Behavioral economics is an interesting and potentially useful field of academic inquiry. At its best, real behavioral economics attempts to address some of the concerns I raise here. But many if not most assumptions about human behavior and response to incentives are not representative of behavioral economics at its best.
Specifically, I’m increasingly concerned with what I see as the simple-minded projection of economic thinking onto everyone and anyone else, leading to ridiculous policy recommendations – that amazingly – get taken seriously – at least by the media and punditocracy.
See, for example, Roland Fryer’s experiment on loss aversion as a strategy for incenting teachers to make sure that their students gain a few extra test score points in limited content areas. Indeed, if we pay you up front, and threaten to take your salary away if you don’t get those test score points out those kids, the data suggest a greater likelihood of squeezing the kids for a few more points. Whether that tells us anything about the motivation and morals of teachers, or of the economists framing this argument is an entirely different question. This tells us little or nothing of the appropriate policy response. Thankfully, the policy implications of this paper were sufficiently absurd that they gained little traction.
Let’s assume classic economic assumptions about human behavior really hold steadfast and can be grossly simplified to an anything for an extra buck, or not to lose one, position. I would argue that it is perhaps economists themselves that are most stereotypical in this regard - at least as represented in the thinking the project onto others. In fact, I would argue that many, born out of a culture that self-selects into economic professions, are simply going out of their way to project their own thinking on others.
Further, many of these economists operate in a world where they can influence/control public policy and they too have an incentive in how they behave in this system. They are not impartial observers by any stretch of the imagination. Their goal is to use their economic research to shape public policy to their own advantage.
Put simply, just because the average morally bankrupt economist might do pretty much anything for an extra buck (or a billion), doesn’t mean the average teacher, doctor, nurse, fireman or police officer would!
This issue has been on my mind for some time, but recently came to a head when I read this completely ridiculous Washington Post article on health care policy – specifically – how to remove the incentive for hospitals and physicians resulting from surgical complications.
I should note, I come from a medical family, so some of my arguments herein are drawn from dinner table conversations (across generations), coupled with my tendency to read health policy research out of personal interest in exploring connections with education policy.
It was implied in the WaPo article… well… actually it was explicitly stated in the article that hospitals and physicians have a big financial incentive for their patients to have serious complications, leading to extended hospital stays and additional procedures.
Now, the average economist might be so morally bankrupt such that if he/she were in an operating room (OR) considering the implications of complications relative to potential earnings that they might intentionally introduce infection or other complication, but thankfully the average economist is not in the OR. Thankfully, they self-selected into economics and not medicine (likely foreseeing greater opportunity to earn more for much less work and upfront investment).
The WaPo article does make the following statement, to head off this argument:
The study does not imply that hospitals intentionally complicate surgeries to bring in more revenue.
But, I would argue that this is actually a rather half-hearted disclaimer (to a half-assed argument) for an article that very much implies just that.
Certainly the economists’ policy response – how to employ crude economic assumptions of human behavior to fix this dreadful perverse incentive – implies that cutting off this financial benefit for malpractice would improve hospital and physician behavior [meanwhile conflating the hospital and physician incentives & roles in the various related processes]. Here is the policy solution recommended by the economists cited in the WaPo article:
If hospitals receive a set amount for every heart surgery they perform, for example, they suddenly have an incentive to reduce complications — they know the extra medical spending will come out of their own budget.
Lost in the economists’ reasoning here are a) the potential longer term financial and career implications to the physician repeatedly entangled in litigation over post-surgical complications, b) and the stress/mental toll on the physician arising from managing complications in tense moments in the OR.
Indeed this is anecdotal, but I’ve not met a physician – surgeon or anesthesiologist – who prefers a day when things go bad in the OR – or would be likely to see dollar signs in those moments of stress. What kind of sick bastard even thinks that way? Well, perhaps the average economist does.
Economists rarely – uh… NEVER face comparable professional stress to managing a patient’s life on the edge – even when they make a massively stupid spreadsheet error stimulating economic turmoil across the globe. Nor do they pay hefty malpractice premiums to shield themselves from such egregious malpractice (despite measurable financial damages). I would assert that the economist never faces the stress of having to care for a classroom of 20 to 40, 5 to 15 year old kids, whose immediate safety and well-being, as well as their long term futures is on the line. This is in part, why they get away with such ludicrous thinking.
It’s all freakin’ game (Freakin’ used here in a technical freakonomics sense)… a game of playing with big data – several layers removed from reality – from people – from real human consequences.
Perhaps that’s the central issue… even more so than economists’ financial self-interests?
Taken in perspective, it’s a fun game and a pretty cushy lifestyle to have opportunity to ponder policy implications of big data, as long as we don’t start thinking that what we do is so freakin’ important and indispensable and as long as we understand where we sit in this big messy puzzle of human behavior and incentives.
Of course, the other interesting piece here is the leap we often see these days between what the study behind the headlines actually said, and the resulting spin in the media headlines. We also often see the economists themselves engaging in the spin. This was equally true in the famed Chetty, Rockoff, Friedman Fire Teachers First, Ask Questions Later study.
For example, here’s what the original study – in the Journal of the American Medical Association – on reimbursements associated with complications actually said:
Depending on payer mix, many hospitals have the potential for adverse near-term financial consequences for decreasing postsurgical complications.
It takes one hell of a leap of logic to get from this measured finding to the policy recommendation above. It takes projecting economists thinking –amoral greed – onto all actors involved. It also takes ignoring entirely a multitude of contextual factors and perverse consequences (economist thinking – first, we assume none of that stuff exists). Indeed, many complications relate to preexisting conditions and/or overall health of the incoming patient. Do we really want to incent risk aversion? (avoiding those far more likely to have complications?). Well, if it leads to lower premiums for and taxes paid by economists, then perhaps?
Tangentially (or not?), there is an equally ill-conceived movement afoot to apply to healthcare management the brilliance of what we have supposedly learned from measuring teacher effectiveness with value-added models, as explained in this policy brief from Mathematica. Notably, I tend to think Mathematica does pretty good work on education policy (better than most. See here, but for more critical perspective, see here) Put in its best light, this policy brief is merely Mathematica researchers engaging in another I’ve got a Hammer… where’s the freakin’ nail exercise.
Put in the light of economic thinking about human behavior – which many economists prefer to project on all others, the incentive here is for Mathematica to broaden its market, gaining contracts to develop value-added metrics for health care systems and for state and Federal government – to ultimately be used in reducing payments for healthcare, and reducing the tax burden and healthcare premiums paid by Mathematica researchers – their funders and their peers. It’s a win/win. More contracts and higher income, and lower taxes and health benefits expenses (not costs, but expenses*).
That is, as long as they are never in need of surgery.
*Cost reduction implies that quality of service remains constant, whereas expenditure reduction may lead to service quality reduction.