Freakonomics Read My Post!

Or maybe they didn’t and my thoughts aren’t as unique as I thought they would be.

I did learn a couple of new things here; I didn’t know that the AMA was fighting so hard to exclude nurse practitioners from independent practice (and presumably other highly trained but under-utilized medical professionals including physician’s assistants, pharmacists, and physical and occupational therapists). I also didn’t know much about retail health clinics though in retrospect I’ve seen the signs for services from CVS and Rite Aid everywhere.

Nurses to the Rescue | Freakonomics

A Slight Rant about the College Football Playoff

I’ve gotten into it from time to time with other college football fans about the nature of championships in Division I-A (ahem, Bowl Championship Subdivision). The basic controversy is this: since teams are assigned to the four playoff berths by a committee, how should they evaluate the teams? Is it the best teams? Or the most deserving teams? To put it more concretely, should it be the teams who have the best chance of beating any other team at the end of the season? Or should it be the team that has put together the best resume over the course of the season?

Of course, the two characteristics are not mutually exclusive; Alabama this year is probably the best team, and has (so far) put together an undefeated record. Accordingly, they are ranked #1. However, they are about to play a team (Auburn) who, lately, have shown they are a threat to beat anybody at this point in the season, but earlier this year lost two games. Even after this game is played and there is a winner and a loser, who is to say the result didn’t hinge largely on luck? The committee may still find a way to put Alabama in the playoff even if they lose this game. And there are many other cases like this to consider.

Meanwhile, University of Central Florida is undefeated this season so far. If they continue, they will be their conference champions. However, for a combination of reasons they will never be considered for the playoff because

  1. They have no football pedigree or brand. Unlike Alabama or Ohio State, or even Pitt, they have no history of football success and have only recently begun playing at this level.
  2. Their opponents have been weaker so the team hasn’t yet been tested to the limits of their ability; however, a school of rising strength will consistently struggle to schedule stronger opponents because the stronger opponents don’t have any incentive to play them.
  3. Their opponents similarly have no pedigree, and thus Central Florida gets less credit for beating University of South Florida than they would for beating Tennessee, though Tennessee is undoubtedly weaker.
  4. When it gets right down to it, the College Football Playoff is about money, and Central Florida will never draw as much money as the blue-chip programs at Alabama, Texas, Notre Dame, and so on.

This is a fundamentally flawed system, because it relies on a “beauty contest”–a systematic beauty contest, but a beauty contest nonetheless. In other words, a team has to rely on the opinions of a handful of people to even get into the game–people who carry into the room their own preconceived notions and biases. This is anathema to American sports–a team can do everything right and still get left out.

I propose a simple system for an eight-team playoff that a) ensures that non-legacy teams (e.g. Utah, Central Florida, Boise State) have an avenue to compete for the championship, b) still allows for an advantage based on strength of schedule, and c) makes conference championships matter. Here is how it goes:

  • There are eight playoff teams.
  • Five of the teams will be the champions of the SEC, ACC, Big 10, Big 12, and Pac-12, however the individual conferences decide to determine their champions.
  • One of the teams will be the highest ranked undefeated conference champion from the group of all other conference (American, Mountain West, etc.) If no teams meet this criteria, this will be an at-large team (see next bullet).
  • The other two teams will be selected by the committee at-large based on the committee’s assessment of the teams’ strength/deservedness.
  • Finally, the seeding will be selected by the committee based on the committee’s assessment of team strength as well as considerations of matchups and bowl venues.

This year, what that might look like, assuming conference championships are won by the current frontrunners, would be (1) Alabama vs. (8) Central Florida, (2) Clemson vs. (7) Stanford, (3) Oklahoma vs. (6) Wisconsin, (4) Georgia vs. (5) Ohio State.

In this scenario, Alabama gets an “easy” first round game as reward for a great regular season and conference championship; UCF and Wisconsin get a shot at the title despite the apparent weakness in their schedules; highly talented teams that may have just had a hiccup get a second chance; and most importantly we get 6 or 7 high-stakes marquee matchups that college football fans would drool over to end the season.

Riddler, 9 November 2017

For once, it seems like the Riddler Express is actually more challenging than the main Riddler puzzle:

The regular hexagon below has an area of 1. What is the area of the shaded region?

If you draw lines from every vertex of the hexagon to every other, you can see that these lines form another, smaller regular hexagon which is composed of 6 of the shaded regions from above. This smaller hexagon has a minor diameter equal to the length of one of the larger hexagon’s sides.

The area of a regular hexagon with respect to one of its sides is

A = \frac{3 \sqrt{3}}{2} s^2

For a unit area hexagon, the side length is thus

s = \sqrt{\frac{2}{3 \sqrt{3}}}

And the minor radius is half that amount. The area of a hexagon with respect to its minor radius is

A_{small} = 2 \sqrt{3} r^2 = 2 \sqrt{3} \left(\frac{1}{4} \frac{2}{3 \sqrt{3}}\right) = \frac{1}{3}

And the area of the shaded area is one sixth of that, or 1/18.

On to the big Riddler:

The largest circle below has a radius of 10, the medium circle has a radius of 5 and the small, orange circle has a radius of 2. The orange circle crawls counterclockwise along the edge of the largest circle until it meets the medium circle, at which point it crawls up along the edge of the medium circle until it reaches the crest. What is the area of the shaded orange region in the right image?

This is pretty simple, if you notice that the shaded area can be split into two regions: the large radius and the small radius. Each region’s area is equal to the difference of two circles. Finally, it’s unclear whether the total area includes the two orange circles, or not, or if the area includes half of each circle.

A = \frac{\pi (12^2 - 8^2)}{2} + \frac{\pi (7^2 - 3^2)}{2} = 60 \pi

If we include the area of the small circles, we add 4 \pi, and if we don’t, we subtract it.

Why the American Health Care System is a Mess: Perspective from an Ordinary Schlub

When I was an undergraduate, USMA required an introductory course in microeconomics. I can’t say I remember all that much; yadda yadda supply and demand curves, marginal cost vs. marginal revenue, yadda yadda. But one thing I do remember this idea of an ideal market, where it was proven that resources are allocated optimally; I also remember that no real markets are ideal, but the general consensus is that small deviations from ideal result in small inefficiencies, and large deviations result in large inefficiencies.

This whole set of ideas rushed back to mind when I saw a video from Vox about how much having a baby costs (side note: we just had a baby, too!), and more importantly, how hard it is to find out how much it costs before you get the final bill. It also reminded me of health care coverage for me and my family while I was a graduate student; we had only one choice for insurance with a premium that cost more per year than our rent, with a significant deductible, and an insurance company that could not be reached by phone even after they mistakenly drained my bank account the day before our rent was due. Yet, even with all those problems, any other path would have been even more exorbitant and out of reach for a graduate student.

The Vox video is embedded at the bottom of this post, but I want to go a little bit beyond. I don’t know what the exact right policy changes are, the below named problems of the health insurance market that can be seen by a consumer in this market need to be fixed before we start getting rid of the large market inefficiencies. To that end, I’ll suggest a few general policy ideas.

Because I only focused enough in Econ 2001 to do well in class but not enough to remember everything, I leaned on Wikipedia to remind me all the pieces of a perfect market. Briefly, I’ll the parts of the health care/insurance market that SEEM aligned with a perfect market before pointing out the non-ideal aspects.

How the health care market is ideal:

  • A large number of buyers and sellers.
  • Every participant is a price taker (no participant has market power to set prices). You might argue this one, but I would claim that even the biggest of Big Pharma or biggest insurance company doesn’t have anything near monopoly power.
  • Perfect factor mobility, i.e. Doctors and medical equipment/facilities can move to where the demand for care is.
  • Non-increasing returns to scale and no network effects. In the health care context, there’s a limit to the number of patients a single doctor or other provider can see.
  • Homogeneous products. This is iffy, I guess, but fundamentally doctors provide care that is far more similar to each other than, say, cars or furniture.
  • Well defined property rights that determine what may be sold, as well as what rights are conferred on the buyer.

And how it is not:

  • Perfect information: All consumers and producers know all prices of products and utilities each person would get from owning each product. This is first and most obvious on the list, and the one that prompted the Vox video. It’s not just having a baby; I challenge you to find out how much an appendectomy, or even a simple exam or prescription costs without actually having to buy it. You can’t! And how can a consumer shop for the right price if you can’t find it out ahead of time. What’s worse is that even if the hospital could tell you the price, they’re probably not able to tell you the price the insurance company will actually pay for it, nor your responsibility in that bill.
  • No barriers to entry or exit. The American health care system has huge barriers to entry and exit: doctors and other providers pay huge sums for education and licensure. Health care consumers, since they typically rely on employer-based insurance coverage, face the obstacle of having to have a good enough job to get decent coverage. A critical and currently insurmountable barrier is that insurers are not legally allowed to sell insurance outside the borders of the state they are registered in. While there are national health corporations (Blue Cross/Blue Shield and Kaiser Permanente being two examples), if you read the fine print, there are individual corporations for each state, and you can only buy the coverage for the state where you live.
  • Zero transaction costs. Any provider, whether they see 1 patient a year or 10,000, pays malpractice insurance; the multi-million dollar claims on this insurance make the premiums borderline exorbitant.
  • Profit maximization of sellers. Many (maybe most?) health providers are non-profits to begin with; even if you asked doctors in for-profit hospitals whether they were trying to maximize profit, I’m sure they would blanch.
  • Rational buyers: Buyers make all trades that increase their economic utility and make no trades that do not increase their utility. This element could go in either category; patients by and large can make good decisions about the health care they consume if presented with the right incentives, but I also know how irrational people can be when they are presented with certain circumstances.
  • No externalities—Costs or benefits of an activity do not affect third parties. This criteria also excludes any government intervention. Obviously based on most of the previous points I’ve mentioned, there are externalities in this market.

These deficiencies naturally point to a handful of policies that would actually help:

  • Require providers to make an itemized cost list available to patients BEFORE they select a provider or request care; and insurers must make it easy to see any deductible or cost share for this list. This would allow people to shop for care and make informed decisions. Require providers to charge those who pay cash for medical services be charged at the lowest rate negotiated by insurance companies, so that the price charged truly reflects the price of service.
  • Create “Yelp for doctors” that makes it easy for patients to register feedback and other patients to see that feedback, as well as some kind of quality rating of various doctors. Granted, measuring doctor quality is an inexact science at best, but there IS useful information to be had.
  • Allow patients to own their medical data and easily port it from provider to provider, as well as read and reference it themselves. This allows people to, if they choose, make themselves better informed to make the right decisions.
  • Everybody pays something for care, maybe means-adjusted. Even a nominal fee of a few dollars would provide incentive for people to consider whether it’s worthwhile to consume healthcare.
  • Give patients the option to cap potential malpractice claims in exchange for reduced cost.
  • Allow national, or at least regional competition of healthcare providers.
  • Tax healthcare benefits to untie health insurance. While I understand that health insurance is not required to be employer based, it almost always is so, because of the tax benefits, which makes the non-employer-based market tiny. New taxes are never popular, so this could be possibly offset by relief in marginal rates or a higher standard deduction, for example.

Finally, and I realize these kind of go against my free-market tendencies, but

  • emergency medical services should be fully government-funded, if not government-run. I say this because services to save life, limb, or eyesight are time-critical, cannot be “shopped around” and a benefit afforded to all people through a basic moral obligation. This is a life-saving service, same as the police or firefighters, and the immense cost of an ambulance and ER stay needs to be out of the equation when somebody needs to call 911; to keep a cap on people abusing the free service, people who frivolously call an ambulance could be subject to the justice system, same as those who call in fake crimes or fires. Perhaps this government service could be a uniformed service, with those who serve accorded the honors given to policemen and firefighters and soldiers.
  • basic pre-natal care, childbirth, and pediatric care to say, age 12, should be free. I don’t think that there is a right to health care for kids, but a) we should minimize the damage done to a person’s whole life by their parents’ inability to provide for them, and b) the investment in childhood wellness is a productivity boost and cost savings as those people are healthier as they age.