Kim Hong-Ji / Reuters

American health care is weird and expensive. Those are about the most sure-fire things anyone can say about the sprawling behemoth of insurers, health professionals, hospitals, clinics, and pharma companies in the United States. Is it good at actually making people healthy? For the most part, but sometimes it depends. Is it the best system in the world? Probably not, but some pieces are really good. Why is it so expensive, and can we fix it? Well, those are the three-trillion-dollar questions. Good luck.

A new study suggests that we might consider taking some tips from our pet health-care system. A NBER working paper by Liran Einav and Atul Gupta at Stanford University and Amy Finkelstein at MIT finds that pet health care in the United States has exhibited growth, accessibility, and end-of-life spending patterns that almost directly mirror patterns in the American human health-care system. Their work suggests that pet health care is a useful comparison point for analysis and research. The high cost of (human) healthcare in the U.S. is often attributed to the labyrinthine structure of our insurance-based system. But a close study of the pet health-care system hints that something else might be at work, because it still suffers from the same high cost problem, despite being streamlined by comparison.

I spoke with author Liran Einav about the study, the role of health insurance in driving costs, and the potential usefulness of using data on pet health care to predict trends in human health. Our conversation has been edited for length and clarity.

Vann R. Newkirk II: So how'd you start researching pet health care?

Liran Einav: A lot of the health-economic research is focused on U.S. health care, and one of the common discussion points is 'how can we imagine a different system or how would a different system operate without all the regulations and particular details of the U.S. system?’ One thing, it prompts you to look at other countries, and another thing, to try to think of other industries that somehow have similar features. And the two industries that make me think—in a very abstract way—that there are some similar features with health care are pets and cars.

One way to think about [these industries] is that 'shit happens' and then you need to deal with it. You go find an expert—a vet or a doctor—and hope that he or she knows what they're doing, and then you need to pay for it. The nice thing about the pet industry is that insurance is very rudimentary, so for the most part people pay out of their own pockets. You talk to dog owners and some of them feel just as emotional about their dogs as they are their human loved ones. It made me think, with all the caveats associated with comparing pets and people, [pet health-care] has features that are kind of similar, so we should take a look.

Newkirk: So in a lot of health-policy debates, a lot of people paint insurance as the bad guy. But we've seen with pet health care that it's grown without insurance playing the same role. What's that say to you?

Einav: We're trying to really not overstate anything, because again if you compare two different things you really want to be careful. But the one thing that strikes me is exactly that: Insurers, government regulations and restrictions, and monitoring are much less prevalent in pet health care. But the aggregate patterns between the two are still somewhat similar. At the minimum, those similarities do make me pause and think about how people tend to blame rising health-care costs on insurance and the government—the different data points that suggest inefficiencies in the sector and waste—and then you see similar aggregate patterns going on in a similar industry that has none of that. It makes you think whether we should attribute the bad things to insurance and government, or maybe health is special. Maybe technology in health care is special.

Newkirk: Is there a next phase for this line of research, or is this just a one-off?

Einav: That's a good question. I'm not sure. The nice thing about doing academic research is you don't need to have a completely predefined strategy to raise money. You can decide whatever you want, whenever you want. I can see myself maybe having a follow-up paper with better data. But I could also imagine that this is a one-off and I never go revisit this again. Or maybe it could be other people! One thing about this research is hopefully it triggers other people to go dig up data.

One problem with the pet care that we encountered when we were trying to get better data is because there's less insurance and less regulations, the documentation is not as clean. In American health care, every disease is codified and has a number for every disease. So you can aggregate lots of doctor notes into a set of quantifiable objects. When you look at a pet’s doctor note, it's mostly just text. I guess you could start quantifying text in a variety of ways, but there's no standards about it so you need to reinvent the wheel a little bit. So the data is a bit more messy to begin with and probably less organized because for the most part there are no third parties that need to see the documentation. Vets just provide the service, provide the payment, and that's it. I wouldn't be surprised if over the next five years either we or someone else picks the topic up again and has something deeper and more granular associated with this data. I would certainly hope so.

Newkirk: The temptation with this kind of research is to use it to predict or explain our health-care system. Do you think that's a useful thing here?

Einav: Prediction is for fortune-tellers. I don't know. This is not about prediction; this is about understanding better what drives things, and what doesn't drive things. You're not going to get me involved in any predictions. There are so many other factors that affect health-care spending, health-care outcomes from politics to macroeconomic factors, to technology. Maybe someone will find some useful predictions using this research, but I don't find it to be a particularly useful exercise.

Newkirk: One of the things that caught my eye was the citation of a study from Robert Hall and Charles Jones about health care as a luxury good. Can you talk more about that and how your work relates to that idea?

Einav: Two of my colleagues at Stanford wrote this piece and the basic idea is simple. It started with the observation that as a share of GDP, the U.S. spends more on health care than other developed countries in the world, and our paper is somewhat motivated by the same observation. And the question was: Why is it? Do we like health more? Do we have a more expensive health-care system? Is there something else? The hypothesis that my colleagues contemplated is that the U.S. is richer than other countries, and health care is a luxury good, so as you get richer, you start spending a bigger share of your income by definition on luxury-good items. You could argue that our paper is somewhat consistent with this in the sense that presumably pet care is also a luxury good, and the share of spending relative to income is growing with income across people and probably across countries. So you could argue [our paper] supports it.

If you asked me, I don't think that this aspect is my top candidate for the explanation for why the U.S. is spending so much more, because we are not much richer than some Scandinavian countries in terms of GDP per capita. We're a total outlier in terms of health-care spending as a percentage of GDP. We are not a total outlier at all in terms of income relative to some other rich countries. Again, health care in general is not just one thing. Health care is a bundle of so many different things, and I can imagine that some of the theories will work better for some components of health care and not so great for others. If we knew everything, we wouldn't need to do research, so the good thing is we don't know the answer.

Newkirk: Do pets need health reform?

Einav: [Laughs] Reform of what? Right now, there's very little intervention in pet health care. It's fairly close to a free market. There's some regulation, presumably on becoming a vet, but right now I haven't heard too many dog owners complain too many times about what's going on in the system, partly because there are fewer third parties and fewer restrictions. Once you pay out of pocket for everything, nobody tells you what you can do and what you cannot do. So it's like ... do diamonds need health reform? No, because people who have money buy diamonds for some reason, and people who don't, don't. Nobody complains. I wouldn't spend any effort and political capital in reforming the pet industry.

We want to hear what you think about this article. Submit a letter to the editor or write to