How much would expanding Medicaid help in states that haven’t accepted the expansion?

Perhaps the single biggest story about the implementation of the Affordable Care Act has been the battle in states deciding whether to accept the Medicaid Expansion. The expansion is perhaps the single most important tool in the ACA’s coverage expansion tool kit. It takes 50 state-based single payer systems and drastically expands eligibility for them, which is the single largest progressive victory in politics since the Great Society. Other important Medicaid reforms drastically streamline the application procedures and eliminated asset tests to draw out formerly eligible people who might have gotten tangled up in the system or not bothered applying because the state made you apply in person on Tuesday between the hours of 2:30 p.m. and 3:04 p.m.

However, the expansion is not some magical talisman that instantly enrolls all eligible individuals. Some people will remain ignorant of the program despite the best outreach efforts, while other will not enroll for any variety of reasons. And more to the point, 24 states hadn’t fully taken advantage of the Medicaid expansion by the beginning of 2015. Pennsylvania, Indiana, Montana and Alaska have all signed on this year, leaving 20 holdouts (half of whom were in the former Confederacy – but I digress).

The cool thing that we have some real data of how ACA has actually performed on the ground over the last two years, we make some interesting dynamic projections of what would have happened had some states accepted the Medicaid, instead of simply discussing the number of people who would be eligible for help under the expansion.

I mean, heck, this Charles Gaba fellow has been counting the people who actually signed up for coverage for two years, I might as well take one shot at figuring out who would have signed up if they could have. 

To accomplish that, follow me below the fold, where I build a simple interactive regression model to project the reduction in the uninsured population in states that haven’t expanded Medicaid. Don’t worry; I’ll label the scary part where I work through the model so you can skip the simple summary where I discuss the results in plain English (but you really should read the model section, it’s rather of important and it makes fun of Bobby Jindal).  

Some follow-up technical notes on State-based exchanges

When I reposted the last post as a DailyKos diary, a commentator engaged me a bit on a few ideas to improve the basic concept. He suggested for controlling for partisan control of the state government as a binary variable (cooperative == full Democratic control=1; non-cooperative==full Republican control=0). I thought a tertiary variable (0=GOP 1=split 2=Dem) might be a bit better.

Having an independent variable for partisan control might soak up some of the variation for cooperation with implementing the ACA, but the problem is that any partisan control variable would be extremely highly correlated with State-based exchanges. There were very few complete Democratic states without at least a shared exchange (West Virginia and Illinois jump to mind), and only one GOP controlled state (Idaho) with an exchange. I'm not sure how useful it would be in practice at sorting out variation not associated with exchanges since the two are so strongly correlated.

In any case, it was a thoughtful piece of feedback and definitely worth the brief conversation we had.

Another idea I thought of was to take quarterly data for each state to increase the number of observations for each state from one to six to track changes from the end of 2013 through the second quarter of 2015. Taking the data from cross sectional data to cross-sectional-time series in this manner would create 300 observations (vs. 50) and increase leverage dramatically, while allowing us to track over-time change in states setting up and taking down exchanges and or expanding Medicaid at different points.  Of course, as my old methods prof John Jackson used to stay "No good deed goes unpunished" and we'd have to control for the serial correlation in the states with either a fixed or random-effects model.  At least we wouldn't have to worry about panel-corrected standard errors, seeing that we're dealing with the universe of states, not a subsample.  And there would also be the problem of increase error within states for each quarter, since the Gallup sample would slip to alarmingly low levels for some states, increasing the margin of error around the uninsured rate for a given quarter.

And that's assuming that I could even get the more specific data out of Gallup.

I don't really have the time to try out either of these ideas right at the moment, but anyone in Internet land can feel free to try and report back. I'm rather interested.