# The Samovar

Fairness and equality

There has been a lot of discussion of a new report by the Joseph Rowntree Foundation on attitudes towards tackling economic inequality (at Directionless Bones, Left Luggage, Sunder Katwala on CiF, Don Paskini at Liberal Conspiracy, and David Osler). Quoting Alderson at Directionless Bones, one of the key findings of the report is:

People didn’t seem to endorse the idea of ‘equality’ as a general principle as much as they endorsed ‘fairness’.

This is a point that several of the posts linked to above considered, and there has been a feeling that the left needs to find a new way to promote their view of the world to people (which traditionally is based on equality).

I find this interesting because for the last few years I’ve been coming to the view that the case for a left-wing politics should be based rather on the principles of fairness and freedom than on equality. Equality is important and essential, but I think it’s a consequence of fairness and freedom. I’ve argued this in much more detail in an earlier entry on capitalism. Essentially, a very unequal society will, in practice, also be necessarily an unfair one.

The report has caused quite a lot of distress because it showed that people are not against what they term ‘fair inequality’ – but I don’t think the left should be dispairing over this. The report also clearly showed that people think there are substantial levels of unfair inequality, and that this is a bad thing. This suggests to me that most people basically get and agree with the final point of the previous paragraph – high levels of inequality lead to an unfair society.

I suggest then that what the left needs to do is to push this analysis further and address the misconceptions that the JRF report showed that people have. Most people substantially underestimate the level of inequality that actually exists and overestimate the level of social mobility. Changing perceptions of these is difficult, but could make a significant difference.

As a final point, the report appears to be more of a blow to a state-centric form of socialism where equality is considered more important than fairness and freedom, and much less of a blow to an anarchist form of socialism which takes freedom and fairness to be fundamental. This is important and suggests the left should be considering a change of direction towards anarchist conceptions – and thankfully much of the left does, slowly, seem to be doing this (even if they don’t call it anarchism). In particular, the point about support for ‘fair inequality’ is very interesting with respect to the remuneration mechanism of parecon (which I written a few things about on this blog). Parecon allows for a certain amount of precisely ‘fair inequality’ – that is, inequality that comes from a choice to work harder or at more onerous labour. It is fair because anyone can make that choice (whereas not everyone can choose to be a doctor, banker, etc.). On the other hand, it absolutely rejects unfair inequality. As such, it seems that many people’s fundamental views of what society should be like resonate more with a pareconish or anarchist conception than a state-centric socialist one.

Yeah, I think another point is that if inequality gets related to power, then its connection with freedom is direct – if someone else have more power over me than I have over them, they’re in a position to make me do what they want, even if I’d rather not, and this makes me less free. Whereas simply having one person perhaps eating 5 cakes and another eating 2 (over a week, not in one go) doesn’t. Merging the idea of equal power and equal consumption obscures this connection with freedom – conveniently for authoritarians.

Comment by Alderson Warm-Fork

One difficulty with fairness is that it is a somewhat fickle criterion; there are people so depressive or in such poor mental health that they cannot be made to work at the present moment, and probably under any economic system such people would exist. Yet many people, perhaps most, would agree that it’s fair to not pay “idlers” like this. I seem to recall even Michael Albert once saying that he personally didn’t see any obligation of society to produce a vagabond’s wage, once society was organized more justly at any rate. (Although he also said that this was a matter independent of parecon, which I agree with.)

I think the thing to focus on, beyond simple material needs, is that each person is able to feel a sense of dignity about herself. And under gross inequality this cannot happen — all the more so when this inequality is entirely uncorrelated with the amount of work being done. (Perhaps this is just a permutation of words though?)

This post was reposted at Liberal Conspiracy, and you might find some of the discussion there interesting. Some of it is – some of it is pretty stupid.

Alderson – yes I thought that was a very good point in your article you wrote. It might almost be the key insight of anarchism?

Brad – I think in the things I’ve read, Mike Albert has suggested that he thought it would be likely that a civilised parecon society would choose to provide for people like that, but I guess he recognises that it might not. I think that’s a reasonable point from his perspective of trying to promote parecon.

I suppose the question is about what you include in ‘fairness’ – is it unfair that some people are unable to walk? or is it something other than fairness?

But in the end, it’s OK if fairness doesn’t cover everything. I would rather have fairness + basic decency as first principles than having equality as a first principle, even if the latter implied the former (which I’m not sure that it does).

Comment by Dan | thesamovar

[...] the social… on Left at a losswageslavedave on Stressing the social in anti-s…Fairness and equalit… on Inequality and the battle of…For Revolutionary Mo… on Inequality and the battle [...]

i’m uncomfortable with having “fairness” as some kind of founding principle. i think freedom is a much more powerful notion. fairness has something rather condescending (in a sort of liberal-english kind of way) about it. fairness also suggests to me a notion of power that is withheld, perhaps artificially, which i think is problematic. freedom has enough richness within itself as a concept to ultimately lead to any notions of ‘fairness’ and ‘equality’ that are necessary while simultaneously rejecting statist or other restrictions on freedom that might be imposed in the name of ‘equality’. of course, i’m using freedom in the ‘subversive’ sense rather than as a shorthand for personal (capitalist) irresponsibility. true freedom is not even necessarily defined at the level of individuals — i would say that the conceptual challenge facing the left is to properly extricate the meaning of freedom.

Comment by sean h

Hey Sean, I’m very much with you as regards freedom. In my earlier article (which we discussed I think already) I argued that fairness+freedom=>equality. I do think that fairness is an important principle though – I’m not sure that just freedom on its own does everything we’d like it to do. I don’t see why you’d say that the notion of fairness is problematic. I mean, in reality, society has almost never been fair, the dice have been loaded in favour of particular classes, and that’s something we’d like to do differently. No? I don’t really get what you mean by the notion of power withheld.

Comment by Dan | thesamovar

I’ve had a look at your ideas here and in your earlier post on capitalism, which I thought were thoughtful and open-minded. I expect you’re probably not going to agree with my own views – and I don’t have an issue with that – but you look like the sort who might like to hear about arguments from other viewpoints, even if only to be able to argue against them better.

In a perfectly free market, the wealth distribution tends towards the scale-free power law distribution. This is because the economy has a fractal structure with self-organising criticality – groups of people acting as a single economic agent – and renormalisation group theory predicts that as you look at it at larger scales, all the small-scale complexity cancels out apart from a distribution that has to be invariant under scaling.

Basically, if you plot number of people against wealth, you get a sharp peak at the minimum survival level, that tails off at a certain rate for increasing wealth. The rate at which it tails off varies, depending on the propensity to invest, but the generic shape stays the same.

So with this model you have a choice between a rapid decay – everyone just above the survival level and nobody rich – or a slow decay – a lower, broader peak that tails off into the rich and the super-rich and so on. In the second option, the average wealth of society is much greater, and there is a much smaller number of poor people. However, the scaling law requires that there always be more poor people than rich people, and that the richer society is, the more unequal it is. In the context of a free market, inequality is good!

That said, there is another situation where the market isn’t free, but split. Bureaucracy, barriers to trade, and the distortions of tariffs and subsidies place a barrier to trade that you need a certain amount of expenditure to overcome. This splits the market into a legal market of the rich, and an illegal ‘black market’ of the poor, each with its own survival minimum and decay rate. The ‘survival minimum’ for the legal market is equal to the price of satisfying the bureaucracy. The black market, without the legal protections and documentation for long-range property rights, is inefficient and therefore has a very steep decay. For a thorough documentation of the considerable potential wealth hidden and inaccessible in this market, see the Peruvian economist DeSoto’s book ‘The Mystery of Capital’.

Free market capitalists are strongly opposed to the inequality imposed by market splits and barriers to trade. The inequality that results from rich people creating more wealth and making themselves and indirectly society richer, we approve. We would say it was only ‘fair’ that those who earned it get to keep it, although we admit that having the opportunity is largely a matter of luck.

Unfortunately, the scale-free Pareto distribution – with its peak of poor – is not something we have the ability to do anything about yet. You would have to restructure the economy on non-fractal, non-self-organising lines: a command economy in which trillions of decisions must be made every day. Every attempt so far has ended in death and ruin.

My apologies for the length. I hope you’ll give it some thought.

Comment by Pa Annoyed

Hi, and thanks for your thoughtful post too. As you guessed, I don’t agree.

For a start, let’s consider only free market capitalism and suppose that income is really distributed according to a power law. You say that for slow decay rates the average wealth of society is higher. But why?

I think I know why you say it. If we take what seems to be the standard definition of the Pareto income scaling rule, then we get that the probability density function is $p(x)=km^kx^{-(k+1)}$ where x is income, k>1 is the scaling coefficient and m is the minimum income. This has the property that $\int_m^\infty p(x)=1$ and the expected value of income $E[X]$ is finite. The total income is proportional to the expected income, which is $E[X]=km/(k-1)$. Now what you’re saying is that for slow decay (k close to 1) the total income is larger.

The mistake is in the assumptions of the model, not the mathematics. The assumption you’re making is that m, the minimum income, is a fixed quantity. If instead we make the alternative assumption that the total wealth is fixed at value T, then we have to choose a minimum income depending on the scaling coefficient, $m(k)=T(k-1)/k$. With this alternative assumption, as k gets closer to 1 the minimum income gets closer to 0 and it’s obvious that the larger k is the better.

I see that you describe the minimum income m as the ‘survival minimum’ which suggests to me that you want to argue that whatever the minimum income necessary for survival is, there will be people at that income, and that consequently the ‘minimum income’ in the model should be set at that level. But what does that argument entail? Does it mean that some people will presumably drop below that income level and therefore die? So how do we factor that into the model? What is the argument that the minimum income will necessarily have to be at the survival level? If it could be higher than the survival level, then a consequence of decreasing the value of k to close to 1 would be to push the minimum income to the survival level point, and therefore presumably increase the number of people who died because they were below the survival level. Perhaps, indeed, this is an accurate description of free market capitalism?

You have to be very careful of arguments based on mathematical models, confusing the model with the reality it describes. For a start, it is very easy to conceal very strong hidden assumptions which can radically change the result or interpretation of the result. Also, it is easy with a mathematical model to extrapolate beyond the range of its validity. This is especially true in economics which is both theoretically and empirically a very weak field.

Moving on to empiricism, the first thing to note is that income distributions do not actually follow a power law. Take a look at the distribution of incomes in the UK. And one for the US. They don’t follow a power law. OK, you may say, this is because they are not perfectly free markets or they follow this split distribution. So in fact we have no evidence at all about what a truly free market would be like in terms of income distribution, all we have is mathematical models extrapolated beyond their range of experimental validity, and thought arguments based on dubious premises.

To finish with, I want to go back to the assumption I made right at the beginning of this comment, that we’re considering only free market economies. The argument about scale free distributions and what not, even if it were scientifically valid, says nothing about how non-capitalist alternatives compare. You dismiss these by posing the dichotomy: either have your economy not be centrally organised, or have it be centrally organised. You identify the former with free market capitalism, and the latter with command economies (which you then further identify with Soviet Russia, etc.). I think there are several mistakes here.

Firstly, the dichotomy is not straightforwardly between centrally organised and not centrally organised, it’s a matter of degree. Thoughtful free marketeers like Friedrich Hayek appreciate this point, for example “To prohibit the use of certain poisonous substances, or to require special precautions in their use, to limit working hours or to require certain sanitary arrangements, is fully compatible with the preservation of competition.” He says this applies to all “measures merely restricting the allowed methods of production, so long as these restrictions affect all potential producers equally and are not used as an indirect way of controlling prices and quantities.” (From chapter 3 of “The Road to Serfdom”.) In other words, he argues that it is right and good that there should be democratically decided state intervention in the economy (in certain cases).

Secondly, your characterisation of the two poles is problematic. Free market capitalism is not and cannot be free of the state. It requires a strong state to enforce property rights and (through policing of monopolies) ensure competition. I could go on all day about that, but suffice it to say here that if you’re right about power law distributions and the scale free nature of the economy, how can there be effective competition? Monopolies and oligopolies are inevitable in a power law distributed market. This is one of the second order effects that this simplistic power law argument fails to consider. On the other side, you assume that any attempt at a planned economy must fail as was the case in Soviet Russia. I’m not sure this is good economics or good history. Did Russian Communism fail because its planners couldn’t make enough good decisions or for political reasons (both national and global)? In an era when computers programmed by financial analysts make near instantaneous decisions controlling the world economy, is it reasonable to make the argument that central planning must fail because the number of decisions to be made is too large? There’s a lot to say about this too, but I’m not going to go down that road here because I don’t even want to argue for a centrally planned economy.

In fact, there are models of economies that are neither centrally planned nor free markets. Take a look at participatory economics. I wrote about it already on this blog. True, these alternatives have never been tried and may have fatal flaws in them. That’s a good reason not to instantly implement them in an overnight revolution. But, we can try to make moves towards better alternatives such as these, and as problems arise we can try to solve them. We must try to do this because the alternative, of going along as we are doing at the moment, would be tragic for most of the people of the world. I’ve tried to argue that point in my blog entry above and the one on capitalism you mentioned, and I would say that, if anything, the idea that free market economics necessarily leads to a scale free distribution of incomes is a strong argument of the insanity and dangerousness of capitalism.

Comment by Dan | thesamovar

Good response!

OK, first you’re right about the argument for the mean being higher with a slow decay. I wasn’t quite assuming that the minimum is constant – I had mentioned that there were two distinct ones in the split market – but it tends to be dependent on factors external to the market.

I’m glad to see that you make your assumption explicit: “If instead we make the alternative assumption that the total wealth is fixed at value T”. Yes, Marxist economics was based on this assumption, and a lot of the left wing since has followed it. It’s sometimes called the zero-sum fallacy – that what one person gains another must inevitably lose. If you assume it, then wealth creation is impossible, the average wealth of a society is constant, and wealth redistribution is your only option. However, even allowing for the fact that I’m simplifying the position for clarity and modern Marxist economists wouldn’t go that far, I don’t believe that it’s true. That’s because of observations like that we are eight times richer today than we were a century ago, and you can’t do that with redistribution. It is, if you like, the core difference of opinion between left and right economics, and I don’t propose to try to change your opinion about it.

The minimum income is based on supply and demand: and in particular, that the supply curve for the labour market has a minimum. Nobody will work for a wage they cannot survive on.

The reason the peak is always there is to do with the difference between investment and consumption. Consumption is spending to get things we want or need directly, that get used up in the process. Food, housing, clothes, etc. Investment is spending money on the means to make more money. Education/training, tools, advertising, R&D, innovation, etc. Investment creates wealth. Consumption consumes it. Consumption isn’t a bad thing – it is after all the entire point of the whole exercise – but investment drives the positive feedback that makes more consumption possible. The fundamental reason that the rich get richer is that they have more to spare for investment. The poor have only just enough to meet their needs, and it is all consumed in the process. They get trapped there.

I agree that one has to treat mathematical models cautiously, and I do. But at the same time, there is a reason we use mathematics so much in the modern world. Power law relationships have been found to be robust, in the same sort of way that the central limit theorem makes Normal distributions pop up everywhere.

Actually, I was sort of impressed by how closely both the UK and US wealth distributions fit the general form. There is a shap peak at the left that tails off. The fit is not exact, largely I suspect because both are welfare states, and the government takes away from the upper end of the distribution to push the peak of poor a little to the right – the UK very obviously more than the US. (50% tax/GDP does that.) Because the US graph goes all the way down to zero, I suspect it might not be including welfare as income. But I wouldn’t have expected an exact fit, anyway. I don’t expect error distributions to be exactly Normal, for the same sort of reason.

Incidentally, I’ve no strong objection to a welfare state so long as everybody understands what it costs. Society loses more than it gains on average, and you can only afford so much before it reduces your ability to pay it below viability. The poor are slightly less poor, but there are more of them.

You are correct that the dichotomy between economy types is not so straightforward, and I accept your point there.
The distinction I ought to be making is between the fractal structure – where groups of people can act together as a single economic agent, and groups of groups act as a larger one, and so on – and other structures generally. It’s not true that the only alternative to the fractal self-organisation is a command economy.

So far as I can see, parecon doesn’t change the fractal nature, it only changes the details of the rules between economic agents. That won’t work – it would lead to a Pareto distribution again, just with different people at the top. But I may have misunderstood it.

One final disagreement: “We must try to do this because the alternative, of going along as we are doing at the moment, would be tragic for most of the people of the world.” It depends what you mean by “going along as we are doing at the moment”. If you mean stasis, exactly as we are now, I’d agree. But if you mean continuing the economic advance of the past two hundred years – the burgeoning food supply, the advance of life expectancy, the steady rise in virtually every measure of global welfare – then I would disagree. Both the number and the proportion of the starving poor has dropped, because of Capitalism. I want to see more of that. Capitalism is not perfect, in that it can never on its own drop the number to zero, but it’s pretty good. You’d have a job to come up with something better.

Comment by Pa Annoyed

Thanks! I enjoyed writing that last comment actually. I didn’t know about this power law stuff in economics and it was fun to look it up, although I’m very dubious about it.

I didn’t want to make the assumption that the total wealth is fixed (which incidentally I wouldn’t characterise as Marxist), just wanted to pose it as an alternative to illustrate how the conclusions depend on hidden assumptions. In this case, I think it showed that your argument is not arguing a consequence of power law distributions, but rather the more well known argument that free markets and their consequent massive inequality lead to higher economic growth – the ‘trickle down’ theory of the 80s. I’m not an expert on this sort of stuff, but my impression was that the evidence was against that, that the economic growth of those years largely benefited the rich only, that the less well off saw their real incomes stay the same or possibly get slightly smaller. Regardless, this is a very different argument and I don’t think the power law stuff adds anything to it.

On the subject of mathematical models, you don’t need to convince me of their usefulness – my job is mathematical modelling (specifically, theoretical models of the brain). My experience of the whole business though is that you really cannot be too cautious about them. They are, of course, incredibly successful in the physical sciences and the proof of that is all around us. As yet though, I don’t think mathematical modelling has been really successfully applied in any social scientific context. It’s (relatively) easy to do the deductive reasoning part of mathematical modelling, but the difficult thing is the empirical part – fitting the model to reality. Mathematical models in economics are rather like those mediaeval maps with unexplored areas labelled “Here be dragons” – rather beautiful but useless. (A nice example is the Hereford Mappa Mundi which I saw this summer. Unfortunately the labels are in mediaeval French but the pictures give you the idea.)

I don’t agree that social welfare costs society. It may in its current incarnation, with its poverty traps, etc. I would favour more universal mechanisms like a citizen’s basic income scheme and un-means tested free access to all health, education, etc. With that in place, society ought to substantially benefit. Security and access to education would mean that everyone could invest in their own human capital rather than just the rich as is the case at the moment. Surely a society where everyone is able to be educated will experience faster economic growth.

Parecon couldn’t develop to a power law distribution of incomes because it’s structurally impossible for large inequalities to develop in it. There can be relatively small inequalities of income, but there is no ownership of capital and therefore these don’t magnify over time.

On your last point, although I’m not a Marxist, I have some sympathy with Marx’ idea that historically capitalism was a necessary stage to go through before moving on to something better. Capitalism has indeed created a base of enormous wealth. However, we now have sufficient wealth, sufficient education, and sufficiently developed infrastructure (travel, communications, etc.) that it is possible and desirable to find a better way.

Comment by Dan | thesamovar

I’m happy with that.

The trickle-down theory has its points, and is correct to an extent, but it both misses the important point and is far too optimistic. The idea was to try to persuade people that Capitalism achieved the aims Socialism aimed for. It doesn’t. The results it achieves transcend the assumptions of Socialism, and in some ways more effectively achieve what Socialism possibly ought to want, but are actually completely different from what it says. I’d like to avoid it, personally. The DeSoto book I mentioned earlier is excellent in this respect – it describes Capitalism from the perspective of a third worlder, and explains why it would work in principle, but what goes wrong with it in practice.

Anyway, the incomes of the poor have been increasing, albeit not as fast as one might hope. Speaking as a member of a family who was one of the less well off in that time, I can say for definite that we are all considerably better off now. But more generally, it depends how you measure it. Relative poverty is not the same as absolute.

The zero sum fallacy is at the heart of several economic systems – the Marxist one is only the most popular and the origin of any number of others. You don’t often see it so explicitly stated, so I was pleased to be able to address the question without having to explain what I thought the other person was doing first. It’s not the only assumption I’m making either, but I’m trying hard to be brief.

Bear in mind that power law distributions can give rise to equality – simply by giving a fast decay rate. The power law is the result of a fractal structure. A fractal structure that does not allow large inequalities to develop would necessarily result in everybody being poor. Personally, I’d call that undesirable. That’s my worry – that in desperately trying to achieve equality, that you’ll pick a set of rules that makes everybody poor. I think that rather than equality, your metric should be to reduce the number of poor as much as possible. I don’t care if it results in some people getting very rich, so long as nobody goes hungry any more. I don’t know, but I feel that the idea that it’s better to go short than to see someone else do so much better is the sort of thing that would be said by someone who has never really gone hungry themselves. That may be unjust. But we’ve done tremendous things over the past two or three centuries, and I think it would be a crying shame to throw that all away.

It’s been a pleasure talking to you. My best wishes.

Comment by Pa Annoyed

I agree that a system that creates equality by making everyone poor is a bad thing. If the choice were really equality and a poor life for everyone, or inequality and wealth for everyone, then sure I’d go for inequality. I don’t think that it’s really like that though. In fact, I think a free and equal society would experience faster economic growth than an unequal one.

Best wishes to you too – an interesting discussion.

Comment by Dan | thesamovar