Democracy, Bureaucracy and the Fear of Statisticians

Often when I am trying to explain problems in the modern political landscape or my own approach to political philosophy, I will return to Max Weber’s account of bureaucracy as more efficient than private office. Yes, I’ve heard all the jokes about “efficiency” in bureaucracy, but Weber’s argument rested on the contrast between private and capricious office-holders and the public and publicly accountable form of governance that characterizes both state and business organization. Weber’s concern was that bureaucracies were too efficient, that their tremendous instrumental rationality obscured a real stupidity about the best ends to pursue. Weber’s theory of bureaucracy was vindicated in the way that Nazi Germany efficiently murdered its Jews, Rom, homosexuals, and communists, and ever since there’s been a tendency to get distracted by the Nazi example whenever the amoral efficiency of bureaucratic regimes are mentioned.

My own interest is in the tension between proceduralism and participation, but this strikes many people as odd and potentially pernicious, since some of my concerns about the administrative state are echoed by populists and the Tea Party activists. (I like to point out to my fellow philosophers that Glenn Beck’s writers have clearly been reading Giorgio Agamben.)

The thing is, I  think the most pressing kind of political philosophy, the research that really needs to be done right now, is a philosophical investigation of the contemporary formulation of bureaucratic governance. In short, it’s time to resurrect the tension between the predictive power of social scientists and human freedom. This is the stuff of mid-20th Century existentialism: Heidegger’s criticisms of technology, Sartre’s anxieties about freedom, and to-and-fro of structuralism and post-structuralism and the crisis and critique of the human sciences that runs through Adorno, Foucault, and Derrida. All these Continental philosophical debates largely occurred at just about the same time that Anglo-American philosophers had become obsessed with freedom and determinism, modal logic and counter-factuals, artificial intelligence and qualia, and the question of scientific realism. These questions, it seems to me, are all very much of a piece. Even as they presented themselves as valuable avenues of research and live debates of scholastic importance, they also captured a epochal anxiety about singularity and freedom, the global battle between communist “technocrats” and capitalist “risk-takers.”

I’m simplifying mightily, of course. But when you read with the question of bureaucracy in mind, it’s amazing how often it shows up in surprising places. Take my favorite thinker of the period, Hannah Arendt. In a 1964 lecture on “Cybernetics,” Arendt said:

When I grew up, it was still very common and very fashionable to believe that people who knew how to play chess very well were very intelligent indeed. If today we know that some kinds of these machines — I’m not going to say and names — can play a reasonably good game of chess, then I think it is a question of human dignity to say that this kind of intelligence apparently has not the same status as other kinds of intelligence, as other kinds of thinking. In other words, it is still something technical and it resides still in such a thing which we may accurately call brain power…. but it does not say anything about the level, or about the special particularities of this human being as such.

Is this really only about computers, or “human diginity”? The best way to think about her anxiety here is through the lens of two kinds of rationality: calculation and practical wisdom.

Many decisions are easily calculable in terms of costs and benefits, risks and probabilities. Because many very difficult decisions depend on the evaluating one’s own costs and benefits alongside the cost-benefit calculations of others, it can be tempting to think of these decisions as incalculable. After all, the mind of the Other is unknowable, just as the future is full of surprising and incalculable risks. Unfortunately, this temptation is dangerous. When the stakes are high, human beings tend to act in highly instrumental ways, and to adopt strategies that are easily calculable in the same way as chess moves would be. Because most moves are easily dominated, their real options are remarkably limited, and a well-programmed chess game can out-predict even the best grandmasters.

We’re all technocrats now, capitalists more than anybody. Consider the case of Bruce Bueno de Mesquita, who uses rational choice theory to predict the future. For $50,000, he’ll answer two questions, and many people would think the money well spent:

His first foray into forecasting controversy took place in 1984, when he published an article in PS, the flagship journal of the American Political Science Association, predicting who would succeed Iran’s ruling Ayatollah Khomeini upon his death. He had developed a rudimentary forecasting model that was different from anything anyone had seen before in that it was not designed around one particular foreign-policy problem, but could be applied to any international conflict. “It was the first attempt at a general mathematical model of international conflict,” he says. His model predicted that upon Khomeini’s death, an ayatollah named Hojatolislam Khamenei and an obscure junior cleric named Akbar Hashemi Rafsanjani would emerge to lead the country together. At the time, Rafsanjani was so little known that his name had yet to appear in the New York Times.

Even more improbably, Khomeini had already designated his successor, and it was neither Ayatollah Khamenei nor Rafsanjani. Khomeini’s stature among Iran’s ruling clerics made it inconceivable that they would defy their leader’s choice. At the APSA meeting subsequent to the article’s publication, Bueno de Mesquita was roundly denounced as a quack by the Iran experts—a charlatan peddling voodoo mathematics. “They said I was an idiot, basically. They said my work was evil, offensive, that it should be suppressed,” he recalls. “It was a very difficult time in my career.” Five years later, when Khomeini died, lo and behold, Iran’s fractious ruling clerics chose Ayatollah Khamenei and Hashemi Rafsanjani to jointly lead the country. At the next APSA meeting, the man who had been Bueno de Mesquita’s most vocal detractor raised his hand and publicly apologized to him.

Using game theory, mathematical modeling, and a panel of regional experts, Bueno de Mesquita can beat the best estimates of individual experts and the entire US intelligence community. His work has only one premise: “In the future, we’re still all raging dirtbags.” The CIA has claimed that his estimations are 90% accurate. Wow!

Rational choice theory promises us a world in which decisions are easily calculable because their results are precisely calibrated. The right choice would then be the one given by a calculation, not a decision taken by a free agent. We might wrangle over values and normative claims, but even these disputes can often be solved by making utility maximizing decisions that remove either-or decisions and make them both-and decisions: if we need not choose between our values, our pluralism can go unchallenged. That means that Bueno de Mesquita can resolve problems that would otherwise be unresolvable:

Recently, he’s applied his science to come up with some novel ideas on how to resolve the Israeli-Palestinian conflict. “In my view, it is a mistake to look for strategies that build mutual trust because it ain’t going to happen. Neither side has any reason to trust the other, for good reason,” he says. “Land for peace is an inherently flawed concept because it has a fundamental commitment problem. If I give you land on your promise of peace in the future, after you have the land, as the Israelis well know, it is very costly to take it back if you renege. You have an incentive to say, ‘You made a good step, it’s a gesture in the right direction, but I thought you were giving me more than this. I can’t give you peace just for this, it’s not enough.’ Conversely, if we have peace for land—you disarm, put down your weapons, and get rid of the threats to me and I will then give you the land—the reverse is true: I have no commitment to follow through. Once you’ve laid down your weapons, you have no threat.”

Bueno de Mesquita’s answer to this dilemma, which he discussed with the former Israeli prime minister and recently elected Labor leader Ehud Barak, is a formula that guarantees mutual incentives to cooperate. “In a peaceful world, what do the Palestinians anticipate will be their main source of economic viability? Tourism. This is what their own documents say. And, of course, the Israelis make a lot of money from tourism, and that revenue is very easy to track. As a starting point requiring no trust, no mutual cooperation, I would suggest that all tourist revenue be [divided by] a fixed formula based on the current population of the region, which is roughly 40 percent Palestinian, 60 percent Israeli. The money would go automatically to each side. Now, when there is violence, tourists don’t come. So the tourist revenue is automatically responsive to the level of violence on either side for both sides. You have an accounting firm that both sides agree to, you let the U.N. do it, whatever. It’s completely self-enforcing, it requires no cooperation except the initial agreement by the Israelis that they are going to turn this part of the revenue over, on a fixed formula based on population, to some international agency, and that’s that.”

A real crisis is one where sides cannot easily be chosen. If a formula always comes up with better decisions than I and my fellow citizens do, I think we’d all rather that the formula and its statistician-caretakers do the governing. That kind of proceduralism takes the average citizen out of the picture, or rather reduces her to a datapoint alongside others.

In a sense, the increased effectiveness of bureaucracies translates into reduced freedom for me. If political engagement is an important public good (and I believe it is) then bureaucratic predictiveness will lead to a maldistribution in that good: statisticians will get more of it, and regular folks less. For that reason, I’m really, really uncomfortable with predictive rational choice theory. It’s not  the resolutely self-interested view of the world Bueno de Mesquita advances that troubles me, the whole “raging dirtbag” shtick. It is, as Arendt put it, a “question of human dignity.” This is the very reason that Bueno de Mesquita refuses to handicap elections. The intersection of polling data and predictive technologies already contribute to a drastic narrowing of political outcomes in US elections. The capacity to know becomes the requirement to know.

In contrast, practical wisdom doesn’t depend on a person’s predictive powers. It allows one to surf that wave between the good and the possible that is characterized by intense risk and unforeseeability. For this reason, many philosophers see it as a corrective to our overly managed and predictable world. In exercising practical wisdom, I don’t want to act ‘rationally,’ according to a ratio or pre-ordained measure: I want to act wisely, i.e. with a view towards a Good that we can all only see in part. Those endowed with great practical wisdom ask: how can I act in a way that every model would call contrary to ‘self-interest,’ that can rocket us out of the one realm of calculation and into another? This will involve a great deal of reasoning, measuring, and calculating, but it should also entail a risk, a chance, fortuna out of which virtu can appear.

All the objections about incomplete information, lack of normative scope, and observer/participant problems pale before my anxiety that something essential about our freedom is lost when it can be reduced to a statistical abstraction that is nonetheless accurate. That’s why I love black swans, emergent events, and secret revolutions so very much. But that love would be pathological if it came at the expense of the well-being of the least-advantaged. I’m increasingly persuaded that attachment to civic engagement requires a non-epistemic justification, and that we must temper our love of administrative governance with the cautions of mid-20th Century existential phenomenologists like Hannah Arendt.


9 responses to “Democracy, Bureaucracy and the Fear of Statisticians”

  1. Andrew Avatar

    I really liked this piece: it crystallized several things that I’ve been thinking about lately. This is my fourth semester teaching Master’s students in Social Work, and the class I’ve been teaching that whole time is a basic introductory theory class. Rational choice theories show up, as does social constructionism, which I had to try and “teach” in about two hours last week. (This week I’ll cover conflict theory, so you can see why I might have been excited by your post.) It’s a basic survey class, but it’s also the first class that students take in social work school, and I’m pretty aware that several of the ways that problems are framed, social work constituencies are defined, and theories are arrayed will likely be ongoing issues that the students have to deal with during the rest of their education. The first two, the way that problems are framed in the social sciences and the way that clients are described, are interconnected and have very large implications for social work practice.

    Or they don’t. There’s a fair amount of talk about “practice wisdom” among the other professors, but mostly from those whose favored theories are in decline. The cognitive/behavioral theory professor never mentions it, and seems the most cocky of those that I interact with. It’s hard to cultivate the proper balance between introducing the material and also introducing the concept of practice wisdom. It helps that the predictive value of most of the theories social work has taken up is very low.

    I’m curious about how you’ve gone at the tension between “the predictive power of social scientists and human freedom.” I like your concern about what we do when things get too controlled by “known” outcomes, and I’m embarrassed to say I’d never heard of Bueno de Mesquita and I’m fascinated, but in my experience the larger tension is between the misplaced faith in social scientists’ predictive power and human freedom. Certainly in most practical social sciences, like social work or teaching or public health, the ability of individual practitioners to choose their modality is disappearing, replaced by methods that claim far more validity than they have (in practice, or in research). This seems to me to be the problem with NCLB, or with the push toward “evidence based practice” in social work or public health. The research models are so poor that quite stringent criteria get adopted without good rationale. It’s positivism without even the positive. I understand that your argument is partly built around the (currently) counter-intuitive notion technocrats might make us less free even as they solve very large problems that would contribute to the greater good, but I’m not sure we’re yet living in a world where the main danger from technocrats is that we’ll get too much of a good thing.

  2. Andrew Avatar

    Or, to put it another way, is there a difference between arguing for more freedom when the social scientists are wrong, and arguing for freedom when they're right?

  3. Joshua Avatar

    Hi Andrew: I think the tension you're describing is very different than the one I'm thinking about, but that the difference might actually be helpful for both of us. You hit it well in your last question.

    Let me start with the case for evidence-based practice: time and time again, you've pointed me to studies that demonstrate the evidence-based case against the psychiatric/medical approach to mental health or in favor of pluralism for treatment modalities. If technocrats aren't responding to those studies or are wrongly interpreting the evidence (like the CBT craze apparently generated by the number and ease of studying CBT treatment regimes) then that's the fault of the technocrats and probably best falls under the kind of Foucaultian critique in The Birth of the Clinic and his lectures on governmentality: the doctor's visualization trumps other interests, experiments masquerade as treatment, and testability and statistical representation stands in for other kinds of knowing. We might say of the trends you describe that we're witnessing an analogue of early medicine, where the desire to know in certain ways trumped patient welfare: a little learning is a dangerous thing. But if you turn your attention away from mental health, doesn't the evidence-based model appear more attractive? Consider Great Britain's concerns about homeopathy and acupuncture…. If so, then what changes in mental health?

    I should probably write separately about the triumph of bureaucracy in health and mental hygiene after Foucault, however: I worry that Foucault's criticisms largely fail to reach a non-welfare state provision of medical services like the one most people face in the US. There ought to be a way to discuss the recent health care reform, for instance, in terms of the privatisation of normative knowledge production and the economization of the disciplinary/control society, right? Though I can't quite put my finger on the issue, I just don't feel comfortable with following Foucault any longer there: I think it sounds far too much like Glenn Beck. I guess Foucault (like Arendt and Agamben) is too easily appropriated for a bad form of libertarianism.

    Another way to get at this is simply to argue that rational choice and game theory don't help us with individuals or with therapeutic practice. This is just the corollary of the real constraints for predictive rational choice/game theory modeling, where you need stake-holders with visible preference structures contesting instrumentally over high stakes. So we can make highly accurate aggregate evaluations without reaching individuals, but *then* the temptation is to point to black swans, emergent events, and secret revolutions!

    Perhaps we should also note that formal choice theory rarely helps much in evaluating the difficulties of the very disadvantaged or very well-off, except insofar as it helps us to see that the mentally ill or those who suffer from addictions, like those with more money than they know what to do with, must have very different preference/incentive structures than non-sufferers and the non-wealthy do. We could contrast this with the remarkably incentive-responsive (in aggregate) behavior of the middle-class. (I'm thinking here of Charles Karelis's work in _The Persistence of Poverty_ on certain "relievers" like drugs and alcohol.)

    In this sense, however, I've just been arguing for "good technocracy" instead of "bad technocracy." One way to avoid the tension is simply to argue that all good technocrats must depend on ordinary citizens to supply data about their preferences and needs, while bad technocrats ignore this data in favor of some theoretical abstraction even when the abstraction is wrong. But this still assumes a power/knowledge privilege in favor of technocrats, somehow understood. If our opinions don't make much of a difference (because we're subordinated to experts) we're increasingly likely to vote or report preferences only to register affiliations, as is sometimes suggested both Kansas and coastal elites do. The non-epistemic case for civic engagement still has to be made.

  4. Andrew Avatar

    But this still assumes a power/knowledge privilege in favor of technocrats, somehow understood. If our opinions don’t make much of a difference (because we’re subordinated to experts) we’re increasingly likely to vote or report preferences only to register affiliations, as is sometimes suggested both Kansas and coastal elites do. The non-epistemic case for civic engagement still has to be made.

    Well, I agree, but it seems to me that while you've made things more difficult in one respect by talking about good technocrats (in the sense that we're forced to confront why civic engagement might trump some raw measure of utility or appeal to rational choice), you've made your job easier by limiting the problem that way.

    I think part of my issue is that it's frequently quite hard to determine when a technocratic social intervention has worked. Your example is clear and useful, but in other realms (education, public health, social welfare), results are often ambiguous and contested. You suggest this yourself, but, in the public sphere, there are all sorts of reasons why that type of problematic experience is more prevalent than the perfect expert one. Rather than expert consensus, one is frequently faced with dueling experts. And, I would argue, this drives engagement toward affiliation registering. Interest groups recruit social science to their positions, rather than the other way around. This leads to the same problems of un-engagement that you spoke of, or, it leads to a non-useful type of participation. This is related to, but different than, the problem of perfect expertise, in that it breeds both distrust of, and reliance upon, imperfect expertise. And then we have a failure to discuss public policy, or possible policy choices, as a matter of preference, or conviction, or morality, because competing experts are made to carry on the discussion in the stead of people. We end up with a non-rational, rational choice model.

    But I keep having the uncomfortable feeling that I’m missing your important points, and getting hung up on my own concerns. If that’s so, I guess I need something more in order to understand the distinction you’re making. (And perhaps a remedial class.)

    1. Joshua Avatar

      I've been trying to think through this concern "that it's frequently quite hard to determine when a technocratic social intervention has worked."

      I'm not sure we're in agreement on that, but we'd really need to work through some cases and examples. One version of this criticism is associated with what I'd call "agonistic politics," popular with radical democrats and cultural conservatives. But there's a relativism in agonistic politics that I don't accept. I guess it really depends on whether your concern is actually epistemic ("Can we know if we've succeeded?") or metaphysical ("Is there a fact of the matter about success, whether or not we can easily access it?") Your further concerns about factions and experts driven by ideology rather than inquiry is a good warning, but I think it's a concern that can easily be overstated.

      I'm interested to try to hash this out, though, so perhaps you could say a bit more?

  5. Andrew Avatar

    Sorry, I didn't mean to abandon this discussion. I got caught up in some other, non-internet-related things. (I think some people call that real life, in a dismissive sort of way.)

    I think part of my lag in getting back to this has simply been my confusion. I'm not sure where we agree and where we disagree, or perhaps I'm simply not sure I understand your argument. Your original post seemed to be concerned with the idea that as technocrats become truly expert, there might be no role for public political participation in the process of deciding on approaches to given problems. You called for a "non-epistemic" basis for civic engagement. I tried to argue, in contrast, that I'm not convinced that we're in danger from too precise technocratic expertise, and, further, that the struggles for power having to do with technocracy have more to do with competing decisions about who actually is an expert. The very fact of this competition suggests that the kind of omniscient expertise you're concerned about is not yet a problem.

    I agree with your overall call for a civic engagement predicated on something beyond the solution to particular social problems. But I also think that the philosophical problem presented by the available solution to particular social problems is different than (at least) I understand you to be arguing. In part this may be because your example in Bruce Bueno De Mesquita is so attractive, but remains unconvincing to me as evidence of a larger problem. Again, I worry I'm not quite getting it, but I find that people simply ignore experts they disagree with, and that they often use their civic engagement to do so with extreme prejudice. Creationists have no truck (indeed, ~50% of Americans have no concern) with the expertise of evolutionary biologists, and are frequently engaged in politics precisely to the extent that they are non-plussed by science. On today's front page of the Post, we're treated to yet another story about Michelle Rhee (although I understand that the bona fides of her expertise is in doubt), this one directly affecting DC democracy, where private foundations are seeking to control the governance of the DC school system. I just don't see the overweening expertise.

    I'm troubled, too, by strict agonistics in politics, but I'm not sure my arguments lead necessarily to only an agonistic model.

    1. Joshua Avatar

      Well, thanks for coming back to discuss it further. I think you've given a good summation of the issues. Let me try to tease out my concern in terms of the examples you gave.

      To start, I think we agree that ordinary citizens have high rates of ignorance of issues that are essential for self-governance. As a result, many decisions made in a democratic manner would benefit from a better informed citizenry. But given the ignorance and irrationality of the median voter, it's tempting to prefer better technocrats to better education: the marginal return on small increases in competence in the administrative state is higher than it is for much larger increases in voter competence. In your example, it seems better to have informed voices on the Texas Board of Education than it does to have informed voters, and we might sacrifice a million informed voters for just two Texas Boardmembers.

      So: there is institutional pressure to take decisions out of the hands of voters and put them in the hands of technocrats, where they can be monitored and improved. In cases where public health, economic growth, or matters of fundamental justice are at stake, experts promise to supply better outcomes. There actually are better and worse policies to reduce infant mortality, there actually are better and worse policies to get the US out of its recession, and there actually are better and worse ways to reduce social inequality. That's a pretty strong incentive to side with the technocrats, in-field disagreements among experts end up being fairly minor compared to their disagreements with non-experts. (There's much less difference between Charles Murray and Stephen Jay Gould than there is between the biologists and Creationists.)

      This produces feelings of coercion and domination in voters, who feel that experts have formed a tyrannical coalition in order to ignore their (allegedly less informed) views. I take strict agonism such as that of Chantal Mouffe to side with the disenfranchised in this situation: experts really are asserting an exclusive and undemocratic privilege when they insist that the correct ought to trump the popular.

      I'm having a harder time parsing the private foundation demands regarding Michelle Rhee, but let me take a shot. Part of the problem is that fixing schools is not something there's much expertise on. Educational inequity is still almost completely a mystery, insofar as we can't seem to predictably generate dependable results, with the exception that rich kids do well. A bunch of unaccountable private foundations believe they have met an expert in this area, Rhee, who will help them achieve a desirable goal. I take it that these foundations agree that the key to a good education is a good teacher, and that union protections unduly provide protections that prevent Rhee from removing bad teachers. Or that the paucity of the data requires us to engage in controlled experimentation that's only possible in charter schools. Or that if outcomes can't be improved then at least we can lower costs. (I don't really know, because Watson and Swinburne don't seem to feel that they owe us their reasons. We can infer possibilities from Rhee's actions, however.)

      But note that in the absence of expertise in school reform, there's really only a general sense of budget management and executive capacity at stake: a strong negotiator or an incorruptible leader who can at least perform as an expert on the administrative side by fixing the incentive structure, lowering costs, rooting out placeholders and clockwatchers, and generally getting more data to inform future expertise.

      But education is also a good space for this civic engagement discussion, because one thing that correlates highly with good schools is parental and community involvement in governance and as volunteers. So if you cut back on the technocracy, you can get better results on at least one technocratic measurement… even if the parents demand intelligent design be taught in biology class or want Thomas Jefferson removed from the list of Founding Fathers. Hell, there's even some limited evidence that participatory budgeting is an effective model, because citizens who participate in making a budget are better attuned to corruption and waste than the professional auditors! But all of this is the kind of limited scope data that may well be ineffective or untranslatable outside of the context where it is gathered. So I'm not so sure that the civic engagement agenda is necessarily just the technocratic winner.

      Plus, even if this kind of participation helps in local governance, it seems like it hurts in global governance. If the question is health care reform or the Federal Funds discount rate or the amount of money to spend on economic stimulus or defense contracts or renewable energy, civic engagement breaks down and we get competing populists fighting for attention using messages that fit on a bumper sticker or can be shoe-horned into a sound bite. So we're back to preferring effective disenfranchisement through governance by career bureaucrats who don't change policies when new politicians are elected.

    2. Joshua Avatar

      Peter Levine has a bit more on the kinds of debates that really trouble claims to expertise in school reform here:

      The main debates concern incentives or pedagogy. That is, the two main strategies for improving schools are to change the rewards and punishments, or else to convince/educate teachers to act differently.

      And in fact there are probably similarly deep and provocative debates in climate science about solutions (carbon tax, cap and trade, geoengineering) even if the facts are less disputed.

  6. […] the humanities even when they are massively popular and profitable. As I’ve been trying to argue, this was always entailed by the growing rent-seeking class of administrators who we have […]

Second Opinions