[candu]

Government investment in science is...the only way basic science happens, really. I'd recommend reading The Entrepreneurial State [1] here: in essence, basic science pays off too slowly to interest even the most deeply-pocketed capital interests, but it pays off, so wise societies invest in it; Silicon Valley owes its existence to massive formative public investments in underlying technologies.

Not to mention that smart people generally prefer to live in places that value and protect science, so it's _also_ an indirect form of geopolitical talent recruitment. (See brain drain + brain gain impacts of science policy, for instance. There's a strong argument to be made that US mid-20th-century dominance in science and engineering was largely driven by a lot of very smart people fleeing Nazi Germany.)

Basic science isn't so much a lottery ticket as a bond with unknown maturity measured in decades, a _very_ high rate of return, a high minimum investment, and dividend-like payouts created by adding skilled scientists, engineers, etc. to your tax base.

[1] https://marianamazzucato.com/books/the-entrepreneurial-state...

[godelski]

  > a high minimum investment

This is the only part I actually disagree with.

Science is incredibly cheap. It can have a long time to mature but interestingly that is dependent on the number of "bonds", with quicker returns when there's more "bonds" issued.

I'd say there's 4 common classes of misinterpretation:

  Perception bias:
  ----------------

Most of science is performed by grad students and academics. Neither of which are known to make much money and the former is known to make poverty wages lol. I can say as a recent graduate that one summer internship at a big tech company gave me more money than my university's spend for the rest of the year. And as an intern I was still much cheaper than a full employer. My equivalent yearly salary was higher than most professors in my department too.

I'd say 80+% of research is being done at this scale. A few hundred grand per year, if even that.

  Amortization (time bias):
  -------------------------

We often hear about the big science projects and this creates the notion that it's expensive but it's usually misleading. You might hear news like the $5.2 billion Europa Clipper mission, but that's spread out over many years. Work began in 2015, construction in late 2019, full assembly in early 2022, and launch in late 2024, where there's 6 years of flight and the budget is for a mission life until late 2034. Amortized that's $5.2bn over 19 years, so $274m/yr ($347m if we conservatively count from 2019).

  Distribution:
  -------------

Most mega projects have a cost that's distributed over many funders. Take CERN. It cost about $10b to build, took 10 years to construct, and costs $1bn/yr to operate. That's distributed through many countries, the largest contributor being Germany, which only accounts for ~20% (so $200m/yr), followed by the UK (15%), France (13%), and Italy (10%). There are also occasional contributions by the US.

  Scale:
  ------

All these numbers are large, but they're also the biggest projects and there's few projects that big. $100m seems like a lot of money to us because we're imagining it in our bank accounts. But that's not the same as money in a government's bank. The US budget is $6.8 Trillion! $100m is 0.0015% of that! In other words, if you had a million dollars to spend each year you're talking about $1.5k (or $1.47 of a $1000 budget). This is not a big ticket item.

  ===

I'm sure you agree with most of what I've said but I wanted these points "on the record" since we live in a time where we're frequently arguing about $1 from a $10000 budget instead while ignoring the $1000 items. We need to get our heads straight. It's like someone complaining about the cost of your bus ticket while they're buying the latest fully loaded Macbook Pro. I don't think their actual concerned is the budget...

[candu]

100% agree that spending on R&D is _very_ efficient in terms of just about anything you could conceivably care about - downstream economic outcomes, quality of life, geopolitical strength / prestige, etc.

For instance, in 2023 the US spent ~$190B in federal funding on R&D [1], compared to a budget of ~$6T [2] - i.e. about 3%. It's really really not a lot when you consider the aggregate impact over decades.

But it is still a lot in an absolute sense. This funding supports an entire ecosystem across both academia and industry that directly creates hundreds of thousands of jobs, many of which require highly specialized skills. I mention this not to create a sense of sticker shock, but to drive home the point that making this investment is a big and complex task - and one that takes a long time to rebuild. I firmly believe that the current chaos in the US will take at least a generation to repair.

[1] https://ncses.nsf.gov/surveys/federal-funds-research-develop... [2] https://www.cbo.gov/publication/59727

[MattPalmer1086]

Just a note you may find helpful. I feel your post is too long to be digested and responded to in a forum like this.

Maybe I'm wrong, and if so I apologise! But as soon as I saw the essay like format, I knew I wasn't going to spend time on it. I think shorter points that provoke discussion may work better here.