Contextualizing Health Care Reform: The Case of Exchange Premiums

When Americans talk about most aspects of their life, they are generally good at putting the discussion in context.  The quality of the weather on a given day is considered relative to what is expected for this time of year.  The record of our favorite sports team is considered relative to pre-season expectations.  Our chidrens’ performance in school is considered relative to their past performance and our sense of their underlying ability.  Yet this ability to contextualize often gets completely lost when we discuss policy issues.   For example, a deficit of $500 billion is often considered “enormous” regardless of the size of the economy to which it is compared.

This issue becomes particularly important when evaluating health care reform.  Any changes in our health economy that are associated with health care reform must be placed in the proper context.  In terms used by economists, changes must be compared to the relevant “counterfactual,” or what would have happened absent health care reform.

An excellent example of this issue is interpreting the data that is now being released on premiums in the health insurance exchanges.  Critics of the ACA have been calling attention to examples of double-digit premium rises on state exchanges from 2014-2015, while supporters have pointed out cases of premium increases for the same period, sometimes equally sizeable.  Neither side puts these changes in context to ask: how are the premium changes in the individual market in 2014-2015 comparing to what we saw before the ACA?

The purpose of a project that I recently completed for the Commonwealth Fund was to address this very question.  I used excellent data collected by Jon Gabel and his associates at the National Opinion Research  Center (NORC) from state insurance web sites over the 2008-2010 period.  I focused on this period because ACA regulations began to impact the individual insurance market in an important way in 2011, through features such as medical loss ratio limitations and regulatory review of premium increases in excess of 10%.   These data are not complete; they are only collected for a subset of states, and in some states for just a sample of insurers.  But the conclusions that they deliver are clear and don’t change even as the sample is restricted in various ways.

This research yielded two important conclusions.  First, before the ACA, there was 10-12% annual inflation in the individual insurance market.   This sets a clear baseline for interpreting the changes that are now emerging from the exchanges: the context for interpreting these changes should be the double-digit rate increases that existed before the ACA.  As a result, premium increases of less than 10-12% should be considered an improvement, not a failure.

Second, before the ACA there was enormous variation in the rate of premium growth, both across states and within states.   Premium growth in individual states varied from 3% to over 20% during the 2008-2010 period.  And variation in premium growth across insurers was even larger, with some insurers seeing premiums fall by 10% or more, while others had premium increases on average of 50% of more.  Therefore, it is hard to draw any strong conclusions from one insurer’s, or even one state’s, rate releases.  To get a full national picture of the ACA requires data from many insurers in many states.

Of course, this analysis ignores other aspects of the ACA which will lower premiums and reduce their variation.  Large tax credits, which apply to the vast majority of exchange purchasers, will lower effective premiums.  And probably the most important source of premium variation across individuals was not average changes in premium for a given insurance product, but massive changes in premiums as individual health varied; such “health rating” is no longer allowed under the ACA.  (A side note: one additional benefit of the ACA is that the collection of premium information in the individual market is no longer such an enormous burden, which will allow for more careful premium comparisons going forward).

Variation in exchange premiums is just one of many changes that the ACA will bring to bear on the U.S. health economy.  But the general point holds that for evaluating any such change, we must contextualize it by considering the “counterfactual” change that would have occurred in the absence of the ACA.  Doing so for individual insurance market premiums shows clearly that any strong conclusions about premium growth from 2014 to 2015 will require waiting until data are available from a large number of insurers in many states.   And when such data is available, the relevant comparison will be with the double digit increases featured in these markets before the ACA.


Dr. Jonathan Gruber is a Professor of Economics at the Massachusetts Institute of Technology, where he has taught since 1992. He is also the Director of the Health Care Program at the National Bureau of Economic Research, where he is a Research Associate. 

Morning Consult