You need a certain scale to capture the value created by technology development
Of course, this scale varies by technology.
At the end of the day technology development is expensive. You either need to be creating it for the sheer joy of it or philanthropic duty/satisfaction (which is absolutely a thing but I will ignore that here) or expect that you you will be able to get some benefit from it that is worth the money (or time equivalent of money) that you spent on it. The benefit could be raw ROI or it could be non-monetary.
This definitely applies to incremental process improvements - if you make something 1% better you need that 1% to be of a very large number in order for it to be impactful. This is why Bell Labs was only profitable in conjunction with AT+T.
This is also why technology startups want to grow quickly and get big: they can only really capture the value of their technology at scale.
Even if a technology makes something 10x or 100x better, or even enables entirely new things, you still need to be at a certain scale eventually for it to be worthwhile. Even if you are the only person in the world with an airplane, the amount you would make running your airline service minus the amount of time and money to invent it probably wouldn’t be much more than a non-inventive job if you only have one airplane. But what about patents? For a patent to recompense the inventor scale needs to happen somewhere so eventually there needs to be a large entity that is deploying the technology at scale. The need for eventual scale doesn’t have any baring on the initial funding though, but it does connect the initial funding to a belief that scaled deployment is possible.
The argument is also a post-hoc reason why countries are incentivized to fund research - if you create a new industry (especially one that has to do with national defense), the only entity that benefits from it is the country that plays host to that industry and can tax it (and use its output for defense.)
Looking a little deeper at countries, the statement ‘countries can capture the value of research because of their scale’ needs a bit of unpacking. If the value is mostly tax revenue then one would expect cities and states to be sponsoring more research. There doesn’t seem to be a reason why a country would be able to capture the value of a new industry more than a city or a state. So either the main reason that countries fund research is for defense purposes or for reasons other than capturing value.
There is also something about our notion that discovery and invention is a public good.
It feels like the historical narrative around this is a little muddled. Nation-states only started directly sponsoring research that was not directly associated with the military (like guns) in the mid-20th century. Of course, there are plenty of technologies that are related to the military with varying degrees of directness from better ship-building technology to (arguably) better food production.^1 Roughly contemporaneously patents seem to have become less useful (Patents held back aviation, In most industries outside of pharma and chemicals, patents are a net drag on innovation, Patents have become more BS over time) and governments took on much more research funding. One explanation for the latter is that the returns to research are much larger. Another explanation is that WWII showed that science has become tightly coupled to defense in unexpected ways so the state needs to have a massive portfolio. Another (unsatisfying) explanation is that the role of governments in funding research is a molten pile of different contingencies that solidified into current practice: a combination of the government taking on a greater role through necessity and ability (Great Depression, WWII, communication technology, etc.), the increasing importance of research to public good generally and to defense especially, the post-WWII science system, cold-war cultural flexing, rent seeking, etc.
It’s also notable that the majority of development expenses are by private companies and the “basic research” is sponsored by the government.