Digressions&Impressions

14 March 1979: Foucault on Programming Human Capital (XXVIII)

I would like to bring out some aspects of American neo-liberalism, it being understood that I have no desire and it is not possible to study it in all its dimensions. In particular, I would like to consider two elements which are at once methods of analysis and types of programming, and which seem to me to be interesting in this American neo-liberal conception: first, the theory of human capital, and second, for reasons you will be able to guess, of course, the problem of the analysis of criminality and delinquency.
First, the theory of human capital. I think the interest of this theory of human capital is that it represents two processes, one that we could call the extension of economic analysis into a previously unexplored domain, and second, on the basis of this, the possibility of giving a strictly economic interpretation of a whole domain previously thought to be non-economic.
First, an extension of economic analysis within, as it were, its own domain, but precisely on a point where it had remained blocked or at any rate suspended....Starting from this criticism of classical economics and its analysis of labor, the problem for the neo-liberals is basically that of trying to introduce labor into the field of economic analysis. A number of them attempted this, the first being Theodore Schultz, who published a number of articles in the years 1950–1960 the result of which was a book published in 1971 with the title Investment in Human Capital. More or less at the same time, Gary Becker published a book with the same title, and then there is a third text by Mincer, which is quite fundamental and more concrete and precise than the others, on the school and wages, which appeared in 1975.--Michel Foucault, 14 March, 1979,  translated by Graham Burchell, Lecture 9, The Birth of Biopolitics, 219-220.

Perhaps, some other time I return to neoliberal history of classical economics that Foucault summarizes in the part I skipped in this quote. And in what follows I also ignore Foucault's acuity in grasping the inner meaning, as it were, of intellectual trends in a different rather technical discipline in real time. 

In the first quoted paragraph, Foucault introduces the central theme of the lecture and, perhaps, of the whole book without alerting his audience (or himself) that he has done so. There are features of American neo-liberalism (in his discussion he mostly confines himself to Chicago economics) that are "methods of analysis and types of programming." That papers and books in economics contain methods of analysis is no surprise. But Foucault insists, without initially making a big deal about it, that they are simultaneously types of programming. And so the question is what (or who) is being programmed and what counts as a program in this context?

While Foucault uses programming (and its cognates) throughout the lecture, Foucault only offers his answer in the final two sentences of lecture nine: "programming of policies of economic development, which could be orientated, and which are in actual fact orientated," toward the development of human capital. (233) So, it is humans and, perhaps, whole societies that are to be programmed. The purpose of this programming can be varied. For, example the focus on developing human capital is to stimulate economic development (locally and internationally) from which many things may follow. (The analysis of delinquency may serve other ends.) 

I do not know if in the 1970s, a French audience, would have heard 'programming' as the thing we do to write software or make computers and robots run. Perhaps, they heard it terms of scheduling. Or perhaps they heard it as an ironic counterpart to the programmatic political projects ordinarily associated with five year social planning. But it is quite clear that Foucault treats these economists not merely as offering (who?) new kinds of descriptions of society, but also offering a new kind of praxis. 

Foucault's claim here sits uneasily with the self-understanding of many Chicago economists of the period; they, following Milton Friedman's lead , generally claim that they make a sharp description between the descriptive ('positive') and prescriptive ('normative') elements in their work; while leaving practice ('art') largely to the side. So, for example, 

THE new economic approach to political behavior seeks to develop a positive theory of legislation, in contrast to the normative approach of welfare economics. The new approach asks why certain industries and not others become regulated or have tariffs imposed on imports or why income transfers take the form and direction they do, in contrast to asking which industries should be regulated or have tariffs imposed, or what transfers should be made.
Both the normative and positive approaches to legislation, however, generally have taken enforcement of laws for granted, and have not included systematic analyses of the cost of enforcing different kinds of laws. In separate studies we recently formulated rules designed to increase the effectiveness of different laws. We proposed that offenders convicted of violating laws be punished by an amount related to the value of the damages caused to others, adjusted upwards for the probability that offenders avoid conviction. Gary S. Becker & George J. Stigler. "Law enforcement, malfeasance, and compensation of enforcers." The Journal of Legal Studies 3.1 (1974): 1.*

Stigler and Becker (both Chicago Nobels eventually) present themselves as doing something on the 'positive'/descriptive side of things. In their work they explore the costs of enforcements. Yet, even if one grants that their work is firmly grounded in empirical analysis, it is not so odd that Foucault thinks they are doing what he calls programming. For, they offer a rule of punishment intended to decrease 'cost' of punishment and increase the 'effectiveness' of these. To an outsider that does seem prescriptive.  Why do Becker and Stigler not call that normative analysis? To an outsider this seems like obscurification. 

The beauty of Foucault's treatment in lecture 9, is that he shows that part of the confusion stems from the way Robbins' definition of what economists do gets (ahh) operationalized in economics. I quote Foucault (initially quoting Robbins):

“Economics is the science of human behavior as a relationship between ends and scarce means which have mutually exclusive uses.” You can see that this definition of economics does not identify its task as the analysis of a relational mechanism between things or processes, like capital, investment, and production, into which, given this, labor is in some way inserted only as a cog; it adopts the task of analyzing a form of human behavior and the internal rationality of this human behavior. Analysis must try to bring to light the calculation—which, moreover, may be unreasonable, blind, or inadequate—through which one or more individuals decided to allot given scarce resources to this end rather than another. Economics is not therefore the analysis of processes; it is the analysis of an activity. So it is no longer the analysis of the historical logic of processes; it is the analysis of the internal rationality, the strategic programming of individuals’ activity. (222-223)

Economics stops being the large scale study of historical processes. Rather, economics becomes the study of the scarce means given ends. As Foucault notes in the passage I quoted at the top of this post, this definition is topic neutral. And facilitates what (already (recall) back in 1934) Talcott Parsons called 'economic imperialism' (that is, "the extension of economic analysis into a previously unexplored domain, and second, on the basis of this, the possibility of giving a strictly economic interpretation of a whole domain previously thought to be non-economic.") The setting of ends is treated as normative or prescriptive activity. And so when one studies means in light of scarcity one is doing descriptive work.

So, in the just quoted passage, the rules offered by Stigler and Becker are rules that govern the means, but not the ends criminal laws are supposed to serve. They are a form of instrumental rationality or conditional/hypothetical rationality. 'Programming' is an excellent term to capture what is going on here because these 'rules' offered by Becker and Stigler are, in fact, supposed to be legally binding on subjects.

Now, a critic may well worry about the immaculate conception of the ends. Who decides what they are? And by what right do Stigler and Becker assume them as given? (Answering these questions has informed my own work on them.) But Foucault, however notices or is interested in something else. (I think this is really spectacular.) I continue with Foucault's analysis of human capital theory:

This means undertaking the economic analysis of labor. What does bringing labor back into economic analysis mean? It does not mean knowing where labor is situated between, let’s say, capital and production. The problem of bringing labor back into the field of economic analysis is not one of asking about the price of labor, or what it produces technically, or what is the value added by labor. The fundamental, essential problem, anyway the first problem which arises when one wants to analyze labor in economic terms, is how the person who works uses the means available to him. That is to say, to bring labor into the field of economic analysis, we must put ourselves in the position of the person who works; we will have to study work as economic conduct practiced, implemented, rationalized, and calculated by the person who works. What does working mean for the person who works? What system of choice and rationality does the activity of work conform to? As a result, on the basis of this grid which projects a principle of strategic rationality on the activity of work, we will be able to see in what respects and how the qualitative differences of work may have an economic type of effect. So we adopt the point of view of the worker and, for the first time, ensure that the worker is not present in the economic analysis as an object—the object of supply and demand in the form of labor power—but as an active economic subject. (223)

To put the point as a paradox: the very theory that seems to objectify humans as capital and that is often accused of treating citizens as optimizing paws in engineering problems, is in fact a theory that treats humans as subjective agents. Now, as my former colleague John Davis has shown in a number of books, as this theory gets developed (in ever more subtle mathematical ways), it becomes difficult to see where, in the effervescent moments of a second derivative, agency ends up being located. But that's for a different time.

Admittedly, the cognitive content of such agency is reductive in character because human motivation is treated in fairly simple fashion. Quoting Foucault again,

People like Schultz and Becker say: Why, in the end, do people work? They work, of course, to earn a wage. What is a wage? A wage is quite simply an income. From the point of view of the worker, the wage is an income, not the price at which he sells his labor power. Here, the American neo-liberals refer to the old definition, which goes right back to the start of the twentieth century, of Irving Fisher, who said: What is an income? How can we define an income? An income is quite simply the product or return on a capital. Conversely, we will call “capital” everything that in one way or another can be a source of future income. Consequently, if we accept on this basis that the wage is an income, then the wage is therefore the income of a capital. Now what is the capital of which the wage is the income? Well, it is the set of all those physical and psychological factors which make someone able to earn this or that wage, so that, seen from the side of the worker, labor is not a commodity reduced by abstraction to labor power and the time [during] which it is used. Broken down in economic terms, from the worker’s point of view labor comprises a capital, that is to say, it as an ability, a skill; as they say: it is a “machine.”And on the other side it is an income, a wage, or rather, a set of wages; as they say: an earnings stream.
This breakdown of labor into capital and income obviously has some fairly important consequences. First, if capital is thus defined as that which makes a future income possible, this income being a wage, then you can see that it is a capital which in practical terms is inseparable from the person who possesses it. To that extent it is not like other capitals. Ability to work, skill, the ability to do something cannot be separated from the person who is skilled and who can do this particular thing. In other words, the worker’s skill really is a machine, but a machine which cannot be separated from the worker himself, which does not exactly mean, as economic, sociological, or psychological criticism said traditionally, that capitalism transforms the worker into a machine and alienates him as a result. We should think of the skill that is united with the worker as, in a way, the side through which the worker is a machine, but a machine understood in the positive sense, since it is a machine that produces an earnings stream. An earnings stream and not an income, precisely because the machine constituted by the worker’s ability is not, as it were, sold from time to time on the labor market against a certain wage. In reality this machine has a lifespan, a length of time in which it can be used, an obsolescence, and an ageing. So that we should think of the machine constituted by the worker’s ability, the machine constituted by, if you like, ability and worker individually bound together, as being remunerated over a period of time by a series of wages which, to take the simplest case, will begin by being relatively low when the machine begins to be used, then will rise, and then will fall with the machine’s obsolescence or the ageing of the worker insofar as he is a machine. We should therefore view the whole as a machine/stream complex, say the neo-economists—all this is in Schultz is it not—it is therefore a machine-stream ensemble, and you can see that we are at the opposite extreme of a conception of labor power sold at the market price to a capital invested in an enterprise. This is not a conception of labor power; it is a conception of capital-ability which, according to diverse variables, receives a certain income that is a wage, an income-wage, so that the worker himself appears as a sort of enterprise for himself. Here, as you can see, the element I pointed out earlier in German neo-liberalism, and to an extent in French neo-liberalism, is pushed to the limit, that is to say, the idea that the basic element to be deciphered by economic analysis is not so much the individual, or processes and mechanisms, but enterprises. An economy made up of enterprise-units, a society made up of enterprise-units, is at once the principle of decipherment linked to liberalism and its programming for the rationalization of a society and an economy. (224-225)

So Foucault understands human capital theory as a theory focused on agents, which are simultaneously programmable machines, that is, as possible generators of income streams. Given the recent focus on risk and uncertainty, and the many social and political challenges (pandemic, climate, financial instability) which lay in the future, it is natural to wonder about how we should think of possible income of these machines with, as the logicians teach, necessarily finite time spans. And this is also natural because Foucault (correctly) treats Chicago economics as a return to the radical utilitarian program of Bentham (248) and given that utilitarianism is as a technique of decision-making resolutely oriented toward the future (and spectacularly useful to forget crimes from the past).

Even so, agents don't fall like manna from heaven ready to produce income streams; they need to be educated and cultivated, and constituted by norms and institutions. Capital-ability is the product of history, and the machine-stream ensemble is built over time in society. So, lurking in this theory is a complex analysis relating past and (possibly discounted) future time. And while from one perspective wages merely reflect supply and demand at a given time, in other perspective they become a way to connect individual decisions over time to larger social ends (e.g., development). And as Foucault recognizes, the irony is that a relatively libertarian, intellectual community with distrust of collective planning, pioneers a set of analytical techniques by which policy-makers may come to shape (if not program) human agents toward collectively imposed ends.  

 

Climate Change: Central Bankers are Worried about Insurance

More frequent or severe extreme weather events and/or a late and abrupt transition to a low-carbon economy could have significant impacts on the financial system, with potential systemic consequences. Extreme weather events could lead to damage of physical assets, including real estate, productive capital and infrastructure, and loss of life with consequent property and casualty insurance losses, damage to balance sheets of households and firms, increases in defaults, and potential financial sector distress. A late and abrupt transition to a low-carbon economy could lead to a sudden repricing of climate-related risks and stranded assets, which could negatively impact the balance sheets of financial institutions....

Real estate and agriculture suggest themselves as sectors that are both particularly important and more immediately exposed to physical impacts of climate change, which could affect banks and insurers exposed to these sectors on both the asset and liability sides of their balance sheet. Furthermore, as physical risks rise or become more unpredictable, insurers will most likely increase premiums or stop insuring some risks. The implied decrease in coverage leads to increased uninsured losses (‘protection gap’) in case of a catastrophic event which could negatively impact the collateral value of properties, leaving banks exposed to credit risk. Understanding the trends in insurability, the cost of insurance, the protection gap, and their spillovers to the banking sector and the economy more broadly is an important direction of research. This requires an analysis of insurers’ role and behaviour, including the effectiveness of the catastrophic risk market. (p. 4)  "The Macroeconomic and Financial Stability Impacts of Climate Change Research Priorities," June 2020  NGFS: Central Banks and Supervisors Network for Greening the Financial System[HT Anne Kervers]

The "NGFS Technical document" quoted above is worth reading because, while it names no names, it is a forthright exploration of the limitations of existing scientific approaches to understanding the impact of climate change on the financial system and it contains several calls for more research in lots of areas that have become politically and morally salient. Here I am interested in using it as a granular, but useful window into what we may call the central bankers' consensus view about how to think about the significance of climate change to their jobs. So, before I get to insurance and reinsurance, let me just provide a two paragraph summary of that.

What is immediately noticeable is that the central bankers view their job as handling, that is anticipating and mitigating, the impact of "shocks" to the economy. And it is quite clear that they view the economy as a kind of autonomous system with a trendline, and their role is to mitigate and ameliorate the deviations to that trend line  caused by shocks external to the trend line. Behind the fancy language we can still discern the  idea that ordinary "fluctuations in economic activity are seen as the result of small, white noise shocks-impulses-that affect the economy through a complex dynamic propagation system," (127) and that major fluctuations have to be understood in terms of large demand or supply shocks.*

Crucially, in the central bankers' perspective, "demand shocks are typically manageable from a monetary policy perspective, [while] supply shocks are generally more challenging as they generate a trade-off for central banks between stabilising inflation and stabilising output fluctuations." (6)  The significance of climate change is viewed through the lense of a destabilising series of supply shocks that potentially overwhelm monetary policy. Strikingly, in the report this is viewed as a danger to the central banks' ability to fight inflation (6-9). Somehow, and this is remarkable in light of recent experience, the risks to deflation are apparently not worth noting at all.** 

There is, of course, plenty be said about the utility and normative significance of treating the economy in terms of deviations from a trendline, but let's turn to quoted passage(s). The first thing to notice is that the 2020 report highlights insurance and reinsurance. I noticed it before looking at last year's report, which is longer (and more detailed engagement with scientific literature); there insurance and reinsurance are entirely absent. Second, central bankers admit that from their perspective, at the moment, some of the best research on the effects of climate change on the economy is taking place in the insurance world, especially, I gather, integrating climate modelling with economic and geographic analysis. ("To this end macroeconomic modellers could borrow well-advanced methodologies used by (re-)insurance firms to quantify the economic impact of physical risks from extreme weather events, including spatial analysis.") I know nothing of this research, but will try to learn about it before long.

Third, the central bankers explicitly worry that in light of climate change insurance and reinsurance (henceforth when I use 'insurance'  and its cognates I mean both insurance and reinsurance) will become more expensive and, subsequently, cover less of economic activity. I think they are identifying a number of distinct concerns. And it may be useful to list them. [I] If insurers misjudge the underwriting risks because they are now in the realm of true uncertainty ["unpredictable"], they may go bankrupt. Since historical data are becoming less useful, underwriting is becoming more art than science reliant on the sensitivity to assumptions in complex modeling. If major insurers cannot pay out, or are perceived not to be able to pay out, then significant bits of the financial and real economy are exposed to serious damage [" financial sector distress."], if not collapse. Since AIG this is not just science fictions.

In fact, [II] the central bankers are worried about the future of insurance as a profitable enterprise ["longer-term climate change implications for the profitability/viability" (6)] What they do not say is that most risk modeling and portfolio management in finance presupposes that profitable insurance is possible. So, the significance of this concern is much wider.

[III] As premiums increase the central bankers assume that ordinary firms and individuals will be exposed to inflationary pressures and the decision to forego coverage. Strikingly the central bankers do not assume (a) that these will stop economic activity in the exposed areas; or (b) that governments will step in to provide insurance or to mandate/subsidize insurance. On (a) this means the central bankers assume that say due to path dependencies economic agents will persist in their activity without insurance and will tolerate considerable more uncertainty in their lives. And this will percolate through the system because a lack of insurance in one sector will impact its interactions with others. On (b) this strikes me as unrealistic. Governments will be lobbied extensively. Of course, at some point governments will have to make tough decisions about which parts of the economy to cover or not.

[IV] Central bankers are still very worried about bank balance sheets. Here the main concern is that the collateral that itself becomes worthless and is under-uninsured. It's one thing if the waves topple a hotel tower in Southern Florida and makes a loan non-performing; another if there is no insurance to pay back the principle of the underwriter. That leads to write-downs and deleveraging, etc. ["the collateral value of properties, leaving banks exposed to credit risk. Understanding the trends in insurability, the cost of insurance, the protection gap, and their spillovers to the banking sector and the economy more broadly "]

[V] The central bankers assume that the effects of climate change are not felt uniformly ["climate change could impact geographical/economic areas differently within a country or currency area." (7)] In particular, throughout the document they show awareness of the significance of  (what is known as) spatial economics; and that their existing models and policy instruments sit uneasily with a spatial or geographic approach to the financial system.

Finally, [VI] it is noticeable that whole areas of insurance that impact the vast majority of consumers and citizens directly (medical, car, travel, home-owner/renters, etc.) are either discussed obliquely or not mentioned at all. And while in many cases this will not matter much, it is important that somebody alerts the central bankers to the risks of cognitive capture by various industries. In the end they must be servants of the wider public, too. 

To sum up. Catastrophic risk insurance is a part of the financial system that generally is ignored by the public. Given the centrality of insurance in modern political economy, it is a good thing that central bankers are starting to reflect on what happens when agents start to wonder if they cannot count on insurance in a crisis.  The danger is, of course, this will very thought accelerates the possibility.

 

 

On Ellsberg's defense of Knightian Uncertainty, Arrow, and Ambiguity (II)

In most of this literature, the term ‘ambiguity’ has been treated as a synonym for what Knight called ‘uncertainty’ namely the fact that relative likelihoods are not characterized by well-defined numerical probabilities. The standard method of dealing with ambiguity in decision theory is to endow the decisionmaker with multiple priors as in Gilboa and Schmeidler (1989). This approach may be combined with a variety of preference models, notably including the maxmin model of Gilboa and Schmeidler (1989) and the smooth model of Klibanoff et al. (2005).
For a non-specialist this is puzzling; there is no obvious link to the ordinary meaning (or meanings)1 of ambiguity as a characteristic of propositions with more than one interpretation. In its normal usage, ambiguity is a linguistic concept, but in decision theory it is typically treated as a property of preferences.
The now-standard usage is quite different from that in Ellsberg’s (1961) original article. Ellsberg treated ambiguity, not as a property of preferences or relative likelihoods, but as a property of the information on which judgements of relative likelihoods might be based.

“Responses from confessed violators [of the expected utility (EU) axioms] indicate that the difference is not to be found in terms of the two factors commonly used to determine a choice situation, the relative desirability of the possible payoffs and the relative likelihood of the events affecting them, but in a third dimension of the problem of choice: the nature of one’s information concerning the relative likelihood of events. What is at issue might be called the ambiguity of this information, a quality depending on the amount, type, reliability and ‘unanimity’ of information, and giving rise to one’s degree of ‘confidence’ in an estimate of relative likelihoods.”

---Simon Grant, Ani Guerdjikova, John Quiggin "Ambiguity and awareness: a coherent multiple priors model." (17 April, 2020) [quoting p. 657 of Ellsberg's (1961) "Risk, Ambiguity, and the Savage Axiom]

In Yesterday's post I noted that while writing in defense of Knight, Ellsberg (1961) innovatively identified Knightian uncertainty with a class of decisions that can be specified precisely as empirical violations of the (Ramsey-)Savage axioms.* In particular, in such instances (when the formalization is violated) there is "no way to infer meaningful [subjective] probabilities for those events from their choices, and theories which purported to describe their uncertainty in terms of probabilities would be quite inapplicable in that area." (646; [parentheses added].) This identification was an innovation made possible and intelligible by formalization.  

As I remarked, Ellsberg noted that it is possible to treat the very same formalization as a [A] normative standard for a certain kind of ideal/idealized agent or [B] as a device to make predictions about agents. Ellsberg recognizes that Savage was (primarily) tempted by the normative interpretation whereas Ellsberg was primarily focused on the empirical interpretation. In particular, he wanted to identify classes of choice circumstances in which the possible rationality (669) of one's decision could not be articulated in or explained by reference to the Savage axioms. 

The Savage formalization allows the attribution to agents of precise subjective probabilities about distributions in the world. That is, in certain choice circumstances, the formalism allows for an interpretation of behavior that reveals if not inner states than at least structures that are compatible with such inner states. (It may be useful to compare this with the Samuelsonian strategy of revealed preference.)

With that in place, let's turn to ambiguity. As Grant/Guerdjikova/Quiggin note, in much of decision theory 'knightian uncertainty' is treated as synonymous with 'ambiguity' and has come to mean the absence of well-defined quantitative probabilities.+ And as they correctly note this is not quite what Ellsberg had in mind (and about that more in a second). Before I get to Ellsberg and ambiguity, it is worth noting that the absence of well-defined quantitative probabilities is indeed what Knight meant by uncertainty in the contrast between risk and uncertainty.

However, what the literature means by the absence of well-defined quantitative probability is (and now I quote Gilboa and Schmeidler (1989)) circumstances in which one "considers a set of priors as possible." (emphasis in original). It is important that this set is limited. Now, this is not quite what Knight, generally, intended to mean  by the contrast between risk and uncertainty.  In paradigmatic cases of uncertainty for Knight (and Keynes), we really have no right to make claims about the priors let alone treating them as well defined possibilities.

As an aside, this is not to deny that, as Richard Pettigrew noted to me, that "there are situations in which your evidence and/or current stage of scientific theorising is not going to deliver a unique probability assignment over the states you consider" but in which one has reasons to limit one's subjective probabilities to a set of possible distributions. And if forced to choose one may well come to think of your own (again quoting Pettigrew) as "your best subjective probabilities and use those while knowing that it would be equally compatible with your evidence to adopt an alternative set of probabilities."** If you and the other involved are experts in more or less the same situation, we can call this situation 'reasonable expert disagreement.' In fact, Savage himself noted (without having to contemplate violations of his axioms) that that “we must be prepared to find reasoning inadequate to bring about complete agreement.”  (1954 [1972]:7; see also 3, 67ff.) This is not trivial in the historical context, as I have noted elsewhere, in which the very idea of science involves the arrival of expert consensus. 

Now, as Grant/Guerdjikova/Quiggin note, Ellsberg thought ambiguity something else: namely it is a property of the "nature of one’s information."  And throughout the paper, Ellsberg goes on to deny that this feature "cannot be conveniently described in terms of a sample distribution." (659) Rather the point of using ''ambiguity" is to try to describe circumstances "when there are questions of reliability and relevance of information, and particularly where there is conflicting opinion and evidence." (659) So, one measure , but it is only one such measure, that one's information is ambiguous in Ellsberg's sense is the kind of reasonable expert disagreement [that is, situations one knows an expert's subjective probability about a distribution is compatible with another expert adopting an alternative set of probabilities.] But it is only one such measure because ambiguity itself cannot (to repeat) be reduced to and articulated in one distribution or another. As Ellsberg noted (660), Savage (57-60), too, had tried to characterize the nature of such ambiguity; Savage had explicitly acknowledged that formal theories known to him lacked resources to do so in satisfying ways.

Now, so for Ellsberg ambiguity is a "subjective variable" (660) about the nature of one's information. But the circumstances under which it arises can be characterized objectively. The proxy he uses to characterize these circumstances are violations of the Savage axioms can be predicted to occur and, perhaps, thought (somewhat) reasonable.

But that is not all of it (and this connects to Gilboa/Schmeidler).*** In circumstances of ambiguity one has, according to Ellsberg, some subjective grounds to "eliminate" some "possible distributions," but not all such possibilities. (661) For, Gilboa/Schmeidler (and the Bayesian literature) that is treated as the occasion to look for solutions to decision problems in terms of sets of priors as opposed to unique priors. In so doing, they end up echoing what I have called the displacement strategy of Knightian uncertainty. 

It is worth noting that there is an interesting difference of emphasis between trying to characterize an optimal decision strategy (in the manner of Gilboa/Schmeidler), say because one must choose or wish to guide policy-makers, and between wishing to promote understanding of how to think about or study one's information about possible distributions. (I have jokingly called the latter modal economics.) Ellsberg is clearly less interested in the former, and more interested in the latter.  Ambiguity in Ellsberg's sense does influence choice (664); yet Ellsberg is interested in finding out when and how, and to create "testable propositions" about such states [669], not to tell you what strategies to use in choice.

That is to say, Knight thought of uncertainty in terms of not knowing the (range and nature of) possible distributions. Ellsberg thinks we can develop rigorous thought  about what a theorist taught me to call intermediate circumstances in which we have some information about the possibilities. Somewhat surprisingly, according to Ellsberg -- and here he refers also to Chipman's experiments++ -- this thought has to be developed not from normative first principles, but rather experimentally ((655) within what in reviewing Donald Davidson, Suppes, and Siegel, he calls (in 1957) "the new field of experimental economics;" (1009). Of course, Ellsberg soon got preoccupied by different concerns. But that's really for another occasion.

So, Ellsberg has been remembered as generating a "paradox" in subjective expected utility theory. (This seems to be due to Harry V. Roberts.) This way of thinking about him says more about the economists' attitude toward their own past than Ellsberg. We have seen that alone among the responses to Knight that are still acknowledged by contemporary mathematical economists, Ellsberg wrote to vindicate Knight.+++ But he did so by using the canons of rational decision-making to specify when some circumstances of true uncertainty might occur. We could call them Ellsberg circumstances. These Ellsberg circumstances are characterized by constrained decision options in which information about world has the character of what he called ambiguity. In Ellsberg circumstances one grasps, we might say, the possible contours of the underlying distributions. Whether, in Ellsberg circumstances, it might pay to take bets on these contours is a tough challenge (one may well be in context of what he called "treacherous possibilities." (666)) He then opened the door not to better decision rules, but rather to the epistemology of information, which uniquely he thought ought to be explored experimentally.  

 

 

 

On Ellsberg's defense of Knightian Uncertainty, Arrow, and Ambiguity (I)

There has always been a good deal of skepticism about the behavioral significance of Frank Knight's distinction between "measurable uncertainty" or "risk," which may be represented by numerical probabilities, and "unmeasurable uncertainty" which cannot. Knight maintained that the latter "uncertainty" prevailed - and hence that numerical probabilities were inapplicable - in situations when the decision-maker was ignorant of the statistical frequencies of events relevant to his decision; or when a priori calculations were impossible; or when the relevant events were in some sense unique; or when an important, once-and-for-all decision was concerned.' Daniel Ellsberg "Risk, Ambiguity, and the Savage Axioms," The Quarterly Journal of Economics, Vol. 75, No. 4 (Nov., 1961), pp. 643-669* 

Before Daniel Ellsberg become a household name for leaking the Pentagon Papers, he was already known among economists and subsequently philosophers for a paper, published after his service as  marine and during his graduate study, that has become associated with a 'paradox'  because it is seen as a challenge to Subjective Expected Utility theory, despite the fact that 'paradox' is not to be found in the paper. And, so, while paradoxes naturally attract attention, I'd like to focus the attention of this digression on Ellsberg's way of motivating and situating his own paper. 

The quoted paragraph is the first in the essay. After citing Knight's Risk, Uncertainty and Profit, the identity of the main skeptic toward Knight's distinction is identified as Arrow in Ellsberg's first footnote: "see Arrow's comment: "In brief, Knight's uncertainties seem to have surprisingly many of the properties of ordinary probabilities, and it is not clear how much is gained by the distinction. . . Actually, his uncertainties produce about the same reactions in individuals as other writers ascribe to risks." Ellsberg is quoting an important and lengthy (1951) review article by Arrow that both helped discredit Knight among mathematically inclined economists and also was key in creating the (recall Alchian inspired; and here) identification of uncertainty with randomness that became central in theories used in risk and price models in finance. So, the question I wish to pose is in what sense is Ellsberg vindicating Knight?

He is responding, in particular, to those who think  that, and Ellsberg is now quoting Ramsey's great (1926) "Truth and Probability," "for a "rational" man all uncertainties can be reduced to risks.3"  That is to say, Ellsberg's paper is meant to meet those who prefer to displace uncertainty with probable risk (and so eliminate true uncertainty) on their own ground. In particular, it is meant to engage those who think that (our gambling) "choices are influenced by, or "reflect," differing degrees of belief...[and] to infer those beliefs from the actual choices." It's at this point that Savage's postulates are introduced because these help distinguish systematically in the analysis of observed outcomes whether they track "relative expectations...[or] relative preferences for outcomes." As Ellsberg puts it, "if one picks the right choices to observe, and if the Savage postulates or some equivalent set are found to be satisfied, this distinction can be made unambiguously, and either qualitative or, ideally, numerical probabilities can be determined." (645)

At this point Ellsberg notes that Savage's postulates/axioms can be interpreted normative and empirically. They can be treated as a rational standard for how one ought to behave and as an empirical prediction of how people with "plenty of time to reflect" (646) will generally behave or (as Ellsberg puts it "the rules will be commonly satisfied." (645)) As others have noted, in the 1950s both approaches were not unusual, although in general, as Wade Hands has noted, when Savage was writing alone he tended to emphasize the normative interpretation in terms of an ideal agent. By contrast Savage's sometime co-author, Milton Friedman, tended to treat these as a useful (as-if) device for making predictions.** (More recently, as Hands notes, the normative interpretation has predominated.)

At this stage one may well wonder what the previous paragraph has to do with Knightian uncertainty for we seem to have drifted into an interpretation of rational choice theory. In fact, it is crucial: Ellsberg's strategy is to suggest that "the meaning" of true or Knightian uncertainty  "would be that with respect to certain events they did not obey, nor did they wish to obey - even on reflection - Savage's postulates or equivalent rules." In such circumstances "the postulates failed to be acceptable in those circumstances as normative rules, or that they failed to predict reflective choices." (646) Ellsberg adds that he is primarily interested in failures of prediction (and notes that Savage himself is more interested, "no doubt" in their status as "normative rules"). 

That is to say, Ellsberg's intention, or strategy, is to identify Knightian uncertainty with a class of decisions that can be specified precisely as empirical violations of the Savage axioms. Again, on may wonder, why would that count as an instance of Knightian uncertainty? And the answer is: "there would be simply [a] no way to infer meaningful probabilities for those events from their choices, and [b] theories which purported to describe their uncertainty in terms of probabilities would be quite inapplicable in that area." (646; [parentheses added].) This [a-b] is true, as Ellsberg notes, absent "different operations for measuring probability." (646; [emphasis added.])  Here, Knightian uncertainty is associated with well defined choice circumstances in which it is impossible to attribute (with the Savage postulates) subjective probabilities to agents.

There is an instructive irony here. As mentioned above, Ellsberg noted that Arrow was a fierce critic of Rawls. When, later in 1973, (recall) in reviewing Rawls' Theory of Justice in the Journal of Philosophy, Arrow (correctly) discerned Rawls' adherence to Knightian uncertainty, but he misdescribed Knight's position as resting not on the denial of subjective probabilities (in the axiomatic tradition of "Ramsey and by L. J. Savage"), but on the denial of "objective probability." So, from Arrow's later (1973) perspective, Ellsberg can't be seen as defending Knight at all!+  I mention the irony because while the later (1973) Arrow is reductive about Knight, it is true that Knightian uncertainty was not limited to the class of interest to Ellsberg (subjective probabilities).

At this point, I am in the position to explain the relationship between Knightian uncertainty and what Ellsberg called ambiguity. But before I get to that (in a future post, alas), I need to spell out why the irony is instructive. One ground on which Arrow (1951) had criticized Knight is that his "analysis" is "lacking in formal clarity." (416) And in larger context, the implication is that without such formal clarity, Knight's position ought not be taken seriously. In fact,  in 1951 Arrow's hopes that then recent "developments"  in "the formalization of the theory of choice under uncertainty and its relation to many business phenomena" represent "dramatic breaks in continuity and have given hopes of a much clearer understanding of the problem." (405)++ That is to say, the formalization presents "a revolution in though" (to quote Samuelson, who in turn echoes J.N. Keynes anticipating Kuhn)  that is, the "Mathematical-Economics Revolution." 

While framed as a defense of Knight, Ellsberg's contribution is in the spirit of the Arrow-Samuelson hopes for economics. For, in fact, Ellsberg makes the very intelligibility of Knightian uncertainty dependent on a prior formalization (e.g. Savage's postulates). For, Knightian uncertainty can in his hands only be constituted by choice instances where this formulazation, when treated as a means to assign subjective probabilities to thoughtful (and willing to bet) agents,  is empirically falsified or (if one dislikes the Popperianism) inapplicable. For, Knightian uncertainty now just is or is constituted by the very violation of the (particular) formalization. Whatever one thinks of that it's pretty clear this far removed from Knight's (and Keynes's) own ideas about the nature and epistemology of uncertainty even if it clearly is a case of Knightian uncertainty. Obviously, as Ellsberg notes, the violation is itself also a research opportunity to develop other formalizations that may be predictive of such choice situations, or their systematic violations.*** 

 

On Enthusiasm and Expected Utility in Grouchy

Last year I wrote post (recall) about Sophie Grouchy's analysis of enthusiasm and inserted her into a historical narrative. Shortly thereafter I received a long letter from my friend the mathematical economist, M.A. Khan, showing fatal flaws in my analysis and sketching an alternative interpretation. One nice feature of these digressions, is that every day I can make a fresh start at thinking. What follows is very indebted to his (unpublished) letter.* I rely on Sandrine Bergès' excellent translation (which is available to order from OUP USA today!)

Grouchy turns to enthusiasm in Letter 3 of Letters on Sympathy (LS) when she confronts a possible objection, “the effect here is too great for its cause, and you will no doubt ask yourself why personal sympathy is sometimes so strong while its motives are so weak and nebulous.” (p. 79) The objection relies on a principle, what I like to call David Hume’s ninth rule of causal reasoning, that causes and their effects must be proportional.[1] This principle is assumed throughout Hume’s science of man, and explicitly articulated in his essay “Of Interest:” “An effect always holds proportion with its cause,”. Grouchy does not object to the principle.

The objection amounts to the idea that in describing and explaining various social phenomena, Grouchy has been illicitly appealing to personal sympathy. Personal sympathy (“sympathie individuelle”) is a directed sympathy at a particular person. The reason why the appeal is thought to be illicit (by the hypothetical objection) is that such directed sympathy appears to be ungrounded in the motives said to cause it. These causes are too weak to explain the intense directed sympathy. So, the effect is out of proportion with its cause(s).

Grouchy meets the objection by appealing to enthusiasm, which fills the ‘proportionality’ gap: “enthusiasm, mixing with our soul’s first observations, extends those observations beyond the point to which our factual knowledge alone is able to bring them.” (79) The response to the objection is by no means self-explanatory.[2]

Enthusiasm is the kind of mental or imaginative mechanism that adds, as it were, causal heft to motives such that these are proportionate to an effect like directed sympathy. It does so by reaching beyond the observed facts. In other words, enthusiasm discloses possibilities. The effect of a person in the grip of enthusiasm, and the motives that occasion them,[4] is “intense and sudden personal sympathies.” (LS 3; I use ‘occasion’ rather than ‘trigger’ because it is not the motives that cause enthusiasm, but the person to which sympathy is directed or our aspirations pertaining to that person.) So far so good.

But one may well wonder if this explanation does not simply push the problem back a level. What is enthusiasm such that it can play this role? Grouchy’s response is extraordinarily subtle (that is, this is the passage I screwed up last year):

Enthusiasm comes from the degree to which our soul is able to represent to itself, at the same time and in an indeterminate manner, all the pleasures or all the pains we would gain from a particular situation, or from a certain person and our relationship with him or her.  This picture brings together in one instant what should in reality span months, years, and sometimes an entire lifetime. Enthusiasm, therefore, conceives of its object in an exaggerated sort of way; and because it presents the mind with a greater number of objects than it is able to consider distinctly, it is always vague in some respects. (L3, 79-80)

Enthusiasm is itself the effect of mental representations of anticipated, future utility attached to a situation or a person and our relationship to it or him/her. (For other posts on Grouchy's political economy, see here; here; here). First a point about  terminology. When Grouchy uses ‘utility’ in LS, she tends to mean -- in common with the usage of Hume, Smith, and Beccaria -- of ‘public interest.’ While there are consequentialist elements in her moral psychology, she is not a Benthamite utilitarian. So, the use of ‘utility’ is anachronistic and potentially misleading.[3] Second, in the passage Grouchy uses ‘enthusiasm’ to refer to (i) the effect of the mental representation and (ii) the feeling associated with the representation. Later she argues that (iii) when enthusiasm has become habitual, she uses it to refer to a disposition.

It is not immediately clear what Grouchy means to convey with the claim that the representation (of such future utility) comes in degrees.  Perhaps she means to suggest that the more determinate or (to put it in the idiom of the age) clear and distinct  the representation becomes the more enthusiastic we may be. Of course, by her lights, the representation can never be fully clear or distinct (or determinate); in her work on demagogues this matters a lot (recall; and also here). For, “enthusiasm…presents the mind with a greater number of objects than it is able to consider distinctly.” These ‘objects’ turn out to be future states of affairs of a particular kind.

For, the content of the representation is itself something of an impossible idealization of utility: it compresses in time (“it brings together in one instant what should in reality span months, years, and sometimes an entire lifetime”); and it represents, perhaps, mutually exclusive (logical) possibilities (“all the pleasures or all the pains we would gain”). Either way, enthusiasm is in one sense not truth apt. It “exaggerates.”

In her analysis, Grouchy then makes an important distinction between how the enthusiastic representation is experienced by the person under its sway and the person theorizing it.[**] This becomes clear when she describes one of the effects of enthusiasm on “our sensitivity.” It is “subject to another form of amplification born out the multiplication of pains and pleasures we imagine.” Our (enthusiasm, that is) representation of future utility produces “fears and desires” that are “either impossible in reality or at least cannot be found together.” Crucially, “”in the midst of our soul’s turmoil, we cannot untangle this impossibility.” But from the theoretical vantage point she adopts, one can discern “there is even actual error involved here.” That is, conceptually, she anticipates the recently popular distinction between experienced and decision utility (see Kahneman et al). 

As noted above, Grouchy also treats habituated enthusiasm as a “disposition:” “if a circumstance or a person has provoked it in us on several occasions, that person or circumstance retains the power of provoking it, independent even of our thinking about it, and we can then consider enthusiasm as a passion of the soul.” (LS 3; later, in LS 5 she emphasizes the role of enthusiasm in reinforcing habit (p. 109.) That is to say, the triggering cause of enthusiasm need not be present actually in order for the mental mechanism to do its work. Merely representing the trigger may lead a predisposed or habituated mind to represent future utility associated with that person/situation.

Grouchy is here presupposing something like a Humean associative mechanism.[10] In particular, she relies on resemblance: “enthusiasm toward certain qualities disposes us to sudden and rash sympathy for the people in whom we think we recognize them.” (80) Here enthusiasm has the effect of intensifying (directed) sympathy for the whole (person) based on properties/features (“qualities”) of it (the person). We may say then that the disposition of enthusiasm is a cognitive heuristic likely to be error prone.

Grouchy is clear that the tendency  toward enthusiasm  is ground in three personality traits:

1. The strength of imagination, which embraces with more or less haste great displays of sensations and events.
2. The strength of sensitivity, which is more or less affected by those displays and preserves them with more or less constancy.

By ‘sensitivity’ Grouchy means something like a property of our bodies that facilitates our capacity to experience the world through the senses. The third trait is our tendency to reflect on the triggering cause, which prepares “our hearts…to feel affection” toward the triggering cause. This tendency is a consequence of “the need or desire to find an object for this affection.”

So, she agrees with Adam Smith (recall yesterday's post) that enthusiasm is the cause (and effect) of emotional intensification. But, while Smith makes the kind of enthusiasm in a person be a consequence of their moral character, for Grouchy enthusiasm is a consequence of (morally) generic character traits such as imagination, sensitivity, and our longings. As she sums up her own argument (with a competing account of sympathy), “I have shown how moral pains and pleasures are born out of physical sympathy that has become personal, strengthened by diverse circumstances, rendered more active and energetic by enthusiasm.” 

But the key difference with Smith is that for Grouchy, enthusiasm is a representational phenomenon with considerable cognitive complexity and that is essentially connected to the future-orientation of agents and their hopes/aspirations connected to other persons and situations. From this it follows, and this is another important difference with Smith, that Grouchy links enthusiasm to a wider range of social phenomena (including love and demagoguery). 

The PSR, Due Diligence, Plutarch, and Debt (HT Graeber)

§ 1  Plato in his Laws permits not any one to go and draw water from his neighbor's well, who has not first digged and sunk a pit in his own ground till he is come to a vein of clay, and has by his sounding experimented that the place will not yield a spring. For the clay or potter's earth, being of its own nature fatty, solid, and strong, retains the moisture it receives, and will not let it soak or pierce through. But it must be lawful for them to take water from another's ground, when there is no way or means for them to find any in their own; for the law ought to provide for men's necessity, but not favor their laziness. Should there not be the like ordinance also concerning money; that none should be allowed to borrow upon usury, nor to go and dive into other men's purses, — as it were into their wells and fountains, — before they have first searched at home and sounded every means for the obtaining it; having collected (as it were) and gathered together all the gutters and springs, to try if they can draw from them what may suffice to supply their most necessary occasions? But on the contrary, many there are who, to defray their idle expenses and to satisfy their extravagant and superfluous delights, make not use of their own, but have recourse to others, running themselves deeply into debt without any necessity. Now this may easily be judged, if one does but consider that usurers do not ordinarily lend to those which are in distress, but only to such as desire to obtain somewhat that is superfluous and of which they stand not in need. So that the credit given by the lender is a testimony sufficiently proving that the borrower has of his own; whereas on the contrary, since he has of his own, he ought to keep himself from borrowing...

[§2] ...Thus we ought in our affairs, as in a besieged town, never to admit or receive the hostile garrison of a usurer, nor to endure before our eyes the delivering up of our goods into perpetual servitude; but rather to cut off from our table what is neither necessary nor profitable, and in like manner from our beds, our couches, and our ordinary expenses, and so to keep ourselves free and at liberty, in hopes to restore again what we shall have retrenched, if Fortune shall hereafter smile upon us...

§ 3 ....The Goddess Artemis in the city of Ephesus gives to such debtors as can fly into her temple freedom and protection against their creditors; but the sanctuary of parsimony and moderation in expenses, into which no usurer can enter to pluck thence and carry away any debtor prisoner, is always open for the prudent, and affords them a long and large space of joyful and honorable repose. For as the prophetess which gave oracles in the temple of the Pythian Apollo, about the time of the Persian wars, answered the Athenians, that God had for their safety given them a wall of wood, upon which, forsaking their lands, their city, their houses, and all their goods, they had recourse to their ships for the preservation of their liberty; so God gives us a table of wood, vessels of earth, and garments of coarse cloth, if we desire to live and continue in freedom. Aim not at gilded coaches, steeds of price, And harness, richly wrought with quaint device; for how swiftly soever they may run, yet will usuries overtake them and outrun them....

§ 4 ... And as King Darius sent to the city of Athens his lieutenants Datis and Artaphernes with chains and cords, to bind the prisoners they should take; so these usurers, bringing into Greece boxes full of schedules, bills, and obligatory contracts, as so many irons and fetters for the shackling of poor criminals, go through the cities, sowing in them, as they pass, not good and profitable seed, — as did heretofore Triptolemus, when he went through all places teaching the people to sow corn, — but roots and grains of debts, that produce infinite labors and intolerable usuries, of which the end can never be found, and which, eating their way and spreading their sprouts round about, do in fine make cities bend under the burden, till they come to be suffocated. They say that hares at the same time suckle one young leveret, are ready to kindle and bring forth another, and conceive a third; but the usuries of these barbarous and wicked usurers bring forth before they conceive. For at the very delivery of their money, they immediately ask it back, taking it up at the same moment they lay it down; and they let out that again to interest which they take for the use of what they have before lent.

§ 5  It is a saying among the Messenians, Pylos before Pylos, and Pylos still you'll find; but it may much better be said against the usurers, Use before use, and use still more you'll find. So that they laugh at those natural philosophers who hold that nothing can be made of nothing and of that which has no existence; but with them usury is made and engendered of that which neither is nor ever was.--Plutarch, Moralia (translated by  William Watson Goodwin) "Against Running In Debt, or Taking Up Money Upon Usury." [ES: I have not checked the Greek, but hopefully this does not undermine the points I wish to make today.]

In Debt: the First 5000 Years (pp 231-2), David Graeber calls attention to the passage quoted from Plutarch. His excerpt is briefer than my version and he uses the passage to quote Plutarch's likening the role of debt to "a foreign invasion."* Here I have different (ahh) interest in the passage. Before I get to that it's only fair to note Plutarch's main argument, which is that debt is the off-spring of the desire for luxury.  Luxury is something one does not need, and so superfluous.+ And so Plutarch's main argument is a prudential plea for moderation.

This commitment helps explain, in part, what would sound strange to us: "that usurers do not ordinarily lend to those which are in distress, but only to such as desire to obtain somewhat that is superfluous and of which they stand not in need." (Hereafter, the maxim of ordinary usury; or, simply, the maxim.) I suspect Plutarch knew (even lacking Graeber's rich account) that there existed plenty of usury (to the distressed) which was 'non-ordinary' in his sense.++ Even  so the passage provides nice evidence for the thought that due diligence into the ability to repay ("the borrower has of his own"), and perhaps character, was part of the "ordinary" practice of usury among the Greek aristocrats (who could be taken or be seen to afford the superfluous) that Plutarch would have known.

Of course, Plutarch's larger argument is a proposal to regulate debts ("that none should be allowed to borrow upon usury, nor to go and dive into other men's purses, — as it were into their wells and fountains, before they have first searched at home and sounded every means for the obtaining it.") This puts a requirement on any would-be-debtor, and any would-be-creditor, to fund any expenses that would (ahh) ordinarily be funded by credit out of assets (or savings). Once such due diligence has been done, there is de facto no proper need for credit/debt. This suggests, however, a more far-reaching implication. If the due diligence shows that the would-be-borrower is short of collateral or funds, then from Plutarch's perspective there is no grounds for a debt. So, the point of the regulation alongside the Maxim is clearly to prevent the creation of credit to the distressed that become debts that cannot be repaid. The underlying (political) rationale is clear enough because a debtor is, for Plutarch, somebody un-free, that is, halfway down the path to slavery.*

So much for the political angle. Plutarch also notes that the very existence of interest makes a mockery of one of the versions of the Principle of Sufficient Reason, nothing can be made of nothing and of that which has no existence. We tend to associate the principle with Parmenides; writing much later, Plutarch treats it as a common principle among the natural philosophers. Interest violates a principle of equality built into the nature of debt (recall this one); because what is returned is more -- sometimes dis-proportionally more -- than the principal.** 

Of course, there are a whole battery of claims and arguments that interest does represent something (cost of borrowing, opportunity costs, inflation expectation, underlying rate of return in a society, etc.) So, the charging of interest can be made intelligible. And, perhaps, that's sufficient to undermine the idea that interest is somehow un-caused. I like that in the version of the PSR quoted by Plutarch, the PSR is committed to the claim that no thing can be produced by nothing in the sense of not existing. (For, of course, it cuts off a natural temptation to say, in response, that social fictions cause interest!) 

Now, Graeber (recall) has quite interesting things to say about the genealogy of philosophy. He argues that philosophy's origins are grounded in paradoxes of coinage, and that philosophy is the complicit offspring of violent eradication of forms of life. In reflecting on this (and recall this post on the PSR), I had the following free association.

As Graeber emphasizes, the Greek word for "interest" literally means "offspring." And, from the perspective of the PSR, one may say that interest is, thus, either un-caused offspring or the product of unusually fertile principal. The latter route is (recall) the Piketty-ian world of disproportionate capital growth, perhaps, characteristic of financial capitalism; the former is a miracle of the sort that founds Christianity.   

 

 

 

 

 

 

On the Essence of Graeber's Debt: Modal Economics Meets Hierarchy and Ideology

It seems to me that this agreement between equals to no longer be equal (at least for a time) is critically important. It is the very essence of what we call "debt."…

Debt is a very specific thing, and it arises from very specific situations. It first requires a relationship between two people who do not consider each other fundamentally different sorts of being, who are at least potential equals, who are equals in those ways that are really important, and who are not currently in a state of equality — but for whom there is some way to set matters straight….

During the time that the debt remains unpaid, the logic of hierarchy takes hold. There is no reciprocity. As anyone who has ever been in jail knows, the first thing the jailors communicate is that nothing that happens in jail has anything to do with justice. Similarly, debtor and creditor confront each other like a peasant before a feudal lord. The law of precedent takes hold. If you bring your creditor tomatoes from the garden, it never occurs to you that he would give something back. He might expect you to do it again, though. But always there is the assumption that the situation is somewhat unnatural, because the debt really ought to be paid….

This is what makes situations of effectively unpayable debt so difficult and so painful. Since creditor and debtor are ultimately equals, if the debtor cannot do what it takes to restore herself to equality, there is obviously something wrong with her; it must be her fault.…

A debt, then, is just an exchange that has not been brought to completion. David Graeber Debt: The First 5000 years, pp. 120-1. [This is the third post on Graeber's Debt; recall here and here.]

According to David Graeber, debt is inter alia a cause of hierarchy. In particular, debt generates (further) hierarchy among those that are (at least tacitly) recognized as equals in some formal sense.  For, according to Graeber those that are not even potentially equal cannot be indebted to each other; this is why slaves, animals, and children cannot (legally) be supplied with credit.

Of course, this equality is merely formal. For, in conditions of general equality the very institution of debt -- not unlike the institution of property -- would not arise. Some material difference is required to generate debt in the first place: the would-be-creditor is in some financial sense superior to the would-be-debtor. So, the institution of debt presupposes considerable formal equality alongside some non-trivial material inequality.

According to Graeber debt makes those formerly recognized as equal de facto unequal (at least for the duration of the debt). For the debtor owes the creditor. Here we leave aside the nature of this obligation (is it properly moral, merely legal, or something else altogether?) The de facto superiority of the creditor is also revealed in the fact that she can decide to forgive the debt, or not.

As an aside, in his logic (not in his history, however) Graeber tends to ignore cases where unpaid debts ruin the creditor. There is an old capitalist joke (attributed to J. M. Keynes, but worth reviving in the age of Donald Trump) Owe your banker £1000 and you are at his mercy; owe him £1 million and the position is reversed. [Obviously, the joke needs to be adjusted to inflation.] But that's because on the whole Graeber assumes+ that debts are backed up by the threat of violence that can force repayment. 

It is central to Graeber's argument that the very idea of a debt presupposes that it can, in principle, be repaid in some form or another (sometimes in kind, or by a life, a symbol, etc.)--and so equality can be restored.  We may say, then, that in Graeber's analysis it is the foreseeable consequence of the institution of debt that it generates further, temporary hierarchy among those that are recognized as equals in some formal sense while aiming at returning to considerable equality.

In fact, given the necessary logic of compound interest, it is a foreseeable consequence of much debt that while debt formally aims at a return to equality, it can be expected that the hierarchy -- which is intended to be merely temporary -- only gets further entrenched and, thereby, eventually undermine even formal equality because the indebted are reduced to real slavery, imprisoned, or deemed the unworthy. This much is all in Graeber.

From a modal (economics) perspective, we can say, then, that [A] debt takes us from worlds in which equality is presupposed and at least thought possible in the future, to worlds in which equality becomes thought impossible. In this new world, relations are not governed by formal reciprocity, but by precedent that is inscribed in some kind of institutional memory (the law, custom, etc.).* So, Graeber's analysis of debt is a kind of foreseeable, unintended consequence explanation. It is unintended in the sense that the the very intelligibility of the institution of debt presupposes aiming at some form of equality (that's the would-be-repayment part).** The outcome is foreseeable because Graeber models his agents -- and this is the main virtue of his theorizing -- as not deceived about this; they understand their own world-all-too-well. 

It would be sufficient to stop here (in order to re-think Liberalism). But it is worth noting that for Graeber [A] is inherently unstable in its modern, capitalist form. It is not unstable because hierarchy is intrinsically unstable. (Graeber analyzes hierarchical societies and finds them remarkably stable over very long periods.) Rather, it is unstable because the moment one thinks and acts that [A] is an eternal, stable equilibrium, that is one thinks it "really will be around forever, everything goes haywire." (p. 358) That is, as long as the agents that inhabit [A] consider it fragile, [A]=[A]. But the moment that the agents that inhabit [A] endorse [A] as occurred during, say, 'the great moderation,' it becomes self-undermining (that is, [A]-->[-A]). That is to say, for [A] to work, it may not be perceived to be true, or, a ruling ideology.

 

 

 

A note on Modal Economics and Interest; or a thought on Thanksgiving

The most common forms of debt are loans, including mortgages and auto loans, and credit card debt. Under the terms of a loan, the borrower is required to repay the balance of the loan by a certain date, typically several years in the future. The terms of the loan also stipulate the amount of interest that the borrower is required to pay annually, expressed as a percentage of the loan amount. Interest is used as a way to ensure that the lender is compensated for taking on the risk of the loan while also encouraging the borrower to repay the loan quickly in order to limit his total interest expense. Debt Definition@Investopedia

 

Interest is charged by lenders as compensation for the loss of the asset's use. In the case of lending money, the lender could have invested the funds instead of lending them out. With lending a large asset, the lender may have been able to generate income from the asset should they have decided to use it themselves.Interest Rate Definition @Investopedia 

Common debt connects the present to the future. From the creditor's perspective, interest represents, among other things, the reward for deferring present consumption or other use (not to mention hiding the money under a mattress), including a reward for risk of default and some protection against expected inflation (etc.). That is to say, expectations about possible outcomes into the future enter into the rate of interest she will charge. (I return to this below.)

From the debtor's perspective debt allows one to shape the future. So, with a mortgage one can, say, turn a logically and physically possible house, into a real house. With a loan, an ecological entrepreneur can turn a dream of a wind-farm that generates clean energy into reality. That is to say, particular loans connect the present to possible worlds that are actualized in virtue of, at least in part, these loans.* It's possible, of course, that there are other means to the roughly same possible future without these loans (the house was build with a bequest not a mortgage, etc). So, I am not suggesting that the existence of this loan can help provide identity conditions or precisification of a possible world.

During a guest lecture on the Greek debt crisis by a high ranking civil servant of the Dutch Treasury, Michel Heijdra (a trained and rather gifted philosopher) remarked (something to the effect) that interest on debt also represent an optimistic view of the future. For, its rate tends to exceed risk-free income as well as other uses (present pleasure in consumption, etc.) that represent a status quo bias. This made realize that within the class of possible worlds from any given production possibility frontier that may be actualized in virtue of contracted debt (recall also this post on Piketty's idea of a technological frontier), interest-bearing debt actualizes a particular sub-set of these. In fact, the sub-set of these are biased toward the generation of income that can cover and repay the principal and interest on it. In a larger equilibrium framework, these are optimal, Leibnizian worlds. I mention Leibniz not just because the optimization techniques he (and his followers, Bernoulli, Euler, etc.) pioneered are still central to how economics thinks of such matters, but also because these remarks are very much in the spirit of his remarks on possible worlds and the role of God as optimizer and economizer in it (see, e.g., Rescher).**

Obviously, not all debt will be repaid. Interest on common debt also represents a reward of risk and this entails some probability of default. So, in practice debt, makes us aim for Leibnizian possible worlds while allowing for the genuine possibility that locally we land on sub-optimal worlds, including worlds that are worse than the present one in which the debt is not repaid and the borrower scrambles for other funds. (The preceding sentence is not an attack on Leibniz (who can easily accommodate the insight.)) So, we may say, then, that common debt connects us to possible worlds that our close to optimal*** given present endowments and possible worlds that are rather annoying from the present perspective. The former ought to predominate, but the latter cannot be ruled out.

 

 

When the Economics of Ignorance Cannot Model Politics

Moreover, when there is a serious enough shortage of collateral, the government will find it optimal to alleviate the situation by supplying collateral that is backed up by taxpayer money. This is more efficient than having the private sector invest in safe assets to cover exceptional severe shortages. The advantage of the government is that it can act ex post, when it is clear that such a shortage is at hand. No funds are tied up in advance....(22)

By now, the methods out of a crisis appear relatively well understood. Government funds need to be committed in force (Geithner (2014)). Recapitalisation [of banks] is the only sensible way out of a crisis. But it is much less clear how the banking system, and especially shadow banking, should be regulated to reduce the chance of crisis in the first place. The evidence from the past panic suggests that greater transparency may not be that helpful....Relying on market discipline and price discovery is unlikely to be effective in money markets generally. Letting the successful pricing of systemic risk in stock markets be a guide for money markets seems misguided. Bengt Holmstrom (2015) Understanding the role of debt in the financial system. (24)

There are five important points in these passages  just quoted; two are  (forgive my pun) transparent in the text and connected, the three others less so. Holmstrom is a recent Nobel laureate and Paul A. Samuelson Professor of Economics at MIT. The paper I quote from is very accessible -- surprisingly so given that he talking to professional bankers not the general public [his modeling and data are submerged in the footnoted papers] -- and does a nice job conveying the conceptual and policy relevant insights from what is otherwise rather technical and demanding modeling (as well as non-trivial data-driven inquiry). His writing is useful both to learn from and as a sociological proxy to find out and reflect on what the professional, technical economics elite thinks when they are talking in a policy-relevant environment (Bank of International Settlements is the place where central banks work together on standards and rules).

The core insight of the paper is that one should not model how to understand collateralized debt (which is information insensitive) -- and exists in order to provide liquidity and find a home for surplus savings -- on how one understands equity (which is information sensitive and pools risk)--in equity markets prize discovery is central; in debt markets providing  liquidity is. When one recognizes this, one can make sense, so Holmstrom argues, of non trivial differences between equity and debt markets (especially the shadow banking system) both in good times and during/after panics. (One very interesting consequence of the core insight is that debt becomes preferred collateral of debt leading to a kind of unstable, open-ended system of musical chairs.) Let me stipulate, for the sake of argument, that Holmstrom is right about this (I really find it fascinating, although it is notable that in his model equity markets are siblings of insurance markets--about that some other time).* Okay, let me now get to the five important points.

First, one consequence that Holmstrom hammers home is that attempts to make debt/money market instruments transparent reflects "a serious misunderstanding of the logic of debt and the operation of money markets" (2). If the collateral is generous enough you don't worry about repayment; if you don't have to worry about repayment you don't have to worry about the details of the collateral; (etc.) If you start making information readily available, you turn debt into equity because you start to have regular occasions to wonder about repayment and collateral, and you will start to make fine-grained distinctions among different debt repayment capacities and collateral values. So, you turn a low monitor, low volatile instrument into its opposite. Greater transparency may facilitate monitoring in good times, but it makes matters a lot worse in bad times (so his argument goes); and monitoring is expensive. (To be precise in times of panic he allows some transparency, but in ordinary times debt markets thrive on opacity.) So, Holmstrom deviates from the (liberal) consensus that more market transparency is desirable and efficient. He (and his co-authors) understand debt as a regime which involve "symmetrical ignorance," (see "Ignorance, Debt and Financial Crises.") This is an insight worth returning to in the future.

Second, he defends tax-funded bailouts. They are more "efficient" than trying to prepare to prevent panics. He also thinks that private firms are not capable of being the proper backstop. It is notable, however, that (i) silent about situations where the government lacks control over its own currency or (ii) finds itself confronted with seriously unwilling tax-payers. Because many governments are not in control over their own currency, the prescriptive force of his remarks is limited to a small number of policy environments. (He is also strikingly silent about bail-outs by supranational institutions [as several of my students noted].)

Third, he does not model the policy side. He simply assumes that (a) policy can ensure that collateral is ultimately "taxpayer backed." In addition, (b) he does not model the distributional effects of his policy prescriptions. Basically he seems to think that the benefits of keeping the system around seem to swamp whatever distributional side-effects. But, of course, electorates and politicians are less sanguine about this.+ (He is also uninterested in the moral evaluation of this, but that's less remarkable.) Basically if politicians were to appear to follow his advice they would probably not get re-elected.

Fourth, it is extremely notable that Holmstrom is modeling capital markets as they are, and have been, but that he does not really explore alternatives to the status quo. (He is very good at rejecting bad models of  the status quo, but he does not allow alternatives to the status quo.) Perhaps such status quo bias is to be expected when you are engaged to talk to Bank of International Settlements (is this an incentive free environment?), but it is odd to be making general efficiency claims about an institutional regime without having done comparative institutional analysis (and when the down-side risks onto those not at the table are not being modeled at all). More important. I am also a bit mystified what the evidence for the efficiency claim is supposed to be even within the status quo--the welfare loss of the financial crisis are still unfolding so, it's a bit odd to take the question as settled. (No source or data are cited, no back of the envelope model provided.)++

Finally, Holmstrom is pretty explicit that panics can't be eliminated. (As he says in the more technical paper, "A financial crisis is a manifestation of the “tail risk” that is endogenously created by agents in the economy in order to trade."(37) Although there is some ambiguity if a financial crisis is inevitable.) At one point he remarks that "Everything that adds to liquidity in good times pushes risk into the tail." It follows that if you can keep adding liquidity (at an even slow pace), you can try to keep pushing the probabilities of undoing of the system into the tail. (He never explains why you can't add liquidity indefinitely, but it's clear that he thinks that even low probability events will occur eventually.) What's striking about Holmstrom's point is that his argument here is not in terms of mechanism of debt markets, but entirely based on the nature of probability distributions. For Holmstrom uncertainty is a psychological state of investors/creditors; but it's not a feature of his model--that reality is composed of probability distributions in which there is no place for true uncertainty, even though the [onset of a] "crisis is a bad enough shock to cause information-insensitive debt to become information sensitive."** Here 'shock' stands for (exogenous) 'bad news.'  One other policy implication follows from his paper: some species of censorship would be good for debt markets.

 

On Whitehead and Hayek (On Modal Economics)

The problem which we meet here is by no means peculiar to economics but arises in connection with nearly all truly social phenomena, with language and most of our cultural inheritance, and constitutes really the central theoretical problem of all social science. As Alfred Whitehead has said in another connection, "It is a profoundly erroneous truism, repeated by all copy-books and by eminent people when they are making speeches, that we should cultivate the habit of thinking what we are doing. The precise opposite is the case. Civilization advances by extending the number of important operations which we can perform without thinking about them." This is of profound significance in the social field. We make constant use of formulas, symbols and rules whose meaning we do not understand and through the use of which we avail ourselves of the assistance of knowledge which individually we do not possess. We have developed these practices and institutions by building upon habits and institutions which have proved successful in their own sphere and which have in turn become the foundation of the civilization we have built up. Hayek (1945) The Use of Knowledge in Society The American Economic Review, 528.

It's a bit surprising, even after repeated re-reading, to see Whitehead show up in Hayek's argument. Whitehead is describing, what in science studies is called 'blackboxing.' (I suspect the idea can be traced back to Leibniz's remarks on the calculus.) Hayek turns it into a broader point -- not just about technologies, scientific and material --, but about norms, habits, and institutions which become the repository of (or are the) means toward effective problem solving (in a limited domain); these means are not really understood by the folk whose problems are thereby solved. It's not a throwaway remark ("profound significance"). But it would be a mistake to see this only as an affirmation of the significance of tradition (a la Hume).

In the article, Hayek makes us think of markets as epistemic devices: "the price system" is a "mechanism for communicating
information." (527) The information is not, as such, available to anybody, but dispersed: "in a system where the knowledge of the, relevant facts is dispersed among many people, prices can act to coordinate the separate actions of different people." (526) There is a lot to be said about the contextual knowledge that Hayek thinks is central to the argument, but here I am focused on the issue raised by Whitehead.. I quote an illuminating passage:

There is hardly anything that happens anywhere in the world that might not have an effect on the decision he ought to make. But he need not know of these events as such, nor of all their effects. It does not matter for him why at the particular moment more screws of one size than of another are wanted, why paper bags are more readily available than canvas bags, or why skilled labor, or particular machine tools, have for the moment become more difficult to acquire. All that is significant for him is how much more or less difficult to procure they have become compared with other things with which he is also concerned, or how much more or less urgently wanted are the alternative things he produces or uses. It is always a question of the relative importance of the particular things with which he is concerned, and the causes which alter their relative importance are of no interest to him beyond the effect on those concrete things of his own environment. (525)

Prices are effects of what may be totality of human causes that relieve any agent from the need to keep track of all these causes. In fact, on Hayek's account the economic agent need not concern herself with nearly all of these causes; prices (the effects), or even just the change in prices, convey all the relevant information for decision. We can understand the relations among prices and price-changes as the phenomena (in the Kantian sense), and we can discern in their underlying causes the unknown noumenal reality (for the economic agent). Here the market is understood as an epistemic device whose inner mechanical operation is not so much a blackbox, but an enigma.

One last thought on the passage. Prices collapse a rich modal structure of possible causes ("anything that happens anywhere in the world that might not have an effect") into a single coarse-grained dimension. That is, while price formation is an extremely useful epistemic device, we should not ignore the fact that it can also obscure bits of reality to those focused on (changes in relative) prices only. (On this picture you can't even really reverse engineer the causes from the effects.) We can understand bonds, futures, and insurance as ways to recapture some of the information about that modal structure, but they do so as imperfect instruments.