To Be Fair, This Was Probably The First Time

I've heard that it takes a big man to admit when he is in the wrong. And so doing my best impression of a big man, I will admit once again that I was indeed mistaken, incorrect, blind to reason, or, with the humble clarity of Alan Greenspan, "I found a flaw in the model that I perceived is the critical functioning structure that defines how the world works, so to speak." Except unlike Greenspan, I actually mean it. So to speak.

In the comment I left on Dave's first post written on the subject of my incontrovertible wrongness, I wrote the following (and yes, I am not only stooping to pull a quote from the comment thread of this blog, but my very own): "My reason I objected to the bill [sic], both in political and policy terms, was that I assumed it would not work as promised." Which I realize is a bit like offering as an excuse, "I know that I was wrong, but to be fair, I was making incorrect assumptions." And with a grammatical typo to boot.

The point I was trying to make however is that I wrote that first blog post without having seen the CBO evaluation of the bill, or Nate Silver's evaluation of their evaluation, or anything resembling reasonably concrete estimates of any kind. Which was of course very unprofessional of me. But more importantly, what I am laboring to say is that my reaction to the bill was based solely on what I read about its contents. And though I am no expert on health care economics or policy, I concluded from a purely deductive approach (uniformed, if you want to be a dick about it, but since the dickishness is all mine, let's say a priori) that there was no way, no-how that this bill would be able to put any downward pressure on premiums and would most likely do just opposite.

But according to the CBO (and notable angry internet number cruncher, Nate Silver) that is 100% objectively incorrect. "But I still don't really get it," I wrote. "How do premiums and overall costs get reduced without a) more competition or b) stricter regulations?" And I wrote that not as a way to deflect the fact that I was wrong, but because I was and continue to be legitimately puzzled. The numbers themselves, as graphically presented in Dave's post below, provide incontrovertible proof (unless you want to argue with the methodology of the CBO, which I do not) that America looks less screwed with the bill than without.

I arrived late to the number party, but now that I've finally shown up, better late than never, I'm curious where the numbers come from. Because think about it as if you didn't know better: the mandate assures that the insurance industry has a captive market (if demand for health insurance wasn't inelastic before 2014, it sure as hell will be after), restrictions on anti-consumer activity will force insurance companies to accept higher costs and, there is no longer a single public entity that can increase competition at the national level--how do premiums not shoot through the roof?

I've tried my best to read through the CBO report on the Senate bill's affect on premiums. Or at least, through as much as my attention span would carry me. Here is what it says about premiums in general:

In general, the premium for a health insurance policy equals the average amount that an insurer expects to pay for services covered under the plan plus a loading factor that reflects the insurer’s administrative expenses and overhead (including any taxes or fees paid to the government) and profits (for private plans).
[...]
Although the factors affecting premiums are complex and interrelated—and thus can be difficult to disentangle—this analysis groups the effects of the proposal on premiums into three broad categories:

• Differences in the amount of insurance coverage purchased
• Differences in the price of a given amount of insurance coverage for a given group of enrollees, and
• Differences in the types of people who obtain coverage in each insurance market.

The most popular and, at least for me, obvious issue here is that of market concentration. The public option was supposed to increase competition, and the absence of a viable replacement (though as I understand it now, licensed non-profits are permitted by the bill to operate on a national level) seemed to many, myself included, like the removal of a vital restriction on premiums. The issue of competition is opaquely categorized under the second category "Differences in the price of given amount of etcera." And what struck me is how minimal a factor it is assumed to be under the CBO analysis. So called "rents" or, to use an inconvenient term, "premiums," higher prices derived simply from the fact that consumers don't have many other options in a particular market, are implicitly considered, but are not given much weight.

But that doesn't mean that the report dismisses the issue entirely. It should also first be noted that, again according to the CBO, the establishment of an exchange itself, which gives consumers the ability to compare the various health insurance plans available to them side by side, in which information would at long last be centralized and its presentation standardized, and in which the administrator of the exchange (the government) has ability to curtail abusive practices, namely sudden premium increases, will, with or without a public option, increase competition and therefore both lower cost and perhaps improve quality.

So yes, competition does seem to play a role in reducing premiums, but not by much. Why? As far as I can tell, that isn't explained in this particular CBO report. As many of you may recall, Uwe Reinhardt had his particular explanation. But easy explanations aside, I suppose it may just be an empirical fact:

[T]he authors [Leemore Dafny, Mark Duggan and Subramaniam Ramanarayanan]

then considered the specific impact of the 1999 merger of two insurance industry giants, Aetna and Prudential Healthcare. The two insurers were active in most local markets, but their share of the markets varied significantly in different areas. They find that after the merger, premiums increased in accordance with the overlap of market share. Extrapolating from that result, the researchers estimate that consolidation increased premiums "in a typical market" by 2.1 percent. "Our results confirm that Americans are indeed paying a premium on their premiums," the authors conclude. "However, consolidation explains very little of the steep increase in health insurance premiums in recent years. While 2.1 percent is large in absolute terms -- and large relative to industry profits -- "it pales in comparison to the doubling in real premiums for our sample during the same 1998-2006 time period."

So evidently, at least below a certain threshold I would imagine, market concentration doesn't exert a major impact on prices--at least not one like the do-or-die proponents of the public option insist exists.

Moving on to the issue of the mandate, this is only discussed in terms of how it will reduce and not raise premiums. According to the report, the mandate will draw in predominantly those who occupy the venn-diagram of people who explicitly choose to go without health insurance and people who are not covered by their employer, rather than, as I assumed, those who have otherwise been excluded from the market for financial or health reasons. This first group described by the CBO report is made up disproportionately of young people, who in turn are disproportionately healthier and thus less likely to incur major medical costs, as opposed to the excluded second group mentioned above, those whom I assumed would be making up the bulk of the new insurance industry customer base, who are in opposition to the first, very high-risk. Additionally, it is also stated by the report that the influx of the 30-40 million insurance customers will increase slightly the economies of scale at the disposal of certain insurance companies. Both these factors should reduce premiums if they affect them at all.

Lastly, there is the issue of certain prohibitions and restrictions on current anti-consumer cost-saving measures used by the insurance companies, namely recision and discrimination from coverage based on risk. Without concurrent premium controls of some kind, I had assumed, the logical result would be higher household health care costs, either in the form of higher premiums, lower quality care, higher deductibles, high co-pays, etc.*

In some respects, the CBO report agrees (imagine my surprise). For "nongroup" purchasers of healthcare specifically (individual buyers who don't get their insurance through their jobs), the mandated higher standard of coverage** would indeed increase premiums: "premiums would have to increase to cover the resulting costs."

On the other hand, this requirement would evidently have very little affect on so-called "group" markets, both small and large, since those basic standards of coverage are usually met anyway when insurance is purchased en masse by an employer. In addition, administrative costs can also be expected to decline, according to the report, since a higher degree of standardization will limit the legal, claim review, and other costs required to implement discriminatory care of this kind. I would assume, however, (though this isn't stated either way in the report and as mentioned above, my deductive thinking has been flawed lately) that the savings achieved by this latter effect (the standardization of plans) won't completely offset the higher costs incurred by the insurer in covering the other-wise excluded group of elderly diabetics (costs that are presumably transferred at least in part to the customer); if that were the case (if it were the case that the insurance companies spent more on risk-mitigating administrative costs than they saved through exclusion of coverage as a way of mitigating that risk) those companies would presumably have standardized contracts on their own as a practical cost-cutting measure.

On the other hand, and to come back to a previous point, some of the net cost increases incurred by insurance companies forced into offering standardized, non-exclusionary contracts would be absorbed by the influx of those same young folks alluded to above (that's people like us by the way), who will enter the individual market and make that particular risk pool a little less risky than it otherwise would be, thus offsetting at least in part the sickly geezers.

I should maybe take this moment to point out the obvious: from the perspective of many individual buyers, this whole discussion is moot. As Dave pointed out, the subsidies provided under the Senate bill are fairly generous, with those at up to 400% of the poverty line receiving some kind of assistance, and so for households, the price tag will be lower regardless. That means less bankruptcy, less medical debt, and less overall misery. I am usually anti-misery. But as explained in the CBO report, the total cost (the sum cost for the insured individual plus that absorbed by the provider plus the subsidies) will decline or remain about the same everywhere except in the much smaller individual market (up to a -3% decline for the "large group" market, between +1 and -2% change for the "small group" market, and a 10-13% increase for the 57% subsidized nongroup market).***

So there it all is: how you get lower or constant premium prices without a comprehensive and easily understood approach (e.g. the public option). Factors cited in the overall analysis include increased competition resulting from the creation of the exchange, largely cost-reducing externalities resulting from the mandate, increased economies of scale, reduced administrative costs through standardization, and the taxation of higher cost plans.

My overall impression is the following: the bill recognizes (or at least implies the Senatorial perception) that the bulk of the problem with American health insurance is in the relatively small nongroup market. As philosophically flawed as employer-based health insurance might be, those who get insurance from where they work receive on average pretty good healthcare. The bill therefore has very little affect on the group market premiums, though the affect that it does exert seems to be largely downward. The market that is seriously affected, the nongroup market, will consequently see its premiums increase. But along with that increase comes an enormous improvement in quality of coverage (in terms of transparency and stability certainly) and a boatload of subsidies. Because the nongroup market represents such a small (though under the bill expanded) portion of the entire market, those subsidies can, according to the larger CBO assessment, be paid for and them some with the revenue acquired through exchange participation fees, taxes on certain plans, penalties on non-participants, and fat shaved off Medicare and elsewhere. The larger fiscal issue wasn't really what I was looking at in this post, but I thought I should mention all that if for nothing else than logical continuity.

This is all getting a little complicated, I realize, and forgive me for all the details. If there was a single provision in the bill that I could point to and declare, "That is what I missed all along!" I would do so. But fundamentally I think it was the lack of elegance to Senate's solution that led me to assume from the get go that the whole endeavor was, as I put it in my previous post, "horseshit." Without a satisfying single cost-reducing mechanism, it was hard for me to understand how a series of piece-meal reforms could really stand in the way of that generational trend, premium inflation. I suspect that this is the reason that so many people on the left assume the bill can't work. Big glaring problems call for big glaring solutions.

But at a certain point, and I suspect that this point comes when the vast majority of experts have crunched the numbers and demonstrated beyond a reasonable non-conspiratorial doubt that there is a flaw in the model that you perceived is the critical functioning structure that defines how the world works, so to speak, it's probably best to concede the point.

So, having satisfied myself, I happily concede the point.

*I should also admit here that I certainly overstated (out of ignorance, not bad faith) the lack of regulation regarding premium that I expected to result from a ban on discriminatory coverage. As the report explains on page 13: "Administrative costs would be reduced slightly by the general prohibition on medical underwriting, which is the practice of varying premiums or coverage terms to reflect the applicant’s health status; nongroup insurers incur some administrative costs to implement underwriting." (emphasis added)

**Page 1: "Policies would have to cover a specified set of services and to have an “actuarial value” of at least 60 percent (meaning that the plan would, on average, pay that share of the costs of providing covered services to a representative set of enrollees). In addition, insurers would have to accept all applicants during an annual open-enrollment period, and insurers could not limit coverage for preexisting medical conditions."

***These estimates were made before the exclusion of the public option, but as is explained on page 15 of the report: "the provisions regarding a public plan would not have a substantial effect on the average premiums paid in the exchanges."