In other words, the incomes of people who were initially at the bottom rose more than the incomes of people who were initially at the top. This is the direct opposite of the picture presented by Census data, based on following income brackets over time, instead of following the people who are moving in and out of those brackets.
Similar patterns appeared in statistics from the Internal Revenue Service, which also followed given individuals. The IRS found that between 1996 and 2005 the income of individuals who had been in the bottom 20 percent of income tax filers in 1996 had increased by 91 percent by 2005, and the income of those individuals who were in the top one percent in 1996 had fallen by 26 percent.{316} It may seem almost impossible that the data from the Bureau of the Census and the data from the IRS and the University of Michigan can all be correct, but they are. Studies of income brackets over time and studies of individual people over time are measuring fundamentally different things that are often confused with one another.
A study of individual incomes over time in Canada turned up patterns very similar to those in the United States. During the period from 1990 to 2009, Canadians who were initially in the bottom 20 percent had both the highest absolute increase in income and the highest percentage increase in income. Only 13 percent of the Canadians who were initially in the bottom quintile in 1990 were still there in 2009, while 21 percent of them had risen all the way to the top quintile.{317}
Whatever the relationship between one income bracket and another, that is not necessarily the relationship between people, because people are moving from one bracket to another as time goes on. Therefore the fate of brackets and the fate of people can be very different—and, in many cases, completely opposite. When income-earning Americans in the bottom income bracket have their incomes nearly double in a decade, they end up no longer in the bottom bracket. There is nothing mysterious about this, since most people begin their careers in entry-level jobs and their growing experience over the years leads to higher incomes. Nor is it surprising that people whose incomes are at the peak of the income pyramid seem often also to be at or near their own peak incomes and do not continue to rise as dramatically as those who started at the bottom.
Some Americans reach the top one percent in income—approximately $369,500 and up in 2010{318}—in a given year because of some particular boost to their income during that particular year. Someone who sells a house may have an income that year which is some multiple of the income received in any year before or since. Similarly for someone who receives a large inheritance in a given year, or cashes in stock options that have been accumulating over the years. Such spikes in income account for a substantial proportion of those whose incomes in a given year reach the top levels. More than half the people in the top one percent in income in 1996 were no longer at that level in 2005. Among those in the top one-hundredth of one percent in income in 1996, three-quarters were no longer at that level in 2005.{319}
Many people who never have a spike in income that would put them in the top one percent may nevertheless end up in the top 20 percent after many years of moving up in the course of a career. They are not “rich” in any meaningful sense, even though they may be called that in political, media or even academic rhetoric. As already noted, the amount of income required to reach the top 20 percent is hardly enough to live the lifestyle of the rich and famous. Nor will being in the top one percent, for that half of the people in that bracket who do not remain there.
Just as there are spikes in income from time to time, so there are troughs in income in particular years. Thus many people who are genuinely affluent, or even rich, can have business losses or off years in their professions or investments, so that their income in a given year may be very low or even negative, without their being poor in any meaningful sense. This may help explain such anomalies as hundreds of thousands of people with incomes below $20,000 a year who are living in homes costing $300,000 and up.{320}
The fundamental confusion that makes income bracket data and individual income data seem mutually contradictory is the implicit assumption that people in particular income brackets at a given time are an enduring “class” at that level. If that were true, then trends over time in comparisons between income brackets would be the same as trends over time between individuals. Because that is not the case, however, the two sets of statistics lead not only to different conclusions but even to opposite conclusions that seem to contradict each other.
The higher up the income scale people are, the more volatile are their incomes. “During the past three recessions, the top 1% of earners (those making $380,000 or more in 2008) experienced the largest income shocks in percentage terms of any income group in the U.S.” the Wall Street Journal reported. When the incomes of people making $50,000 or less fell by 2 percent between 2007 and 2009, the incomes of people making a million dollars or more fell by nearly 50 percent.{321}
Conversely, when the economy grows, the incomes of the top one percent “grow up to three times faster than the rest of the country’s.” This is not really surprising, since incomes at the highest levels are less likely to be due to salaries and more likely to be due to income from investments or sales, both of which can vary greatly when the economy goes up or down. Similar patterns apply to wealth as to income. “During the 1990 and 2001 recessions, the richest 5% of Americans (measured by net worth) experienced the largest decline in their wealth,” the Wall Street Journal reported.{322}
Differences in Skills
Among the many reasons for differences in productivity and pay is that some people have more skills than others. No one is surprised that engineers earn more than messengers or that experienced shipping clerks tend to earn higher pay than inexperienced shipping clerks—and experienced pilots tend to earn more than either. Although workers may be thought of as people who simply supply labor, what most people supply is not just their ability to engage in physical exertions, but also their ability to apply mental proficiency to their tasks. The time when “a strong back and a weak mind” were sufficient for many jobs is long past in most modern economies. Obvious as this may seem, its implications are not equally obvious nor always widely understood.
In those times and places where physical strength and stamina have been the principal work requirements, productivity and pay have tended to reach their peak in the youthful prime of life, with middle-aged laborers receiving less pay or less frequent employment, or both. A premium on physical strength likewise favored male workers over female workers.
In some desperately poor countries living close to the edge of subsistence, such as China in times past, the sex differential in performing physical labor was such that it was not uncommon for the poorest people to kill female infants. While a mother was necessary for the family, an additional woman’s productivity in arduous farm labor on small plots of land with only primitive tools might not produce enough food to keep her alive—and her drain on the food produced by others would thus threaten the survival of the whole family, at a time when malnutrition and death by starvation were ever-present dangers. One of the many benefits of economic development has been making such desperate and brutal choices unnecessary.
The rising importance of skills and experience relative to physical strength has changed the relative productivities of youth compared to age, and of women compared to men. This has been especially so in more recent times, as the power of machines has replaced human strength in industrial societies and as skills have become crucial in high-tech economies. Even within a relatively short span of time, the age at which most people receive their peak earnings has shifted upward. In 1951, most Americans reached their peak earnings between 35 and 44 years of age, and people in that age bracket earned 60 percent more than workers in their early twenties. By 1973, however, people in the 35 to 44-year-old bracket earned more than double the income of the younger workers. Twenty years later, the peak earnings bracket had moved up to people aged 45 to 54 years, and people in that bracket earned more than th
ree times what workers in their early twenties earned.{323}
Meanwhile, the dwindling importance of physical strength also reduced or eliminated the premium for male workers in an ever-widening range of occupations. This did not require all employers to have enlightened self-interest. Those who persisted in paying more for male workers who were not correspondingly more productive were at a competitive disadvantage compared to rival firms that got their work done at lower costs by eliminating the male premium, equalizing the pay of women and men to match their productivities. The most unenlightened or prejudiced employers had higher labor costs, which risked the elimination of their businesses by the ruthlessness of market competition. Thus the pay of women began to equal that of men of similar qualifications even before there were laws mandating equal pay.
While the growing importance of skills tended to reduce economic inequalities between the sexes, it tended to increase the inequality between those with skills and those without skills. Moreover, rising earnings in general, growing out of a more productive economy with more skilled people, tended to increase the inequality between those who worked regularly and those who did not. As already noted, there are striking differences between the numbers and proportions of people who work and those who don’t work, as between the top income brackets and the bottom income brackets. A simultaneous rise in rewards for work and a growing welfare state that allows more people to live without working virtually guarantees increasing inequality in earnings and incomes, when many of the welfare state benefits are received in kind rather than in money, such as subsidized housing or subsidized medical care, since these benefits are not counted in income statistics.
One of the seemingly most obvious reasons for different individuals (or nations) to live at very different economic levels is that they produce at very different economic levels. As economies grow more technologically and economically more complex, and the work less physically demanding, those individuals with higher skills are more in demand and more highly rewarded. The growing disparities between upper level income brackets and lower level income brackets are hardly surprising under these conditions.
Job Discrimination
While pay differences often reflect differences in skills, experience, or willingness to do hard or dangerous work, these differences may also reflect discrimination against particular segments of society, such as ethnic minorities, women, lower castes, or other groups. However, in order to determine whether there is discrimination or how severe it is, we first need to define what we mean.
Sometimes discrimination is defined as judging individuals from different groups by different standards when hiring, paying or promoting. In its severest form, this can mean refusal to hire at all. “No Irish Need Apply” was a stock phrase in advertisements for many desirable jobs in nineteenth-century and early twentieth-century America. Before World War II, many hospitals in the United States would not hire black doctors or Jewish doctors, and some prestigious law firms would not hire anyone who was not a white Protestant male from the upper classes. In other cases, people might be hired from a number of groups, but individuals from different groups were channeled into different kinds of jobs.
None of this has been peculiar to the United States or to the modern era. On the contrary, members of different groups have been treated differently in laws and practices all around the world and for thousands of years of recorded history. It is the idea of treating all individuals the same, regardless of what group they come from, that is relatively recent as history is measured, and by no means universally observed around the world today.
Overlapping with discrimination, and often confused with it, are employment differences based on substantial differences in skills, experience, work habits and behavior patterns from one group to another. Mohawk Indians, for example, were long sought after to work on the construction of skyscrapers in the United States, for they walked around high up on the steel frameworks with no apparent fear or distraction from their work. In times past, Chinese laborers on rubber plantations in colonial Malaya were found to collect twice as much sap from rubber trees in a given amount of time as Malay workers did.
While preferences for some groups and reluctance or unwillingness to hire others have often been described as due to “bias,” “prejudice,” or “stereotypes,” third-party observers cannot so easily dismiss the first-hand knowledge of those who are backing their beliefs by risking their own money. Even in the absence of different beliefs about different groups, application of the same employment criteria to different groups can result in very different proportions of these groups being hired, fired, or promoted. Distinguishing discrimination from differences in qualifications and performances is not easy in practice, though the distinction is fundamental in principle. Seldom do statistical data contain sufficiently detailed information on skills, experience, performance, or absenteeism, much less work habits and attitudes, to make possible comparisons between truly comparable individuals from different groups.
Women, for example, have long had lower incomes than men, but most women give birth to children at some point in their lives and many stay out of the labor force until their children reach an age where they can be put into some form of day care while their mothers return to work. These interruptions of their careers cost women workplace experience and seniority, which in turn inhibit the rise of their incomes over the years, relative to that of men who have been working continuously in the meantime. However, as far back as 1971, American single women who worked continuously from high school into their thirties earned slightly more than single men of the same description,{324} even though women as a group earned substantially less than men as a group.
This suggests that employers were willing to pay women of the same experience the same as men, if only because they are forced to by competition in the labor market, and that women with the same experience may even outperform men and therefore earn more. But differences in domestic responsibilities prevent the sexes from having identical workplace experience or identical incomes based on that experience. None of this should be surprising. If, for example, women were paid only 75 percent of what men of the same level of experience and performance were paid, then any employer could hire four women instead of three men for the same money and gain a decisive advantage in production costs over competing firms.
Put differently, any employer who discriminated against women in this situation would be incurring unnecessarily higher costs, risking profits, sales, and survival in a competitive industry. It is worth noting again the distinction made in Chapter 4 between intentional and systemic causation. Even if not a single employer consciously or intentionally thought about the economic implications of discriminating against women, the systemic effects of competition would tend to weed out over time those employers who paid a sex differential not corresponding to a difference in productivity. This process would be hastened to the extent that women set up their own businesses, as many increasingly do, and do not discriminate against other women.
Substantial pay differentials between women and men are not the same across the board, but vary between those women who become mothers and those who do not. In one study, women without children earned 95 percent of what men earned, while women with children earned just 75 percent of what men earned.{325} Moreover, even those women without children need not be in the same occupations as men. The very possibility of having children makes different occupations have different attractions to women, even before they become mothers. Occupations like librarians or teachers, which one can resume after a few years off to take care of small children, are more attractive to women who anticipate becoming mothers. But occupations such as computer engineers, where just a few years off from work can leave you far behind in this rapidly changing field, tend to be less attractive to many women. In short, women and men make different occupational choices and prepare for many of these occupations by specializing in a very different mix of subjects while being educated.
The question as to whether or how much discrimination women encounter in the labor market is a question about whether there are substantial differences in pay between women and men in the same fields with the same qualifications. The question as to whether there is or is not income parity between the sexes is very different, since differences in occupational choices, educational choices, and continuous employment all affect incomes. Men also tend to work in more hazardous occupations, which usually pay more than similar occupations that are safer. As one study noted, “although 54 percent of the workplace is male, men account for 92 percent of all job-related deaths.”{326}
Similar problems in trying to compare truly comparable individuals make it difficult to determine the presence and magnitude of discrimination between groups that differ by race or ethnicity. It is not uncommon, both in the United States and in other countries, for one racial or ethnic group to differ in age from another by a decade or more—and we have already seen how age makes a big difference in income. While gross statistics show large income differences among American racial and ethnic groups, finer breakdowns usually show much smaller differences. For example, black, white, and Hispanic males of the same age (29) and IQ (100) have all had average annual incomes within a thousand dollars of one another.{327} In New Zealand, while there are substantial income differences between the Maori population and the white population, these differences likewise shrink drastically when comparing Maoris with other New Zealanders of the same age and with the same skills and literacy levels.{328}
Much discussion of discrimination proceeds as if employers are free to make whatever arbitrary decisions they wish as to hiring or pay. This ignores the fact that employers do not operate in isolation but in markets. Businesses compete against each other for employees as well as competing for customers. Mistaken decisions incur costs in both product markets and labor markets and, as we have seen in earlier chapters, the costs of being wrong can have serious consequences. Moreover, these costs vary with conditions in the market.