David Anderson

Via Charles Gaba’s exceptional write-up of the Biden healthcare plan, there is one segment that has me scratching my head:

Governors and state legislatures in 14 states have refused to take up the Affordable Care Act’s expansion of Medicaid eligibility, denying access to Medicaid for an estimated 4.9 million adults. Biden’s plan will ensure these individuals get covered by offering premium-free access to the public option for those 4.9 million individuals who would be eligible for Medicaid but for their state’s inaction, and making sure their public option covers the full scope of Medicaid benefits. States that have already expanded Medicaid will have the choice of moving the expansion population to the premium-free public option as long as the states continue to pay their current share of the cost of covering those individuals. [my emphasis]

Why is there a maintenance of effort (MOE) requirement? It effectively rewards laggard states. It buys a slightly better Congressional Budget Office (CBO) score on federal spending. It is bad macro-economic policy.

And far more importantly, we’re in the second negative interest rate recession in the past thirteen years. States have balanced budget requirements. The federal government does not. State funded programs are fundamentally pro-cyclical. We, as a society, should want to have strong counter-cyclical programs that leverages the single society wide entity with almost unlimited risk eating capacity, cheapest access to almost infinite credit and a multi-generational shadow of the future to spend money in big recessions. From a stabilization point of view with any understanding of the incentive structure of the American political system, we also should want those stabilizers to be automatic with minimal need for Congress to haggle over a deal on an FMAP bump.

In December 2018, I did not get the reasoning for including a maintenance of effort requirement in Medicare Extra put out by the Center for American Progress:

And I mean that from a policy angle and a political angle. Why does CAP want to charge states for a portion of low income health care expenses? From a policy angle, I am scarred by the 2008-2010 Great Recession. I want as many automatic stabilizers that are not tied to balance budget requirements as possible. One of the big chunks of the stimulus bill was an enhanced federal Medicaid payment rate which allowed states to not cut budgets as deeply as they normally would have. We don’t want 50 Mini-Hoovers squeezing out Medicaid.

Secondly, and this is a political angle, state level buy-in to a proposal like this is probably likely to be a whole lot higher when the Feds could tell a state “You know, don’t worry about low income health care programs, we’ll take care of it and you can use that $3 billion a year on schools, roads, tax cuts or hookers and blow; your choice”

We should leverage the federal government’s ability to absorb massive credit and shock risk by not requiring states to engage in counter-cyclical spending cuts at the time when more people need public services.

Bidencare, public options and maintenance of effort requirements?Post + Comments (6)

Missouri voters approved Medicaid Expansion via a state plan amendment last night.

Missouri voters have PASSED Amendment 2, the ballot initiative expanding Medicaid to approximately 230,000 low-income Missouri residents, @DecisionDeskHQ projects. Missouri joins Oklahoma as the second state to expand Medicaid via ballot initiative in 2020 https://t.co/AB2Vrh6lml pic.twitter.com/8OpdLQliW3

— Grace Panetta (@grace_panetta) August 5, 2020

The State Plan Amendment Expansion will go into effect July 1, 2021. The State Plan Amendment means waivers that enable shenanigans like work requirements are not applicable to the expansion. The cities voted strongly for expansion while the suburbs and rural areas voted far less intensely against it than they typically vote against the cities.

12 states including the Big 4 of Texas, Florida, Georgia and North Carolina still have not expanded Medicaid.

Congratulations Missouri for doing something to help your own residents and aid keeping your rural health networks plausibly viable.

Missouri accepts Medicaid ExpansionPost + Comments (73)

This afternoon at 3:30 EST, I’ll be part of a pre-recorded panel at ARM-2020 where I will be presenting a paper in progress: Association of Insurer Characteristics and Plan Quality Measures in the ACA Health Insurance Marketplaces.

My co-authors, Sih-Ting Cai, Coleman Drake and Jean Abraham, and I,  asked a basic question: Does quality ratings vary by insurer characteristics in the ACA individual market?

We looked at four indicators from the CMS Public Use data set and matched on HIOS-ID to the Robert Woods Johnson HIX Compare and Plan Participation data sets.  We also matched the HIOS ID to the National Association of Insurance Commissioners data set to get ownership structures.  We categorized insurers in four manners:

• Non-profit
• Offered a PPO
• Member of the Blue Cross/Blue Shield Association
• Primarily Medicaid Managed Care

The last two are mutually exclusive.  The rest are potentially mixed.

We looked at four quality of care indicators:

• Are kids with upper respitory infections treated correctly?
• Are kids with pharyngitis tested correctly?
• Are antibiotics avoided for adults with bronchitis
• Are imagining studies used appropriately for complaints of lower back pain

We had a prior that non-profits and Blues would do better than average, and Medicaid based insurers would do worse.

That is not quite what we found.

Non-profits were a little bit better on two measures.  PPOs had lower measured quality on two metrics.  Medicaid and Blues had lower quality on one metric.

Why does this matter?

Most people are buying ACA plans primarily on premium and if there is a secondary attribute, it is network.  Quality does not matter a whole lot. We think that there is reason to suspect that there is a quality-premium trade-off but there are no easy markers for people to figure out quality on their own.  If we want to encourage quality, the ACA price linked subsidy system is likely to be a barrier.

Effectiveness of care quality rating in the ACAPost + Comments (2)

Missouri is voting today.  The big question on the ballot is whether or not Missouri will expand Medicaid to all adults earning under 138% FPL.

The Yes campaign has been active from getting signatures to get the question on the ballot to actively campaigning.

The No campaign has been fairly quiet except for taking advantage of the machinery of the state to increase barriers to voting by holding this vote on a mid-summer, mid-pandemic primary day instead of the general election.

Results should come in tonight.

If you are a registered voter in Missouri who has not yet voted, please go vote.

Show me the votes for Medicaid ExpansionPost + Comments (17)

This week, New Jersey took a step to wrap around subsidies to more individuals on the ACA individual market for the 2021 plan year.

The Center Square reports the details:

A bill that adds a 2.5 percent tax on health insurance premiums in New Jersey is headed for Gov. Phil Murphy’s desk…

The money collected from the tax will create a $300 million “Health Insurance Affordability Fund” which will be allocated to New Jersey’s state-based exchange.

Effectively,what New Jersey is doing is claiming some of the revenue that had been allocated to the now repealed 3% Health Insurance Tax (HIT) that went to the federal government before it was repealed in the December 2019 budget deal. The 2020 HIT was built into the premiums even as it is no longer being collected. Some of it might eventually be sent back to beneficiaries through Medical Loss Ratio (MLR) rebates in a few years. What New Jersey and a few other states have looked into, is add a state based tax that will grab most of the revenue that had been built into 2020 premiums and removed from 2021 premiums and redirected it to state coffers. From a current law baseline, this increases premiums by 2.5% while from a current pricing baseline, this is no change while current law would have led to a premium reduction.

New Jersey intends to use this money to improve affordability for both subsidized and non-subsidized individuals. The non-subsidized market will benefit as a portion of these funds will be allocated to pay for the New Jersey share of its 1332 reinsurance waiver that lowers gross premiums by 15%. The subsidized market will see lower net premiums as additional wrap-around subsidies will be offered via the soon to open New Jersey state based marketplace.

However, New Jersey is leaving policy leverage on the table. New Jersey has very tight premium spreads for both their cheapest Bronze and cheapest Silver plans compared to all other Healthcare.gov using states in 2020. This is not surprising. New Jersey has a multi-insurer market which usually will reduce spreads as well as Medicaid expansion. However many other Healthcare.gov states have these attributes and wider spreads.

New Jersey has an intentional set of policy choices that reduces premium spreads.

New Jersey requires Bronze plans to have comparatively low deductibles and very high actuarial values for its class. 64% and 65% actuarial value Bronze plans have far less potential to create a wide premium spread from silver plans that have a baseline nominal actuarial value of 66% to 72% than a not allowed in New Jersey 59% AV Bronze plan. New Jersey silver plans have minimal variation in plausible configurations.

These are deliberate policy choices which may ease the information processing challenge of buying individual market health insurance and limit top-side financial exposure. Those are legitimate goals. However, these goals are in tension with minimizing gross and net premiums.

If New Jersey allowed for lower actuarial value and thus higher deductible/out of pocket Bronze plans, both subsidized and non-subsidized buyers would have more affordable options due to drops in both gross and net of subsidy premiums. This step could be done with limited increases in catastrophic exposure as co-insurance dollars would be transformed into deductible dollars. Individuals with large but not catastrophic expenses such as a knee replacement would pay more cost-sharing while individuals with massive expenses (say a bout with malignant cancer) would still be hitting the same or lower out of pocket maximum.

Everything has a trade-off. New Jersey is electing to make a set of trade-offs that result in higher premiums and modestly lower “normally expensive” out of pocket cost sharing. This policy choice has resulted in New Jersey trying to smooth off some of the trade-offs’ implications with other funds. It could also be smoothed by allowing for lower actuarial values to be sold as Bronze plans.

New Jersey and ACA affordabilityPost + Comments (8)

There are two very short passages in a pair of ACA papers that are tertiary findings of those particular papers that have greatly influenced how I think about the ACA since the first time that I saw them.  These two passages would have seemed bizarre to me in September 2013 when I was first started to write at Balloon-Juice and was working way too many hours putting together the UPMC Health Plan provider networks.  But these two passages have significantly shaped my thinking.

First, Coleman Drake and Jean Abraham** looked at the affordability of plans after silverloading in the July 2019 Health Services Research.  They conducted a sensitivity analysis that had a null result that is utterly fascinating:

This was a sensitivity analysis to eliminate possible alternative explanations from their primary finding.  They wanted to know if insurers were acting strategically.  I think this is a reasonable  question.  Were insurers recognizing that they were in monopoly counties and acting differently than if they were in deeply competitive markets?  A profit maximizing insurer should act very differently.  In a competitive market, premium spreads are likely to be low while in a monopoly market, premium spreads are whatever the insurer wants them to be.  Yet Drake& Abraham found little evidence of deliberate strategic behavior.

That is fascinating!

Paul  Shafer, Sarah Gollust, Seciah Aquino, Laura Baum, Ericka Franklin-Fowler ## and I recently published a large study on the effects of advertising on ACA enrollment in the Russell Sage Foundation Journal of Social Sciences (open access).  We found some small effects of advertising on enrollment:

These models demonstrate that both sponsor and the specific product marketed are consequential for whether advertising volumes are associated with Marketplace enrollment rate (plan selections per hundred thousand population younger than sixty-five). Specifically, model 1 finds no relationship between health insurance ad volume overall and the Marketplace enrollment rate. Model 2 disaggregates by sponsor type and we see large positive estimates of the marginal effect of federal and both types of state ads (own and other state) on the Marketplace enrollment rate; however, only the volume of ads from other states reaches statistical significance. Finally, model 3 disaggregates the private ads further by product, showing that private non-Medicare, non-Medicaid, and private Medicaid airings are positively associated with enrollment, offering evidence of a spillover effect across products….

However the finding that has changed my thinking is also a secondary finding:

Higher premiums are counterintuitively good for enrollment.  This is a finding about the unusual gearing of price linked subsidies.  A 2% gap at $500 a month is significantly smaller and less valuable than a 2% gap at $1,000 per month for a benchmark premium.  Premium spreads for the cheapest silver and cheapest subsidized plan are often far greater than 2%.  It is the intuition that prompted my interest in zero premium plans and really gets to the functionality of the CSR silverloading mechanics that have fascinated me since Spring 2017.  This little factoid gives a number to an inkling.

Neither of these facts are overwhelming, but they have shaped how I look at the ACA pricing dynamics over the past two years and how I think about the market as it is today and how it could be in 2022 or 2023.

* * The three of us write together frequently and, in the past 20 months, prolifically.

## Paul and I write together a lot.  We’re in the midst of cooking several papers for late 2020 or early 2021 submissions. I’m co-authoring with Paul, Sarah, Laura and Ericka another paper that is currently under review.

A few sentences that shape my thinking on the ACAPost + Comments (7)