Missing Link Friday – Inequality edition

You’ve read about the floods, you’ve given to the flood relief appeal and now you need a break. So instead of talking about the distribution of water, let’s talk about the distribution of income. Thanks to Christopher Joye it’s been a hot topic over the past week.

People are getting too worked up about income inequality argues Christopher. In a piece for The Drum he writes:

I don’t think there is anything wrong at all with a rise in income inequality if one assumes that: (a) we have equality of opportunity; (b) we are committed to combating extreme poverty; and (c) we are vigilant in protecting those members of the community who are fundamentally and irreversibly disadvantaged through, say, mental or physical disabilities. In fact, I think we should be focussed on dealing with (a), (b) and (c) rather than drumming up hysterics about inequality. It turns out that Dr Leigh’s own research backs up this view.

Matt Cowgill isn’t convinced. At his blog We are all Dead he argues that inequality of opportunity is unsustainable without some limits on inequality of outcomes. He writes:

Over time, vast inequality of outcomes erodes equality of opportunity. Wealth, privilege and connections are handed down through generations. Last generation’s meritocrats … become this generation’s entrenched, quasi-aristocratic elite, able to secure their children’s place in the hierarchy by paying for them to attend expensive schools, or by buying them houses or providing start up capital for entrepreneurial ventures.

As Ilya at Beats and Pieces writes, Christopher responded by posting "an uncharacteristically angry-sounding response to Matt Cowgill’s analysis". In this response Christopher insisted that "if you have talent combined with patience and persistence, there are few real barriers to progress in contemporary Australia (again, there are clearly exceptions found amongst various minorities)."

Other bloggers soon joined the debate. Both Ilya at Beats and Pieces and Alister Air link to Paul Krugman’s recent New York Times column on economics and morality. Krugman writes:

… when you hear conservatives talk about how our goal should be equality of opportunity, not equality of outcomes, your first response should be that if they really believe in equality of opportunity, they must be in favor of radical changes in American society. For our society does not, in fact, produce anything like equal opportunity (in part because it produces such unequal outcomes).

According to Ilya, inequality of opportunity is the product of policy choices. In the US "the choices made over the last thirty years with regard to tax, health care, and education policies have produced rising inequality and, in parallel, eroding social mobility". And presumably that’s why Krugman’s post is relevant to the Australian situation. Our policy makers also have choices to make.

Alister questions whether equality of opportunity really exists in Australia. While he concedes that differences in socio-economic status among university entrants might not make much difference to their outcomes, he suggests that differences in family background make a big difference earlier in life. For example, do children in families where nobody has worked for three generations really have the same opportunity to get into university as those from more advantaged families? 1

Elsewhere …

Christopher’s piece at the Drum was inspired by two reviews of Andrew Leigh’s new book Disconnected. Tone at Tone’s Random Thinks writes: "it seems Christopher Joye hasn’t actually read the book by Andrew Leigh, just two reviews of said book. I suppose next time I could just read the comments on one of Mr Joye’s articles, and then proceed to frame my own response – fair enough?"

Inequality has been a hot topic in the US media and blogosphere. Ilya has links.

Catherine Rampell at the New York Times Economix blog explains Why So Many Rich People Don’t Feel Very Rich. The reason is, "there is much greater inequality at the very top of the income scale than at the bottom or in the middle." When people in the middle look at those who are richer than they are, the gap isn’t all that large. But when the rich look at the incomes of the super-rich, the gap is huge. Daily Intel at the New York Magazine looks at how this applies to Brad Pitt (an actor worth an estimated $150 million):

… compared to Mayor Bloomberg, who is worth $18 billion at last count, Brad Pitt is pathetically poor. Poor poor poor. If he felt like it, Mayor Bloomberg could probably come up with an offer that would make Brad Pitt actually consider working as his personal butler for a year. How do you think that makes Brad Pitt feel? Not rich, that’s how.

Ezra Klein joins the debate, Brad DeLong redraws Rampell’s graph, and Paul Krugman argues that the widening gap between the rich and super-rich has produced "a society of winners as whiners, where people who are not only doing fine but doing much better relative to the median than they were a generation ago nonetheless feel left behind."

Note: I’ve just updated this post to include a few more links about Rampell’s post as well as correcting her name (I accidentally called her Charlotte Rampell).

  1. update: see Alister’s comment below[]
Subscribe
Notify of
guest

30 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Alister
13 years ago

While he concedes that differences in socio-economic status among university entrants might not make much difference to their outcomes…

What I’m noting is that while two students with the same ENTER but different SES backgrounds appear to do as well as each other, ENTER correlates to SES in ways which suggest meaningful equality of opportunity doesn’t exist. Christopher Joye’s argument is that income inequality isn’t important if we have (among other things) equality of opportunity. I’m going to do some digging to see what more data I can find to support the argument, and go from there.

conrad
conrad
13 years ago

“if you have talent combined with patience and persistence, there are few real barriers to progress in contemporary Australia (again, there are clearly exceptions found amongst various minorities).”

You forgot to mention the two anecdotes about his friend’s success, which must prove the claim — Why people dredge up qualitative anecdotes at the extremes to try and prove quantitative claims when there are happily quantifiable measures that could be used really beats me.

derrida derider
derrida derider
13 years ago

Conrad, it’s called the salience effect by the behavioural economists and psychologists. We tend to believe events we know first hand to be much more representative of the universe of events than they are actually likely to be.

Either that or they’re just ignorant of the quantifiable measures and prefer not to relieve that ignorance precisely because it might upset comfortable assumptions.

Social scientists have of course studied the question of the effects of inequality at great length. Yes, high inequality of outcomes correlates very strongly with lack of social mobility – just as you’d expect. After all, what’s the point of being rich if you can’t give your own kids a head start over the hoi polloi? But there’s another reason to be concerned about gross inequality of outcomes – it makes democracy difficult. We can all see how the US now has the best government money can buy, f’rinstance. And not coincidentally it has about the lowest intergenerational mobility of incomes in the OECD.

Andrew Norton
13 years ago

If you are bright and hard-working and do well at school your prospects are very good regardless of your family background. But Alister is clearly right that family background makes a big difference to whether you will satisfy those conditions.

We’ve probably gone backwards for the lowest 10% by SES, due to the increase in seriously dysfunctional families.

But functional low SES families do have social mobility, such as increased rates of uni attendance.

Nicholas Gruen
Admin
13 years ago

Don, you were making light and there are floods. You were not making light of the floods and were being overly influenced by political correctness methinks.

Conrad, there is not good quantitative data to answer Chris Joye’s claim that if you’re talented, patient and persistent (I think the middle term is largely redundant) you’ll get on. You might be able to get data that disproves the claim. But how do you measure talent and persistence? It’s not easy. I’ve heard lots of stories about people who you could argue were talented and persistent and didn’t get on – but then in a lot of cases I’d argue that they weren’t what I’d call talented and persistent.

Stories are concrete, data is always categorised by someone else before it becomes data. If the categorisation is ‘black hair’ or ‘blond hair’, no problem. But ideas like ‘persistent’ and ‘talented’ are much harder to define indisputably, or to define for one purpose in such a way that the definition is useful for other purposes.

trackback

[…] Club Troppo provides a meta summary of the debate started by Christopher Joye, and continued by Matt Cowgill's post and my comments […]

Andrew Norton
13 years ago

I’m a bit rusty on the research, but I think we can can measure both talent and persistence in the way Joye means. Not talent in the sense of exceptional ability, but any of the standard academic tests. And persistence and self-control can be measured many ways – the famous Stanford marshmallow test being one of the more memorable methods. These indicators do have quite good predictive power. That’s why there is such a fuss about early childhood.

Fred Argy
13 years ago

Don, yes, inequality is a big topic at present, with both sides at it. You missed two recent exchanges.

First, there is an interesting debate being held between Greg Mankiw, Paul Krugman and Eric Schoenberg (see http://economistsview.typepad.com/economistsview/2011/01/zombie-economics-and-just-deserts.html) on how people should be rewarded for their efforts.

Mankiw asserts that based on “just deserts theory”, under which people should get what they deserve, people would be paid the full value of their marginal product. And, he says, tax theory should be based on that belief. Krugman queries this – partly saying that “economics is not a morality play”.

One could raise a semi-moralist argument against Mankiw about the growing inequality of incomes in the USA, but let me avoid this horny issue. I would prefer to use three main arguments against the Mankiw case. First, it fails to allow for distortions in the distribution of income (e.g. market failure and political power). Secondly, it assumes a high degree of substantive equality of opportunity (the US is way down the world on most education and health standards). Thirdly, Schoenberg is right to suggest the introduction of an inheritance tax, where inheritors have contributed nothing themselves and should pay higher rates of tax than others who contribute more to the creation of wealth.

Mankiw also raises the “hereditary” issue as explaining why much of the recent biological differences in education outcomes give a dominant view to nature over nurture (http://gregmankiw.blogspot.com/2011/01/half-full-glass-of-economic-mobility.html).

There are of course conflicting views on this last issue, as I have argued in my piece on Equality of Opportunity in Australia. Inequality must play a role in education outcomes (as Mankiw concedes) and there are many other factors bearing on equality of opportunity (such as health and housing). The fact remains that social mobility is weaker in the USA than in most other developed countries such as Australia and Europe.

Labor Outsider
Labor Outsider
13 years ago

A lot of James Heckman’s research (and related stuff) on early childhoood development and later life outcomes is relevent to these questions. My reading of that, and the general social mobility literature, is that, at least as far as you can measure it, individuals’ material circumstances (in relative terms) when young have a significant effect on later outcomes in most countries, though the magnitude of that effect varies depending on various institutional (and cultural) factors. It probably isn’t a coincidence that US has both high inequality and low social mobility. But then, Australia has somewhat high inequality and average social mobility. Gini coefficients aren’t everything.

Moreover, persistence itself (however you want to measure it) is itself endogenous to those material circumstances. As Andrew says, that is one of the forces behind the whole early intervention literature. Of course, we still don’t have a great grasp of exactly which type of interventions work best (the best results for early intervention have been found for small programs and don’t necessarily hold when scaled up). And then one has to deal with all the identification problems that plague this literature.

Philosophically, I (and most progressives I know) like the idea of positive (as opposed to negative conceptions) of liberty. The sense that inequality (especially in initial conditions) constrains individuals ability to exercise genuine choices in their lives. Freedom to, not just freedom from. As such, one thing I have always hated is the claim that just because some people overcome difficult circumstances to make the most of their talent, that all people facing those circumstances can and should be able to do the same thing.

Of course, translating those sentiments into policy isn’t a straightforward thing. And that is where this debate often breaks down – it is conducted in vague terms, or over marginal policy changes. People express a view about inequality generally, without being specific about how much inequality is too much, and which types of inequality matter most to the things they care about, or the institutional changes that are necessary to alter things in the direction they want, or even how those ends can be achieved politically.

Indeed, that was the biggest problem I had with Matt’s response to Chris. Philosophically I sympathised but in practical terms it left me cold.

Matt C
13 years ago

Don,
Thanks for the link.

LO,
My post was intended to make the case in a fairly abstract way as to why we should care about inequality at all. It was not intended to be an all-encompassing manifesto of desired public policy changes that would bring about my desired state of the world.

conrad
conrad
13 years ago

“Conrad, there is not good quantitative data to answer Chris Joye’s claim that if you’re talented, patient and persistent (I think the middle term is largely redundant) you’ll get on. You might be able to get data that disproves the claim. But how do you measure talent and persistence? It’s not easy.”

I don’t see why it can’t be answered. You could do this very easily superficially if you were interested in particular talents. Say, pick a group of kids who are all precociously good at playing the violin, and then see the probability that they end up as masters in a decade and the extent that SES affects that. I know where money lies given the cost of music lessons.

At a less superficial level, you could certainly look at relatively stable personality characteristics such as conscientiousness and some measure of persistence at tasks (not being a personality guy, I can’t name the measure off the top of my head, but I’m sure people must have looked at this), and then look at some more general measures of talent (e.g., mathematics ability). I don’t think any of those would be especially hard to measure in a reasonably valid way (indeed, there are already very good conscientiousness and mathematics ability measures). Based on the claim that anyone could succeed given these characteristics, you could simply select young kids high on conscientiousness, persistence and mathematics ability (say the top 5%), and then see if they end up in decent professions that use the great mathematics ability that they had when they were five. According to the claim, there should be little correlation with SES, since those low on SES should succeed anyway, and that should be true even with some error in the scales, such as if the persistence measure is good but not great. After being affiliated with some people that went to Syndal Tech in my younger days (where about 2 people went to university every year from Year 12), I think I’d probably be willing to bet my house there would be a significant correlation with SES.

Off the top of my head, it might well be the case that this data already exists, or at least data good enough to test the claim — I know there is a huge population database with many variables from Sweden(?) that might have some of these variables, although I’d be far more interested in the results from countries with more SES based school segregation like Australia.

Labor Outsider
Labor Outsider
13 years ago

Conrad, you are going to run into pretty big identification problems with the analysis you have in mind, not least of which is coming up with a credible way of determining whether any correlation you find between SES and later achievement is due to SES or other factors that you are going to have difficulty controlling for that are correlated with both SES and kids’ achievement. I certainly can’t think of an Australian database rich enough to test those claims properly.

Andrew Norton
13 years ago

There are a couple of British birth cohort studies (one from 1958, one from 1970) which have tracked the people born in a particular week through to the current time. They include class, housing conditions, school, parent’s aspirations for child, interest in child’s education, child’s ambition, qualifications achieved, and academic tests at age 7, 11 and 16. The child’s academic ability at age 7 and 11 are by far the most predictive of occupational status at age 33. However father’s class was also important, though through indirect affects on other variables (a quick summary of some work by Peter Saunders).

Andrew Norton
13 years ago

A point Labor Outsider makes at point 10 is important. Debates over income inequality are conducted at too high a level to be very useful, and send controversies off down very long and circuitous routes to the micro-level issues that influence whether a person’s life will go well or not.

For example despite the efforts of Wilkinson etc I find it very difficult to believe that the mere presence of very rich people in society is seriously harmful to psychologically normal other persons. I find it very plausible that severe material deprivation, chaotic and incompetent parenting, bad schools, poor healthcare etc are seriously harmful.

Generally speaking it makes sense to go for the least contentious arguments for the highest-impact policies in your broad political agenda.

conrad
conrad
13 years ago

“not least of which is coming up with a credible way of determining whether any correlation you find between SES and later achievement is due to SES or other factors that you are going to have difficulty controlling for that are correlated with both SES and kids’ achievement.”

You don’t need to control for these to answer the simple question of whether SES makes a difference — since according to the claim it doesn’t as long as you have the right characteristics. Alternatively, if you want to know why, which is clearly the more interesting question, you do — but that’s a brutally hard question — even trying to work out which way causation goes is exceptionally hard for many possible models you might like to consider.

I’m sure we’ve discussed similar things before, but it’s also worthwhile noting that if you pulled out people from a high and a low SES group that were matched on the factors suggested, then it should in fact be easier to find a result that would suggest that low SES people achieve as well as high SES people than if there were no confounds (under the assumption that this is the true result). This is because most variables related to SES make it harder for people of low versus high SES to succeed (i.e, SES correlates negatively with some of the factors mentioned, like mathematics ability). Thus if you got a group of low and a group of high SES kids that were initially matched on some of the variables, you would in fact have a much more exceptional group of low SES kids than high SES ones because you would be selecting low SES kids who had already been good enough to make up for the difference with the high SES group. Thus, if you found out that the low SES kids didn’t do as well as the high SES kids later on, that would suggest that the disadvantages not only outweighed the advantages in a truly matched sample, but also in addition outweighed any advantages that the low SES kids would have got from the non-random selection.

Richard Tsukamasa Green

I’m actually a bit concerned about Joye’s comments on ideology. Whilst he is right to dismiss as foolish the tendency to last oneself to a mast with that label and allow it to then influence one’s thinking, he neglects the essential role that ideology has in a descriptive sense. Anyone’s thoughts, no matter how independent they are, are based on certain assumptions and thoughts, and these labels are useful as a rough way of identifying what these assumptions are. Quite simply, there is nothing in human thought that comes from pure logic, and this is even more the case when it comes to human affairs and morality. Matt tried to determine what assumptions Joye was working from (and how they differed from his own), and Joye evidently thinks that Matt got it wrong. Given Joye never attempted to say what he might be working from, I think Matt can hardly be blamed, and I’m afraid Joye may not have explained because he’s never bothered to think about it, under the impression that he was truly ideologically free. But no-one ever is, the best we can be is aware of our own assumptions, to which we can then affix a rough label for shorthand description.

We also may need to establish that we come from different ideas of what inequality of opportunity is. As someone who was born into a privileged family, went to a privileged school and is acquainted with people in the corporate and higher income levels of society, it’s natural that Joye thinks of it as the ability of the public/catholic school majority to make it into these circles, which they evidently can, although there may be a lower probability of doing so.

On the other hand, whilst I was born into a comfortable middle class family, I was in a primary school environment where fellow students might be from three generations of unemployment or low skilled labour. From my perspective, I’m less concerned about whether they can get into corporate circles than whether they can merely approach the median income, or even get a job.

L Outsider does mention that “perseverance” as a virtue that is rewarded, is endogenous. Just like we only work in the expectation of pay or invest with the expectation of profit, perseverance is only manifested when there is an expectation of return. This expectation is based on the bounded rationality of experience and traits learned from family and peers. This means that historic inequality of opportunity can self perpetuate even if the historic reasons are gone. If a child has seen that work has not been rewarded in previous generations, there is no reasons from them to expend this effort. It’s not laziness, but a rational action based on limited information.
Then of course there’s the large literature on how stereotypes become internalized – test results can be affected by reminding people of their ethnicity, gender or class (or caste in this case). That last case also includes an experiment that controled for whether subjects thought there would be scope for discretion on the part of the tester – namely authority. If you’re in Redfern or Macquarie Fields, and your background has mainly experienced authority in the form of the trained skepticism of Centrelink officers or police harassment, you’re unlikely to see any authority as a source of potential support, and are ths unlikely to seek and receive it the way middle class kids routinely receive help from teachers etc. Even though class boundaries are weaker in Australia than most places, they nonetheless exist and can be internalised.
And again, even if the success of the prior generation is entirely arbitrary, it can lead to a greater expectation of return to effort, and thus promote it, such as was the celebrated case of Argentinian squatters who randomly received title to their lands.
These are just a few other ways that inequality of outcome begets inequality of opportunity in ways apart from direct income.

Matt C
13 years ago

I find the ‘ideology’ thing extremely frustrating. I use the term to indicate a general bundle of preferences and biases. I do not intend for it to mean a rigid set of beliefs regarding means and ends, such that a person who professes some sort of ideological adherence is necessarily blind to evidence. I see ideology as relevant to this debate, as I think your judgement about how much inequality to tolerate is a moral and philosophical, as opposed to empirical, question. The notion that any of us is, or can be, a bias-free automaton who merely follows the data is absurd. Ultimately you need to choose some measure of welfare to use as a yardstick when assessing policy: pareto optimality, Kaldor Hicks, the difference principle, whatever. Any of those choices is defensible, but it is a choice that is grounded in moral intuition and ideology.

Andrew Norton
13 years ago

Matt – I agree with this, though I think there is useful distinction to be drawn between ‘methodological’ and ‘normative’ ideology. I think ideologies tend to include both empirical assumptions and normative judgments. The most fruitful debates are likely to be about the former, as the latter tend to be grounded in socialisation, life experience and personality type, producing beliefs that are unlikely to be altered much by argument alone.

chris joye
13 years ago

A few quick comments, without provoking too much angst:

*Don: I like Converse’s description: my beliefs are characterised by a lack of conditioning constraint (eg, many of the policy proposals and/or positions I have outlined–eg, opposition to the NBN (methodology) but support for government intervention in markets to provide pricing transparency when these markets fail, as they do, due, for example, to problems associated with information asymmetries and raw behavioural biases (ie, government’s need to support liquidity’s public good characteristics)–are endorsed and opposed by virtually all respected folks I know that put themselves into the centre-left and centre-right camps.

My analysis and opinions are governed by logic, not ideology. And my logic is not based on a belief system that conforms to a set of assumptions that can be associated with specific political philosophies, or derivatives therein. What am I missing here?

When I confront a question I seek the best possible answer. At times those solutions seem to fall into one ideological domain. In other instances, they do not. For example, I do not believe in rational expectations or efficient markets (I published a dissertation disproving conventional forms of the latter in 1999). But, on the other hand, it is not clear to me that aggressively progressive income tax regimes are the optimal redistribution solution, all things considered (I don’t have a strong view on this, fwiw). The first statement puts me well and truly in the activist social-democratic tee-pee. The second obviously does not. Neither of these statements need be logically inconsistent with one another. So where is my ideology?

*Everybody else: I am very happy to say that I come to the inequality and social capital debates without having thought much at all about either subject. Almost every commercial and academic insight I have had in my life was yielded by looking at problems from a fresh vantage. I found in my academic work in particular that not having been exposed to a subject area, and been conditioned by prevailing schools of thought, was an enormous advantage.

From where I stand today, I see inconsistencies and logical flaws in the inequality debate. Having read Matt and Ilya’s work, and a book chapter by Andrew Leigh, I am even more confident of that position today. These debates echo the same vulnerabilities I have seen in so many other areas of academic and commercial life: fuzzy thinking.

So I don’t mean to be blunt. And I absolutely reserve the right to say that I have been proven wrong–indeed, I eagerly await that moment of new learning. But it has not arrived yet, perhaps because I have not spent a great deal of time pursuing it.

chris joye
13 years ago

As another example, add this advocacy of the ABC on the basis that news information is a public good into the social-democratic camp:

http://www.abc.net.au/unleashed/29930.html

chris joye
13 years ago

Don,

*I find talk about ‘ideology’ unimpressive. Your comments above have not influenced my position in this regard. All I get from your post is that different people have different understandings about what ideology means (and futher what specific ideologies represent), which is actually what I said in my original article. Ideology is a simplifying framework for understanding the way the world works. If ideology broadly corresponds to a model of human behaviour, its principal value is using it to reject the null that it has any durable explanatory power across behaviours.

A great example here is behavioural economics. Critics of behavioural economics correctly argue that it offers no alternative ex ante model or theory of human behaviour. But behavioural economics in nonetheless extremely informative as an ex post empirical interpretation of human outcomes.

My simple view of the world is that behaviour is too complex to be usefully explained by crude ideology and limited, axiomatic laws. As I said above, when I face a problem, I search out the empirical truth. In my experience, the truth do not fit comfortably with any given ideology.

*On your second point about coming to problems without conditioning knowledge. It is funny that you raise housing. Many of my earliest and most important housing insights–eg, regarding the value of resolving the indivisibility of the housing asset from an investment perspective–came to me before I had any knowledge of housing markets. I think it was a tremendous advantage that (a) I had never owned a home, (b) I was not a housing economist, and (c) I could think about problems in the housing market without having been contaminated by the extant literature.

I am not arguing here that the extant literature, and the existing base of knowledge, is not very valuable. But I think absorbing that existing pool of learning inevitably ends up impeding your ability to think in a truly unencumbered way.

This is not an overly novel position to hold. Many academics and mathematicians, for instance, have similar views.

chris joye
13 years ago

Oh and on searching out the “empirical truth”. I guess what I am saying is a bit more nuanced. Given a problem, I try to apply unconditional, non-ideological analysis or logic to get to the highest probability answer. But, at the end of the day, what happens in practice–that is the empirical truth–has the final word, and I will defer to that if it conflicts with my ex ante logic.

chris joye
13 years ago

Oh and on searching out the “empirical truth”. I guess what I am saying is a bit more nuanced. Given a problem, I try to apply unconditional, non-ideological analysis or logic to get to the highest probability answer. But, at the end of the day, what happens in practice–that is the empirical truth–has the final word, and I will defer to that if it conflicts with my ex ante logic.

chris joye
13 years ago

Oh and on searching out the “empirical truth”. I guess what I am saying is a bit more nuanced. Given a problem, I try to apply unconditional, non-ideological analysis or logic to get to the highest probability answer. But, at the end of the day, what happens in practice–that is the empirical truth–has the final word, and I will defer to that if it conflicts with my ex ante logic.

derrida derider
derrida derider
13 years ago

Chris, with all respect, you really need to read some of the more sensible post-modernists (not to mention some behavioural psychology), and apply it to yourself as well as others.

Your stance of “nnon-ideology” is a fine ambition to aspire to – but for God’s sake don’t think for a moment it’s achievable. And especially don’t kid yourself that you come to consider a new question with a blank slate.

It’s a bit like one of those verbs you conjugate, isn’t it? “I am rational, you are an ideologue, and he is an ignorant bigot”.

trackback

[…] far in Inequalityfest 2011 we’ve focused largely on moral and ethical issues, as well as on the distinction (if one can […]