Government Priorities: Compare And Contrast

Part One: The federal government funds a study to monitor bad political speech on Twitter:

The National Science Foundation is financing the creation of a web service that will monitor “suspicious memes” and what it considers “false and misleading ideas,” with a major focus on political activity online.

The “Truthy” database, created by researchers at Indiana University, is designed to “detect political smears, astroturfing, misinformation, and other social pollution.”

Part Two: The federal government ignores, for a generation, its statutory obligation to gather and report on data about excessive force used by law enforcement:

But in the 20 years since the law went on the books, very little has happened and no one's holding the AG or any of the law enforcement agencies below him accountable for the lack of input.

Last 5 posts by Ken White

Comments

  1. Roy says

    Those aren't directly comparable priorities, though. An NSF grant to create a non-government run service vs. a report that's supposed to be generated by the Department of Justice itself only comes together in the top-level federal budget, and I hardly think that the NSF gets too much money in relation to the DoJ. Perhaps interested parties could try to get an NSF grant to start a crowd-sourced database on excessive force.

    Do you find the Indiana university project objectionable? Isn't the remedy for bad speech supposed to be more speech? For that to work, a light has to be shone on that bad speech. The linked article implies that the result will be untrustworthy because of the personal politics of the lead researcher, which seems to be to be a pretty weak ad hominem. That said, I would hope (as I hope for most publically funded research) that the low level source be made publically available, which is the best way to find and counter any intentional or unintentional bias introduced into the results.

  2. Rob says

    The National Science Foundation is financing the creation of a web service that will monitor “suspicious memes” and what it considers “false and misleading ideas,” with a major focus on political activity online.

    The “Truthy” database, created by researchers at Indiana University, is designed to “detect political smears, astroturfing, misinformation, and other social pollution.”

    Well that's not creepy and liable to be abused by the party in power at all.

  3. says

    I would like to know who gets to define the terms "suspicious memes" or "false and misleading ideas" for this study. And more important, what kind of personally identifiable information is collected or retained. Then most important who controls access to the database?

    The scientist in me doesn't necessarily oppose studies like the first when it's done for actual academic purposes. However, the libertarian in me gets majorly bent out of shape seeing any sort of government monies used to fund this kind of research or the possibility that they may be given any more access to the data than what is published in the peer reviews.

    As far as tracking police use of force? Sorry, but that's not a federal job either, except where federal agencies are concerned. The States are supposed to protect us from the Federal government, not the other way around. Now that may be appropriate for a State to do, and certainly county and local, but fed? No. That said, I would seriously like to see that kind of data on the states.

  4. Wyrm says

    @Rob
    Your sarcasm and criticism have been flagged by the thought police.
    Please register at your local police station immediately.

  5. FSE says

    Like Rob wrote, NSF grants are a general indicator of academic priorities, not political priorities. And they span a pretty wide range of topics. Look, here's one that you might like:

    American Cities and Police Responses to Social Movements

    In a prior study, one of those NSF investigators concluded:

    Countering terrorism requires that the state win the battle for “hearts and minds”among bystander populations. By treating populations with suspicion, the state may be discouraging cooperation and even reinforcing terrorist narratives and recruitment efforts

    There, your tax dollars at work. Happy?

  6. Rob says

    Wyrm
    August 25, 2014 at 3:28 pm

    @Rob
    Your sarcasm and criticism have been flagged by the thought police.
    Please register at your local police station immediately.

    Nuts!

  7. sorrykb says

    There's more info about the study here . It looks like it could be interesting. The researchers plan to develop a publicly-available open-source platform (that could then be adapted or expanded to other areas of research), and the data will be made available as well. They say, "We hope to provide for the first time a large-scale infrastructure to delve deeply into a broad set of questions about how and why information spreads online." (So, yes, the platform they develop could be used to track the diffusion of certain political speech, or it could be used for other less nefarious purposes. Or it could — and most likely will — be used by advertisers. ugh.)

    *hums* We can't go on together, with suspicious memes…

  8. Roy says

    Access to the data: http://truthy.indiana.edu/apidoc

    From looking at the site, it seems that they are taking publically accessible data, via the Twitter API, and outputting publically accessible data, through their own API. I DIDN'T see anywhere where you could request raw source, which is unfortunate, but I think all the pearl clutching about this is overblown.

  9. Jason says

    This is great once the NSF is done with their truth detector, they can hand it over to a new organization (lets call it the Ministry of Truth) which can monitor the internet and protect us from all the untrue things out there.

  10. Roy says

    Step 1: Use a twitter bot to detect sock-puppeting
    Step 2: ???
    Step 3: A boot stomping on a human face, forever

    Good thing you are onto their cunning plan.

  11. Rick says

    Well, we have the best government that money can buy.

    Oh wait…no we don't. More like "Meet the new boss/Same as the old boss."

  12. sorrykb says

    I wouldn't mind the NSF funding a study on excessive force by police, provided the data and results were public. It would make a whole lot more sense that giving the job to the International Association of Chiefs of Police.

  13. DRJlaw says

    @Raul

    "[when] any sort of government monies used to fund this kind of research or the possibility that they may be given any more access to the data than what is published in the peer reviews."

    Concerning the first part, so long as government is a significant sponsor of academic research, including sociological research, I disagree. There's no reason to cut off funding simply because the research may have political applications. Concerning the second part (access), I certainly agree and think that the typical protections applied to collected research data (pseudoanonymization via randomized ID or aggregation without individualized data retention) should be applied and applied with respect to all seeking that data.

    "As far as tracking police use of force? Sorry, but that's not a federal job either, except where federal agencies are concerned."

    Actually, it is, pursuant to certain civil rights laws which are based upon the Fourteenth Amendment (note section 5 of that amendment). In the abstract, it is not the Federal government's job to enforce standards for police use of force or demand general information concerning it, but it can research and ask all it likes, and tie its grants to law enforcement based upon that information if it chooses to (thank that pesky general Welfare clause).

  14. Deniable Sources says

    Well, it would be rather difficult to gather data on the excessive use of force by police, since the general supposition, strongly endorsed by those in power, is that there isn't any such thing, or at least not really. Not that you'd notice, anyway.

    Seriously, it's hard. It requires actually collecting information that's not always readily accessible. You have to call people. You have to standardize terms and normalize data. It would take real time and real effort, and the returns wouldn't justify further effort – in fact, the more you collect, the less people want you to collect anything else.

    The Twitter study, on the other hand, just requires that you have a college, a network, a little API work, a press release every few months, and voila: a paper-generating grant-justifying machine. You don't ever even have to step off campus. As a bonus, you can have a student do most of the difficult programming work, pay them virtually nothing, and leave the money to sponsor the professor's salary. Good work if you can get it, and you can apparently get a lot of it.

  15. TimL says

    In my experience, NSF's funding decisions are made without political influence. Most people that I know at NSF are academics on temporary leave from their home institutions, working at NSF in positions with contracts of less than 3 years.

    Funding decisions, for the most part, are made by external review panels, not government employees. Research grants, like this one, are HIGHLY competitive; in my experience, NSF panels fund 5-10% of the project proposals that are submitted to them (and it was much worse in the year when we had the sequester).

    NSF's mission is to fund basic research leading to fundamental understanding; the mentioned project seems to be within this scope. With absolute certainty, the researchers are publishing their work in high-level scholarly journals — which is NSF's standard metric for productivity.

  16. Jacob Schmidt says

    It seems rather silly to compare two very different parts of the federal government, then lump them together to make a point. That's like blaming the chemistry department for poorly maintained lab space in the biology department.

  17. LDouglas says

    I like this site a lot, Ken, but your attempt to imply there's something nefarious about the NSF grant- which relies on leaving out critical context- is way beneath you.

  18. says

    From a quick glance at the site, this looks like a pretty routine data mining and computer science research project. They're attempting to spot disinformation by looking at the diffusion patterns of references and retweets.

    For example, normal celebrity (or politician) information radiates outward from the celebrity as his tweets influence an expanding discussion which spreads throughout the twittersphere. Natural political discussions, on the other hand, originate across thousands of people all expressing their views on a subject. The stuff they're calling "fake" resembles neither of these patterns. Instead, dozens or hundreds of accounts with little history all simultaneously start tweeting into other discussions. It's a coordinated attempt by multiple accounts to change the subject to a particular talking point or web site.

    So they're not really looking for wrong information as for information that's not spreading in a natural way. Frankly, this seems likely to undermine control by the major political parties by detecting their attempts to manipulate the media.

  19. Jonathan says

    As someone who works in research administration for a top-tier research university, this article looks like one of those gell-mann amnesia instances you wrote about. It's extremely unlikely that the NSF is funding the "outcome" of a database to track what political speech is true vs. false. Almost certainly, they are instead funding fundamental research theories which speak to semantic analysis or information theory, and this project is one output from a research team with many diverse funding sources.

    But, "the government is funding a dissent database" is a far more sensational headline.

  20. says

    Roy
    "Those aren't directly comparable priorities, though. An NSF grant to create a non-government run service vs. a report that's supposed to be generated by the Department of Justice itself only comes together in the top-level federal budget"

    Read those links. The one time the DOJ sorta semi pretended to gather the excessive force data… they shuffled it off to a "non-government run service" (IACP; the folks running the departments that would have to self-report their own misconduct – no conflict of interest there).

    "Do you find the Indiana university project objectionable?"

    Abso-frickin'-lutely.

    "Isn't the remedy for bad speech supposed to be more speech?"

    This puts the gov in control of what gets defined and reported as "bad speech." Hmm, what could go wrong with that? It isn't as though they support regional fusion centers that release reports calling Constitutionalists terrorrist/extremists, or run a "terrorist watch list" and "no-fly list" with which they get to define "terrorist" with effectively zero oversight or due process (I seem to recall a recent court case about that).

  21. Deniable Sources says

    @Jonathan,

    It's extremely unlikely that the NSF is funding the "outcome" of a database to track what political speech is true vs. false.

    I completely agree. That's not the NSF's modus operandi. The NSF is, instead, funding a lazy and pointless survey of a conveniently located source of hopelessly skewed data – to wit, Twitter – in the completely understandable process of using federal tax dollars (or rather, in 2014, obligations on our children's children) to perpetuate the salaries and office space of the investigators. It's not conspiracy, it's just the sludge that consumes far too much NSF money these days.

    Almost certainly, they are instead funding fundamental research theories which speak to semantic analysis or information theory

    Bet me. The cutsey "Truthy" name alone gives it away. I'd be stunned if this produced anything more fundamental than maintaining a standard of living for the researchers.

    this project is one output from a research team with many diverse funding sources

    Now THAT I'll give you. Damn things are like kudzu.

  22. says

    There should be a high wall of separation between science and state.. Outside of a clear constitutional function, such as national defense, the government has no business funding any "science". This particular project might employ scientific methods but its stated goal is purely political. "Social pollution" and " hate speech" aren't objective terms.

  23. CJK Fossman says

    Outside of a clear constitutional function, such as national defense, the government has no business funding any "science".

    Right. Because, for example, the tobacco companies did such a great job.

    its stated goal is purely political

    No, it isn't. This is the goal:

    Truthy is a system to analyze and visualize the diffusion of information on Twitter. The Truthy system evaluates thousands of tweets an hour to identify new and emerging bursts of activity around memes of various flavors. The data and statistics provided by Truthy are designed to aid in the study of social epidemics: How do memes propagate through the Twittersphere? What causes a burst of popularity?

    We also plan to use Truthy to detect political smears, astroturfing, misinformation, and other social pollution.

    Further, Truthy tracks left-leaning memes and right-leaning memes equally. Go see for yourself. Anybody can submit a meme to Truthy. Truthy publishes the twitter content driving each meme. It's transparent. Even somebody calling himself @SIV can go there and see what the evil Dems are doing.

    Finally, Truthy is not rocket science. The major political parties and well-funded campaigns could easily already have something like it and probably do.

  24. Paul Baxter says

    I think for that law to actually become effective in any way we would need the active involvement of some sort of non-profit group which would both harass and cooperate with the relevant authorities to see that progress is made.

  25. sorrykb says

    CJK Fossman wrote:

    Finally, Truthy is not rocket science. The major political parties and well-funded campaigns could easily already have something like it and probably do.

    Exactly. This grant expands the work and will allow the rest of us to take advantage of it.

    And, for clarification, Truthy was created before the grant. The grant was awarded to broaden the work (beyond Twitter) analyzing public data in social networks. It's a bunch of researchers geeking out on social network analysis, but it's a useful tool that could have many positive applications. (It could also have negative applications, but that's how tools tend to work. This one at least is out in the open.)

    BTW, this grant began three years ago and is now in its final year. (It expires June 30, 2015.) Not sure why it is suddenly making news now. Although, come to think of it, we could probably use the research funded by the grant to figure that out.

  26. Devil's Advocate says

    @SIV

    There should be a high wall of separation between science and state.

    If by that you mean that politicians shouldn't influence scientific results or funding decisions, I agree with you. Fortunately, the government grant process does an amazingly good job of shielding the fundees from the funders.

    Outside of a clear constitutional function, such as national defense, the government has no business funding any "science".

    Surely making legislative, executives, and judicial decisions are all clear constitutional functions! I can think of several of the funders who would say that such decisions should be grounded in the best research available.

    This particular project might employ scientific methods but its stated goal is purely political.

    Close. Its stated goal is purely sociological. Political groups take advantage of sociology, but they're hardly the only ones.

    "Social pollution" and " hate speech" aren't objective terms.

    I imagine that you could actually get pretty good inter-coder consistency using a fairly small rubric for those two terms. Do you have a better measure of objectivity that that?

  27. Matthew says

    Outside of a clear constitutional function, such as national defense, the government has no business funding any "science".

    If you don't want the government funding research, then who *do* you want funding research? And under your model, how are people like Issac Newton and Albert Einstein supposed to get funding?

    "Mr. Einstein, if our company finances your research, how will it benefit financially?"
    "Well…my work on Relativity will form the mathematical foundation for a network of satellites forming a Global Positioning System."
    "When do you foresee this System of yours becoming functional?"
    "Er… the necessary advances in rocketry, engineering, mathematics, and computer science won't occur until decades after I die."
    "So Mr. Einstein, how will your research benefit our company in a time-frame relevant to our investors?"
    "…"

  28. CJK Fossman says

    Here's a great recipe for sweet irony:

    Write a blog post advancing the premise that the media generally get things wrong.

    Follow it with another based, in part, on a media report about the evil of a government-funded research program.

    Top it off with a number of comments by people who have read the media report but failed look at, let alone tried to understand, the subject matter.

    Because, you know, you can always trust the media.

  29. DRJlaw says

    @SIV
    "Outside of a clear constitutional function, such as national defense, the government has no business funding any 'science.'"

    Said no government of an industrialized country in modern times. Of course, you're welcome to give me an example of a government that's adopted that philosophy.

    The Federal government has the power to tax and spend to promote the general welfare. It has had that power virtually since the beginning (Article I sec. 8). We tried the Articles of Confederation war-and-diplomacy-and-precious-little-else government, it failed, and it's not coming back.

    And if you don't like that, then State government is going to give you a conniption fit. No more limited government of enumerated powers. Full-on soverign state with carve-outs here and there in favor of Federal powers, but spending is not a carve out. Do you have a public research university in your state? I bet you do. I bet that you have had one for quite a long time.

    So there's the legal side of "no business." I'll reserve the practical side of "no business" for something other than a provocative but entirely conclusory statement.

  30. Devil's Advocate says

    @Deniable Sources

    federal tax dollars (or rather, in 2014, obligations on our children's children)

    Fun fact: the vast majority of outstanding US debt matures within the next ten years.

  31. Cthippo says

    I have absolutely no problem with the study.

    If the end result is a better understanding of how plausible sounding* but entirely counter-factual things start spreading around the internet. If you read Politifact (which I highly encourage) you've seen them, the chain emails with "pants on fire" ratings that spring up and people pass along as truth. Sure, some of it is satire misinterpreted as truth, but some of it appears to be intentional attempts to spread false information. The university is performing essentially a traffic analysis to see if the pattern by which a meme spreads can identify how it started.

    As far as how my tax dollars are spent, I would much rather they go to this than, say, running an indefinite detention center in Gitmo or paying for someone to go to a religious school on a voucher or funding further tax cuts.

    *OK, some of it is plausible sounding, depending on what your pre-existing beliefs are, but most of it is just batshit crazy.

  32. Meaculpa says

    By my read, Ken's assertion isn't that the NSF researching twitter memes is diabolical, or even especially noteworthy beyond the fact that he makes note of it, but that the feds' priorities in funding research are terrible misaligned if on the one hand we can find up-to-the-minute data about trending topics on a terribly current medium, but on the other hand a legally mandated tool for providing data is ignored entirely.

  33. says

    Meaculpa gets it.

    It doesn't matter whether you find the study dull, scary, or silly. The point remains that the government can get involved in spending a million bucks to study memes on Twitter, but cannot get it together in 20 years to fund and advance a statutorily-required report to the American people about cops beating the shit out of them.

  34. En Passant says

    Meaculpa wrote on August 26, 2014 at 8:45 pm:

    … misaligned if on the one hand we can find up-to-the-minute data about trending topics on a terribly current medium, but on the other hand a legally mandated tool for providing data is ignored entirely.

    Ditto. Political paranoia about a diabolical plot with NSF to create a "Ministry of Truth" is entirely misplaced. The NSF project is interesting technically and sociologically, and might yield some insights ordinary people can use. But it's unlikely to become a useful government tool to further enslave the huddled masses yearning to be free.

    OTOH DOJ's failure to gather and publish information that it was mandated to publish should inspire angry mobs to storm a few castles with pitchforks and torches. But it's just business as usual. Nothing to see here. Move along, citizen.

    Oh look, a squirrel! Somebody just tweeted proof that Joe Biden is a Martian!

  35. Devil's Advocate says

    To be fair, if the DOJ did spend a million bucks studying police violence, the results would be pretty much indistinguishable from spending zero. Combining the cost of supplies, office space, travel (you do want them to travel to places where violence occurred, right?), and salaries, you could get maybe ten people working on it.

    There are roughly 18,000 police departments in the United States, which means that each employee would be in charge of 1,800 departments, so each employee would have to scrutinize about 7 department each workday. That doesn't strike me as a recipe for being thorough.

  36. sorrykb says

    [Dratted hamster ate my comment.]
    Ken White wrote:

    The point remains that the government can get involved in spending a million bucks to study memes on Twitter, but cannot get it together in 20 years to fund and advance a statutorily-required report to the American people about cops beating the shit out of them.

    Fair point.

    Maybe the NSF should take the lead on the excessive force study. Competitive awards process, public process, specific timeline for results, and more accountability than we ever get from DOJ. (And for once I'm being serious. Maybe this could work.)

  37. says

    @DRJLaw

    Sorry for the delay. Been swamped with life. :-(

    On the NSF study, I noted the conflict between my scientific side and libertarian side. The result is generally that great consideration need be given before allowing the government to fund scientific research. The politicization of science rarely has a good outcome. Political implications of research isn't a problem. It's the political use and abuse of that research that concerns me. When the government acts as a bridge or facilitator for worthwhile research that couldn't otherwise be accomplished, then I'm sanguine about it – having been involved in such myself back in my research days.

    On the second point, I certainly concede the point regarding the legality. My comments were statements of opinion, not of law. As you say; that pesky General Welfare clause…

  38. says

    @Devil's Advocate:

    The Constitution provides for the relationship between individuals and the federal government and the states and the federal government.

    Frankly, you can just as legitimately ask who will protect the states from the people? Considering the states on danger of bankruptcy, that may well be the better question.

    Beyond that, what's wrong with John Galt???

  39. Roy says

    Meaculpa gets it.

    It doesn't matter whether you find the study dull, scary, or silly. The point remains that the government can get involved in spending a million bucks to study memes on Twitter, but cannot get it together in 20 years to fund and advance a statutorily-required report to the American people about cops beating the shit out of them.

    Before I got sidetracked, that was the main thing I was trying to address. I don't think you can treat "the federal government" as a monolith here. One part of the federal government (the NSF) is doing what the law directs it to do, funding basic research. Another part, the DOJ, is NOT doing what the law mandates it to do, and I don't think it's a matter of "cannot get it together" but "doesn't want to, doesn't try". Priority doesn't come into it, it's not like doing less basic research funding will someone motivate the DOJ to do what the law says.

  40. CJK Fossman says

    @Raul Ybarra,

    You really mean this?

    Gov't funded research = bad
    Private entity research = good

    Like there's a magical wall between privately funded research and political abuse?

  41. Kratoklastes says

    @Devil's Advocate

    Fun fact: the vast majority of outstanding US debt matures within the next ten years.

    Weak sauce, Holmes: as the debt matures it is rolled over (and added to): it can be said with absolute confidence that the US Federal debt will be almost twice as large in ten years, as it is now. (That does not square with budget forecasts, but budget forecasts are lies – non-GAAP, zero-accrual bullshit from top to bottom).

    Your "fun fact" is designed to convey the impression that debt maturity is the same as debt repayment: it's not, by a long shot.

  42. Devil's Advocate says

    @Kratoklastes

    Fun fact: the vast majority of outstanding US debt matures within the next ten years.

    Weak sauce, Holmes: as the debt matures it is rolled over (and added to): it can be said with absolute confidence that the US Federal debt will be almost twice as large in ten years, as it is now. (That does not square with budget forecasts, but budget forecasts are lies – non-GAAP, zero-accrual bullshit from top to bottom).

    Your "fun fact" is designed to convey the impression that debt maturity is the same as debt repayment: it's not, by a long shot.

    It actually does change the nature of the beast. The popular misconception is that we've already obligated "our children's children", while in fact with a sufficiently draconian change in fiscal policy, we could be substantially debt-free in ten years. We couldn't do that if we had obligated "our children's children" without some extremely difficult or questionable tactics. You're right that it's likely that we'll continue the game of kick-the-can-down-the-road, but we don't have to, so it's politically dangerous to talk like there is no other option.

    What we currently do is different from obligating "our children's children" in another way: it's more risky because it relies on perpetually renewing our debt. We can't lock in our current low interest rates, and we're reliant on there always being a huge market for short term US debt. While there's currently such a huge demand for it that the demand exceeds the supply (see the low interest rates), we can't rely on it staying like that indefinitely.

Trackbacks