Indifference and Fake Indifference - Theoretical Edition, Part 2

I was really trying to move on to a new issue this week but this whole business of fake distributional indifference and the concept of Pareto improvements strikes me just now as rather fun and interesting, which I suspect may be a somewhat unusual state of affairs, so I thought I might do one more on that issue first. I actually tried to just go back and add a little something to my last post to dispose of this issue, but I got myself a tad confused, for reasons I’ll mention here maybe at the end as a bit of comic relief, so I put my previous post back the way it was and resolved to do another one today. I’ll save the rhetorical device of references to possibly non-existing distributional mechanisms for next time.

Last time out I discussed how the concept of Pareto improvements can generate fake indifference if one falls into the trap of adopting some sort of informal ethical proposition to look only at Pareto improvements from any given distributional starting point because, of course, according to utility as it is defined and used in economic theory, one should really be indifferent between those Pareto improvements and other policies or changes that involve redistribution, with at least some people becoming worse off relative to that starting point and those resources going to other people who some people may suppose have a stronger ethical claim to them. I believe the proposition one should consider only Pareto improvements is not a part of economic theory proper, which is why I think it belongs in the outer layer of bad economics with the other rhetorical tricks associated with defeating distributional indifference, but one can well understand how some people may come to suppose it is part of economic theory. Distributional indifference is ostensibly designed to avoid controversial ethical issues and if one isn’t paying close attention one may come to suppose a restriction to look only at Pareto improvements does the same thing.

However, there is another way the concept of Pareto improvements can generate fake indifference. If one considers two policies both representing Pareto improvements relative to any given starting point because they make at least one person better off and no one worse off, the choice of those Pareto improvements involves resolving an interpersonal conflict or distributional issue, namely, who should get the raise. People supporting one set of distributional ethics may believe one policy justified while those supporting some other set of distributional ethics may believe the other policy justified.  In other words, Pareto improvements are rather unsurprisingly like Pareto optimums, everyone may like them in the abstract, and one can talk about the class of Pareto improvements and Pareto optimums without raising ethical conflict relating to distributions, but everyone doesn’t like the same one, and as soon as you start talking about instances of Pareto improvements and Pareto optimums you’re in a world of distributional indifference or should be anyway.  Economic theory is indifferent between making Pareto improvements and redistributing resources, making one or another Pareto improvement, attaining one or another Pareto optimum, and maintaining any given Pareto optimum or moving to certain other non-Pareto optimums (those would be the non-Pareto optimums that are not dominated by that particular Pareto optimum, that is to say, the non-Pareto optimums that have the quality that one cannot get to that particular Pareto optimum from that particular non-Pareto optimum via Pareto improvements). Economists exhibit fake distributional indifference when they explicitly or implicitly treat one side of these oppositions differently than the other.  If they consider or discuss or analyze or look at only one side of these oppositions, that’s fake indifference.  If they give policy advice predicated upon or consistent with or accepting of only one side of these oppositions, that’s fake indifference.

So just by way of summary so we don’t lose our way entirely, what does economic theory actually say in this area anyway? Well, I think the simplest way to say it is that it shows for any given person, with a given set of distributional beliefs, there will be some path from any given distributional starting point through Pareto improvements consistent with that person’s distributional beliefs to a Pareto optimum consistent with that person’s distributional beliefs. That means for any given person there should be some Pareto optimal, economically efficient, perfectly competitive market outcome that should look better than other market structure under the famous conditions for perfectly competitive markets and setting aside some of the other ethical issues we’ll be discussing in other posts relating to utility and market systems. However, when one adds another person with different ethical beliefs relating to distributions one ends up with what should be very far reaching indifference indeed, but which is often ignored on various bases in practical situations, a phenomenon I’ve been calling fake indifference or more properly fake distributional indifference. Fake distributional indifference defeats real distributional indifference and makes economic theory appear to be a player in disputes about distributional ethics, supporting or accepting some distributional results, shooting down or blocking arguments for redistribution or other distributional results. It’s really a misuse of economic theory for rhetorical purposes. Adding explicit ethical propositions, necessarily relying on some basis other than utility, that address distributional ethics would be fine, propositions like one should accept any distribution one happens to find because it was there when one got there, or it was the result of some particular political or social mechanism, or it was supported by some group of people, etc. However, if one intends to add something like that in a context of a theory that makes as big deal about utility and distributional indifference based on utility that economic theory does, then one will want to be as clear and explicit as possible. One wouldn’t want to slip it in unawares, at least if one were being intellectual honest.

And just for fun, what was the issue I alluded to earlier that got me mixed up when I tried to do a quick revision of my previous post? Well, I started talking about Pareto improvements in terms of money or economic power and all manner of oddness ensued. It seems apparent Pareto improvements defined in terms of money or economic power are not necessarily real Pareto improvements because when one changes relative economic power one can change how interpersonal conflicts are resolved in markets. For example, if A has $1 and B has $2, and they both want unique good G, then B will get G, but if we make an apparent Pareto improvement based on money or economic power so A now has $3 while B stays at $2, A gets G and B is worse off. It wasn’t a real Pareto improvement. So what is the proper basis for defining Pareto improvements? Utility, I suppose, although talking about the level or amount of utility sounds like it involves a cardinal conception of utility that would not be consistent with the common ordinal only “preference utility” definition of utility often used in economic theory. Well, doesn’t really matter for what I’m talking about, so I won’t bother looking into it. Economic theory is funny. It’s been around for years but delving into it is like walking through Grimpen Mire or Fangorn Forest by moonlight. One false move and you may be lost forever. The Dark Forest of Economic Theory.

Indifference and Fake Indifference - Theoretical Edition

I know I meant to take up the add-on rhetorical device of references to non-existing distributional mechanisms this week, but I had an interesting discussion with a fellow economist the other day that made me think I could probably say a few more words about the rhetorical device of fake distributional indifference before moving on, specifically for those a little more familiar with economic theory. Normally, I try to write for what I suppose is an averagely educated person in the street who may remember a thing or two about economics from their all but obligatory dose of neoclassical welfare economics in Economics 101, so I assume some background, but I try to eschew economic jargon as much as possible. However, that can actually complicate communication with those accustomed to talking about those issues using that jargon. So this week I’m doing the same thing but with a bit more jargon. By all means give it a shot. I’m still trying to explain it as I go as much as I can, although I cannot of course deliver a course on economics in the space of this blog post. However, if you’re put off by the unfamiliar terms or indecipherable references to theory please rest assured I’m just doing another version of what I did last time in a different context.

So-called “Pareto improvements” are changes that make at least one person better off and no one worse off. They’re of theoretical interest in neoclassical welfare economics because we can use them to show the perfectly competitive market structure performs better (under the usual assumptions and conditions) than other market structures without bringing up potentially controversial distributional issues or ethics. Applying this concept gets one to an “economically efficient,” often somewhat confusingly just called “efficient,” “Pareto optimal” market result, which is a result from which no further Pareto improvements are possible. The concept plays an important theoretical role in establishing that for any non-Pareto optimal result there is some Pareto optimal result that is preferable to it, or “dominates” it. Basically, under any given distribution of economic power, there will be some efficient, Pareto optimal, perfectly competitive market result that dominates the other market structures under that distribution of economic power and hence consistent with the ethical beliefs supporting that particular distribution of economic power.

However, some funny things can happen with the concept of a Pareto improvement. One thing that may happen is one may incorrectly suppose any Pareto optimal result dominates any and all non-Pareto optimal results. That is incorrect because starting from any given distribution Pareto improvements can only get one to certain Pareto optimal results, not all Pareto optimal results. In particular, any Pareto optimal result that corresponds to a distribution of economic power in which at least one person is worse off compared to the starting distribution will not be available via the process of making Pareto improvements. In other words, economic efficiency or Pareto optimality are not global concepts. Any given Pareto optimal market result will dominate certain non-Pareto optimal market results, but not others. Any given economically efficient market result will dominate certain non-economically efficient market results, but not others. Any given perfectly competitive market result will dominate certain non-perfectly competitive market results, but not others. In the cases in which one result does not dominate the other we have distributional issues we cannot resolve using the definition of utility used in economic theory, that is to say, we have distributional indifference. So within the constraints of economic theory one should be indifferent or neutral not only between different Pareto optimal points on the so-called Pareto frontier (the collection of all such points across all distributions of economic power) but between certain Pareto optimal results and certain non-Pareto optimal results, certain economically efficient results and certain economically inefficient results, and certain perfectly competitive market results and certain non-perfectly competitive market results. 

Another funny thing that can happen in this area is the ethical or value proposition that economists should consider or discuss or analyze or offer policy advice only on the basis of Pareto improvements may be introduced into economic theory or in discussions ostensibly based on economic theory. Restricting oneself to Pareto improvements breaks distributional indifference because it treats existing distributions differently than other distributions to which one should be indifferent under economic theory. It gives special preference to existing distributional arrangements and the ethics supporting those distributional arrangements. Restricting oneself to Pareto improvements, by precluding consideration of changes that make anyone worse off, are consistent with ethical beliefs relating to distributions that suggest everyone has an ethical claim to whatever he or she happens to presently have, so we should not consider any changes that make anyone worse off, and are inconsistent with ethical beliefs relating to distributions that suggest some people presently have rather more than is ethically justified and should, in fact, be made worse off in order to make others who have stronger ethical claims to those resources that much better off. Introducing the ethical proposition economists should only consider Pareto improvements contradicts distributional indifference based on the definition of utility and really requires explanation of whether it is meant to supplant and negate distributional indifference in terms of neoclassical welfare economics proper, or whether it’s meant to be some sort of subset or sideshow that investigates what happens when one adopts that particular ethical perspective. 

Neither of these two awkward bits appears to be stressed in common accounts of economic theory, for what one can only suppose are rhetorical reasons, leaving student, observers, and one suspects even some economists understandably confused about the status of distributional indifference in neoclassical welfare economics. Interestingly, the natural result of both are the same: an exaggerated attraction to arriving at or maintaining perfectly competitive markets and preserving at least some features of existing distributional patterns, and a failure to appreciate or acknowledge what should be indifference or neutrality to doing otherwise.

Just to be clear what I’m arguing here; it’s perfectly fine for economists to hold ethical beliefs relating to distributional issues, and it’s perfectly fine for them to develop a theory expressing their beliefs. That’s not the problem I’m addressing at all. People can have ethical beliefs and argue for them. But if they do, they should be clear they’re doing it, own it, explain it, take steps to avoid any confusion on that point. It’s misleading to make a big deal about distributional indifference in the context of defining utility then go beyond it later based on some other and unstated ethical reasoning without highlighting the fact. Economists have certainly had plenty of time to do that, and their failure to do so can I think only plausibly be ascribed to incompetence as far as the ethical philosophy they’re dabbling in or bad intent and a willful impulse to deceive. We need to fix bad economics and in particular confront the rhetorical device of fake distributional indifference. Distribution indifference is either there or not there. Choose one. No one can have his or her cake and eat it too, not even the mighty economist.

Indifference and Fake Indifference

Last time out I talked about the three layers of the metaphorical onion of bad economics and I said the outer layer was composed of composed of rhetorical tricks of the economics trade that although not really part of economic theory, per se, are commonly added on to it. I mentioned these informal add ons generate a great deal of ambiguity and confusion because they create a split between what economic theory actually says and what it is commonly presented as saying or implying. Two such bits of rhetoric I mentioned specifically were fake indifference and nonexistent mechanisms to address distributional concerns. I thought I might take up the first one today and save the other for my next post. I should clarify that by fake indifference I mean fake distributional indifference specifically.

So what is real distributional indifference? Well, let’s say we have two distributional mechanisms or results, an existing one A and one of a potentially infinite number of alternatives we’ll call B. If one were truly indifferent between these distributions, if one really expressed distributional indifference, one would treat them exactly the same. One would be entirely neutral, impartial, disinterested between the two. A critic would never be able to suggest one gave one special status relative to the other. Supporters of both A and B would be equally happy and satisfied with one’s policy advice and recommendations. In the real world, in which one is constrained to either retain some particular distributional mechanism or result or move to some other distributional mechanism or result, and in which any practical policy advice one might give inevitably involves implicitly accepting or not accepting some distribution or redistribution, a theory that expresses real distributional indifference is irrelevant in a practical sense.

What then is fake distributional indifference? If one exhibits fake distributional indifference, one does not treat A and B exactly the same. One treats them differently albeit perhaps subtly differently. One grants one of the two, usually A, some sort of special status, usually implicitly through such means as giving policy advice consistent with A but not B, predicated on A but not B, accepting of A but not B, etc. The basis of granting differential status to A under fake distributional indifference could be and usually is quite general in nature and typically involves propositions like it was here when I got here, some other people like or support A, some legal or political mechanism led to A, etc.

The reason I took pains to clarify at the outset I was talking about fake distributional indifference rather than simply fake indifference in general, and the likely the reason this issue creates so many problems, is that one can, of course, be indifferent to A and B in a sort of second order way in the sense one might be prepared to support either A or B depending on which of the two happened to be here when you got here, or other people liked or supported, or was the result of some legal or political mechanism. However, that second order indifference relating to the results of some process or fact ostensibly granting special status to A is conceptually distinct from true distributional indifference between A and B. It’s not distributional indifference, per se, but something else. It constitutes an additional ethical proposition one should acknowledge and address to avoid confusion. What if someone says they think there’s nothing ethically special about what was here when economists got here, what some other people accept or support, or what was generated by some legal or political mechanism? Would that be relevant to assessing practical policy advice ostensibly based on neoclassical welfare economics, or would it be an irrelevant consideration that has nothing really to do with the matter at hand? If one grants any sort of special status to a distributional mechanism or result on any basis, that proposition represents a potentially controversial ethical proposition that should be acknowledged.

A great deal of confusion is generated in bad economics by economists pretending to express true distributional indifference in accordance with economic theory when, in fact, what they’re actually expressing is fake distributional indifference that adds ethical or normative content to economic theory. The natural result is economists seeming to talk out of both sides of their mouths, claiming to take no stand on distributional issues, but then indirectly or implicitly supporting some distributional mechanism or result by granting it special status and giving policy advice or recommendations predicated on that distributional mechanism or result and neglecting other potential distributional mechanisms or results that really should have equal billing.

Properly interpreted, neoclassical welfare economics says some interesting things, but in the same way a logical or mathematical parlor game says some interesting things. It doesn’t say things that have practical relevance to the real world unless the theory is informally combined with additional ethical propositions added on at the level of explication or practice. Economists should either interpret and use neoclassical welfare economic theory correctly and treat it as a parlor game with very limited and possibly no practical significance to real world situations, or they should explicitly acknowledge and defend the additional ethical or value propositions required to make it relevant to the real world, and designate those ethical propositions a proper part of economic theory. It’s just too confusing to make such a big deal about the implications of modern economic theory’s famously attenuated definition of “utility,” and in particular distributional indifference, and then introduce into the theory unrelated ethical or normative propositions involving property rights and the use of market mechanisms, and then add to the theory when using it random ethical or normative propositions relating to distributional issues. At some point economists will have to start getting a little more rigorous in their ethical philosophizing or just stop trying to tell people what economic arrangements are socially optimal and so on.

The Onion Of Bad Economics

It occurred to me the other day another potentially useful and visually evocative way of expressing what I said a few posts ago about bad economics is to say it has layers, like an onion. Also like a parfait, which I know everyone loves and is also a perfectly reasonable choice. But let’s go with onion. Seems a more apt metaphor for bad economics.

How many layers are we talking about? Well, I’m counting three right now, but I wouldn’t be at all surprised if I think of a few more later. Seems to happen every time I make a taxonomy of anything. I make a few nice categories that seem to cover everything and then something else crops up in short order to ruin the beauty of my scheme, usually the very next day. Awkward. So let’s just say our onion has three layers, at the moment.

The first layer, or core if you will, of the onion of bad economics is all the funny business associated with the term “utility,” which in contemporary neoclassical economic theory serves no purpose really beyond acting as a red herring to get people thinking and talking about ethical utilitarianism. The actual ethical propositions explicitly expressed through the ostensible objective of “maximizing utility” using the term “utility” in this layer aren’t particularly controversial. We have the implied ethical proposition that economists qua economists will not address the thorny ethical issues associated with resolving interpersonal conflicts of needs and desires, at least not officially on the basis of utility or economic theory. And then we have the rather more substantive ethical proposition that, if someone is acting in isolation and has no effect on anyone else, we should let him or her do whatever he or she prefers to do, or in the fanciful terminology of modern neoclassical welfare economics, we should act in a way that maximizes (social) utility by allowing the person in question to maximize his or her own (personal) utility. Seems fairly innocuous. However, even this most benign layer isn’t entirely devoid of potentially controversial ethical content. One might, for example, be dealing with someone who has preferences involving harming animals, or planet Earth, or some other entity or thing someone might reasonably have ethical beliefs about even when he or she is not around to be directly involved. But one has to do a bit of work to think of potentially controversial ethical situations in this layer. Takes a bit of imagination. In normal situations, most people would probably have no problem with the ethical propositions involved. This layer is, of course, a core part of neoclassical welfare economics even when that theory is properly interpreted.

The second layer of the onion of bad economics is composed of ethical propositions not based on “utility” nor even expressed using that term that are introduced into economic theory by the suggested, implied, assumed use of markets to address interpersonal conflicts on the marketplace, that is, property rights and the use of the market mechanism. There are a few somewhat more potentially controversial ethical issues introduced in this layer. Like what, for example? Well, questions like the following. Should we support the market mechanism as a way to resolve interpersonal conflicts of needs or desires based on economic power in the marketplace in situations where the typical behavioral assumptions associated with economic theory are not met? For example, let’s say one party to a market transaction is not thinking rationally, or doesn’t have full information? There are also situational ethics issues that may not relate directly to explicit behavioral assumptions typically associated with economic theory. For example, should we support the market mechanism as an ethical way to resolve interpersonal conflict on the basis of economic power when one side to a transaction is desperate, or has less bargaining power, or there are power imbalances of any sort, or in the extreme case where one party has no economic power at all, which is the case involving future generations? This second layer also involves ethical issues relating to the property rights that are necessary for a functioning market system, including at least the implied indifference to changing or revising property rights, which seems a logically necessary corollary of distributional indifference but which sometimes seem to go unremarked creating no end of awkwardness. Indeed, as with “utility” the language associated with property “rights” seems a bit odd. A right one is indifferent to changing? Funny sort of right, that. This layer I think must also be considered a bona fide part of neoclassical welfare economics properly interpreted.

The third layer of the onion of bad economics is composed of misinterpretations and errors that make it seem as though the theory says a lot more about distributional issues than it really does. This outer layer is associated with the most ethical controversy and is by far the most offensively pungent of the three layers. Yes, I realize that’s not necessarily how actual onions work, but I’m talking about a metaphorical onion, so please allow me the artistic license to suppose my virtual onion is more pungent on the outside than the inside. Honestly. What difference does it make? This third layer includes various bits of nonsense like fake indifference, references to nonexistent means to address distributional issues, and various other rhetorical tricks of the trade. I talk about them all the time and will undoubtedly go over them again, many times, during the course of this blog series, so if you’re not sure what I’m talking about you have that to look forward to. Or you can read my little book on the subject, which probably needs some revising by now, but anyway the biggest and most annoying bits of rhetorical nonsense are there in rather scattershot fashion. This layer, unlike the others, is not really a part of neoclassical welfare economics properly interpreted but consists of commonly seen and commonly accepted or at least uncommonly refuted add-ons, which is a feature of bad economics that creates a great deal of ambiguity and confusion on its own. Indeed, I suppose the difference between what economic theory actually says and what many people, including many economists, pretend it says may be the feature of bad economics that has done the most to confound its critics and allow bad economics to carry on unchanged and unperturbed for as long as it has. In many way, bad economics is like one of those shape shifting creatures of myth; hard to pin down or get a firm grip upon. Fine, too many metaphors. It’s an onion.

It helps when addressing the opaque, inconsistent, and just bad ethical philosophizing commonly associated with bad economics, flawed popular presentations of neoclassical welfare economics, if one tries to keep in mind which layer of the onion of bad economics one is dealing with. Otherwise they’ll have you going in circles. I may just try to do that in this blog series to add some method to the madness. I mean keeping in mind which layer I’m talking about, not going in circles, although I’ll probably end up doing that as well. So, if you hear me later referring to layers of bad economics, this is the scheme I have in mind. And by the way, just for the record, I have no problem with actual onions. They’re delicious and healthy. Who doesn’t enjoy a nice onion? The virtual onion of bad economics is something else entirely.

Addendum

I’ve lately moved away from the idea of implicit normative propositions not based on “utility” being a legitimate part of neoclassical welfare economics in favor of the line we should treat as exogenous any normative proposition going beyond what we can say based on “utility.” However, it should be noted taking that line implies neoclassical welfare economics expresses indifference to whether we choose to use markets to resolve any particular interpersonal conflict of preferences, or any such conflicts at all, or indeed whether we even have the legal framework to create or sustain such a market. I think it clarifies the normative content of neoclassical welfare economics, but at the expense of removing much of the normative relevance of that theory for the real world. For example, see the post Law Over Anarchy In Neoclassical Welfare Economics from June 2, 2021.