Make the Endless Frontier Act great again

A once-fantastic bill has seen its ambitions shrink

(Xinhua/Xu Yong via Getty Images)

Scientific progress is the fundamental wellspring of human betterment in ways that the Covid-19 pandemic has really underscored — the development of highly effective vaccines based on a whole new technological platform in record time is going to deliver the planet from this scourge, and will likely bring us more and better vaccines for other diseases in the near future.

But the overall pace of productivity growth in advanced countries has slowed down over the past generation or two, and that’s at least plausibly connected to an overall slowdown in scientific progress.

One potential view of the situation is that all the good science has somehow already been done, and we’re now running out of things to learn. But a new initiative pending in Congress and moving to the U.S. Senate floor this week rejects that. It’s called the Endless Frontier Act and, per the name, it’s based fundamentally on the idea that we can accomplish more if we raise our ambitions, and that a large increase in federal science funding is warranted.

This bill existed during the last Congress but had been in a weird kind of legislative purgatory — seen as a mostly Democratic measure, though with a Republican cosponsor in Todd Young, and just not really going anywhere. But it turns out to have been proceeding on the “secret Congress” track where Joe Biden never mentions it; it doesn’t get much mainstream media coverage; it isn’t polarized; but it actually might pass — it secured a bipartisan stamp of approval from the relevant Senate committee last week! Now it’s ready to continue proceeding by “regular order” — the congressional process that everyone says they favor. In other words, it’s not a partisan bill being pushed through on a party-line vote using the reconciliation process. And it’s also not a mega-deal negotiated in secret by the leadership of both parties and then sprung on the members at the last minute. It’s going through a real, honest-to-god amendment process in which all kinds of members can participate.

Unfortunately, this also turns out to mean opening the door to changes that people behind the idea think might undermine its main objectives, and in its current version, the bill is a shadow of its former self.

The original plan: A big new National Science Foundation directorate

The original Endless Frontier Act was a bipartisan, bicameral initiative by Sen. Young (R-IN), Sen. Schumer (D-NY), and Reps. Ro Khanna (D-CA) and Mike Gallagher (R-WI).

The bipartisan, bicameral bill is a storied congressional tradition even in this polarized age. But it’s a little unusual for such a bill to have the Senate Majority Leader as one of its bipartisan sponsors since that tends to make it look maybe not bipartisan. On the other hand, that also means maybe it doesn’t just get left to die on the vine in committee, deemed unworthy of precious floor time like so many of these bills.

Either way, the main idea of the bill was to pump $100 billion over five years into the National Science Foundation. The NSF is basically a big, semi-independent government agency divided into a few directorates that give grants to universities to support various kinds of academic research. Endless Frontier was going to create a new Technology and Innovation directorate to make grants in the fields of AI, semiconductors, quantum computing, robotics, advanced manufacturing, “natural or anthropogenic disaster prevention or mitigation,”¹ biotechnology, genomics, synthetic biology, cybersecurity, data storage and management, “advanced energy,”² batteries, materials science, and engineering.

This is both a big change in the sense that it’s a huge injection of money and also not a big change in the sense that it’s basically just pumping up what’s already the main federal funding stream for research. As you’d expect, the status quo has various critics, and the original legislation sort of nods at all their ideas, suggesting that maybe the new directorate can copy DARPA’s program manager structure (but without otherwise implementing any of the things that make DARPA unique and, to its fans, effective). It also suggests in a nonspecific way that the new directorate “could partner with the rest of NSF and other federal research entities, including the Department of Energy, NIST, and National Labs, to advance its objectives in the key technology areas.”

But mainly it takes the NSF, asks it to do more stuff, and then gives it a bunch more money. Then as a smaller initiative, it invests $10 billion over five years for the Department of Commerce to create regional innovation hubs. Then there’s some smaller stuff about scholarships and technical programs at community colleges and other stuff like that which doesn’t cost much money.

The bill changed a lot in committee

This legislation then got the treatment that Senators are always saying they favor, where it had a genuinely substantive march through the Senate science committee, with amendments from many corners adopted. That turns out to recall the old saw about it being ugly to see the science getting made, as it became a bit of a Christmas tree for random stuff.

One framing of the Endless Frontier proposal is that we need to boost science funding to stick it to China. But there are lots of ways to stick it to China. For example, many people object to shark fin soup on humanitarian grounds. And shark fin soup is primarily consumed in China. So banning the sale of shark fins is arguably a means of sticking it to China, and Senator Brian Schatz of Hawaii got that added to the bill.

Tammy Baldwin and Dan Sullivan teamed up on a logroll to get protectionist nation-of-origin labeling rules but exempt some kind of crabs that people in Alaska care about:

Democratic Sen. Tammy Baldwin, for example, introduced an amendment that would require online sellers to disclose the country of origin of the goods they're selling, providing more transparency about products that are made in China. That amendment prompted a lengthy back-and-forth over how such a requirement would affect the seafood industry. Republican Sen. Dan Sullivan of Alaska offered to help Sen. Baldwin “get this over the goal line,” if only she would exempt certain types of seafood from those disclosure requirements.

Specifically, Sullivan asked Baldwin to create carve-outs for “cooked king crab and tanner crab.” Both the Baldwin and the Sullivan amendments were accepted.

This whole thing actually underscores one of my key points about the filibuster — the reason all these amendments got added is that Senators believe this bill will pass, so they want to fight to get their ideas in it. But another reason all these amendments got added is that Senators are willing to actually support the bill if they can get their key changes made. This is why I don’t think a majority-rules Senate would mean an endless parade of 51-50 bills. Vulnerable members will still shy away from highly partisan fare, but since it’ll be more likely that bills will pass, members will have more incentive to engage constructively with the process.

But the changes weren’t all as trivial as the shark fins. As this graphic from the American Institute of Physics shows, a “substitute amendment” was written by Roger Wicker (R-MS) and Committee Chair Maria Cantwell (D-WA) that drastically scaled down the new directorate’s funding relative to the original proposal. That amendment was adopted without objection, and then on top of that, Democratic Sen. Ben-Ray Lujan of New Mexico proposed an amendment to move a third of the already-reduced directorate funding to the Department of Energy. His amendment was adopted on a 23-5 vote.

What’s up with that? Well, the Department of Energy runs a network of national labs, and there are two of them in New Mexico.

Lujan pushed to get money taken out of the new directorate and put into the labs instead. Sen. Young (R-IN), who hates this idea, said it would jeopardize Republican support for the bill. That may or may not prove to be true, but the dispute really isn’t partisan — Marsha Blackburn (R-TN), one of the most right-wing Republicans, supported this idea because Oak Ridge National Laboratory is in her state.

Ideally, we’d spend on new stuff

It would be nice to think that if scientific research funding helps to get us scientific breakthroughs, then doubling funding would double our breakthroughs. But it’s obviously more complicated than that. And over the years, a bunch of people have developed serious critiques of the basic model of research science that is funded by the federal grant process and passed through American universities.

Stripe founder Patrick Collison and the scientist and author Michael Nielsen wrote a good piece in the Atlantic several years ago about declining scientific productivity, arguing that we are getting diminishing returns out of our science funding and that the NSF itself is dangerously uninterested in that problem. You can also see Nielsen’s writings on the dangers of centralized mechanisms for evaluating science and the true scarcity of expert attention relative to money per se.

I’m not sure I can do full justice to the smart critique of the status quo, but it goes something like this: The current science funding paradigm in the United States rewards competent bureaucrats, not brilliant researchers. The game is to get really good at writing grant proposals, managing staff, and playing the research equivalent of small ball. You want lots of publishable results. That means incremental findings that basically agree with other important people’s existing findings so that at peer review and in various committees, everyone likes you. As a junior researcher, the way to get ahead is for more senior people to decide they like you. If you swing for the fences you’ll annoy people who’ll want to cut you down, and you’ll end up striking out a fair amount of the time, so they’ll have the opportunity to cut you down. So we’re cutting a lot of checks to people who cut checks to subordinates and whose grants get taxed by university administrators to subsidize other stuff, but we’re not getting home runs. In fact, the process is antithetical to home runs, which come from annoying people pushing hard on weird ideas until one of them pans out.

I find this critique fairly persuasive; the existing system is really not ideal, and while I wouldn’t recommend burning it to the ground, I would rather see new dollars pushed into new ideas rather than increasing the volume of the existing flows.

This bit by Robert Root-Bernstein explaining how work from the heroic era of physics would play in today’s grant world captures it well:

Imagine that Roentgen discovers X-rays today, or that Lise Meitner, Otto Hahn, and Fritz Strassman have just reported the existence of atomic fission. What happens? First of all, in our system, all of them risk losing their funding if they follow up their results, since these are detours from their preplanned research programs. At best there will be a delay of several months while they apply to their funding agency for permission to detour. On the other hand, if they complete their current grant commitments before following up their odd results, they may be “scooped” or they may lose interest. Second, to continue their research, they have to apply for further funding, which may take several years. They might not be funded at all. Roentgen’s results were met initially with almost universal disbelief, and the Meitner-Hahn-Strassman work flew in the face of dogma concerning the inviolability of atomic structure.

Alex Guzey has a project called New Science (focused initially on life sciences) that is aiming to develop a different kind of way of funding research. If you are reading this and happen to be incredibly wealthy and interested in science, I really hope you’ll consider helping him out with this or with funding comparable ideas. And as long as Congress is throwing vast sums of money around, they really ought to be looking at diversifying research funding, not just doing more more more more of what’s already happening.

But the NSF is pretty good!

That said, Guzey also wrote a great article a year ago explaining that even though most of what the critics say is true, they’re also kind of wrong. He’s got a long and substantive take on this that’s worth your time, but to summarize it very briefly, there are a bunch of dumb problems and perverse incentives in the existing system. People bend or break the rules, and smart people with promising research mostly get money.

Root-Bernstein, it’s worth saying, eventually became some kind of AIDS contrarian who decided that HIV isn’t the real cause of the disease.

Everyone in every field probably agrees that people like them deserve more funding to pursue open-ended projects with minimum meddling and bureaucratic oversight, and obviously every writer loves the story of the rebel who defied conventional wisdom and changed the world. But for every Roentgen whose results were met initially with almost universals disbelief, there’s someone saying HIV didn’t cause AIDS.

And in particular, I think getting the government to be more creative and less risk-averse with its money is tough because of the basic nature of democracy.

William Proxmire, a well-regarded progressive Senator from Wisconsin from 1957 to 1989, had the kind of populist streak that usually works for politically successful midwestern Democrats. That included the unfortunate mid-1970s invention of the “Golden Fleece” award that he would bestow upon government agencies he believed were wasting the taxpayers’ precious money. And that often meant targeting the NSF:

I’m referring to the Golden Fleece Award, which more than anything else he did brought Proxmire into the national spotlight. In 1975, he issued a press release announcing the National Science Foundation was the first recipient of the award, which he devised to bring attention to what he believed was frivolous government spending for dubious causes, especially in the realm of scientific research. The NSF had spent $84,000 in taxpayer dollars to fund a study on the origins of love. Proxmire didn’t think that was a good use of $84,000. And thus, a new generation of headline-grabbing anti-intellectualism was ushered into an institution where such posturing had a storied legacy on which to build: Congress.

That’s an unfortunate legacy, and I think we should all try to push back on it. The actual cost of saying “yes” to goofy-sounding research is very low as long as the money gets into the hands of talented people.

In other words, I applaud everyone who is trying to create funding streams that break with the existing NSF/NIH model, and I would encourage non-government funders to explicitly look to zig where the government zags. At the same time, I don’t think large sums of taxpayer money are going to be the place to innovate in funding models. And there is reason to believe that giving more to the NSF is a good idea.

The current version of the bill is much worse

I am a realist about politics, and I think that obviously if a bill that involves a huge increase in science funding is working its way through Congress, then members who have a national lab in their state or district are going to fight for a piece of the pie.

New Mexico, in particular, is a sort of weird state that is well below the national average in terms of the share of the population with a college degree but about average in advanced degrees, largely thanks to those labs. If the bill can address shark fin soup, it can boost national lab funding too.

But the overall bill has been greatly shrunk — instead of $100 billion in new science funding, it now boosts funding up to $100 billion, a much more modest increase.

Caleb Watney @calebwatney

@hamandcheese Other main change is that this subsumes NSF's normal budget which wasn't included in the topline 100B before. NSF would have gotten around ~42.5B over the next five years under normal budget anyway, so even NSF's yearly budget is only getting around +2B/per year in new money.

What’s now happening is that the boost in NSF funding is much smaller, and some of that new money is going to the existing directorates, so the proposed new directorate is a relatively tiny $4.3 billion.

Now while the core purpose of the bill shrunk, Schumer has also folded in a bunch of other bits of legislation that are related to the themes of technology and/or China. Some of this is good stuff, like money to boost semiconductor construction capacity. There’s also some space stuff that I’m lukewarm on, and some provisions related to alleged Chinese intellectual property theft that I don’t think I can really evaluate properly.

But the main thing is that while the bill would still amount to something like a $30 billion boost in science funding, that’s much less than what we originally started with. And it’s really clear how much Congress let their ambitions shrink. But the situation seems at least somewhat fluid (for this to happen, there has to be a whole other legislative process in the House), so I’m hoping, perhaps in vain, that with a little public attention we can maybe get some of that money back in there.

The China angle

Meanwhile, not everyone loves the anti-China framing that the bill has acquired.

I see that critique operating on two levels. One is just the view that it’s a little silly that Congress needs to work itself up into a head of steam about China in order to agree to boost science funding. And on that level I agree — it is silly. But this is politics. Almost half of the members of Congress are Republicans. They are ideologically predisposed to be skeptical of government spending, and they are institutionally hostile to universities. At the same time, they are favorably disposed to nationalism and national security arguments.

It is just good sense to let people see the convergence of values in these things, and frankly, I wish we had more legislation like that. Democrats should probably mention that their proposed American Families Act investments would reduce abortions. You don’t do policy by first converting everyone to cosmopolitan secular liberal egalitarianism and then passing bills. You need to try to build coalitions and do things.

A stronger argument, floated last week in Politico, is to ask “Could the Endless Frontier Act fuel anti-Asian hate?”

I think the idea that strong criticism of the People’s Republic of China amounts to anti-Asian racism is just really wrongheaded in a more profound way. The PRC regime is, in fact, quite bad. And since it is located in China, the bad things they do mostly afflict Asians. Criticizing the destruction of democracy and political freedom in Hong Kong is not anti-Asian. Supporting Taiwanese autonomy is not anti-Asian. Trying to forge ties with Vietnam is not anti-Asian.

Overall, I do not think of myself as a “China hawk” or really much of a national security hawk at all. By that, I mean that I’m not really into bombs and aircraft carriers and spending $800 billion on a global military presence that we then turn around and call “defense.” But it is normal and natural for the biggest countries in the world to have some kind of rivalry, especially when their underlying political systems and historical relationships with other countries are so different. To me, making the point that things like investing in science or opening up to more immigration are the best ways to maintain our lead is the alternative to a highly militarized relationship. America is a huge and diverse country that is bad;y suffering these days due to culture-war politics. I don’t want us to find unity in war with China, but finding a measure of unity in a bit of rivalry — especially competition over something benign like science funding — seems good and sensible to me.

1. I think this is a way of saying climate change without saying “climate change.”

2. Again, climate change.