Basic Research As Signal

From a recent Fortune article, a clear example of why corporations do basic research:

For [Bill Gates] this is a triumphant visit to China, a victory lap of sorts, on which I’ve been invited to tag along. The country is his.  No other Fortune 500 CEO gets quite the same treatment in China. … It was not always so. Microsoft bumbled for years after entering China in 1992, and its business was a disaster there for a decade. … But it was a relatively small step in 1998 – the opening of a research center in Beijing – that proved a turning point. "We just started it here because we thought they’d do great research," says Gates, who raves about the quality of the country’s computer scientists. The lab was what Gates calls a "windfall" for Microsoft’s image. It began accumulating an impressive record of academic publications, helped lure back smart migr scientists, and contributed key components to globally released products like the Vista operating system. The lab soon became, according to polls, the most desirable place in the country for computer scientists to work.

Basic research would go on even with no government or charity funding.  Its main function is not research progress, however, but signaling impressive abilities.

GD Star Rating
Tagged as:
Trackback URL:
  • Robin, Gates’ celebrity status in China is also linked with the remarkably bad deal Microsoft negotiated with the Chinese government. $3/seat for Windows licenses for everyone in China, full source code access for the Chinese government, zero commitment to the platform in any official matter, and Microsoft helps perpetuate the Great Firewall of China.

    Don’t get me wrong — I think your point stands. But Microsoft also gave away the farm in other ways to secure preferred status in China. Mostly they didn’t want top-quality research from Chinese computer scientists happening on Linux platforms.

  • Henry V

    Signaling impressive abilities? What about signaling a willingness to play ball? Or, simply PR?

  • Robin, tagging on a sentence like this at the end “Basic research would go on even with no government or charity funding. Its main function is not research progress, however, but signaling impressive abilities.” corrupts your enterprise here in my opinion, because it makes it look like you’re not attempting in good faith to make this blog about ‘overcoming bias’. Instead, moments like these make it feel like this blog is more an intentional usurpation of that space (like calling a slanted news channel “fair and balanced”).

    I think the best solution would be a competitor blog that is more a good faith attempt to overcome bias without attaching other agendas to it.

    What’s annoying here is not so much that you have bias, but that I don’t see even a good faith effort on your part to identify it, be transparent about it, and to overcome it. Some of your other contributors at least seem to be engaged in that struggle.

  • TGGP

    HA, why do you think he is not in good faith? It doesn’t seem that different from other posts by Robin.

  • Hopefully, understanding the social pressures on and function of academic research seems to me important to interpreting the results of academic research, and to identifying biases in such research. You clearly dislike something I have said, but what exactly you disagree with is far from clear to me.

  • Bruce Britton

    the purpose of basic research is to gather basic knowledge about the world. Like electricity and things like that

  • Bruce, you seem to use “purpose” in the sense of what you wish it would do. I meant “function” in the sense of what social context best explains its existence and details.

  • It is certainly true that basic research would continue without government funding. However, its character would be fundamentally different. There would be lots of computer science research, plenty of materials science, a good amount of chemical engineering (and petroleum engineering for sure)… and no particle physics, astronomy, linguistics, or evolutionary biology.

    Well, there would probably be some of those things, but they would be drastically, nearly completely, reduced, along with lots of other sciences that have no fortune 500 companies that need to signal excellence involved.

    But sure, *some* basic science would still happen.

  • Carl Shulman

    Some alternative functions of corporate basic research:

    1. Internal morale: techie employees will be pleased that their organization is contributing to ‘cool’ research. Also, employees can be rewarded with rotations in the research group (McKinsey rotates consultants to its McKinsey Global Institute, a development think tank), improving employee retention in profitable segments. Basic research may also provide opportunities for training and development of human capital in employees.
    2. Name recognition (aside from ability signaling): getting research into the news earns free media among potential employees and business partners.
    3. Appeasing governments with public service to forestall threatening regulation, e.g. Bell Labs during the Bell monopoly (similar to the news programs of the major television networks, which were trotted out for license renewals).
    4. Signaling a science-friendly or intellectual corporate culture (separately from ability) to help attract employees from academia.
    5. Technophile managers exploiting principal-agent problems to extract funds from shareholders to advance research they support. (David Shaw has set up a protein biology team inside his investment management firm.)
    6. Companies may be overconfident about their ability to capture the fruits of basic research, e.g. Xerox.
    7. Basic research may be funded by consortia, monopolists, or just large players that really are able to capture a worthwhile chunk of the benefits of the research or take unusually long-term bets.

  • Robin, this line seems pretty arbitrary to me “Basic research would go on even with no government or charity funding” except as invoking a type of libertarian belief-as-attire vibe on the blog. Building this sort of vibe, that this blog is more a space for a certain type of belief-as-attire cohort (which I’d roughly describe in contributors as hierchically not-quite-first rung academic white guys and in solicited readers as hierchically not-quite-first-rung knowledge worker white guys) corrupts an overcoming bias mission, in my opinion.

    We as a species could use a space focused on overcoming bias that guards itself against the sort of corruptions in the writings of people like contributors to The Corner group blog of National Review (Jonah Goldberg, etc.) or bloggers like Steve Sailor. In some of the commenters and fewer of the contributors I see a good faith effort to keep this blog honest in that regard. But I don’t see much of a good faith effort on this from you, Robin.

  • michael vassar

    Carl: Was Xerox really overconfident? It’s clear after the fact that they didn’t capture profits from PARC, but the value created was HUGE. If they had a reasonable chance of creating that much value and a reasonable chance of capturing as much as M$ or even Apple, their investment may have been sound.

  • Hopefully, your complaint is that one sentence of mine gives you the wrong “vibe”? You don’t even say whether you disagree with it, or if you think it relevant in context. With such a weak complaint, it is hard to take you seriously when you say this is evidence that I am corrupting the whole blog.

  • Robin, I think it’s pretty straightforward from my post that I’m referring to a larger trend than one sentence. I’m not really asking you to take me seriously since I think the point is simple and I think it’s more likely you understand it than not. Thanks for leaving my comment up for a larger audience to draw their own conclusions from my expressed concern.

  • Bruce K. Britton

    Does anyone know how Robin found out that the “main function” of basic research is “not research progress.” A separate question is how Robin found out what its main function is, namely “signalling impressive abilities.”

    Are not these empirical questions? I have an anecdote: Based on my 30 years of doing basic research, it was my impression that the “main function” of it was exactly “research progress.”

    There was also some “showing off” by researchers, but most often in a half-joking way, but it was not the main function, I think I and others would have noticed a researcher whose main purpose was anything other than research progress.

    And we would have noticed a person who was just signalling impressive abilities.

    But of course, I do not know what is going on in other minds, just my own, and I was trying to make research progress.

    But how does Robin know what the main function is not,and how does he also know what the main function is?

    Perhaps Robin just has anecdotal evidence, like my anecdote, though I would be delighted if he had more than anecdotes, because anecdotes are not very persuasive, usually. Perhaps there is some other research on the subject, which Robin has found or done, that would advance our knowledge on this subject of interest.

    This raises some interesting methodological questions, like “how do you decide what the main function of something is?” I suppose the most direct way would be to ask the people who were doing the “something.” For example if we want to know what the main function of a nuclear power plant is, the operators might say “to generate power.” More to the present point, what methods did Robin use to establish not only what the main function of research is, but also what its main function is not.

    But can you trust the researchers to tell you the truth? And if not, who do you ask?

  • Carl Shulman


    Good point, Xerox is a poor example. Even in evaluating success in hindsight, it matters enormously how we characterize the research leading up to the laser printer.

  • Carl Shulman


    If donors to universities and national grant-making agencies fund research with the aim of producing research progress, while the grant recipients seek (at some level of description) to signal ability, which would you characterize as the ‘main’ function? If you give pride of place to the perspective of the researchers, do you think that signaling ability is more important than research progress, financial incentives, and displaying achievement? It certainly seems that achievement is important to status, separately from ability.

  • Bruce K. Britton

    Robin is asking a “why” question: Why do people do basic research?
    And the word “why” has two meanings. One applies to inanimate entities and the answer is antecedent conditions, so if we ask ‘why is the sky blue’ or why does virus x make us sick, the answer is something that happened antecedent to when we got sick, or antecedent to light reaching the part of the sky we can see>

    But the other meaning of “why” refers to only animate beings, and only animate beings who have “intentions.”

    Here the answer can be in future terms, why did we do basic research? In order to, in the future, have something else, like scientific progress.

    Or in order to in the future have some consequence of signalling impressive abilities, which would be what?
    I suppose we can rule out any kind of reproductive consequences of signalling impressive ablities, unless Nobel Prize winners have lots more children than other people, or much more sex, or something like that? If its not that, what could it be, a higher salary, a job offer, or narcissistic levels of self-esteem?

  • Carl, my guess is that research progress is mainly a side effect both to the researchers and to the funders.

    Bruce, I am acting as a social scientist here, trying to account for all of the relevant behavior data of which I am aware. The evidence in this post is of course only one small part of the total evidence. What the people involve claim to be their motive and functions is of course relevant evidence, but hardly decisive.

  • Carl Shulman

    “Carl, my guess is that research progress is mainly a side effect both to the researchers and to the funders.”


    This is a different claim from saying that both researchers and funders aim to show off impressive abilities. Does the NIH really aim to help scientists show off? Or does it seek to maintain a process for which it will not be criticized, that will advance science, and that will maintain bureaucratic privileges?

    Would you say that U.S. public companies’ main function is profitability? Managers wish to retain their jobs and enjoy high compensation, directors wish to maintain their pleasant gigs and relationships while avoiding embarassment or lawsuit, and shareholders seek to increase the value of their investment, but no one has corporate profitability as their main goal. Even though the goals of the individual actors are only imperfectly correlated with profits, and corporate behavior very frequently diverges from profit-maximization, it seems that there is still an important sense in which corporations’ function is indeed earning profits.

  • I agree that research performs a signaling function, but the evidence you present here is pretty weak. It’s easy to imagine that Chinese attitudes would have shown the same change without Microsoft opening a research center there. E.g. the mounting evidence that increased globalization and acceptance of big business correlated with increased wealth might have caused a change in opinions about businesses.

    Your statement that “research would go on even with no government” seems superfluous if bias is your only interest, but not superfluous if you are also interested in signaling libertarian sympathies. But HA appears to exaggerate when he calls this part of a trend.

  • I first suggest that we strongly distinguish between the motives of RESEARCHERS, and the motives of GRANTMAKERS.

    Marvin Minsky, speaking at the recent annual convention of the World Transhumanist Association, said that it is no longer possible to get funding for genuinely basic research in Artificial Intelligence. If this was an exaggeration, it was only a slight one, from what I’ve seen.

    Oh, there’s plenty of programs out there which say they want to fund brilliant, innovative, fundamental, outside-the-box research, but what they actually fund is research with a high probability of a small return. The committee members who vote to approve a genuinely basic grant – meaning that there’s a 90% subjective probability it will fail – do not share in the profits if the grant creates ten billion dollars of value. Their incentive is solely to protect their resumes. As long as committee members are not fired for failing to produce major revolutions, they have no incentive to fund major revolutions.

    I have seen nothing to contradict the belief that many researchers would like to do more basic research with a longer-term time horizon.

    I have seen plenty to support the belief that the top echelons of grantmaking agencies want to be seen as doing blue-sky research and put out PR to this effect, while the middle echelons have no incentive for supporting anything that looks weird or that doesn’t have a 90% probability of producing at least some results.

    All these comments apply to government and charity funding just as much to corporate funding, if not more so. The only return to government bureaucrats is PR. Very few philanthropists can look beyond reputational returns to actual utilons created (Peter Thiel is the only example that springs to mind). Corporations have at least some chance of capturing the value of a Black Swan from basic research.

  • Flynn, a valid point.

    Peter, yes, this one case is by itself weak evidence.

    Eliezer, I think you mean that many researchers would like to be paid to do more ambitious basic research. The fact that that they do not pay for such research themselves, however, is important evidence about their primary motivations.

  • TGGP

    I don’t think the statement qualifies as belief-as-libertarian-attire. It doesn’t state that it would be preferable if government did not fund research, only that basic research would still be done. He does not state that what research is done by private companies like Microsoft is more productive or efficient or altruistic than the government funded variety, as he seems to classify all research as primarily functioning to signal expertise and talent. Most libertarians look much more kindly on charities than government, but Robin lumps them together.

    I don’t know if the Corner allows comments, but have you tried mentioning to other bloggers what you see as corruption in their writings? Or do you only see folks here or at Marginal Revolution (recalling the nanny-state comment) as possibly being receptive? I tried telling Sailer once that he was acting petty, but it got other commenters irritated. The comments section there does tend to be rather deranged.

  • Bruce K. Britton

    Robin, you say there is ‘evidence’ in your original post.

    So far that is the only evidence we have, as far as I know, except other anecdotes.

    When I look at that post, what would be interpreted as ‘evidence’ for your claim?

    I’m not expecting some bold claim in the quote from the media report on Gates’ trip to China, like

    ‘Gates announced today that the main function of basic research is signalling of impressive abilities.’

    But I would like to see something in the quote that I can interpret as evidence for your claim. You must be looking at something in the quote when you conclude that, and I don’t see what it could be.

    Could you tell me please where the evidence is?

  • Bruce Britton

    Robin, Gates himself says in the quote that he started the labs to do great research.

    Where do you get signalling of impressive abilities from that?

    You have to have some counter-evidence to his claim, don’t you?

    Or can you just state the contradiction and leave it at that?

    But if you leave it at that, then all the evidence is against you.

    And Gates is of course the prime authority on why he is funding basic research, as only he can have access to his motives.

    Is there something in the quote that I have missed?

  • Bruce, in this example it appears that the Chinese liked to associate with an impressive research institute. By providing the Chinese with an opportunity to do this he gained their favor, helping induce them to buy his product.

  • It’s also not “how much” they spend on research but “how” via leveraging in four areas: ideation, project selection, product development, and commercialization.

    Robin: Basic research would go on even with no government or charity funding. Its main function is not research progress, however, but signaling impressive abilities.

    About signaling abilities. Here is some data, that might be useful to analyze. If you look at some average measures, the number of science citations in in Italy and France are rather similar, yet, proportionally, the France government inputs about two times Italy’s government of GDP into basic research. Italy has almost no basic research in the private sector, either. Some explanation? The only two explanations I’ve found is that First: Italy is unusual in the amount of support that the families give to their children. Most Italian scientists I know are living in property that were either given to them, or acquired with substantial familial financial support, so in essence, the Italian families are subsidizing Italian science. Second: After decades of steady decrease in government funding, those who remain in the country are unusually passionate about their subject.

  • Bruce K Britton

    Robin, You claimed that the quote was ‘evidence’ for your claim, but the quote says nothing about the chinese being induced to buy gates’ product, which you seem to claim the quote says, but it does not say.

    Moreover, the quote has no infomation at all about signaling impressive abilities, or the relation between basic research and signaling impressive abilities, all it does have, relevant to your claim, is evidence directly contrary to your claim, namely gates’ statement that the institute was created to do great research.

    So the quote you are claiming as evidence for your claim provides no evidence for it, only evidence against it.

    In what way can evidence against your claim be interpreted as evidence for your claim? Do you see what I am saying?

  • Pingback: Overcoming Bias : Academia’s Function()