Sanjay Reddy on economics

by Ingrid Robeyns on January 13, 2012

Hear hear! What a wonderful short interview with Sanjay Reddy by Perry Mehrling from the Institute for New Economic Thinking (INET):

Reddy defends the position that economics is a profoundly value-entangled science, and that “Good theory is theory which illuminates the world, and good theory cannot start from a-priori premises which are disconnected from the world. Good theory has to start in part from observation from the world.”

I agree with every word Reddy says, but am a bit puzzled why Mehrling sees Reddy’s position as ‘a strong position’. In my view, if it is regarded (by economists?) as a ‘strong position’, that is just because economics has so forcefully tried to distance itself from any evaluative or otherwise ethical concerns; but in truth, economics has never been value-free, it has only fooled itself that it could be so. I’m really glad that Reddy is contributing to a better understanding of economics as value-entangled. Can’t wait to read the result of his INET project, “a book making a broad case for the resurrection of normative reasoning in economics”.



René 01.13.12 at 10:25 pm

Great interview Ingrid, thanks for sharing. I would just like to add this recently published book (link below) which collects key articles on the debate between positive and normative economics. It includes essays by Putnam, Walsch, Sen and new contributions by Nussbaum and Harvey Gram. Although it is mostly a compilation of already published articles it is nice to have them all in the form of a book. I believe the only one that would be missing, on the side of the defenders of the distinction, is Dasgupta’s “What do Economists Analyse and Why: Values or Facts?” Economics and Philosophy, 2005, 2 (2) of which there is a revised and extended version in the Oxford Handbook of Philosophy and Economics.


John Quiggin 01.13.12 at 10:33 pm

Also interesting is his discussion of the gap between theoretical and applied work – this is a problem everywhere I guess, but I certainly find it to be a big one in economics. Work on DSGE macro is a prime example, where internal theoretical concerns dominate at the expense of practical relevance. The same is true of decision theory where I do most of my work – the gap between theory and application is much wider than it needs to be.


UnlearningEcon 01.13.12 at 10:55 pm

‘The Skeptical Economist’ is a fantastic book for exploding myths about economics being value-free. Consumer sovereignty over worker sovereignty is an obvious value judgement. Presupposing barter, private property and contracts as somehow ‘natural’ is another. It also strikes me that ideas like taxation as theft also have their roots in ‘the market’ as a separate entity in which the government ‘intervenes’.


Substance McGravitas 01.13.12 at 11:22 pm

I’ve been reading Debt which also has some ideas about the origins of the market. I’d love to see a review of it, given that David Graeber has had a mention or two here before.


Jim Harrison 01.14.12 at 1:09 am

I’d like to believe economics will go through a transition rather like what happened in the 17th Century when astronomy separated itself from astrology and chemistry separated itself from alchemy. The problem is that the legitimate practice of economics, economics as a branch of sociology, is necessarily a rather modest operation compared to the bogus physics of money that currently rules the universities under the patronage of the oligarchs.


Joshua Broady Preiss 01.14.12 at 1:15 am

Best line comes at the end: “One of the most traumatic experiences of my life was my first week or two as a graduate student in economics at Harvard… Because the set of presumptions about as to what is important and how the analysis should be undertaken were rather alien and alienating to me coming back to economics after exposure to a broader social science which took reality much more seriously.


Antonio Conselheiro 01.14.12 at 3:15 pm

I have argued elsewhere that any applied work is by definition normative, and that any application is potentially an experiment capable of yielding new knowledge. Academic non-normativity tends to be associated with formalization and non-empirical or anti-empirical systems building.

Some of these ideas were developed in Donald Stokes’s Pasteur’s Quadrant, which argues that Pasteur’s scientifically most important discoveries came from application oriented research (prevention and treatment of disease, suppression of mold and spoilage, and control of industrial fermentation processes).

Some of the impetus toward the claim of purity and value-freedom came from nuclear scientists who were trying to obscure the actual aim of their research, which was oriented toward killing as many people as possible as quickly and efficiently as possible. Some came from technocrats who were trying to marginalize politics. Some came from rightwingers who wanted to obscure their designs. Some of it was a resurgence of the age-old distinction between the sublunar and corrupt and the superlunar and pure, as manifest here on earth in the distinction between clergy and laity.


Mr.Violet 01.14.12 at 4:16 pm

I didn’t know anything about Mr.Reddy before!
It’s interesting because I had a very similar experience following, as an alien, the recent economic debate.
Here I have explained my last point of view on the issue:

Unfortunately I do not know the entire work of Mr.Reddy, so I have to base my comment just on the interview.
I perfectly agree with him that economics are imbued of assumptions about what is good and what is bad. And I perfectly agree that the first step is to acknowledge this.
I didn’t do any research on economics so I have no experience at all if it could be possible to work on economic problems without these kind of assumptions.
Anyway I don’t think any kind of science could be science when it is based on trying to establish what is good and what is bad, because is simply impossible to establish objectively what is good and what is bad.
I think that we can establish what is good and what is bad subjectively, when we’re looking in ourselves, or inter-subjectively when we are interacting with others. But an inter-subjectively established good is just a conventional good, not an objective one, so it’s not science, it’s politics.
Personally I do not think there is something bad in deciding in someway a conventional good and trying to pursue it. But the point is that this is a political decision. When we take political decisions in democratic systems my view of the good counts 1 as that of everyone else. It’s not that a certain dude, just because of applying what the economic model says to be The Good, counts 1 million, no that dude still counts 1 like me and anyone else.
Otherwise we get a kind of economic-totalitarianism, it doesn’t matter if it is a Keynesian dressed totalitarianism or a Neoclassical dressed totalitarianism, a fresh or salt water totalitarianism and so on. We are trying to decide what is good and what is bad for all of us on an objective base, but there isn’t any objective base for that.


Antonio Conselheiro 01.14.12 at 5:42 pm

#8 displays a most remarkable discomprehension of what’s at stake here,


Mr.Violet 01.14.12 at 7:33 pm

@Dear Mr.Conselheiro
you’re very kind in pointing out my misunderstanding, but as you put that I cannot understand anything more. If you wish to explain something more to me it would be good, at least in my opinion. Of course you’re not compelled to do that in anyway.


Antonio Conselheiro 01.14.12 at 8:48 pm

Essentially you just expressed the rationale justifying economists’ fraudulent claims to objectivity and neutrality, which is the claim that value judgments are subjective and cannot be part of science. Once value-freedom was established as the goal and standard some sort of disastrous miscomprehension of economics was inevitable. Also, you misused the word totalitarianism. Essentially your post is a simple restatement of the point of view Reddy was arguing against.


Henri Vieuxtemps 01.14.12 at 9:13 pm

I wouldn’t frame it as “what is good and what is bad”, but rather ‘who it serves’, what political/ideological ends.


Antonio Conselheiro 01.15.12 at 2:00 pm

The people who have realized that economics is intrinsically political are mostly conservatives. This was very explicitly stated by Nobelist James Buchanan who describes his work as intermediate between objective science (public choice theorist, somewhat Chicago Schoolish) and political thought (little-government Madisonianism). Hayek was also hostile to scientism, and according to Mirowski the difference between classical liberalism and neo-liberalism is neoliberalism’s recognition that liberalism sometimes requires a strong state. even unto Pinochet. (Neoliberals, however, have learned the deviousness of Straussianism and are quite willing to claim to be pure scientists).

Corporate pluralist liberalism’s naive trust in neutral scientific technocrats was its downfall. Greenspan was neither competent nor neutral and wrecked the global economy while protecting finance, and that’s the end of the story. What comes next I don’t know, probably le temps des assassins deux.


Bruce Wilder 01.15.12 at 7:09 pm

I was tempted, earlier, to complain that I found the interview with Reddy pretentitious and insipid — more the fault of the interviewer than Reddy, perhaps, but I didn’t find much that explained the poster’s enthusiastic affirmation, beyond the quip about reality.

Has capital-e Economics distanced itself from “any evaluative or otherwise ethical concerns”? Does it pretend to be value-free? I’m sorry, but I simply do not recognize the referent for such claims about Economics.

Way back when, textbook economics made a formal distinction between the “normative” and “positive”, but it was never to exclude the “normative”.

There’s a division in Economics between the orthodox and everyone else. I don’t know of any other social science with such a strongly patterned “orthodoxy”. That division reflects the way in which the right-wing has repeatedly over the decades won the struggle over methodological criticism in Economics, so that they define “rigor”, that all-important boundary between the orthodox and the barbarians at the gate. That “rigor” (some might say, “rigor mortis”, given the production of Zombies) orders values and valuation in Economics, but doesn’t exclude it, per se. That ordering, from Pareto efficiency to the EMH to the Lucas Critique, not surprisingly, favors strongly conservative convictions and views. In fact, much of the apparatus of the policy-oriented Macroeconomists consists of little more than mathematical formalizations, to justify strongly typed normative conclusions about whether a particular, stylized policy is pareto superior, a purely qualitatitve judgement at base.

The idea that economics, while studying how societies organize the production and distribution of goods, could be anything, but entangled — indeed, obsessed with — value, should be immediately recognized as absurd. But, is there really anyone making such a silly claim?

It seems to me that it is not values, but judgment and institutions, which have been under attack from the political and ideological right. The “orthodox” have their heads so far up their posteriors, that they cannot recognize a trillion-dollar housing bubble, or understand it as a threat to well-being. And, they have elaborate and formal justifications for such ignorance and stupidity. I see a detachment from reality, but not from “values”, per se.


Antonio Conselheiro 01.15.12 at 9:03 pm

But, is there really anyone making such a silly claim?

Basically, it’s the foundation of the technocratic approach. Neutral experts make wise decisions uncorrupted by politics. That’s the idea behind the Federal Reserve. How anyone could think of Greenspan as neutral is beyond me.


John Quiggin 01.15.12 at 9:19 pm

@Bruce – there’s an ambiguity here (one exploited very much by the right) between two concepts of orthodoxy (or, as it’s called from the inside, ‘mainstream economics’). One, includes essentially everyone except the self-consciously heterodox (Marxists, Austrians, institutionalists and so on) while the other is what used to be called the Chicago School, in which EMH, Lucas critique and so on are accepted. Someone like Krugman (or me for that matter) would count as mainstream on the first definition but not on the second.

Of course, as you say, the methodological insistence on rigor favors the right within the mainstream, as well as excluding the heterodox. A comparable instance is the use of the claim “capitalism has prevailed everywhere” to argue for the specific form of capitalism prevailing in the US as against social democratic alternatives.


Michael Harris 01.15.12 at 10:44 pm

Agree with John. If “neoclassical” is the term representing the orthodoxy, at least keep in mind that neoclassical can be used variously to be refer to (a) a very broad church of economists and economics (excluding, as JQ says, the deliberately self-styled non-neoclassical heterodox schools but including almost everyone else), or (b) capital-n Neoclassical, meaning Chicago-style “free market” economics, including but not limited to freshwater macro (excluding a fair chunk of the profession).

Ideology and methodology become fused in the latter case. Strict adherence to a fact/value distinction, strict adherence to optimisation and to microfoundations in aggregate models, strict adherence to ideas of stable preferences and the use of axiomatic foundations tend to be the sort of things that maintain zombie economics ideas that really should wither and properly die under the weight of empirical evidence.


Paul H 01.16.12 at 12:06 pm

It’s interesting to hear talk of ‘normative foundations’ for economics. But isn’t it the role of democratic politics to reconcile the implications of applied economic practice to ordinary practical/moral knowledge?

Also, I’m suspicious of the idea that it’s possible to list a set of ‘values’ that should then go on to inform the application of economic theory.
It assumes there is agreement on a set of moral rules that (apparently) exist outside the economics profession, and that the only thing necessary is to embed applied economics within these rules. I’d question whether there’s much agreement about the moral rules in the first place.

How do normative rules emerge? How are they generated? Do they come via a Rawlsian Original Position? Or some kind of continuous intersubjective Habermasian discourse?

Also, whilst I am not an expert, I wonder whether economists aren’t being a little harsh on themselves (and as a consequence further misleading the rest of us) when they self-flaggelate about ‘not predicting’ this present crisis of capitalism. As I understand it, some economists did predict it. The reasons the US, UK etc failed to heed these predictions are nothing to do with economics. The reasons are political/ideological.

And it’s the latter that seem to me to be in want of moral justification.


Michael Harris 01.16.12 at 1:37 pm

How do normative rules emerge? How are they generated?

I may be misunderstanding your questions, but the standard model for normative economics is consequentialist (utilitarian) ethics, traditionally focused entirely on outcomes as opposed to anything concerning the process that led to the outcomes.

This gets weird when the more libertarian economists get het up, because the whole incessant focus on markets as instruments of liberty is only incidentally, secondarily an argument about their efficiency properties (the consequentialist stuff).

Again, not sure whether this is what you were asking about, or whether you knew all of the above already.


Antonio Conselheiro 01.16.12 at 1:56 pm

Any challenge to pluralist liberalism will be met with the reiteration of the truisms of pluralist liberalism, as though the challenger had never encountered them, or had carelessly forgotten them and needed to be reminded.


Antonio Conselheiro 01.16.12 at 2:10 pm

As I understand it, some economists did predict it.

The economists who did predict it were not among the acknowledged leaders of the profession, and they have not been promoted since they have been proven right The economists who got it wrong have not acknowledged that they were wrong, and they have not lost their leadership positions. The doctrines which led to the error have not been revised or discarded.


Bruce Wilder 01.16.12 at 7:03 pm

Michael Harris @ 19

Economics has always been oriented to “process”, where the process in question is “the market”. All the most basic concepts, including “efficiency”, are defined within the conceptual apparatus of the perfectly competitive market and a canonical catalog of deviations, or “market failures”.

The philosophical framework of evaluation may be utilitarianism, but it is most definitely a utilitarianism in which the subjective enjoyments supposed as outcomes are treated as not just fundamentally unobservable, but as almost unquestionable and taboo. “Efficiency” is not a matter of measurable outcomes — it is strictly a process concept. And, indeed, the immunity to evidence and reality, which many economists exhibit is deeply related to this religious devotion to a faith in efficient market processes, the outcomes of which can not be questioned by the faithful.

What’s “weird”, I think, about this “process” orientation, is the absense of mechanism: the invisibility of the invisible hand in what is usually a merely metaphoric market.

I will venture the hypothesis that it is in the denial of mechanism and determined ignorance of institutional reality that technocratic economics may earn its curious reputation for being proudly value-free.

The glaring example in the lee of global financial crisis is a technocratic macroeconomic policy, devoid of any substantive concern for the regulation of banks and financial instititution. Jamie Galbraith has remarked that when he raised the question of financial fraud at an INET conference and at an IMF conference on the causes of the financial crisis, he was greeted with silence.

These are very much questions of values, calling for righteousness, but they call in vain, in a room full of economists, even economists ripe with “new ideas”.


Michael Harris 01.17.12 at 1:59 am


I’m at risk of getting further into defending-status-quo territory than I would probably have wished to. I was responding to Paul H’s query, but without knowing much of his background knowledge or understanding.

I still think your response is at risk of the conflation that John Q and I have both pointed out. It’s fair enough (IMO) to say that certain sectors of the economics profession link “markets” (individual agents endogenously and voluntarily interacting?) and “desirable consequences” (overaall welfare gains from individual agents endogenously and voluntarily interacting?) almost tautologically (see e.g. Steven Landsburg). I don’t think it’s true of the entire mainstream profession, except maybe to the extent that economists have historically tended to talk about markets relatively more than other processes, and in comparison to other social scientists.

But for a long time, other disciplines have had some fairly dodgy models of human behaviour, and their reach has far exceeded their grasp. I talked to an academic architect not long ago who ruefully reflected on the (now discredited) tendency architects had to build buildings whereby, they assumed, their amazing designs would result in predictable changes in human behaviour in and associated with those building designs. “If you build it [this way], they will [do this].”

Of course, people responded to their designs, but often in entirely unpredictable — or maybe, simply, unpredicted — ways. (Having spent months at the UC Davis Humanities and Social Sciences building, and having listened to the architect speak about it, all I can say is, man, no kidding. He gave an “I’m such a visionary” talk I attended, completely oblivious to the hostile reaction his design had engendered amongst the faculty and grad students who had to work in it.)

But back to economics: I don’t think the profession is nearly as self-aware as it should be about the nature of the ethical space it inhabits as a discipline. Geoffrey Brennan — a not-philosophically-unsophisticated economist — once described welfare (i.e. explicitly normative) economics as a branch of applied ethics. That’s not too bad a starting point for enriching the discussion, I would think.

But nor do I think that it would be a bad thing if our regular policy processes required more explicit and transparent uses of cost-benefit analyses, say. (I believe this is far more built into the system in the US re environmental regulations, say, than it is in Australia where I am.) Yes, cost-benefit analysis is not the only defensible paradigm or useful tool, but I’d generally like to see more of them done than fewer.


Michael Harris 01.17.12 at 2:38 am

Follow-up (I got lost in my reply as I was typing it):

The philosophical framework of evaluation may be utilitarianism, but it is most definitely a utilitarianism in which the subjective enjoyments supposed as outcomes are treated as not just fundamentally unobservable, but as almost unquestionable and taboo.

Here’s the default position, as I understand it. (By all means, critique away.)

Individuals are assumed to be the appropriate judges of their own well-being. Individual agents endogenously and voluntarily interacting are assumed to deliberately promote their own well-being, and non-deliberately improve the well-being of those they interact with. So, yes, utility is unobservable, and increased utility is presumed to be the outcome of voluntary action (“otherwise why do it?”).

Therefore the default position is to allow individuals to interact voluntarily unless there is deemed to be an a priori reason to believe that the individual action is welfare-reducing somewhere else (“market failure”). Then the question is, what to do? And then there will be lots of discussion/argument about what to do, with options ranging from “do nothing” through “do complex exercise in mechanism design” to “find something to nudge” to “implement big top-down policy initiative.”

This will be obvious to many, I am sure, but I’m spelling it in hopes of getting you to be more explicit about your own argument. Yes, human welfare is fundamentally unobservable, and yes, obviously “revealed preference” logic can be taken to extremes (as in, self-evidently individual X is optimally obese because if they weren’t optimally obese, they would choose to not be obese).

But this non-observability is intrinsic, it seems to me, so I think I am missing a key point here.

“Efficiency” is not a matter of measurable outcomes—it is strictly a process concept. And, indeed, the immunity to evidence and reality, which many economists exhibit is deeply related to this religious devotion to a faith in efficient market processes, the outcomes of which can not be questioned by the faithful.

Well, I work/teach in the broad areas of agricultural and environmental economics. Agricultural economics is good for learning about “government failure” (poor policy design, unintended consequences, rationalisation of redistributive policy actions), while environmental economics is good for learning about market failure, and taking it seriously.

You can’t do environmental economics and have lots of faith in market processes. We do things like surveys asking people to consider various trade-offs in potential environmental policies (“non-market valuation”). I submit that this is “efficiency” as a matter of measurable outcomes, not purely as market process. Of course, if you’re e.g. Mark Sagoff, you think this whole enterprise is misguided anyway, and that’s a non-trivial conversation to have.

Anyway, I don’t know if I’ve misinterpreted much of what you’ve said, or whether I’m just defending the status quo a little too enthusiastically.


john c. halasz 01.17.12 at 5:10 am


Briefly, two points:

It’s not really “individuals”, which are the main transactors of economic exchanges; it’s organizations. And it doesn’t follow that because economic exchanges tend to result in the realization of economic surpluses, that they are thereby unquestionably “voluntary”. Rather agents act under constraints, (which is why they are intentional agents at all, which is a much larger issue that just economic exchange), and the results of economic transactions can much better be traced to the different structural constraints under which agents are interacting and transacting than to “revealed” preferences and “voluntary” agreements. And I’m frequently under the impression that the quantitative estimates of economic gains or “dead-weight losses” are inaccurate, because of the failure of neo-classical marginalism to take into account the institutional context and the fact that the constraints of production systems are not those of market exchanges, based basically on a model of individual consumption demand.


Michael Harris 01.17.12 at 5:56 am


Point 1: agreed at least in principle. I was attempting to present a favourable kind of straw-man position in order to try to pin down the nature of the critique(s) above me. I had meant to make a side-note (particularly about firms, but organisations/institutions more generally) that we use “individual” very loosely, and too broadly, in many of the meta-narratives of economics (as though firms or governments really have nicely defined objective functions), but like I say, I got a bit lost in my own other bits of detail.

Point 2: quantitative estimates. This is interesting, although I am no sure what to make of it. As in “So, what are the implications of this for how I might do a benefit-cost analysis?”

Anyway, as far as dealing with my undergrad students goes, I try to downplay efficiency (to some degree) and place more stress on the more limited but practical notion of cost-effectiveness. There will usually be multiple ways to skin a cat, but some of those ways will cost much more than other ways, and it’s probably sensible to investigate those costs. Particularly if you’re dealing with a finite government (e.g. agency) budget constraint.


js. 01.17.12 at 6:54 am

But for a long time, other disciplines have had some fairly dodgy models of human behaviour, and their reach has far exceeded their grasp.

If you’re going to assert this, I’d like some evidence beyond your encounters with one out-of-touch architect.


Michael Harris 01.17.12 at 10:13 am

Two architects. One was (is) a professor at a major Australian university, and one who built a huge building at a major Californian university and then talked a boatload of utter bollocks about it at a public forum while its occupants seethed.

I could talk about the integrated social science initiative at my university (now defunct) that I was involved with. It was actually interesting in large measure — I mean, I read CT (rarely commenting) because it is a collection of above-average social scientists who do interesting work and have interesting things to say — but there is a wide range of quality in social science research (including in economics), not least because it’s difficult to do well, easy to do at a mediocre standard, and it doesn’t have the relatively settled methods of the physical sciences. Social sciences and humanities have been particularly parochial/insular in Australia because historically many of our academics have been trained at home and then often employed by the department they studied in.

(Economics, in Australia, internationalised some time ago. If you think all economics is rubbish by definition, this won’t matter, but in effect, the average university department here is closer to international standard in economics than it is in various other disciplines.)

Anyway, one of my jobs was to be on a committee evaluating social science research around the university for purposes of awarding prizes. Some of them made fascinating reading, although it was not clear in a number of cases whether there was a distinctly identifiable model of human behaviour at all. The empirical pieces tended to be inductive rather than deductive, but they were interesting to read and thought provoking. But the kicker was the PhD award, where we read three. I read one in detail, because while it wasn’t my discipline, it was in a topic area and literature I was familiar with and had a number of references in the bibliography I was familiar with. I started reading it sympathetically, hoping I’d find the take on a topic I was familiar with to be a fresh and challenging one.

All I’ll say (since I’m posting under my own name) is that had I been an external examiner on this PhD — which, don’t forget, was being put up for a prize — I would have failed it. It was a travesty. I even got to read the examiners’ reports, none of them by international scholars. Lazy, uninsightful and self-satisfied.

I had a getting-to-know-you conversation with my university’s new Vice-Chancellor soon after he started. He himself expressed some consternation about social sciences at the university, giving the example of an academic who had sent a young post-doc to discuss their work with the VC. The VC was immediately struck by some key weaknesses in the work which the young scholar seemed oblivious to. But the deeper concern was not that a sub-standard post-doc was working at the university; it was that there was a senior scholar who thought this post-doc was sufficiently impressive to send to meet the new VC.

I could regale you with the tale of a science colleague who is a born-again social scientist and believes he is doing “complex systems non-equilibrium economics”. He’s done this by (with his co-authors) imposing ad hoc behavioural rules of thumb (when they’re not arbitrarily assuming exogenous border prices). I wrote him about 10 pagges of critical comments ending with 20 questions about the paper, to which he has yet to respond. But if you don’t want to take my word for it, I could send the paper and my comments on it to John Quiggin. Since it purports to model an agricultural river basin, and John’s interested in Australia’s Murray-Darling Basin, I’m sure John would be interested in how someone else is analytically representing an economic model of a productive basin.

I could also talk about the planner, working for a state government department, who I did some sustainability-related consulting work some time back. He was full of ideas, full of notions, but he was all over the shop. Incoherent as hell. The small consulting team would silently laugh at me in our several-hour meetings as I’d grapple with him and try to keep a clear eye on what were trying to do; at some point, they’d see my eyes glaze over as my brain just gave up and shut down, and they would then take over our end of the discussion and bring it some kind of conclusion.

Of course, this is all anecdotal, and about individuals more than about entire disciplines. But those individuals exist — and in some cases rise quite high — because their scholarly disciplines have been pretty undisciplined.

If economics is to get fingers pointed at it — and it should — it doesn’t deserve to be singled out. Every social science discipline, particularly the ones with pretensions to policy relevance, need to lift their game.


Paul H 01.17.12 at 10:55 am

@19 Thanks.

Comments on this entry are closed.