some of those businesses will find it more advantageous to switch to a level-funded plan or other alternative plan, depending on the health claims experience of the employees and their family members.

Other factors are altering the situation at the same time. For example, market concentration among health care providers or among insurers can drive up health care costs and premiums, and state and federal policy changes affect the rules governing health insurance plans. Those factors make determining the causal relationship between premium growth and the share of plans that are community-rated difficult.

Other Empirical Work

Recent research has established that adverse selection does occur in the small-group market but provides inconclusive evidence on the effect on premiums. Studies have examined offer rates for health insurance and premiums to consider how they relate to decisions to retain or move to an experience-rated plan. One study of states with variation in experience rating in the pre-ACA small-group market showed that small businesses in industries with healthier employees were more likely to self-insure than those in industries with less healthy employees.72 Similarly, another study found that small businesses in California were less likely to self-insure (or offer any health insurance at all) after regulation increasing thresholds for stop-loss insurance made it riskier for small firms to self-insure in the state’s small-group market but found no effect on premiums in the fully insured market.73 Finally, a third study found that businesses with healthier employees were more likely to retain grandfathered and transitional plans after the introduction of ACA community-rated plans, resulting in higher premiums for those community-rated plans.74

Considerations in Regulating the Small-Group Market

Rules governing the small-group health insurance market require policymakers to balance several policy objectives (see Table 2). Those rules affect insurers, small businesses, and employees and their dependents in different ways, and their effects vary among states and at the federal level. Differences in regulatory approaches reflect differing priorities regarding affordability, access to coverage, consumer protections, and market stability. As policymakers consider alternatives, information about trends in enrollment and premiums can help identify relevant developments in the small-group insurance market.

Trade-offs are inevitable when trying to pursue diverse goals. Making more affordable options available to some businesses may raise the cost of coverage for others, or it may lead to the introduction of plans that cover a very limited set of services or that have inadequate provider networks. Additionally, some people enrolled in plans that provide fewer benefits or that have caps on covered spending may find themselves faced with high out-of-pocket costs. However, allowing businesses to select only very comprehensive plans may make it more difficult for them and their employees to afford coverage and may require them to purchase more coverage than they would otherwise choose.

---

1. 1. The term “large business households” refers to households that include at least one person working (either full or part time) at a large business, and “small business households” refers to households in which someone works at a small business but nobody works at a large business. Households without a worker are excluded from both categories. The term “household” refers to a health insurance unit (HIU). A single-person household is its own HIU. Otherwise, an HIU is the set of individuals who could be covered by a family plan offered by the employer, that is, a plan that covers an employee and the employee’s dependents.

2. 2. See Centers for Medicare & Medicaid Services, “Market Rating Reforms: State Specific Rating Variations” (September 10, 2024), https://tinyurl.com/4fuuc5xy. Starting in 2026 (after the period of analysis covered in this report), Colorado narrowed its definition of the small-group market by changing it to include businesses of 50 or fewer employees. See Colorado Department of Regulatory Agencies, Division of Insurance, “Average Premium Increases of 13.6% Requested for Small Group Market, Small Employer Definition to Change” (July 31, 2025) https://tinyurl.com/38t6km22.

3. 3. The share of total employees working in businesses in the small-group market was 27 percent in 2011 and in 2024. With the increase that is attributable to the redefinition of the small-group market excluded, the number of workers in just small businesses with fewer than 50 workers grew by 6.8 million over the period and fell to 26 percent of private-sector workers at the end of the period.

4. 4. Individuals who elect not to take up an affordable offer of coverage from an employer are ineligible for ACA marketplace subsidies. Those provisions are specified in 26 U.S.C. § 36B.

5. 5. See Congressional Budget Office, Federal Subsidies for Health Insurance: 2026 to 2036 (July 2026), www.cbo.gov/publication/62539.

6. 6. See Holly Wade and Madeleine Oldstone, Small Business Health Insurance Survey (NFIB Research Center, March 2023), www.nfib.com/small-business-trends-research.

7. 7. John A. Graves and others, “Breadth and Exclusivity of Hospital and Physician Networks in U.S. Insurance Markets,” JAMA Network Open, vol. 3, no. 12 (December 2020), e2029419, https://doi.org/10.1001/jamanetworkopen.2020.29419; Matthew Rae and others, How Narrow or Broad Are ACA Marketplace Physician Networks? (KFF, August 24, 2024), https://tinyurl.com/8tzwkvfm; and Mark A. Hall and Michael J. McCue, Expanding Consumer Health Insurance Options by Easing the Boundaries Between Individual and Small-Group Markets (Commonwealth Fund, October 2021), https://doi.org/10.26099/e50v-eb64.

8. 8. Because most states with income and payroll taxes base those taxes on federal adjusted gross income, employers’ payments for health insurance premiums are typically exempt from state income and payroll taxes.

9. 9. All small businesses could arrange for their employees’ premium contributions to be excluded from taxation. However, doing so requires setting up a Section 125 cafeteria plan, which allows workers to choose between a taxable benefit, such as cash wages, and nontaxable fringe benefits. See Internal Revenue Service, Employer’s Tax Guide to Fringe Benefits, Publication 15-B (April 30, 2026), https://tinyurl.com/mvwcj98r. Not all small businesses go through the process of setting up such a plan. CBO analyzed tabulations from the MEPS-IC to calculate the share of small business employees whose employers had set up such a plan.

10. 10. See Internal Revenue Service, “About Form 7206, Self-Employed Health Insurance Deduction” (June 27, 2026), www.irs.gov/forms-pubs/about-form-7206.

11. 11. See Chris Wheat and Chi Mac, The Burden of Health Insurance Premiums on Small Business (JPMorgan Chase Institute, June 26, 2024), Figure 5, https://tinyurl.com/yc2facj2. The authors calculated an estimated “health insurance payroll burden” by dividing the monthly health insurance payment by the sum of the monthly payroll and the health insurance payment. The analysis found that small businesses with revenues below $600,000 per year had a median payroll burden for health insurance of 11.8 percent, compared with a median burden of 7.1 percent for small businesses with revenues of $2.4 million or more.

12. 12. CBO calculated those amounts using MEPS-IC data for private-sector businesses from 2024. See Agency for Healthcare Research and Quality, MEPS-IC Complete Table Series (2024), “Average Total Single Premium (in dollars) per Enrolled Employee at Private-Sector Establishments That Offer Health Insurance by Firm Size and Selected Characteristics: United States, 2024,” Table I.C.1, and “Average Total Family Premium (in dollars) per Enrolled Employee at Private-Sector Establishments That Offer Health Insurance by Firm Size and Selected Characteristics: United States, 2024,” Table I.D.1, https://tinyurl.com/48kz2bxm.

13. 13. CBO calculated that amount using data from the Bureau of Labor Statistics. See “Employer Costs for Employee Compensation–December 2024” (news release, March 14, 2025), Table 6, www.bls.gov/bls/news-release/ecec.htm#2024.

14. 14. See Agency for Healthcare Research and Quality, MEPS-IC Complete Table Series (2024), “Percent of Private-Sector Employees in Establishments That Offer Health Insurance by Firm Size and Selected Characteristics: United States, 2024,” Table I.B.2, and “Percent of Private-Sector Employees Eligible for Health Insurance That Are Enrolled in Health Insurance at Establishments That Offer Health Insurance by Firm Size and Selected Characteristics: United States, 2024,” Table I.B.2.a.(1), https://tinyurl.com/74826va3.

15. 15. For research finding that low-income workers often value health insurance at less than its cost, see Amy Finkelstein, Nathaniel Hendren, and Mark Shepard, “Subsidizing Health Insurance for Low-Income Adults: Evidence From Massachusetts,” American Economic Review, vol. 109, no. 4 (April 2019), pp. 1530–1567, https://doi.org/10.1257/aer.20171455.

16. 16. See Michael S. Dworsky and others, “The Effect of the Minimum Wage on Employer-Sponsored Insurance for Low-Income Workers and Dependents,” American Journal of Health Economics, vol. 8, no. 1 (Winter 2022), pp. 99–126, https://doi.org/10.1086/716198.

17. 17. See Irena Dushi and Marjorie Honig, “Price and Spouse’s Coverage in Employee Demand for Health Insurance,” American Economic Review, vol. 93, no. 2 (May 2003), pp. 252–256, https://doi.org/10.1257/000282803321947146.

18. 18. Premium tax credits are calculated as the difference between the benchmark premium (that is, the premium for the second-lowest-cost silver plan available in a region) and a specified maximum percentage of household income that is expected to be paid out of pocket toward the premium. For what constitutes an affordable offer of health coverage by an employer, see HealthCare.gov, Glossary, “Affordable Coverage” (accessed May 22, 2026), www.healthcare.gov/glossary/affordable-coverage/.

19. 19. For a discussion about such workers, see Richard A. Hirth and others, “Worker Preferences, Sorting and Aggregate Patterns of Health Insurance Coverage,” International Journal of Health Care Finance and Economics, vol. 6, no. 4 (December 2006), pp. 259–277, https://doi.org/10.1007/s10754-007-9007-x.

20. 20. CBO analyzed data from the National Compensation Survey, conducted by the Bureau of Labor Statistics, to identify industry categories in which fewer small businesses had access to health coverage. See Bureau of Labor Statistics, “Data Retrieval Tools” (accessed August 12, 2026), www.bls.gov/data/tools.htm.

21. 21. See Jean Marie Abraham, Thomas DeLeire, and Anne Beeson Royalty, “Access to Health Insurance at Small Establishments: What Can We Learn From Analyzing Other Fringe Benefits?” Inquiry, vol. 46, no. 3 (Fall 2009), pp. 253–273, https://doi.org/10.5034/inquiryjr‌‌nl_46.03.253.

22. 22. For an analysis of how competition among agents and brokers can affect offers of health insurance and premiums paid by small businesses, see Pinar Karaca-Mandic, Roger Feldman, and Peter Graven, “The Role of Agents and Brokers in the Market for Health Insurance,” Journal of Risk and Insurance, vol. 85, no. 1 (March 2018), pp. 7–34, https://doi.org/10.1111/jori.12139. The ACA called for the establishment of exchanges under the Small Business Health Options Program in all states to make it easier for small businesses to select among health plan options; however, in many states, no such platform is available. In those states, interested small businesses are directed to agents, brokers, and insurers. See Centers for Medicare & Medicaid Services, “SHOP Resources” (updated September 10, 2024), www.cms.gov/marketplace/agents-brokers/shop-resources.

23. 23. CBO’s analysis of coverage was restricted to households that contained at least one worker, including small business owners who had at least one employee.

24. 24. For a discussion of the importance of health insurance to entrepreneurship, see Congressional Budget Office, Federal Policies in Response to Declining Entrepreneurship (December 2020), pp. 19–20, www.cbo.gov/publication/56906. For a description of the number of small business owners and self-employed people covered by marketplace plans in 2022, see Department of the Treasury, Office of Tax Analysis, Affordable Care Act Marketplace Coverage for the Self-Employed and Small Business Owners (September 20, 2024), https://tinyurl.com/yd54dmme.

25. 25. Those changes were part of the Families First Coronavirus Response Act (Public Law 116-127). The Consolidated Appropriations Act, 2023 (P.L. 117-328) ended the continuous eligibility provisions on April 1, 2023.

26. 26. Those changes were included in the American Rescue Plan Act of 2021 (P.L. 117-2) and later extended through calendar year 2025 in the 2022 reconciliation act (P.L. 117-169).

27. 27. Congressional Budget Office, Who Went Without Health Insurance in 2019, and Why? (September 30, 2020), www.cbo.gov/publication/56504.

28. 28. For its analysis of the uninsured population, CBO used its models that produce baseline budget projections. Those models classify uninsured people into categories based on what subsidies they would be eligible for if they were insured—information that is generally not available from other data sources.

29. 29. See Congressional Budget Office, Federal Subsidies for Health Insurance, 2026 to 2036 (July 2026), p. 6, www.cbo.gov/publication/62539.

30. 30. To analyze the cost of state and local premium taxes, CBO examined 2024 data from the Supplemental Health Care Exhibit filed annually by insurers with the National Association of Insurance Commissioners. CBO accessed the data using the Health Coverage Portal TM by Mark Farrah Associates, available at www.markfarrah.com/products/health-coverage-portal/.

31. 31. For additional details on the risk adjustment program, see Centers for Medicare & Medicaid Services, “Premium Stabilization Programs” (accessed April 30, 2026), https://tinyurl.com/mpss8vnu.

32. 32. Guaranteed issue is the requirement that insurers offer policies to all small businesses that apply regardless of the health status of employees and their dependents, and guaranteed renewal requires that insurers offer to renew a policy as long as premiums are paid on time and the insurer still offers the policy to other businesses in that market.

33. 33. The ACA requires health insurance plans to include coverage for 10 essential health benefits. Additionally, states establish their own benefit requirements. For example, in 2014 and 2015, 5 states required weight-loss services; 19, infertility treatments; and 45, chiropractic care. See Janet Weiner and Christopher Colameco, Essential Health Benefits: 50-State Variations on a Theme (Leonard Davis Institute of Health Economics and Robert Wood Johnson Foundation, October 2014), https://tinyurl.com/4madxx6r.

34. 34. To be eligible, plans must be certified by the Small Business Health Options Program. See Internal Revenue Service, “Small Business Health Care Tax Credit and the SHOP Marketplace” (September 30, 2025) https://tinyurl.com/mtza8e2f; and HealthCare.gov, “The Small Business Health Care Tax Credit” (accessed May 22, 2026), https://tinyurl.com/y9cbassr.

35. 35. In 2024, about 55 percent of private-sector businesses (of all sizes) were less than 10 years old and thus were not in existence when the ACA was implemented in 2014. Such businesses could not have sponsored a grandfathered or transitional plan. See Bureau of Labor Statistics, “Number of Private Sector Establishments by Age, Total Private,” Table 5 (accessed May 22, 2026), https://www.bls.gov/bdm/bdmage.htm.

36. 36. For more information about grandfathered plans, see HealthCare.gov, “Grandfathered Health Insurance Plans” (accessed May 22, 2026), https://tinyurl.com/5xzbn7n5.

37. 37. To determine the number of states that allowed transitional plans, CBO analyzed bulletins issued by states’ insurance departments. Transitional plans are sometimes called grandmothered plans. For more information about transitional plans, see Mark A. Hall and Michael J. McCue, “How Do Noncompliant Health Plans Affect the Market?” To the Point (blog entry, November 15, 2017), https://tinyurl.com/2zpw6hn3.

38. 38. See Vanessa C. Forsberg and Ryan J. Rosso, Applicability of Federal Requirements to Selected Health Coverage Arrangements, Report R46003, version 7 (Congressional Research Service, November 13, 2019), Appendix, www.congress.gov/crs-product/R46003.

39. 39. Many states impose restrictions on stop-loss coverage for small businesses. See Department of Labor, Employee Benefits Security Administration, “Guidance on State Regulation of Stop-Loss Insurance,” Technical Release No. 2014-01 (November 6, 2014), https://tinyurl.com/yfrejwwc.

40. 40. Although self-insured plans are exempt from many of the ACA’s regulations, they are required (with some exceptions for grandfathered plans in place since 2010 or earlier) to refrain from imposing exclusions for those with preexisting conditions, ensure no discrimination based on health status, adhere to annual limits on out-of-pocket maximums, limit waiting periods to no more than 90 days, support participation in clinical trials without undue restrictions, eliminate lifetime limits on the dollar value of essential health benefits if such benefits are provided, refrain from imposing annual limits on the dollar value of essential health benefits for plans put in place in 2014 or later, provide 30 days’ notice and evidence of fraud before rescinding coverage, cover recommended preventive health services without imposing cost-sharing requirements on enrollees, extend coverage to dependents until age 26, provide clear explanation of coverage documents, establish internal claims appeals processes for review, ensure choice of health care professionals, and uphold other protections outlined in the Public Health Service Act. For additional information on the ACA’s rules for self-insured plans, see Bernadette Fernandez, Vanessa C. Forsberg, and Ryan J. Rosso, Federal Requirements on Private Health Insurance Plans, Report R45146, version 11 (Congressional Research Service, March 12, 2026), Table 1, www.congress.gov/crs-product/R45146.

41. 41. AHPs and union plans are types of multiple-employer welfare arrangements, as defined by the Department of Labor, affecting how the plans are regulated at the federal and state levels. See Department of Labor, Employee Benefits Security Administration, Multiple Employer Welfare Arrangements Under the Employee Retirement Income Security Act (ERISA): A Guide to Federal and State Regulation (April 2022), https://tinyurl.com/2cjkjnn5.

42. 42. See Congressional Budget Office, How CBO and JCT Analyzed Coverage Effects of New Rules for Association Health Plans and Short-Term Plans (January 2019), www.cbo.gov/publication/54915.

43. 43. See 29 U.S.C. § 1002(5). The criteria used by the Department of Labor to determine whether an association qualifies as an employer include “(1) whether the group or association has business or organizational purposes and functions unrelated to the provision of benefits (the 'business purpose' standard); (2) whether the employers share some commonality of interest and genuine organizational relationship unrelated to the provision of benefits (the 'commonality' standard); and (3) whether the employers that participate in a benefit program, either directly or indirectly, exercise control over the program, both in form and substance (the 'control' standard).” See Definition of “Employer”—Association Health Plans, 88 Fed. Reg. 87968, 87969 (proposed December 20, 2023), footnote 18, https://tinyurl.com/mukt9b4j.

44. 44. See Definition of “Employer” Under Section 3(5) of ERISA—Association Health Plans, 83 Fed. Reg. 28912 (June 21, 2018), https://tinyurl.com/nhjpm2vu. For more information, see Fritz Busch and Jason Karcher, Association Health Plans After the Final Rule (Milliman, August 2018), http://tinyurl.com/y7nmfoyv.

45. 45. New York v. U.S. Dep’t of Labor, 363 F. Supp. 3d 109 (D.D.C. 2019).

46. 46. See New York v. U.S. Dep’t of Labor, No. 19-5125 (D.C. Cir. May 30, 2024); Department of Labor, Employee Benefits Security Administration, “Fact Sheet: Department of Labor Rescinds Invalidated Rule on Association Health Plans” (April 29, 2024), https://tinyurl.com/3arrfrns; and Definition of “Employer”—Association Health Plans,” Final Rule, Rescission, 89 Fed. Reg. 34106 (April 30, 2024), https://tinyurl.com/2p9wmb2c.

47. 47. See Congressional Budget Office, How CBO and JCT Analyzed Coverage Effects of New Rules for Association Health Plans and Short-Term Plans (January 2019), www.cbo.gov/publication/54915.

48. 48. See Sabrina Corlette, “Final Rule Rescinds Trump-Era Association Health Plans,” Health Affairs Forefront (blog entry, April 30, 2024), https://doi.org/10.1377/forefront.20240430.42514.

49. 49. According to survey data for 2024, 9 percent of enrollees in union plans worked in businesses in the construction industry. See Agency for Healthcare Research and Quality, MEPS-IC Complete Table Series (2024), “Percent of Private-Sector Employees That Are Enrolled in Plans Offered Through a Union (Multi-Employer Health Plan [MEHP] or a Trade or Business Association [AHP]) or Neither by Selected Characteristics: United States, 2024, Table I.B.2.d, https://tinyurl.com/3zez4j62.

50. 50. For more information about multiemployer health plans, see International Foundation of Employee Benefit Plans, “Understanding Multiemployer Plans” (accessed May 12, 2026), https://tinyurl.com/5c983e3u.

51. 51. See ERISA Working Group of the Regulatory Framework (B) Task Force of the Health Insurance and Managed Care (B) Committee, Health and Welfare Plans Under the Employee Retirement Income Security Act: Guidelines for State and Federal Regulation (National Association of Insurance Commissioners, January 2022), p. 47, https://tinyurl.com/mryrudwp.

52. 52. That approach is not possible in Maine, where PEOs are required to enroll small businesses in ACA community-rated products. See Emma Hoo, California Policy Perspectives on Association Health Plans (California Health Care Foundation, March 2021), p. 7, https://tinyurl.com/mujmcw8t. Maryland and Wyoming also prohibit PEOs from offering large-group insurance to small business clients. See Maryland Insurance Administration, Professional Employer Organizations (PEO) Study (January 3, 2025), https://tinyurl.com/yh22cufy. 

53. 53. See Laurie Bassi and Dan McMurrer, PEO Clients: Who They Are, Where They Are, and What They Do (National Association of Professional Employer Organizations, October 2025), Table 1, https://napeo.org/intro-to-peos/industry-research-data/. 

54. 54. See Laurie Bassi and Dan McMurrer, PEO Clients: Who They Are, Where They Are, and What They Do (National Association of Professional Employer Organizations, October 2025), Table 4, https://napeo.org/intro-to-peos/industry-research-data/. 

55. 55. See HealthCare.gov, “Health Reimbursement Arrangements (HRAs) for Small Employers” (accessed May 22, 2026), www.healthcare.gov/small-businesses/learn-more/qsehra/.

56. 56. See Health Reimbursement Arrangements and Other Account-Based Group Health Plans, 26 C.F.R. parts 1 and 54, 29 C.F.R. parts 2510 and 2590 and 45 C.F.R. parts 144, 146, 147, and 155, https://tinyurl.com/5absbd23. See also Katie Keith, “Final Rule on Health Reimbursement Arrangements Could Shake Up Markets,” Health Affairs Forefront (blog entry, June 14, 2019), https://doi.org/10.1377/forefront.20190614.388950. After the implementation of the ACA but before the QSEHRA and ICHRA options were introduced, employers could offer employees funds through an HRA only if the HRA was paired with an offer of group health insurance.

57. 57. Employees’ premium contributions toward plans in the ACA marketplace cannot be excluded from income and payroll taxation. See Department of Health and Human Services, “New Health Coverage Options for Employers and Employees: Individual Coverage and Excepted Benefit Health Reimbursement Arrangements” (June 13, 2019), Question 9, https://tinyurl.com/bnhx8e2y.

58. 58. For information about who qualifies for premium tax credits, see Internal Revenue Service, “The Premium Tax Credit—The Basics (July 22, 2026), https://tinyurl.com/3py8af43. For current affordability standards, see HealthCare.gov, “Affordable Coverage” (accessed May 14, 2026), www.healthcare.gov/glossary/affordable-coverage/.

59. 59. For information about requirements for employers offering ICHRAs, see HealthCare.gov, “Individual Coverage Health Reimbursement Arrangements (HRAs)” (accessed July 24, 2026), https://tinyurl.com/49bx4mnx.

60. 60. In the four states that consider businesses with up to 100 employees to be small businesses, some of those businesses are subject to the ACA’s employer mandate requirements. If such businesses offer an HRA contribution rather than a health plan and the contribution does not meet the affordability threshold set by the ACA, they can face penalties. See Internal Revenue Service, “Employer Shared Responsibility Provisions” (November 4, 2025), https://tinyurl.com/3hbvy25f.

61. 61. Employees cannot combine an ICHRA contribution and ACA premium tax credits to purchase nongroup coverage. To receive premium tax credits, the employee must decline the ICHRA contribution.

62. 62. See Internal Revenue Service, “Questions and Answers on the Premium Tax Credit,” Question 23 (accessed July 22, 2026), https://tinyurl.com/bzcftjjr.

63. 63. Employers’ contribution amounts are capped for QSEHRA plans. See HealthCare.gov, “Deciding Between Group Coverage and an HRA?” (accessed July 24, 2026), https://tinyurl.com/2vhp47yr.

64. 64. As noted earlier, in the four states with the small-group cutoff at 100 employees, some small businesses with more than 50 employees are considered “applicable large employers” under the ACA and are therefore subject to the ACA’s employer mandate requirements.

65. 65. For more detail about CBO’s approach, see the appendix.

66. 66. For more information on the data collection process for the MEPS-IC in 2020, see Samuel H. Zuvekas and David Kashihara, “The Impacts of the COVID-19 Pandemic on the Medical Expenditure Panel Survey,” American Journal of Public Health, vol. 111, no. 12 (December 2021), pp. 2157–2166, https://doi.org/10.2105/ajph.2021.306534. 

67. 67. A plan can lose its grandfathered or transitional status if it changes significantly, for instance, by increasing cost sharing or deductibles beyond certain thresholds or removing coverage for treatment of a particular medical condition. Once the status is lost, it cannot be regained (see 45 C.F.R. § 147.140(g)). For more information about grandfathered plans, see HealthCare.gov, “Grandfathered Health Insurance Plans” (accessed July 19, 2026), https://tinyurl.com/bp556e2h.

68. 68. The Agency for Healthcare Research and Quality sent CBO special tabulations of the percentage of enrollment in AHPs and union plans by state along with standard errors for the estimates. The standard errors were too high to rely on the state-specific percentages, but the data were reliable at the national level.

69. 69. KFF’s Employer Health Benefits Survey reports some information on offers of support for nongroup coverage (including ICHRAs and other options) and plans to offer such support in the future. The MEPS-IC began asking about offers of funds through QSEHRAs and ICHRAs but did not report the responses during the years CBO analyzed. Neither survey has information about take-up of those mechanisms to purchase nongroup coverage.

70. 70. See HRA Council, Growth Trends for ICHRA & QSEHRA, 2024–2025 (June 17, 2025), p. 6, www.hracouncil.org; and Gary Claxton and others, Employer Health Benefits 2024 Annual Survey (KFF, October 9, 2024), pp. 58–60, https://tinyurl.com/5bvc4hhn.

71. 71. Information on the premiums for ACA community-rated plans is available in risk adjustment data compiled by the Centers for Medicare & Medicaid Services. Data are omitted for states that merged their nongroup and small-group markets for the risk adjustment program. CBO estimated growth in overall premiums in the entire small-group market using MEPS-IC data. CBO estimated growth in experience-rated plans by calculating the weighted average for premiums in the entire small-group market and subtracting enrollment and average premiums for ACA community-rated plans. For details, see the appendix.

72. 72. Erin Trish and Bradley Herring, “Does Limiting Allowable Rating Variation in the Small Group Health Insurance Market Affect Employer Self-Insurance?” Journal of Risk and Insurance, vol. 85, no. 3 (September 2018), pp. 607–633, https://doi.org/10.1111/jori.12184.

73. 73. Mark K. Meiselbach and Matthew Eisenberg, “Financial Risk and the Decision of Small Employers to Self-Fund Health Insurance: Evidence From Stop Loss Regulation in California,” American Journal of Health Economics, vol. 11, no. 2 (April 2025), pp. 274–299, https://doi.org/10.1086/728588. 

74. 74. Sebastian Fleitas, Gautam Gowrisankaran, and Anthony Lo Sasso, “Incumbent Regulation and Adverse Selection: You Can Keep Your Health Plan, but at What Cost?” Journal of Public Economics, vol. 205 (January 2022), article 104556, https://doi.org/10.1016/j.jpubeco.2021.104556. The authors also estimate that employers with higher turnover were more likely to retain health insurance.

Appendix: Data Sources and Methods for Analysis of Market Segments

For this analysis, the Congressional Budget Office developed a new methodology that combines several data sources to measure enrollment in three market segments: the community-rated plans established by the Affordable Care Act (ACA), grandfathered and transitional plans, and self-insured and other plans that do not fall into those two categories.1 CBO chose the market segments for comparison because the plans in them need to meet different legal requirements for premium rating and benefits. Consequently, differences in enrollment and costs can help identify policies’ differing effects. CBO developed its approach to help overcome concerns about underreporting of enrollment in self-insured plans, to examine more clearly the decline in enrollment in grandfathered and transitional plans over time, and to estimate enrollment in some less common types of plans.

To estimate the size of each market segment, CBO combined survey and administrative data sources that cover some or all of the markets and compared their enrollment. Because the data from the sources are collected for different purposes and have different units of measurement, CBO adjusted the data in each source to make the sources comparable.

Key Data Sources

To measure enrollment and premiums in the small-group market, CBO primarily relied on the following three data sources:

• Medical Expenditure Panel Survey-Insurance Component. The Medical Expenditure Panel Survey-Insurance Component (MEPS-IC) provides information on total enrollment in the small-group market across all three market segments. An annual survey of employers conducted by the Agency for Healthcare Research and Quality (AHRQ) and the Census Bureau, the MEPS-IC provides national and state-level estimates of total enrollment in health insurance sponsored by small employers.2,3
• Medical Loss Ratio Public Use Files. Medical loss ratio (MLR) data, compiled by the Centers for Medicare & Medicaid Services (CMS), provide information on combined enrollment in ACA community-rated plans as well as grandfathered and transitional plans. The data consist of insurers’ filings with the Department of Health and Human Services (HHS), which the ACA requires to verify that insurers in the small-group market are complying with rules requiring them to spend at least 80 percent of premium dollars on medical care and quality improvement. (Employers who self-insure are not considered insurance issuers and are not included in the MLR data.) Each insurer reports enrollment in all fully insured health plans in the individual (nongroup), small-group, and large-group markets.
• Summary Report on Permanent Risk Adjustment Transfer. The risk adjustment data provide information on enrollment in ACA community-rated plans. CMS publishes these data in an annual report on payment transfers between insurers as part of HHS’s risk adjustment program for the individual (nongroup) and small-group markets.

CBO also used other survey data, including the Medical Expenditure Panel Survey-Household Component (MEPS-HC) and the 2018 Current Population Survey (CPS), to help convert enrollment in the MEPS-IC to a unit of measurement for enrollment comparable with those in the MLR and risk adjustment data.

CBO also used survey data from KFF’s Employer Health Benefits Survey to provide additional benchmarks for comparing CBO’s estimates of the prevalence of self-insured and level-funded plans, to identify the types of health plans offered to small business workers, and to answer other questions.4

Methods for Estimating Enrollment by Market Segment

Measuring enrollment in the small-group market by segment is challenging because administrative requirements differ, data sources often do not distinguish between markets, and employers sometimes do not respond accurately to survey questions and reporting requirements.5 Moreover, as a survey source, the MEPS-IC probably undercounts self-insured plans because, according to AHRQ, small employers in level-funded plans often do not realize they are self-insured when responding to the survey.6 KFF also acknowledges issues with reporting by self-insured and level-funded plans in its annual Employer Health Benefits Survey.7

To overcome those limitations, CBO combined several data sources to estimate the size of each market segment (see Figure A-1). First, the agency obtained information on enrollment in ACA community-rated plans from the risk adjustment data. Then, CBO estimated enrollment in fully insured grandfathered and transitional plans by calculating the difference between enrollment in the MLR and the risk adjustment data.8 Next, CBO estimated enrollment in self-insured and other plans as the difference between enrollment in the MEPS-IC and MLR data. That difference captures enrollment in self-insured plans and other experience-rated options, such as many association health plans (AHPs) and union plans.

Enrollment in AHPs occurs in different market segments. AHPs that meet the requirements of establishing a group health plan that is an employee welfare benefit plan (consistent with 29 C.F.R. § 2510.3-1 or with other applicable guidance) and that cover more than 50 employees for multiple small firms would file in the large-group market rather than the small-group market. (In CBO’s presentation of small-group enrollment, enrollment in such AHPs shows up in the “self-insured and other” category; see Figure 8.) Fully insured AHPs that do not meet those requirements would be subject to MLR rules for the small-group market. (Enrollment in such plans would be included in the ACA community-rated segment or, if they have been in operation long enough, in the grandfathered and transitional segment.) AHPs that formed under “Pathway 2” rules during the brief period when the 2018 rule for AHPs was in effect most likely would be included in the “self-insured and other” category because they were not subject to MLR rules or risk adjustment.

Many union plans are self-insured; enrollment in such plans by small business employees and their dependents would show up in the “self-insured and other” category. Fully insured union plans might show up in the ACA community-rated or grandfathered and transitional segments if they cover 50 or fewer employees.

Because AHPs and union plans can fall into more than one market segment, CBO estimated enrollment in them using the MEPS-IC national-level estimates of the share of plans obtained through a trade association or through a union. That breakdown was feasible only at the national level, not the state level, because of concerns with the reliability of survey estimates of both types of plans at the state level. In this report, state-level analyses use only the broader category of “self-insured and other” plans rather than breaking those plans up into AHPs, union plans, and self-insured plans.

Although offering health benefits to retirees is not common at small businesses, a few do. Those retired workers (and their covered dependents) who are enrolled in fully insured small-group plans are included in the MLR and risk adjustment data; however, they are not included in the usual count of enrolled workers in the MEPS-IC. Separately, the MEPS-IC includes questions about the number of retired policyholders in single and family plans sponsored by small businesses. CBO assumed each family plan for a retiree covered two people. CBO added those policyholders and dependents to its estimates of total enrollment in small business health plans (see Figure 8).

As noted later in this appendix, CBO is unable to estimate enrollment in self-insured and other plans in 2020 because of the effect of temporary furloughs on the MEPS-IC estimate of the total number of employees enrolled in health plans in that year.

Comparison With Other Estimates of Enrollment by Market Segment

CBO’s estimates of the share of workers who are policyholders of small business health plans that are categorized as self-insured and other plans differ from estimates of the share of small business workers covered by self-insured and level-funded plans from KFF’s Employer Health Benefits Survey as well as from estimates of small business workers enrolled in self-insured plans from the MEPS-IC (see Figure A-2). (For comparability with other estimates, CBO’s estimates here are of policyholders, rather than enrollees, including dependents.)

The comparisons show that CBO’s methodology can capture the increase in self-insured plans that is not discernible in the MEPS-IC data. For 2015 through 2017, CBO’s estimates of self-insured and other plans are slightly higher than the MEPS-IC estimates of the share of employees in businesses with self-insured plans—which probably reflects the fact that some union plans and AHPs are included in CBO’s estimates. For later years, as level-funded plans became more prevalent, CBO’s estimates of self-insured and other plans increase, whereas the MEPS-IC estimates remain relatively steady. The most likely reason for the discrepancy is that, in the MEPS-IC, small businesses offering level-funded plans did not realize that they should identify their plans as self-insured plans. Level-funded plans look very much like a traditional health insurance plan and are often administered by the same companies that sell fully insured plans.

KFF’s estimates of workers at businesses with 3 to 49 workers enrolled in self-insured or level-funded plans increased rapidly from 2018 through 2021, in contrast to the MEPS-IC’s estimates of very modest growth in the share of workers in plans that were self-insured in those years. CBO’s estimates of self-insured and other plans for 2018 and 2019 are higher than KFF’s estimates. For 2021 through 2023, CBO’s estimates are roughly comparable to the KFF estimates. CBO’s estimate for 2024 is somewhat higher than KFF’s and three times as high as the MEPS-IC’s estimate.

How CBO Used the Data Underlying Its Estimates of Enrollment in Market Segments

CBO also made other small adjustments to the data to improve the quality of estimates in each market segment, particularly at the national level, where state-level inconsistencies in the data are less influential.

MEPS-IC Data

To compare the MEPS-IC with other data sources, CBO converted the MEPS-IC’s estimate of workers enrolled in health plans offered by their small employer to life-years using estimates of enrollees per policy from the household component of the MEPS, or MEPS-HC, and Current Population Survey (CPS) by the Bureau of Labor Statistics. The MEPS-IC reports the number of single, employee-plus-one, and family policies but does not collect data on covered dependents. If the counts of policyholders reflect annual averages, total covered life-years (member-months divided by 12) can be estimated by multiplying policies by the enrollees-per-policy ratio.

To estimate that ratio, CBO used the MEPS-HC’s counts of dependents. Single plans have one enrollee per policy. For nonsingle policies, CBO combined self-plus-one and family policies and calculated the enrollees-per-policy ratio in two steps. First, using MEPS-HC data, CBO determined the ratio of dependents per small-group nonsingle policy, weighting dependents by months enrolled to approximate life-years.9 Because of small sample sizes, data from 2014 to 2021 were pooled using a constant ratio. Second, CBO adjusted the MEPS-HC’s national measure to state-level estimates using the 2018 CPS’s health insurance unit sizes for small-group enrollees, with family sizes remaining stable over time. Those ratios were applied to the MEPS-IC’s counts of policies to estimate the total covered life-years by state and year.10

Except for answers in 2020, CBO interprets answers to survey questions on enrollment “during a typical pay period” as reporting the yearly average and assumes that enrollment remained stable after the survey.11 The MEPS-IC’s questions to employers about enrollment during a typical pay period may introduce measurement error in two ways. First, respondents may interpret “typical pay period” differently from the actual yearly average. Second, responses may be biased by the survey’s timing. If enrollment changes significantly in subsequent months, the MEPS-IC may reflect a point-in-time estimate rather than an annual average.

Although the MEPS-IC excludes some employers and includes incorporated self-employed sole owners with no employees, CBO assumes that the omissions and additions have only a negligible effect on estimates of the small-group market.12 Incorporated self-employed sole owners with no employees generally cannot buy small-group plans.13

Medical Loss Ratio Data

MLR data from CMS report enrollment in life-years. That measure includes both policyholders and dependents. CBO adjusted MLR enrollment numbers in particular states to account for filing practices there and for obvious filing errors. Before 2017, for businesses with 100 or fewer employees, some states reported enrollment in the small-group section of the MLR data, even though their state small-group laws applied only to firms with 50 or fewer employees.14 CBO identified those states using various methods, including consulting with CMS, examining state department of insurance bulletins, and identifying unusually large decreases in a state’s MLR enrollment or policies per group in 2016 or 2017 that may have been caused by correcting the business sizes reported in the small-group MLR filings.15 MLR enrollment for those states was deflated using the MEPS-IC ratio of enrollment in businesses with fewer than 50 employees to businesses with fewer than 100 employees. Lastly, enrollment numbers in Minnesota in 2021 and Wisconsin in 2014 are questionably low and high, respectively; consequently, CBO used enrollment figures from the Supplemental Health Care Exhibit, which is a preliminary MLR filing, instead of MLR data for those state-years.16

Risk Adjustment Data

CBO converted risk adjustment data on enrollment to covered life-years using data provided by the Center for Consumer Information and Insurance Oversight at CMS. Public datasets on risk adjustment report enrollment in billable member months.17 To facilitate the conversion, CMS provided state-level data from which CBO calculated the ratios of billable member-months to total member months (including nonbillable individuals) by state for 2021.18 Assuming that those ratios did not change over time, CBO applied them to all years. Total member months were then divided by 12 to get total covered life-years.

For some states in some years, CBO needed to separate out enrollment in the small-group and individual (nongroup) markets because those states merged data for those two markets in the states’ risk adjustment programs.19 To estimate enrollment in small-group community-rated plans, CBO applied the ratio of individual to small-group market enrollment in the MLR data to those states’ merged figures. That approach relied on the assumption that the share of covered lives in small-group plans is roughly the same for both community-rated plans and grandfathered and transitional plans.

How CBO Resolved Inconsistencies When Comparing Enrollment Across Data Sources

Inconsistencies in state-level data are small, and they average out nationally in CBO’s estimates. Some state-year combinations show higher enrollment in MLR data than in MEPS-IC data, creating a negative measure for the market segment for self-insured and other plans. Such instances accounted for 2 percent of total small-group enrollment in 2014 and less than 1 percent in later years, declining from 16 cases in 2014 to none in 2024. Early mismatches likely stem from states’ defining the small-group market differently for MLR filings. Some may also reflect the stable ratio of total enrollees (including dependents) to policyholders that CBO assumed when estimating covered life-years from the number of policyholders in the MEPS-IC data. If in reality that ratio changed over time—for example, if workers included more or fewer dependents on family policies—the assumption may have introduced some error. Finally, such instances may reflect variation in survey responses in the MEPS-IC data.

Additionally, for some states in some years, the risk adjustment data show higher enrollment than the MLR data, resulting in negative figures for enrollment in the market segment for grandfathered and transitional plans, though differences are small—averaging 2 percent of total covered life-years in a state’s small-group market. The negative figures occur mainly in states that banned transitional plans, leaving only a small number of grandfathered plans remaining in the MLR data. CBO attributes the mismatches to minor reporting differences and considers them insignificant for national estimates.

Finally, both the risk adjustment and MLR data include retirees and their dependents in plans in the small-group market, but retirees and their dependents are excluded from state-level MEPS-IC measures of workers enrolled in a plan through their employer. Few small businesses offer health benefits to their retirees, and MEPS-IC estimates of the number of retirees with such coverage vary substantially from year to year even at the national level. CBO made an adjustment at the national level but was unable to do so at the state level. Therefore, in a small number of instances, state-level enrollment in the MLR data exceeded the MEPS-IC estimates.

How CBO Estimated Premiums by Market Segment

Premiums for ACA community-rated plans are drawn from CMS’s risk adjustment reports and are calculated as the national average of state average premiums per billable member month, weighted by total billable member months in each state. Premiums in the overall small-group market are based on CBO’s estimate of average premiums per life-year using MEPS-IC data for small businesses (those with fewer than 50 employees in most states and, starting in 2016, businesses with fewer than 100 employees in California, Colorado, New York, and Vermont), converting policyholders to total life-years of enrollment using average life-years per policy from CMS’s medical loss ratio data. Benefits and provider networks may differ for plans in different market segments as well as within the same market segment.

Employment-Based Coverage During the COVID-19 Pandemic

The beginning of the COVID-19 pandemic in 2020 brought disruptions in employment and economic activity. Initial estimates of the effect of the pandemic on employment and health insurance coverage predicted that many people would lose employment-based coverage and become uninsured.20 Several government interventions helped mitigate those effects. Most notably for small businesses, the Paycheck Protection Program provided businesses with funds to cover up to eight weeks of payroll costs, including employers’ payment of health insurance premiums for workers.21

According to the Business Response Survey in 2020, conducted by the Bureau of Labor Statistics, many employers continued to provide health insurance benefits for furloughed workers for at least some period of time.22 Workers who were furloughed but still receiving health insurance from their employer may not have been counted as employees of businesses surveyed in the MEPS-IC nor as enrollees in employer-sponsored plans during that time period. The omission of those furloughed employees still covered by small business health plans may explain the drop in MEPS-IC estimates of the total number of small business workers enrolled in coverage through their employer, when administrative data sources showed much smaller decreases.

---

1. 1. The analysis of enrollment by market segment excludes enrollment in most professional employer organizations as well as enrollment in the nongroup market with support from a qualified small employer health reimbursement arrangement (QSEHRA) or an individual coverage health reimbursement account (ICHRA). In addition, data limitations make it difficult to differentiate between grandfathered and transitional plans and to analyze association health plans and union plans at the state level.

2. 2. MEPS-IC estimates of private-sector enrollment at small businesses are publicly available. For estimates of such enrollment at the public-sector agencies small enough to fall into the small-group category, AHRQ provided CBO with special tabulations from the MEPS-IC.

3. 3. The MEPS-IC provides state-level estimates for both businesses with fewer than 50 employees and businesses with fewer than 100. The availability of state-level data allowed CBO to follow each state’s definition of the small-group market in most of the analysis done for this report. Unless otherwise specified, any reference to the "small-group market" includes businesses with 50 or fewer employees in most states. In California, Colorado, New York, and Vermont, starting in 2016, “small-group” refers to businesses with 100 or fewer employees. The categories in the survey data differ slightly, but they are close to the legal definitions.

4. 4. See KFF, Employer Health Benefits 2024 Annual Survey (October 9, 2024), https://tinyurl.com/466kxenc. For information on KFF’s public use files from the Employer Health Benefits Survey, see KFF, “KFF EHBS Data Use Form” (accessed April 30, 2026), www.kff.org/ehbs-data-use-form/.

5. 5. For example, as an administrative source, the MLR data do not distinguish between enrollment in ACA community-rated plans and grandfathered and transitional plans and do not include any information on self-insured and other plans. CBO investigated another potential administrative data source on health plans collected by the Department of Labor. Employee welfare benefit plans such as health plans sponsored by employers are required to file a Form 5500 with the Department of Labor, but plans with fewer than 100 participants are often exempt; therefore, Form 5500 does not provide good data on self-insured plans sponsored by small businesses. See Department of Labor, Employee Benefits Security Administration, “Form 5500 Series” (accessed May 5, 2026), https://tinyurl.com/2txhkcvd.

6. 6. Agency for Healthcare Research and Quality, correspondence (September 18, 2023).

7. 7. Gary Claxton and others, Employer Health Benefits: 2024 Annual Survey (KFF, October 2024), p. 167, https://tinyurl.com/3e4psmu4.

8. 8. Some grandfathered plans are self-insured and thus are not included in the MLR data and are instead counted in the market segment for self-insured and other plans. If an AHP organization does not qualify as an “employee welfare benefit plan,” data for enrollees working at small businesses (and their dependents) must be reported in the section of MLR data on small-group plans. However, those plans are regulated as any community-rated plan and will also be included in the risk adjustment data. Therefore, the residual measures only fully insured grandfathered and transitional plans. See Centers for Medicare & Medicaid Services, “Medical Loss Ratio (MLR) Annual Reporting Form: Filing Instructions for the 2022 MLR Reporting Year” (accessed August 11, 2026), pp. 6–7, https://tinyurl.com/3jx26bfw. If an association health plan is self-insured, the plan is exempt from community rating and MLR filing regardless of the association’s qualification as an employer.

9. 9. Policies in the MEPS-HC were limited to those for private-sector workers who reported having coverage through a business with fewer than 50 employees and only one location. Including respondents who reported working at a business with fewer than 50 workers and more than one location would likely include workers not in the small-group market—for example, workers in a fast-food chain that has a small number of employees at a worksite but that is a business with more than 50 employees. For the states with an expanded definition of small businesses, CBO followed the same procedure for workers in businesses with fewer than 100 employees.

10. 10. The MEPS-HC and CPS statistics were calculated for both businesses with fewer than 50 employees and businesses with fewer than 100 employees, corresponding to each state’s definition of a small business.

11. 11. Census Bureau and Department of Health and Human Services, “Health Insurance Cost Study Plan Information Questionnaire,” OMB No. 0935-0110 (accessed May 8, 2024), https://tinyurl.com/38rpw844.

12. 12. Agency for Healthcare Research and Quality, correspondence (September 12, 2023). Excluded employment types, such as employment by federal agencies, private households, and certain financial institutions, are unlikely to be in the small-group market and have a negligible effect on CBO’s assessment. In particular, excluded entities include U.S. post offices, the federal government, private households, public administrations, insurance and employee benefit funds, trusts, estates, agency accounts, offices of bank holding companies, offices of other holding companies, offices of notaries, and a small number of other entities with a history of nonresponse.

13. 13. Agency for Healthcare Research and Quality, MEPS-IC Glossary of Health Insurance Terms (accessed August 11, 2026), https://tinyurl.com/my5hu44s. Some states allow self-employed sole owners with no employees to buy small-group plans. In the states where they are allowed to buy in the small-group market, they are included in risk adjustment data but may not be included in MLR data (Meinan Goto, Centers for Medicare & Medicaid Services, Office of Legislation, personal correspondence [February 6, 2024]). For more information, see Sabrina Corlette and others, Perspective From Brokers: The Individual Market Stabilizes While Short-Term and Other Alternative Products Pose Risks (Urban Institute, April 2020), https://tinyurl.com/2yefd3yh.

14. 14. Meinan Goto, Centers for Medicare & Medicaid Services, Office of Legislation, personal correspondence [February 6, 2024]. See also Centers for Medicare & Medicaid Services, Center for Consumer Information & Insurance Oversight, “Frequently Asked Questions on the Impact of PACE Act on State Small Group Expansion” (October 19, 2015), https://tinyurl.com/2p9mv2yd.

15. 15. The states identified are Alaska, Arkansas, California, Florida, Georgia, Indiana, Louisiana, Nebraska, Ohio, Texas, Tennessee, and Utah. All but two had a negative measure for enrollment in self-insured and other plans before the MLR data were deflated.

16. 16. For the difference between filings for the Supplemental Health Care Exhibit and MLR data, see National Association of Insurance Commissioners, “Supplemental Health Care Exhibit (SHCE)—Preliminary Medical Loss Ratio (MLR) Cautionary Statement” (2015), https://tinyurl.com/ycy53c8p.

17. 17. Billable member months measure the total months that a billable individual was enrolled; billable individuals generally include the policyholder, a spouse, and the three oldest children on a family plan. See 45 C.F.R. § 147.102. New York and Vermont use a family tiering method in which billable individuals on a family plan include the policyholder, a spouse, and only the oldest child. For additional information, see Centers for Medicare & Medicaid Services, Center for Consumer Information and Insurance Oversight, “Market Rating Reforms: State Specific Rating Variations” (September 10, 2024), https://tinyurl.com/bzmwjksp.

18. 18. Centers for Medicare & Medicaid Services, Center for Consumer Information and Insurance Oversight, correspondence (May 4, 2023). The ratios were below 0.98 for New York and Vermont, probably because of their different tiering method, and for Utah, probably because of the state’s relatively large family sizes.

19. 19. State and year combinations with merged individual and small-group markets in CMS’s risk adjustment data are Massachusetts, 2017 to present; Vermont, 2014 to 2021; and Maine, 2023 to present. Massachusetts operated its own risk adjustment system from 2014 to 2016 and is omitted from CMS’s risk adjustment data in those years.

20. 20. Jessica Banthin and others, Changes in Health Insurance Coverage Due to the COVID-19 Recession: Preliminary Estimates Using Microsimulation (Urban Institute, July 2020), pp. 1–9, https://tinyurl.com/m2jahs8v.

21. 21. For details, see Department of the Treasury, “Paycheck Protection Program” (accessed May 22, 2026), https://tinyurl.com/yvxsfpax.

22. 22. Among small businesses, the shares varied by their size: 33 percent for firms with 1 to 4 employees, 42 percent for firms with 5 to 9 employees, 52 percent for firms with 10 to 19 employees, 62 percent for firms with 20 to 49 employees, and 73 percent for firms with 50 to 99 employees. See Bureau of Labor Statistics, “BRS 2020 Tables,” Table 12 (accessed August 26, 2026), www.bls.gov/brs/data/tables/2020/home.htm.

About This Document

The Congressional Budget Office produced this report at the request of the Ranking Member of the House Committee on Education and the Workforce and the Ranking Member of the Committee’s Subcommittee on Health, Employment, Labor, and Pensions. In keeping with CBO’s mandate to provide objective, impartial analysis, the report makes no recommendations.

Allison Percy wrote the report with assistance from Ben Hopkins and from Robert Lindsay, Julianna Mack, and Rajan Topiwala (all formerly of CBO) and with guidance from Alexandra Minicozzi, Tamara Hayford, and Chapin White (formerly of CBO). Justin Falk, Kyoung Mook Lim, Jeffrey Schafer, Ellen Steele, Chad Shirley, Emily Vreeland, and James Williamson offered comments. Ravza Aykan (formerly of CBO), Nicholas Johnson, Ruth Payne, and Ian Shayne fact-checked the report.

Carrie Colla of the Geisel School of Medicine at Dartmouth (a consultant to CBO), Sabrina Corlette of the McCourt School of Public Policy at Georgetown University, Jason Karcher of the American Academy of Actuaries, and Brian Blase of the Paragon Institute commented on earlier drafts. The assistance of external reviewers implies no responsibility for the final product; that responsibility rests solely with CBO.

Jeffrey Kling reviewed the report. John Skeen edited it, and R. L. Rebach illustrated the cover, created the graphics, and prepared the text for publication. The report is available at www.cbo.gov/publication/57984.

CBO seeks feedback to make its work as useful as possible. Please send comments to communications@cbo.gov.

Phillip L. Swagel

Director

September 2026