Digressions&Impressions

Riker on The Calculus of Consent; some Rawls and Strauss and the confused and unproductive state of Political Theory

As numerous writers have pointed out, political theory as a field of academic concentration has been in a confused and unproductive state for at least the last generation. Most of the academicians who have written about it have been either historians of ideas who speculated about the influence of writer A on writer B or literary critics who simply explicated and criticized the texts of A and B. Very few academicians have had the intellectual courage to play the role of A and B themselves; so that, in consequence, most contemporary political theory has been produced by politicians and journalists. Refusing thus to theorize themselves, most academicians have been limited to a consideration of intellectual primitives or of, what amounts to the same thing, the rantings and ideological tergiversations of half-educated modern pamphleteers. In almost no other field of study do the professionals thus abase themselves before the ancients and the amateurs.

    Not unsurprisingly, there has been much dissatisfaction with this state of affairs and in the last decade or so a number of writers have tried to reorient the study of political theory. Some, like Leo Strauss, have urged a return to the great tradition; but unfortunately the process of return, as it appears in the work of both master and students, has turned out to be an even more sterile historicism than that against which they revolted. Others, like T. D. Weldon and Carl Friedrich, have urged a close and systematic study of words; but again unfortunately, word-study, while a useful preliminary to theorizing, has produced no more than literary criticism and is not properly theory at all. Still others, the political behaviorists, have turned to the theories of contemporary psychology and sociology for a new start in theorizing about politics; and, while these writers have undoubtedly succeeded in generating descriptive theories that can actually be tested empirically, still their theories seem to have more in common with social psychology than with the traditional subject matter of politics. Furthermore, their positivistic character, wholly appropriate for descriptive theories, is far from satisfying to those who believe political theory should be at least partially normative. These three attempts at re-orientation have been useful in the study of history, philosophy, and behavioral science; but they have not added much to political theory itself. Indeed, if these three schools were all we had, it would be fair to say that political theory was no better off than before.

    There is still a fourth school, to which the book here reviewed belongs, which offers some hope of a really successful reorientation of political theory. Since its inspiration is more from economics than any other source and since economics is the social science closest to politics both in the nature of the events abstracted for study and in the process of abstraction, it happens that the works of this school are clearly political in nature. Since economics like politics has a long tradition of normative theorizing, these writings contain normative elements alongside of positive ones. Most important of all, however, they seem to be capable not only of reformulating traditional problems in a more sophisticated way but also of dealing with problems beyond the ken of even the great theorists of the past. Some of the main works in this fourth school of re-orienters of political theory are: Duncan Black, The Theory of Committees and Elections; Anthony Downs, An Economic Theory of Democracy; Morton Kaplan, System and Process in International Politics (although this work has its roots as much in sociological as in economic theory); Thomas Schelling, The Strategy of Conflict; and, as an important addition, James Buchanan and Gordon Tullock, The Calculus of Consent.--William H. Riker (1962) Midwest Journal of Political Science, Vol. 6, No. 4 (Nov., 1962) [Reviewing The Calculus of Consent. by James M. Buchanan and Gordon Tullock], pp. 408-409.

Because I am working on a side project (recall here; here) on the (non-)reception of Leo Strauss within analytic philosophy, I stumbled on this passage in Riker's review of The Calculus of Consent. I also want to situate the passage in the black hole in our collective narrative of what exactly political philosophy could be within analytic philosophy and its sister-discipline political theory pre-Rawls (which was re-opened for me by Katrina Forrester's In the Shadow of Justice: Postwar Liberalism and the Remaking of Political Philosophy (recall (recall here; here; here; and here). Somewhat to my own surprise these issues intersect with my interest in neoliberalism and the split between ('positive') economics and ('normative') philosophy.

I suspect most of my readers know who William Riker is (not to be confused with the Star Trek character). He is now widely seen as one of the founders of rational choice theory, and among other achievements introduced game theory and social choice theory into political science. He is, what in an earlier generation disparagingly would have been called (by the sociologist Talcott Parsons [recall]), an exponent of 'economic imperialism.' We call it 'intellectual arbitrage' now. He founded the so-called Rochester School that's associated with the 'behavioral revolution' in political science. When he published the review of Buchanan and Tullock, Riker had just published The Theory of Political Coalitions, which was a major contribution to orientation he promoted.

About half of the review is quoted above. And only after this set-up does Riker discuss the contents of the book. So, the review itself is self-consciously a methodological intervention and cannon creation. 

So, Riker distinguishes four approaches to political theory. First, there is the return to the classics which he (correctly) associates with Leo Strauss. This is treated -- much as Bacon treated final causes -- as sterile. Riker correctly represents Strauss's school as critical of historicism, but then misrepresents, either from confusion or from sincere insult, Strauss as instantiating historicism. (I am unsure he ever engaged with Strauss in substantive fashion elsewhere.)

Second there is a school devoted to "a close and systematic study of words." Riker's criticism of it echoes Isaiah Berlin's (1958) criticism of ordinary language philosophy (recall). T.D Weldon I know primarily as a Kant scholar and as being the supervisor of Wilfrid Sellars. But he is also the author of The Vocabulary of Politics (1953), which in reviews is described as an attempt to apply ordinary language philosophy to politics. A book I, thus, have to read because it is clearly in the event horizon of the black hole of analytic political philosophy whose shadow Rawls has (ahh) blotted out.*

I have to admit that I am puzzled by Riker's inclusion of Friedrich in this category. Friedrich's approach is so eclectic, that I would think he would be hard to categorize. This is not to deny he sometimes focuses on words, too. But this comment (alongside the Strauss as historicist 'mistake') makes me wonder, if Riker actually had carefully studied the schools he is rejecting. Or whether his polemics derive from supreme confidence in his own (sophisticated!) methods. That confidence is not without robust intellectual foundation, but one does wonder if in his polemics with opposing schools there isn't an element of bluster and bluff. (This combination is not unfamiliar to us within analytic philosophy.) I actually suspect that Riker is confusing Friedrich with Felix Oppenheim (who can, perhaps somewhat unfairly, be slotted into this category.)

No exemplar of the third school is mentioned. I suspect Talcott Parsons (or those influenced by him) is one of the targets. This school is also engaged in intellectual arbitrage from other social sciences, but Riker finds them wanting. Riker's two criticisms of this approach (not very far removed from his own) are very interesting: first, they produce work that fits social psychology but is in some sense not properly political. One would have liked an extra sentence here because it would have illuminated his own stance toward the subject (or the discipline). Second, on Riker's view political theory ought to be descriptive/empirical (or at least fruitful in generating theoretically salient empirical generations and be an engine for theorizing) and normative/positive. So, the right kind of theory can help us study the political world and tell us how to improve it. (I return to this in the next paragraph.)

These two criticisms do not apply to the fourth school (which is Riker's own). From my perspective what is most interesting here is that Riker rejects the very firm distinction entrenched within economics (in word, less so in deed) since the time of Marshall, Sidgwick, and John Neville Keynes, and articulated by Lionel Robbins (and Milton Friedman) between the positive and normative projects.+ The final paragraph of the review actually explains Riker's own stance on this:

If it is assumed that most men are rational and that the appropriate rational behavior is transparently obvious, then the analysis is entirely positive. If it is assumed that rationality is a goal rather than a fact, then the calculus is normative. If it is assumed that the logic of the calculus is not obvious until it is deduced, then again the analysis is normative. (411)

Of course, the problem here is not the willingness to combine empirical and normative aims, but (to use scientistic terminology) that how to think about circumstances of what is 'transparently obvious' to the theorist/expert may not be so to (ahh) 'most men' and, more important, if, in a political context, the latter have to rely on the former for the deductions in the 'logic.' [As a slogan: Riker takes the relationship between theorist and theorized for granted.]

Be that as it may, lurking in the final paragraph that I quoted at the top of the post is the sense that the fourth school is generating a scientific revolution: "Most important of all...they seem to be capable not only of reformulating traditional problems in a more sophisticated way but also of dealing with problems beyond the ken of even the great theorists of the past." The fourth school has moved beyond the past and is progressive. (Such claims are common among economists of the age.) Interestingly enough the Nobel committee in economics has also recognized the contributions of this school of 'political theory' (including, in addition to Buchanan, Schelling).

After the polemical first half, the rest of the review of Calculus of Consent "list[s] what are for [Riker] some of the most interesting notions in the essay." (409) One important point is that Riker treats Calculus of Consent (not unfairly) as an advance beyond Downs on the question of what the effect is "of differing intensities of concern about particular issues under the rule of simple majority voting." (410)

From the vantage point of thinking about the nature and evolution of political theory at the middle of the twentieth century, the first, main question is actually rather important:

The main question the authors ask is: On what basis does a rational man, situated in a society created by unanimous agreement, consent to the transfer of a particular activity from private to collective decision-making? This question is, of course, a new and more sophisticated form of the question to which the theory of the social contract provided the answer. (409)

In the spirit of progress (we may say), Riker understands Buchanan and Tullock as having surpassed and made obsolete social contract theory. While the historian in me hesitates to disagree with such a well-informed contemporary of them,  Riker somehow misses (presumably for his own polemical purposes) that Buchanan and Tullock themselves present their own work, en passant (but not obliquely) as a contribution to social contract theory. For, throughout Calculus of Consent, they imply both that individuals cannot withdraw from the social contract, as well as (and  more important) that "Insofar as participation in the organization of a community, a State, is mutually advantageous to all parties, the formation of a “social contract” on the basis of unanimous agreement becomes possible." (ch. 17; the whole chapter is relevant, including for understanding why they frequently use scare-quotes around 'social contract,' but not always.)** So, the main question by Riker's lights is answered in terms of a social contract analysis according to Buchanan and Tullock. (The character of their social contract is worth exploring in a digression some time.)

Two final quick promissory thoughts connected to Rawls, since I click-baited you here with Rawls. First, there was a pre-Rawlsian political analytic philosophy. (Riker's second school.) Weldon is an example of this. (So is, perhaps, Felix Oppenheim.)++  And it was noticed by outsiders. But Rawls did not respond to it in The Theory of Justice.*** Oppenheim is cited once respectfully in TJ, and Weldon isn't mentioned. Rather, Rawls took the fourth school, the re-orienters, as his interlocuters and rivals. In fact, in TJ, Rawls also cites Black, Downs, Buchanan and Tullock, and Mancur Olson as authorities. And whatever else he is doing, he is also trying to offer an alternative to their project.+++ That's partially obscured to later readers because he doesn't name them much as targets. But it is worth noting they inherited their own normative project from utilitarianism, which is Rawls' official target. And partially obscured by the fact that after the mid to late 1970s Rawls himself emphasized the normative project in Kantian terms and so it's much harder to see TJ as building on and in competition with Rochester and Virginia.

 

 

On Leaving Unpublished Papers on the Bus (Abba Lerner)

In 1932, Lerner received his bachelor's degree from the L.S.E. and published his first paper. The following year, he and a group of colleagues founded The Review of Economic Studies. Impatience with articles long in form and light in content was a prime motivation for the creation of a professional journal whose success has now extended over more than half a century. Also in the early to mid-1930s he played a leading role in organizing a London-Oxford-Cambridge Joint Seminar for younger economists in the three universities--a rare example of collaborative effort among them. By the time he completed his graduate studies in 1934, he was recognized as one of the future leaders of his profession with eight major articles to his credit, and the award of a prestigious Post Graduate Fellowship that permitted him to spend the year 1934-35 in Cambridge and in Manchester. But even more remarkable than his eight publications is a paper written in 1933 and unpublished for 17 years, stating and proving the international factor price equalization theorem. The result, re-discovered in 1948, became one of Nobel Laureate Paul A. Samuelson's most celebrated contributions to international equilibrium theory.

Lerner's first writings showed several of the main characteristics of his prolific output during a period of almost 50 years: originality and lucidity of thought, allied with an exceptional gift for exposition. They also displayed his virtuosity at using simple geometric arguments in economic analysis.--G. Debreu B. Hansen C.J. Hitch

A few days ago, I mentioned Abba Lerner (who is being discovered because of political debates over MMT) in passing, and quietly linked to the quoted obituary. Not unlike his ideological opposite, Gordon Tullock, Lerner is relatively easy to read and the (relative) simplicity of his prose hides profundity. No less an authority than Paul Samuelson (1964) pointed out that 'Papers like those of Lerner's are so packed with results that few readers have ever gleaned all their fruits" (quoted by Landes). 

When I started to write this post I intended to comment on Debreu's (to me surprising) appreciation of Lerner. But I got reminded of the great story around Lerner's unpublished seminar paper. The story is told with more detail by Landes, and I quote him first:

Samuelson's demonstration found its way to the desk of Lionel Robbins, who recalled that he had heard the same argument from a student in seminar some fifteen years earlier; and he still had a copy of that seminar paper, by Abba Lerner. At Robbins's urging, Lerner published the piece as originally written: "Factor Prices and International Trade," in Economica in 1952. The question remains, what had Lerner done with his copy of the paper during all those years?

The story, as told by Scitovsky, is that it had been lost; that Lerner had given the corrected typescript to a fellow student, who had volunteered to type it for submission to a periodical and then left it on a bus. It was never recovered, and Lerner was too busy working on other papers and perhaps too embarrassed by the circumstances to recover the text. At least that was the talk among Lerner's students in 1935, when Scitovsky was one of them.

Landes's account is worth reading not the least because he hints at the genteel British, academic antisemitism that prevented Lerner from obtaining a position. Be that as it may, Robbins was a central player in economics in the 1930s (and I have discussed him (recall) before not only because of his impact on Rawls), and LSE one of the key nodes of the profession. That he was no dummy is shown by the fact that he remembered Lerner's paper fifteen years later when he encountered Samuelson's paper. (I often don't even remember my own papers that long ago!) But it is notable that he seems to have missed its original significance (even if we can speculate that he may have been the one to nudge Lerner to submit the paper "to a periodical.") 

Of course, Robbins fastidiousness in keeping seminar papers also saved the day. As I contemplate the story of the missing paper, I was struck by the fact that the fellow student, the culprit, never gets named. I think he (she?) would make a fine character in a short story, or (recall) histories of walk-in characters in the history of science.

As regular readers know I am fascinated by ways in which purported epistemic peers can miss the significance of a central contribution to the field (recall this episode on Bennett). Samuelson offers a generous footnote to Lerner's (then still unpublished piece) in his famous article, and he clearly suggests that he recognizes its importance (calling it "masterly, definitive treatment of the question, difficulties.") Samuelson was, at the time, redefining economics. Some other time I return to these papers (which connect nicely to my reading of Adam Smith).

I close on a minor key. Landes suggests that even Lerner "himself was not always aware of what he had done." This raises the interesting question to what degree one can fail to discern the significance of one's own contribution.While, of course, it's quite possible one fails to foresee every downstream implication of one's contribution, I am, nevertheless a bit suspicious of this idea; it lets us, lesser lights (who miss stuff all the time), off the hook a bit too easily. I prefer to think Lerner was simply busy tackling the next big problem.

 

 

 

Should Policy Relevant Researchers Stay Out of Politics?

To lock yourself up in an ivory tower is impossible and undesirable. To yield subjectively, not merely to a party machine, but even to a group ideology, is to destroy yourself as a writer. We feel this dilemma to be a painful one, because we see the need of engaging in politics while also seeing what a dirty, degrading business it is.--Orwell (1948) "Writers and Leviathan."

Liam Kofi Bright ignited discussion on Facebook of a (fairly) recent paper by (BHLer) Bas van der Vossen. This paper argues for a normative conception of the division of labor, between "consumers," that is, "activists," and producers, that is, political philosophers (19), of political philosophy such that pure specialization between them should be maintained:

for precisely those academics that work on politically relevant topics, most prominently among them political philosophers. For them, the university should become more like an Ivory Tower, not less.... the problem with these [political] activities is that they encourage us to think about ourselves in partisan terms. And this is incompatible with our academic professional responsibilities.--Bas van der Vossen (2015) "In Defense of the Ivory Tower: Why Philosophers Should Stay Out of Politics" (1-2)

Before I get to his main argument, I want to note, first, that throughout his paper Van der Vossen equivocates among (i) 'working on politically relevant topics' and (ii) "those who are serious about thinking through political issues" and (iii) being "paid to think about politics." (On (iii) he notes that political philosophers do not have a monopoly and he also lists "sociology, political science,  economics, gender studies, psychology." (2)) By this I do not mean that (i-iii) are not synonymous in the strict sense (they are not). Rather, there are quite a few sciences (genetics, climate-science, epidemiology, civil engineering, IT, etc.) that are politically extremely relevant, but that are not, in the first instance, about thinking about political issues (although, interestingly enough, it does happen).  Because Van der Vossen ignores those latter sciences he can escape having to think about difficult, hybrid contexts where the distinction between consumer/producer is not so easy to draw. In what follows, where necessary, I distinguish between policy relevant research and research about politics and political issues.

Second, Van der Vossen's paper presupposes non-trivial commitments about the good life: "Many people stay out of political activism and they do just fine. Activism is not a necessary ingredient of a good life." (18) One need not be an elitist, a Republican-thinker, or a follower of Hannah Arendt to see that appealing to what "many people" do and that they are doing fine is an odd standard. The paper is also oddly complacent about the status quo (b) "many people in even the healthiest democracies around the world are not politically active and, in the grand scheme of things, these democracies seem to do just fine." (19) Even if this were true (when the paper was published in 2015 [i would reject this]) a cosmopolitan, political philosopher may take an interest in the health of transnational political subjects and promote these via transnational activism.

Third, and this is more important, Van der Vossen takes for granted that the bridge between consumers and producers of political philosophy can be bridged without the producers being politically active. His is a magical theory of dissemination; if you produce it, it will be  consumed. This may be true for (lots of) highways, but it is rarely true otherwise. While there are many different ways in which the gap can be bridged between policy relevant research and politics, a key way is by way of what Merel Lefevere and I call aggregators. Aggregators are distinguished researchers, whose professional function is, in part, to act as an interface between research and policy. There are many different kinds of aggregators (journal editors, science policy-advisors, science-journalists, etc.) and not all of them maintain active research profile or need to do so. But because cutting edge research is often extremely complex/subtle it is important that some of the  best researchers are also aggregators. In practice, these researchers-aggregators often have high prestige within the field (think of Singer, Pogge [sic], Nussbaum, Baroness O'Neill of Bengarve, Norman Daniels, etc.). (This is not the place to explore all the tricky issues pertaining to this class of researcher-aggregator, especially if they deviate from professional consensus and advocate their own views, but recall this post.) In some grant-driven environments, researchers are, in fact, expected by the grant agency to play some role as an aggregator or actively disseminate their research.

That he misses the very issue becomes clear when we turn to his argument (which he has helpfully summarized):

(1) People who take up a certain role or profession thereby acquire a prima facie moral duty to  make a reasonable effort to avoid those things that predictably make them worse at their tasks

(2) The task of political philosophers is to seek the truth about political issues

(3) Therefore, political philosophers have a prima facie moral duty to make a reasonable effort to avoid those things that predictably make them worse at seeking the truth about political  issues

(4) Being politically active predictably makes us worse at seeking the truth about political issues

(5) Therefore, political philosophers have a prima facie moral duty to avoid being politically active (17-18)

Aggregators clearly violate the  conclusion of the argument, But one could accept all the premises of the argument, even its conclusions, and still see a need for aggregators in practice. That is, (5) is at best a defeasible duty. It becomes defeasible for two interesting, interdependent reasons: (a) research is esoteric, and so requires expertise to navigate; (b) political activists lack time to explore and master an esoteric realm.

In addition, it is, of course, also an empirical question if aggregators really become worse at their research. (This is not to deny that Van der Vossen draws on rich empirical research on the effects of partisanship and also about motivated reasoning and rationalization--including work by Dan Sperber that I happen to be reading this week, too.) I bet the most admired 20th century professional economists -- by other economists -- were Samuelson, Arrow, Friedman, and Keynes.* While the first two cultivated a decidedly technocratic image, it's hard to say they stayed out of politics--they certainly functioned as aggregators in the sense I use the word. Would they have been better at their craft if they stayed clear of policy relevance? That's hard to imagine. Their biases animate their research. Would they have been even better researchers if they stayed clear of policy and the role of aggregator? That's a tricky counterfactual. I am not confident it can be answered.

So, far I have pretended to accept  all his premises. Regular readers, who may discern a skeptical strain in these Digressions, may suspect that I would hesitate at (2). But, for the sake of argument, I accept it. As it happens, I think there are also aesthetic goals of research that are not properly captured by 'truth.' (Not unlike Marcus Arvan I think there are also excellent reasons to think that there are other very respectable, epistemic and moral goals of research that may have  to be balanced.)

But I close with a final objection that is directed at the way Van der Vossen cashes out premises (3-4). For, he thinks that biased research is incompatible with the pursued of truth. That is, he conflates unbiased (and partisan) research with truth-apt research.**  As the previous paragraphs make clear individual researchers are part of a larger epistemic process in which motivated reasoning may well have some use. In fact, this theme is a feature, not a bug, of the law, which is an adversarial process in which the conflict between motivated reasoning, advocacy, helps uncover the truth. Obviously, academic research is not fully analogous with that (and being lawyer-ly is, in an important sense, inimical to a proper, collaborative research ethos). But if we think of research as a community-activity, then we can immediate see that's compatible with individual bias--as long as these individual biases contribute to a social, epistemic process that is truth-conducive. (Regular readers will recognize in the previous sentence my interest in combating biases -- purportedly truth-apt status quo and hierarchy biases -- with other biases in research.) That is, the process requires mechanisms and filters that help to transform the biases in truth-conducive activity. There is a nice formal (see here for useful discussion) and empirical literature on the role of diversity in  research, when it is or is not apt. [So, I am certainly not saying that all individual biases are good!] So, for (4) needs to be rewritten as follows:

(4+)Being politically active predictably makes us worse at seeking the truth about political issues unless the research institutions in which we operate are designed to prevent our biases from undermining the epistemic process.

 

I started this post with a quote from Orwell. Throughout his writings, Orwell discusses the problems of partisanship and its relationship to (artistic) integrity (recall, for example, this post; and here). And he discerned that we should not conflate partisanship and group-think with all political activism. That is to say, he recognizes that party-politics and group-enforced ideology do corrupt (he was primarily interested in writing, but his arguments carry over easily to research). But the answer to this need not be political quietism or withdrawal into the ivory tower. The problem is party-politics and ideologies.

For, there are lots of ways one can be politically active in which one's research (artistic vision) need not be undermined at all. Sometimes this is due to functional role (e.g., recall the stuff above aggregators above). But sometimes is due to the fact that corrupting elements can be avoided. This is most clear in advocacy work on single issues (death penalty, abortion, same-sex marriage, euthanasia, basic income, human rights, racism, etc.). The researcher can be uncompromising in her single-issue advocacy and maintain her integrity. (Grant me the distinction between single-issue and systemic ideology.) Often this generates a kind of lack of moderation, even fanaticism in the name of truth. While there are problems with that, these are political problems not research problems.

 

 

On Chicago Economics and Samuelson's Technocratic Conception of Economics

Samuelson's exposition is consistently mathematical, and presupposes a knowledge of advanced calculus, higher algebra, and differential equations...I consider this failure to provide translations for the "literary" economist a serious shortcoming of his work. He dismisses translations into words as "mental gymnastics of a peculiarly depraved type." I disagree. There is no depravity, nor is there virtue, in telling other competent economists things in a language they all can understand-there is simply responsibility to the canons of scholarship.--G.J. Stigler (1948) reviewing Samuelson. [HT David M. Levy]

Even extremely well informed readers do not always recognize the significance of the books they review; or if they do, they may wish to suppress it in the hope to prevent a certain outcome. George J. Stigler, one of the leading Chicago-Economists of the 20th century (Nobel 1982), insists at the start of  his 1948 review of Paul Samuelson's Foundations of Economic Analysis that "much of [the] subject matter is currently of much diminished interest." Samuelson's (1947) Foundations is the book that (together with Samuelson's textbooks that followed) brought together the formal languages and techniques that created the mathematical 'revolution'* and its accompanying technocratic ideal which has been its self-image of economics since. (This is not to deny later contributions by Arrow, Debreu, McKenzie, etc.) After Samuelson (Nobel 1970) -- by his own lights "the last 'generalist' in economics," -- economics became an esoteric field for specialists and specializations. 

Stigler's review was not very influentual (most citations to it are by recent scholars) and is less known than, say, Boulding's [see this essay by Khan]. Even so, Stigler review is instructive because it articulates a number of themes that become distinctive of 'Chicago' (and, thus, a crucial feature of the intellectual edifice what is often known as 'neo-Liberalism.') Before I elaborate on that, Stigler's review is at its best when it explains Samuelson's methodological strategy. One often hears that mathematical economics is not empirical (as well as not realistic). But as Stigler notes (correctly), "The leitmotiv of Samuelson's treatise is the search for meaningful theorems, that is, theorems which are in principle capable of empirical contradiction" (either introspectively** or by way of social scientific data). In the context of commenting on Samuelson's empirical methodology, Stigler asks, 

what if the theorem is contradicted by observation? Samuelson says it would not matter much in the case of utility theory (p. 117); I would say that it would not make the slightest difference. For there is a free variable in his system: the tastes of consumers. These tastes are not observationally (or operationally) defined, so any contradiction of a theorem derived from utility theory can always be attributed to a change of tastes, rather than to an error in the postulates or logic of the theory.

 

On Keynes, Chicago Economics, and Social Engineering

"It is quite another thing to conclude that therefore ends of good policy are beyond the realm of scientific discussion. For surely the primary requisite of a working social system is a consensus on ends. The individual members of society must agree upon the major ends which that society is to seek. If any large share of the population actively disagrees with the society's ends, and in particular if it believes that the system is unfair by the group's criteria, the social system will surely disintegrate, probably with violence.

This is almost axiomatic in modern social theory…

At the level of economic policy, then, it is totally misleading to talk of ends as individual and random [as new welfare theorists do—ES]; they are fundamentally collective and organized.

If this conclusion be accepted, and accept it we must, the economist may properly exceed the narrow confines of economic analysis. He may cultivate a second discipline, the determination of the ends of his society particularly relevant to economic policy. This discipline might be called, following J.N. Keynes, applied ethics.7

7 See his Scope and Method of Political Economy (4^th edition: London 1930) especially chap. II. I did not realize how neglected this excellent book has become until, when I recently referred to it, several friends expressed surprise that I did not know Keynes's middle name was Maynard! –GJ Stigler (1943) "The New Welfare Economics" American Economics Review, 33(2): 357-8.

At the very start of his famous (1953) methodology paper, Milton Friedman calls attention to a three-fold distinction he inherits from the logician-economist, J.N. Keynes (yes, father of that more famous philosopher-economist, J.M. Keynes): "In his admirable book on The Scope and Method of Political Economy, John Neville Keynes distinguishes among "a positive science ... a body of systematized knowledge concerning what is; a normative or regulative science...[,] a body of systematized knowledge discussing criteria of what ought to be...[,]an art... [,] a system of rules for the attainment of a given end"; comments that "confusion between them is common and has been the source of many mischievous errors"; and urges the importance of "recognizing a distinct positive science of political economy."* Friedman admits that he is primarily interested in articulating the nature and methodology of positive economics, and along the way he wishes to argue that "positive economics is in principle independent of any particular ethical position or normative judgments." (By contrast, according to Friedman normative economics presupposes positive economics.) In what follows, I accept these distinctions by stipulation.

Neither Stigler nor Friedman, follow Keynes slavishly (recall this post, too). In particular, in the passage quoted at the top of this post, Stigler offers a world-historical innovation over Keynes that will have far-reaching consequences in economics and the way it understands itself as a policy science. In Keynes, "applied ethics" (that is, normative or regulative economics) is the business of philosophers not economists. And one of the main reasons it is the business of philosophers (and not economists) is that Keynes does not expect consensus over the ends of society to emerge. As he writes, "There is a further reason why a positive science of political economy should receive distinct and independent recognition. With the advance of knowledge, it may be possible to come to a general agreement in regard to what is or what may be in the economic world, sooner than any similar agreement is attainable in regard to the rules by which economics of individuals and communities should be guided. The former requires only that there as to the facts; the latter may be prevented by conflicting ideals." (50-51) So, Keynes's proposed strategy is to remove (applied) ethics as much as possible from economics so that economics can quickly hasten to become a mature (in a proto-Kuhnian sense) science.

How I Learned to Love Derrida by Talking to An Economist (I), [No, Not Paul de Man, too.]

[Due to a technical problem with Typepad this could not post yesterday.--ES]

"one cannot usefully say "I love you" in mathematical language"--Paul Samuelson (1965) Foundations of Economic Analysis, new foreword. (Quoted in Khan 1993: 795.)

Beginning as it did in the writings of philosophers, theologians, pamphleteers, special pleaders, and reformers, economics has always been concerned with problems of public policy and welfare. And at least from the time of physiocrats and Adam Smith there has never been absent from the main body of economic literature the feeling that in some sense perfect competition represented an optimal situation.—Paul Samuelson (1947) Foundations of Economic Analysis, Chapter VIII. (Quoted in Khan 1993: 787)

In Samuelson’s second passage, philosophy belongs to the disreputable far origins of economics (“special pleaders” is not a compliment). It’s an origin that is not wholly overcome in these mathematical and scientific Foundations, for a part of economics is continuous both with the far origins and its more recent origins ca 1750-76 (that is, “the time of physiocrats and Adam Smith”). Samuelson (an early Nobel in economics), who was not only the most influential mathematical economist within economics of the twentieth century, but also a first-rate historian of economics, is aware, of course, that the economist, Smith, is a philosopher not devoid of theoretical interest in feeling within the economy and (echoing Hume) in theorizing about the economy as well as even in theorizing about the social practice of theoretical reflection on the economy. For Samuelson, the continuity is topical (exhibited in concern with “problems of public policy and welfare”) as well as doctrinal (perfect competition represents an optimal situation).* Samuelson presupposes that what mathematical economics is about is wholly distinct from the mathematical language deployed in representing it. In his commentary on Samuelson, Khan, who does not deny the continuity, challenges this form-content distinction in mathematical economics.

Part of Samuelson’s achievement is to move away from feeling and articulate very precise circumstances in which perfect competition is an optimal situation. (To be clear, these circumstances occur in a model-world or representation; here I ignore their way of being, but it is their odd status that generates Khan's challenge.) In doing so, Samuelson manages to make economics something distinct from philosophy (a process initiated, perhaps, by Sigdwick and Marshall in the late 1880s) while remaining continuous with it. For the friends of Thomas Kuhn: Samuelson creates the paradigmatic normal economic science--a concern by many of his generation (see here). As Samuelson puts it in his 1965 foreword: his work is "part of the dawn, [with] serious economics a mathematical subject, just as physics...had become a century ago." Surprisingly enough, Samuelson’s achievement does not eliminate concern with problems of public policy and welfare within philosophy (reviving this, is part of Rawls’ achievement).