Thinking about equity

I had a comment on one of my pieces recently:

“I'm curious, why do you think "maximum equity" should be a goal of Australian (or any other) society? I know that inequality is a popular buzzword at the moment; watch Q&A for more than a few minutes and someone is bound to mention it. But I rarely if ever hear anyone enunciate why societies should pursue this end. It is seems like an unspoken assumption that "equity" is a good in-and-of itself.

         It seems like one could make a pragmatic argument for increasing equity. For                        instance, where inequity is great there tends to be serious poverty at the lower                      extreme. Serious poverty, in a society where there is also affluence, can lead to                    increased crime. So if you want to decrease crime then increasing equity might be a              solution. But then this could also be achieved by lifting the lower extremes without                bringing down the higher extremes. 

         I don't really see anyone making this argument though. What tends to be voiced is                either a flat assertion that "inequity is bad because... inequity", or else an appeal to                emotion. "Don't you care that people are poor/disenfranchised/in need?" We might call          this the 'empathy argument'. The problem with this line is that assumes that plight of              the least well off confers a moral responsibility on everyone else. And as philosophical          grounds go this is far more shaky than I think people realise.”

I have just finished teaching a module on inequality in China which covered a lot of this material, so I thought I write up a generalised response.

To start, I must address the ‘maximum equity’ point by emphasising that maximum equity does not mean perfect equity. I made that comment in the context of trade-offs between efficiency and equity. There are good reasons for tackling inequality, as I will outline below, and so if there is no cost to doing so we should proceed with tackling it. However, there are usually trade-offs to tackling it, also discussed below, so we need to weigh the costs and benefits.

I will start with the arguments for addressing inequality, then the counterarguments, and then discuss how to operationalise this stuff in public policy.

Arguments for why equity is important can be classified into two broad categories—intrinsic and instrumental. Intrinsic arguments concern why inequality is bad in and of itself. Instrumental arguments suggest that inequality should be avoided because it leads to negative contingent outcomes, like crime, low growth or social dislocation.

In my opinion, the instrumental arguments for why inequality is bad are quite weak, especially in terms of the empirical evidence for them, whereas the intrinsic arguments are quite strong. I will consequently start with the instrumental arguments, and keep the discussion brief.

The most comprehensive analysis of the instrumental arguments against inequality is The Spirit Level by Richard Wilkinson and Kate Pickett. This book has subsequently came under intense criticism, notably by Christopher Snowden in The Spirit Level Delusion. I think it is worth noting that even staunch advocates of instrumental arguments against inequality, like Andrew Leigh, don’t think The Spirit Level is very good. Its analysis mostly consists of correlations not causal relationships, and their data appears cherry picked on occasion. You can access a lot of Snowden’s takedowns here: http://spiritleveldelusion.blogspot.com.au/ and here http://www.wsj.com/articles/SB127862421912914915.

Leaving aside the data, the theoretical arguments are at least reasonable. I’ll only cover a few big ones here:

Inequality leads to lower growth. There are typically two conduits for this effect. The first involves poor people not being able to save as much, thereby depressing investment, which reduces capital accumulation and slows growth. The second is that poor people can’t invest in mobility in the form of education, health and transport, which depresses human capital formation and thereby, productivity.

This argument recently drew some support from an IMF study. I was at a seminar last week on a paper that basically debunked the IMF study. I can’t link to the paper because it is forthcoming. Statistical evidence suggests that the effect the IMF observes is a poverty effect, not an inequality effect. If you are impoverished, you can’t save or invest in human capital, and this weakens growth. But you can be relatively poor in a rich but unequal society like the United States and still be able to save and invest.

Inequality undermines democracy. Stiglitz makes this argument at length in The Price of Inequality. He suggests that the very wealthy can co-opt the political process through lobbying and the financing of political campaigns, skewing outcomes to favour them rather than the broad mass off the public. Recent empirical research suggesting that legislation in the US resembles Oligarchic outcomes corroborates this argument. What’s less clear is whether the channel is inequality or some breakdown in the institutional structures of the United States. Would greater equality fix this? Similar analysis needs to be conducted for more equal states before we can draw a clearer conclusion.

China provides some interesting case data for this argument (albeit a more generalised form concerned with social unrest rather than democracy). Since reform and opening up under Deng, inequality has shot up in China, but poverty has also plummeted. Protests and other instances of instability and disaffection have risen proportionally to inequality, suggesting a relationship. It gets weird when you see that the Communist Party is doing a lot about it though, which suggests that inequality only foments unrest until that unrest because political salient, at which point the ruling class acts on it. In China’s case decisively, and I think we’re about to see relatively decisive action in the West as well (if we aren’t already).  

Inequality leads to crime, bad health and social dislocation. This is the main argument made in The Spirit Level and the main one that has been debunked. The data just don’t support it. The causation is deeper. I certainly think things like public education, health and transport help, but it’s not clear why.

Those are the main instrumental arguments. I don’t find them very convincing, but maybe I haven’t presented them as robustly as possible. I turn now to the intrinsic arguments, which I find thoroughly convincing. 

The first intrinsic argument comes from utilitarianism. Theory and empirical evidence suggests diminishing marginal returns to income for happiness. That means that a dollar is worth more in terms of happiness to a poor person than a rich person. Ergo, taking a dollar from a rich person and giving it to a poor person increases the total utility of the population.

The second intrinsic argument, one that even staunch defenders of inequality like Hayek acknowledged, is that life outcomes are substantially determined by the genetic lottery and one’s birth class. That is to say, a large part of the reason why someone is rich or poor has to do with luck rather than effort. It seems unjust for some to be condemned to a life of relative scarcity by accident, so there is an argument for redistribution, especially insofar as it relates to equality of opportunity (i.e. equal access to transport, jobs, education and health).  

The third intrinsic argument is similar. It says that we are all human and so have at least a close-to equal claim to collective resources. This argument is about enhancing cooperation within the species, whereas arguments in favour of inequality tend to be about encouraging individual achievement. Both are important for our survival.

The final intrinsic argument for equality is that people care about it. As my commenter said, some people take it on assumption that equity is good. Recent research in evolutionary biology suggests that this is wired into our genes to help us compete as a group against other groups. While selfish people outcompete altruistic people within groups, groups with more altruists outcompete groups with relatively more selfish people at the group level. It is thus important to have people wired for collectivist values. Inequality thus makes a large number of people unhappy—it incenses them. There is therefore a reason to act on it, even if there is no instrumental reason for doing so. I should stress that these people do not want perfect equality, but surveys of the kinds of income distributions that people would be satisfied with always show preferred distributions that widely diverge from actual distributions. The free ebook Thinking About Inequality reports a lot of this data.

That covers off the argument for taking action against inequality. What about the argument against taking action? These tend to take two forms: the first is property-rights, the second is disincentive effects.

The rights-based argument is that people are entitled to their property. Taxing them for redistribution purposes (as opposed to Pigovian taxes or the funding of public goods like defence) is thus basically a kind of theft. Ironically, this was the first rights-based argument ever advanced, long before rights became a vehicle for social justice, their language was used by aristocrats to defend their privileges.

A related and in some ways stronger argument is the ‘just desserts’ principle. This argument, which is well made by Mankiw in his paper ‘Defending the 1 per cent’, is that people are entitled to what they earn. If you work you but off to get qualified as a doctor and then work your butt off to earn a bunch of money as a doctor, the government shouldn’t come in to redistribute your earnings to someone who doesn’t go to all that effort.

In his book The Righteous Mind, psychologist Jonathan Haidt provides empirical evidence that both left and right-wing people care deeply about inequality. However, right-wing individuals have a just-desserts concept of inequality, which means that they see unemployment benefits as unfair rather than equity-enhancing. In contrast, the left emphasises relative inequality and sees unemployment benefits (among other welfare payments) as critical to equity. No wonder our politics is so fractious.

Just desserts is related to the third argument against taking action on inequality, which is the disincentive effects. If you tax someone when they work, they will be discouraged from working. In the extreme case, they will instead opt to receive transfer payments instead.

The formal articulation of this idea is the Mirrlees model in economics. Therein, everyone has a level of productivity, W, and a level of effort E. Their output is ExW. If the government taxes high productivity individuals and transfers some of their output to low-productivity individuals, it creates an incentive for high-productivity individuals to hide their productivity by exerting less effort and either being taxed less as a result or receiving a transfer payment. The result is less wealth overall. This situation, where redistribution results in less total money to go round, is referred to by economists as ‘Okun’s leaky bucket’.

Now in recent times we have found that disincentive effects are not as strong as many people think. Marginal tax rates only seem to kick in around 70%. Up until that point, the status effect of being middle class keeps people working and chasing promotions rather than settling into a cushy dole-bludging lifestyle. Tax evasion by corporations seems similarly inelastic at the top end because high-tech and advanced firms need to attract very highly qualified talent and that talent wants to live in nice, developed places like Copenhagen.

The leaky bucket also doesn’t have legs in the context of wealth taxes like capital gains and inheritance tax. There is no effort involved in making money off existing wealth, so there can’t be much of a disincentive effect to taxing it. That said, tax cuts to the rich often result in increases in total tax receipts because people take their money out of low-interest, tax free state securities and put them into more productive assets that are taxed.

How to operationalise all this? Libertarians like Mankiw would argue that tax should only be collected to pay for public goods (i.e. non-rival and non-excludable) like street lights and defence, and to offset externalities like pollution. In other cases, things should be left to the market. Implicit in this argument is the notion of market justice: in a pure market, the person who wants a good the most is willing to pay the most for it. The problem with this line of argument is that if people have unequal wealth then the person who wants a good the most might not be able to pay the most for it. There is also the issue of the genetic lottery, human rights and utility enhancement.

That said, perfect equality is not desirable either. Such a system discourages innovation and basically results in an equal share of poverty. So there must be some balance, and this balance requires value judgements.

This is what I was on about in my post that was commented on. When designing public policy, we can start by picking the low-hanging fruit, which means trying to get as much easy efficiency and equity as possible before making value judgements. An example is the Australian superannuation system as opposed to the old defined benefit scheme or the ludicrous American model. Defined benefits are too generous and inefficient while the American pension system leaves too many people without retirement benefits. The Australian model is both efficient and equitable. After you’ve picked this low-hanging fruit, then it is up to the public to determine how the rest of the cookie crumbles—more progressive publics will push more equity (e.g. in Scandinavia), more libertarian publics will push for more efficiency (e.g. in America). An example is the HECS system for funding university education in Australia. There is a subsidy for course fees in this system. How large that subsidy is gets determined by public values. But before this determination, the system already picked the low hanging equity and efficiency fruits by employing income-contingent loans and the tax system to manage repayment.

There is a lot more that I could write, but I think this is enough.