Digressions&Impressions

On Information and the State: from Justin Smith to George Stigler

[Over time I plan to phase out D&I at typepad. This post was first published at  digressions.impressions.substack here. To receive new posts and support my work, consider becoming a free or paid subscriber at <digressionsimpressions.substack.com>]

A recent wide-ranging, erudite, and politically salient Substack post by Justin Smith defends the thesis that "we may identify a vast process currently underway whereby we are moving, in our efforts to understand reality and to harness it to our ends, from its to bits. It seems to me the real story of the past few decades, a story that easily explains the rise in popularity of the simulation argument...is of a fairly total regime change, from physics to informatics —here I am adapting the French informatique, which names the research domain in question far more appropriately than “computer science”— as the Prima Scientia or ultimate science of reality." (The post may be gated so you will have to trust me on its interest or take a subscription to Justin's excellent substack, Hinternet.) Presumably, by 'regime' Smith is suggesting that the wide world of learning is fundamentally ordered and (he makes an explicit nod to Hacking here) ordering and authoritative in some sense. 

In these digressions, I have been joking over the years that economics will soon be displaced by computer science. But as is usual while I have been focusing on a change in one tree, Justin is noticing the shifts in the wider forest.

Unfortunately, Smith never defines 'information' (and its non-trivial cognate,) 'units of information processing' despite it playing a key role in his overall argument. That's probably a feature and not a bug of his argument because part of his point is that in virtue of the significance and fluidity (within constraints) of symbolic representation information is not (ahh) ontically stable, eternally out there. 

Inform has Latinate roots (to give form to, to shape, educate, etc.), and, perhaps, there are echoes of the scholastic doctrine of hylomorphism lurking here. (Not so long ago, I was pleased to learn that computer scientists use 'hylomorphism' to refer to a certain recursive function which struck me as very funny and profound at once.) As I was barely conscious contemplating the possible significance of this, through the mystery of the associative process and the rapidity of modern networks (and Google), I looked up and downloaded George Stigler's (1961) "The Economics of Information." This paper has over 12000 citations, and alongside his work on economic regulation, crucial to his importance to the development of economics (for which he got a Nobel).

Somewhat amusingly, Stigler also fails to give a definition of information. He tends to use 'knowledge,' 'advertising,' subjective beliefs, and the comparison of prices by consumers as ways of understanding and describing information. Information can be 'pooled' and has a 'flow.' Most famously, information is costly (and can be bought and sold).

That information is costly is both part of the cause and part of the effect of the dispersion of prices. (In an older tradition of economics -- as David M. Levy taught me, one can think of Wicksteed -- prices were thought to be uniform in a market at a given time; although earlier, during the 18th century, Adam Smith and his contemporary, James Steuart, thought that prices dispersed, too although their analysis is not identical.) I don't mean to suggest that this exhausts the role of information in Stigler's analysis of economic life because it connects means and ends, but about that some other time more.

In Stigler's paper, I encountered a fascinating passage that is worth quoting and discussing (and this is the real point of this digression):

The maintenance of appreciable dispersion of prices arises chiefly out of the fact that knowledge becomes obsolete. The conditions of supply and demand, and therefore the distribution of asking prices, change over time. There is no method by which buyers or sellers can ascertain the new average price in the market appropriate to the new conditions except by search. Sellers cannot maintain perfect correlation of successive prices, even if they wish to do so, because of the costs of search. Buyers accordingly cannot make the amount of investment in search that perfect correlation of prices would justify. The greater the instability of supply and/or demand conditions, therefore, the greater the dispersion of prices will be.
In addition, there is a component of ignorance due to the changing identity of buyers and sellers. There is a flow of new buyers and sellers in every market, and they are at least initially uninformed on prices and by their presence make the information of experienced buyers and sellers somewhat obsolete.
The amount of dispersion will also vary with one other characteristic which is of special interest: the size (in terms of both dollars and number of traders) of the market. As the market grows in these dimensions, there will appear a set of firms which specialize in collecting and selling information. They may take the form of trade journals or specialized brokers. Since the cost of collection of information is (approximately) independent of its use (although the cost of dissemination is not), there is a strong tendency toward monopoly in the provision of information: in general, there will be a "standard" source for trade information. (Stigler 1961, 220)

Chicago economists are not exactly known for assuming that any market will tend toward a monopoly. As I noted last week, in the 1950s Stigler's work was, in fact, prominent for undermining empirically the Marxist (and Marx-inspired) thesis that markets tend naturally tend to monopoly. 

Now, later when (in part inspired by the work of the sociologist Morton and in part because of his interest in which economics Phds would be hired where) Stigler turned to the sociology of scientific knowledge, he tackled this tendency toward monopoly in knowledge production head on. And so I will return to it some other time. But in the 1961 paper he leaves the point in the quoted passage alone.

It would be tempting here to discuss the role of google or wikipedia or the stanford encyclopedia of philosophy as standard sources of information in our ecologies. (And to what degree ChatGPT can displace these or become one itself.) But I want to close with an important broader context in which i situate Stigler's passage.

In the liberal tradition — with which Stigler was very familiar — which understands, as a a core function of the state, it (the state) as a machinery of record, a collector and disseminator of accurate public data. A lot of our social practices, inside and outside the market, presuppose a social infrastructure in which the machinery of record is reliable, allowing public authorities and private actors to plan their activities. For that to happen, the public must be well-informed, and the only way citizens can possibly be well-informed on complex matters of policy is for state experts to organize and process information. Elsewhere, also in joint (forthcoming) work with Nick Cowen I have argued that Walter Lippmann's (1922) Public Opinion puts this vision forward.[1]

In fact, the germ of Lippmann’s idea is expressed in the closing paragraph of John Stuart Mill’s and Harriet Taylor’s On Liberty (1869, Ch. 5), where they claim that “the greatest dissemination of power consistent with efficiency” should be allied with “the greatest possible centralisation of information, and diffusion of it from the centre.” In context, it is clear (recall) that they are responding to Tocqueville’s fears that in order to prevent democratic despotism through local self-government, incompetent art of government is inevitable.

So, lurking in Stigler (1961) is a causal account that both explains the origin of the state and justifies the state's continued existence as a (perhaps even natural) authoritative monopoly provider of the machinery of record. (In fact, this is how Smith explains the epistemic origin of the state: as guarantor of measures and weights and the soundness of currency. ) And that even in a bit-rich world this will remain so.

Of course, Marxists and Libertarians may object that the account hinted at here effaces the violence of the state. This is undoubtedly so, but I consider this a feature not a bug of the argument.

• This first appeared at: <On Information and the State: from Justin Smith to George Stigler (substack.com)> a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Chicago Economics, MIT, and Monopoly Capitalism

[Over time I plan to phase out this blog at typepad. This post was first published at  digressions.impressions.substack here. To receive new posts and support my work, consider becoming a free or paid subscriber at <digressionsimpressions.substack.com>]

Marxism predicts that left to its own devises capitalism tends toward monopoly. Let's call this the 'monopoly capitalism thesis.' I suspect the most famous articulation of this thesis was develop in the context of debates over the nature and fate of early twentieth century imperialism. For, both (recall; and here) Hobson (not a Marxist) and Lenin noticed that (militaristic) imperialism and monopoly were mutually reinforcing. For Hobson (in his 1902 Imperialism), monopoly capitalism was the effect of rent-seeking political behavior (by corrupt, imperial, financial interests). Lenin (ca 1916) didn't deny this, but while drawing on German data, especially, thought that capitalism had an inevitable tendency toward monopoly and cartels, and that imperialism (the search for monopoly markets overseas) is an effect of this tendency.

Arguably, (recall) Schumpeter (in Capitalism, Socialism, and Democracy (1942) synthesized the two positions by suggesting that in conditions of modernization, monopoly would be caused (perhaps not by an internal logic of market economies, but rather) by, the "concentration of decision-making centers of the administration and the state" who seek out social counterparts in capitalist enterprises, who in turn take advantage of rents. While Burnham (in The Managerial Revolution 1941) is not interested in exploring the dynamics of the monopoly capitalism thesis, one can find a similar conclusion in it (see also my comment about Berle & Means below).

In what follows, I don't mean to suggest that the theoretical structure of the monopoly capitalism thesis was not developed after mid-century. Arguably Joan Robinson, Paul Sweezy and Baran, and Ernest Mandel all did so in various ways, and have inspired renewed work. From the left one could mention (recall this post) Durand, Cédric Durand's (2017) Fictitious capital: How finance is appropriating our future; from the more libertarian right one could mention Eric A., Posner and E. Glen Weyl (2018) Radical Markets: Uprooting Capitalism and Democracy for a Just Society, which strongly implies our age is characterized by monopoly capitalism.

As an aside, with the revival of scholarly interest in the roots of neoliberalism, it is quite natural to come away thinking that the major point of contention between Marxists and liberals within political economy centered on the socialist calculation debate. This debate is theoretically very interesting because it brings together epistemology, social theory, political economy, and (after Oskar Lange) artificial intelligence. And because Hayek got an early Nobel (in part) for his contribution to it, and Austrian economics shaped American libertarianism, there has been impressive continued interest in it even after central planning lost its luster. But even critics of Marxism can acknowledge the soviet style central planning and administrative pricing need not be the only way really existing Marxism can be tried out. This observation made me return to the fate monopoly capitalism thesis: why did it not haunt mid twentieth century economics?

Now, a key part of the answer has to be in the success of Keynes & Keynesianism and also the very idea of a mixed economy (the genealogy of which I have not studied, but I assume involves Samuelson). But in a way this pushes the problem back a level--why think a mixed economy is stable so as to block the monopoly capitalism thesis (or road to serfdom thesis)?

Part of the answer can be found in a number of empirical, inductive studies associated with 'Chicago economics,' especially Warren Nutter’s (1951) The Extent of Enterprise Monopoly in the United States, 1899–1939, which originated in his (1949) dissertation and the fifth lecture, “Competition in the United States,” of George Stigler's (1949) Five lectures on economics problems first delivered at LSE at the behest of Robbins. (For detailed description of the material I am about to discuss see my paper, but back then I studied it through lens of paradigm building.)

I doubt I need to re-introduce Stigler (a future Nobel), who was at Columbia, to my audience. But despite the recent political and polemical focus on the Virginia School, Nutter is understudied (even his Wikipedia page is scarce), and only Daniel Kuehn has written much on him. Nutter, who during the cold war was one of the pre-eminent scholars of the Soviet economy Stateside (together with my very teacher Franklyn Holtzman), was at Yale when his dissertation work made an impact on American public life (it was featured in Fortune magazine). He later joined and co-founded the Virginia school of public choice and ended up a significant figure in the Nixon administration. In 1969, he published an extension of the dissertation with Henry Einhorn: The Extent of Enterprise Monopoly: 1899-1958. 

Now, as my comments above suggest, the 'monopoly capitalism thesis' drifted out of Marxism into mainstream-ish opinion during the Great Depression. Nutter's official target the “hypothesis that monopoly is automatically generated in a private-enterprise system” was also defended by Arthur R. Burns "in his book, The Decline of Competition.” Burns was an institutionalist economist at Columbia. (This Burns should not be confused with another economist at Columbia, Arthur F. Burns, who became more prominent as Chair of the Federal Reserve.) In his book, our Burns argues that technological developments alongside patent laws (which generate rents) made monopoly inevitable (and that anti-trust was imperfect).

Another, politically more important target (for Nutter and Stigler) is The Modern Corporation and Private Property (1932), co-authored by A.A. Berle and Gardiner Means. This was a major work that shaped the New Deal, corporate governance, and corporate law for  a generation. Berle and Means analytically (and legally) separated corporate ownership (which, under modern conditions, was diffused) from corporate control (which was concentrated). They thereby anticipate Schumpeter and Burnham's focus on a relatively narrow managerial-corporate class (albeit not to be identical simply with capital in a Marxist sense).

Nutter and Stigler have different empirical ways of tackling the underlying question, and they use different proxies to get a handle on the empirical problems (of operationalization concentrations within a industry and measuring its change over time, and of the whole economy). And their empirical findings don't always agree. But in a rather critical review (1953) article, "Is monopoly increasing?" Solomon Fabricant (an economist associated with NBER [see this portrait drawn by Lipsey]), who notes Nutter's and Stigler's partiality toward competition, and hostility to anti-trust policy, and implies they may be systematically biased,  concludes, nevertheless, “Yet, whatever the outcome, by the essential validity of their conclusion must stand. All the doubts that can be raised do not destroy, rather they support, the conclusion that there is no basis for believing that the economy of the United States is largely monopolistic and has been growing more monopolistic.” (92. Nutter himself had suggested there might have been very modest growth in monopoly, but mostly as an effect of government policy.) 

That is to say, by the early 1950s the economics profession thought the monopoly capitalism thesis was, if not empirically falsified, then at least decidedly unpromising empirically. This is especially notable because the data they covered included the impact of the Great Depression and all the New Deal activity.

In work (1951) that covered later time-series data for manufacturing through 1949, "The measurement of industrial concentration," another economist, M.A. Adelman, also decried the imperfection of the data (although he praised Nutter's measures explicitly) and diagnoses underdetermination of theory by data. (He also acknowledges that Nutter is more interested in monopoly, while he is more interested in concentration.) At the end of his survey Adelman concludes: 

"But although a given set of facts is compatible with more than one policy conclusion, this does not make it compatible with any conclusion, i.e., the facts are not simply irrelevant. Stated somewhat differently: our conclusions, if correct, should mean something for the atmosphere in which policy is formed. References to "the growth in economic concentration," or statements that "during the past I6 years [I932-48] big business has been getting bigger and little business littler," or that "the forces of concentration [are] growing stronger by the hour," or that "concentration of industry is increasing so fast that no half-measures can stop it," and many others of like tenor - these must now be dismissed as unfounded. Concentration may be a problem, but for better or worse it is not threatening to engulf the economy. The moral, therefore, is simply that there is time to stop, look, and take thought. This may detract from the color and excitement of our public life, but it may also contribute to better results."  

Adelman has, I think, attracted no scholarly attention. But he became one of pillars of the MIT economics department that Samuelson built to pre-eminence. (After his early work in anti-trust he became somewhat well known in the 1970s as a leading expert of energy economics and how to tackle the oil crisis.) And so he very much represents the middle of the road of the economics profession of the era.

I quote Adelman's passage at length because it gives a sense of how he takes the empirical results to license considerable de-politicization. (One can also find this in Fabricant's conclusion.) While economists remained highly interested in theories of monopoly and imperfect competition, and how to deal with (ant-trust) policy questions emanating from actual monopolies and economic concentration (and a whole range of potentially fascinating technical questions), the monopoly capitalism thesis was effectively dead, or at least temporarily pushed to the margins. 

As I noted in my original paper, this conclusion remained intact even if this inductive approach to the question of monopoly as pursued by Nutter and Stigler got displaced by a very different approach articulated by Harberger in “Monopoly and Resource Allocation.” (1954) And while this, in turn, was quite consequential for the shape of American anti-trust (not the least compared with the Ordoliberal approach), it was unconcerned with the monopoly capitalism thesis.

• This first appeared at: <On the controversy in the philosophical blogosphere. (substack.com)> a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

 

Some Pre-History on the History and Sociology of Multiple Discovery: Merton, Dicey, Stigler- (etc.)

It may very well, owing to the condition of the world, and especially to the progress of knowledge, present itself at the same time to two or more persons who have had no intercommunication. Bentham and Paley formed nearly at the same date a utilitarian system of morals. Darwin and Wallace, while each ignorant of the other’s labours, thought out substantially the same theory as to the origin of species.--A.V. Dicey [2008] (1905) Lectures on the Relation between Law & Public Opinion in England during the Nineteenth Century, Indianapolis: Liberty Fund, p. 18 n. 6 (based on the 1917 reprint of the second edition).

As regular readers know (recall), I was sent to Dicey because he clearly shaped Milton Friedman's thought at key junctures in the 1940s and 50s. So, I was a bit surprised to encounter the passage quoted above. For, I tend to associate interest in the question of simultaneous invention or multiple discovery with Friedman's friend, George J Stigler (an influential economist) and his son Steven Stigler (a noted historian of statistics). In fairness, the Stiglers are more interested in the law of eponymy. In his (1980) article on that topic, Steven Stigler cites Robert K. Merton's classic and comic (1957) "Priorities in Scientific Discovery: A Chapter in the Sociology of Science." (Merton project was revived in Liam Kofi Bright's well known "On fraud.")

When Merton presented (and first published it) he was a colleague of George Stigler at Columbia University (and also Ernest Nagel). In his (1980) exploration of the law of eponymy, Steven Stigler even attributes to Merton the claim that “all scientific discoveries are in principle multiple." (147) Stigler cites here p. 356 of Merton's 1973 book, The Sociology of Science: Theoretical and Empirical Investigations, which is supposed to be the chapter that reprints the 1957 article. I put it like that because I was unable to find the quoted phrase in the 1957 original (although the idea can certainly be discerned in it, but I don't have the book available to check that page).

Merton himself makes clear that reflection on multiple discovery is co-extensive with modern science because priority disputes are endemic in it. In fact, his paper is, of course, a reflection on why the institution of science generates such disputes. Merton illustrates his points with choice quotes from scientific luminaries on the mores and incentives of science that generate such controversies, many of which are studies in psychological and social acuity and would not be out of place in Rochefoucauld's Maximes. Merton himself places his own analysis in the ambit of the social theory of Talcott Parsons (another important influence on George Stigler) and Durkheim. 

The passage quoted from Dicey's comment is a mere footnote, which occurs in a broader passage on the role of public opinion in shaping development of the law. And, in particular, that many developments are the effect of changes in prevaling public opinion, which are the effect of in the inventiveness of "some single thinker or school of thinkers." (p. 17) The quoted footnote is attached to the first sentence of remarkably long paragraph (which I reproduce at the bottom of this post).* The first sentence is this: "The course of events in England may often at least be thus described: A new and, let us assume, a true idea presents itself to some one man of originality or genius; the discoverer of the new conception, or some follower who has embraced it with enthusiasm, preaches it to his friends or disciples, they in their turn become impressed with its importance and its truth, and gradually a whole school accept the new creed." And the note is attached to 'genius.'

Now, often when one reads about multiple discovery (or simultaneous invention) it is often immediately contrasted to a 'traditional' heroic or genius model (see Wikipedia for an example, but I have found more in a literature survey often influenced by Wikipedia). But Dicey's footnote recognizes that in the progress of knowledge, and presumably division of labor with (a perhaps imperfect) flow of ideas, multiple discovery should become the norm (and the traditional lone genius model out of date).

In fact, Dicey's implicit model of the invention and dissemination of new views is explicitly indebted to Mill's and Taylor's account of originality in chapter 3 of On Liberty. (Dicey only mentions Mill.) Dicey quotes Mill's and Taylor's text: "The initiation of all wise or noble things, comes and must come from individuals; generally at first from some one individual." (Dicey adds that this is also true of  folly or a new form of baseness.)   

The implicit model is still very popular. MacAskill's account (recall) of Benjamin Lay's role in Quaker abolitionism (and itself a model for social movement building among contemporary effective altruists) is quite clearly modelled on Mill and Taylor's model. I don't mean to suggest Mill and Taylor invent the model; it can be discerned in Jesus and his Apostles and his been quite nicely theorized by Ibn Khaldun in his account of prophetic leadership. Dicey's language suggests he recognizes the religious origin of the model because he goes on (in the very next sentence of the long paragraph) as follows: "These apostles of a new faith are either persons endowed with special ability or, what is quite as likely, they are persons who, owing to their peculiar position, are freed from a bias, whether moral or intellectual, in favour of prevalent errors. At last the preachers of truth make an impression, either directly upon the general public or upon some person of eminence, say a leading statesman, who stands in a position to impress ordinary people and thus to win the support of the nation."

So far so good. But Dicey goes on to deny that acceptance of a new idea depends "on the strength of the reasoning" by which it is advocated or "even on the enthusiasm of its adherents." He ascribes uptake of new doctrines to skillful opportunism in particular by a class of political entrepreneurs or statesmanship (or Machiavellian Virtu) in the context of "accidental conditions." (This anticipates Schumpeter, of course, and echoes the elite theorists of the age like Mosca and Michels.) Dicey's main example is the way Bright and Cobden made free trade popular in England. There is space for new directions only after older ideas have been generally discredited and the political circumstances allow for a new orientation. 

It's easy to see that Dicey's informal model (or should I say Mill and Taylor's model?) lends itself to a lot of Post hoc ergo propter hoc reasoning. So I am by no means endorsing it. But the wide circulation of some version of the model helps explain the kind of relentless repetition of much of public criticism (of woke-ism, neoliberalism, capitalism, etc.) that has no other goal than to discredit some way of doing things. If the model is right these are functional part of a strategy of preparing the public for a dramatic change of course. As I have noted Milton Friedman was very interested in this feature of Dicey's argument [Recall:  (1951) “Neo-Liberalism and its Prospects” Farmand, 17 February 1951, pp. 89-93 [recall this post] and his (1962) "Is a Free Society Stable?" New Individualist Review [recall here]].

I admit we have drifted off from multiple discovery. But obviously, after the fact, multiple discovery in social theory or morals can play a functional role in the model as a signpost that the world is getting ready to hear a new gospel. By the end of the eighteenth century, utilitarianism was being re-discovered or invented along multiple dimensions (one may also mention Godwin, and some continental thinkers) as a reformist even radical enterprise. It was responding to visible problems of the age, although its uptake was not a foregone conclusion. (And the model does not imply such uptake.)

It is tempting to claim that this suggests a dis-analogy with multiple discovery in science. But all this suggestion shows is that our culture mistakenly expects or (as I have argued)  tacitly posits an efficient market in ideas in science with near instantaneous uptake of the good ideas; in modern scientific metrics the expectation is that these are assimilated within two to five years on research frontier. But I resist the temptation to go into an extended diatribe why this efficient market in ideas assumption is so dangerous.

Milton Friedman on Marx and Mill (pt 1): The Road to Serfdom and the Art of Government

Judged by the course of events of the last century, rather by the avowed aim of Mill and Marx, there is much for reversing the stereotyped roles assigned to the two men. If collectivism ultimately triumphs over individualism, it will be in no small measure a result of the influence of the ideas first popularized and made respectable by Mill; whereas, if individualism ultimately triumphs, it will be in no small measure a result of the ultimate effects of the belief in revolutionary action to which Marx and Engels gave such vivid expression in the Manifesto…

The great defect of the Benthamite liberals among whom Mill grew up was the absence of any theory or doctrine of the positive role of the state in the organization of economic activity. Benthamism was at bottom a fervent belief in the possibility of improving the condition of mankind through legislative enactment devoted to achieving the “greatest happiness of the greatest number.” These central premises do not themselves prescribe any particular content for legislative action. They are, in strict logic, consistent equally with fargoing collectivism and paternalism or with the ”laissez faire” doctrines with which they were in fact combined. The acceptance of laissez faire as a guiding principle was far less the product of explicit analysis or comparison of any exhaustive set of alternatives[;] prohibited were largely assaults on person or property overwhelmingly regarded as clearly indefensible and the appropriate subject for punitive legislation. In this way, the success of laisses faire removed one of the chief factors responsible for the initial acceptance of laissez faire. By the end of John Mill’s life, the state was no longer what it was during his father’s or Bentham’s time—a corrupt, inefficient instrument whose enactments were widely held in low repute. It had become a relatively honest and efficient body, whose enactments were held in high esteem by the body politic.

The sweeping away of the hindrances to the free movement of men, goods, and capital was followed by the great improvements in economic well being. Yet there obviously remained much misery and poverty to which a passionate humanitarian like Mill could not remain blind. It was perhaps not unnatural that we was willing to sanction action by an honest and much improved state administration to redress grievances. He had no principles of state action by which to test proposals for reform. He was almost certain to minimize or reject entirely the argument—if it were made—that direct interference by the state would threaten that private liberty he prized so highly.  For this argument conflicted with his deep, though naïve, belief in the perfectibility of human beings through education. Once men were educated, he believed, they would become not only wise but also good.--Milton Friedman (September 10, 1948) "Discussion of Paper by V.W. Bladen The Centenary of Marx and Mill" at The Eight Annual meeting of the Economic History Association. Hoover Institution, Collection Title: Milton Friedman papers Container: box 39. [HT David M. Levy]

Bladen's paper can be found here. Originally Friedman had been invited to comment on a paper on laisser faire by J. Bartlett Brebner (which was turned into an influential article: "Laissez Faire and State Intervention in Nineteenth-Century Britain." The Journal of Economic History, 8(S1), 59-73.) From the correspondence at Hoover it's unclear what prompted the move, but Bladen's paper was the opening and keynote to the conference, and Friedman did not object. The other commentator on the program is Elizabeth Schumpeter--the schedule adds in parenthesis: "Mrs. Joseph A." And it would be lovely to locate her comments. 

Bladen was late sending Friedman his paper, and this may help account for the fact that much of Friedman's discussion reads as a riff on A.V. Dicey's (1905 [1914]) Law & Public Opinion rather than a detailed criticism of Bladen (although Friedman added a passage on trade unionism that clearly is critical of Bladen). Throughout Friedman's writings Dicey is an important source, not the least his better studied (1951) “Neo-Liberalism and its Prospects” Farmand, 17 February 1951, pp. 89-93 [recall this post] and his (1962) "Is a Free Society Stable?" New Individualist Review [recall here] some other time I will return to that. (The 1962 piece draws on themes from the manuscript I am discussing today.) And while in today's post 'Marx' is clickbait in the title, I will return to Friedman's comments on Marx, too. Okay, with that in place, let me turn to the text.

Friedman attributes to Dicey (1835 – 1922) a kind of road to serfdom thesis in which Mill's good intentions lead not just to the prevailing support for collectivism that Dicey diagnoses as the effect of Mill's writing at the end of the nineteenth century, but also that the collectivist reforms proposed by Mill would (now quoting Friedman) "seriously threaten political liberty" in virtue of the gradualism that Mill advocated and the tendency to attribute difficulties consequent intervention to the "defects of the price system." Crucially, for Friedman it's the historical experience of Marx's effect on the Russian revolution that halts the English road to serfdom. 

Now, in his analysis Friedman ignores the role of imperialism, and the opportunities for rents this provided, in changing the political culture of nineteenth century liberalism. Hobson, for example, argued that this undermined the pacific, free trade coalition. And while Dicey has less nostalgia for this coalition, he concurs with Hobson's diagnosis. {Of course, given Mill's own advocacy for a civilizational mission of British imperialism, it's not as if this lets Mill off the hook.} It is worth noting that Dicey thinks that imperialism (and high taxation that is the effect of it) may well have slowed the road to serfdom process that Friedman attributes to him (see, especially, the 1914 introduction to the second edition, and chapter XII).

It's a bit odd that Friedman misses the significance of (financial and military procurement) rents to imperialism. Because earlier, in describing the rise of laissez faire, Friedman argues in a public choice vein, that “The Benthamites devoted much attention to improving public administration. Their success in this connection was as great as in establishing a large measure of laissez faire, and the two achievements are not of course unrelated. The establishment of laissez fair enormously reduced the benefits which civil servants could confer on private individuals and greatly lessened the incentive or opportunity to break laws.” 

As an important aside, I am pretty confident that Friedman had read Hayek's Road to Serfdom by 1948. And there is no sign in his 1948 argument that he is as critical of Hayek in the way that his later use of Dicey in 1962 suggests (recall here). 

What's neat about the material I have quoted above is that according to Friedman the key defect in Mill's political economy is that "he had no principles of state action by which to test proposals for reform." That is, the central problem that Mill faces in using state action to ameliorate the plight of the poor and miserable, is that according to Friedman Mill lacks -- and now I am using terminology common to Mill, J.N. Keynes, Friedman, and Foucault -- an art of government.

In fact, (recall) we know from his correspondence with Stigler (Hammond & Hammond) that by 1948 Friedman had started working on his (1953) "Methodology of Positive Economics" paper that uses that very terminology. And that Friedman ends up echoing Mill by treating the art of economics as dependent on empirical science. For it’s this science that provides the knowledge that constitute at least part of the rules of how one gets from given ends to proper outcomes. That is, the dependence of the art on positive science is epistemic in character. And so lurking here is a more fundamental (Marshallian) criticism of not just Mill's art of government, but his political economy more generally one that attributes to Mill a kind of violation of a do no harm principle in political life. To be continued.

 

 

 

 

 

 

On Foucault's Discipline and Punishment & Chicago Economics

It is understandable that the criticism of the public execution should have assumed such importance in penal reform: for it was the form in which, in the most visible way, the unlimited power of the sovereign and the ever-active illegality of the people came together. Humanity in the sentences was the rule given to a system of punishment that must fix their limits on both. The ‘man’ that must be respected in the sentence was the juridical and moral form given to this double delimitation.
But, although it is true that reform, as a penal theory and as a strategy of the power to punish, took shape at the point of coincidence of these two objectives, its stability in the future was due to the fact that, for a long time, priority was given to the second. It was because the pressure on popular illegalities had become, at the period of the Revolution, then under the Empire, and finally throughout the nineteenth century, an essential imperative, that reform was able to pass from the project stage to that of an institution and set of practices. That is to say, although the new criminal legislation appears to be characterized by less severe penalties, a clearer codification, a marked diminution of the arbitrary, a more generally accepted consensus concerning the power to punish (in the absence of a more real division in its exercise), it is sustained in reality by an upheaval in the traditional economy of illegalities and a rigorous application of force to maintain their new adjustment. A penal system must be conceived as a mechanism intended to administer illegalities differentially, not to eliminate them all.--Michel Foucault (1975) Discipline and Punish: The Birth of the Prison [Surveiller et punir: Naissance de la prison] translated by Alan Sheridan, pp. 88-9 [in the Vintage & Penguin classcis edition]

The passage quoted from Discipline and Punish (hereafter D&P) caught my attention because the final sentence expresses an idea that later in the (1979) lecture series, Birth of Biopolitics, Foucault associates with the 'Chicago' (that is, Becker and Stigler) approach to crime. Let me explain.

At the end of the second lecture of 17 January 1979, with an explicit nod to Beccaria, Foucault had summarized  the"new regime" of "the well-known principle of mildness of punishment appears:" "...government is basically no longer to be exercised over subjects and other things subjected through these subjects. Government is now to be exercised over what we could call the phenomenal republic of interests." (p. 49) In fact, in 1979, Foucault here explicitly denies that this change has anything to do with a new humane sensibility.

He then returns to the theme of crime as a major subject (recall this post) in his tenth lecture of 21 March 1979, which is the second one devoted to American (Chicago) neoliberalism (in context he explicitly mentions Ehrlich, Becker, and Stigler).* There he starts his discussion by claiming that "Their analysis of criminality at first appears to be the simplest possible return to the eighteenth century reformers like Beccaria and especially Bentham." (248)+ He then goes on to summarize a kind of synthesis of the reformers as follows (and uses Coase-ian language):

A good law is necessary and almost sufficient, was nothing other than the desire for what could be called, in economic terms, a reduction in the transaction cost. The law is the most economical solution for punishing people adequately and for this punishment to be effective. First, the crime must be defined as an infraction of a formulated law, so that in the absence of a law there is no crime and an action cannot be incriminated. Second, penalties must be fixed once and for all by the law. Third, penalties must be fixed in law according to the degree of seriousness of the crime. Fourth, henceforth the criminal court will only have one thing to do, which is to apply to an established and proven crime a law which determines in advance what penalty the criminal must suffer according to the seriousness of his crime. An absolutely simple, apparently completely obvious mechanics constitutes the most economic form, that is to say, the least costly and most effective form for obtaining punishment and the elimination of conducts deemed harmful to society. (p. 249)

So, in 1979, Foucault implies that the late eighteenth century penal reformers (Beccaria & Bentham) invent the economic analysis of the law in terms of transaction costs and cost benefit analysis. It is worth noting that this reflects a modest change from Discipline and Punishment, where Beccaria is not treated as a progenitor of law and economics, and where (with nods to Capital) Bentham is treated as someone who helps create or constitute the homo economicus through a 'disciplinary mechanism' (D&P, pp. 205-209), but does not take homo economicus as given (in the way, say, Chicago economics does). This change, reflects a rethinking on Foucault's part of the role of interest (and of a decrease of Marx's significance and an increase of Hume's significance to his overall argument), but about that some other time more.

Now, in the passage quoted from p. 249 it seems as if Foucault allows that the 18th century reformers allowed for the complete elimination of conduct deemed harmful to society. But recall (again here) that this will come at a cost. And so, de facto, the way Foucault introduces the eighteenth century reformers is that in their law & economics of crime, one can calculate a kind of optimum crime rate at the lowest cost to society. And, after explaining that the reformers did not have a pure law & economics, this is, in fact, the view he ends up attributing to Stigler as "the objective of a penal policy: “The goal of law enforcement,” he says, “is to achieve a degree of compliance with the rule of prescribed behavior that society believes it can procure while taking account of the fact that enforcement is costly.” This is in the Journal of Political Economy in 1970." (256; as the editors note, this is a reference to “The optimum enforcement of laws,”) And then Foucault glosses Stigler's point as follows: "Consequently, good penal policy does not aim at the extinction of crime, but at a balance between the curves of the supply of crime and negative demand. Or again: society does not have a limitless need for compliance. Society does not need to conform to an exhaustive disciplinary system." (256)**

That a "penal system must be conceived as a mechanism intended to administer illegalities differentially, not to eliminate them all" is, thus, a key insight of Stigler (1970), and Foucault makes a point of emphasizing this in his 1979 lecture where he treats it as a return to the reformist sensibility of the late 18th century (before it was, as it were, corrupted by criminology's search for the anthropology of the criminal). Foucault thinks this is highly significant because, looking forward, "On the horizon of this analysis we see instead the image, idea, or theme-program of a society in which there is an optimization of systems of difference, in which the field is left open to fluctuating processes, in which minority individuals and practices are tolerated, in which action is brought to bear on the rules of the game rather than on the players, and finally in which there is an environmental type of intervention instead of the internal subjugation of individuals." (259-260, emphasis added on the Hayekian, 'rules of the game'). While I would not go so far as claiming that Foucault endorses this horizon in normative fashion, I do think he is intrigued by the opportunities this affords.

Before I conclude. You might think I have made way too much of Foucault's claim that the "penal system must be conceived as a mechanism intended to administer illegalities differentially, not to eliminate them all" as an anticipation of his analysis of Chicago law & economics. After all, the quoted paragraph leading up to it does not seem very law & economics at all. Fair enough. However, in his own gloss, he goes on to write the following:

Shift the object and change the scale. Define new tactics in order to reach a target that is now more subtle but also more widely spread in the social body. Find new techniques for adjusting punishment to it and for adapting its effects. Lay down new principles for regularizing, refining, universalizing the art of punishing. Homogenize its application. Reduce its economic and political cost by increasing its effectiveness and by multiplying its circuits. In short, constitute a new economy and a new technology of the power to punish: these are no doubt the essential raisons d'etre of penal reform in the eighteenth century. (D&P, 89)++

So, it's clear that in 1975, Foucault understands the significance of the late 18th century penal reformers in terms of a political economy. And he presents them as anticipating the very point that in 1979 he attributes to Stigler (who he treats as returning to the best insights of the reformers). And this theme is really introduced as early as D&P p. 82, so it's not a mere  throw away remark.

Let me conclude with two speculations. First, it is not impossible that Foucault was familiar with Stigler's work before he published Discipline and Punish. There is an annotation on Stigler 1970 in his hand (see here). After all, Foucault was enmeshed in prison reform activism, so not just doing historical research on prisons in the early 1970s. This hypothesis is strengthened by the fact that Stigler 1970 is absent in Lepage (1978), which is often treated as Foucault's source to Chicago neoliberalism.** So, if that's right, Foucault was already interpreting his source material in light of some familiarity with the Chicago approach to crime.

Second, it's also possible, of course, that Foucault's familiarity with Stigler and the Chicago approach to crime came after he wrote Discipline and Punish. And then that opens up the possibility of not just an epynomous disovery (a term coined by George's son, Stephen Stigler), but also a kind of convergence between Foucault and Chicago economics on a non-trivial issue. This suggests that Foucault's purported late affinity with American neoliberalism (a much debated issue among Foucault scholars) may have roots in his own intellectual development in reflecting on the role of crime and punishment in society.

 

Mill's Elitism and Progressivism, The Art of Government, Tocqueville (and Foucault), Part I

The mode of conducting the practical business of government, which is best under a free constitution, would generally be best also in an absolute monarchy: only, an absolute monarchy is not so likely to practise it. The laws of property, for example; the principles of evidence and judicial procedure; the system of taxation and of financial administration, need not necessarily be different in different forms of government. Each of these matters has principles and rules of its own, which are a subject of separate study. General jurisprudence, civil and penal legislation, financial and commercial policy, are sciences in themselves, or rather, separate members of the comprehensive science or art of government: and the most enlightened doctrines on all these subjects, though not equally likely to be understood or acted on under all forms of government, yet, if understood and acted on, would in general be equally beneficial under them all. It is true that these doctrines could not be applied without some modifications to all states of society and of the human mind: nevertheless, by far the greater number of them would require modifications solely of detail, to adapt them to any state of society sufficiently advanced to possess rulers capable of understanding them. A government to which they would be wholly unsuitable, must be one so bad in itself, or so opposed to public feeling, as to be unable to maintain itself in existence by honest means.--J.S.Mill (1861) Considerations on Representative Government, Chapter 2 ("The Criterion of a Good Form of Government.)

One can find passages in Mill's writings that suggest he was a skeptic about the very possibility of an art of government. But some of the works that endured endorse the art, and even suggest it, as above, that it is a "comprehensive science." In Mill, this tends to mean, amongst other things, that it is a body of rules (see System of Logic). But in the quoted passage he also emphasizes that these involve certain "doctrines."

In virtue of the fact that these rules are context insensitive, Mill is willing to call these rules a 'science.' Presumably the rules, which are built on a body of empirical generalizations (in the indicative), have the character of (what he and his followers) would call 'imperative statements' (recall this post). These have the task of providing one the means for given ends, that is, a kind of instrumental or conditional rationality. This is actually something meta, because government itself has this kind of instrumental rationality baked into its purpose according to Mill; it exists (no surprise here given his utilitarianism) to promote "the interests of any given society." 

Interestingly enough, and this anticipates Foucault's observation (recall here) on the liberal art of government as such in the first few sentences of the Birth of Biopolitics, for Mill there is a version of the art 'which is best under a free government.' Now undoubtedly he does not mean to say, in the first instance, that this is the optimal art of government, but in the second instance he notes that is best sans phrase. It turns out that there is a normative art of government that would be context invariant.

Obviously, I overstate Mill's explicit position. As the ancients teach there are more modes of tyrannical government, where, as he notes, the normative art of government is impossible by definition because it would be self-underming ("A government...so bad in itself, or so opposed to public feeling, as to be unable to maintain itself in existence by honest means").* But what Mill calls a 'free constitution' and an 'absolute monarchy' do not exhaust the non-tyrannical options. (By a 'free constitution' Mill means something like what we call a 'liberal democracy' to considerable degree.) I return to this issue below. But Mill allows (and there are echoes of Hume here), that good government is, in principle, possible in an absolute monarchy even if it is rare in practice.

I take it that the 'art of government' just is, for Mill, well grounded skill at what he calls 'the practical business of government.' And this happens to be subdivided into a whole number of practical disciplines commonly taught we might surmise in law schools ("General jurisprudence, civil and penal legislation") and public policy or and/or economics departments ("financial and commercial policy.") The point is not just that within the art of government there is a cognitive division of labor, specialization, but also that these disciplines are ultimately subordinate to a larger purpose.

Some of you may be eager to remind me that strictly speaking the art of government is not really context insensitive. Mill quite clearly presupposes something like a distinction between advanced civilization, and those cultures that still require some tutelage. As he puts it at the start of the chapter that I have been quoting, "the proper functions of a government are not a fixed thing, but different in different states of society; much more extensive in a backward than in an advanced state." This stadial idea of cultural development is deeply woven into Mill's social theory (and, obviously, is a part of his support for a certain kind of reformed imperialism.) Strictly speaking, and this is a further strain of elitism in Mill, only those that would have to apply the art of government need to 'advance' to the level familiar from civilized culture. 

As an aside, every time I confront these features of Mill's thought I always despair a little bit for the liberal tradition. But I also find in it a salutary lesson that great moral sensitivity is compatible with great obtuseness, and this should make us skeptical in our dispositions about our own certitudes. (No, I won't preach more of my skepticism today.)

Be that as it may, somewhat paradoxically, where the art of government is most needed (the underdeveloped state), according to Mill, because there government must have many functions, it is unavailable because (to put in modern slightly more neutral terms, well only slightly) of a lack of human capital to develop and implement it. Mill himself uses puts it as follows: "the most important point of excellence which any form of government can possess is to promote the virtue and intelligence of the people themselves." So, lurking here is the idea that real development presupposes considerable state capacity and expertise that rests on considerable human capital. And as a people develop government functions disappear or wither away. (I put it like that because there is something analogous in Marxism.) A nation of super-intelligent spinozist sages would not need any government. But the tragedy of human life (and development) is, on Mill's view, that (knowledgeable) investment in human capital only really gets going after decent enough government has been achieved.

For Mill, "one criterion of the goodness of a government, the degree in which it tends to increase the sum of good qualities in the governed, collectively and individually, since, besides that their well-being is the sole object of government, their good qualities supply the moving force which works the machinery [of government]." When I started this post, I wanted to say something about how we can discern here in Mill a response to Tocqueville but about that some time more. I do note that we can see here some non-trivial kinship between Becker and Stigler and Mill. So, when discussing the programming of human capital by Chicago, Foucault's tendency to treat Becker and Stigler as the natural culmination and intensification of the utilitarian-philosophical-radical traditon (in context Mill is clear that his views on this point have an affinity with Bentham's) is actually well grounded. (I have a suspicion that Halevy and Lepage have something to do with this. But again, some other time more.)

One element of the art of government is guided by the following doctrine (which looks back to Hume and Spinoza and Madison): "The ideally perfect constitution of a public office is that in which the interest of the functionary is entirely coincident with his duty." And this, for Mill, scales up to what Rawls calls the 'basic institutions:' "What we have said of the arrangements for the detailed administration of the government is still more evidently true of its general constitution. All government which aims at being good is an organization of some part of the good qualities existing in the individual members of the community for the conduct of its collective affairs." 

That is, we may say then, that Mill divides what we can call political theory/philosophy in two. One is engaged with  the basic structure or as he puts it, the "fundamental constitution of the government." This is institutional design. (Recall this post.) And the other, is the art of government, that is, the understanding of the mode of conducting the practical business of government, Obviously, there is an important connection between the two because there are going to be many interaction effects between them. While it is tempting to see the designing of the 'fundamental constitution' as the more important institution (since it molds the rest), as the argument above has revealed, without the proper clay it won't get off the ground. And so that means that, in practice, the constitution, and even the most general principles that enter into the art of government must respect context. (The rest of the chapter articulates this very point.) In this chapter, Mill treats the art of government as part of political philosophy.** 

Here I want to close with an observation. One may well think in light of the significance of professional schools above, that fundamentally the art of government is something technocratic for Mill. But I don't think that's quite right. Not only because of the ways in which empirical and normative elements blend in it. For the aim of the art is something programmatic: the capacity to articulate and implement modes of government that "favor and promote, not some one improvement, but all forms and degrees of it." And while that  presupposes an ideal constitution, which is to be developed theoretically, it also presupposes a program of open-ended intellectual, moral, and social progress in our capacities for (and this is the significance of my Spinozist joke above), self-government. 

 

 

 

Foucault on Hume on Interest and (Adaptive) Preference

What English empiricism introduces...and doubtless for the first time in Western philosophy, is a subject...who appears in the form of a subject of individual choices which are both irreducible and non-transferable....

Second, this type of choice is non-transferable. I do not mean that it is non-transferable in the sense that one choice could not be replaced by another. You could perfectly well say that if you prefer health to illness, you may also prefer illness to health, and then choose illness. It is also clear that you may perfectly well say: I prefer to be ill and that someone else is not. But, in any case, on what basis will this substitution of one choice for another be made? It will be made on the basis of my own preference and on the basis of the fact that I would find someone else being ill more painful, for example, than being ill myself. In the end the principle of my choice really will be my own feeling of painful or not painful, of pain and pleasure. There is Hume’s famous aphorism which says: If I am given the choice between cutting my little finger and the death of someone else, even if I am forced to cut my little finger, nothing can force me to think that cutting my little finger is preferable to the death of someone else.
So, these are irreducible choices which are non-transferable in relation to the subject. This principle of an irreducible, non-transferable, atomistic individual choice which is unconditionally referred to the subject himself is what is called interest.--Michel Foucault, 28 March 1979, The Birth of Biopolitics, pp. 271-272. 

This is the third of my recent posts on Foucault's reading of Hume in The Birth of Biopolitics (recall here; and here). In the first paragraph of the quoted passage I have removed Foucault's reference to Locke as the world-historical origin of English empiricism as the fount of Utilitarianism. Foucault is almost certainly inspired by Élie Halévy's classic (1901-1904) The Growth of Philosophic Radicalism, which I now know Foucault studied. (We have some of Foucault's note-taking on volume 2 [see here].) I have also removed Foucault's taxonomy of different kinds of approaches to identify a subject in 'Western philosophy.' Here I am also skipping Foucault's treatment of Appendix 1, “Concerning Moral Sentiment” to Hume's (1751) An Enquiry Concerning the Principles of Morals, which illustrates  what it is for a choice to be irreducible in empiricism, according to Foucault. I do so not because these topics are uninteresting (on the contrary, they are too interesting), but because I want to focus on Foucault's analysis of Hume's account of the non-transferability of choices. (I am unfamiliar with other work on this topic so feel free to share your suggestions.)

In the paragraph I skipped Foucault had shown that for an empiricist (like Locke and Hume) "The painful or non-painful nature of the thing is in itself a reason for the choice beyond which you cannot go." (p. 272. This is irreducibility of choice, or what Foucault calls "the regressive end point in the analysis.") While a lot can be said about this, let's stipulate it's true for the sake of argument. But for what follows, you need to keep in mind that at bottom the agency of an empiricist subject consists in choices that are ground in (the desire for) pleasure or the aversion to pain. And according to Foucault this is what creates a subject with an interest (as a kind of characterization of agency.) That's a big claim, but I am interested here in the details. 

The non-transferability principle is cashed out by Foucault in terms of the fact that preferences are one’s own and that another has no fundamentally no access to them. Foucault really uses the modern (somewhat technical) language of ‘preference’ on p. 272. (It's also there in the French: "À partir de ma préférence à moi et à partir du fait que je trouverai plus pénible, par exemple, de savoir qu'un autre est malade que de l'être moi-même." Naissance de la biopolitique, p. 276.) While Hume certainly has ‘preference’ in his vocabulary, in Hume it generally means something like ‘favoring’ a particular outcome or state of affairs (which is also the use one can find in Ricardo). In Hume 'preference' is not used a technical term to track what is motivationally foundational. In my view it is only in Bentham that the use of 'preference' starts to obtain it's technical, familiar sense. (Something one might pick up from Halévy, who emphasizes Bentham's self-preference principle in various contexts but not in Hume i think.)

I mention this not to castigate Foucault's anachronism. But to remind ourselves that the treatment of Hume in the eleventh lecture follows on Foucault's better known analysis of Becker and Stigler (1977) for who (famously/infamously) hold "the hypothesis of stable and uniform preferences." While I have  no evidence that Foucault read Becker & Stigler (1977, although we know he read papers/works by both), it figures prominently in a fascinating book by Henri Lepage Tomorrow, Capitalism: The Economics of Economic Freedom [first published in French as Demain le Capitalisme in 1978.) As the editors of Birth of Biopolitics note several times (but not in context), Foucault almost certainly read Lepage ahead of his lectures (sadly I have been unable to find any annotations by Foucault on Lepage). I mention this because the stability of preferences in Becker & Stigler (1977) is highlighted by Lepage in his account of human capital more generally (see p. 165 of the English translation). And human capital had been central to the previous lecture of Birth of Biopolitics.

In fact, and as a kind of aside, the 11th lecture is used to tie two major narrative threads in Birth of Biopolitics together: first, Foucault provides a historical eighteenth century framework for his own account of the development of utilitarianism from Bentham to Stigler/Becker. This framework has the shape of a donut in whose hole one can discern the shadow of Halévy's classic The Growth of Philosophic Radicalism. (Basically Foucault skips the heart of Halévy's narrative, and only changes the origin and terminus ad quem of philosophical radicalism/Utilitarianism.) Second, Foucault had claimed that this radical tradition is one of two traditions within liberalism that has an account of liberty. Strikingly, in the material I ommitted, Foucault alludes to this. (I return to that another time, too.)  Now, Foucault does not claim that for Hume preferences are stable. But Foucault's anachronism here alerts us to Hume's role in these two threads. This is exactly made explicit by a passage shortly after the one quoted above:

What I think is fundamental in English empiricist philosophy— which I am treating completely superficially—is that it reveals something
which absolutely did not exist before. This is the idea of a subject of interest, by which I mean a subject as the source of interest, the starting point of an interest, or the site of a mechanism of interests. For sure, there is a series of discussions on the mechanism of interest itself and what may activate it: is it self-preservation, is it the body or the soul, or is it sympathy? But this is not what is important. What is important is the appearance of interest for the first time as a form of both immediately and absolutely subjective will.
I think the problem and that which gets the problematic of homo oeconomicus underway is whether this subject of interest or form of will called interest can be considered as the same type of will as the juridical will or as capable of being connected to the juridical will. (273)

Be that as it may, in the present context 'non-transferable,' thus, refers to the brute fact that we cannot fundamentally feel another's pain or let them feel ours. Even sympathy's role is, as Foucault stipulates, a mechanism "to activate" or shape what we feel and thereby our interests not a means to share or trade one's feelings. Pleasures and pains are not portable from one subject to another. Even Adam Smith, who has a much more capacious understanding of the power of sympathy will grant this point. (For him mutual sympathy requires mutial modulation of our original feelings.) Now, as I mentioned above the point of the analysis is to connect Hume to the development of the formal features of homo oeconomicus as found in the radical tradition. Somewhat peculiarly, in Birth of Biopolitics Foucault shows no interest in Hume's economic writings (including a highly salient essay called "Of Interest.")

Now, the textual support that Foucault offers for all of this is rather fascinating (and that's what I'll end on today). As the editors of Birth of Biopolitics, who recognize in Hume's "famous aphorism" a reference to Treatise 2.3.3.6, and then (as I noted before) quote Hume: “Where a passion is neither founded on false supposition, nor chuses means insufficient for the end, the understanding can neither justify nor condemn it. ’Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger.” This is indeed a famous aphorism (one that clearly exercised Hume's friend Smith at The Theory of Moral Sentiments, 3.3.4, 136–7.)

But while Foucault gets the spirit of Hume's remarks at Treatise 2.3.3.6 right, he is not slavishly following the text of Hume. In Hume the famous aphorism, is introduced to argue that feelings are, in a certain sense, beyond rational scrutiny. However, I use in a 'certain sence' because they are not wholly beyond scrutiny as Hume explicitly notes. For they are often accompanied by judgments and these may well be scrutinized and found wanting. And, more interesting for present purposes, as the quote eveals, when the feeling itself is triggered by "false suppositions" say in cases of adaptive preferences or wishful thinking (or simply bad epistemic luck), or, when it is action guiding and it "chuses means insufficient for the end" and so leads to failures of instrumental rationality. 

Because in the context of 2.3.3.6, Hume is talking about passions more generally and not just about the bedrocks of pain and pleasure, I don't think Foucault really is misunderstanding Hume's general point. But it is a shame that he does not dwell on the specifics because, as I note, 2.3.3.6 is rather important for Hume's ideas on thinking on the contexts in which even a picture that takes preferences as irreducible and non-transferable, one need not treat them as merely given, nor beyond all critical scrutiny. That is to say, Foucault comes incredibly close, but somehow misses, that were the radical tradition and Utilitarianism more Humaen from the start some enduring problems internal to it could at least be diagnosed.

Foucault, Lepage, Anarcho-capitalism and anti-Marxism

In his fifth lecture of 7 February 1979 Foucault introduces "American anarcho-capitalism" as a synonym for "contemporary American liberalism." (The Birth of Biopolitics p. 104.) In context, the intended contrast is with German Ordoliberalism and Hayek's Austrian liberalism. In the next lecture, of 14 February, 1979, Foucault contrasts "contemporary American anarcho-capitalism" with the "classical liberalism of the nineteenth century." (p. 133) Later in the sixth lecture, Foucault claims that "American anarcho-capitalism" develops from the rejection of "Keynesian economics, and others from the Beveridge plans or European social security plans." (pp. 144-145) Foucault doesn't use the phrase "Anarcho-capitalism" again. He does use 'anarcho-liberal' (and its cognates) a few times (p. 117; p. 161) as apparent synonym. 

It's pretty clear, however, that he is thinking of the Chicago school, especially the strand inspired by Henry Calvert Simons (1899 –1946). Because Foucault makes clear that that "Simons drafted a number of critical texts and articles, the most interesting of which is entitled: “The Beveridge Program: an unsympathetic interpretation,” which there is no need to translate, since the title indicates its critical sense." (p. 216) In addition, Foucault had just introduced Simons to his audience as follows, "The first, fundamental text of this American neo-liberalism, written in 1934 by Simons, who was the father of the Chicago School, is an article entitled “A Positive Program for Laissez-Faire.” (p. 216) And, in fact, Foucault explicitly identifies 'anarcho-liberalism' with the Chicago school (p. 161).

Foucault's adoption of the phrase 'anarcho-capitalism' is a bit strange because Simons is no anarchist. And, in fact, to the best of my knowledge most real anarchists (Left and Right) would not treat any Chicago economist as an anarchist (since they tend to accept non-trivial state institutions). Milton Friedman's defense, for example, of central banking often gets him criticized by certain libertarians and anarchists alike. 

Unusually, the editorial team of The Birth of Biopolitics is silent on Foucault's use. I mention that not to criticize them. For, in a note attached to a cryptic remark by Foucault -- "The ideas of North on the development of capitalism, for example, are directly in line with this opening up made by the neoliberals" -- in lecture of 14 February 1979 they call to "H. Lepage, Demain le capitalisme (Paris: Librairie Générale Francaise, 1978; republished “Pluriel”) p. 34 and chapters 3 and 4 (this book was one of the sources used by Foucault in the last of these 1979 lectures)." Lepage treats North at length in this fascinating book, translated by Sheilagh C. Ogilvie* as Tomorrow, Capitalism: The Economics of Economic Freedom (Open Court, 1982).

Lepage introduces the term 'anarcho-capitalism' near the end of the first chapter ("The New Economists") in a section, "The Libertarians," as follows:

This renaissance in modern economic thought has given rise to the appearance on the American political chessboard of a new ideological movement—the Libertarian movement. It does not yet represent more than an intellectual fringe, but it embodies an astonishing reconciliation between a distinctly controversial social libertarianism (favoring legalization of drugs, moral freedom, freedom from military conscription, international neutrality, suppression of the CIA, abolition of the corrupt ‘‘military-industrial complex’’) and an economic libertarianism that calls for a radically laissez-faire capitalist society in which a maximum of transactions would take place through voluntary contractual exchange.

Often having close ties with the ‘‘radical”’ philosophy of the New Left and mostly less than 40 years of age, these libertarians are the direct descendants of the great libertarian revolt of the 1960s. They hold the very existence of the state to be the supreme evil that is to be actively resisted. Opponents of all manifestations of Gulag, down to the least offensive, their objective is to abolish the state in favor of a social system in which all public functions would be made private. This has given them the name ‘‘anarcho-capitalists,’’ a name assumed by certain of their leaders such as David Friedman, son of Milton Friedman, also a professor of economics, now at UCLA and formerly at the Virginia Polytechnic Institute and State University in Blacksburg." Lepage Capitalism, Tomorrow, p. 25

A few pages later Lepage clarifies that "Milton Friedman, for instance, passes for “‘libertarian’’ in most circles, but he is not recognized as libertarian by militants of the Libertarian Party such as his own son." (p. 27) And a few lines down, he is equally clear that "not all Chicago school adherents or the new economists are libertarians." (p. 27) It's natural to read this page as suggesting that Nozick and David Friedman are anarcho-capitalists. In the section on anarcho-capitalism (near the end of chapter 7) only David Friedman is mentioned and quoted (and no Chicago economist).+ 

I suspect Foucault gets the term 'anarcho-capitalism/liberalism' from Lepage. And it is pretty clear that Lepage's discussions of Becker, Stigler, and North shape Foucault's analysis (about this some time soon more). But Foucault does not follow Lepage slavishly (or may have conflated Milton and David Friedman). Another important difference between Foucault and Lepage is that while Lepage identifies Simons as one of the founders of the "Chicago tradition," he treats Knight (correctly) as the "true founder." (p. 5) Lepage devotes no space at all to discuss Simons (who was long dead when Lepage wrote his book). By contrast Foucault largely ignores Knight and focuses on Simons. 

Okay, Capitalism, Tomorrow (Demain, le capitalisme) is a superb survey of anti-Keynesian economics and political philosophy of the 50s-70s. (It was a bestseller in France.) In his "foreword" to the English translation, James Buchanan notes that Lepage visited in September 1976. Throughout the foreword Buchanan treats Lepage as a "journalist." But I understand from Guillaume Yon that Lepage wrote important policy briefs for the regulatory framework of French electricity industry. And in France Lepage quickly became associated with 'new economists' and 'anarcho-capitalism.' And I suspect that Foucault used 'anarcho-capitalism' to alert his audience that he would be talking about the whole ensemble of characters that Lepage discusses in his book, rather than conflate the more narrow work of David Friedman with the much broader movement that Lepage discusses.

One reason Douglass North is so important to Lepage's argument, and why I think Foucault mentions him in that otherwise obscure passage when he first introduces 'anarcho-capitalism,' is that running through Lepage's book are a whole number of set-pieces in which known criticisms by Marxists of capitalism, markets, and/or liberalism are answered (if not refuted) rather thoroughly. As Lepage puts it in his introduction, "capitalism and economic freedom cannot be defended successfully...[without] new theoretical and scientific insight into the economic bases and implications of their own political philosophy," (p. 3; emphasis in original.) In fact, Capitalism, Tomorrow would still work very well in an 'intro to PPE' course or as the right-wing response to (say) Polanyi's The Great Transformation (or some such book) or as an introduction to economic way of thinking.

An example of this occurs right at the start of the book. After briefly introducing the significance of human capital theory and the new theory of consumer behavior to his audience, Lepage writes,

Formulated separately by Gary S. Becker at Chicago and Kelvin J. Lancaster at Columbia, the new theory of consumption sees the consumer as an individual who not only consumes but also ‘‘produces’’ his own satisfaction, using inputs consisting of his market purchases and his time. This is in itself—as we shall see'’—an intellectual revolution, yet despite its controversial political implications it is still practically unknown outside economic circles. Its new view of consumer demand finally makes it possible to reply effectively to the critics of capitalism who base their view of society on a supposed distinction between ‘‘true”” and ‘‘false’’ needs. In their view, advertising, by changing people’s desires, makes consumers the slaves of the producers, and the proliferation of new products becomes proof of the wasteful and suicidal nature of consumer society. But consumer behavior can be explained convincingly without invoking ‘‘false needs’’ or consumer irrationality. (Capitalism, Tomorrow, Chapter one, pp. 9-10, emphasis added)

My present point is not to endorse Lepage's claim that the new theory of consumption does respond adequately to the Marxist criticism. To evaluate that we will need to explore the details of his argument (in the very near future!) But rather that responding to anti-capitalism systematically is clearly Lepage's purpose. And I suspect this is what made the book attractive to Foucault, who, as is well known, runs with the idea that what is distinctive about American neoliberalism (as exemplified by Becker and Stigler) is the "programming of policies of economic development, which could be orientated, and which are in actual fact orientated," toward the development of human capital. (Birth of Biopolitics, p. 233) And this programming can, as Foucault clearly discerns, be developed through social policy and one's own activities.

Finally, Becker's move is at odds (recall; see also here) with Frank Knight’s position who claimed that “wants which impel economic activity and which it is directed toward satisfying are the products of the economic process itself.” (This is a view that is not just Marxist, but goes back to Smith and Rousseau and probably Aristotle.) Arguably this is one of the central oppositions between 'old' and 'new' Chicago.  TBC.

 

 

Milton Friedman (Schumpeter, Foucault, Dicey, and Hayek) on the Possibility of a Free Society

THERE IS A STRONG TENDENCY for all of us to regard what is as if it were the “natural” or “normal” state of affairs, to lack perspective because of the tyranny of the status quo. It is, therefore, well, from time to time, to make a deliberate effort to look at things in a broader context. In such a context anything approaching a free society is an exceedingly rare event. Only during short intervals in man’s recorded history has there been anything approaching what we would call a free society in existence over any appreciable part of the globe. And even during such intervals, as at the moment, the greater part of mankind has lived under regimes that could by no stretch of the imagination be called free.

This casual empirical observation raises the question whether a free society may not be a system in unstable equilibrium. If one were to take a purely historical point of view, one would have to say that the “normal,” in the sense of average, state of mankind is a state of tyranny and despotism. Perhaps this is the equilibrium state of society that tends to arise in the relation of man to his fellows. Perhaps highly special circumstances must exist to render a free society possible. And perhaps these special circumstances, the existence of which account for the rare episodes of freedom, are themselves by their nature transitory, so that the kind of society we all of us believe in is highly unlikely to be maintained, even if once attained.--Milton Friedman (1962) "Is a Free Society Stable?" New Individualist Review.

In yesterday's post, I said something about the intellectual significance of New Individualist Review to the political development of what came to be known as 'fusionism' Stateside and its Hayekian context (and so the mainstreaming of neoliberal ideas, 'neoliberal' in its historical sense, that is). I also suggested that Friedman's essay (I quoted the first two paragraphs above) should be understood in context of the surprisingly critical reception of the Road to Serfdom thesis in this context. (Friedman's friend and colleague, Stigler, savages it.)

I wish I had read Friedman's essay before I blogged on Graeber and Wengrow's recent Dawn of Everything because one way to understand the significance of their book is that it challenges the idea that a genuinely free society is rare. Now, obviously Friedman and Graeber/Wengrow don't mean the exact same thing with a 'free society.' While Friedman is a right or 'individualist' libertarian he is clearly not an anarchist, so it's not the case that Graeber/Wengrow refute Friedman's claim here. (And it's pretty clear that Graeber & Wengrow would not accept Friedman's claim that we're living in a free society.) But since Graeber/Wengrow and Friedman agree that legally sanctioned coercion is itself generally problematic (in ways that mainstream liberals often think not, alas), my suggestion is not altogether farfetched. Because it's pretty clear that the many urban polities that Graeber and Wengrow admire are open to commerce while rejecting state organized/sanctioned/rent-seeking (etc.) capitalism (which is what draws Friedman's ire, too).

So, one way to react to Friedman's set-up of his argument is that he is (tacitly) working with a biased sample or a historically limited data-set (something Friedman may well appreciate given his kind of research in monetary history) because, and this is the interesting part, he is drawing on a literature (shared by Marxists and certain classical liberals alike) that presupposes something like an idea of progress (that has already occurred) and that is distinctly Eurocentric/Western. This can be checked because Friedman's main source is Lectures on Law and Public Opinion in the Nineteenth Century (1905) by A. V. Dicey. Dicey clearly impressed Friedman because he also frames Friedman's argument in his important earlier (1952) essay "Neo-Liberalism and its Prospects” that I have discussed here.

What makes Friedman's "Is a Free Society Stable?" interesting are  two important observations. First, and this one (I am pretty sure) he has repeated elsewhere but here clearly credits Dicey with noticing, is that the consequentialist arguments that tend to support classical liberalism all violate what we might call 'naïve empiricism' or 'lived experience' because they require a kind of statistical or aggregate perspective:

The argument for a free society, he goes on to say, is a very subtle and sophisticated argument. At every point, it depends on the indirect rather than the direct effect of the policy followed. If one is concerned to remedy clear evils in a society, as everyone is, the natural reaction is to say, “let’s do something about it,” and the “us” in this statement will in a large number of cases be translated into the “government,” so the natural reaction is to pass a law. The argument that maybe the attempt to correct this particular evil by extending the hand of the government will have indirect effects whose aggregate consequences may be far worse than any direct benefits that flow from the action taken is, after all, a rather sophisticated argument. And yet, this is the kind of argument that underlies a belief in a free or laisser-faire society.

So, that also means (and this is Dicey's initial observation) that arguments for a free society have a kind of utopian quality ahead of their existence. And even while such societies exist, given the abstracted/philosophical nature of the argument it is "says Dicey...really amazing and surprising...that for so long a period as a few decades, sufficiently widespread public opinion developed in Britain in favor of the general principle of non-intervention and laisser-faire as to overcome the natural tendency to pass a law for the particular cases." (The natural tendency is, thus, rooted in what I called the naïve empiricist assumption which is attributed to ordinary life.) Obviously, as an aside, Marxists would take Dicey's argument, if they cared to take it seriously, and suggest that it is violence (and then ideology) that has to do the job to get people to accept a free society.

I don't mean to suggest such utopianism is required. For, as Friedman notes there is another route that is at least partially ground in experience:

one of the factors that led Bentham and the Utilitarians toward laisser-faire, and this is a view that is also expressed by Dicey, was the self-evident truth that if you wanted to get evils corrected, you could not expect to do so through the government of the time. The government was corrupt and inefficient. It was clearly oppressive. It was something that had to be gotten out of the way as a first step to reform. The fundamental philosophy of the Utilitarians, or any philosophy that puts its emphasis on some kind of a sum of utilities, however loose may be the expression, does not lead to laisser-faire in principle. It leads to whatever kind of organization of economic activity is thought to produce results which are regarded as good in the sense of adding to the sum total of utilities. I think the major reason why the Utilitarians tended to be in favor of laisser-faire was the obvious fact that government was incompetent to perform any of the tasks they wanted to see performed.

So, you might say that on Friedman's view Bentham and the radicals that followed backed themselves into laisser-faire as a kind of second best (or n-th) best option. What's undoubtedly correct here is that there is no intrinsic connection between free society and Utilitarianism. But Friedman ignores how important Adam Smith (and Hume) is to Bentham's intellectual development and the more important political radicals.

The second really interesting observation by Friedman is set up by him noting that "Dicey’s argument is enormously strengthened by an asymmetry between a shift toward individualism and a shift away from it." (For all his admiration, Friedman doesn't mince words about a kind of Victorian complacency in Dicey.) By contrast, in virtue of accepting the rarity of a free society (and a kind of explanatory parsimony), Friedman treats the mechanisms one might posit in a road to serfdom thesis as symmetrical (even potentially identical) with those that lead to a free society. Friedman is motivated to do so, also, by his reflection on Schumpeter's acceptance of the Marxist idea that capitalism is doomed to failure (but with this twist that "[W]hereas Marx’s view was that capitalism would destroy itself by its failure, Schumpeter’s view was that capitalism would destroy itself by its success."). 

As an aside, if you accept my claim that Hayek is Friedman's real target in his 1962 article (which I claim is plausible on contextual grounds), what's neat about this is that he inscribes the road to serfdom thesis in a larger debate, inspired by Marxism, about the historical necessity of capitalism, and its subsequent demise. And in Friedman (recall here and here) not unlike Foucault in Lecture 7, The Birth of Biopolitics, 21 February 1979. Birth of Biopolitics, Schumpeter is crucial for linking the debates about monopoly capitalism and the road to serfdom thesis. (And unlike Foucault, Friedman, in turn, treats Schumpeter as a variation on Veblen and Burnham as a variation on Schumpeter--it is really a virtuoso performance by Friedman.) Some other time I will address Friedman's criticism of Schumpeter's thesis.

The mechanism, which has some unacknowledged roots in Marx's analysis, is as follows:

the development of laisser-faire laid the groundwork for a widespread respect for the law, on the one hand, and a relatively incorrupt, honest, and efficient civil service on the other, both of which are essential preconditions for the operation of a collectivist society. In order for a collectivist society to operate, the people must obey the laws and there must be a civil service that can and will carry out the laws. The success of capitalism established these preconditions for a movement in the direction of much greater state intervention.

But then Friedman adds:

The process I have described obviously runs both ways. A movement in the direction of a collectivist society involves increased governmental intervention into the daily lives of people and the conversion into crimes of actions that are regarded by the ordinary person as entirely proper. These tend in turn to undermine respect for the law and to give incentives to corrupt state officials....the erosion of the capital stock of willingness to obey the law reduces the capacity of a society to run a centralized state, to move away from freedom. This effect on law obedience is thus one that is reversible and runs in both directions. It is another major factor that needs to be taken into account in judging the likely stability of a free system in the long run.

And this argument is directed against both Marxist necessity and the road the serfdom. Because countervailing forces are naturally developed as one goes down the road to serfdom: "Once government embarks on intervention into and regulation of private activities, this establishes an incentive for large numbers of individuals to use their ingenuity to find ways to get around the government regulations." (And this, in turn, starts to swallow up time of the more earnest government officials, etc.) Friedman recognizes his claims are speculative (much more historical work is needed), but a reader of Friedman comes away thinking that free societies are rarely stable, if one means by that, in equilibrium; that is, they are always in some kind of crisis (again this anticipates Foucault's idea that liberalism is co-constituted with crisis); but these crises of a free society have, on the whole also a self-correcting quality over the medium term.

It's easy to mock Friedman's position as I have reconstructed it as a kind of providential optimism. But perhaps such providentialism is required within political life; the great danger of it is that encourages complacency about the possibility of a crisis that fatally undermines a free society.

 

 

Foucault, Stigler, Weber, and some pre-history going back to Thomas More, Mandeville, Grouchy and even Spinoza

The cost limitation upon the enforcement of laws would prevent the society from forestalling, detecting, and punishing all offenders, but it would appear that punishments which would be meted out to the guilty could often be increased without using additional resources. The offender is deterred by the expected punishment, which is (as a first approximation) the probability of punishment times the punishment-$100 if the probability of conviction is 1/10 and the fine $1,000. Hence, increasing the punishment would seem always to increase the deterrence. Capital punishment is cheaper than long term imprisonment; and seizure of all the offender's property may not be much more expensive than collecting a more moderate fine.

To escape from this conclusion, Becker introduces as a different limitation on punishment the "social value of the gain to offenders" from the offense. The determination of this social value is not explained, and one is entitled to doubt its usefulness as an explanatory concept: what evidence is there that society sets a positive value upon the utility derived from a murder, rape, or arson? In fact the society has branded the utility derived from such activities as illicit. It may be that in a few offenses some gain to the offender is viewed as a gain to society,' but such social gains seem too infrequent, small, and capricious to put an effective limitation upon the size of punishments.

Instead we take account of another source of limitation of punishment, which arises out of the nature of the supply of offenses. It is no doubt true that the larger the punishment, the smaller will be the expected net utility to the prospective offender from the commission of a given offense. But marginal decisions are made here as in the remainder of life, and the marginal deterrence of heavy punishments could be very small or even negative. If the offender will be executed for a minor assault and for a murder, there is no marginal deterrence to murder. If the thief has his hand cut off for taking five dollars, he had just as well take $5,000. Marginal costs are necessary to marginal deterrence.2 The marginal deterrence to committing small crimes is also distorted if an otherwise appropriate schedule of penalties is doubled or halved. One special aspect of this cost limitation upon enforcement is the need to avoid overenforcement. The enforcement agency could easily apprehend most guilty people of we placed no limits upon the charging and frequent conviction of innocent people. In any real enforcement system, there will in fact be conviction and punishment of some innocent parties, and these miscarriages of justice impose costs of both resources and loss of confidence in the enforcement machinery. The costs of defense of innocent parties, whether borne by themselves or by the state, are part of the costs of enforcement from the social viewpoint. The conviction of innocent persons encourages the crime because it reduces the marginal deterrence to its commission.--Stigler, George J. "The optimum enforcement of laws." Journal of Political Economy 78.3 (1970): 527-528.

Foucault quotes the Stigler (1970) article in the tenth lecture of 21 March 1979 (Birth of Biopolitics, p. 256.) It is quite natural, if you read the editorial apparatus of the Birth of Biopolitics, to think that Foucault may not have read the original source, but that he is relying entirely on Jenny, F. "La théorie économique du crime: une revue de la littérature." Vie et sciences économique 73 (1977): 7-20. The wording of the editorial footnote (18) -- "(Foucault draws on information provided in this article.)" -- is sufficiently ambiguous to give the editors plausible deniability here. But the editors go on to imply it in a later footnote (25) when they tell the readers that "This phrase is not found in Becker’s article. Foucault relies on the synthesis of Becker’s and Stigler’s works provided by Jenny." Foucault's vexed relationship to underlying sources came up in discussion over lunch with Bruce Caldwell and Joel Isaac yesterday.

So, I was pleasantly surprised yesterday when during a hallway conversation Vincent Carret -- yes I am visiting Duke's Center for the history of political economy [hence all the name-dropping] -- informed me that there is a project, Le projet Foucault Fiches de Lecture (FFL), that is collecting and making digitally available many of Foucault's annotations on his readings which he did to prepare his lectures and books. So, when I went to check it out, I thought it made sense to use it as my first foray to see what annotations he had left on Becker and Stigler. While zero annotations would not be decisive -- after all notes get lost and re-used, maybe he relied on others for summaries and checked them against original sources -- it would naturally be taken to reinforce the editors' point.

To give a non-trivial example that a false negative is not decisive: as far as I can tell, there are no annotations on Jenny's article (or the underlying book in which it appeared). Since there are Foucault annotations on papers by Becker (here) and Stigler, that would have made a nice punchline for this digression, but I have other fish to fry!  In addition, the annotations do not seem to be dated, and as far as I can tell [after a few hours of poking around on the website] the project has not yet tried to date the manuscripts. So, I am not going to try to connect Foucault's scribbling to particular passages in the lecture course. (I will leave that to a different kind of scholar.) Foucault read and made a few annotations (see here) on Stigler's "The Optimum Enforcement of Laws." That should be no surprise because Foucault had a deep interest in criminology, and its history, and Becker's and Stigler's work were creating important waves in the field. (I would be amazed if these annotations were written after the Birth of Biopolitics lectures.)

When Stigler wrote the passage I have quoted, Becker may well still have been at Columbia. (Becker returned to Chicago in 1970.) And, at first sight the passage has the aura of a senior scholar setting the young upstart straight on marginal analysis (then exactly a century old), although the real disagreement is to explain the absence of widespread and enormous numbers of capital punishment. For, while Stigler's (and Becker's) project has a normative component (Stigler is explicit about this), it's also meant to do a decent job with the empirical world.

And, in fact, Stigler's analysis ("If the thief has his hand cut off for taking five dollars, he had just as well take $5,000.") echoes pre-marginalist thinking going back to Thomas More (in part I of Utopia [recall this NewAPPS post]), and, as Foucault explicitly notes (repeatedly in the lecture), in the writings of Beccaria and Bentham (and (recall this digression on) Sophie Grouchy). This is why Foucault treats Stigler and Becker as an updated version of Benthamite utilitarianism (although as the reference to More suggests cost-benefit analysis of criminal decision making and how to create rational punishment in light of it precedes Benthamite utilitarianism). 

Now, I want to close with more substantive observations. Foucault himself makes an important point in a passage (which I have quoted in passing before (recall here) in which he explicitly quotes Stigler (1970):

The regulatory principle of penal policy is a simple intervention in the market for crime and in relation to the supply of crime. It is an intervention which will limit the supply of crime solely by a negative demand, the cost of which must obviously never exceed the cost of the supply of the criminality in question. This is the definition that Stigler gives of the objective of a penal policy: “The goal of law enforcement,” he says, “is to achieve a degree of compliance with the rule of prescribed behavior that society believes it can procure while taking account of the fact that enforcement is costly.” This is in the Journal of Political Economy in 1970.30--Foucault 21 March, 1979,  translated by Graham Burchell, Lecture 10, The Birth of Biopolitics, p. 256.*

The 'definition that Stigler gives' is from the first section of the essay labeled 'The Goal of Enforcement.' Stigler returns to the regulatory principle in the final (third) section labeled 'The Enforcement Agency: A Normative Approach.' Some other time I want to discuss how we should think about the normativity involved in Stigler's analysis. Foucault does not remark on it. It's worth nothing that 'Chicago' is by no means the first to think about the role of enforcement costs in law enforcement: Spinoza (in the Theological Political Treatise) and Mandeville (in the Fable) recognize that Amsterdam's practice of tolerating brothels is explicitly shaped by consideration of enforcement costs.

Now, 'the art of government' is the major theme of Birth of Biopolitics. And the phrase (and its cognates) are used repeatedly throughout the lectures. But one peculiarity of the lectures is that the phrase does not occur in the lectures on Chicago economics ('governmentality' does in case you want to play gotcha with me). And this has let commentators, I think, to miss that Chicago school does have an account of the art of government: in Friedman it's called the 'art of economics' (which is recall here and here the phrase he inherits from J.S. Mill and J.N. Keynes) and Stigler with less flair, calls it "advice on policy...[in light of] understanding of the regulatory process" in the introduction to his (1975) Citizen and the State. Now, in Friedman (and Mill, and even Adam Smith) the content of the art of government is articulated in terms of maxims or principles. Foucault echoes this language when he describes Stigler's 1970 in terms of 'regulatory principles.'** 

To put my first important point simply: in the context of the modern administrative state, the liberal art of government, as understood at 'Chicago' is partially codified in terms of regulatory principles that constrain how one should think about the relationship between policy aims and fundamental liberal principles (and constitutional constraints, etc.) in light of experience with the ordinary workings of the democratic political process which is often also a market in political regulation. 

The second point is that Foucault uses 'regulatory principle' as the content of the art of government in the context of the modern administrative state. This is actually confirmed or illustrated in an interview held with him (by folk who are actually confused on this) in a passage that has attracted quite a bit of critical attention (because it involves Foucault's attitude toward consensus politics), where Foucault corrects his interviewers' terminology by saying that 'X is a regulatory principle' always presupposes that X is subject to "governance" and to "be organized." (The Foucault Reader, p. 378)

And lurking here is the most important point often lost in twentieth century debates over markets vs states is that much of neoliberalism, that, as Foucault correctly recognizes, is about -- and Becker and Stigler are explicit about this -- a process of rationalization of the administrative state on liberal principles not its abolition. And so while Foucault repeatedly ties the ORDO liberal approach to Max Weber's shadow, he could have also done so with Chicago (even if Stigler got his Weber through Knight and Talcott Parsons).