qotd: Gruber and colleagues call for more research on health plan competition

The National Bureau of Economic Research
NBER Working Paper No. 20140
May 2014
More Insurers Lower Premiums: Evidence from Initial Pricing in the
Health Insurance Marketplaces
By Leemore Dafny, Jonathan Gruber, Christopher Ody

Abstract

First-year insurer participation in the Health Insurance Marketplaces
(HIMs) established by the Affordable Care Act is limited in many areas
of the country. There are 3.9 participants, on (population-weighted)
average, in the 395 ratings areas spanning the 34 states with federally
facilitated marketplaces (FFMs). Using data on the plans offered in the
FFMs, together with predicted market shares for exchange participants
(estimated using 2011 insurer-state market shares in the individual
insurance market), we study the impact of competition on premiums. We
exploit variation in ratings-area-level competition induced by United
Healthcare's decision not to participate in any of the FFMs. We estimate
that United's nonparticipation decision raised the second-lowest-price
silver premium (which is directly linked to federal subsidies) by 5.4
percent, on average. If all insurers active in each state's individual
insurance market in 2011 had participated in all ratings areas in that
state's HIM, we estimate this key premium would be 11.1% lower and 2014
federal subsidies would be reduced by $1.7 billion.

From the Conclusion:

Given the incipiency of these markets, this study is but a first step in
what will surely become a deeper and broader literature on insurance
exchanges and the nature and significance of competition among exchange
participants. There is substantial room for further research on how
competition affects pricing and other outcomes in this market. Future
studies will be easier to execute once information about consumer
enrollment decisions has been released, and once the market is in
longer-term equilibrium. These conditions will allow researchers to
apply well- established supply-side methodologies to studying
competition on the exchanges. Such research will permit more-nuanced
conclusions and recommendations regarding the impact of competition and
competition-related policies on various outcomes of interest. Given the
large federal role in developing and regulating the exchanges, and in
subsidizing the purchase of plans offered on the exchanges, research on
how competition affects consumer choice and insurer behavior is of
critical importance.

http://www.nber.org/papers/w20140

****


Comment by Don McCanne

This study makes an attempt to support the theory that as the number of
competing insurers increases within each of the 395 ratings areas for
federally facilitated marketplaces, the premiums decrease. Though the
title is quite clear that "more insurers lower premiums," the actual
data are quite fuzzy.

If you look at Figure 2 in the report, there is considerable scatter
around the weighted regression line, and if you eliminate just a few of
the outliers from the 395 ratings areas, there is no obvious correlation
between numbers of insurers and premiums.

Nevertheless, to make their point they used two what-ifs. They project
that premiums for the second lowest cost silver plan were 5.4 percent
higher than they would have been simply because one insurer - United
Healthcare - declined to offer plans through the exchanges. Further,
they project that if all insurers already active in each rating area
were to participate in the exchanges then the premium for this silver
plan would have been 11.1 percent lower. It seems to be a leap to state
that competition has been shown to be successful in lowering premiums
because it would have been so with other competitors participating in
the exchanges even though those other competitors did not actually
participate. At least it demonstrates an unwavering belief in the
competition theory being tested, which really wasn't tested, but, what if?

Of greater concern is their concluding comment. They state, "There is
substantial room for further research on how competition affects pricing
and other outcomes in this market." Also, "Given the large federal role
in developing and regulating the exchanges, and in subsidizing the
purchase of plans offered on the exchanges, research on how competition
affects consumer choice and insurer behavior is of critical importance."

Jonathan Gruber, who helped develop both the Massachusetts plan and the
Affordable Care Act, is a co-author of this NBER paper. He had told us
that the proven model of single payer had to be rejected because of lack
of political feasibility, yet in its place we had to adopt a model using
health plan competition - a model that needs further research to see how
it will affect consumer choice and insurer behavior. What? We enact this
for the entire nation, and then we'll see how it works?

Actually we do not need further research on private insurance plans
competing within an exchange. We have had decades of experience with the
Federal Employees Health Benefits Program, and with California's CalPERS
for public employees and the Pacific Business Group on Health for
employer-sponsored plans. Each of these exchanges offered a choice of
competing private plans. Yet health care costs continued to escalate at
intolerable rates. Plans competing within exchanges have not been
effective in slowing health care inflation to levels found in other nations.

The authors contend that more research on health plan competition is of
"critical importance." Baloney. We've had all the research we need on
health plan competition. What we need is to get rid of the private
health plans, fix Medicare, and then provide it to everyone. Now.