The Case For The NSF No Longer Funding Political Science
    By Hank Campbell | May 16th 2012 12:32 PM | 12 comments | Print | E-mail | Track Comments
    Last year, fiscal hawk Senator Tom Coburn stuck his economic talons into waste at the National Science Foundation. Scientists, alarmed at the prospect of losing funding, circled the wagons while the usual kooky progressive suspects claimed because Coburn had an 'R' in his political party, he must be anti-science.

    In actuality, he was being incredibly pro-science.  He was actually trying to get the NSF to fund more science.  He isn't some lawyer on a cultural bender, he is a medical doctor who understands the value that science research has brought to America.   He just didn't think playing Everquest 2 was science.  And he was right.   While the loudest and shrillest in science media sought to make it a 'Republicans hate science' issue, actual scientists who aren't media mavens quietly agreed - the NSF, with 'science' in its name, should not be spending money analyzing the habits of Farmville players, since the creators of Farmville have clearly done just fine understanding their users.

    In the aftermath of World War II, the government saw the value of what science could do and set out to formally guide it.  They do just that with dozens of overlapping agencies funding basically the same stuff, with no clear boundaries, but their reach has grown and now the majority of research is funded by taxpayers indirectly, but the changing whims of political administrations directly. Because the goals of politicians are political, that has gradually meant including social science as 'science'.   Obviously the social sciences have value but if you go to an older university they are located in the humanities buildings and not the science ones.  There is a reason for that.  They have different missions than science.

    Tom Hartsfield is a physics Ph.D. student at the University of Texas and argues that while political science, for example, has value, he and his peers in physics or life sciences or earth sciences should not have part of a very finite budget siphoned off for philosophical issues like "how power affects empathy" and the "outlook on life and political ideology". Fund them, to be sure, but not as part of the NSF because it "does not and cannot follow the rigorous requirements of reproducibility, testability and objective truth required of science."

    NSF Should Stop Funding Social 'Science' by Tom Hartsfield, RealClearScience


    Ignorance repeated is more ignorance. Hartsfield emphasizes that science is "repeatable, exact and quantitative." I would add disprovable. Hartsfield's hypothesis is easily disproved by looking at an recent issue of the American Political Science Review, the American Journal of Political Science, the Journal of Politics, and so on. There is a battle in political science right now between traditionalists and those who pursue questions with research designs that are repeatable, exact and quantitative (also disprovable). One of the MAIN ALLIES on the right side of this battle is the National Science Foundation. If much of political science looks like what Hartsfield disparages, then why not champion funding for the portion he would recognize as his type of science? Perhaps he is not applying science to his own opinions.

     One of the MAIN ALLIES on the right side of this battle is the National Science Foundation.
    But that is the problem, right?  The NSF is interested in enhancing its budget and authority and not thinking about science.  No one is saying political science has no value - but it is not science. Declaring that 'the right side' happens to be the side you like is the problem in political science.  It is opinion.

    If a political science agency were funding genomic studies, would you argue that worthy political science projects denied funding due to that shouldn't complain, because genomics has future policy implications?
    Gerhard Adam
    Oh boy ... give me more of this:
    Scholars have already begun discovering specific genes associated with political behavior, which may be the first few pieces in the puzzle to understanding the biology that underlies it. For example, two studies (Fowler and Dawes 2008; Dawes and Fowler 2008) recently identified variants of three genes that are positively correlated with voter turnout.
    I guess it's a good thing that our genetics [and biology] are predisposed to the American form of government.  Otherwise, there's no telling what the data might show.
    Mundus vult decipi
    You're gonna get called a Nazi being all rational and science-y like that.
    Gerhard Adam
    Stay focused on the hat .... :)
    Mundus vult decipi
    Confusing politics and political science. Is biology not a science because natural selection involves winners and losers? Having opinions and studying those who do are two separable things.

    BTW: saying something is not science is not particularly scientific unless you specify the criteria by which something is scientific. I think when you do you will find that much of political science IS science (i.e. it specifies disprovable claims based on logic which are then tested, and not infrequently disproved).

    Many people confuse "science" for "truth". Truth is what science is pursuing, not what it is. It is the careful identification of mis-truth that is at the core of the scientific endeavor. Is economics a science? Many political scientists are using the same theoretical tools and methods (in fact, they are often criticized by humanists for doing so).

    Gerhard Adam
    Is biology not a science because natural selection involves winners and losers?
    No, but it does demonstrate that you're conflating your values with natural selection.
    Truth is what science is pursuing, not what it is.
    No, at best that might be a philosophical interpretation, but even that [I would argue] is an incorrect interpretation.
    It is the careful identification of mis-truth that is at the core of the scientific endeavor.
    Again, no.  That's based on Popper's notion of falsification.  So, what is the over-arching theory that political science is based on?  What is it that can be falsified?  What predictions does it make?

    BTW ... economics is not a science either.
    Mundus vult decipi
    This is simple. To be scientific, the assertion that something is not a science has itself to be scientific. This means it has to be treated as hypothetical until demonstrated. Since anecdotes are not evidence of a trend, it is not clear how Hartsfield or others have demonstrated anything. Further, the method applied is not science. I don't know what Garhard's definition of science might be. Hartsfield mentions three criteria:

    reproducibility: Most major journals in Political Science require replication as a standard for publication. You can go to the journal pages, download the data, read the studies and reproduce their findings. Full stop.

    testability: Again, go to any major journal and you will see tests applying standard notions of statistical inference. There are also experiments in the classical sense. It is not clear that this criterion can be applied in all "hard" sciences, since of course research on the origins of the universe only has one case, so the tests must be over-determined.

    Objective truth: This is a bit bizarre as a standard, since science never has truth to compare itself to, just observations that themselves are more or less closely tied to underlying causal mechanisms (ever heard of an "error"?). But anyway, objectivity is again present. How many members of Congress are there? Is this subject to interpretation? What is size of the US budget (or the NSF Political Science budget --- if this is not objective then what exactly are we fighting over?)

    So, claiming that something is not science in the absence of systematic confirmation is not science. QED.

    Gerhard Adam
    How many members of Congress are there? Is this subject to interpretation? What is size of the US budget (or the NSF Political Science budget
    What does that have to do with anything?  Mathematics is chock full of explicit calculations and values, but it isn't scientific.

    Testability is a strange way of saying "falsifiable". 

    Similarly, reproducible is essentially predictability.

    So, again ... what does political science do to satisfy that criteria?

    Mundus vult decipi
    I'm baffled by his persistence that statistics equals science.  People may get tired of me harping on postmodernism but this horseshit subjectivity regarding science started in the humanities and they should be the first ones up against the wall for it.
    This is just nonsense. Hartsfield's opposition cannot be to method, since this is not demonstrably different. Much of political science is doing EXACTLY what Hartsfield is advocating -- applying principles of reproducibility, testability and objectivity (objective truth is a misnomer -- what objective truth do we have on the edge of knowledge? That comes afterward, as a product of science). Further, NSF funding is behind the impetus to apply scientific methods to political science. The portion of political science he opposes is the portion that is NOT funded by the NSF. It appears as if Hartsfield has assumed that one cannot apply science to the SUBJECT MATTER of politics, and then proceeded to his conclusions, something demonstrably false.

    We are also talking about $10.5 millions for ALL of political science out of a $6.8 billion budget. How much physics is Mr. Hartsfield going to get done (with indirect costs) for this figure?

    Gerhard Adam
    ...something demonstrably false...
    How so?
    Mundus vult decipi