Inequality, evolution & complexity

Why has mainstream neoclassical economics traditionally had little to say about the causes and effects of inequality? This is the question raised in an interesting new paper by Brendan Markey-Towler and John Foster.

They suggest that the blindness is inherent in the very structure of the discipline. If you think of representative agents maximizing utility in a competitive environment, inequality has nowhere to come from unless you impose it ad hoc, say in the form of "skilled" and "unskilled" workers.

But there's an alternative, they say. If we think of the economy as a complex (pdf) adaptive system - as writers such as Eric Beinhocker, Cars Hommes and Brian Arthur suggest - then inequality becomes a central feature. This is partly because such evolutionary processes inherently generate winners and losers, and partly because they ditch representative agents and so introduce lumpy granularity. Markey-Towler and Foster write:

Inequality is a phenomenon that should arise naturally in a complex, networked economy as value‐generating connections (transactions, business relationships etc.) are formed, consolidated and broken. Concentrations of market power, skill differentials, luck and rent‐seeking can all be dealt with in such an analytical framework.

This is a point touched on by Ben Fine in a paper (pdf) claiming economics is unfit for purpose.He cites some econophysics research which shows a few dozen multinationals control the world's economy.

If you think that's a bit Bilderbergian, consider this new paper by Pablo Torija. He shows how, since the 1980s, western politicians have stopped maximizing the well-being of the median voter, and have instead served the richest few per cent.

If the economy is an adaptive ecosystem, it is one in which a few predators are winning at the expense of the prey.

Comments

[Stuart]

"...since the 1980s, western politicians have stopped maximizing the well-being of the median voter, and have instead served the richest few per cent."

Doesn't that also jive with the period where global inequality has fallen to its lowest ever (post-industrial revolution, anyway) level?

[Greg vP]

Stuart, surely you mean "highest", not "lowest".

Taking the global population as a single collection, as is appropriate when making global claims about a globalised world, inequality is greater now than ever.

Comparing country means is tendentious fakery.

[Stuart]

...Hah. I was assuming that the rise of China (and India) would have ineveitably reduced global inequality. Quickly searched for articles, sure they would back me up, only to find that no one seems to really know what global inequality is doing. That's what happens when you comment in ignorance, I guess. Sorry about that! At least most people agree that global poverty is falling...

[Perry Ismangil]

Exactly what I was wondering! Very informative, though a bit disheartening because managing inequality in a complex system is, um, complex.

[Blissex]

«since the 1980s, western politicians have stopped maximizing the well-being of the median voter, and have instead served the richest few per cent»

OH NO! They have been maximizing the perceived well-being of the median voter, because the median voters (especially older and female ones) think that their interests as property speculators and rentiers are aligned with those of the upper classes.

That is the very big "win" of the conservative strategy of persuading middle-income, working class people, that they poor are exploitative parasites and the rich are what middle-income, working class people will become if only they speculate on property.

It is middle-income, working class median voters who have been electing with great conviction politicians espousing policies for lower wages and welfare, and lower taxes and higher capital gains.

[Blissex]

«since the 1980s, western politicians have stopped maximizing the well-being of the median voter, and have instead served the richest few per cent»

As to this the usual numbers on tax-free capital gains for median voters:

http://www.bbc.co.uk/news/business-19288208
«In 2001, the average price of a house was £121,769 and the average salary was £16,557, according to the National Housing Federation. A decade on, the typical price of a property is 94% higher at £236,518, while average wages are up 29% to £21,330»

In those 10 years landlords of average UK houses have enjoyed (usually tax-free) capital gains of around £12,000 per year, nearly doubling their £14,000 after-tax average earnings.

Think of the political implications of such an enormous redistribution of income from the working poor upwards.

The median voters sitting on those huge tax-free capital gains probably think that politicians have done very well for median voters. They are wrong, but the game is not yet over.

[Luis Enrique]

I think ABM is very promising, and I probably agree that mainstream economists aren't sufficiently troubled by inequality, and could speculate that's because they think in terms of models in which inequality arises from differing abilities (although Lord knows what empirical basis I have for that belief) ... otoh, I don't see what's ad hoc about modelling heterogeneous agents with different skill levels. Wouldn't ABM models also incorporate some worker heterogeneity w.r.t skills? Okay, you can shut down that dimension if you want to study how inequality could arise from other sources, but mainstream econ can do that do.

My guess is if you wanted a more satisfactory mainstream model, you'd need heterogeneous agent OLG model (so unequal endowments can evolve over generations) with some "luck" determining individual outcomes, something that links human capital acquisition (broadly interpreted to include things like having the right accent) to initial wealth, perhaps assortative mating ... and you'd have quite a rich model of inequality. Do we know such models don't already exist?

[Andrew]

Excellent post Chris.

@Luis, mainstream economics doesn't explain inequality because it is incoherent at its most fundamental microeconomic level of utility maximizing rational agents. So it is nonsense.

In addition, it makes assumptions that bear no relation to the real world.

In addition it is a laughable attempt at equilibrium analysis of a complex non-linear system!

In all honesty, how can you *possibly* wonder why it doesn't adequately account for increasing inequality?

[Andrew]

Also @Luis

- By the time you have finished adding ad-hoc mechanism after ad-hoc mechanism to your ill-begotten classical economic model, what classical model assumptions do you have left, exactly?

[Luis Enrique]

Andrew

why do I care how many "classical model assumptions" I have left?

what is phrase "ad hoc" doing here? If I was to sit down and try to model inequality, I would probably want to use a mechanism whereby which the offspring of wealth parents enjoy certain advantages. If you call that "ad hoc", how would you go about modelling inequality without introducing "ad hoc" mechanisms

on what do you base your claim that mainstream econ does not "adequately account for increasing inequality". Have you surveyed all the mainstream econ accounts of increasing inequality?

[fostering agencies]

That is nice information. I got this right off. I will back again.
fostering agencies

[info]

what did u say

[Andrew]

@Luis - Because you always seem to be defending neoclassical economics, and it is, as far as I can see, nonsense. If you agree with me on this then I am mistaken and apologise for misreading you.

The phrase "ad hoc" is doing there because it would refer to any inelegant bolt-on modifications to a fundamentally inadequate class of model. Traditional neoclassical economics is such a class of models. If you are in fact operating a totally different set of models, that are in fact not complete nonsense, then I congratulate you, but you are not doing traditional mainstream economics, to which this post refers.

You could also ask Chris, Brendan Markey-Towler and John Foster whilst you're at it.