Root canals and tax advantaged accounts

I scheduled a root canal for the end of the month.  Once that tooth is taken care of, I should be able to sleep and think better.  Duke dental insurance will pick up a small portion of the expense.  Most of the cost will be out of pocket.  I will throw it on a credit card and then submit most of the personal component of the procedure’s cost to a health reimbursement account.   By mid-March, I should get a nice big check back from that account to pay off the credit card while still collecting airline points.

An HRA is a tax advantaged vehicle that employers can set up for their employees.  The money can be spent on a wide variety of medical costs including dental/vision services as well as cost sharing obligations.  I get a significant tax advantage from it.  I don’t  have to pay income or FICA taxes on the part I’ve added to the account on top of what Duke added.  The part that I throw into the account is immediately taken off the top of my total compensation for tax purposes.

I am getting about a 30% net tax discount for my root canal. Thanks for the help everyone!

2019 Me is doing well.  I have a great job that pays well while having me do things that I love to do.  My wife has an awesome job.  We are secure. We have both liquid and quasi-liquid reserves.

2010 Me was stretched in fifteen different directions. I was out of work, my wife was massively underemployed and we had a one year old at home. My wife and I were uninsured. My daughter was on CHIP (best insurance we ever had!). I got to spend the year with my daughter as not paying for day care and surviving on unemployment was the best of bad options.

Assuming that 2010 Me had an HRA with a non-zero balance, and if I needed a root canal, I would be getting a 17% discount (FICA and 10% tax bracket) from the Feds.

That year was probably the most financially stretched year that I will experience as an adult. This year looks to be the easiest year financially. I’m not going to turn down the help but this is a jacked up system. I am getting the most federal aid for a root canal in a year when I have the most capacity to grumble, mumble and bear the cost. In 2010, a root canal done by a licensed and trained dentist would have been a major financial emergency. I would have been flipping the coin behind my mouth and my mortgage. And I would be getting very little help.

Deductions to taxable income are asinine when those deductions are applying to medical expenses. Credits or capped deductions with bump-ups to send the help to the people with the highest marginal utility of the last dollar would make way more sense from a policy and a moral perspective.

States, policy innovation and proofs of concept

Many states are proposing a series of experiments with their health care markets that are aimed at expanding coverage, increasing actuarial value, and limiting provider payments.

Medicaid Buy-ins

  • New Mexico
  • Nevada

Public Options with Medicare-like rates

State based mandates

  • New Jersey
  • Massachusetts
  • Vermont
  • California
  • Maryland

Downpayment Plans

  • Maryland

Expanded Subsidies

  • California


I agree completely with Adrianna.

These states will provide evidence of what can work, what trade-offs are real versus illusive, what some of the unexpected interactions may be, and the challenges of figuring out how to cover more people for roughly the same cost.  The liberal  experience in health policy from 1994-2007 was a long consensus building session as to what could be done within self-identified political constraints and limitations.  Massachusetts with a large Democratic super-majority in both chambers of the legislature was the proof of concept of the three legged stool approach.  The three legged stool was a combination of guaranteed issued/community rated insurance that was backed by significant low-income subsidies to make the insurance affordable and a mandate to get and keep healthy people in the risk pool.  Medicaid was the base of the coverage expansion with the private market taking more of the load up the income scale.  The three major Democratic primary contenders in 2007 all bought into variants of this plan and the major veto players in the Democratic Senate caucus were also on board.

I think that the states are limited in what evidence they can provide on a pure single payer system.  They don’t have the counter-cyclical fiscal capacity nor the expectation of seeing waivers approved to unlock significant federal fund flows for that project.  However they can test the impact of expanding subsidies, offering government price leveraged plans and using Medicaid further up the income scale.  These are all needed and worthwhile policy experiments.


Disclosing contextless factoids in Box 12-DD

The Affordable Care Act requires the disclosure of the cost of employer sponsored insurance (ESI) on employee’s W-2s.  This is reported in Box 12-DD.  It is a big number that is purely informational.  The theory of change is that people will see that big number and realize that health insurance is a substitute for cash compensation and then demand better value for their ESI health insurance dollars.

I don’t think there is enough context in the large group market for this theory of change to make any sense.

Health insurance premiums are a function of how many services are used by a group and the price of those services.  Benefit design, network configuration, incentives and rewards can all shape both the quantity and price of received services.  In fully insured market segments that are risk adjusted, the premium a group replay will be reasonably reflective of the choices a group makes as well as its underlying demographic composition.  This is not the case in large group, administrative services only (ASO) contracts.

ASO contracts are when the insurer takes on no financial risk from an employee group.  ASO is the dominant means of providing insurance to most Americans who get ESI coverage.  The insurer “merely” processes claims, builds networks, handles customer service complaints and manages contracts.  The employer takes on the full cost of the claims.

Two employers with 1,000 covered lives where 999 of them are identical and have their employees in the same exact plans with the same exact network and the same utilization pattern at the same exact prices for those 999 covered lives will see very different Box 12-DD numbers if in Firm A that last person is a 25 year old male in perfect health or at least no utilization past a flu shot for the year while the last person in Firm B is a seven year old living with treatable Cystic Fibrosis.

Firm B’s Box 12-DD will show a number that is $400 per covered life per year higher than Firm A’s Box 12-DD.

Box 12-DD for ASO/self-insured firms is a partial reflection of the health status of everyone in the pool.  Firm B could show “spending discipline” if they find some way to legally not cover that seven year old kid even if they expanded their network and lowered the cost sharing by $100 per member per year.

For people who get their insurance through work and whose work is self-insured with an ASO contract, Box-12-DD is a contextless factoid that informs little.

Market functionality in insurance types

Yesterday, Gin & Tonic rightly called me out for a big assumption:

is extremely attractive assuming deep and well functioning individual markets

Quite an assumption there.


I think that I have an understated assumption that the health insurance markets are far less functional of a market compared to most other insurance product markets unless there is a massive regulatory thumb on the scale. I think and believe that the big differentiator of health insurance compared to home, auto, fire and life insurance is the combination of politically and economically strong counter-parties and very localized monopolies with significant barriers to entry.

Next week, my wife and I are closing on a new house. We’ve spent most of this week chasing down all the final threads of paperwork. My wife had to take care of the homeowner’s insurance policy. She was able to look at several national carriers that operate in North Carolina. They all offer functionally similar coverage. The big differentiator is customer service and bundling discounts. She chose to go with Allstate because we get our auto insurance through them so the bundled discount is significant and the friction cost of setting up another bill and keeping track of it is not worth any marginal savings.

If a triggering event occurs for the home owner’s policy, they will send an adjuster and then write a big check. After the check is deposited into our account, we then get to go argue with local general contractors to fix whatever the problem was. The same basic story applies for an auto insurance policy. The local auto repair market and the local home repair markets are fairly fragmented, price-taking markets with modest information asymmetry. Almost all general contractors and mechanics will take a personal check and all of them will take cash. The counter-party/doer barrier to entry for new insurers to enter a region is reasonably small for home and auto insurance. It is very small for life insurance as a check is written and accepted for deposit in any US bank.

Health insurance is more complex. There is massive informational uncertainty and asymmetry. I can evaluate reasonably well if my roof has been repaired but I am guessing for a while if my knee will be right after an intervention. I, as a patient, have a harder time evaluating quality as well. I can assess whether or not the check cleared the bank. I can assess whether or not Joe’s Autobody did a good job of replacing my right front fender. I have a harder time determining if a hospital/doctor/pharmaceutical intervention fixed the underlying problem, masked the symptoms or alleviated the worst of the situation without making things go back to the status quo pre-event.

Health insurers build a network of preferred contracted providers. When something goes wrong that triggers a health insurance claim, that claim comes through the contracted network. I think this is a significant barrier to entry when there are locally concentrated medical markets. An insurer can only get a good price on services with either massive government shoulder throwing as in Medicaid, Medicare and Medicare Advantage, or by being able to steer large populations to preferred rate providers and away from not-preferred entities. Building a network is a chicken and an egg problem. An insurer gets good pricing with a big membership base. They get a big membership base because they have good pricing on a good network. It is a natural neck deep moat for dominant local insurers against new entries. New entries, as we saw with co-ops in the ACA, have a hard time getting competitive pricing on their provider networks until they can build up the membership base. This means selling loss leaders for several years and lighting a lot of money on fire.

This is true for the individual market.  This is true for large group markets.  It is not as true for Medicare Advantage as price setting regulations in Medicare sets a ceiling of roughly 110% of Medicare Fee for Service as a pragmatic anchor point in pricing.  Medicare Advantage has other start-up  challenges in risk adjustment but building a non-exorbitantly priced network while having a low membership base is not one of those challenges.

I think that I assume that local/regional health insurance markets that don’t have significant price regulations are not the most functional markets.  They reward size and incumbency so new entries need to be able to climb over some very large hurdles.  I think this is less true in the individual market as the decision making agent is an individual with only family needs in mind and more true in large group where the decision making agent is someone in HR with a strong budget and a moderately strong scream constraint to work against.



Blue Wave — Arriving

The 116th Congress is currently being sworn in.

The gavel is being handed over.

The Blue Wave arrived.

Open thread

Image of one of the Healthcare Gavels from Napa Valley Register

HRAs and downward nominal wage rigidity

Earlier this year, I was talking with a very astute healthy policy observer. They pointed out that the Trump Administration’s push towards much wider use of Health Reimbursement Accounts (HRA) to pay for employer coverage on the individual market is extremely interesting in a wide variety of ways. The basic HRA concept is that the employer would place an age/geography/family size adjusted sum of money into employee accessible accounts. Employees would then use the exchanges to purchase insurance on the individual market. The employer effectively makes explicit healthcare costs and off-loads the expense of administration of healthcare benefits to the exchanges.

This is interesting in a wide variety of ways. We talked about how the current system of requiring the reporting of insurance premiums on income tax W-2s (Box 12-DD) does not make people too sensitive. Right now, at most box 12-DD is an interesting factoid as the sum of money that Duke spends on my health insurance is intermediate through multiple opaque layers of value creation and destruction. I don’t know if other large employers can provide my family with the same or better realized value on health insurance. If I had to shop on the Exchange and Duke offers $12,000 for family coverage and Other Employer offered $10,000 in an HRA for family coverage, I can make a real estimate of the value trade-off. But right now, if Duke holds premiums constant while increasing deductibles, I have a hard time determining if my value proposition and trade-offs have changed.

In 2012, Paul Krugman was banging the drum on downward nominal wage rigidity.


Via Mark Thoma, a new paper from the San Francisco Fed offers stunning evidence on downward nominal wage rigidity, a topic I’ve written about before.

What the paper shows is that many, many workers are getting precisely zero wage growth in dollar terms:

This stuck with me. People don’t like taking nominal pay cuts as their debts are mostly denominated in nominal dollars. Employers don’t want to demoralize their workforce with nominal pay cuts so they cut head count and fringes to reduce costs and then rely on several years of nominal zero percent changes in an inflationary environment to produce real pay cuts.

Health insurance in an employer sponsored world is a major source of potential savings as Box 12-DD numbers aren’t real. Employers could go to a tiered network or higher deductibles or more restrictive plans while still providing something that most of their remaining employees would consider “similiar” enough to a 0% raise. Those moves are invisible or close enough to invisible so that the scream minimization constraint is satisfied. ESI health insurance premiums are very well hidden compensation for employees although it is very clear compensation for employers.

A question that sticks in the back of my policy brain is “What if the 2008/2009 Great Recession happened again?” This question does not dominate my thinking but it gets asked at least a couple of times a year as I think about possibilities.

Moving to an common HRA arrangement for employer sponsored health insurance should make the cost of premiums far more explicit in a repeat version of 2008-2009 employment shocks. An employer who is seeking to cut compensation costs by taking an average of $800 per employee per year out of the health insurance budget can’t do that any more by narrowing the network, restricting the formulary and switching from a PPO to an HMO. Instead it is an explicit cut where the HR rep has to convince the workforce that this year they received on average $10,000 in employer premium support and next year they will receive $9,200 in average premium support and that is still a great deal. And it may be a good deal for individuals who qualify for subsidies as they will get topped up but for folks who don’t qualify for subsidies, this is an explicit wage cut.

The idea of HRA breaking the employer role of selecting health insurance is extremely attractive assuming deep and well functioning individual markets or at least individual markets that are no more dysfunctional than the current large group markets. However, making explicit the cost of health insurance may exacerbate a repeat of 2008-2009.

Expansion and elections

This happened yesterday:

Next month, Maine will expand Medicaid as well with coverage being retroactive to last July.

Elections matter for expansions. And right now our initial evidence is that an expansion is mostly a one way ratchet with some backsliding. Governors and political parties that don’t like expansions can prevent an expansion from happening as we see in North Carolina, Georgia, Florida and Texas. However governors that don’t like expansions can not fully roll back an expansion that has already been implemented. Kentucky is the most aggressive state in trying to limit its expansion of medical insurance coverage (private and Medicaid) after a change in political regimes, but the roll-back is by hassle and hurdle of work reporting requirements and informational FUD rather than explicit repeals.

We should be seeing at least three more states expand Medicaid in the next 365 days, so the ratchet will continue forward.