Fewer Radiologists, or More Scans? January 28, 2018 Snippets
As always, thanks for reading. Want Snippets delivered to your inbox, a whole day earlier? Subscribe here.
This week’s theme: It’s not the prices, it’s the volume, Part 2. Plus Autonomic joins forces with Ford, and Social Capital is hiring!
Over the past few weeks in Snippets, we’ve been talking about Cost Disease: why is it that a college degree, a night in the emergency room, or a mile of subway track get more expensive every year? Is there something special about those industries that is driving costs up, or is it a general phenomenon of specialized labor-heavy services in an age of automation and efficiency? We looked at a theory called Baumol’s Cost Disease, which is an elegant explanation for how productivity gains in some industries make other industries less efficient, and for why innovation’s impact on the overall economy can sometimes become a self-defeating cycle. But Baumol’s wasn’t a totally satisfactory explanation: it explains wage increases, but not the piling-on of extra bells and whistles that seem to compound in our tuition and hospital bills every year.
So last week, we looked at Higher Education as patient zero of “escalating cost spiral” syndrome. The evidence was pretty clear: higher education spending doesn’t have a prices problem so much as it has a volume problem. The culprit is something called the Bowen Effect: a one-way ratchet of spiralling cost increases that is driven by two factors. On the supply side, it’s that as non-profit institutions, universities have a mandate to provide as much service as possible for each dollar available, thereby ensuring that new revenue will always be put into providing additional services as opposed to covering future cost increases. On the demand side, it’s that consumers commit early on to paying a huge sticker price with money that isn’t theirs (it’s typically either coming from their future self, in the form of student loans, or from their parents) and are thereby incentivized to consume as much as possible during the time that they’re there. When budgets are set each year, the ratchet moves up and the cycle repeats:
As we said last week, what we’re left with isn’t a prices problem, it’s a volume problem. Now we have to ask the question: is that just a higher-education thing? Or is true for that other hugely expensive area of our economy where no one seems to know how to control costs: health care?
It’s no secret that the United States spends way more money on health care per capita than any other nation. The question, of course, is why: do we have a prices problem, a volume problem, or both? There’s an argument that’s been around for a long time that it’s the prices: our complex private health care markets are failing us, and it’s because we have no way to control costs on unit items like gloves and syringes at hospitals (let alone hours billed by professionals):
Again, if this were true, then it would reflect neither Baumol’s nor Bowen’s as a culprit of runaway health care prices, but rather a complete failure of markets to work at all. (No amount of Baumol’s Cost Disease would explain why a hospital would pay $50 for a pair of surgical gloves, no matter how many pay raises the factory employees have received, because those gloves can obviously be obtained more cheaply elsewhere. Nor can Bowen’s explain it, because there’s no inherent scarcity to surgical gloves: there’s no reason why consuming more of them should make the marginal cost per glove go up.) A lot of people share this point of view on both a logical and emotional level: US Health Care is fundamentally broken, it’s because we have no way to control prices, and it can only be fixed through some sweeping change like moving to single-payer.
The problem with this point of view is that it’s not totally supported by the evidence. A great recent post on Random Critical Analysis walks us through why American health care spending may not actually be so different from higher education spending or other cost disease-afflicted industries: it’s not the prices; it’s the volume. Sure, you can find anecdotes of outrageously-priced items in a hospital. But in general, the culprit isn’t a failure of markets to work, it’s our runaway consumption. This shouldn’t be that controversial an idea!
Of course, what this doesn’t settle is why Americans spend so much more money on health care than any other country. But the answer there may actually be so obvious we don’t see it. In all countries, individual health care consumption increases faster than household consumption or disposable income, and the United States is the richest country in the world.
If this is correct, then what that means is pretty clear: had Americans continued to consume the same volume of health care services that they did a generation ago, then our total health care spending relative to personal income would have fallen, or at least held constant. So when we try to figure out what’s wrong with American health care spending, maybe we shouldn’t be asking “Why is the United States an outlier”, but rather, “Why does this happen at all?” (Of course, the United States is an outlier: in total household wealth. Occam’s Razor strikes again.)
So the actual question we should be asking is, why do nations and their health systems around the world keep getting out over their skis? Why does health care spending so consistently outpace household income, not just in the US but in every country as they get richer? Well, at a basic level we may have answered the question last week. Health Care actually isn’t so different from higher education after all: it’s subject to the same one-way ratchet of mandated care and service on the supply side (“do everything for the patient we have with the resources available”) and “well, we’ve committed; and it’s not our money anyway” on the demand side that lead to cost overruns each and every year.
Given all of this, let’s wrap up the week with a question: what happens when software finally eats health care? Does software make this better, or does software make it worse? The reflexive answer is, “well, in industries like retail and transportation where software is running its course, productivity goes up and prices go down.” But the answer in these cost disease industries may be the exact opposite. For instance, we’ve all heard the rumblings about software and Artificial Intelligence automating and potentially replacing many medical functions. There are serious people who are suggesting we stop training radiologists, because they’ll soon find there are no positions available for them. But which do we really think is a more likely future: that we’ll employ fewer radiologists, or that we’ll do ten times as many scans? If what afflicts health care is truly a price problem, then software will eventually turn the Cost Disease tide and bring real, genuine cost-saving efficiencies to health care. But if it’s a volume problem, then we can see what’s going to happen: tech isn’t going to make things cheaper. It’s going to make it even easier for us to consume an order of magnitude more care, more services, more tests, and more scans. We’ll consider this possibility next week.
Elsewhere in the world:
China: a return to the ‘kingdom of bicycles’? | Guoqing Hu, British Medical Journal
European court suggests relaxing gene-editing rules | Alison Abbott, Nature
Africa’s art boom | Lynsey Chutel, Quartz Africa
The largest early world map is unveiled for the first time | Ahmed Kabil, The Long Now Foundation
Crypto, my friend, you are entering a world of pain (it’s a league game, Smokey):
Opening remarks at the securities regulation institute | Jay Clayton, SEC
Preston Byrne: Crypto-pocalypse | Invest Like the Best podcast
Momentum in energy that’s hard to stop:
Will utilities keep investing in solar after Trump’s tariffs? | Krysti Shallenberger, Utility Dive
Who says newspapers can’t make great online-native content?
40 days of devastation: the Thomas Fire interactive project | Joe Fox, LA Times
The Follower Factory | Nicholas Confessore, Gabriel J.X. Dance, Richard Harris & Mark Hansen, NYT
Senior oversight:
We need to talk about servant leadership | Jonathan Nightingale
Other reading from around the Internet:
Here’s a sweet recipe for cheap, green plastic: sugar and corncobs | Roni Dengler, Science
Uber, Lyft and the roads of hell | Om Malik
The playing field | Graham Duncan
Dude, you broke the future! | Charlie Stross
And just for fun:
In this week’s news and notes from the Social Capital family, we’ve got a few big announcements to make. First of all, the team at Autonomic is officially joining Ford:
Autonomic joins Ford to build the future of transportation | Sunny Madra, Autonomic
Ford expands its mobility empire by acquiring Autonomic and TransLoc | Andrew J Hawkins, The Verge
Since 2016, the team at Autonomic has been hard at work building the future of connected transportation in urban centers; from the very beginning, they’ve worked alongside Ford in order to make it happen. We saw a preview of their partnership a few weeks ago at CES, with Ford and Autonomic’s shared vision for smart cities centered around mobility as a focal point. So it’s no surprise to see them join forces to work on the Ford Transportation Mobility Cloud, and build on their initial partnership as a unified team moving forward.
From Sunny’s post, he has one more request for the whole community at large: if you’re solving a hard problem around cities and transportation, please join them! He writes, “We would like to take this opportunity to invite the entire transportation community to join us in shaping the [Ford] Transportation Mobility Cloud. We want to work with transportation stakeholders of all kinds — automakers, large-scale fleet operators, public transit operators and communities everywhere — to develop the mobility platform for the cities of tomorrow. Whether it’s to route self-driving cars, manage large-scale fleets or help residents plan transit journeys, join us as we continue to build the platform that will achieve mobility’s full potential for millions of people in cities around the world.”
Congratulations again to Sunny, Amar, Caleb, Ben, Nithin, Gavin and the entire Autonomic team!
Earlier this week, Chamath was interviewed by the legendary Christiane Amanpour, and it’s a must-watch:
And finally, we’re hiring here at Social Capital! We’re looking to hire an exceptional full stack engineer and a data engineer; if you’re one of those and (especially) if you read Snippets regularly, we’d love for you to apply. As you know, we believe the utilization of data at scale will fundamentally shift the nature of capital allocation in the future. Given the exploratory nature of the current space, you’ll do best if you get excited about discovering and delivering a wide range of unconventional and aggressive value generating products with a tight group of exceptional researchers and expert capital allocators. The specific types of challenges you will focus on will depend on your passion, technical skill set and cadence with our teams. So:
We’re looking for a full-stack engineer who can work with our product and research teams to dissect and solve difficult open-ended challenges by rapidly designing, prototyping, productizing and evolving solutions.
We’re also looking for a data engineer who will work with our data team and analysts to help create our data platform — the mechanism that force multiplies the ability of our researchers and partners by giving the access to an ever growing body of data and actionable insight.
As an added bonus, you’ll get to work for Mike Sela, one of the secret awesome people at Social Capital who does incredible work behind the scenes on our platform engineering team. (If I could pick anyone at Social Capital to work for on a daily basis, Mike would definitely be a candidate to make the Final Four… just saying.) Check out our other job openings at Social Capital while you’re at it, and please forward along to anyone you know you think might be a good fit!
Have a great week,
Alex & the team at Social Capital
