Read on Twitter
COVID was a preventable disaster. Many ideas existed that could have helped that were not tried because no one had an incentive to test them. The government can provide incentives to develop innovations that make people healthier while letting the market discover what works.
Specifically, consider the following financing model:
1) Private firms propose an innovation they think will work
2) Private firms finance testing by a third party
3) The govt. contracts to scale up the program nationwide and pay the firm *only if it works*
1) Private firms propose an innovation they think will work
2) Private firms finance testing by a third party
3) The govt. contracts to scale up the program nationwide and pay the firm *only if it works*
This is the "social impact bond" model that Jeff Liebman has long championed ( https://www.americanprogress.org/issues/general/reports/2011/02/09/9050/social-impact-bonds/)-- it is desperately needed in healthcare and epidemiology.
This is also similar in spirit to Michael Kremer's "Advanced Market Commitments". But we need such a program everywhere, not just for vaccines, and not just in low-income countries where funding is otherwise unlikely to materialize (although it is urgently needed there!).
Imagine if this had been in place during COVID. Think of the possible innovations:
1) Programs to pay nursing home workers so they stay in just one nursing home
2) Programs to promote and distribute masks and encourage mask uptake
1) Programs to pay nursing home workers so they stay in just one nursing home
2) Programs to promote and distribute masks and encourage mask uptake
3) Programs to provide pulse oximeters onsite at every nursing home for early detection of COVID
4) Programs to improve contact tracing
And hundreds of others...
4) Programs to improve contact tracing
And hundreds of others...
There are so many ideas that could have saved thousands of lives that private firms didn't pursue because there was no clear route to profitability.
But imagine that the pulse oximeter producer signs a contract with Medicare and Medicaid that says: "We will finance a trial where half the nursing homes purchase pulse oximeters. If the impact on COVID cases or deaths is at least X, the govt. will scale it up nationwide."
The innovator does not have to be the provider of the service. Suppose I have an idea to pay nursing home workers to work only at a single site. I believe this will reduce COVID, but I (and potential investors) currently gain nothing if it works.
But if the government said, "If you fund the trial and it gets results, we'll agree to scale it up nationally and pay based on the results", funding to test this idea would materialize, and the route from testing to practice would be dramatically shortened.
Note that programs like this are possible whether you have single-payer healthcare, private firms, or a hybrid system like the US. This is not just a problem that arises when public payers replace private payers.
Imperfect information and adverse selection (by firms) mean that consumers are often skeptical of new healthcare innovations, even when they are valuable (e.g. https://www.nber.org/papers/w27578 ). The market alone doesn't solve these problems.
This idea is related to CMMI, a part of the Affordable Care Act, but the key question is *who finances the trial*. https://twitter.com/darshaksanghavi/status/1347580401108267008
If the government needs to finance such trials, they will be few and far between. But if the private sector can finance them with the promise if large financial rewards if they succeed, they will be much more commonplace.
Of course, there are many important details to fill in. Will the trial results be externally valid and scalable? How much will social welfare be improved? How much should the government pay for good outcomes? But these are not insurmountable.
The bottom line is that we need need more innovation and experimentation, and we need to identify and scale up programs that work. Social impact bonds achieve both, but are virtually unheard of in healthcare. This needs to change.
