4 ms·

Real federal expenditures per capita have grown, so it's not merely population growth. https://research.stlouisfed.org/fred2/graph/?chart_type=line&recession_b

Real federal expenditures per capita have grown, so it's not merely population growth. https://research.stlouisfed.org/fred2/graph/?chart_type=line&recession_bars=on&log_scales=&bgcolor=%23e1e9f0&graph_bgcolor=%23ffffff&fo=verdana&ts=12&tts=12&txtcolor=%23444444&show_legend=yes&show_axis_titles=yes&drp=0&cosd=1947-01-01&coed=2014-10-01&width=670&height=445&stacking=&range=&mode=fred&id=GCEC96_POP&transformation=lin&nd=&ost=-99999&oet=99999&lsv=&lev=&scale=left&line_color=%234572a7&line_style=solid&lw=2&mark_type=none&mw=1&mma=0&fml=a&fgst=lin&fgsnd=2007-12-01&fq=Quarterly&fam=avg&vintage_date=&revision_date= https://research.stlouisfed.org/fred2/graph/?chart_type=line... This means that the government is doing more stuff per person. There is no reason to expect basic research funding or other types of funding to track either GDP growth or population growth. Population may double, but that doesn't mean there are twice as many good hypothesis to test. Once you've achieved economies of scale, programs like basic research should grow sublinearly.

nmrm 12y ago
I think the point of the bulk of the document was to demonstrate that lots of good hypotheses aren't being tested (in the US).
seanmcdirmid 12y ago
Strange, the chart doesn't go before 1950, we would definitely see a huge spike in the 40s. Government grows with GDP because people want more services, less crime, less poverty, and stuff like that; there is a reason libertarians can't get elected. One thing people like is going to space, having scientists trained in public/private universities. Even if they go into private industry, they are trained with lots of federal dollars. Cut off all that, and China will just pass us in a decade or two. And really, why not spend additional GDP on public R&D? The benefits have been net positive over the long term, spurring new industries and economic growth. This is a no brainer.
- pandaman 12y ago
  >And really, why not spend additional GDP on public R&D? The benefits have been net positive over the long term, spurring new industries and economic growth. This is a no brainer. Is there any proof for such a claim? Please, before attacking me, realize that there are many states between "more money for R&D" and "no money for R&D". I am merely asking if there is any evidence that adding more money will produce the benefits you asserted, not proposing to cut funding completely.
  - seanmcdirmid 12y ago
    Obviously there is a point where the economy can't absorb it or there is no one to do it, which is why I think keeping it steady with GDP is reasonable. 1% for public non millitary R&D is sustainable, 10% os probably not unless there is a big problem to be solved (e.g. An asteroid tumbling toward earth or something). That also allows us to keep training a healthy number of scientists to do something useful with the extra GDP (as opposed to training more day traders/speculators, capital has to go somewhere, might as well be somewhere useful).
    - xamuel 12y ago
      Sure, there is hypothetically such a point. But if you've looked at the basic research job market, you'd realize we're a long, long way below that point.
      - seanmcdirmid 12y ago
        Sure. 1% might be too low, but hopefully private industry picks up the slack at some point. The government only needs to prime the pump.
        xamuel 12y ago
        A lot of basic research is unprofitable. Private industry has little incentive to pick up that slack. Individual philanthropists might throw basic research a bone here and there, but the most fundamental basic research is completely orthogonal to profit motives. We know about the basic research which led to enormous profits, but that's because of survivor bias. Almost all basic research leads nowhere or takes centuries to mature, and in the rare exceptional cases, it isn't like you can enforce a patent on laws of physics. What we'll get from industry funding is an even bigger circus of 'applied' science (highly credentialed people drawing lines through points and obfuscating everything they say and selling it all as science)
        seanmcdirmid 12y ago
- yummyfajitas 12y ago
  Government grows with GDP because people want more services, less crime, less poverty, and stuff like that; there is a reason libertarians can't get elected. There is no reason for that stuff to grow with (real) GDP, unless you define "poverty" as a relative quantity. It doesn't cost more (in real $) to provide someone with housing and food simply because other people also consume caviar (which raises GDP). It doesn't cost more to put someone in jail simply because other people have an XBox. That stuff would grow linearly with population, however, unlike research. Research doesn't grow because 1 scientist can make 1 discovery which then works for either 10 people or 10B people. And really, why not spend additional GDP on public R&D? Concretely speaking - the population increases from 150M to 300M and that population now enjoys XBoxes, MRIs, restaurant meals and larger houses. Why have the number of feasible research projects suddenly quadrupled? Why should there be a relation between restaurant meals and research projects?
  - seanmcdirmid 12y ago
    > Concretely speaking - the population increases from 150M to 300M and that population now enjoys XBoxes, MRIs, restaurant meals and larger houses. Why have the number of feasible research projects suddenly quadrupled? Why should there be a relation between restaurant meals and research projects? Those Xboxes and MRIs came from research. Tomorrow they probably want holodecks and cures for cancers (selfish peasants!). > Why should there be a relation between restaurant meals and research projects? Future GDP growth depends on basic research, and so vice versa.
    - yummyfajitas 12y ago
      If the population doubles, why do holodecks require 2x as many resources to invent? Alternately, if Keynesian stimulus or welfare reform causes Americans to return to the workforce and raises GDP, why do holodecks require 2x as many resources to invent? That's the core question which you are ducking. No one, least of all me, disputes that research is useful - that's all you've actually argued. I'm merely disputing your assertion of a linear scaling law between GDP and the funds needed for research, and you haven't provided an argument why such a law would exist.

nmrm 12y ago

I think the point of the bulk of the document was to demonstrate that lots of good hypotheses aren't being tested (in the US).

seanmcdirmid 12y ago

Strange, the chart doesn't go before 1950, we would definitely see a huge spike in the 40s. Government grows with GDP because people want more services, less crime, less poverty, and stuff like that; there is a reason libertarians can't get elected. One thing people like is going to space, having scientists trained in public/private universities. Even if they go into private industry, they are trained with lots of federal dollars. Cut off all that, and China will just pass us in a decade or two. And really, why not spend additional GDP on public R&D? The benefits have been net positive over the long term, spurring new industries and economic growth. This is a no brainer.

pandaman 12y ago
>And really, why not spend additional GDP on public R&D? The benefits have been net positive over the long term, spurring new industries and economic growth. This is a no brainer. Is there any proof for such a claim? Please, before attacking me, realize that there are many states between "more money for R&D" and "no money for R&D". I am merely asking if there is any evidence that adding more money will produce the benefits you asserted, not proposing to cut funding completely.
- seanmcdirmid 12y ago
  Obviously there is a point where the economy can't absorb it or there is no one to do it, which is why I think keeping it steady with GDP is reasonable. 1% for public non millitary R&D is sustainable, 10% os probably not unless there is a big problem to be solved (e.g. An asteroid tumbling toward earth or something). That also allows us to keep training a healthy number of scientists to do something useful with the extra GDP (as opposed to training more day traders/speculators, capital has to go somewhere, might as well be somewhere useful).
  - xamuel 12y ago
    Sure, there is hypothetically such a point. But if you've looked at the basic research job market, you'd realize we're a long, long way below that point.
    - seanmcdirmid 12y ago
      Sure. 1% might be too low, but hopefully private industry picks up the slack at some point. The government only needs to prime the pump.
      - xamuel 12y ago
        A lot of basic research is unprofitable. Private industry has little incentive to pick up that slack. Individual philanthropists might throw basic research a bone here and there, but the most fundamental basic research is completely orthogonal to profit motives. We know about the basic research which led to enormous profits, but that's because of survivor bias. Almost all basic research leads nowhere or takes centuries to mature, and in the rare exceptional cases, it isn't like you can enforce a patent on laws of physics. What we'll get from industry funding is an even bigger circus of 'applied' science (highly credentialed people drawing lines through points and obfuscating everything they say and selling it all as science)
        seanmcdirmid 12y ago
yummyfajitas 12y ago
Government grows with GDP because people want more services, less crime, less poverty, and stuff like that; there is a reason libertarians can't get elected. There is no reason for that stuff to grow with (real) GDP, unless you define "poverty" as a relative quantity. It doesn't cost more (in real $) to provide someone with housing and food simply because other people also consume caviar (which raises GDP). It doesn't cost more to put someone in jail simply because other people have an XBox. That stuff would grow linearly with population, however, unlike research. Research doesn't grow because 1 scientist can make 1 discovery which then works for either 10 people or 10B people. And really, why not spend additional GDP on public R&D? Concretely speaking - the population increases from 150M to 300M and that population now enjoys XBoxes, MRIs, restaurant meals and larger houses. Why have the number of feasible research projects suddenly quadrupled? Why should there be a relation between restaurant meals and research projects?
- seanmcdirmid 12y ago
  > Concretely speaking - the population increases from 150M to 300M and that population now enjoys XBoxes, MRIs, restaurant meals and larger houses. Why have the number of feasible research projects suddenly quadrupled? Why should there be a relation between restaurant meals and research projects? Those Xboxes and MRIs came from research. Tomorrow they probably want holodecks and cures for cancers (selfish peasants!). > Why should there be a relation between restaurant meals and research projects? Future GDP growth depends on basic research, and so vice versa.
  - yummyfajitas 12y ago
    If the population doubles, why do holodecks require 2x as many resources to invent? Alternately, if Keynesian stimulus or welfare reform causes Americans to return to the workforce and raises GDP, why do holodecks require 2x as many resources to invent? That's the core question which you are ducking. No one, least of all me, disputes that research is useful - that's all you've actually argued. I'm merely disputing your assertion of a linear scaling law between GDP and the funds needed for research, and you haven't provided an argument why such a law would exist.