By now you’ve probably read about why Obamacare won’t “spiral into fiery, actuarial doom“—at least, not on account of unconvinced young invincibles. That’s the main takeaway from a recent Kaiser Family Foundation analysis, which found that low enrollment among the young adult set would have a minimal impact on premiums. The 3:1 age band actually prices in risk-by-age quite well, so—all else equal—having fewer young people couldn’t drive up premiums much.
But age is only one dimension of the adverse selection story. Health status is critically important: regardless of age, the risk pools need to achieve some balance of “healthy” and “sick” individuals to be stable. Failure to enroll enough healthy people into exchange plans could result in adverse selection. The media has seized on this economic phenomenon, and its devil child: the adverse selection “death spiral”, wherein rising premiums drive the healthy out of the risk pool in a cycle of self-perpetuating oblivion.
Of course, the law anticipated that some adverse selection might occur in early years, so it has mechanisms to protect itself. I wrote about the role of risk corridors and subsidies in buffering adverse selection. There’s also limited evidence for health insurance death spirals in the literature. People like to point to premium hikes New York’s individual market, which enacted “pure” community rating, without flexibility for age, in the early 1990s. But Thomas Buchmueller and John DiNardo found nearly identical enrollment shifts (toward an older risk pool) in two “control” states during the same time: Connecticut, which enacted more modest reforms, and Pennsylvania, which enacted no reforms.
Moreover, the “conventional wisdom” about death spirals fails to account for one minor detail: market forces. At least, that’s the argument Linda Blumberg and John Holahan make in a new RWJF policy brief :
[W]hile insurers may experience some losses in 2014 if adverse selection occurs, market competition will make it difficult for them to recoup those losses in 2015 by increasing premiums substantially. If enrollment grows throughout 2014 as technical problems are overcome and outreach efforts continue, leading over time to a broader mix of health care risks enters the Marketplaces, then competitive pressures are likely to dissuade insurers from ratcheting up premiums. In a competitive market, insurers must set premiums for 2015 based on expected enrollment in 2015, not based on any losses that occurred in 2014. Simply put, insurers cannot recoup losses without achieving significant market share, and achieving market share requires that they price their products competitively for expected enrollees in the coming year. [...]
Bidding high makes a plan less likely to be the second lowest cost plan in an area, which means that enrollees would be required to pay more for that coverage than the percent of income cap provided for by the federal subsidies. This would decrease plan enrollment and hurt profits. Insurers presumably do not look at their business over a two-year horizon where any year’s shortfalls must be made-up in the following year; they cannot recoup losses if they have no market share, so they must make decisions for the longer term.
There remains an outstanding possibility that insurers could exit the exchanges as a reaction to financial losses in 2014, or due to future uncertainties. That’s not something the authors address, but I’ll try: I’m skeptical that they’d pull out after year one. Death spirals, by their nature, require time to unfold; to borrow a quip from a friend, there’s a reason we don’t call them “death slingshots”. For the market to unravel, you need fundamentally broken risk pools, not a bad year chalked up to a bad website. Considerable time and resources have been invested in the ACA (see also: the industry bending over backward to accommodate the administration’s mercurial deadlines).
And market power is at play here, too. The new exchanges represent a pretty substantial slice of the potential individual market consumer base; the more enrollees an insurer has, the more power it wields for negotiating prices with providers. Exiting the exchange is likely to be accompanied by a pretty substantial blow to that market power.
There’s one last crucial variable: we don’t actually know what level of risk the insurers baked into their premiums. Obviously they couldn’t anticipate website woes on the scale that we’ve observed in the last three months, but I find it hard to believe that they’d draw up projections that assume a perfectly balanced risk pool from the start. Call me crazy, but something makes me think that actuaries are better at hedging than the blogosphere.