Wednesday, February 21, 2024

How the Trump administration handled the ACA marketplace, Part 1

 Note: All xpostfactoid subscriptions are now through Substack alone (still free), though I will continue to cross-post on this site. If you're not subscribed, please visit xpostfactoid on Substack and sign up.


In the wake of Trump’s vow to repeal the ACA if elected, Larry Levitt, Kaiser Family Foundation, outlines the former president’s past and purported future healthcare agenda.

One of Trump’s biggest political failures as president was his inability to persuade Congress to repeal the Affordable Care Act (ACA). However, the Trump administration did make significant changes to the ACA, including repealing the individual mandate penalty, reducing federal funding for consumer assistance (navigators) by 84% and outreach by 90%, and expanding short-term insurance plans that can exclude coverage of preexisting conditions. And, the Trump administration supported an ultimately unsuccessful lawsuit to overturn the ACA.

In one of the stranger policy twists, the Trump administration ended payments to ACA insurers to compensate them for a requirement to provide reduced cost sharing for low-income patients. At the time, Trump said this would cause Obamacare to be “dead” and “gone.” But, insurers responded by increasing premiums, which in turn increased federal premium subsidies and costs to the federal government, likely strengthening the ACA.

In the current campaign, Trump has vowed several times to try again to repeal and replace the ACA, saying he would create a plan with “much better health care.”

Trump certainly meant harm to the ACA. His comments in the wake of abruptly cutting off direct reimbursement of insurers for the value of Cost Sharing Reduction, cited by Levitt above, show his intent, as does his pressure on Republicans in Congress to pass legislation gutting its core programs. Should Trump regain the presidency, there is no question that he will pursue the agenda that Levitt outlines in his conclusion, including the ACA-related parts:

Trump’s record as president from 2017 to 2021, combined with recent comments on the campaign trail, suggest he would pursue policies to weaken the ACA, reduce federal spending on Medicaid, restrict access to abortion and family planning, and scale back benefits for immigrants if reelected as president in 2024.

Moreover, should Trump regain the presidency, he would lead a Republican party even more subservient to his will than in his first term. A Republican Congress would almost surely roll back the ACA Medicaid expansion and impose sharp spending caps on surviving Medicaid programs, as well as deregulating and largely defunding the ACA marketplace, as failed Republican legislation aimed to do in 2017. Should Democrats control one or both houses of Congress, an HHS Department filled with MAGA partisans, in line with plans currently being laid by well-funded right-wing organizations like the Heritage Foundation to root out technocratic expertise and install Trump loyalists at every level in all federal departments, would doubtless pull out all stops to undermine the marketplace and reduce Medicaid enrollments.

In Trump’s first administration, his appointments to HHS and CMS also were hostile to the structure of the ACA marketplace and the Medicaid expansion. Most notably, CMS administrator Seema Verma encouraged states to impose work requirements on “non-disabled, working age Medicaid enrollees — with some success, although the measures were largely checked by the courts. She also pushed states to conduct more frequent income and eligibility checks on Medicaid enrollees, encouraging the kind of procedural disenrollments (often of people who never received demands for information) now plaguing the post-pandemic Medicaid unwinding. '

But Verma and HHS Secretaries Tom Price and Alex Azar were also more constrained by conventional political incentives and the needs of corporate, state and individual constituents than their successors in a second Trump administration would likely be. The administration’s record with respect to ACA marketplace administration was mixed. Some measures harmed product quality and enrollment; some measures boosted enrollment and retention.

 Here I want to review the effects of several Trump administration initiatives (and one legislative act):

  • Reinsurance

  • Cutoff of direct reimbursement of insurers for Cost Sharing Reduction

  • Defunding the Navigator enrollment assistance program; boosting brokerage participation

  • Zeroing out the individual mandate penalty for going uninsured

  • Standing up a medically underwritten alternative market

I’ll need two posts to get through this agenda. This post will cover reinsurance, the CSR cutoff, and treatment of navigators and brokers.

Subscribe now

Reinsurance. When the ACA marketplace launched in November 2013 (selling plans effective in 2014), premiums came in lower than expected, as insurers jockeyed for position in a new market. Premium growth was modest through 2016. In 2017, however, a three-year federally funded reinsurance program established by the ACA expired, and concurrently, insurers had enough data to recognize that they had underpriced initial offerings, as the risk pools were smaller, older and sicker than anticipated. A sharp correction ensued, with benchmark silver premiums increasing by an average of 20% from 2016 to 2017. In March 2017, then-HHS Secretary Tom Price invited states to submit innovation waiver proposals to establish state reinsurance programs, partially funded by the “pass-through” of federal savings resulting from lower premium and therefore lower premium subsidies. By 2020, HHS had approved 15 such state waivers. In most of these programs, the state pays a percentage of claims for enrollees whose costs pass a certain threshold, up to a cap — e.g., in Colorado, the state pays 60% of individual claims between $30,000 and $400,000.

Many progressives have soured on reinsurance programs, as they tend to weaken options for subsidized enrollees (the vast majority of enrollees). By reducing premiums, reinsurance reduces “spreads” between the benchmark second cheapest silver plan — the plan that determines subsidy size — and plans that cost less than benchmark (that is, one silver plan, most bronze plans, and in some states and regions, some or all gold plans). In the era of subsidies enhanced by the American Rescue Plan (March 2021 through at least 2025), when almost 90% of enrollees are subsidized, reducing unsubsidized premiums can be viewed as a poor use of state funds. In the wake of the large premium spikes of 2017 and 2018, however, the ACA-compliant market bled unsubsidized enrollees. According to KFF estimates, unsubsidized enrollment in ACA-compliant plans dropped by essentially half from Q1 2016 to Q1 2019, from 6.7 million to 3.4 million. Reinsurance programs probably helped staunch that bleeding. At a time when about 40% of enrollees were unsubsidized, the tradeoff (to the extent that the effect on subsidized premiums was understood) seemed worth it.

CSR cutoff. This really is the strangest chapter in the Trump administration’s handling of the ACA. The ACA statute stipulates that the federal government reimburse insurers separately for the value of Cost Sharing Reduction, a secondary benefit that attaches only to silver plans in the ACA marketplace and only for enrollees with income up to 250% of the Federal Poverty Level (FPL). But the drafters neglected to make funding for CSR mandatory, and the Republican Congress refused to fund it. Through 2016, the Obama administration found the necessary funds in couch cushions, and the Republican House sued in 2014 to stop the payments. A judge upheld the suit but stayed the ruling pending appeal. When Trump was elected, it was widely anticipated that he would cut off the payments. In October 2017, following the final death knell for legislative ACA repeal, he did so — abruptly, not only ending reimbursement for the upcoming 2018 plan year, but stiffing insurers for the last three months of 2017. As Levitt notes, he then crowed that he’d killed the marketplace.

The thing was, though, the likely effects of cutting off CSR reimbursement were by that point long understood, and insurers and state regulators were prepared (though time was short to adjust for the 2018 coverage year, with open enrollment for 2018 just two weeks away). Regulators in almost all states allowed or encouraged insurers to price CSR into silver plans only, inflating silver premiums and therefore premium subsidies, which are set so that enrollees pay a fixed percentage of income (on a sliding scale) for the benchmark second cheapest silver plan. Inflated benchmark premiums widened spreads between the benchmark and cheaper plans, rendering bronze plans free for millions of enrollees. Moreover, the CSR now loaded into silver premiums makes silver plans roughly equivalent to platinum for enrollees with income up to 200% FPL — the vast majority of silver plan enrollees. When fully implemented, “silver loading” — pricing silver plans in accord with the average actuarial value obtained by silver plan enrollees — makes silver plans more expensive than gold plans.

This dynamic was first laid out in an HHS ASPE brief in December 2015, followed by an Urban Institute analysis in January 2016. In August 2017, the Congressional Budget Office forecast that cutting off direct CSR reimbursement would increase federal spending on premiums by $194 billion over ten years. That is, Trump might have anticipated (and HHS officials surely knew) that his purported death blow to the ACA would inject almost $200 billion into a marketplace that had been hobbled from the start by inadequate subsidization.

The only question was whether abrupt cutoff — without reimbursement for CSR already provided in Plan Year 2017 — would drive out a critical mass of insurers at a time when insurer participation in the marketplace had shrunk to dangerous levels (policymakers feared that the marketplace might be plagued with “bare counties,” where no insurer offered plans). David Anderson scoped out this danger in May 2017, responding to a Politico story warning that abrupt CSR cutoff “could send the health law’s insurance markets into a tailspin.” While concluding that being stiffed for the bulk of 2017 could drive many insurers out of the market by thinning out their reserves, Anderson concluded, in his terse way, “The CSR threat loses its ability to blow up the market by sometime in the fall.” Trump dropped his hammer too late. (In a followup post, Anderson detailed the financial risk that an early cutoff of CSR payments would impose on insurers.) And the courts ultimately ordered the federal government to make insurers whole for the three months that Trump stiffed them (as Anderson also implicitly predicted).

As it turned out, “silver loading” —pricing the full value of CSR into silver plans — never really took full effect in most states, as insurers continue to underprice silver plans. That’s apparently because a) they compete to offer lowest-cost and benchmark silver, still the dominant metal level; b) low-income enrollees in plans with strong CSR utilize medical care less than anticipated, inhibited by relatively low but still substantial out-of-pocket costs; and c) the ACA’s risk adjustment formula doesn’t reflect this low utilization and thus tends to penalize insurers who sell a lot of bronze or gold plans. In about fifteen states, however, average gold premiums are lower than average benchmark silver premiums, some by regulatory or legislative requirements, and some by insurers choices. And in all states, silver loading has generated significantly lower-cost bronze plans and at least somewhat lower-cost gold plans, net of subsidy, than would be the case were CSR still reimbursed separately. In 2018, KFF estimated that 4.2 million uninsured people were eligible for a free bronze plan. As a result, according to an estimate by Aviva Aron-Dine (linked to by Levitt above), the CSR cutoff had boosted enrollment by about 5% by 2019.

About eight states now effectively require marketplace insurers to price silver plans either in strict accord with the average actuarial value obtained by silver plan enrollees (thanks to CSR, the majority in most states are in plans with 94% or 87% AV) or, more radically, under the assumption that all silver plan enrollees have income under 200% FPL and therefore obtain AV of 94% or 87%. The latter (pioneered by New Mexico) is meant to be a self-fulfilling prophecy, since this assumption prices gold plans (80% AV) well below silver, so no one ineligible for the two highest levels of CSR has any reason to choose silver.

These policy initiatives have largely been spurred by progressive policy advocate Stan Dorn and a pair of conservative actuaries, Greg Fann and Daniel Cruz — who, unlike most conservative policymakers in healthcare, are committed to making the ACA marketplace work as well as possible. While strict “premium alignment” has been implemented mainly in blue states, Fann and Cruz help the idea to take root, incredibly, in Texas, where the average lowest-cost gold premium in 2024 is 14% below the average benchmark silver premium. This premium alignment policy, first implemented in Plan Year 2023, probably is a factor in the 89% enrollment growth the Texas marketplace experienced from 2022 to 2024, an increase of more than 1.6 million. I told the rather remarkable story of how strict premium alignment took hold in the Texas legislature - -which passed the measure unanimously — here.

Defunding navigators/supporting brokers. The Trump administration had a rooted hostility to the Navigator nonprofit enrollment assistance program, established by ACA statute and funded by insurer user fees paid to the federal marketplace in states that use HealthCare.gov. (State-based marketplaces collect their own fees and fund their own enrollment assistance programs.) Seema Verma’s CMS gutted navigator funding, cutting it in two stages from $63 million in 2016 to $10 million in 2018 and years following. For justification, CMS released a report (the link appears to be dead) asserting that the program was ineffectual, as few enrollments on HealthCare.gov were credited to navigators. The report ignored key facts: a) unlike brokers, navigators were not oriented toward nor incentivized to be credited as assisters in HealthCare.gov enrollments, and not authorized to positively recommend a particular plan, and b) navigators, who primarily serve a low-income population, by their own report helped far more people enroll in Medicaid than in the marketplace (state exchange reports also bear this out). (On the first point, a Kaiser Family Foundation reported, “The healthcare.gov online application includes a field where Navigator staff can enter their identification number for each consumer whom they assist. Navigators report that program staff have not been trained on this data entry and did not consistently enter it.) Verma’s CMS also cut advertising for the marketplace from $100 million to $10 million.

CMS’s assault on the navigator program may have reduced enrollment to some degree, and it deprived many low-income enrollees of guidance to enrollment in a program, the ACA marketplace, that remains poorly understood by vast swaths of the U.S. population to this day. It may also have reduced Medicaid enrollment. But the harm should be put in perspective on two fronts.

First, while federally funded Navigator organizations play a sort of quarterbacking role among nonprofit enrollment assistance programs in some states, numerically they represent a relatively small part of the enrollment assistance ecosystem. In 2016, according to another KFF survey, Navigator organizations constituted only 10% of some 5,000 assister programs, accounting for 17% of the 5.3 million people served by such programs. (As the Biden administration has boosted Navigator funding to about $100 million per year, they may now account for a somewhat larger percentage.) Dedicated funding for enrollment assistance at Federally Qualified Health Centers, hard-wired into FQHC funding to the tune of $150 million in 2016, dwarfed Navigator funding and doubtless still does. Independently funded programs fielding Certified Application Counselors (CACs), who undergo a federal training program somewhat more limited than the program for Navigators, accounted for 65% of assister programs in 2016. In the Trump years, navigator programs survived on a shoestring (program leaders and veteran navigators described their cutbacks and adaptations to me in this post), cutting staff, relying more on remote assistance (spurred by the pandemic), and cutting back on the outreach/education programs that are part of their mandate.

Second, while the Trump administration starved the navigators, it fed the brokers — who, even in 2016, accounted for more marketplace enrollments than did nonprofit assisters. In the Obama years, HealthCare.gov’s “find help” feature separated nonprofit assisters and brokers (and, if I remember right, defaulted to nonprofit help). The Trump administration mingled the two, and stood up a “help on demand” program — continued by the Biden administration — in which a site visitor could enter her zip code and phone number or email address and be contacted promptly by a certified broker.

Perhaps more importantly, Trump’s CMS actively encouraged and facilitated development of the commercial Direct Enrollment and subsequent Enhanced Direct Enrollment (EDE) program, through which commercial online brokers can be certified by CMS to take applications and process subsidies without appearing to shift applicants to the federal exchange. EDE-enabled e-brokers, led by the market-dominating HealthSherpa, provide brokers with a range of tools enabling them to track and serve clients. EDE has also streamlined the enrollment process itself: An application can be completed more quickly on HealthSherpa than on HealthCare.gov.

EDE, which so far is enabled only in the 32 states using HealthCare.gov, may at least partly explained why enrollment growth in HealthCare.gov states in the pandemic years has far outstripped enrollment in states running their own exchanges. As I’ve noted recently, in HealthCare.gov states in 2023, brokers assisted more than 70% of active enrollments on HealthCare.gov (excluding passive auto-re-enrollment), and 81% of active broker-assistance enrollments were via DE or EDE. HealthSherpa processes more new enrollments and active re-enrollments than HealthCare.gov does directly. 83,000 brokers are now registered with the federal exchange, up from 49,000 in 2018.

Some brokers are unscrupulous. But the extensive industry infrastructure that pre-dates the ACA ensured that they would play an essential role in marketplace enrollment, and that role has grown. There are far more brokers selling ACA plans than there are nonprofit enrollment assisters, and the latter help more people eligible for Medicaid than for the marketplace. The Biden administration has continued to support EDE development — and brokers.

The Trump administration’s fondness for DE and EDE probably has roots in a conservative goal, as old as the ACA, to decentralize the marketplace. Effecting enrollments through commercial sites — a goal reflected in a Georgia waiver proposal, to be discussed in Part II, to do away with any centralized state-sponsored exchange — was ideologically affiliated with breaking the ACA-compliant plan mold, founded on guaranteed issue and mandatory Essential Health Benefits. EDE platforms are anchored to HealthCare.gov, however, and subject to strict regulation — regulation that CMS has proposed to tighten in preparation for enabling state-based marketplaces to embrace EDE.

It should also be noted that CMS’s gutting of navigator funding was paired with sharp reductions in marketing and advertising and complemented by rhetorical and ideological hostility to the marketplace as currently constituted. HealthCare.gov messaging in the Trump years — e.g., in email marketing to people with HealthCare.gov logins — never had anything good to say about the coverage on offer; the enrollment reminders were bare-bones and grudging. The Trump administration also cut the Open Enrollment Period in HealthCare.gov states to six weeks. That measure probably reduced OEP enrollment, while it also apparently reduced enrollment attrition, as only those who knew they needed coverage found their way to enrollment. Retention improved steadily through the Trump years, also spurred by the widespread availability of zero-premium plans boosted by silver loading. In fact, when improved retention is taken into account, the reduced end-of-OEP enrollment in the Trump years looks like something of an illusion.

Next up (later this week or early next): mandate penalty repeal and the ACA-noncompliant market.


No comments:

Post a Comment