October 14, 2009

Effective Substitue for the Public Option (& Slowing Medical Inflation)

Congress faces a political quandary: how to make health insurance work better yet still be affordable, without raising taxes for subsidies too sharply.

Senator Olympia Snowe is focused on a couple of the most crucial aspects of reform -- general cost inflation and whether middle class families will be able to afford to their health insurance premiums in 2014 or 2016.

At the moment, Senators Baucus, Reid, and Snowe have great influence on what form reform will take, as they work to blend the two Senate committee bills.

Senator Baucus says a Public Option (which would help control costs) is not going to get 60 votes in the Senate, and Senator Snowe only proposes a "trigger" where a Public Option would come into effect in an individual state only after competition in the state is shown to be inadequate.

Does the idea of a trigger sound effective? It would be amazing for an effective trigger to emerge in the legislative details. But if the triggered public plans are limited to individuals states, the resulting public plans would often be relatively small in size and thus lose ability to control costs well.

I think the best proposal is to set up a nation-wide public option plan, and then allow individual states to opt in and opt out, as they like. This way, states where sentiment dislikes the idea of a public option (if there are any states where a majority does in fact feel that way) could opt out of the national public option. Later, if the voters of the state feel a public plan option is better, they can elect different state representatives and opt into the national public plan.

But...in case none of these variations on a public plan are chosen, there is another way to control costs quickly (within a year of implementation) that Congress should use in the absence of any public option plan.

Supporters of a Public Option should think about "medical loss ratio" regulation as a very good alternative strategy.

The medical loss ratio (or "health benefit ratio") is the portion of total revenues from health insurance premiums that an insurer pays out for health care claims. The average in the U.S. is about 80% now. It was well above 90% in the 1990s.

A strong regulation of the medical loss ratio would be as good as a strong national public option in its immediate effect on costs.

New regulation of medical loss ratios is already in the proposed reform, but hasn't been in the spotlight, and so lobbyists have kept this provision weaker than it should be.

Strengthening this regulation is something that both Snowe and Baucus can do politically.

Reform proposals at the moment suggest regulating the medical loss ratio up to a 85% minimum.

That is not high enough.

The insurance lobby claim on Monday that premiums will go up more quickly under reform isn't all smoke and mirrors (although it has plenty of smoke and mirrors). Premiums will go up for simple reasons -- like closing loopholes, for instance. When you pay less out-of-pocket, then you must pay more in premiums, even before including increases from health costs inflation. But of course, there will also be subsidies for lower income households to allow them to afford this coverage.

The real bottom line elements for costs in the medium and long term are health system inefficiencies and the bigger issue of health care inflation itself.

When I think about the political tradeoffs between insurance reform and costs (the costs of health insurance premiums and federal subsidies), I keep coming back to the implication that two kinds of cost reduction *must* happen for reform to succeed.

1. Near Term Cost Reduction, which requires either:

  • A) "health benefit ratios" (medical loss ratios) must go *much* higher in exchange for the individual mandate -- such as 90% instead of only the meager 85% currently on the table.
  • OR
  • B) some method of powerful cost competition must come into effect quickly, such as a national public option default that states can opt out of by state choice, and/or a powerful opening of the exchanges such as Senator Wyden proposes (with effective regulation that designates certain types of policies offered ("bronze," "silver", etc.) so that shoppers are able to make real apples-vs-apples comparisons of actual benefits covered).

  • We really need either option A) or option B) to help control costs near term.

2. Long Term Cost Reduction, which is the type of reform I've focused on in my blog, such as how to get to more "integrative" medical practice (like the Mayo clinic) by changing how treatment is paid for.

But near-term cost reduction is key to the Congressional political quandary I think.

Insurers must give to get.

With the mandate, they will get millions of new customers.

In exchange for continuing to have their profitable niches with the new captive customers the necessary individual mandate will create, and without public competition, they should raise their payout ratios much more sharply than only from the current 80% up to 85%.

They will be able to increase payout ratios easily. They will no longer need to spend huge amounts of money to weed out and exclude sick people or struggle with providers over every claim detail, under reform.

Under reform, their administrative costs are going to go down sharply. That much is already in the cards.

We should be talking about regulating the medical loss ratio up to 92-94% (90% at a absolute minimum). I lay out in this post how a "health benefits ratio" payout level of 94% is a good example of reasonable efficiency.


(This part below is meant for policy wonks or Congressional staffers: There is one important wrinkle in such a regulation to allow for....

First, the adequate regulation of medical loss ratios (e.g. above 90%) is needed when the insurance exchanges are not extremely active and powerful -- either by lacking sufficient participation (without Senator Wyden's plan to open up the exchanges), or by lacking sufficient ability for shoppers to compare policies if the regulation specifying benefit levels doesn't really make comparing "bronze" policies real apple-to-apple comparisons. In other words, when option B) above is not in full effect, then option A) becomes necessary, due to lack of competition to hold prices in check. But option A) has to allow for whatever provisions gradually bring about option B) over time.

This note is about a further, crucial issue though.

When an insurer considers whether to try delivery system innovation to lower medical costs, they must be able to expect to gain extra profit for some period of time.

If an insurer is required to pay out for instance 92% of their premium revenues to claims, and then that insurer considers whether to spend money to innovate and set up a pay-for-outcome system, and this new system might in turn quickly lower actual claims costs while maintaining or improving quality... then a perverse effect from the regulation could occur. Lower total claims would force the insurer to rebate more of their policyholder premiums at years end, actually decreasing their profits unless they rapidly expand by picking up a lot of new policy holders. This temporary fall in profit would create a disincentive for this kind of innovation. The danger from the insurer's point of view is that before their new lower payout medical costs allow them to gain more new customers in their marketplace (how active is that marketplace, with how many potential customers?), there will be a lag in time, during which they will not yet have much increase in new policyholders but will have lower payout ratios, and thus must pay rebates, and thus lower their profits below normal.

Of course, the outcome of a good innovation that lowers costs should instead be an increase in profits, at least for a few years.

Therefore, a regulation of medical loss ratios must include a provision to allow insurers to *lower* medical costs through delivery innovation, without being required to rebate the excess retained premiums above the required payout ratio (as described below) that result during some fixed period of time, such as 2-4 years (or until the entire industry adopts their innovation). Instead, they would be rewarded with more profits for their beneficial innovation.

One way to accomplish this provision would be to allow insurers to use an industry average payout cost (per type of procedure or condition), or their previous payout costs per condition, instead of only their own new payout costs. They could optionally use an annually established payout level average for certain conditions or treatments, these averaged basis numbers (for each condition or treatment) calculated annually by the regulator for the entire region health industry, and maintained in an online table, etc. In short, the insurer would be able to temporarily use the old costs levels to calculate their legal payout ratios even while they have lower payout ratios in reality due to the savings in the new delivery system. They can act for a specified time period as if they are paying the old costs that existed previous to the new cost levels their innovation creates.

In this way, an insurer is motivated to try to improve value by increasing medical effectiveness per dollar, as it will increase their medium term profits directly (and also in the longer run as their better cost levels allow them to win new policy holders for the long run).

Something for insurers themselves to consider: This would payoff even on the level of individual conditions. For instance, if a knee replacement costs on average in claims paid in the medical region $19,500, but an insurer manages to innovate and lower costs to $17,100, then the insurer would be able to use the average number -- $19,500 -- for the purpose of accounting for what they paid versus the regulatory requirement of payout ratio vs. premiums, keeping the difference as profit for a specified time period (such as 3 years). Thus good innovation in just one procedure would be rewarded by profit. The same applies to any costly and common treatment. Finally, since this insurer can lower overall premiums in response, they can attract a larger customer base, spreading their overhead costs and thus increasing long term profits also.


  1. Hal, thanks for your post. I have a question: if the issue is supporting a public option, and costs of private plans are already too high anyway, what's wrong with taxing those plans a little bit more to cover the public option's cost?

  2. There are many different possible versions of a Public Option being considered at the moment, but most have the quality that there is no ongoing subsidy for the public option plan entity from the federal government.

    Instead, all the likely public option plans would only receive a one-time, initial infusion of money to help get started. After that, the public option plan would rely on premiums alone to cover it's costs, and would not receive any further federal subsidy.

    The only subsidies that would be ongoing would instead be for individuals/families that cannot afford to purchase the minimum credible coverage required by the mandate, regardless of whether they choose a public or private plan.

    These subsidies to individuals and families will be part of reform regardless of whether some kind of public plan is even established.

    If a public plan exists, it would be only be one among many optional insurers that individuals with subsidies could choose. In this way the playing field between the public and private plans is level.

    As to the entirely separate question of whether to tax high-cost private plans as one form of revenue, this is not a question I've done any research on.

    There is a blog post here though that addresses the broad philosophical issue of whether subsidies for lower-income people to buy insurance is a just idea. For that, look in the archive on the right under September for "The Ultimate Social Issue..."