Among the most exciting news in health research is that the Advanced Research Projects Agency—Health (or ARPA-H) was finally enacted into law some five years after the idea was first floated. Modeled after DARPA (which for decades has sponsored cutting-edge research on everything from the original Internet, to cryptography, to mRNA vaccines), ARPA-H is intended to fund innovative health research and launch new technologies in health. It’s being launched with a billion dollars in funding, short of the $6.5 billion that was discussed at one point, but still no small amount of money.
But ARPA-H might end up being less exciting and innovative than its proponents have hoped.
That is, the appropriations legislation declared that Secretary Becerra was to decide whether ARPA-H should be an independent agency or whether it should sit within the National Institutes of Health.
This should immediately strike us as odd. The law didn’t give the slightest guidance as to how Becerra was supposed to make this decision, nor did it say that anyone else can be involved. There are few, if any, precedents for Congress take the trouble of creating a brand new agency, while delegating such an important organizational decision to a single person, with no oversight or criteria whatsoever.
To make matters worse, Becerra was between a rock and a hard place: Powerful congressional leaders like Pelosi and Eshoo wanted ARPA-H to be completely separate from NIH. Other powerful leaders (particularly Francis Collins, the former NIH Director who recently retired only to be named Biden’s Chief Science Adviser) wanted to keep ARPA-H within NIH’s turf.
Thus, Becerra split the baby: He decided that ARPA-H would be set within NIH so as to use NIH’s facilities, software, etc., but that the ARPA-H Director would report to him directly (rather than to the NIH Director).
The wisest choice for Becerra would have been to set up ARPA-H as an independent agency.
The whole point, after all, is to “pursue ideas that break the mold on how we normally support fundamental research and commercial products in this country,” in the words of President Biden. Or, as Mike Stebbins and Geoff Ling wrote in their proposal with the Day One Project, the point is to “develop new capabilities for disease prevention, detection, and treatment and overcome the bottlenecks that have historically limited progress.”
The law does exempt ARPA-H from the normal standards of peer review, and gives its director the authority to issue grants or contracts in the same way that DARPA does, and to hire people without the normal civil service requirements. If the director reports to Becerra, we might hope that even if set within NIH, ARPA-H could still operate independently to the full extent of the law.
But that might be too optimistic, as there is often a difference between the law as written and the law as implemented in practice. No matter what the law says, there is a substantial risk that ARPA-H could slowly be dragged into the usual practices of NIH—the very practices that inspired the idea of ARPA-H in the first place. As Arati Prabhakar, a widely-respected former DARPA Director, said, “I struggle to imagine how you build the culture you need to build an effective ARPA-H, that actually ends up triggering change.”
There’s even a recent and fairly on-point precedent for this situation.
In the wake of the financial crisis of 2008-09, some observers argued that we should establish a National Institute of Finance to track key financial data that no other agency systematically monitors. But it turned out that the Treasury Secretary at the time (Timothy Geithner) was a staunch opponent.
As Simon Johnson, former chief economist for the International Monetary Fund, wrote in 2014, “Mr. Geithner fell back on a standard bureaucratic trick – he took the Office of Financial Research into the Treasury and set about ensuring that it would never be particularly effective.”
Thus, as scholar Hilary Allen of American University wrote last year, some think that “housing the OFR within the Treasury Department politicized it, hobbled its efficacy, and made it beholden to the interests of the financial industry.” And as a Wall Street Journal headline noted in 2018, “Washington’s $500 Million Financial-Storm Forecaster Is Foundering.”
Let’s not repeat that mistake.
ARPA-H has the opportunity to show that DARPA’s model of moving quickly with aggressive targets and minimal bureaucracy can actually work at a large scale within biomedical and health research. At least two pending congressional bills would give ARPA-H more independence from NIH, and Congress should address them by considering what’s best for medical research and ultimately for the health of patients. And beyond that, Congress and the White House should ensure that any new ARPA-H Director has the backbone to run the new agency with the full independence it needs.
UPDATE on 4/22/2022: The Federal Register now contains Becerra’s notification. It includes this line: “NIH may not subject ARPA-H to NIH policies.” So that’s something.