Episode 85: Rawls on Social Justice

On John Rawls's A Theory of Justice (1971), most of ch. 1-4.

What makes for a just society? Rawls gives us a thought experiment: Imagine you don't know whether you're rich or poor or any of the other specifics of your situation (he calls this going behind "the veil of ignorance" into the "original position"). Now what principles would you pick to determine basic social institutions? Would you choose a caste society where you might be born an untouchable and be screwed? Rawls thinks that in this position you would instead support his basic rules of justice, which are (in short) to make sure everyone has basic liberties, and, more controversially, to allow only such inequalities as bring up the fortunes of those least well off. So you can allow super riches so long as doing so means that the poorest will be less poor than with any other arrangement; the default position is everyone getting an equal share of society's wealth unless you can demonstrate that letting some have more will benefit all.

This theory has been massively influential, and you can easily read it into Obama's speeches. Even many defenders of free-market capitalism do so on basically Rawlsian grounds. The founding fathers (Mark, Seth and Wes) and an especially energetic Dylan debate whether this original position is really coherent and whether it yields the principles that Rawls wants it to.

Read more about the topic and get the text, then listen to Seth's introduction.

End song: "Yours to Keep," by Mark Lint & the Madison Lint Ensemble, featuring Bob Linsenmayer. Read about it.

Please support the podcast by becoming a PEL Citizen or making a donation. Remember please to do any holiday shopping at Amazon you may do through PEL's Amazon link in the right margin of partiallyexaminedlife.com.


  1. Profile photo of Wayne Schroeder says

    I originally struggled with Plato’s Republic in the original greek, and now I struggle with Rawl’s Republic in the original rationalistic english.

    Thank you PEL for your patience and ability to bring Rawls’ content into a reasonable form for consideration, which is ironic because his project is supposed to start not only from a rational thought experiment, but ipso facto assuming that the individual having the thought experiment is rational (although some might argue that into the night).

    Equal liberty and fair distribution of inequality were summarized by Seth. He said, “If Rawls just said if you and ten other people were going to form a community in order to work together to benefit each other, everybody is going to work together because the overall benefit is greater when you work together than when you work individually, and you understand that when people work together they take on different roles that there are structures that need to be put in place, there are general rules of conduct that need to be observed, and there needs to be someway of adjudicating disputes and of distributing the benefits that get generated when everybody works together.”

    ” Now, imagine that you were going to do this but you didn’t know in advance that you were going to be in position one or position ten, how would you go about doing that? And his conclusion is that anybody that is faced with that position would want to structure society in such a way that they wouldn’t necessarily want people to be equal, and they would deny workers the fruits of their labor, all they would want is that you would have some measure against being exploited or forgotten by everybody else.” One person cuts the birthday cake and someone else selects the piece (p. 85), although “pure procedural justice is impossible, ” and slavish obedience to procedural justice may not lead to what is fair.”

    Mark then stated:
    “There’s some actually really interesting things in here, reflective equilibrium [coherence of one’s beliefs], ideal [the thought experiment] versus non-ideal ethics [failed states of the ideal], compliance theory [start with the ideal] versus partial compliance theory [we do not have to comply with unjust laws, etc].”

    Regarding the veil of ignorance, what appears to be the fulcrum for all of Rawls’heavy lifiitng, the Stanford Encyclopedia states:

    “Parties do not know:

    The race, ethnicity, gender, age, income, wealth, natural endowments, comprehensive doctrine, etc. of any of the citizens in society, or to which generation in the history of the society these citizens belong.

    The political system of the society, its class structure, economic system, or level of economic development.

    Parties do know:

    That citizens in the society have different comprehensive doctrines and plans of life; that all citizens have interests in more primary goods.

    That the society is under conditions of moderate scarcity: there is enough to go around, but not enough for everyone to get what they want;

    General facts about human social life; facts of common sense; general conclusions of science (including economics and psychology) that are uncontroversial.”

    Rawls’ theory is a yeoman’s effort which was rewarded during a time when analytic thinking held sway, but just doesn’t have enough horsepower to carry into this brave new world–the integration of rationality and experientiality. However his “veil of ignorance” is foundational to a new approach incorporating his very useful ideas.

  2. Roy Spence says

    I have some sympathy for your opening remarks on the difficulty in reading A Theory of Justice. (However, I had to read it twice.) As an alternative to reading the original A Theory of Justice from 1971, there is a more mature statement of the theory in Justice as Fairness: A Restatement written in 2000 and edited by Erin Kelly. The latter is only 202 pages and presents the main arguments with a revision to his principles of justice. There is also a chapter on various social systems with their political, economic and social institutions.

  3. dmf says

    Rawls, and the reactions against him, can be understood as an interesting pivot point in Philosophy as many were turned off by his attempts to valorize principles/concepts and wanted instead to focus more on matters relating to the accomplishing/negotiating of specific tasks of actual politics/living (not unlike some of what Kuhn did for science studies), is this the end of Philosophy or just a practice “turn”?
    let me throw http://plato.stanford.edu/entries/williams-bernard/
    into the PEL mix.

  4. Jake Z. says

    I was reading the SEP entry on Rawls after listening to the podcast and found this interesting bit: “On this topic Rawls is adamant: unless there is public funding of elections, restrictions on campaign contributions, and substantially equal access to the media, politics will be captured by concentrations of private economic power, making it impossible for equally-able citizens to have equal opportunities to influence politics regardless of their class.”

  5. Tom V. says

    I think it was a good thing that you guys spent a lot of time on discussing the merits of the method Rawls uses to come to his theory, and not on the details of his theory. Making explicit and examining the most basic assumptions is a big part of what philosophy does after all.

    I think Mark – i believe this was also Seth’s conclusion in the end – was overall spot on with his criticism on Rawls approach when he compares it to the approach taken in the federalist papers. If the theory doesn’t take into account enough of the world as it seems to be working, what is the use and value of such a theory in the end?

    I once heard a very smart guy say that ‘A theory of Justice’ was one of the best attempts at a rationalistic theory because of it’s detail and meticulousness (is that a word?), and that the real value of it, is that it exemplifies best why rationalism doesn’t work. I tend to agree, being able to elliminate certain methods is no small thing.

  6. DJ says

    Really enjoyed this episode. During your discussion, the use of terms such as “equilibrium” and “rational agents” really brought game theory to mind. Do you think Rowl’s project is actually just an application of game theory (tried and tested in economics and diplomacy) to the field of ethics?

  7. Mitch Hampton says

    I think there has been some profound misreading, confusion and indulgence in non sequitur among the many critics of Rawls, whether they be communitarians, Marxists, or conservative traditionalist. What they fail to notice is that the good and goods are inextricably intertwined with justice and fairness. The artificial separation of the good and rights is a reactionary narrowing of what is meant by fairness. The panel mentioned genital mutilation and it is actually best condemned under Rawlsian principle I.e. it is profoundly unfair to violate and change the futures of females in ways that create unfairness and inequality from the beginning. Rawls is a deeply moral thinker and fairness is an issue that gets at the heart of I believe most violations of good. And that is but one example. Another form of critique is to be moralistic and overreaching and describe as objective good those things that are primarily personal preferences rather than innate or universal law. Yet this criticism fails. The original position is something we should be all striving towards, that is, to transcend current social limitations. He wants to discuss minimal human goods and by definition has to leave out certain important values, not because they aren’t valuable but because they are not absolutely mandatory and for all people in all times. His theory is an elegant balance between the extreme Left and Right which is one of the strengths.

  8. Mitch Hampton says

    I wanted to say that this is one of the most sophisticated and thorough philosophy programs out there. No matter the topic justice is done to that topic. I appreciate the blend of humor and serious rigor. Bravo to all involved.

  9. Profile photo of John Pellow says

    Question: Assume Rawl’s thought experiment was workable through some special technology which could make a group of individuals forget their race, place in society etc. but know all other aspects.

    Is it even feasible for this special group of people to achieve a prescribed set rules/laws which could work as well as what America has now? Would there be real collective deliberations where all conflict gets resolved?

    Note: I do not think what we have works well enough, but also don’t think anything achievable exists as a better alternative – yet.

    PS: Expression of thanks to Mark Wes Seth & Dylan. I find your podcasts immensely helpful and humorous.

  10. Profile photo of Victoria Adams says

    My problem with Rawl’s has always been that I just don’t think it would convince anyone that that did not buy the thought experiment of the original position. Surely a dedicated believer of some religion would reply that all this might be true if you were behind the vile of ignorance but they are not – they have revealed truth. Once they have this they just don’t need all these intellectual gyrations. It seems to be it is easily enough accepted by people that agree and unpersuasive to people who don’t

  11. Andrew Davies says

    I just listened to this episode. I hadn’t read A Theory of Justice since my class on Rawls in the the mid-90’s. I loved the course, and returning to it with my older self was illuminating.

    First, I really enjoy Rawls’ thought process. It’s dry and laborious at times, but it sticks with me. Even when I disagree with him, I feel like his process enables my my critiques to teach me further rather than simply disregard what I’ve read. For me, that’s what is good about even the most “wrong” good philosophers. They bring you to a new place you wouldn’t be otherwise, even when you disagree with most of what they’re saying.

    Now for my critiques:

    The difference principle seems fundamentally flawed. It is Rawls’ way to mitigate the potential damage utilitarianism sometimes prescribes. But in doing so, he has a built-in principle of loss aversion. In real life, we have very few choices that don’t lead to a negative effect that we could have avoided by not making that choice. Are we to assume, then, that no scientific progress, energy efficiency, or medical intervention can ever be justified because it inevitably harms some people? Should no doctor ever prescribe chemotherapy since it will definitely kill some of its patients? Should the speed limit on freeways be limited to 30 MPH because we can calculate how many more people will die for each increase in the speed limit? No. In practical terms, the limits on utilitarianism intended by the difference principle are worth very little as stated.

    That being said, I think Rawls could have modified his difference principle to take “risk” (as discussed in the episode) into account. For example, he could have required that any harm caused by a utilitarian intervention (1) must be mitigated by placing the burden/costs of moral hazards on those who benefit, but only to the extent that the cost of the burden is less than the positive effect. (2) All actors in the intervention must be transparent about any unmitigated harm, which would (a) avoid corruption and (b) make it more likely that a solution (technological or otherwise) may be studied and found.

    However, stated that way, it’s still utilitarianism. A utilitarian would agree to any mitigating act with a net positive effect. Rawls’ difference principle fails because it seeks to reconcile net public advantages with individual advantages, but in the real world there is no way to draw a hard and fast rule to do no harm. Almost everything does harm to someone in some way.

    Ironically, Rawls’ original position could allow this utilitarian approach to risk. Throw out the difference principle, and people would generally agree that some acts, interventions, and policies carry a risk, but would still agree to them because of the net benefit.

    The law of diminishing returns already mitigates Rawls’ concern over individual harm. On the benefits side, doling out twenty million dollars to 100 millionaires has less utility than doling out ten million dollars to 1000 poor people. A millionaire simply doesn’t get a significant marginal benefit from 200 thousand dollars compared to the significant marginal benefit experienced by a poor person who gets only one thousand.

    The cost side also has diminishing negative returns. Experiencing excruciating pain one hour per week is horrible, but experiencing the same pain two hours per week is not twice as horrible. Maybe it’s 1.5x or 1.8x, but it’s not 2.0x.

    Consequently, Rawls’ difference principle itself is fundamentally loss averse. He would rather stick with the status quo than do something that hurts someone. Even doctors who take an oath to “first do no harm” don’t believe in that level of intervention avoidance. In fact, Rawls pretty much ignores the principles of behavioral economics that we are beginning to take for granted today. To be clear, it’s not that he doesn’t see that others have biases. He does. He just doesn’t see how his own loss aversion is coloring the underlying theory itself.

    Rawls would know what I’m talking about if he read about prospect theory, loss aversion, and other cognitive biases in the 2011 behavior economics book “Thinking, Fast and Slow” by Daniel Kahneman, winner of the Nobel Prize in Economics. #gratuitousnamedrop

    • John Halloran says

      Hi, Andrew,

      So I was really interested in your comment. I do, however, have a critique of your critique that I think is directly related to a bit muddle that was made out of Rawls’s project in the episode. In your paragraph talking about the difference principle you criticize it using examples of medical research and other safety related stuff, but such things would not really be subject to the explicit difference principle as stated by Rawls.

      What Rawls is doing out of the original position is not creating universal rules for evaluating the effect of individual decisions or individual pieces of legislation, instead he is deriving principles that guide the establishment of large-scale political theory and large-scale political institutions. In the American context, we can think of those as the principles that would guide the establishment of the Constitutions of the Federal and state governments (in the hypothetical). There is no mandate in Rawls that the difference principle be baked into individual moral decision making or even into broader political decision making (so long as the underlying structure complies with the difference principle)–that’s not the project here. In fact, Rawls takes pains to distance his theory of justice from individual moral systems of justice.

      That said, I do agree that you could certainly articulate the difference principle differently and still possibly effect the same conception of political justice. However, I’m not sure that one should read the difference principle as narrowly as you (or the PEL guys) do–I think that there are opportunities to make it more flexible in its current form. And, in spite of the calculations that Rawls uses to keep in conversation with the utilitarians, I think that what he’s doing is not rawly mathematical and, in a sense, doesn’t have to be utilitarian in the strict sense.

      Finally, and this isn’t anything you said, Andrew, more a reaction to the podcast in general: I think that the PEL guys got a little too hung up on liberal democracy as an outcome out of the original position. Though I think that it’s pretty clear that Rawls sought to justify liberal democracy, you can come out of the original position with his principles of justice and turn around and design a society that is a underlying political structure that’s not a liberal democracy in the fashion of the United States in the 20th Century. And, again, I think that the reason this gets a bit muddled up is because the narrow political purpose of Rawls’s theory of justice wasn’t well understood.

      • Andrew Davies says

        Thanks John. Very interesting. You really got me thinking when you wrote, “There is no mandate in Rawls that the difference principle be baked into individual moral decision making or even into broader political decision making (so long as the underlying structure complies with the difference principle)….” I’ve been mulling over the implications of that separation of the original position and its underlying structure (WITH the difference principle) from more utilitarian individual decisions (WITHOUT the difference principle). I’m really going to have to re-read some of Rawls in that context. I’ve noticed before that I sometimes conflate Rawls’ assumptions from different contexts.

        Your example of the U.S. Constitution sounds apt, but I must admit to sharing the biases as the PEL guys in this regard, which I agree are probably not the only possible conclusion of Rawls’ original position. That said, the U.S. Constitution does take great pains to protect the “weakest” factions in many ways. I believe Madison was a genius in his conclusion that the cure for factions was, in fact, more factions. I consider him to be a very early behavioral economist in many ways. If only we would learn from this principle today when it comes to corporations, special interests, political parties, etc. We react too often by limiting and controlling rather than by expanding and diffusing power. It is ironic that this philosophy often tends to protect the weakest quite effectively.

        At the same time, I still think Rawls’ theory may be far too risk-averse. In fact, I found an interesting paper which argues that the Difference Principle is actually “maximin”, or maximizing the minimum value. In other words, Rawls may be perfectly (or infinitely) risk averse. It’s a short but very interesting paper:


        The writer mentions that “Harsanyi is positing that in the original position people would have no level of risk aversion. Rawls seems to be assuming an infinite level of risk aversion. (1974 143)” But later, “Rawls says that one of the things blocked by the veil of ignorance is knowledge of one’s level of risk aversion. (149)” How to reconcile this? How can we make any “underlying structure” without knowing at least how risk averse we are?

        The solution the writer proposes is quite elegant. In the original position, we are risk ignorant, but uncertainty-averse. “So, as a first step, we can say that the uncertainty-averse person would prefer whatever equal distribution gives the most to each person – that is, the maximal equal distribution.” That’s maximin, or the Difference Principle. That’s probably the best economics-based argument I’ve seen for why the Difference Principle is necessary to the Original Position.

        In this context, my biggest problem with Rawls is that his setup of the Original Position is rigged. He has set up a system that will give him the outcome he wants. Not a specific outcome, mind you, but a biased outcome nonetheless. He offers little to convince me (1) that the people in the Original Position would actually prefer these stipulations on what they are or aren’t capable of considering, and (2) that these stipulations are fundamentally necessary to the Original Position. He assumes so much from the very beginning that his conclusions — while brilliantly argued — often seem to be an elaborate form of begging the question.

        Thanks for the feedback! I’m more interested and less certain than ever before. But since I strangely enjoy cognitive dissonance, I’m loving it.

        • Profile photo of John Halloran says

          This is super interesting, and you’ve gotten me thinking about exactly the breadth of moral ends that are allowable in individuals under Rawls’s two principles. He sidesteps that issue with his “reasonable comprehensive doctrine” requirement, but I’m not sure that I’ve been complete enough in my thinking about what the extent of that really is.

          And I really like your formulation of the Difference Principle in terms of risk averseness. I’m going to read that Schroeder piece. My favorite piece on difference principles is Philippe van Parijis, and it’s worth the read. It goes into some detail on the different possible formulations of the Difference Principle: https://www.uclouvain.be/cps/ucl/doc/etes/documents/2002.DP.CUP.pdf

          I like the Original Position as a thought experiment, though I totally agree with you that it would be really interesting to tweak the rules a little bit and see how that would change the results. Some of the things that are thrown into Original Position are a little confusing to me. I’m only an armchair political philosopher, but I would like to see more information in the original position (though maintaining the veil of ignorance around an individuals position in the society, and the nature of the society itself). Nevertheless, I really think the Original Position is a useful exercise in understanding whether a reasonable person would subscribe to the social contract in an extant society.

          Thanks for pushing me a bit. I think this is a useful back and forth.

          • Andrew Davies says

            I finished reading the paper you referenced, let it bake in my brain for a while, then came back to it and read a bunch of other stuff. I find myself both more appreciative of Rawls’ theory and at the same time more convinced that the difference principle is of little use in existing, practical choices whether they be institutional or individual.

            I keep thinking of three ideas that illustrate this:
            1. First, Do No Harm
            2. Loss Aversion
            3. Relativity

            First, Do No Harm — While not officially part of the Hippocratic Oath, this is a common idea among physicians. It has a lot in common with Rawls’ Difference Principle. The biggest difference is that doctors apply it to an individual patient within a very particular circumstance where available options are largely known. Even here, they often fail to follow the principle, especially in dire situations. For example, stage 4 cancer patients are too often given treatments that have a horribly low chance of saving them, and in fact, shorten their life while lowering the quality of that time. In such a situation, a doctor should think of the treatment like a game of Russian Roulette with a 20-chamber revolver where 19 of the chambers are loaded, and the 20th chamber is loaded with a painful salt plug. Inaction in such cases is a far better choice, but one that is very difficult for doctors and patients alike to accept.

            This individual Hippocratic principle can be applied to larger issues, such as the War on Terror. The amount of blood and treasure spent on it has little objective value other than mitigating our fear. As such, a “First, Do No Harm” principle could easily bring us to a much greater appreciation of accepting better choices when our expectations have been crushed. The Difference Principle, on the other hand, either (1) requires too much circumstantial ignorance to be of practical use or (2) doesn’t allow for the practical consideration that almost all decisions have both good and bad effects (even on a very high level), many of which simply cannot be sufficiently mitigated to justify the action or system. As such, it may simply lead to a paradox where no system will ever meet the demands of the difference principle, even though some systems will come closer than others.

            Loss Aversion — I know I’ve discussed this before, but I want to clarify that this is different than risk-aversion or uncertainty-aversion (both of which Rawls addresses). It “refers to people’s tendency to strongly prefer avoiding losses to acquiring gains. Most studies suggest that losses are twice as powerful, psychologically, as gains.” (Wikipedia is my friend). This tendency leads to sunk cost fallacy, where we base decisions on what we’ve invested or expected in the past rather than between real options in the present and likely outcomes in the future. You can see how the stage 4 cancer patient (and their doctor) or the War on Terror hawk are making this mistake. They are stubbornly, yet understandably, refusing to adjust their expectations based on the current reality rather than their preferred reality.

            Is Rawls’ Difference Principle fundamentally loss averse? In other words, does it consider losses to be worse than gains? In the case of the “least advantaged”, the answer is unequivocally yes. If instead he applied a principle of diminishing returns, he could retain the idea that hurting the least advantaged hurts far more than hurting more advantaged (or more than helping the more advantaged offsets it). But as I said, this would simply be utilitarianism. Are Rawls’ critiques of utilitarianism based more on his limited understanding of modern economics? That would be ironic, since understanding economics is one of the things he actually does allow in the original position.

            Relativity — I am fascinated by the philosophical implications and even religious allegories available in theoretical physics. For example, I like thinking that a solution to free will vs. determinism may lie in the two-slit experiment where a single electron, when unobserved, creates a wave’s interference pattern by actually going through slit 1, and also going through slit 2, and also going through both slits, and also going through neither slit. When we observe it, however, the wave “collapses” and the particle only goes through one slit. Does that solve determinism by effectively saying that all possible choices exist as real potentials in the physical universe until a sentient entity actually observes, aka “chooses”?

            But I digress. Relativity tells us that there is no objective time, speed, etc. It only exists when comparing things to one another. For me, this may be a fundamental flaw of Rawls’ theory as a whole. The PEL guys implied this when they criticized Rawls for returning to a “theory first, real world second” method that was more common with ancient and medieval philosophers. In effect, Rawls’ “veil of ignorance” may be creating a logical paradox, or possibly a type of “uncertainty principle” where the information required to actually make “original position” decisions is unavailable, since the framework itself requires comparing things as they actually occur (economics, for example). Effectively, any “Theory of Justice” might need to accept that its conception of the good only exists in relative terms.

            I love this idea of relative measures, because it promotes things I like in making choices and negotiating change: (1) circumstantial considerations instead of ideological ones, (2) present choices matter (call it “ethical superposition”) rather than the present state (loss aversion), (3) “problem definition” or a clear definition of the goal in order to compare alternative, (4) objective and common metrics of effectiveness, and (5) acceptance that nothing is altogether good or altogether evil, but is simply better or worse, and only within narrowly defined metrics.

            Holy crap, that was long and dry. I’m turning into Rawls.


Leave a Reply

Your email address will not be published. Required fields are marked *