Are We Falling Behind on R&D?

Alex Tabarrok is unhappy about the amount of money we’re spending on R&D to fight COVID-19:

The latest relief bill contains another $320 billion in small business relief and $25 billion for testing. Finally, we get some serious money to actually fight the virus. But as Paul Romer pointed out on twitter, this is less than half of what we spend on soft drinks!!!….Despite monumental efforts by BARDA and CEPI we are also not investing enough in capacity for vaccine production so that if and when when a vaccine is available we can roll it out quickly to everyone (an issue I am working on).

The failure to spend on actually fighting the virus with science is mind boggling….Are we more politically divided about PCR tests than we are about unemployment insurance? I don’t think so yet we spend on the latter but not the former. The rot is deeper.

I want to push back on this a little bit. Not completely, though: Generally speaking, I think the evidence suggests there really has been something of a slowdown in the related areas of American entrepreneurship and innovation over the past few decades. But I’m not sure it’s as bad as it’s sometimes made out to be.

First off: spending on COVID-19 testing isn’t R&D. We already have the tests. It’s just infrastructure and technology spending. It is disgraceful that we’ve been so slow to roll this out, but it’s not really related to R&D.

Ditto for vaccines. There are something like 70 or 80 vaccine trials in process right now across the globe and something like 20 or 30 of them are happening in the US, a number fantastically higher than we’ve ever attempted before. But the gating item is mainly testing, not money. There might be something I’m missing here, but I don’t think increased spending is really an issue. As for production of vaccines, that’s more than a year away and I don’t really doubt that funding will be there when we need it.

Finally, Tabarrok’s chart shows a slow decline in federal R&D spending as a share of the federal budget, but who cares about that? What matters isn’t whether we’re spending money on other stuff, but the absolute amount we’re spending on R&D. So here’s another chart:

The blue line is the one of interest: it shows federal non-defense R&D spending adjusted for inflation. In the last four decades, it’s gone up from $40 billion per year to nearly $80 billion per year—though it’s been pretty flat since 2003. Still, that’s a sizeable increase. Here’s a related chart:

The absolute numbers are a little different here thanks to a different baseline for inflation adjustment, and the report I took this from explains some of the reasons to discount the business spending growth a bit. Even taking that into account, though, the overall verdict is obvious: government R&D has gone up somewhat while business R&D has skyrocketed from $70 billion per year to $300 billion per year. Put everything together and total US R&D spending has increased from about $140 billion in 1980 to $460 billion in 2016. That’s pretty big growth.

Finally, here’s a UNESCO estimate of global R&D spending by country:

Should we be spending more? You can make that argument. At the same time, we’re obviously not laggards. Our absolute spending is huge, and there are only a couple of countries that are substantially ahead of us in R&D spending as a percent of GDP.

Now, all this said, there are questions about what we spend our money on. How much is for basic research? How much is wasted? Does business R&D spending really serve the same purpose as government R&D spending?

At the same time, I think you also need to look at spending priorities over time. In the past, we spent a lot of money on things that required gigantic investments in physical infrastructure: supercolliders for particle physics, manned missions to the moon, new energy sources, and so forth. Today our interests have moved into areas like genetic editing and artificial intelligence. This stuff isn’t cheap, but it’s inherently less expensive than some of our previous ambitions.

Bottom line: my pushback against the general theme of stagnation has been the same for a long time. We are disappointed that we lack the spectacular physical inventions that we dreamed of in the past: flying cars, lunar cities, and so forth. And I’ll concede that right now we seem to be in sort of an innovation valley. The two biggest areas of near-future progress are almost certain to be genetic engineering and artificial intelligence, and someday those will blow the doors off anything we’ve done before. However, neither one has really delivered yet. Still, it also took a long time for cars and radio and TCP/IP to change our lives after they were first invented. But eventually they did, and so will genes and silicon.