The United States has traditionally prided itself on being a
country that provides “equality of opportunity”. By that is usually meant that
individuals with ability and ambition who are born into families with modest
means and education have a good chance of succeeding when they become adults.
The experiences of different generations of immigrants coming from many
countries have supported this view, for their children and grandchildren often
succeeded very well as they blended into the mainstream of American life.
That rosy view of opportunities in the US is being
challenged by a growing belief that opportunities are declining for children
from modest backgrounds. This belief is
partly based on the strong evidence that human capital, including education, IQ
and other measures of “ability”, have become much more important in determining
economic success than was the case in the past. This new economic environment
favors educated parents because they have both greater schooling and also
higher average cognitive and non-cognitive skills.
Better-educated and higher income parents have always spent
more on their children’s education and training, including quality of schools
and after school activities. However, this gap in spending between families has
grown in recent decades as the relation between so-called “enrichment”
expenditures on children and family income has become much steeper. In
addition, better off families complement their spending of money on their
children by also spending more time with children through teaching them,
helping with homework, and in other ways.
Countering this growing advantage of children from better
off families is that different levels of government are doing more than in the
past to improve education opportunities from children from poorer backgrounds.
The many policies that work in this direction include subsidized pre
kindergarten activities, head start programs, and greater attention to the
public schools attended by poorer children.
Despite various claims about declining opportunities in the
US, and the indirect evidence that supports this conclusion, there is little
direct supporting evidence. Many commentators have confused the strong evidence
that inequality in the distribution of earnings and other incomes has increased
greatly since 1980 with declining opportunity for children from poorer
families. This is a jump in logic since greater inequality of income is
consistent with stable or even growing opportunities for these children.
To be sure, in comparisons among countries, greater
inequality of incomes in fact does tend to go along with fewer opportunities
for poorer children to rise up in the income scale. However, this finding does
not imply that countries that have growing inequality in the distribution of
incomes also have falling opportunities.
The direct evidence on what is happening over time to
opportunities provides a mixed picture. On the one hand is the evidence on the
IGE, which measures the average effect on children’s earnings of higher
parental earnings. For example, an IGE of 0.5 means that when parents earnings
are say 10% above average, their children’s earnings tend to be 5% above
average. Most of the evidence suggests that the IGE in the United States changed
little as inequality grew during the past 30 years. This direct evidence is
limited in part because different studies provide a range for the IGE in the US
from about 0.3 to about 0.7.
The other hand comes from work on what has happened over
time between the performances on achievement tests of children from families at
the lower and upper ends of the income distribution. In particular, a study in
progress by a graduate student at the University of Chicago, Eric Nielsen,
compares the performances of children in the late 1970s and 1990s that have
parents in the top and bottom 20 percentiles of the income distribution. To
make the comparisons he uses scores on the Armed Forces Qualifying Test (AFQT),
a widely used measure of achievement. Using (“ordinal”) methods of comparison
that he developed, Nielsen finds a considerable narrowing over this time period
between the performances on this test of children from the high and low ends of
the income distribution.
It is somewhat puzzling that the performance gap on
achievement tests has apparently narrowed while the gap between high and low
income families in their spending of time and money on their children’s human
capital has apparently widened. One important piece of missing evidence is on possible changes over time in the link between the AFQT performance and adult earnings.
In summary, it is far from clear whether opportunities of
children from poorer families have been declining in the US. Still, since
abundant opportunity for these children is an important goal of an efficient
and attractive society, more effort might be needed to improve their
opportunities through better schools, better teachers, and perhaps in other
ways as well.
Income inequality in the United States has grown substantially since the 1970s, to the point where today the bottom 20 percent of the nation’s households have 5 percent of total income, the top 10 percent about 50 percent, and the top 1 percent more than 20 percent. The question is whether such a high level of inequality is likely to reduce what is called “relative mobility,” or the likelihood that members of one generation of a family and members of the next generation will end up at different points in the income distribution. You are very likely to earn more than your father; the question is how likely are you to be higher (or lower) in the income distribution. If he is in the bottom quartile, for example, how likely are you to be in the next higher quartile?
Income inequality in the parents’ generation might be expected only to create income inequality in the next generation. Whether income inequality will affect relative mobility will depend on why income inequality in the current population is so high. One possibility is that, because of increased assortative mating as a result of declining discrimination and of the efficiency of online search for potential mates, there are greater differences in IQ across families than there used to be. Another possibility—closer to a certainty—is that as a result of automation in the broadest sense, the economic returns to IQ have risen relative to the returns to strength and stamina, which are the qualities important to such vocations as factory work, construction, mining, and farming.
The combination of assortative mating with higher returns to IQ could have dramatic effects on relative mobility if the effect was to insulate to a significant degree a prosperous family’s children from economic risk. And it may be. The adults in high-IQ families are disproportionately represented in the jobs (professional, managerial, financial, and so forth) that pay well, and their income can and often is used to give their children a boost—for example in the form of payment of tuition to high-quality (and very expensive) private schools, payment to tutors, a variety of other educational enrichments, and entry into high-quality colleges without need for their children to borrow to finance college (or graduate or professional school) and thus assume debt. Colleges like to admit kids from high-income families, seeing such kids as future donors. And high-IQ parents are likely to produce high-IQ children, further enhancing the children’s attractiveness to first-rate colleges. These factors, which loom larger the greater the inequality in the income distribution, because that inequality creates a highly affluent tier of families (a proximed by the income shares of the top 10 percent and within that group the top 1 percent) are likely to reduce relative mobility, by securing a disproportionate number of the top college and university admissions and top jobs for children of the intellectual-economic elite.
These factors can be offset to a significant extent by immigration, because immigrants tend to be more ambitious, bold, and determined than the average member of their nation of origin; refugee immigrants are often drawn from the elite of their nation of origin. First-generation immigrants tend not to have a high income, but to endow their children with the attitudes and abilities that enable the children to achieve economic success. Certainly the United States has benefited greatly in recent years from immigration from countries like China, India, and South Korea. But relative mobility that is the consequence of the artificially depressed income of first-generation immigrant families does nothing to promote relative mobility for the children of low-income native-born Americans.
A 2011 study by Scott Winship for the Brookings Institution reports that the likelihood that an American will rank higher in the income distribution than his parents is lower than in most other wealthy countries. The report states: “If being raised in the bottom fifth [of the income distribution] were not a disadvantage and socioeconomic outcomes were random, we would expect to see 20 percent of Americans who started in the bottom fifth remain there as adults, while 20 percent would end up in each of the other fifths. Instead, about 40 percent are unable to escape the bottom fifth. This trend holds true for other measures of mobility: About 40 percent of men will end up in low-skill work if their fathers had similar jobs, and about 40 percent will end up in the bottom fifth of family wealth (as opposed to income) if that’s where their parents were.” Income inequality is greater in the United States than in our peer countries, and may be responsible for our lower relative mobility. In the limit, an income distribution that produces a very wealthy top tier of earners and a very large bottom tier of poor or low-income families may reduce movement between the tiers in subsequent generations.
Becker points to Headstart and other government programs as possible counters to the effect of income inequality on economic opportunities for children of families that are rank low in the income distribution. This raises the question whether there may be a more efficient way of dealing with the problem of relative mobility than spending government money. A natural starting point would be to increase the very low federal income tax rate (15 percent) on dividends and capital gains, which is a significant factor in the increase in income inequality.
The back page of the Education section of the Wall Street Journal for October 9 features a very interesting debate among three economists; I have given the debate’s title to this blog post. The economists are Rudy Fichtenbaum, Katharine Lyall, and Richard Vedder, all experts on the economics of education and two (Fichtenbaum and Lyall) with considerable practical experience in academic administration. They don’t agree on the causes of the high costs of an American college or university education; and though they agree that those costs are a problem, they don’t agree on the solution. That is characteristic of the academic and journalistic literature on American higher education.
When I was an undergraduate at Yale College in the late 1950s, the cost of tuition plus room and board was $2000 a year; it is now almost $60,000, which after adjustment for inflation represents almost a fourfold increase. Yale is not unique, and the costs of a college education are comparable to those of Yale and other Ivy League schools (and their counterparts elsewhere in the United States, such as Stanford, the University of Chicago, and the University of Texas) even at many much less prestigious colleges.
The causes are various. They include the enormous—I am tempted to say the stifling—increase in legal and other regulation of colleges (and universities, but for simplicity I’ll use “college” to denote all higher education), the decline in financial subsidies to state colleges, the increased cost of scientific equipment, and the expense of computerization and other electronics. But another important cause,paradoxically, is the increased cost of college education, which tilts the student body toward richer kids—and rich kids and their parents expect superior amenities in the way of housing, food, athletic facilities, and police protection. Such students expected to be treated as consumers, rather than as kids with no rights or representation (the situation of students at Yale in the 1950s; there was no student government, and no appeal from expulsion).
In addition, the availability of federal loans to students enables colleges to jack up tuition. This is in part because the loans are subsidized, in part because young people tend to exaggerate their future prospects, and in part because even very intelligent young people often have difficulty comprehending interest rates. Still another cause is the very strong demand for college education. Word has gotten out that the gulf in employment opportunities between college and non-college educated young persons has widened and is likely to continue to do so. Supply lags demand because it takes a long time for a new college to establish credibility in the eyes of prospective employers, and because expansion of existing colleges, while of course both feasible and common, is costly and slow; it requires enlarging faculty, administrative staff, security, and physical plant—there are few if any economies of scale in higher education. When demand outruns supply, price rises.
Another factor in the rising cost of college, perhaps an uncontrollable one, is the shift from a professional to a business model of higher education, a trend similar to that observed in the legal and medical professions. Colleges seem far more competitive than of old, as symptomized by the big-business type salaries increasingly commanded by successful college executives, where success is measured primarily in financial terms. This competition has I think resulted in greater investment in intercollegiate athletics, greater attention in admissions to the income of an applicant’s family, grade inflation, reduction in number of required courses, and other pandering to student preferences. I worry that college trustees do not have the same incentives to hold down costs as corporate directors, who are answerable, at least to some extent, to shareholders. Private colleges (the majority of colleges and the best colleges) don’t have shareholders.
There is also a question about the value added of a college education. Although college seems a good investment, whether it is depends I think on the underemphasized factor of IQ. The average IQ of Americans is (by definition) 100, and about one-sixth of Americans have an IQ of 85 or less. And many kids who have a sufficiently high IQ to benefit, in principle at least, from a college education have character or psychological or other personal problems (including family responsibilities) that prevent them from deriving significant benefits from higher education. Not every career track should begin with a college graduation.
It is relevant to note that college completion rates have declined even though colleges require less effort by students to graduate. This is consistent with colleges’ enrolling more students who can’t benefit significantly from a college education.
I see hope, however, in the MOOCs—massive open online courses, which offer enormous potential cost savings and quality improvements for colleges. They can eliminate most of the living expenses associated with college (students can live at home, presumably cheaply) while enabling a reduction in faculty size (because there is no limit to the number of students in an online course) coupled with an increase in average faculty quality, since there is no limit on the number of students that a superb teacher can teach online. The MOOCs are not a panacea, but they are the most promising response to the problem of the high costs of a college education in America.
As Posner discusses, tuition and other college costs have
risen greatly during the past 30 years. So too have the many benefits from
college, including the greater earnings, health, and even marriage rates of
college graduates compared to high school graduates. Moreover, the return on a college
education has also increased, as measured by the higher benefits of college net
of the increase in college costs. As a result, college is even a better deal
than it was 30 years ago for most of the students who can afford the higher
Hence despite the sizable growth in college costs, more students
want a college education. This increased demand for college education is
obviously one factor that has pushed up college costs. The growth in college
costs means, however, that college may not be a better deal now than in the past
for capable high school graduates who have trouble financing higher college costs.
They and other students have turned to loans to ease the financial burden of
college, which, along with low interest rates, largely explains the sizable
growth in student loans. Many colleges, especially private ones, have also eased
the financial burden to students by increasing aid at a rapid rate.
Colleges are competing harder for better students, so that
the financial burden of going to college has grown more slowly for these
students than for average students. This, combined with the fact that the
benefits of a college education have also grown more rapidly for abler college
graduates, means that changes during past several decades in both benefits and
costs of college have been much better for abler students. This increased financial
cuddling of better students is another factor that has increased the cost of
college for the average student.
Increased competition for better students has gone hand in
hand with greater competition for better faculty. Just as returns to college
have increased more for better students, it is likely (there is little
quantitative evidence on this) that returns from college have also grown more
from an education with better faculty and a more effective education program.
In any case, the stronger competition for college faculty has been one
important source of the growth in college costs since faculty is by far the
dominant cost of a college education.
Although the fraction of young persons who go to college has
increased in all ethnic, race, and gender groups as a result of the higher
return from college, the increase has been rather modest, especially for males. Part of the explanation is that some students have been discouraged by
higher tuition and other college costs, although this does not explain why
young women are now far more likely to complete 4 years of college than are
To ease the burden of student loans, it would be wise to
introduce on a larger scale loans that tie repayments to subsequent earnings. Those
who earn relatively little would effectively have lower interest rates on their
loans. One obvious challenge in running such a loan program on a financially
sound basis is that these loans are more attractive to students who do not
expect to earn a lot, either because they have limited skills, or expect to
enter occupations that do not pay well.
Students cannot go to college if they have not graduated
from high school, and the fraction that drop out of high school, especially
young males, has remained stubbornly high during the past 30 years (with some decline
during past few years). Although the high dropout rate may be indirectly
related to the costs of a college education, it is puzzling since the plight of
high school dropouts has remained pretty dismal during the past several
decades. They have low earnings, high unemployment rates (as made clear during
the past recession), poor health, and bad marital prospects.
The best explanation for the large number of dropouts is
that parents, teachers, and schools have not prepared these students for a decent
high school education, let alone for college. This is being increasingly
recognized as pressure grows on schools and teachers, and also on parents, to
improve the education of the mainly minority students who constitute the great
majority of high school dropouts.
It is imperative to find better ways to help college students
cope with the large growth in tuition and other college costs. Yet, one should not
at the same time conclude that a college education has been oversold, since in
virtually all dimensions of life a 4-year college education generally pays off
In a little over a week the federal government will reach
its debt ceiling of almost $17 trillion unless Congress raises the ceiling. I
am 99% confident that the ceiling will be raised in time to avoid a crisis. It
is difficult to see any useful purpose served by these ceilings. Indeed, I
will show why a focus on debt ceilings may even be harmful.
Federal debt is rising toward the present ceiling because
the government has been running a deficit between its spending and tax revenue.
The government has run a deficit practically every year for many decades, so
that Congress has been forced to raise the ceiling over 90 times during the
past 70 years, and 15 times since 1993 alone.
The financial burden of the federal debt depends not only on
the size of the debt, but also on the interest paid on the debt. The product of
the level of the debt and the interest rate essentially determines the
government revenue needed to service the debt. The low interest rates on
federal debt (and other debt) during the past decade have made it easier to
service even a growing debt. The burden of the debt to the economy also depends
on the size of the debt relative to GDP. Federal debt held by the public, including debt held overseas but excluding debt held by other government agencies, is about 70% of American GDP.
This debt ratio grew rapidly during the past five years
partly because federal spending increased greatly for most of this period, and
also partly because GDP has been growing slowly because of the lingering
recession. Still, a 70% debt to GDP ratio is much lower than the ratio in some
other developed countries. In Japan, this ratio is over 150%, and the ratio is
also higher in some European countries.
Many conservatives have liked having a debt ceiling, even
though it has been raised many times, because they believe a ceiling helps
limit federal deficits. This is why conservative members of Congress in 2011
were able to abolish the “Gephardt Rule". This rule stipulated that when the
House adopted a budget, a separate debt-limit vote was not necessary since the
ceiling automatically increased by enough to accommodate the spending implied
by the budget. Yet since deficits have been common even before the Gephardt
Rule was adopted in 1979, it is not apparent that having a separate vote on the
ceiling reduced the size of the deficits.
Conservatives who have supported a debt ceiling to reduce
deficits are really usually mainly concerned about the size of government.
However, government size depends not on deficits, but rather
fundamentally on the level of government spending. Since deficits can be
reduced either by cutting spending or raising taxes, both liberals and
conservatives can agree on the value of reducing deficits while strongly
disagreeing on how to reduce them.
Liberals want to raise taxes to cut deficits, while conservatives want
to limit many kinds of government spending in order to reduce the size of the
To the extent that debt ceilings mainly induce tax increases
to slow the growth in debt, a focus on debt ceilings and deficits does not help
rein in the size of government. Moreover, the substantial growth in federal
spending during the past 50 years under both Democratic and Republican control
of Congress and the presidency strongly suggests that the many debt ceilings
during this period did little to reduce the size of government. The numerous
deficits over this period even suggest that the ceiling has accomplished
little, if anything, in reducing deficits.
For those worried about the growth of government, there is
no substitute for a focus on the scale of government spending. Having debt
ceilings may not be completely innocuous because they may detract from that
Both Democrats and Republicans are concerned with the growth of the federal debt relative to the growth in GDP. Consider an analogy to real estate. If one owns a house worth $100,000, borrowing $50,000 from a mortgage lender is unlikely to jeopardize one’s solvency. And if the value of the house increases to $200,000, borrowing another $50,000 would not be imprudent either. But suppose the value of the house remains unchanged, and still the borrower increases his borrowing from $50,000 to $100,000; his debt burden will have increased greatly. And likewise if a government increases its borrowing at a rate substantially in excess of the increase in government revenues as a consequence of economic growth or higher tax rates. It is burden, not amount, that is, or should be, the concern. And thus if interest rates are falling, the debt burden may not be increasing even if the amount of debt is increasing. But most federal debt is fairly short term nowadays, and thus the currently low interest rates are not locked in for long. If interest rates increase as the amount of debt increases (and those rates may increase for a separate reason—the greater the debt, which includes interest on the debt, relative to the borrower’s ability to repay, the greater the risk of default and so the higher the interest rate demanded by the lender), the overall debt burden will increase by more than the amount of borrowing increases.
The debt ceiling is not actually a ceiling on borrowing by the federal government. Rather, it is a limit on the amount of money that the Treasury can pay to meet its obligations, and thus to repay, or pay interest on, its borrowings. If the ceiling is low enough to prevent the Treasury from paying what it owes its creditors, the government is in default and interest rates on government debt are likely to soar.
The difference between our political parties is not that the Democrats want to “grow” the federal deficit faster than the economy is growing and that the Republicans want to shrink it. The Democrats want to reduce the deficit—everyone fears the impact of growing entitlement spending on the nation’s fiscal health—but mainly wants to do it, it seems, by increasing federal tax rates, while the Republicans want (or rather say they want) to reduce the deficit by reducing government spending. As Becker points out, if the method used to reduce the deficit is higher taxes, the debt ceiling has no effect. And as for reducing spending, that probably is politically infeasible, because the Republicans, though they say they want to reduce spending, don’t wish to touch entitlements (or at least social security and Medicare) or military or law enforcement expenditures; and there doesn’t seem to be much room as a practical and political matter for further economizing on other federal expenditures, which have already been reduced. Attractive reforms such as eliminating agricultural subsidies (importantly including the ethanol subsidy) do not seem to be politically feasible.
So a debt ceiling is unlikely to reduce the size of government. But it is pernicious, in inviting political tactics that could well be thought to violate the Constitution, or at least the spirit of the Constitution. Republicans want to destroy, or in the short run greatly weaken, Obama’s health care law (“Obamacare”), even though it resembles a health care reform proposed by President Nixon and successfully championed by Mitt Romney when he was governor of Massachusetts. The Republican preoccupation with Obamacare is thus rather surprising, but may reflect a fear that when once Obamacare is debugged and up and running it will prove popular, which will boost the Democrats. Yet the Republicans are not in a position to repeal or even amend the law by constitutionally authorized means, because repeal or amendment would require a majority vote in both houses of Congress (actually a two-thirds vote in both houses, for given a lesser majority Obama could veto a repeal or amendment without fear of being overridden). The intention, which is contrary to the structure of the federal legislative process ordained by the Constitution, is to coerce Congress to repeal (or by amendments to defang) Obamacare by threatening to precipitate an economic crisis by refusing to vote for an increase in the debt ceiling. If the tactic succeeded, it would mean that a minority in Congress had succeeded in amending a federal statute.