Is the Increased Earnings Inequality among Americans Bad? BECKER

Income inequality widened, particularly between urban and rural households after China began its rapid rate of economic development in 1980. At the same time, the fraction of Chinese men, women, and children who live on less than $2 a day--the World Bank's definition of poverty--greatly fell. Few would argue that the poor in China did not become much better off due to the rapid economic development, even though the gap between their incomes and those of the middle and richer classes widened by a lot. A similar conclusion would apply to India as the explosion in its general economic development during the past 20 years widened the gap between rich and poor, but raised the income levels of the very poor. I make this observation in reaction to the great concern expressed by politicians and many others in the United States over the rather substantial increase during the past 25 years in earnings inequality among Americans. The China and India examples illustrate that whether rising inequality is considered good or bad depends on how it came about. I believe that the foundation of the growth in earnings inequality of Americans has mainly been beneficial and desirable. The basic facts are these. There has been a general trend toward rising gaps between the earnings of more and less skilled persons. With regard to education, real earnings (that is, earnings adjusted for changes in consumer prices) earnings of high school dropouts did not change much. Earnings of high school graduates grew somewhat more rapidly, so that the gap between dropout and graduate earnings expanded over time. The main action came in the earnings of college graduates and those with postgraduate education. They both increased at a rapid pace, with the earnings of persons with MBA's, law degrees, and other advanced education growing the most rapidly. All these trends produced a widening of earnings inequality by education level, particularly between those with college education and persons with lesser education. I should also note that while an upward trend in the earnings gap by education is found for both men and women, and for African Americans and whites, the earnings of college educated women and African Americans increased more rapidly than did those of white males. As a result, inequality by sex and race, particularly among college educated persons, narrowed by a lot. As the education earnings gap increased, a larger fraction of high school graduates went on to get a college education. This trend toward greater higher education is found among all racial and ethnic groups, and for both men and women, but it is particularly important for women. The growth in the number of women going to and completing college has been so rapid that many more women than men are now enrolled as college students. Women have also shifted toward higher earnings fields, such as business, law, and medicine, and away from traditional occupations of women, such as K-12 teachers and nurses. The greater education achievement of women compared to men is particularly prominent among blacks and Latinos. The widening earnings gap is mainly due to a growth in the demand for educated and other skilled persons. That the demand for skilled persons has grown rapidly is not surprising, given developments in computers and the Internet, and advances in biotechnology. Also, globalization increased the demand for products and services from the U.S. and other developed nations produced by college educated and other highly skilled employees. Globalization also encouraged a shift to importing products using relatively low-skilled labor from China and other low wage countries instead of producing them domestically. Rates of return on college education shot up during the past several decades due to the increased demand for persons with greater knowledge and skills. These higher rates of return induced a larger fraction of high school graduates to get a college education, and increasingly to continue with postgraduate education. Some of you might question whether rates of return on higher education did increase since tuition grew rapidly during the past twenty-five years. However, increases in tuition were mainly induced by the greater return to college education. Pablo Pena in a PHD dissertation in progress at the University of Chicago argues convincingly that tuition rose in part because students want to invest more in the quality of their education. Increased spending per student by colleges is partly financed by higher tuition levels. This brings me finally to the punch line. Should not an increase in earnings inequality due primarily to higher rates of return on education and other skills be considered a favorable rather than unfavorable development? Higher rates of return on capital are a sign of greater productivity in the economy, and that inference is fully applicable to human capital as well as to physical capital. The initial impact of higher returns to human capital is wider inequality in earnings (just as the initial effect of higher returns on physical capital is widen income inequality), but that impact becomes more muted and may be reversed over time as young men and women invest more in their human capital. I conclude that the forces raising earnings inequality in the United States is on the whole beneficial because they were reflected higher returns to investments in education and other human capital. Yet this is not a ground for complacency, for the responses so far to these higher returns is disturbingly limited. Why have not more high school graduates gone on for college education when the benefits are so apparent? And why did the fraction of American youth who drop out of high school, especially African American and Hispanic males, remain quite constant at about 25 per cent of all high school students? The answer to both questions lies partly in the breakdown of the American family, and the resulting low skill levels acquired by children in broken families. Cognitive skills tend to get developed at very early ages, while my colleague, James Heckman, has shown that non-cognitive skills, such as study habits, getting to appointments on time, and attitudes toward work, get fixed at later, although still relatively young, ages. High school dropouts certainly appear to be seriously deficient in the non-cognitive skills that would enable them to take advantage of the higher rates of return to greater investments in education and other human capital. So instead of lamenting the increased earnings gap by education, attention should focus on how to raise the fraction of American youth who complete high school, and then go on for a college education. These pose tough challenges since the solutions are not cheap or easy. But it would be a disaster if the focus were on the earnings inequality itself. For that would lead to attempts to raise taxes and other penalties on higher earnings due to greater skills, which could greatly reduce the productivity of the world's leading economy by discouraging investments in human capital.

Comments

[Hans Gruber]

Immigration (which was noted by Posner) probably has more to do with rising income inequality than many people think.

It's not that immigration per se would do this; indeed an immigration policy intended to attract educated and skilled workers would decrease income inequality.

Illegal immigration in particular skews our immigrant population to be very low-skilled and uneducated. This is because educated or skilled workers are not willing to pay the costs of illegal status.

We have around 11-20 million illegal aliens in the US (6-10 million workers), most of which are employed in occupations like Construction and food service, jobs that millions of Americans depend on to support their families. This massive influx has undoubtedly lowered the wages in these occupations, and in some cases so low as to drive out certain low skilled workers, like high school dropouts, out of the labor market entirely (current unemployment rate among high school dropouts is 15% or so).

If I were to venture a guess on the biggest causes of the increasing income inequality it would be globalization and immigration, followed by the rise of the importance in technology (many workers have trouble adapting or feel intimidated by these changes).

The increasing demand for college educated workers is at least somewhat troubling. Many people are just not intelligent enough to complete college, and even if they do the utility of finishing is probably much lower. What happens to these individuals? It used to be that one could make a decent living working in a factory, meat-packing plant, or in construction. This reality is disappearing because of globalization and immigration. We can't send everybody to college, Mr. Becker, so the rise of a college education to necessity status should be troubling. The left of the bell curve needs to make a living and provide for their families, too. What does the future hold for them?

[Arun Khanna]

Three points:

1. I agree with your conclusions that rising inequality in income reflects increasing returns to human capital investments. I would add two related points.

2. Rising returns to human capital are not due to generic higher education per se but rather due to any form of specialized human capital investment. Such specialized human capital investments make a first world worker outsourcing-immune.

3. One factor explaining why more high school graduates are not opting for higher human capital investments could be that universities are not geared up to deal with the changing dynamic. I conjecture that a broader 'Summers effect' prevails across American universities. Additional ÔøΩsoftÔøΩ demand for human capital investment in specialized sciences, applied mathematics, engineering and bio-medical fields exist. Hard supply of university education is not responding to this potential new university recruits; rather following Say's law supply (or over-supply) of liberal arts degrees creates its own demand. Two reforms are needed in university education. First, we need 3rd millennium universities that have professional schools at the core competing with 2nd millennium universities that have liberal arts schools at the core. Second, interest rates on financial aid should vary across degree programs based on past payment records of university graduates in that field.

[James]

One thing to sort out is how much of the return to education is merely signalling, as opposed to value added. I'm pretty sure my earnings potential increased because I attended a selective college, but I'm not sure I learned anything there that will make me a more productive worker. This doesn't mean Becker is wrong (employers increasingly need skills that can be identified with college admissions data), although we might think about ways to sort workers at a lower cost.

[Hans Gruber]

"This doesn't mean Becker is wrong (employers increasingly need skills that can be identified with college admissions data), although we might think about ways to sort workers at a lower cost."

I almost made this same point myself. Most individuals pursuing undergraduate education are not doing so because they believe they will learn valuable new skills and acquire special insights--they go to college because that's what employers are demanding of their employees.

And employees are not by and large believers in the value of education per se, they are believers in the value of education as a sorting mechanism (intelligence, initiative, discipline).

[Wes]

There are certain things that are really really bad for an economy, one of them is being micro-managed by a corrupt dictatorship. I would attribute China's economic growth less to an increased gap between the rich and poor and more to a decrease in corrupt dictatorship micro-management. In fact, it is possible that China's economy would be growing even faster if it was taxing the very rich heavily (while refraining from micro-management) and using the proceeds to provide infrastructure and education for the very poor.If China's economy was much more efficient than the United States economy (eg. higher per capita GDP) and the only difference was a greater gap between the rich and the poor, then maybe the United States should be copying the Chinese economic model, but otherwise copying the Chinese economic model would be expected to make the United States economy less efficient.In fact, it is interesting to note that the United States economy has done quite well over the last century and that most of that century saw very high progressive tax rates on the rich (often as high as 90%).Income distributions in unregulated capitalist economies tend toward a distribution where a few people are very rich and everyone else is very poor. There are two questions: 1. Does such an unequal distribution maximize the efficiency of the economy? 2. If not, is it possible to achieve a more efficient wealth distribution without imposing other inefficiencies that offset the gain from the more efficient wealth distribution.In particular, does a highly progressive tax that achieves a more even wealth distribution result in a net increase or a net decrease in economic efficiency?A related question is: for a given level of economic efficiency, what wealth distribution maximizes the general welfare and happiness of the society as a whole? For example, is a society better off when everyone makes $50,000 a year or when 99% of the people make $20,000 a year and 1% of the people make $3,020,000 a year?

[Berend de Boer]

Another issue is tax. I'm guessing that even in America most people don't pay tax anymore. The top 1% pays 80% of the taxes. So for the top to get any meaningful additional income, they will have to earn a lot more money. And such money isn't then available for the people with lower incomes, widening the gap.

It all comes back to the usual "unintended consequences".

[Anupam]

How much of the income inequality is due to "human capital investments" and how much of it is due to hereditary capital (descendant of rich-person - a la Rockfeller, ...).

I don't have access to this kind of data and would be interested in your views.

I would assume that this lack of downward mobility in the heirs of the rich is still a justification for progressive taxation.

[Stockton Gaines]

"increase in earnings inequality due primarily to higher rates of return on education and other skills"

A lawyerly question indeed, in which you assume the conclusion you wish to reach. Do you have any evidence that this is the PRIMARY cause for the increased disparity in income?

[Jason Ruspini]

The Gordian Knot of social justice: how to decrease variance in wealth without also decreasing average/total wealth.

[Tim]

I was curious as to any of the negative externalities that are sometimes associated with rising income inequality. At the moment I am thinking of those pointed out by Chuck Collins and Bill Gates Sr. in their book, "Wealth and our Commonwealth: Why America Should Tax Accumulated Fortunes " especially, the correlation between poor health and income inequality. (I apologize, I can't quite remember the citation or page numbers, it was some time ago I read it.)

[Paco]

Dear Professor Becker:

I just wanted to make two minor points. One is regarding your analysis on rates of return to higher education. I once read an interesting essay by James Buchanan on how education, in addition to raising one's future stream of income, also has an effect on changing a person's tastes and even his self-identity. These intangibles of higher education seem like important 'nonpecuniary' forms of income that one should take into account.

The other point I wish to make is regarding your analysis of the high-school 'drop-out' rate. While I agree with you in principle that it would a good thing to reduce the drop-out rate, I would also argue that there exists an "optimal" drop-out rate, since many students are better off learning a trade or playing sports, than trying to read Shakespeare or learn geometry. Of course, in a perfect would, everyone would finish high school, go to college, be cultured, etc. etc., but that world is not the world we live in.

[Bill]

The question is not whether there should be a differential for educational attainment. The market should determine that. The question should be whether there are equal opportunities for people to obtain education.

[W]

interest rates on financial aid should vary across degree programs based on past payment records of university graduates in that field.

Just wanted to point out what a great idea this is.

[BillD]

This is a fascinating discussion...

First, as some of the previous posters have implied, I believe this is a primarily social, not economic issue. Other successful (Japan and Europe) (maybe not as much as USA) societies have much lower variance of wages (not sure about wealth). Also, some not very successful societies (Venezuela and other banana republics) have much higher variance of wealth (not sure about wages).

To me the biggest question is the equality of opportunity (to both rise AND fall), not results. I believe this historically fundamental trait of American society is becoming less abundant.

Look at ExxonMobil. I'm sure that Lee Raymond is a pretty skilled and intelligent person. But there is only 1 CEO. What is the marginal value of his skills to the company vs. most of the next few layers of management? Based on my experience, not very much and definitely not integer multiples. Not to mention the value of "the team." And also no CEO of a large public company takes real personal risk. Upside: deci- or centi- millionaire, downside: millionaire.

These questions of valuing marginal value of individual worth tend to vary over time. My fear is that our society, with less opportunity up, but particularly down, is producing a permanent upper class leading us to the aristocracy that we purposely shunned in the past.

[monkyboy]

It's interesting to note that China and America have the same Gini index (a measure of wealth distribution).

So America's income pie is growing, but educated people are getting an ever increasing slice.

At the same time, the world's income pie is growing, but America is getting an ever diminishing slice.

When was the last time America's GDP grew at a faster rate than the world's GDP?

Last year, the world's GDP grew 4.5%, America's GDP grew at 3.5%.

By this measure, America's highly-educated are grossly overpaid...but the less-educated paid the price.

Seems like we need more professionals from countries that are outgrowing the U.S. economically to come here and take over things...

Professors at China's top colleges make about $2000/year...seems like a great place to start.

Might even make a college degree more affordable.

[Arthur Gandolfi]

I am largely in agreement with Prof. Becker's analysis that rising income inequality in the U.S. is driven by higher returns to education. However, I do not believe that this means that there is no problem.

The combination of globalization and high rates of unskilled immigration is clearly driving down
the wages of unskilled Americans. Some of these people can, and should persue additional education. But what do we do with the bottom 15-25% of the population? People who have little human capital beyond their raw labor, and who probably have a low return on education.

I don't think, as a society, we are willing to have their wages driven to the market clearing level of China, or Mexico for unskilled labor.

Furthermore, it seems as if globalization is beginning to impact "white collar" jobs, e.g. outsourcing radiology and investment research positions to India, more and more.

If this trend continues, it could be the bottom 50% of our population (by skill) that has it's wages driven to world market clearing levels. We clearly will not tolerate that degree of economic inequality.

This can only lead to more gov't intervention in the economy, more income redistribution, and more "socialist" economic policies.

As a generally free-market conservative I think that the U.S. would be much better off if we implemented reasonable tariffs on low wage countries (to offset at least their cost advantages due to environmental, safety and labor regulations) and reduced immigration, to shield our low skill workers from some of the wage competition.

This will be far less costly in the long run, than the redistributionist policies that will inevitably result from higher income inequality.

Thanks for the interesting topis,

Arthur Gandolif

[Libertarian Propagandist]

I echo the observation of several posters that college is a sorting mechanism where students learn very little. Derek Bok, head of Harvard, admitted in a recent book that if you start with a student in the 50th percentile coming into college, he will only reach the 69th after 4 years of college.

The top 31 people are wasting their time going to college. They really be spend the next 4 years running their own business, or marketing themselves to employers. They have been brainwashed by Gary Becker and others in the educational establishment who accept government funds.

We need to separate school and state. For more info, see www.sepschool.org

[CRRA]

A related question is: for a given level of economic efficiency, what wealth distribution maximizes the general welfare and happiness of the society as a whole? For example, is a society better off when everyone makes $50,000 a year or when 99% of the people make $20,000 a year and 1% of the people make $3,020,000 a year?

The relevant question is:

Which society is better off?

A) Every body makes $40,000.
B) 99% make $45,000 and 1% makes $1,000,000.

[Bill Conerly]

The impact of investment in education on inequality is the second step of the story. Here's the first: rising average productivity with diverse preferences for labor vs. consumption. Let everyone have equal productivity. If output per person = subsistence, then there is perfect equality. At output = 2 x subsistence, some people live at subsistence and spend 1/2 their time in leisure; others keep nose to grindstone and consume 2 x subsistence; and others are in the middle with somewhat more leisure and somewhat more consumption. That's a more unequal distribution of income. With productivity at 10 x subsistence, we have a ratio of 10:1 of top earners to lowest earners. As productivity grows, the ratio of top income to bottom income grows. Call this the pure productivity effect.

I think that diverse attitudes about current consumption versus future consumption (which is the investment in human capital issue) would work with the pure productivity effect in a multiplicative way. Think of a matrix with the leisure-consumption preference on one axis, and time preference on the other. Of the people at the far end of the distribution for preferring consumption over leisure, some of them would also have a low rate of time preference. These people get lots of education and work very long hours, earning a lot. At the other extreme would be some people who happen to have both a high preference for leisure and a high rate of time preference. These people stay right around subsistence. The ratio of high income to low income is far greater than without educational opportunity. Rising productivity with increased returns to education really accentuates the inequality of earnings.

One more complication. Suppose that we throw in inherited wealth, and assume that it is well above subsistence for those who have it. Then rising productivity initially reduces inequality, by bringing at least some workers up to the income level of heirs. Eventually, rising productivity overwhelms the inheritance effect (Bill Gates is richer than any Rockefeller now alive) and inequality rises as discussed above. We would thus see a U shaped path for inequality as productivity grows.

[Haris]

Bill

Your analysis implicitly explains my prior point, which is that income inequality does not necessarily mean inequality. Even in a country where all opportunity is equal, including equal initial wealth (yay estate taxes!), fair access to education, and even equal physical and mental abilities, there will be income inequality because people have different preferences when it comes to consumption and leisure. Thus, two entirely identical people may end up having incomes tens or hundreds of thousands apart because one prefers consumption while another prefers leisure. This is simply a further reason that income inequality is not as big a problem as the numbers would imply, since some of the disparity may be caused by different preferences.

Posters have raised the social aspect of income inequality, and that is not one to be neglected. Even though the poor of today are much better off than the poor of 50 years ago (better goods, safer cars, internet, health care, etc), the social aspect of inequality never changes. Jealousy on one end and condescension on the other always contribute to class tensions, though these are certainly less pronounced in the US than they have been elsewhere. The absence of an ancient aristocracy, which Judge Posner feels is helpful, is actually probably exacerbating the problems of unequal access to education and opportunity. This is because those near the top of the income group feel that their position there is justified, that they have earned to be there. A previous poster used the example of a Harvard and a U of Alabama graduate. While it is probably true that the former Crimson looks down on the latter, many would agree with them because there is a widespread impression that there is sufficient equality of opportunity, and that those who did not make it to Harvard simply didn't work hard enough. I admit this is an oversimplification, but I believe that most people would say that there is sufficient equality and that income inequality is the result of people's individual efforts and talents.

This impression is highly erroneous, however. Studies have shown that socioeconomic status is very strongly correlated with abilities needed for a successful education. One study I recall showed that the very poor have half the reading skills of the middle class even before they are in first grade. Having much of one's future success essentially determined before one even encounters education is certainly not equality of opportunity. The perception that there is such equality, however, prevents real reforms because the voting public, which is ultimately responsible for bringing about such changes, is under the impression that no such reforms are needed.

[michael phillips]

A few commentors discussed the ideal of equality of opportunity. However, I doubt that they mean literally that they support an equality of opportunity, but instead support some type of "opportunity floor" under which no American should exist. "Equality of opportunity" is often contrasted with "equality of outcomes", a concept that seems relegated to the margins of civil discourse. Equality of opportunity, on the other hand, seems to mean something other than equality. It makes me wonder to what extent the virtue of equality continues to matter in America.

Judge Posner and Professor Becker both ignore the strongest case for why rising income inequality ought to be an issue in the United States: the question of justice. Both Posner and Becker write about an ideal market that distributes its rewards based on merit alone. While both scholars note that merit is largely the result of life's natural lottery - Posner cites "differences in IQ, energy, health, social skills, character, ambition, physical attractiveness, talent, and luck" - neither takes up the case for why life's natural lottery is the most just distributor of wealth in this country. It is the contention that the natural lottery does not provide distributive justice that I was most hoping Judge Posner and Professor Becker would rebut.

[Pan]

michael phillips wrote:
It is the contention that the natural lottery does not provide distributive justice that I was most hoping Judge Posner and Professor Becker would rebut.

-------------------------------
Sometimes we complain that the Nature or God or whatever is so unfair. Some people even are trying to correct what the nature has imposed on our human beings.
Being aware of the lessons of China 40 years ago, I can tell you for sure that that kind of experiments can only show you to what an extent manmade disasters can be. I believe you can still be suprised though you had been familiar with Concentration Camps in WWII.

[monkyboy]

Interesting post, michael.

I think most Americans realize that some people will get a head start in the economic "game" and can live with that fact.

It's the people who cheat to "win" the game, whether it's Barry Bonds and his steroids or Dick Cheney giving his old company billions of dollars worth of no-bid contracts, that cause resentment...

How many high-income Americans deserve an asterisk by their salary?

[Wes]

...neither takes up the case for why life's natural lottery is the most just distributor of wealth in this country.That's an excellent point.My view is that it's more about practical considerations than anything else. In order to consume, society must produce. If people who produce more are allowed to consume more then that creates an incentive to produce.While a capitalist system that ties consumption to production is good at motivating people to work, theories that capitalism is optimal in other respects seem to me to be mostly wishful overgeneralization on the part of proponents of capitalism.In particular, it is not clear that the distribution of wealth that results from a purely capitalist system is optimal either in terms of maximizing happiness or even in terms of optimizing total economic output.

[Jay Jeffers]

OK let's see if I can do this without writing a novel...I don't care about income all that much, I simply care about a rough level of social equality. I don't care if some people get rich and others stay poor. I don't care if people, in their free time, choose to associate with people who are carbon copies of themselves. What I care about is the rising inequality in access to health care. I care that poor rural communities are dealing with meth labs in their midst and poor urban communities are dealing with crack dealers and dangerous gangs. I care that educational opportunities are not truly equal. If we could have a basic level of shared responsibility and opportunity, then I wouldn't have a problem with people's natural abilities and preferences sorting everyone into all sorts of differing social and class groups.

The conservative line of attack against affirmative action is that in America, we assure equality of opportunity, not equality of outcomes. A better argument I cannot imagine. But does anyone think that we truly have equality of opportunity? Do children in the ghetto whose parents cannot provide them with quality health maintenance over the long run, and who have to worry about drive-by shootings and a sub-par school system really have the same opportunity that upper-middle class kids in relatively posh Chicago suburbs have with their publicly funded schools and safe streets? These margins are getting closer together as lower-middle and working class folks are exposed to the problems already mentioned.

Take a gander at this book by Mickey Kaus; his big-picture ideas are solid. Like the review on Amazon says, there is plenty to disagree with as far as the details of how to attain this social equality. I'm a little less ambitious than he is about the restructuring of policy, but can we at least distinguish between income on the one hand, and on the other what are more intangible things like equal opportunity and sacrifice, which could lead to social equality even in the midst of rising income inequality?

http://www.amazon.com/gp/product/0465098290/102-6282809-4408912?v=glance&n=283155

http://lilt.ilstu.edu/gmklass/pos334/archive/kaus.html