It’s a new day, with new data. Ohio’s Vax-A-Million lottery is up and running. We’re not just hand-waving anymore. We have one concrete example analyze, attack, defend and revise.
Let’s not get wowed by the W-O-W numbers, but Week-Over-Week vaccination statistics look good, and a million Ohioans signed up before 48 hours were out.
NY and MD have caught the bug, with lotteries that look less well thought out. No doubt, there are discussions behind the scenes in other statehouses. Ohio’s shot in the dark wasn’t perfect, but could be a breakthrough, while BAD1 imitations elsewhere could give the whole thing a bad odor.
Time to look our gifting horse in the mouth. What conformational frailties should a willing sponsor look for before setting one of these creatures loose on the track?
Begin with two objections we can write off as obsolete. “Enough people will get vaccinated anyway” and “the COVID threat might not be serious.” Both so 2020.
But passionate objections to incentive programs — lotteries in particular, and health care rewards in particular — have deep roots.
These roots give rise to the majority of prominent objections of the sort that readily gives “ethics” a bad name — arguments that are strictly PUC2, i.e., they bear no relation whatsoever to the problem context or the proposed solution. Let’s flush a bunch of them out, and reserve our critical facilities to digest more substantive concerns.
“Incentives are not the whole solution.” Nobody, anywhere, at any time in my experience ever remotely suggested they were. If they work, most tools of influence are complementary, not mutually exclusive.
“The focus should be on messaging." The focus should be on everything that works.
“It doesn’t address the underlying issues.” Of course it doesn’t. It circumvents underlying issues for some, and gives others cause to re-examine theirs to see if any are showstoppers. There’s no silver bullet for “the underlying issues”. If that’s a precondition to action, nothing is ever worth doing.
“The most hesitant fraction will never respond.” See preceding point.
“COVID is serious. Lotteries are frivolous — just a game”. COVID is serious. Doing what works is not frivolous. Blissful ignorance is frivolous. Focus: Find out if they work, and how they work best.
“Lotteries are a waste of (my) money.” $1M sounds big. We want it to sound big. Now see it in context. Ohio spends $5M in prize money, and maybe another $3M in expenses. $8M out of about $16,000M in federal COVID relief for Ohio state/local government alone. If it doesn’t work, at worst it’s not a big waste. It’s a de minimis3 waste, and worth the gamble to find out4.
“It’s wasteful if the winner would have gotten vaccinated anyway.” It doesn’t motivate winners alone. It motivates everybody with any hope of winning.
“But the opportunity cost!5” What else could you have done for less than a buck a head? Frankly, not that much. More PSA’s? Ohio is several huge media markets. Pop-up vaccination clinics? Ohio has a lot of street corners. Those measures might have to come out of the remaining $15,992M.
“Rewarding new recipients cheats the early birds.” Ohio makes all vaccinated subjects eligible — as does our proposal in Part I. The sooner they get vaccinated, the more winning chances they get.
“We can’t afford to pay everyone”. That’s the key to incentive lotteries — we hardly ever pay anyone! In the carrots-and-sticks metaphor, it’s a sustainable carrot. Sponsors don’t run out of prize money, and participants don’t get their prospects doused by receiving rewards.
“It rewards public intransigence”. They’ll hold out for another lottery when the next pandemic arrives? Is that even true? If true, is that a problem? If it’s a problem, don’t we have a ready answer?
“Lotteries are morally suspect.” In many cases I do agree, even when part-proceeds go to worthy public purposes. State lotteries tax desperation, raking money out of poor suckers who can’t afford it. This, conversely, is win-win. All who get vaccinated are winners, and the public wins collectively.
“Gambling is addictive.” Indeed it can be, and it’s not a pretty picture, but we do not create the addictive hazards of active repetition, or chum players along with smaller prizes and “so close” number combinations, or create enough false hope to undermine anyone’s motivation to get up and go to work in the morning.
“Incentives take advantage of poor and minorities … apply excess pressure … deny them agency … paternalistic … exploit the disadvantaged … accentuate racial disparities … etc etc etc”. On one hand, monetary incentives arguably exert greater effect on lower-income individuals6. On the other hand, FOR FUCK’S SAKE, PEOPLE, WHAT ON EARTH ARE YOU POSSIBLY THINKING, WHINGING ABOUT ‘AGENCY’ WHILE COVID KILLS THEM OUTRIGHT … AND IN LOPSIDED NUMBERS??? Poor, disadvantaged and BIPOC communities are dying in droves out on your untainted moral high ground.
“Incentives rob the act of moral significance.” FFS again — now we’re on trial for robbery??? (I never realized … is that why I wear a mask?) If I texted a friend to let him know a nearby clinic had appointments open, did I cheat him out of “moral significance” too???
Maybe it’s time for a little break, and a digression on occupational perspectives.
We offer an exchange of cumulatively small individual material gains for a major public good. A good exchange, even if the “herd immunity” game is irretrievably lost.7
In general, economists get paid to overweight material self-interest in the dynamics of human behavior. They look for twists and wrinkles worth writing about, but nothing they can’t restate and restate til it looks like material store-keeping and score-keeping.
Moral philosophers get paid to overweight anything other than material self-interest. That stock in trade doesn’t leave ethicists enough scope for sermons or thesis topics.
Throughout the 20th century, motivational psychologists over-invested in models inspired by 19th-century physics — treating us (and our lab rat cousins) as idealized objects responding predictably to measurable forces.8
The simple truth is, none of it is that simple. Material rewards aren’t everything, or nothing, and neither is intrinsic motivation.9
We step into any discussion of incentives from our personal comfort zones in a broad range of initial orientations. Near the infrared end of this spectrum, “hard-headed” business types might insist everything of value is bought one way or another. Near the ultraviolet end, Alfie Kohn surveys a world of paradox in his valuable but tendentious “Punished by Rewards: The Trouble with Gold Stars”. On the fringe of the visible range, Noam Chomsky regards paying for work or working for pay as profound betrayals of fundamental human values.
In actual practical cases, everybody learns that well-intended incentive schemes can go awry — disturbingly, even hilariously.10
Here are a few pragmatic issues we should take seriously.
Adverse signaling. Compensation raises the notion there’s a downside that must be compensated for. Is it inconvenience? Unpleasantness? Risk of bodily harm?
There will be an effect. The magnitude is an empirical question. The amount of promotional ammunition expended on all sides of vaccination issues suggest this effect won’t mean much in the overall balance.
Calibration and Extinguishment. Initial positive response can fade as once-novel rewards become routine, expected, and fade into the status quo background. Subjects may also become satiated by rewards already received, making them less responsive to the future prospects.
Lottery incentive rewards never become routine, and only the few winners are at risk of satiation.
We might expect motivation to decline as hopeful participants are repeatedly disappointed. State lottery experience suggests few players give up hope — and in this application, our players are already vaccinated anyway.
Crowding-out. The theory holds that external rewards displace intrinsic motivators (such as concern for one’s own health … or the “moral significance” we allegedly stole earlier).11
For those who credit crowding-out theory in general, one health-specific review of relevant literature suggests “no evidence for crowding out of incentivized health behaviors”.
Conspiracy theories. Everything is grist for the conspiracy mill. This is no exception, and may get more attention as it involves government, money, and decisions behind closed doors.
In Part I we suggested doing it without government funds. It probably shouldn’t use Gates Foundation money either, but no source — or use — is above suspicion. Public-relations treatment of this challenge is above my pay grade, except to suggest professionals in opinion research and PR should be engaged in support of any such project.
What else can go wrong? I can see at least one potentially difficult line of attack, which I’ll hold in reserve for discussion.
Footnotes include a selection of references to eminent objections — earnest scholarship, but largely of the PUC variety. 12 We’ll open a discussion thread soon, so those kicked herein can properly kick back.
BAD: Broken As Designed.
PUC: Pure Unadulterated Crap.
"de minimis”, short for de minimis non curat lex — “the law does not concern itself with trifles”. Legalese for “the game’s not worth the candle” — which originates as gambler’s shorthand for “whatever I win here won’t even pay the light bill for the time it takes to win it”.
If it does work, we’ve spent 0.008 billion dollars to learn that $1B (a prize every day for a year, nationwide, all administrative expenses included) can make a major dent in a $10,000B problem. Not bad odds.
“Opportunity cost” is econ-speak for “the kids could’ve had nice new sneakers if you hadn’t bought those nice new speakers”.
I’m not perfectly convinced my friends in the streets want the next $1M more than my friends in the suites — who got there in part by wanting each million more than the next guy.
Herd immunity may be out of reach. When the percentage vaccinated (plus those immune as a result of surviving COVID) reaches the (unknown) threshold of herd immunity, each infected individual in turn infects on average “only” 1.0 other individuals. That doesn’t stop the pandemic at all.
Say we immunize more and knock the re-transmission rate down to 0.9. The next “generation” produces 0.81 cases, then 0.73, then 0.66, and 0.59, 0.53, 0.48 … many turns before the contagion dies out. [All that’s under the vastly over-simplified non-network, non-Markov model of uniform re-transmission.]
If herd immunity is out of reach, almost all of us either get immunized or get infected.
My eyes were opened long ago by a bright middle-schooler’s science project. Mouse #1 navigated a maze in 33 seconds, say, where mouse #2 took 27 seconds. When she placed both mice in the starting blocks together, they hit the cheese in 4-5 seconds.
You can schedule a motivational speaker to share the simple answer that pumps up your team’s intrinsic motivation. Next quarter you schedule another speaker with another simple answer. Do you know why speakers share the answers with you? You pay them.
Consider the exec who gratefully rewarded his company for an exceptional year by trucking in frozen turkeys for the holidays. Next year — a very ordinary year in corporate results — the rank and file grew restless as holidays approached, nobody had a bird in the freezer, supermarket stocks were getting picked over, and no truck showed up in the parking lot.
I hold “crowding out” as a suspect explanatory hack for otherwise genuine observations — in the mode of phlogiston theory or Ptolemaic astronomy — and thought it jumped the shark when advocates coined “crowding in” to account for contrary cases.
https://jamanetwork.com/journals/jama/fullarticle/2775005
https://www.aamc.org/news-insights/dollars-doughnuts-will-incentives-motivate-covid-19-vaccination
https://jme.bmj.com/content/early/2021/04/01/medethics-2021-107235 (abstract page includes a nice set of recommended readings)