This post is part of my cycle of posts on inequality. In the first part I dispelled some of the most eggregious myths regarding the trends in inequality and poverty, now I turn to the problem of aggregation and the misunderstandings it causes in the debate.
Part II: Stop thinking of inequality as a homogeneous aggregate – its contributing factors are radically heterogeneous
The biggest misunderstanding in the inequality debate is the presumption that it can be reduced to one thing. That this complex social phenomenon, which we measure through a variety of aggregate indicators, is just one thing. In Part I some readers may have thought I was unnecessarily confusing the issue by discussing not just income or wealth inequality, but inequality of health, education, and opportunity as well. The reason I started this way is that inequality is very much a multi-faceted issue. To those inequalities already discussed, we could still add many more, such as race, gender, and consumption inequality, for example.
More importantly, all of these inequalities are themselves products of numerous factors. And this holds true even when we narrow our focus to just one type of inequality. I will demonstrate this now by focusing on the most discussed type of inequality: income inequality.
Those most worried about income inequality point out its supposed negative correlation with growth[i] and social trust, on the one hand, and positive correlation with crime rates[ii] on the other. Those least worried, point out that inequality is natural, just as differences in talent or diligence are. Moreover, the inequality of rewards for the work we do has a social function. It rewards those activities most valued by society most, and those least valued least. It is the price mechanism at work, in other words, communicating the supply, demand, and relative scarcities of each type of work.
Both camps will usually add a normative conclusion – the first camp believes inequality is “bad”. The second believes it is “good”, or alternatively that it is just a fact of life, neither good nor bad. Either way, the second camp is likely to tell us that inequality is a non-issue. The policy recommendations will be equally polarizing: the first camp will see anti-inequality policies of all sorts as a priority, the second will argue to do nothing.
One of the camps must be mistaken, right? And as a proponent of free-markets, I will surely say the one mistaken is the first camp? Well, not necessarily. They are both right, and they are both wrong. You get the idea where the title of this cycle of posts comes from now.
Income inequality has numerous heterogeneous contributing factors. It is not just one thing. The aggregation conceals what is actually going on and leaves debaters from both camps arguing past each other. Both are raising some valid points, but inequality is a heterogeneous phenomenon that should not be labeled “good” or “bad” as a whole. When we look at the individual contributing factors to inequality, it becomes apparent the issue is far more complex, and members of both camps will quickly admit that, in fact, there are some drivers of inequality, which we should see favorably, but also some drivers, which both camps will tend to agree, are indeed undesirable.
The normative discussion about which of the mentioned factors is indeed “good” or “bad”, and why, is beyond the scope of my writing here. I assume that a great majority of people would agree with how I have grouped the eight factors listed, but that judgement is not the point of this post – even if you were to disagree with the grouping you will appreciate that the factors listed are very diverse, and crucially: we may want to minimize/maximize some, but not others.
Examples of “good” contributing factors to inequality:
- Wise management, hard work, talent
- Technological progress, innovation
- Emancipation of women
- Demographic trends
The first of these is rather self-explanatory. There exists an inequality of ability, competence and working habits. If you had a magic wand you could one day wave to equalize the world’s incomes and wealth, the incomes (among free people at least) would very soon be unequal again. Because what people (can) do with what they get is unequal. Furthermore, that is actually a good thing! It enables division of labor and specialization, necessary ingredients for the rise in humanity’s living standards.
Technological progress and innovation (not just in technology, but also logistics, for example) will also be welcomed by most people. Without them, the favorable trends I called to your attention in Part I would be impossible. Yet, investment in innovation is often a high risk / high reward type of affair, and the most amazing of innovations make those who bring them to the market very rich indeed. Consequently, high rates of innovation are a contributing factor to inequality, and places that have higher rates of innovation also on average have higher levels of inequality. See for example (Aghion et. al. 2015) for more on this.
That emancipation of women actually contributes to more income inequality, rather than less, is probably the most surprising thing I have learned while researching the issue. You would expect that the rise of gender equality would be an equalizing force, yet this empowerment of women has had an influence on marital and childbearing patterns. More educated and more well off women tend to marry partners who are alike. On average, they also start having children later and end up having fewer in total than poorer women, which can be explained by the trade-offs women face: the higher their potential income, the higher the opportunity cost of having a child (or at the very least of pregnancy).
Hence, a pattern of richer women marrying richer men, creating richer families who have fewer children to divide their wealth among is created. This is an important contributing factor to inequality in the developed world, yet most of us probably wouldn’t lament women’s emancipation. The literature on the issue remains sparse, but (Wolf 2013) and (Greenwood et. al. 2015) are important references to look at for more detail.
Rising life expectancies, which I mentioned already in Part I are another factor most of us will obviously celebrate. Yet, it contributes to income inequality. This is rarely mentioned in the debates over inequality, but once you seriously think about it it’s actually a rather obvious point. People living longer means people on average have more time to accumulate wealth and capital. The gap between those who have had 80 years to accumulate and those who have just started to accumulate, is on average greater than between those who have had 60 years and those who had just started. Obviously then, all other things being equal, rising life expectancies mean greater inquality.
An additional point, which I would not count among the “good” factors anymore, ought to be made about the redistributive nature of pension systems in practically all developed parts of the world. They redistribute from the young to the old. That is from people who have had less time to accumulate to people who have had more. This reinforces the effect of rising life expectancies.
Examples of “bad” contributing factors to inequality:
- Deficient Rule of Law and poor protection of property rights
- Regulation / entry barriers
- Subsidies, bailouts
- Monetary system
These are the reason why I believe the camp saying inequality is not at all a problem is mistaken as well. Free-marketeers who argue as if inequality is never a problem and we might even want to strive to maximize it, overlook these sources of inequality. Surely, inequality, which is a product of monopoly rents enabled by corruption is not something to be celebrated? It is neither free marker nor meritocratic.
Indeed, if we look at which countries in the world have the most extreme levels of income/wealth inequality (South Africa, Haiti, and among OECD countries Mexico for example) it becomes clear that there must be something undesirable which is connected to extreme level of inequality too.
One of the key factors keeping people in underdeveloped countries in poverty is poor protection of property rights. They might accumulate and save in a limited capacity, but without secure property rights and legal title, they cannot leverage their assets and are generally quite credit restrained. They lack key institutions for wealth creation in a commercial society (De Soto 2003). In such societies, those already better off are able to smooth their way through connections or bribery – but not everyone has that option, not everyone “has a go” at becoming wealthy.
Those already in a privileged position may then further entrench their privilege through extractive institutions in the form of regulations and other entry barriers which protect their individual interest at the expense of the majority population (Acemoglu and Robinson 2012).
Indeed, the empirical literature links both more regulation with more corruption (Holcombe 2015), and higher entry-barriers with more income inequality (Mclaughlin 2016). Many policies that are promoted as protecting the consumer or the worker thus end up contributing to inequality instead of primarily helping the poor as people often believe.
The rent-seeking and lobbying connected to subsidies and bailouts probably requires little explanation. The typical Olsonian story of concentrated benefits for the small organized groups of rent-seeking profiteers and dispersed costs of the disorganized large groups of regular citizens who are being exploited. Another clear source of inequality which most would ideally wish to eliminate altogether.
Relevantly thought, government expenditure as a share of GDP has risen by about a factor of four in most developed countries since the first world war. Thus, there are more and more of these rents to be captured through lobbying.
And finally, as an Austrian economist it would be remiss of me not to mention the negative role of inflation in the story of inequality. Under an inflationary system, those who can borrow credit and are most savvy in investing in the variety of financial instruments will do very well for themselves, while those who merely earn a wage or save in a bank account will see the real value of their assets fall. Such a system greatly benefits the haves at the expense of the have-nots (Bagus and Marquart 2016).
To begin talking to each other instead of past each other we must disaggregate the notion of inequality
Discussion of inequality as an aggregate obscures its sources. Consequently, more often than not measures attempting to address inequality will negatively impact factors that are broadly considered good, while they will do little or nothing to address factors which are clearly negative. Indeed, if we look back at the eight factors I have listed, and consider the usual policy recommendations offered by those trying to fight inequality, we must notice that those recommendations are very seldom addressing any of the negative factors. The most commonly suggested policy, higher taxes, is clearly likier to disincentive work, investment, and innovation then it is to sort out corruption, rent-seeking, or the monetary system.
Disaggregation forces both camps to admit that there are “good” and “bad” drivers of inequality. Hence, inequality is or isn’t a problem, depending on what it is primarily caused by. This prompts a very different sort debate than the one we have been seeing so far. I will return to the implications of the debate, restructured in this manner, in the final part of the cycle.
[i] The claims of both OECD and IMF studies that income inequality hinders growth and that greater equality should be pursued to promote more growth, have been challenged (see Fuest 2017 for example) for their inconsistency and failure to show causality. The treatment of inequality as a sweeping aggregate instead of isolating individual contributing factors makes it quite impossible to convincingly show a relationship between growth and inequality in one direction or another. This, however, is exactly the point of my article: disaggregation is needed to pursue productive research in this field.
[ii] Studies purporting to show a causal relationship between inequality and crime have been criticized on similar grounds as the ones related to growth. (Kang 2016) attributes most of the claimed correlation to inadequate aggregation of datasets and their subsuquent misinterpretation.