The Elusive Paradoxes of the ACA

Actuaries can play a key role in market improvement by helping policymakers understand the math behind proposed fixes Greg Fann


At the midpoint of the most recent open enrollment period (Nov. 1, 2019–Dec. 15, 2019), a Health Affairs blog article opened with this counterintuitive claim: “In 2020, Affordable Care Act (ACA) health insurance marketplace premiums will increase for significant segments of the currently enrolled population. This is because gross premiums are decreasing in the ACA marketplaces for the second year in a row” (emphasis mine).1 The successful counterintuitive understanding made me smile. It’s a story I’ve been telling for years, and it was pleasant to finally read about the inherent nature of ACA math2 from another source.

Much has been written about the societal benefits of the ACA as well as the new challenges it has created. This article is not intended to champion the ACA, critique its flaws or discuss its overall impact; instead, it walks through the ACA’s nonintuitive paradoxical dynamics as a result of its unique design. It is loosely segmented to discuss each paradoxical element with the year when it became apparent. As the 10th anniversary3 of the ACA arrives, most Americans have a surface-level understanding and in turn believe that ACA dynamics are directionally logical. Readers of this article will learn to appreciate the ramifications of the unique dynamics and the non-instinctive actualities that are foundational to ACA mechanics.

The purpose of this article is to provide “objective analysis and insight on important societal issues”4 in a marketplace that relies heavily on a deep technical understanding. The nonintuitive nature of ACA mechanics has led to some policy decisions with unintended consequences. In the interest of sustaining challenging markets, it is imperative that actuaries educate other stakeholders on the implications of the ACA’s unique dynamics, and clearly communicate that traditional aims (e.g., lowering costs) do not always align with market success. The ACA marketplace can be understood—but not with a traditional health insurance approach. It requires accepting mathematical dynamics that are plainly true but clearly irrational.


The inaugural year of ACA marketplaces commenced with significant operational challenges that relegated mathematical dynamics and underlying financial sustainability to the back page. When the dust settled, the market was older, sicker and more concentrated at lower incomes than ACA architects expected. “The 18–34 demographic represented 28 percent of the market at the end of the 2014 open enrollment, short of the 39 percent targeted expectation.”5 Intuitively, some commentators concluded the 3:1 age curve was the likely culprit. The more relevant impact was “premium subsidies are allocated in such a way that the net premium costs for older people are actually lower than the net premiums for younger people for the lowest-cost plan options.”6

Figure 1: Net Monthly Premium of Second-Lowest Carrier (275% FPL)

Source: Fann, Greg. The True Cost of Coverage. The Actuary, December 2015/January 2016 (accessed February 24, 2020).

The unbalanced allocation of subsidy dollars, “surprisingly at odds with legislative intent to attract young, healthy enrollees and the noted sustainability requirements,”7 led to skewed enrollment and market sustainability challenges. Contrary to unsophisticated projections that young adults would enroll due to subsidy application,8 an “understanding of the subsidy mechanics is important to understand enrollment ramifications and, ultimately, the long-term implications of the ACA on the individual market.”9

Proposals to widen the age curve have been offered as a solution. Interestingly, a change from a 3:1 curve to a wider curve would increase subsidized bronze and lowest-cost silver premiums (and gold10 premiums in appropriately priced markets) for younger adults and lower the same subsidized premiums for older adults. A change to a wider age curve may result in the opposite of the intended enrollment incentives.11

The same mathematical formula that attracted primarily older adults also benefitted some consumers with higher premiums. In a reversal of what we see in 2020 dynamics, I had observed that “many people with exchange coverage will actually see their net premium rates decrease due to rate increases as the federal government’s share of the cost increases.”12 Warnings were provided that “the rate review process benefits higher-income individuals but actually increases net rates on low-income individuals; an anomaly might alarm regulators who view the rate review process primarily as a consumer protection function with a focus on representing low-income residents.”13

The first section of the aforementioned Health Affairs blog article is titled “The Strange Math of The Marketplaces: Why Premium Decreases Can Increase Costs for Subsidized Enrollees.” The strange inversion of net premium rates is an inherent dynamic in ACA marketplaces. It has always been there, but only recently has it become publicly discussed and accepted by a nonactuarial audience. We should consider that progress, yet there are other ACA paradoxes that are still not well understood.


The lifeblood of ACA markets is taxpayer-funded premium subsidies.14 A sharp reduction in subsidy funding can be dangerous and severely curtail financial incentives for eligible individuals to enroll. Like other health insurance coverage involving federal financing, funds are appropriated based on prescriptive government formulas. In the case of Medicare Advantage, benchmark funding is calibrated from costs of robust Medicare claim data. When ACA markets went into effect, there was no historical government data to utilize for the purposes of determining appropriate funding. Without an established benchmark in place, ACA architects decided to let annual market rates determine the subsidy allocation. Without systematic control, the confluence of subsidy mechanics and unadjusted market premium rates can create environmental market problems.

Somewhat arbitrarily, but also due to funding constraints, premium subsidies were developed from the second-lowest-cost plan offering of the second-lowest (silver) of the four levels of benefit coverage. Rather than directly calculate subsidy amounts, the formula was reverse-engineered to maximize enrollee contributions as a percentage of income (for a “benchmark” plan) and allow premium subsidies to be the difference between premiums and enrollee contributions. Effectively, as gross premiums rise or fall, subsidies move dollar for dollar and the enrollee’s contribution to the benchmark plan is fixed.

As the calibration is on the second-lowest plan rather than an average or median, an increase in insurer participation alone creates a bias for compressed premium subsidies. To test this, roll a pair of dice three times. Write down the second-lowest total. Repeat 10 times. Now roll the dice five times and do the same. Simply having more data points depresses the second-lowest result. As more insurers enter markets (this happened in 2019 and 2020 as favorable conditions15 developed), the benchmark is based on a lower-cost plan.

As benchmarks decline and gross premiums stay the same, consumers pay more. This results in a less attractive market and lower enrollment. The dice example is somewhat analogous to an insurance market with competing insurers that have similar cost structures. Only rolling one die would create a greater bias and represent an insurer with unusually lower costs entering an ACA market; this would certainly lead to different expected results.

ACA market participation varies across the county, but some state markets have been penetrated by health plans with unusually low-cost structures. These plans are often competing with commercial group insurers that offer similar networks and have equivalent reimbursement rates in individual markets. Due to ACA subsidy dynamics, introducing an atypical health insurer with a materially lower cost structure to a commercial market can artificially lower the area benchmark, and consequently reduce premium subsidies below intended16 levels. It then raises net premiums and creates challenges for lower-income consumers to procure health coverage from traditional insurers.

Like an invasive species introduced to a new environment,17 a health plan offering lower costs is not inherently bad. Low-cost plans are actually very helpful in most markets. In the ACA world, however, they usually do more harm than good (not always noticed) by lowering premium subsidies and increasing net premiums. “In the health insurance ecosystem, the paradoxical dynamics of the ACA may allow an invasive species to go undetected and abide in a market misconstrued as a helpful part of the ecosystem.”18 There has been some recent recognition of this dynamic, and stakeholders now are seeking regulatory changes in subsidy formulas rather than cost reductions.19

As robust competition has been the national primary concern in the ACA’s initial years, it’s been rare for states to raise concerns around subsidy funding disruption—but some savvy state exchanges have done exactly that. In 2017, a health plan in Rhode Island was prevented from joining the exchange due to such anxieties. “The exchange was concerned that the additional low-cost plans would have reduced the subsidies available to all exchange enrollees, making coverage less affordable if people chose plans other than the new low-cost options.”20

The news is not all bad with low-cost plans. The compressed premiums result in lower costs for taxpayers, reduced premiums for unsubsidized consumers, and lower cost-sharing for subsidized and unsubsidized enrollees. However, these “good news” items do not detract from the reality that atypical cost offerings in ACA markets are often hurting those they are presupposed and intended to help. Whether or not states actively address this problem is not an actuarial consideration, but it is the responsibility of actuaries to be sure states properly understand these dynamics.


Conventional thinking among the members of the general public is insurance companies want to provide coverage for healthy people who have few medical claims and avoid those with high costs and chronic conditions. Of course, the reality is that most insurers simply want a reasonable matching21 of premiums and expected claims, and they remain agnostic about whom they enroll if that matching can be achieved. Insurers are more than willing to accept high-risk individuals if premiums22 are sufficient, and less interested in enrolling low-cost individuals if premiums are not adequate. The government assumed a role of assuring premium sufficiency with a community-rated, risk adjustment market. “It is therefore imperative that the operational methodology is precise and impartial, as the Centers for Medicare & Medicaid Services (CMS) has assumed accountability for equity among market participants.”23 As CMS does not contribute funds to the risk-adjustment model, any risk adjustment payments must come from other insurers.

2016 was the year when “serious concerns regarding the ACA risk-adjustment methodology became publicly apparent.”24 Many stakeholders believe the ACA risk-adjustment model25 results in imbalanced assessments that penalize the low-cost, well-managed insurers that attract the healthy enrollees the ACA needs to survive. This is not conjecture; health plan actuaries confirmed “healthy doesn’t pay under risk adjustment”26 and the “transfer formula penalizes plans with lower premiums.”27 A state actuary claimed “CMS told us to expect that high-cost case conditions would be overpaid,”28 and the former CMS chief actuary opined, “The current HHS-HCC risk-adjustment model established by CMS is known to understate risk scores for relatively healthy individuals and to overstate them for those with significant health conditions.”29 Compounding the risk-assessment inequities, transfer payments are often magnified as the nature of the statewide average premium formula “severely exaggerates risk transfers for efficient insurers by mandating an inflated transfer amount relative to their cost structure.”30 The dual requirement from an insurance company perspective is enrollment of a profitable segment fed by risk-adjustment funds and a healthy mix of overall marketplace enrollment to provide a source for those funds.

The paradox of risk adjustment is that market attractiveness (both supply and demand sides) is skewed away from what we collectively agree the risk pool needs to maintain viability. As I said on a 2016 podcast, “Health plans are doing better financially with high risk individuals … So we’re sort of in this paradoxical box now where young and healthy individuals don’t have much incentive to enroll, heath plans may not want to enroll them, but we turn on the TV and all we hear about is how we need young adults to enroll in this market to preserve the risk pool.”31 Notably, changes to risk-adjustment methodology have been implemented, but litigation is still ongoing regarding risk-adjustment equity.32 As the methodology is budget neutral and inequities naturally create winners and losers, there in turn are different insurer viewpoints regarding the equity of the current model.


The ACA changed how we fund health care costs for high-risk individuals who are not enrolled in group insurance or eligible for traditional government programs. It shifted the burden from a patchwork of state-based “high-risk pools” to the broader individual market. In a sense, it shifted the financial burden from a general population obligation to a limited population of individual market participants. Consequently, individual market premiums are substantially higher than they were pre-ACA. Notably, much of the burden is shifted again to the federal government via premium subsidies attached to the higher premiums. However, unsubsidized enrollees bear the full brunt of the resulting higher premiums.

A recent preference for states has been to utilize the ACA federal funding and mechanically carve-out high-risk individuals to reduce premiums and relieve pressure on unsubsidized enrollees. This is accomplished through Section 1332 waivers.33 2017 marked the first year that such waivers were permitted. While originally intended to allow more creative innovation by the states, restrictive regulatory guidance34 has made states reticent to explore innovation beyond “invisible reinsurance.”

Alaska had the highest costs in the country, and it was the first state to adopt such a waiver program. The Alaska Reinsurance Program has worked well, and 11 other states have followed suit and adopted a similar model. The adoption of reinsurance has brought attention35 to a paradox—“these waivers have appropriately reduced the high unsubsidized premiums; however, little attention has been paid to the impact of this change on the larger subsidized market.”36

Reducing gross premiums raises net premiums for subsidized enrollees for plans that are priced lower than the benchmark plan. Beginning in 2018, this affects an increasing number of enrollees, as silver premiums rise above gold premiums.37 Reinsurance waivers have directly improved the risk pool by removing the highest-cost enrollees. Beneficially, this reduces premiums for unsubsidized enrollees, the population most harmed by the ACA’s rating rules. Less intuitively, it increases net premium rates for some subsidized enrollees. It’s up to states to decide whether the trade-off is good policy. It’s up to actuaries to let states know that the dichotomy exists. In a general sense, “the need for a holistic perspective extends to more innovative uses of Section 1332 waiver authority beyond carving-out costs of high-risk enrollees.”38


After operational improvements following ACA implementation, financial challenges emerged leading to high premium increases, insurer exits and declining enrollment. 2016 and 2017 were among the ACA’s most challenging years.39 “Obamacare Marketplaces Are in Trouble. What Can Be Done?”40 asked The New York Times. High rate increases in 2017 improved insurer financial performance (82 percent medical loss ratio (MLR) compared to 99 percent average from 2014–201641), but enrollment continued to decline and insurers continued to exit markets—but many were persuaded by states to remain to avoid counties without any insurers.

The boost to ACA markets came from perhaps the most publicized and counterintuitive dynamic of the ACA: the cost-sharing reduction (CSR) paradox.42 Throughout 2017, amidst partisan ACA repeal efforts, there were bipartisan calls to appropriate CSR funding, which had been allowed by the Obama and Trump administrations but never appropriated by Congress. It was often listed as the primary requirement necessary to “stabilize the market.”43

Following a legal recommendation from the Department of Justice, President Trump discontinued CSR funding in October 2017. As expected,44 insurers followed the “fundamental actuarial principle”45 of premium/claims matching and responded by increasing silver premiums, thereby increasing premium subsidies46 and reducing net premiums. The market responded as predicted.47 “Enrollment significantly exceeded expectations, insurer profitability skyrocketed to record levels and premiums decreased for the first time in 2019.48 The beneficial market changes reignited insurer interest, and the number of state-level insurers increased 17 percent in 2019 after a 28 percent reduction in 2017 and a 21 percent reduction in 2018.”49

Many now recognize that CSR defunding has reduced net premiums, boosted enrollment for the impacted population and attracted insurers to the marketplace. Reversing this change now would draw substantial opposition, even from those initially opposed to implementation. It took longer than it should have for an appropriate understanding to occur, but that’s the recurring pattern with nonintuitive dynamics and a highly politicized media culture.50 Perhaps the best advice is to trust objective mathematical analysis.


ACA paradoxical dynamics remind me of the arcade game Whac-a-mole, which involves players using a mallet to hit toy moles that randomly emerge from holes. If a mole is successfully whacked, there is no time to relax or celebrate as another will soon emerge from an unknown location. ACA paradoxes are just as elusive; when we learn to understand one paradox, another emerges without warning, sometimes from where we least expect it.

As ACA repeal efforts have failed during President Trump’s first term, his potential Democratic challengers have floated policy proposals with more government involvement. No candidates are proposing to leave the ACA alone; the mildest proposals include expanding premium subsidies or adding public coverage options. One proposal, which ironically benefits the demographic helped rather than harmed by the ACA,51 is an option to buy into Medicare at ages 50 to 64.

A recent RAND study52 found that allowing older adults to opt into Medicare would cause ACA premiums to increase. This is an interesting dynamic—it means that the 3:1 age slope is not a steep enough reflection of ACA enrollment. While it is generally acknowledged that the theoretical age slope is somewhere between 5:1 and 7:1, “a broad variety of older adults, including both healthy and unhealthy people, tends to enroll in the individual market. Younger adults who enroll in the individual market, in contrast, tend to be unhealthy and expensive.”53 Hence, the resulting cost difference is a measure of all older adults to unhealthy young adults.

Intuitively, pricing age slopes are self-fulfilling. If a narrow pricing slope is applied to a market, it is more attractive to older healthy enrollees and less attractive to younger healthy people. The younger enrolled population is now less healthy. In the case of ACA markets, the pricing slope is effectively narrower than 3:1, as most enrollees are subsidized and older adults receive greater subsidies. For a benchmark plan, it is 1:1 for individuals with the same income level as inverted for plans priced below the benchmark.54 It is mathematically clear that RAND’s findings are logical and a Medicare buy-in would drive up insurance premiums. It will be interesting to see if this understanding of this paradox resonates with stakeholders. If it does, it may result in proposals that seek to preserve the ACA benefits and counteract the problems it created, rather than magnify the problems, increase ACA premiums and relegate ACA markets to a solution for adults age 26 to 49—“a small remnant of the robust marketplaces that ACA architects once envisioned.”55


Rational, intuitive perspectives are everywhere. Proclaiming that “markets need healthy competition and innovations that reduce systematic costs” won’t result in many arguments. It also won’t strike people as profound or insightful. We are in the seventh year of ACA marketplaces, and those vanilla viewpoints are still expressed as viable solutions. They are logical and generally accurate—but do not apply to most consumers in ACA markets. Competition and benchmark premium reductions56 don’t always help; they often just make things worse.

Most proposed solutions take the form of Congressional legislation. With no significant adjustments to the ACA since its 2010 passage, legislative efforts in a divided Congress seem unlikely—but that doesn’t stop the chatter. Talking points for much of the proposed legislation are “building on the ACA.” This generally means increasing and expanding premium subsidies, not addressing some the ACA’s unique dynamics.

Without federal legislation, states literally can do both (increase and expand premium subsidies) through a two-pronged approach of optimization57 and reallocation.58 Actuaries can play a key role in market improvement, and appreciation of actuarial involvement in understanding of ACA dynamics is growing. A recent Health Affairs Blog article discussed how regulatory policy was “making the market more attractive,”59 citing a Strategic Initiative of the Society of Actuaries (SOA) Health Section. It ended with strong encouragement that policymakers take into consideration the objective analysis of actuaries.60

Attempts to improve the current framework are fruitless when the underlying mechanics and resulting challenges are not well understood. As many stakeholders crafting reforms will not fully understand the complex technicalities, they will need assistance to know if their proposed changes are holistically helpful or harmful to markets with nonintuitive dynamics. If you hear them say “premiums will increase because gross premiums are decreasing,” you will know they have moved in the right direction. Actuaries can provide valuable analysis and insight, and help show them the rest of the way through the paradoxical matrix.61 In a world where the mathematical ramifications are plainly true but clearly irrational, stakeholders need the continued guidance of dispassionate mathematical experts more than ever.

Greg Fann, FSA, FCA, MAAA, is a consulting actuary with Axene Health Partners, LLC (AHP) in AHP’s Temecula, California, office. He is also the volunteer leader for the Individual/Small Group Subgroup of the SOA’s Health Section Council.

Copyright © 2020 by the Society of Actuaries, Schaumburg, Illinois.