Wharton's Ashley Swanson discusses her new research on preferred pharmacy networks.

Health care is one of the most pressing issues in the 2020 presidential race, and most of the Democratic presidential candidates are supporting an expansion of coverage through what’s being called Medicare for All or a public option system. An overall fix for the out-of-control costs of health care is tough to craft and implement, but a series of smaller fixes could be a more feasible and effective way of dealing with the system’s problems. One such idea would be preferred pharmacy networks, a relatively new tool used by health insurers to steer consumers to pharmacies where they have negotiated lower drug prices.

This cost-reduction mechanism is the focus of new research by Ashley Swanson, an applied microeconomist and professor of health care management at Wharton. She is author of “Preferred Pharmacy Networks: Health Care Savings on the Margins,” a brief published by the Wharton Public Policy Initiative that examines the potential savings for enrollees in Medicare Part D. Swanson and Amanda Starc, a professor of strategy at the Kellogg School of Management at Northwestern University, are also co-authors of a research paper on the topic, “Preferred Pharmacy Networks and Drug Costs.” Swanson joined the Knowledge at Wharton radio show on SiriusXM to discuss the findings. (Listen to the podcast at the top of this page.)

An edited transcript of the conversation follows.

Knowledge at Wharton: Could you break down the main idea of this research?

Ashley Swanson: The big punch line of the article is that these preferred pharmacy networks have saved the plans that implemented them about 1% of overall enrollee drug costs. Relative to the overall costs of the health care system, that’s chump change, but these kinds of 1% solutions do start to add up.

Knowledge at Wharton: Are there a lot of these preferred networks in place right now?

Swanson: They’re really new. They came into being around 2010, 2011. In 2011, we only saw in our sample about 13% of plans having these kinds of preferred networks, but they really took off like wildfire. By 2014, about 70% of plans had a preferred network, and I think it has increased since then. The plans in the Medicare Part D program have really cottoned onto this idea, and it has taken off.

Knowledge at Wharton: Do you really think small fixes can lead to a significant change in our health care system?

Swanson: Absolutely. This is an idea that I first heard articulated by Fiona Scott Morton, an economist at Yale, and I think it’s really sharp. We have this really complex health care system, and it sounds exhausting to try and make a hundred tweaks to address a hundred tiny inefficiencies. However, in the U.S. and globally, there’s a mass of talented economists and health care researchers who study health care programs, and they’re working hard to find these opportunities and share that information back to policymakers.

Now the counter-argument is that we don’t want to be pennywise and pound foolish. But even if the ultimate goal is to implement Medicare for All or dramatically overhaul the health care system, overhauling is hard and time-consuming and going to take a minute, so we should be addressing the inefficiencies we’re aware of in the meantime.

Knowledge at Wharton: In looking at your work, Medicare Part D is a research focus for you. Tell us why.

Swanson: It’s a really fascinating program. It’s a government benefit, but the plans are offered by private firms, and there are a lot of interesting free market forces at work. Consumers choose their plans. Plans compete for business. Plans try to interact with other health care providers, like pharmacies and pharmaceutical manufacturers, to try and control costs. It’s an interesting marriage of public and private, so it’s really fertile ground for economists to study.

Knowledge at Wharton: Can you take us further into the concept of these preferred pharmacy networks and where the benefits lie?

Swanson: Most people, if they’ve had the misfortune to interact with the health care system, will have heard of narrow networks. This is something that’s a very hot topic these days. When people talk about narrow networks, they’re talking about narrow provider networks like, “Oh, this plan doesn’t have my doctor. This plan doesn’t have my hospital. This stinks.”

In our paper, we study something kind of similar. It’s an analogy in the prescription drug market — narrow pharmacy networks. Narrow pharmacy networks restrict access or at least incentivize enrollees to be steered away from some pharmacies within their prescription drug plans. And the plans, arguably at least, use that power to steer enrollees to lower cost pharmacies and even maybe get some additional leverage in negotiating with the pharmacies.

We should note that pharmacies fly under the radar a little bit in the U.S. relative to drug companies or hospital systems. But it’s a very concentrated market. You can probably think of what the top four pharmacies are in the U.S.

Knowledge at Wharton: Yes, there has been a lot of consolidation in this sector over the last few years, which obviously plays a role in the cost structure for consumers.

Swanson: Yes, absolutely. That’s right.

Knowledge at Wharton: One of the goals of Medicare Part D is to keep costs down. That’s also the benefit of preferred pharmacy networks, correct?

Swanson:  Yes, the argument for these kinds of preferred pharmacy networks — or of any other kind of cost-control levers that exist within these Part D plans — is that they are going to try and bargain for discounts from drug manufacturers, from pharmacies, from any part of the supply chain that gets drugs into the hands of enrollees in the program. They can get costs down. Whether they pass those cost savings onto enrollees or the government really depends on the context, and I’m afraid I don’t have a strong answer for that in this context.

Knowledge at Wharton: Part D doesn’t cover every pharmacy in the U.S. There are independent pharmacies that are not covered, correct?

“We have this really complex health care system, and it sounds exhausting to try and make a hundred tweaks to address a hundred tiny inefficiencies.”

Swanson: Something that we were a little bit surprised by is that it even goes deeper than that once you start talking about these preferred pharmacy networks. There are these network adequacy requirements in the Part D program that say, “We don’t want an elderly enrollee in the Part D program to have to drive 100 miles to get to a pharmacy.” That would not be a well-functioning program. These network adequacy requirements say that if you’re in an urban area, you have to have pharmacies within X miles. If you’re in a suburban area, it’s a little bit higher. If you’re in a rural area, it’s even a little bit higher. Basically, the program requires that these Part D plans have a pharmacy within convenient access of every one of their enrollees — or at least the vast majority of their enrollees.

For the networks overall, you might have some independent pharmacies excluded if you can set up a convenient, adequate network without including every single pharmacy. But typically, you’re not going to see the big chains excluded. You’re not going to see CVS excluded. You’re not going to see Walgreens excluded. The network adequacy requirements I just described don’t apply to them.

You can actually have preferred pharmacy networks where most of the pharmacies in your local area are not included. Suppose someone is enrolled in a plan where the preferred pharmacy network includes CVS, but it doesn’t include Walgreens. Now, if Walgreens is their closest pharmacy, they need to make a decision. Are they going to go a little bit further — whether it’s walking, driving or taking public transit to go to the CVS that’s cheaper — or are they going to pay $6 to $8 more for a month’s supply of their drug to go to their convenient pharmacy?

Knowledge at Wharton: How are the preferred networks in the United States set up and designated? I think that’s something that the consumer probably doesn’t understand.

Swanson: That process plays out as a kind of pairwise negotiation between the issuers of these Part D plans and the pharmacies — maybe a consortium of pharmacies, independent pharmacies, or the big retail pharmacy chains. They’re actually negotiating with each of these pharmacy chains or other entities and saying, “Would you like to be in our preferred network? If so, what prices are you going to give us?”

Knowledge at Wharton: More expensive prescription drugs are a significant hardship for people with lower incomes who are on a tight budget to begin with.

Swanson: Yes, absolutely. It really depends on what part of the income distribution you’re in. For the very low-income enrollees, they’re going to be low-income subsidy enrollees in the Part D program. That means they’re never going to pay more than, say, $4.50 for a one-month supply of a generic drug, or maybe $6.30 for a branded drug. For them, these preferred network distinctions don’t really matter because you hit the co-pay max before that price differential makes a difference. However, if you’re just above the LIS (low income subsidy) threshold, you’re still relatively low-income, and those are the people that are going to get hit hardest.

Knowledge at Wharton: How much comparison shopping goes on when you’re talking about getting the prescription at X price from one pharmacy, compared to Y price from another pharmacy?

Swanson: This is actually really interesting. I teach a lot of classes on health economics, so I see these papers come out all the time. There’s this huge body of evidence in health economics literature that, when it comes to health care, people don’t shop. Even when you’re talking about things that you schedule ahead of time — you have time to shop, you have time to do a little bit of research on, say, lower limb imaging. People don’t, either because they’re not aware of all the providers available, or they’re not aware of the prices at those providers. A lot of information barriers might come into play. People do not shop, so there’s not a lot of opportunity for savings by trying to give stronger price incentives to consumers for their health care, like doctors and hospitals.

However, when it comes to pharmacies, we have really good evidence that, at least when it comes to these preferred pharmacy networks, people do shop. For unsubsidized enrollees — the people who don’t hit those $2 or $6 co-pay maximums — we find that these co-pay differentials between preferred and nonpreferred pharmacies move about 8 percentage points of market share, which is quite large relative to the business of the preferred pharmacies. It’s about 16%. For this context, preferred pharmacies and prescription drug fills, we do see that people will shop a little bit.

Knowledge at Wharton: In the paper, you bring up the concept of steering. What does that include?

“Even if the long-term goal is more of an overhaul, why not chip away at these problems in the meantime?”

Swanson: Steering includes the decision I talked about before: I know that CVS is in my preferred network for my plan and Walgreens is not. Am I going to pay $6 to $8 more to go to Walgreens because it’s more convenient to me? That’s the steering we’re talking about. It’s switching from one pharmacy to another based on this co-pay difference.

When it comes to prescription drug utilization, the fact that we see it for this elderly Medicare Part D population suggests that in other contexts where people have more access to internet resources, more access to information about price differentials across different options, what we find for this context might be a lower bound for what you might see more broadly for prescription drug utilization. However, expanding this notion to hospital utilization or doctors, I don’t think that there are a lot of opportunities there just because there’s such a huge body of evidence suggesting that people don’t shop for doctors or hospitals.

I should say one exception to that rule is one of our former students in the Health Care Management Department, Elena Prager, has done some research in Massachusetts looking at tiered hospital networks. That’s a non-elderly population. And she does find some evidence of steering, so I think more research is needed.

Knowledge at Wharton: Then why are there still limited network plans?

Swanson: What our results suggest is that there are savings available from these limited network plans. We find savings of about 1%, once you combine the unsubsidized enrollees that really do respond and the subsidized enrollees that really don’t respond. There are some savings available from using levers like this. However, there’s a trade-off because it does mean limited access.

Knowledge at Wharton: With these potential savings, are you encouraged that maybe we’re starting to tackle at least one of the problems with health care right now?

Swanson: Absolutely. Whether it’s looking at surprise billing for emergency care — this is work that Fiona Scott Morton has done with her colleagues — or overspending on long-term care hospitals — this is work that Amy Finkelstein, an economist at MIT, and her excellent co-authors have done — there are a number of research teams that have identified these kinds of small fixes that can be applied in the health care system. I think that those would be great to implement. Even if the long-term goal is more of an overhaul, why not chip away at these problems in the meantime?

To be perfectly honest, developing this muscle is never going to be a bad idea because even in a single-payer health care systems, they have to make decisions about how to pay doctors, how to pay hospitals, how to pay for drugs. There are going to be inefficiencies in those systems as well, and researchers are going to try and identify those problems and find solutions to them. So, developing this framework for making small tweaks to the system is still going to be productive if we move to Medicare for All.

Knowledge at Wharton: What reaction have you received to this paper?

Swanson: Most of my conversations about this paper have been in talking to economists, but the Wharton Public Policy Initiative has said that some Congressional staffers have reached out to them and are interested in a conversation. Maybe that will go somewhere.

I should say that what we’re talking about here with the preferred pharmacy networks is something that has been implemented by private plans, and they’ve kind of run with it. I think what we’ve done with our paper is quantify some of the trade-offs involved in the selective contracting. It does save money, but it does also restrict access. That is going to be useful for thinking about the costs of these kinds of programs going forward and evaluating alternative policies. Suppose some entity like the pharmacy lobby proposes that the network out-of-state requirement should be applied to the preferred networks. That would probably limit plans’ abilities to do something like this, so that’s where this paper would fit into the national conversation about health care reform.

Knowledge at Wharton: What is the reaction of the pharmacy companies to some of these ideas that would tweak their structure?

“The pharmacy industry — they’re not going to be huge fans of this.”

Swanson: I think it’s very clear that the existence of these preferred networks is a lever that allows Part D plans to push back against a relatively consolidated pharmacy industry. The pharmacy industry — they’re not going to be huge fans of this.

Knowledge at Wharton: What is the takeaway for the average consumer who is on a Part D plan?

Swanson: As with any kinds of restrictions, like high deductibles in your health care plan or restrictive hospital or doctor networks, the fact that premium costs might go down in the long-term might be a bit of cold comfort. Hopefully, if we can get to a place where these preferred pharmacy networks are so well-designed that nobody feels the pinch of them, but they are in some way lower in cost — I’m not sure exactly what that would look like. Then enrollees might feel a little bit better about it. In the meantime, I think it’s clear that the enrollees do not love these restrictive preferred networks.

Knowledge at Wharton: Again, the financial pinch is harder for those with lower incomes.

Swanson:  Absolutely. When you’re talking about people who are living kind of close to their budget constraint as it is, having to pay $6 more for every drug that you take every month is going to hit you. And that’s on top of all of the other financial strains and decisions that they have to deal with throughout the year. Prescription drugs are a small part of your overall consumption portfolio, but this is real money. It adds up.