Wednesday, February 10, 2016

4 winners and 2 losers in New Hampshire

There were two stand-out, runaway favorites in the New Hampshire primary results: Ned Augenblick and Scott Nicholson. Their names weren't on the ballot per se, but their theory on ballot position and choice fatigue certainly was.
John Kasich was the only major candidate in the top half of the New Hampshire republican presidential primary ballot.
Augenblick and Nicholson found that voters would be more likely to vote for things that appear earlier in the ballot, and this result was very much confirmed in New Hampshire where the official list of republican presidential candidates was much longer than the national list. Kasich was the only nationally known candidate still in the race to appear in the top half of the list, and as the decision-fatigue theory predicted, he massively outperformed the polls and came in second over all.

Of course, Trump and Sanders also won their primaries, for whatever that's worth.

Clinton and Rubio were the biggest losers. Rubio was expected to surge into second place after his better-than-expected Iowa performance. New Hampshire at best interrupts his momentum, at worst confirms that Christie's attack in the last debate--probably the single most effective attack against a major candidate ever launched--was as destructive for Rubio as everyone thought. As for Clinton, for someone who not long ago was supposed to win all the primaries without much trouble, the tie in Iowa and loss in New Hampshire represents a pretty serious upset.

On paid sick days

The United States is one of the few developed countries that still does not have a sick leave program. Most countries have a publicly funded sickness program that pays workers not to go to work. Over 102 countries offer workers more than a month of sick leave, at varying levels of wage replacement. In the US, by contrast, sick leave programs are administered by employers, if they choose to have one at all, and typically only offer 5 days off.

Thus, there's a substantial movement in the US to make sick leave benefits universal by mandating that all employers provide at least 5 paid sick days off to all employees. There are lots of benefits to having sick days. For one, it's the humanitarian thing to do—when someone is violently ill, we should let them stay home and take care of themselves, or to go to the doctor if necessary. For another, having a sick person at work can, for example, increase risk of workplace accidents. There are lots of other benefits as well, but one that is most often touted is that letting sick peoples stay home will reduce the spread of diseases in the workplace.

Unfortunately, I think that last particular benefit is overstated—US-style sick leave policies have only minimal effect on the incidence of disease.

The problem is that the typical sick leave policy allows only 5 days off, and even then we're faced with strong social norms prohibiting workers from using them effectively to prevent contagion. Your employer and co-workers will suspect you of shirking if you call in sick but don't sound violently ill, and you're expected to only take one sick day—2 at the very most—for a particular illness.

Surveys show that for the flu, one of the more aggressive and dangerous of the common illnesses, workers with access to paid sick days take fewer than 1.8 days off on average. The problem is that the flu remains contagious for 5 days, and up to 2 weeks for some people, so taking only 1 or 2 days off from work does not substantially reduce your co-workers' exposure. Even taking all 5 days would not eliminate the risk of spreading flu at work. The result is that access to paid sick days only reduces workplace transmission by less than 6 percent.

Aside from food poisoning and hospital-aquired infections, most common contagious illnesses these days are viral and not amenable to treatment with antibiotics. Most of these viral illnesses have a pretty similar menu of symptoms, differing in severity, including fever, congestion, runny nose, sore throat, nausea, and diarrhea. But epidemiologically, they're all over the map. Some viruses become contagious before people develop any symptoms, others remain contagious after the symptoms disappear, and many viruses can be transmitted by people who never develop any symptoms. The notorious norovirus that infected Chipotle restaurants on two recent occasions, for example, remains contagious 3 days to 2 weeks after their symptoms go away, and 30 percent of people who have norovirus infections never develop any symptoms at all. Most of these viruses remain contagious for longer than the 1 or 2 days workers typically take off from work.

Having more sick days could help. In other countries with public sick leave programs, workers typically take 5 to 10 days off a year for sickness, much more than the 1.8 in the US. And some theoretical and empirical estimates in the US suggest that having access to 7 to 10 days off a year could be associated with a much larger reduction in workplace transmission, especially during larger epidemics. But even then, sickness leave policies will leave large gaps as workers come to work whenever they lack symptoms, even if they're still contagious.

Why is this so hard?

I think it helps to make a comparison to vaccination programs. Vaccines obviously aim to give individual-level immunity to the individuals who get it, but when administered to the entire population they also provide a second level of protection known as "herd immunity," where the transmission of the virus is impeded so much that it is unable to infect even those without individual-level immunities. Even with vaccines that are highly effective at producing individual-level immunity, you still have to vaccinate more than 90 percent of the population before you start to see significant herd immunity effects. Paid sick leave policies aim to create a herd immunity effect—impede the transmission of illnesses to the non-immune—without any of the individual-level immunity that makes this work for vaccines. If we really want to see a significant effect here, we'd need to dramatically ramp up the coverage of paid sick days. Not merely extending those 5 days off to the whole population, but also increasing the number of days and ensuring that workers take weeks, not days, off whenever they get sick. Without that dramatic step, I doubt most work places will see much effect on illnesses from their sick leave policies.

Tuesday, February 9, 2016

Most bets are meaningless

Bryan Caplan announces that he's won his bet with Tyler Cowen. Caplan bet that unemployment would recover, Cowen bet that it wouldn't:
...U.S. unemployment will never fall below 5% during the next twenty years. If the rate falls below 5% before September 1, 2033, he immediately owes me $10. Otherwise, I owe him $1 on September 1, 2033.
The unemployment rate in January dropped to 4.9 percent. Never the less, Tyler Cowen feels that he's the one who won the bet.

No one agrees with Cowen (sorry). Here's Paul Krugman and Noah Smith piling on. Nevertheless, Cowen's position is not totally unreasonable. The essence of why Caplan and Cowen disagree who won the bet is this trend in labor force participation:
The participation rate certainly doesn't, prima facie, appear to have recovered at all from the recession, and since this determines the denominator in the unemployment rate calculation, in Cowen's mind this means he was right overall but merely didn't anticipate that workers would actually drop out of the labor market.

But it's still not that simple. The peak year for baby boomers was 1947 and births remained elevated until 1960. 1947+65=2012, which means that we hit a big wall of retirees at the same time as the recovery from the great recession—labor force participation rate was always anticipated to decline over this period. No doubt some baby boomers retired early because of the economy in 2009—realistically, no one who retired early at 62 is going to rejoin the labor force now that they're 69. In this respect, recessions have a ratchet effect on labor force size that make the recoveries look slower than they really are.
I think that ratchet effect is most of what we're seeing in Cowen's graph here.

One way to adjudicate this is to look at pre-recession projections that estimated what would happen to the labor force as a result of demographic trends alone. These projections aren't perfect—relevant factors could have changed since then—but they are very clean in the sense of being totally immune to our current biases about the recession and recovery since these were not known at the time. BLS has such a projection dating from 2002: they estimated that aging of the baby boomers would cause labor force participation to fall to 64.6 percent by the end of 2015 and that further aging of the population would eventually reduce participation to 60.2 percent by 2045. The labor force participation rate is now 62.7 percent. That suggests that at least half of the fall in labor force participation was just normal population aging. Combined with the ratchet effect described above, early retirements could explain all of the fall.

What have we learned?

My real point though is not to explain that Cowen is wrong about the labor market but that Caplan and Alex Tabarrok are wrong about betting. Tabarrok declared that bets are—pardon my language—"a tax on bullcrap," and force people to reveal their true beliefs rather than cheap talk. And yet here we are with Cowen and Caplan both declaring themselves winners on the opposite sides of the same bet.

I wrote at the time that Tabarrok and Caplan were wrong to claim bets reveal beliefs. The problem, I argued, is beliefs are always necessarily conditional predictions. I believe the earth will warm by more than 2 degrees unless the governments of the world take dramatic action to curb greenhouse emissions, for example. A bet on the earth's warming does not reveal my beliefs about climate change, it reflects a complex combination of beliefs about climate change and geo-politics and lots of other things. My beliefs about what the economy will do (not fall in recession) are conditional on my unrelated beliefs about what the Fed will do (not raise interest rates too quickly), among lots of other things.

Oh, sure you could make conditional bets. Cowen could have bet that unemployment will not fall unless labor force participation falls. Highly conditional bets will usually not payoff either way though, so they don't actually tax incorrect claims. In this case, if Cowen had made such a bet, there'd be no payoff since both Caplan and Cowen's positions allow that unemployment and labor force participation falls. Moreover, listing out every single factor that weighs in your belief system simply isn't practical.

Ultimately, the problem of identifying beliefs through bets is exactly the same problem as identifying causality through econometrics. Econometric models, after all, do nothing more than generate predictions about the data, which we then try to use to form beliefs about the data-generating process. And that's really hard and doesn't work out a majority of the time.

Sunday, January 31, 2016

Why we need insurance networks

Here's one of the questions I often encounter when discussing health insurance markets and single payer: Why do we need insurance networks?

It's an understandable frustration. You get sick, find a doctor you like and BAM! your insurer says it's out of network and won't pay. Narrow network plans can be particularly problematic for a variety of reasons. I've seen plenty of cases where children with rare diseases were unable to get coverage for their doctor because the only specialist in the region that handles their condition is, of course, out of network. And there's a never-ending parade of stories in the media where individuals went to an in-network hospital with an in-network doctor only to find an out-of-network bill from an un-requested "drive-by doctor" while they were unconscious under anesthesia. Wouldn't it be nice if insurers just paid all health providers like in-network providers?

Unfortunately, that just doesn't work.

To understand why, it helps to be specific about the policy we are considering. To abolish out-of-network charges, we'd need a regulation that prohibits insurers from refusing to pay for out-of-network care. Now imagine that you are a healthcare provider in this environment. Insurers are obligated to pay your bills regardless of whether they've negotiated any prices with you. What stops you from raising your prices? Patients don't pay the prices, so they have no incentive not to continue to come to you for care, and insurers legally can't prevent their enrollees from coming to you for care, because of the new regulation. As a result, the new higher price has zero impact on demand for your service. Why not raise prices further? Why not charge infinity. Legally, insurers are obligated to pay you infinity.

Ok, actually healthcare providers wouldn't charge infinity because insured patients do still have to pay a portion of their prices. For a typical private insurance plan, patients pay 10 to 15 percent of total costs, though after the deductible the share is typically smaller than that. Of course, the same single-payer advocates who want to abolish insurance networks also want to abolish co-pays, co-insurance, and deductibles. But even if we keep them, with patients on the hook for only 10 percent of providers' price hikes, providers would be able to hike prices quite substantially before they'd start losing enough patients to harm profits. This in turn means insurers would have to hike premiums quite substantially, and most people would no longer be able to afford any insurance.

Insurance networks are the end result of negotiations between insurers and healthcare providers on your behalf. Insurers insulate you from the full price of healthcare; in exchange, they ask you to only use healthcare providers who have agreed to charge them reasonable prices. In the absence of this bargaining step, providers would have little incentive to charge reasonable prices, and the insurance plan would collapse.

Medicare-for-all, or single-payer, does not solve this problem. In the US, Medicare and Medicaid do have de facto insurance networks, even though they work slightly differently. Instead of the one-on-one negotiations with providers, Medicare instead promulgates a list of maximum prices it will pay for things, and the network forms around these rates as individual health providers decide whether or not to accept Medicare patients. It is then up to Medicare patients to make sure they see only Medicare-accepting doctors. And medicaid is way more complicated than that, since most states run tons of different medicaid plans that each pay different rates for different kinds of care—I often see cases where kids have to be dropped from one medicaid plan and moved to another medicaid plan as their medical condition changes overnight in the hospital.

I definitely think the way we regulate insurance networks could be vastly improved, and the financial liability for non-consensual out-of-network care should never fall on patients (as in case where out-of-network doctor drops in while patient is unconscious). But still, unless we commit to a system where the government directly dictates the prices providers are allowed to charge—with all of the waiting lines and misallocation of resources that government price-setting typically causes—insurance networks will remain a necessary part of the health insurance system.

Wednesday, January 20, 2016

Bernie Sanders doesn't seem to know what Medicare is

In my last post I analyzed whether Bernie Sanders's claim that medicare-for-all would save the average family $5,000 a year was realistic, and ended up concluding that it was at least plausible, and that some amount of savings is likely, due to the lower administrative costs and in particular the lower rates Medicare pays to healthcare providers.

Ezra Klein has some tough criticism of Sanders's plan though. Here's the part that stands out:
Sanders calls his plan "Medicare for all." But it actually has nothing to do with Medicare. He's not simply expanding Medicare coverage to the broader population — he makes that clear when he says his plan means "no more copays, no more deductibles"; Medicare includes copays and deductibles. The list of what Sanders's plan would cover far exceeds what Medicare offers, suggesting, more or less, that pretty much everything will be covered, under all circumstances.
Klein concludes that Sanders has done a bait-and-switch, calling his plan a simple expansion of Medicare, while actually proposing something that works very differently than Medicare. I think the truth is worse than that.

Sanders commits the classic liberal fallacy of assuming that Medicare is great health insurance. He doesn't seem to realize that Medicare has copays, co-insurance, and deductibles, that it doesn't cover everything or every provider, that it still denies a portion of all claims.

Medicare is actually fairly crappy insurance. Here's a break down: Medicare Part A is insurance for hospital in-patient stays, Part B is insurance for "outpatient" care (which, confusingly, includes a lot of in-hospital care, even short stays and many kinds of surgery), while Part D covers prescription drugs. Part C is Medicare Advantage, which is an alternative to Parts A and B.

On deductibles, Medicare is pretty reasonable: $1,288 deductible for Part A and merely $166 for Part B. It's everything after the deductible that's crappy. For Part A, after the deductible your out-of-pocket costs actually grow as the severity of your condition worsens: after 60 days of hospitalization you pay $322 per day, and after 90 days you pay full cost (there are an additional 90 "lifetime reserve days" you can apply there though). Part B is actually worse: you pay 20 percent of all costs after the deductible.

This made some amount of sense when Medicare was designed. Back then doctors typically didn't even bother charging patients for outpatient care, because almost everything was done inpatient and outpatient stuff was just insignificant. But because of technological improvements, healthcare overall is shifting from inpatient to outpatient. Currently, Medicare spends about as much on outpatient care as inpatient, and that ratio is growing—CBO projects that by the end of this decade, outpatient care will exceed inpatient care. My point here is that while you could argue that hospitalizations longer than 90 days are very rare, that 20 percent on the other half of your health costs is extremely expensive. Seniors can easily find themselves owing $20,000 on a $100,000 outpatient surgery. Considering that the average social security benefit is only $14,169.60 per year, I don't know many seniors who can bear that kind of medical bill.

The simple truth is that Medicare is no longer sufficient insurance because it does not cap seniors' outpatient expenses. The program has totally failed to modernize in the face of dramatic changes to healthcare delivery. As a result, the vast majority of seniors—86 percent to be exact—purchase additional insurance to cover the rest of what Medicare won't pay. Some seniors qualify for both medicare and medicaid, so medicaid does actually cover these additional costs. But the rest, a solid majority, purchase private insurance one way or another. This includes Medicare Advantage (Part C), which is a private alternative to Medicare and generally has much better coverage, as well as employer-sponsored insurance (the most common option, typically offered as a retirement benefit), or "medigap" policies which are simply private supplemental insurance coordinated through federal exchanges.

Sanders undoubtedly has supplemental insurance as it is an employment benefit he would have received as a result of his years in the US Senate. He doesn't seem to realize that that's not medicare, and that he in fact relies on private insurers for much of his medical coverage.