8/31/2017

Tom Heintjes: Welcome to another Economy Matters podcast. I'm Tom Heintjes, managing editor of the Atlanta Fed's Economy Matters magazine. Today we're once again joined by Melinda Pitts, an economist at the Atlanta Fed and director of the Atlanta Fed's Center for Human Capital Studies. Melinda coauthored a recently published working paper that looked at the financial impact of people losing health insurance, and she's agreed to discuss her research with us today. Melinda, thanks for joining us again.

Melinda Pitts: Thanks for having me.

M. Melinda Pitts, Research Center Director of Center for Human Capital Studies, at the recording of a podcast episode.

M. Melinda Pitts at the recording of this episode

Heintjes: And I should note—longtime listeners to this podcast might notice that this is your third appearance here. You're the first three-time guest! A few more and I think you win a toaster oven or something like that.

Pitts: I'll hold you to that.

Heintjes: [laughter] Melinda, health care is such a timely topic—and one that inspires such strong opinions. So when you first told me some time ago about the research you were conducting, I was very excited about getting you in the studio to discuss it. Before we get too far into our discussion, I should mention your coauthors: Laura Argys and Andrew Friedson of the University of Colorado at Denver, and Sebastian Tello-Trillo of the University of Virginia. Tell me about the genesis of your research—what led you to look into this specific situation?

Pitts: When my coauthors, Laura and Andrew, and I were working on our debt and mortality paper that I discussed on my last visit here, Sebastian was working on a paper that examined the impact of losing public health insurance on health outcomes. We got together at a conference, and it seemed a natural extension to examine the financial impact of losing health insurance as well.

Heintjes: Right. Well, as far as I'm aware, this is the first study to look at the detailed financial impact of losing public health insurance.

Pitts: That's correct, at least to my knowledge. In addition, we also believe that we are the first to look at the consequences of losing eligibility for any public support program.

Heintjes: Oh, interesting. Melinda, roughly speaking, what did you find was the financial impact on individuals of losing public health insurance?

Pitts: In general, we found the impact was very similar to the impact of the 2001 recession for the representative person in Tennessee. Of course, the impact would be much larger on individuals that actually lost insurance coverage. In addition, we found the impact was larger for those with lower initial credit ratings or credit scores.

Heintjes: Most research and analysis of this topic looks at the effects of extending health care to new populations, but your work looks at the flip side of the coin: the impact of removal of health care from a population that previously had it.

Pitts: That is correct. There is a relatively large body of research that looks at the effect of gaining health insurance, including examining the effect of the recent expansions associated with the Affordable Care Act [ACA]. But given the uncertainty surrounding some of the expansions in health insurance associated with the ACA, we think it is interesting to explore it from both sides.

It's also the fact that it's not necessarily symmetric—i.e., the impact of gaining health insurance could be different from the impact of losing health insurance—and that's for a few different reasons.

Heintjes: What are some of those reasons? Could you cite some?

Pitts: Sure. One issue has to do with how much debt you've taken on. When individuals have higher income—which, when you have insurance it's kind of like driving up your income…you have more financial security, you take on more debt, and that debt doesn't go away when the insurance is lost. So now, you don't have health insurance and you have higher levels of debt than you had before.

Heintjes: Makes sense.

Pitts: People behave differently when they have health insurance. It's called moral hazard.

Heintjes: Like you might go skydiving, or things like that?

Pitts: [laughter] Exactly, exactly. That's the classic example—you take up skydiving. You pick up habits that you wouldn't if you were uninsured, and when they lose insurance they may have a higher latent risk of an adverse event, i.e., breaking a leg from skydiving.

Heintjes: Sure, sure.

Pitts: It also could work the other way. Having had insurance increases the number of face-to-face visits with a physician, and that could have led to the opportunity for more education and case management, and that could carry over into periods even after you've lost insurance, and therefore you may have a lower chance of a bad health outcome.

Heintjes: Right. Well, let's get a little more granular at this point. What set of specific circumstances did you examine to conduct this research?

Pitts: In 2005, Tennessee began the process of cutting approximately 200,000 individuals from its Medicaid program. This is the largest-scale disenrollment that has been documented. It mainly affected childless adults and people that were unable to get private insurance.

Heintjes: So describe TennCare briefly. This was the state's Medicaid program?

Pitts: Yes, TennCare is the Medicaid program in the state of Tennessee. In the 1990s, they transformed the program into a managed-care system with the hopes that the money saved by using managed care would allow them to expand eligibility to other individuals not traditionally covered by Medicaid. These include individuals up to 400 percent of the federal poverty level—which is mainly childless adults—and the uninsurable, which has not ever been done.

Heintjes: Right. It was something of an experiment at the time, if I recall.

Pitts: Yes, it was very much an experiment, and very much praised. And people were hopeful that it would work.

Heintjes: And that was at a time when managed care was maybe going to be something of a panacea for rising health care costs?

Pitts: Exactly. And it did keep health care costs down for a short period of time, but then they went back up, and so it did not turn out the way they had hoped, and Tennessee could no longer continue providing the coverage they had. So in 2005, they began disenrolling people from the plan, and as you mentioned before—about 200,000 people lost coverage from TennCare, and current families…now, you have to have around 200 percent of the federal poverty level to qualify, among other criteria.

Heintjes: Let's discuss briefly the data that you used. Can you discuss the sort of data that resulted from these disenrollments?

Pitts: From Tennessee, all we were able to get was the Medicaid coverage rate in each county. So we have how many people were covered by Medicaid, by county, in Tennessee. On the financial health side, I used the same data I used in the paper I discussed last time I was here: the Federal Reserve Bank of New York's Consumer Credit Panel, which is based on credit histories provided by Equifax, and it's also only available to researchers in the Federal Reserve System. It is a nationally representative sample of 5 percent of the U.S. population with credit reports. And as I note, it is very confidential.

Heintjes: Yes, you always note this.

Pitts: We cannot identify individuals, but it does have rich information and we were able to match each person in Tennessee to the county Medicaid coverage rate.

Heintjes: And no name, address, gender, education, anything like that?

Pitts: Unfortunately, from a researcher perspective, no. But from a privacy issue, that is correct.

Heintjes: Right. I guess it's not big news that losing health insurance has a big financial impact. But beyond that, what did you find out about the effects on people's financial well-being?

Pitts: From an individual perspective, definitely when you lose health insurance it's going to have a day-to-day financial impact. But we were able to explore a little bit more by looking at their credit scores, which is similar to your FICO score, and the amount of severely delinquent debt, which means debt that is 90 days past due or more.

So we compared both across counties—relying on the fact that some counties had much higher rates of TennCare coverage initially than others—and across age groups within counties. By "age groups," I mean we compared those aged 21 to 64 versus those that were Medicare eligible—because yes, there are some people that are Medicare age eligible that had Medicaid, but even if they lost Medicaid they still had Medicare as a buffer, so the impact was going to be less for that group.

Heintjes: Right, that's an important distinction.

Pitts: Yes. And so we found that just across all measures we looked at, credit scores fell, so you were less able to qualify for future credit because credit ratings went down. We found that the amount of debt that was severely delinquent and the number of accounts that were severely delinquent both increased. And we found that the bankruptcy rate increased relative to comparison groups as well.

Heintjes: You've noted that when President Lyndon Johnson signed Medicare into law in 1965, he said that part of the intent was to help insure people against the financial shocks that medical costs can induce. And I guess we see now that, decades later, that really hasn't changed.

Pitts: I agree. Recently a lot of the discussion and focus of research has been on how health insurance affects health outcomes and access. Although these are important topics, at the end of the day one of the main purposes of insurance is to mitigate financial shocks.

Heintjes: Right. Well, you and I talked about this subject a number of times, and I've heard you say that, in a sense, monetary policy is health policy. I guess this is kind of a good reminder of that.

Pitts: Yes, I think this work is a reminder that nothing happens in a vacuum. When you take away something—in this case, health insurance—the impact can be much more widespread and affect areas that may not have been considered when these decisions were being made.

Heintjes: I don't want to use this opportunity to broaden this conversation into a national scope, but while you were conducting this research there was a possibility of millions of people around the country losing their health care. Did that possibility lead you to do any even informal, back-of-the-napkin calculations or anything like that? Is it even possible to infer repercussions on a national level from this sort of research?

Pitts: From my perspective, broaden away. This uncertainty about the future of health insurance nationwide is the motivating factor for this work. If this was only going to happen once in Tennessee, then it was kind of irrelevant, the long-term impact. One of the intriguing things about looking at this Tennessee population is that it is relatively similar to those that gained insurance through the Affordable Care Act and the ones that are most at risk of losing it again—these are largely childless adults and the uninsurable. Given the similarity between the populations, it suggests that we would expect similar outcomes on a national basis.

Heintjes: Well, it's certainly something we're all going to be watching closely down the road. I also want to be sure to note that we have a link to Melinda's paper on our website, frbatlanta.org. I really hope you'll check it out—it's a unique look at a very timely and important topic that will remain, as we've noted, part of the national conversation for some time to come. Melinda, thanks so much for spending some time with us today—I really enjoyed this conversation.

Pitts: Thanks for having me back again, and I look forward to earning that toaster next time.

Heintjes: Well, you're getting closer. [laughter] And that brings us to the end of another episode of the Economy Matters podcast. Please join us next month, when we sit down with Atlanta Fed economist Julie Hotchkiss, who will discuss her new recent research into the impact of monetary policy changes on families. Thanks for being with us.