President Trump recently said that the Republicans in Congress should wait for the Affordable Care Act to “fail” and then replace it. This is a facile rhetorical strategy. First, it shifts the blame: Replace with what? Trump has been simultaneously evasive and vacuous on the question of how he would fix what, in his own words, is a complicated issue. But more interestingly, the strategy is brazen: believing something is clearly broken but declaring that we shouldn’t fix it. This is a horrible way to make policy and, in the case of health insurance, quite possibly nonsensical.
Is waiting for policy failure a good way to prompt policy change? As an argument that it isn’t a good idea, I refer to Hurricane Katrina. Faced with Category 5 winds and surging waves, New Orleans was ultimately swamped by rain and flood waters in 2005. It caused 1,836 deaths and was the costliest natural disaster in history, costing approximately $108 billion.
This was clearly a policy failure — even though the hurricane itself was a random event, the risk was known, as were methods to mitigate those risks. As stated in the report prepared by the bipartisan congressional committee that investigated preparations for and responses to Hurricane Katrina:
Even with its hurricane protection system, it was common knowledge that New Orleans was susceptible to hurricane-caused flooding. The risks of a major hurricane and flooding in New Orleans had been covered in the general media — by Scientific American (October 2001) and National Geographic (October 2004) — as well as in emergency management literature.
Furthermore, the government was aware of the risks (for example, see this 2004 report). Proper preparation and response to events like Katrina are systems that require planning, resources, and maintenance — just like complex markets such as those for health insurance in the various states. When systems clearly collapse, we can’t just hit reset, wipe the board clean, and restart everything anew. Rather, when a system fails, damage is done, and some of it is permanent.
When has an insurance market collapsed? Even if we agree that letting the ACA fail is prudent, how would we know when it has failed? Well, it won’t be when the insurers all leave “the market,” because there is no national insurance market (that’s the first problem). Rather, there are at least 50 health insurance markets: Health insurance is jointly regulated by the federal and state governments. Essentially, each state has a marketplace that insurers can join to offer insurance plans that comply with the ACA’s requirements and thus qualify for federal subsidies based on the purchaser’s family income.
Under the ACA, these plans must be priced according to a community rating system. In a nutshell, community rating systems limit the set of factors that can be used by the insurance company when setting the price of the plan. For example, the price can depend on your age, whether you smoke, and where you live (referred to as your “geographical rating area”). These geographical distinctions are central to the debate about whether the ACA is collapsing.
What you can buy depends on where you live. Over the past two years, some insurers have been pulling out of insurance marketplaces, including in selective ways (i.e., dropping coverage of certain rating areas). A recent report by Avalere documents this, including this very helpful figure:
As the figure makes clear, an increasingly large number of counties have little or no competition on their state exchange. Many of these counties are sparsely populated and have few people participating on their state exchange, making the offering of a profitable plan under a community rating system more difficult.
The figure also demonstrates the complication of determining when the ACA has “failed”: Every state will have a different experience as time progresses. And on that point, compare two (among many) narratives about the cost of health insurance. (The figures below are drawn from this 2016 report by the McKinsey Center for US Health System Reform.) On the one hand, average premiums are increasing. The figure below demonstrates that in each tier of coverage, more than half of counties are seeing the cheapest premium increase by around 25 to 30 percent.
On the other hand, the next figure (also drawn from the McKinsey Center report linked above) demonstrates that there is significant county- and state-level variation in the direction of health insurance costs across the nation. There are several differences between the data in the median cost figure above and the county-level change figure below.
Perhaps the subtlest of these is that the median cost figure is reporting gross premiums, whereas the figure below is reporting net premiums — which takes into account the federal subsidy, which is based on an individual’s family income. Thus, the county-level figure below is illustrating (at least) two types of heterogeneity in the US “national health insurance market”: Demographics such as age, race, and family size vary widely across different counties, and so does income.
This second variation is particularly salient in the recent debates about replacing the ACA and, in particular, reforming Medicaid. The variation in the county-level maps also partially illustrate why Senate Majority Leader Mitch McConnell (R-KY) is having such a hard time crafting a repeal and/or replacement of the ACA. Simply put, how the ACA is performing depends on who you are and where you live.
So what now? It is of course highly unlikely that the health care debate will die away, but the experience of Hurricane Katrina demonstrates that even sustained attention is not sufficient to avert policy failure. According to the congressional committee, a fundamental failure in the response to Katrina was informational: There was too little coordination of policymaking during the preparation for Katrina’s landfall. The analogy with health care reform is the need for clear discussion and preparation of benchmarks by which to judge the “national health insurance market.”
Without such a discussion, it is not only highly likely that we won’t have agreement on when to reform it but also very likely that we will continue to disagree in fundamental ways about how to reform it.