The Causes of Wealth Inequality (31): Automation and the Hollowing Out of the Labor Market

Conventional wisdom has it that automation comes at the expense of low-skilled jobs and aggravates income inequality because of labor displacement at the bottom of the income distribution. It turns out that this is a bit too conventional, and not only because it runs afoul of the lump of labor fallacy (machines need to be built and people can go on and do other things). Mid-level jobs are also hit by automation, and perhaps even more than jobs at the bottom of the skill continuum. This has been called the “hollowing out” of the labor market. This hollowing out, caused in part by automation, in turns causes an increase in income inequality. This is mere arithmetic: if the middle drops, then the extremes become relatively more important and inequality rises. Ryan Avent puts it well:

Work published in 2006 by David Autor, Lawrence Katz, and Melissa Kearney argued that employment and wage growth in America have “polarised” in recent decades, a conclusion that has been reinforced by subsequent research. Employment in high- and low-skill positions has risen substantially relative to middle-skill jobs. The resulting employment distribution generates a distribution of wages that is similarly polarised and more unequal than that which prevailed prior to this period. (source)

Why does technological automation focus mainly on middle skill levels?

Daron Acemoglu and Mr Autor pioneered a “task approach” to labour markets. Tasks can be completed by either labour or capital. The more routine a task is, the more susceptible it is to automation. But whether or not a task is automated depends upon the relative supply—and the real wage—of workers of various skill levels. Subsequent work has shown that automation and trade are responsible for displacement of routine tasks previously done by middle-skill workers, in both manufacturing and clerical or service activities, leading to polarisation of local and national labour markets.  (source)

Technological automation focuses mainly on middle skill levels because it’s relatively easy at that level, easier sometimes than at the extremes of high and low skilled tasks. “Easier” here means both technologically easier and more cost effective. Highly skilled tasks, such as teaching a philosophy course, are difficult for machines to do because they are complex (although we do sometimes see high-skilled jobs being automated, such as legal research for example). In the case of low-skilled tasks, some of these are surprisingly hard to automate, as in the case of truck driving or toilet cleaning. Even low-skilled jobs that aren’t technically hard to automate aren’t always automated because the pay-off may be too low – people doing those jobs are poorly paid so developing expensive machines to do it for them isn’t worth the trouble.

And then there’s the added worry that displacement of many low-skilled workers would create a permanent underclass unable to participate in the economy – unable, in other words, to buy the goods and services produced by machines. There’s a famous anecdote about Henry Ford mocking a labor union president in one of his factories, saying it wouldn’t be easy to get the robots to pay their union dues. To which the union president responded that Ford wasn’t going to get the robots to buy his cars.

The hollowing out of the labor market, driven by mid-level automation, has therefore a direct effect on income inequality, but it also a few indirect effects. For example, automation means lower production costs, and the savings or the added value go primarily to shareholders through capital gains and stock appreciation. Since stock ownership and capital income are concentrated among those already better off, income inequality is further increased.

If technology decreases the relative importance of human labor in a particular production process, the owners of capital equipment will be able to capture a bigger share of income from the goods and services produced. (source)

Another indirect effect: increasing automation of manufacturing jobs pushes unionization rates down, which in turn decreases bargaining power among low-skilled workers. This, in the end, aggravates inequality yet again.

More posts in this series are here.

The Causes of Poverty (72): How Unemployment and the Incentive Theory Aggravate Poverty

Not all poor people are unemployed. Even in developed countries some of the poor are “working poor“, i.e. people who have an income that is below the poverty line. Hence, if some who work for a living can’t make ends meet, it’s obvious that the unemployed are even more at risk of being poor. At best, the latter only have a temporary replacement income in the form of unemployment benefits. This income is often lower than even the lowest wage income, and in most countries it’s also limited in time. Hence, poverty is the likely result of unemployment, even in wealthy countries.

To some extent, this result is intentional: governments want to “incentivize” people and “nudge” them into the labor market. High unemployment benefits that aren’t limited in time – combined with the fact that many jobs don’t pay enough to avoid poverty – may trap some people in unemployment. Low and limited unemployment benefits, on the other hand, invoke the horror of poverty which in turn may force people to look for work.

It’s obvious why many unemployed people are poor – their incomes are just too low, and intentionally so. But why are poor people unemployed? The incentive theory tells us that poor people shouldn’t be unemployed. The horror of poverty should drive them out of unemployment. While some people are probably “incentivized” in this way, this is not something that will work for all the unemployed poor. It will only work for all the unemployed poor if jobs are abundant, and indeed that’s what the incentive theory assumes. Of course, this assumption is wrong. There’s something called the natural rate of unemployment. Even in the best of economic circumstances, there’s always some level of involuntary unemployment, and during recessions, this level is higher than normal. During a recession (starting from 2008 in this case) more poor people than normal indicate that their unemployment is due to their “inability to find a job”.

So, there are no conceivable incentives that will push all unemployed poor into the workforce, not even at the best of times.

Notice how the inability to find a job is just one reason among many, and definitely not the most important reason why poor people don’t work. Choice (“home or family reasons”) is another reason. And that’s one which can sometimes be influenced by incentives. However, the reason that’s cited most frequently, namely “sickness or disability” cannot. The horror of poverty will not force you into employment if you can’t work for health reasons. Hence, even if we assume, unrealistically, that there is an abundance of jobs and no natural rate of unemployment, it’s unrealistic to assume that all poor unemployed people will one day join the workforce.

This has implications for the incentive theory. If you put too much emphasis on incentives, you’re likely to put unemployment benefits at a very low level with a strict time limit. After all, the closer the horizon of poverty, the more likely that people will act and look for a job. However, if there are not enough jobs or if people have other reasons why they won’t work, then they won’t work. And if unemployment benefits are stingy, then these people will become poor. Incentives don’t always work as intended. Sometimes they may even cause poverty.

Now, I do agree that work is important and that we should therefore try to get as many people in jobs as we can (and, by the way, improve the quality of those jobs; that’s also in the book). Not only because working people normally – although not always – are less at risk of being poor, but also because work is important in human flourishing. I also agree that incentives can play a role in getting people to find work, and that the specter of poverty can be used as an incentive. However, we should be careful with this particular incentive. It may work sometimes, but often it doesn’t.

More posts in this series are here.

Migration and Human Rights (42): The Labor Cost Argument Against Open Borders

I’ve argued many times before against the popular view that increased immigration is detrimental to native employment and income. The simple argument about an increase in supply of cheap labor driving down wages and forcing expensive native workers out of the job market is just that: simple, too simple. There’s even evidence that the opposite is true: immigration increases native wages (because it allows native workers to move up the pay scale). But even if immigration did impose a cost on the host country, that wouldn’t be the final argument against immigration, since such a cost could be seen as a form of global redistribution and global justice: improving the lot of the poorest of the world surely justifies imposing a burden on those who have more wealth and who had the good fortune of being born in the “right” part of the world. True, this burden shouldn’t fall on the poorest members of the “right” countries, but if it does that can be corrected by national redistribution.

Still, let’s return to the labor cost argument against immigration. Here’s another piece of evidence that tips the scales yet a bit further against the view that the extremely low cost of immigrant labor results in displacement of low-level native labor. The evidence I want to cite is about internal migration in China, but it’s perfectly possible to use it against arguments favoring restrictions on international migration:

Hundreds of millions of rural migrants have moved into Chinese cities since the early 1990s contributing greatly to economic growth, yet, they are often blamed for reducing urban ‘native’ workers’ employment opportunities, suppressing their wages and increasing pressure on infrastructure and other public facilities. This paper examines the causal relationship between rural-urban migration and urban native workers’ labour market outcomes in Chinese cities. After controlling for the endogeneity problem our results show that rural migrants in urban China have modest positive or zero effects on the average employment and insignificant impact on earnings of urban workers. When we examine the impact on unskilled labours we once again find it to be positive and insignificant. We conjecture that the reason for the lack of adverse effects is due partially to the labour market segregation between the migrants and urban natives, and partially due to the complementarities between the two groups of workers. Further investigation reveals that the increase in migrant inflow is related to the demand expansion and that if the economic growth continues, elimination of labour market segregation may not necessarily lead to an adverse impact of migration on urban native labour market outcomes. (source, source)

More posts about arguments against open borders are here, here and here. More posts in this blog series are here.

The Causes of Human Rights Violations (30): Language

The effect of language on human rights can be straightforward, as in the case of hate speech. Imagine an individual member of a racial minority living among members of the majority. The latter are constantly hurling insults and hateful bile at this individual, making it almost impossible for her to move about the neighborhood, find employment and do many of the other things she has a right to do. In this case, a particular type of language and a particular use of this language has obvious repercussions on someone’s human rights.

But what I’m interested in here are more subtle effects of language on human rights. Take the example of the gender-exclusive pronoun. In most languages, personal pronouns distinguish male from female, and the male pronoun is the default: when we’re not talking about a specific person, or when we’re talking about a mixed gender group, then we use the male pronoun. (This is similar to the equally common rule that children should get the surname of the father). Attempts to invent and promote gender-neutral or gender-inclusive pronouns haven’t quite succeeded, and the habit of using the female pronoun as the gender-exclusive one is often considered awkward. I also fail to avoid the traditional rule in my writing. In general, this problem is often labeled a fake one, invented by people high on political correctness.

And yet, the problem isn’t fake at all. Compared to other, more extreme uses of language such as hate speech, the harm done by the use of gender-exclusive pronouns may be small, but it’s not negligible. There’s a study here that tries to measure the harm:

Three studies assessed whether a common cultural practice, namely, the use of gender-exclusive language (e.g., using he to indicate he or she), is experienced as ostracism at the group level by women. Women responded to the use of gender-exclusive language (he) during a mock job interview with a lower sense of belonging, less motivation, and less expected identification with the job compared to others exposed to gender-inclusive (he or she) or gender-neutral (one) language (Studies 1 and 2). Moreover, the more emotionally disengaged women became over the course of a job interview upon hearing gender-exclusive language, the less motivation and job identification they subsequently reported (Study 3). Together, these studies show that subtle linguistic cues that may seem trivial at face value can signal group-based ostracism and lead members of the ostracized group to self-select out of important professional environments.

Another study focused on the use of gendered words in job ads, and found that ads signal whether a job is typically held by men or women. As a result of this signaling, women are less likely to apply to certain jobs, and this in turn perpetuates gender inequality in the workplace. Wording differences in ads affect the job’s relative appeal to men and women, independent of the type of job. The use of more masculine wording such as “competitive” makes traditionally female-dominated jobs more appealing to men, and vice versa.

In both these examples – gender-exclusive pronouns and gendered language in job ads – women respond – or are made to respond – in an unconscious way so as to perpetuate gender inequality.

A similar example of language affecting human rights is called stereotype threat: when the belief that people like you (African-Americans, women, etc) are worse at a particular task than the comparison group (whites, men, etc) is made prominent through some kind of preliminary “information” or briefing, then you perform worse at that task. For example, if a group of girls about to take a mathematics test, is “reminded” that boys tend to do better on this type of test, it’s likely that the girls will do more poorly on the test than they would have done had they not been told. It’s not difficult to imagine cases in which this can be used in order to perpetuate inequality, submission and domination. (More on stereotype threat here and here).

I just mentioned signaling, and signaling in the more strict definition of the term – engaging in speech or activity not necessarily for the sake of this speech or activity but in order to convey relevant information about yourself (for example, people acquire an education and talk about it not only because they want to be educated but also because they want to signal ability to potential employers) – is yet another example of the way in which language affects human rights. Take the case of capital punishment: I strongly believe that this is not about fighting crime, just retribution or desert, or even anger and revenge. Proponents of capital punishment, by expressing their support for it, signal their own moral rectitude. This is especially important for politicians, elected judges etc. In other words, for those who could, if they wanted, end the practice. (More on human rights and signaling is here).

Obviously, language doesn’t always have a negative effect on human rights. It’s easy to find examples of a positive effect: storytelling can promote empathy, and language aimed at shaming people can rid the world of rights violations when reasoning is insufficient.

Gender Discrimination (25): The Plough as a Cause of Gender Inequality

Gender inequality means different levels of protection of human rights according to gender. No need to say which of the two gender’s rights are usually violated more or protected less rigorously. Gender inequality occurs in many areas of life:

  • in political representation or participation
  • in income or labor market participation
  • in labor sorting (when women are relegated to certain professions)
  • in family life (when women do not have the same marriage or divorce rights, inheritance rights etc.)
  • in criminal justice (when the testimony of women is considered less valuable) etc.

Too many areas to mention, unfortunately.

When you read about the causes of gender inequality, the usual suspects are religion, patriarchy and all sorts of anti-women prejudice. A different and interesting perspective, focused on inequality in the labor market, is the following:

Ester Boserup … argues that gender role differences have their origins in different forms of agriculture practiced traditionally. In particular, she identifies important differences between shifting and plough cultivation. The former, which uses hand-held tools like the hoe and the digging stick, is labor intensive and women actively participate in farm work. The latter, in contrast, is more capital intensive, using the plough to prepare the soil. Unlike the hoe or digging stick, the plough requires significant upper body strength, grip strength, and burst of power, which are needed to either pull the plough or control the animal that pulls it.

Because of these requirements, when plough agriculture is practiced, men have an advantage in farming relative to women. Also reinforcing this gender-bias in ability is the fact that when the plough is used, there is less need for weeding, a task typically undertaken by women and children. In addition, child-care, a task almost universally performed by women, is most compatible with activities that can be stopped and resumed easily and do not put children in danger. These are characteristics that are satisfied for hoe agriculture, but not for plough agriculture since large animals are typically used to pull the plough. …

[T]his division of labor then generated norms about the appropriate role of women in society. Societies characterized by plough agriculture, and a resulting gender-based division of labor, developed the belief that the natural place for women is within the home. These cultural beliefs tend to persist even if the economy moves out of agriculture, affecting the participation of women in activities performed outside of the home, such as market employment, entrepreneurship, and participation in politics. (source)

And there does seem to be a strong statistical correlation between historical plough use and prejudice against women.