Confronting the Facts in American Race Relations
Gerald David Jaynes and Robin M. Williams, Jr.
Many white Americans believe the civil rights era of the 1960s removed all barriers to equal opportunity and that blacks are making steady gains towards equality.
The facts suggest otherwise. We recently led a team of scholars that carried out the most comprehensive review since World War II of the status of black Americans. After reviewing a wealth of data and research, we were persuaded that time alone is unlikely to resolve our nation's racial problems.
The data show that, after several decades of rising economic status relative to whites, blacks' gains have stagnated or declined on many measures since the early 1970s. Blacks remain substantially behind whites in health and life expectancy, educational and residential opportunities, and political participation. Throughout the 1980s, about two-fifths of all black children lived in poverty during any given year.
These findings have important implications for public policy. The large gaps between blacks and whites—which have remained constant or widened during recent years—discredit the view that racial equality can be achieved without planned government actions. To have no policy is, in fact, to have a negative policy in the area of race relations.
Reasonable people may agree that a policy is needed without agreeing on the policy itself. Much debate has focused on whether policies are needed most to change black people's
behavior or, instead, to expand their opportunities. Those who would change behavior often hypothesize that a special "culture of poverty" exists among poor blacks and must be altered if people are to begin living more productively. However, we found little evidence to support this hypothesis.
Does this mean that poor blacks do not engage in behaviors that help perpetuate their poverty? No—crime, teenage pregnancy and the like can be found in any inner city. However, there is little in the record to suggest that these behaviors occur independently of socioeconomic conditions—which can be changed with social policies.
Black-white cultural differences have narrowed, not widened, since 1960. Blacks respond to changes in American society much as whites do. Their poverty rates generally rise and fall with those of whites. Black employment rates, educational attainment and births to unwed mothers do likewise.
Furthermore, we found that much racial inequality continues to be due to discrimination and the segregation of poor blacks from quality schools, neighborhoods and other social institutions.
These findings lead to the clear conclusion that the main thing that needs enhancement is not black attitudes but black opportunities. Cultural values and beliefs are certainly important, but the fundamental problem of poor, black Americans—as of all poor Americans—is their lack of opportunity and their underlying poverty. A large body of research literature shows, for example, that employment
programs that improve poor people's skills have been more successful than those that seek to inculcate ''discipline'' among the poor by placing them in low-paying jobs devoid of meaningful training.
The record is also clear that previous programs by governments and private institutions have made a large difference in the opportunities and conditions of black Americans. These programs have included the Job Corps, Head Start, financial aid for college students and health services for young mothers.
Social policies with demonstrated benefits include the provision of education, health care and other services to enhance people's skills and productive capabilities. In addition, discrimination and involuntary segregation must be reduced. Better income maintenance and other family-assistance social welfare programs are needed to avoid long-term poverty. Above all, what black Americans need is full employment.
Certainly, anyone except an unrepentant segregationist can take pride in the gains our country has made since the days when blacks were forced to ride in the back of the bus. Over the past 50 years, the social status of black Americans has on average improved dramatically. Yet the many problems that remain make it clear that our national agenda in race relations remains unfinished.
Further progress requires public and private programs to enhance the productive capabilities and the opportunities of the poor. Our nation need not debate endlessly whether such efforts can make a positive difference. The empirical record shows that they can, and now is the time to put this knowledge into action.
September 17, 1989
Gerald David Jaynes of Yale University and Robin M. Williams, Jr., of Cornell University are the editors of the recent National Research Council report A Common Destiny: Blacks and American Society.
* * *
Homeless Children: An Emerging Tragedy
Ellen L. Bassuk
At least 100,000 American children will go to sleep tonight—or try to sleep—on sidewalks, in shelters or wherever they can. Each of these homeless children has a uniquely tragic story to tell, but the stories add up to a chaotic tale of Dickensian proportions.
I served recently on a committee of the Institute of Medicine of the National Academy of Sciences that found the growing number of homeless children to be "nothing short of a national disgrace."
Our report echoed what has become a growing body of research on homeless children that confirms some of our worst fears about their future. In a recent study I helped conduct in Massachusetts, a majority of school-age children from homeless families in shelters agreed with the statement: "I think about killing myself, but I would not."
Many of the preschoolers exhibited severe and multiple developmental lags. Compared with poor children who did have homes, these youngsters had serious problems with language development, gross motor skills, and social and personal development. For example, one 19-month-old boy we saw stopped eating and developed nightmares when his family moved into a shelter. Despite the obvious needs of such children, however, many of them receive few services.
As winter settles over the country the plight of homeless children becomes temporarily more visible, even though their year-round struggle is always desperate. Families are the fastest-growing segment of the homeless population, now making up about one-third of the total. Most of these families are headed by a woman with two or three children, a majority of whom are 5 years old or younger.
Emergency shelters and substandard welfare hotels are poor substitutes for the safety and security of a stable home. Small, crowded rooms with little privacy and, in many places,
no dining facilities are inadequate places to raise children. This is particularly true because so many mothers are understandably focused on daily survival and have little energy left to respond empathetically to a distressed child.
Many homeless children rarely go to school or attend only erratically. Transportation is frequently inaccessible or schools are located great distances from shelters. One homeless mother described to me how she had to ride public transportation for more than two hours every morning to take her two older children to schools located in different neighborhoods far from their shelter.
Her case was extreme, but a more common example is illustrated by a homeless mother of three children who had been living in an emergency shelter for six months. Over the past five years her family had moved eight times, doubling up in overcrowded apartments and various welfare hotels. Her three children attended school irregularly and all had repeated a grade. The oldest was failing two subjects and the middle one had memory and speech problems.
Stories like these are not isolated examples. The lives of homeless children are filled with fear, chaos, insecurity and loss. Because their daily activities are not routine or reassuring they do not have the emotional anchors many children take for granted.
What is so discouraging about the situation is that we know ways to ease or solve it for many children. The greatest needs, of course, are for decent, affordable housing and supports to help poor families maintain their homes. Also, various programs should be expanded so that they can be made available to homeless preschoolers. Infant stimulation, enriched day care and Head Start programs are among those that have been shown to make a measurable difference in the lives of severely disadvantaged young children.
Research has also shown that many of the developmental delays brought on by homelessness are reversible. Some of the damage can be undone by providing children with the homes and support they need. However, ignoring the problem today will only result in a higher price tomorrow. In other words, we can act as a society to help homeless children. Successful programs exist that can be expanded to
reach more of these youngsters who are at risk of becoming seriously anxious and depressed, failing at school and losing hope. That has happened to too many homeless children already. We must resolve that this new year will be different.
January 3, 1989
Ellen L. Bassuk is associate professor of psychiatry at the Harvard Medical School and president of the Better Homes Foundation.
* * *
The Gender Wage Gap
Robert T. Michael
Women in the United States earn only about 65–70 cents for every $1 earned by men. Is this due to bias against women—or to differences between women and men in their levels of education, work experience and other measurable factors?
That is the central question behind the continuing—and continually controversial—debate in our country about how to achieve "pay equity" and over whether "comparable worth" is a worthwhile strategy. Comparable worth seeks to eliminate gender bias from the labor market by employing techniques of job evaluation that are designed to use objective criteria to value the content and requirements of jobs, whether those of nurses, truck drivers, teachers or plumbers. The goal is to establish wages that reflect a job's true worth rather than historical salary patterns.
It is a debate subject to hyperbole on both sides, but the outcome is important because the concept could affect the paychecks of millions of Americans. The primary argument from proponents of comparable worth has been that it provides fairer pay for jobs held mostly by women. The main oppos-
ing argument has been that it imposes an arbitrary assessment of a job's relative value, disrupting the workings of the labor market and possibly causing inefficiency and inequity.
One reason the debate has been so contentious has been a lack of conclusive evidence about the heart of the matter, which is whether pay differences really are due to systematic bias against "women's jobs." Comparable worth is designed to address systematic undervaluation of jobs held by women. If, however, the problem is that women are prevented from entering male-dominated jobs, then the proper remedy may be equal-opportunity legislation. Of course, both problems can occur simultaneously.
A recent series of studies on pay equity from the National Research Council helped clarify the situation. It showed that the wage gap is indisputable. The studies confirmed that women receive lower pay than men across a wide range of occupations and employers. A discrepancy remains even after accounting for education, job experience and other measurable differences between men and women.
Whether this wage gap is attributable to systematic bias against women or to other market forces still poorly understood or measured, however, remains unclear. A number of studies have found that wage differences by gender are diminishing, although gradually. It also appears that although jobs held mostly by women do receive lower wages than those held mostly by men, women generally are not paid less for the same job.
One study found that starting salaries for specific civil service positions fell when women or minorities entered those jobs in large numbers, even though the job requirements remained the same. This suggests the jobs were devalued because women and minorities held them, as comparable-worth advocates have argued. Yet, even here, alternative explanations are possible.
Prescribing the medicine of comparable worth in the face of such an inexact diagnosis is a political question rather than a research subject. Yet the recent experience of several states and localities suggests that doing so is less traumatic than some have predicted.
A comparable-worth plan introduced in Iowa's state em-
ployment system in 1985 cut that state's gender wage gap in half, from 6 percent to 3 percent. Minnesota appears to have had success with similar legislation. In Australia, the relative wages of women were raised substantially with no evidence of adverse employment effects or disruptive economic effects.
A number of studies suggest that pay equity plans designed for specific employers in the United States can, at most, correct about 20 percent of the observed gender wage gap. They also have little power to address differences across firms or industries, and it is unclear whether they will affect worker productivity or overall unemployment. Yet they do provide a substantial increase in salaries for women workers and appear not to cause negative side effects, at least in the short run.
It is essential that society debate comparable worth on the basis of all available facts and that it find empirical answers for the remaining questions. One of the fortunate aspects of the controversy is that it lends itself to objective analysis. A just and efficient wage structure for working Americans can—and should—be based on an understanding of the complex reality of what happens in the workplace and in the job market.
October 22, 1989
Robert T. Michael, director of the National Opinion Research Center and professor of education and public policy at the University of Chicago, chaired a National Research Council panel on pay equity research.
* * *
Child Care in Disarray
John L. Palmer
It is late afternoon and two women are picking up their 4-year-old children. The first woman, a married attorney, drives to an attractive child care center to get her child. The second woman, a single mother who works as a waitress, takes the bus to a friend's one-bedroom apartment, where her child spends the day with five other neighborhood children.
If you were in a position to assist one of these two women, whom would you choose?
Uncle Sam now favors the attorney. She claims a tax credit on her income tax and has a Flexible Spending Account with her firm that lets her pay the child care center with tax-free dollars. The waitress, who earns the minimum wage, is too poor to use either benefit. Like four of every five low-income children, her child cannot get into a Head Start program.
Of course, as any working parent knows, the balance sheet on child care is not defined by money alone. There also is the emotional cost of sending a child to school with a fever or of missing the child's class play because of a business meeting.
But, as a panel I chaired for the National Research Council reported recently, the federal government has been providing less and less support to poor, working parents, who need help most. In 1972, 80 percent of federal child care support was targeted to low-income families through subsidies to child care centers and other providers. By last year, these programs accounted for less than 30 percent of the total as federal support had shifted to direct consumer subsidies primarily benefiting middle- and upper-income families.
Our panel did not take a position on the child care bills now before Congress. But we did urge that special attention be paid to the millions of poor, working parents who lack the resources to place their children in safe, healthy and enriching settings.
Children from poor families are especially at risk, but they are hardly alone. Most children in the United States now have working mothers; all too often, quality child care is unavailable. Many families find the care available from neighbors or local centers to be too costly—or simply inadequate. Parents who work non-traditional hours, or those whose children have medical problems, often have special difficulty arranging child care services.
Not that white-collar professionals have it easy. They may worry less about paying the bills, but they also find child care to be a source of stress. Organized programs for infants and toddlers are scarce in many neighborhoods, and an estimated 2.1 million youngsters are "latchkey kids."
A growing body of research confirms that quality care is essential to a young child's healthy development, but our national child care system remains fragmented and inadequate. Services have developed without any rational framework of legislation or policy. State regulations vary widely and often ignore recognized standards of quality. Some states allow a single person to care for as many as 12 infants.
Legislative strategies differ, but several steps clearly are required to improve the quality, affordability and accessibility of child care. The need to assist poorer families and to expand compensatory programs such as Head Start is acute. National standards for the quality of all child care should be set. In addition, most parents should have their jobs protected for up to one year if they choose to stay home and care for a newborn. Research indicates it is important that new families get off to a good start. Other countries have successfully implemented parental leave policies.
It would be foolish to legislate a single kind of child care for all children. One good thing about the current system is its diversity. But parents should be able to choose among options of high quality, including providing care themselves, rather than lurching from crisis to crisis.
The price tag for quality care will not be low. For example, our panel recommended that government at all levels provide an additional $5 billion to $10 billion annually just
as a first step. But attorneys, waitresses and millions of other parents can attest that our children deserve better, and the future of our country depends upon it.
April 1, 1990
John L. Palmer is dean of the Maxwell School of Citizenship and Public Affairs at Syracuse University.
* * *
Effective Drug Treatment
Lawrence S. Lewin and Dean R. Gerstein
Despite the deficit, President Bush and Congress are proposing large increases in funding for drug treatment programs. Why? Because publicly supported drug treatment is the neglected front in the national drug war.
More than 2 percent of the U.S. population over age 12, or 5.5 million people, clearly or probably need drug treatment. Fewer than one in six now get it. There are 66,000 people on waiting lists for openings at public drug treatment programs across the country; many more don't apply. An estimated 105,000 pregnant women have serious drug problems, as do a million people under the supervision of the criminal justice system.
These numbers are growing. Although surveys indicate that casual drug use is decreasing, severe drug abuse and dependence are becoming more prevalent, the menu of illegal drugs is expanding, and the AIDS epidemic makes treatment even more complex. Yet, until recently, drug treatment was regarded too lightly by most officials. Public funds were aimed at drug busts and other law enforcement efforts. Only about one-seventh of the $3.9 billion allotted in 1989 for the federal anti-drug campaign went for drug treatment.
But would increased spending on drug treatment really lead significantly more people to end or even reduce their dependence on illegal drugs? If done right, the answer is yes.
We led a committee of the Institute of Medicine of the National Academy of Sciences that recently studied drug treatment programs, and we found that the better ones not only can get addicts off the street but also help change their lives. The benefits to society in reduced crime, improved productivity and the like exceed the costs. Expanding treatment is both humane and cost effective.
Not every drug user needs formal treatment. Many experimental or casual users just lose interest in drugs or respond to preventive counseling or disciplinary sanctions. Drug treatment is justified and appropriate when an individual shows clinically significant signs of dependence or chronic abuse.
Not all treatment programs are the same; quite the contrary. Our country's treatment delivery system is sharply divided between public and private tiers. Individuals with private insurance generally enter hospital-based programs for inpatient detoxification and treatment. Those whose treatment is paid from public funds are more likely to make weekly visits to a clinic or to enter a spartan residential community, with a daily program of work duties and intensive group therapy. Public programs that maintain addicts on methadone also are common.
Surprisingly, the costs of these different approaches are not based on their effectiveness, although there is a direct relationship between the length of time spent in treatment and long-term effectiveness. Although objective data are sparse, largely because research and evaluation programs were slashed several years ago, there is good evidence that methadone maintenance for heroin dependence and residential "therapeutic communities" are cost effective. The evidence is scantier for outpatient non-methadone treatment, but the much lower costs make even modest benefits worthwhile.
Other treatment approaches, such as self-help groups and prison-based "boot camps," have not been evaluated in sufficient detail. Evidence of effectiveness also is generally slim
for the programs offered by private providers. There is no evidence that expensive hospital-based chemical dependency programs are more effective than similar programs not sited in hospitals. Private payers should insist on reliable evidence of clinical outcomes and not waste money on unnecessary services or overhead charges.
One big concern about drug treatment programs has been whether they have any impact on patients who enter treatment in the community to avoid prison. An encouraging finding is that credible criminal justice pressures do not diminish the effectiveness of treatment. These programs can do a much better job than imprisonment of reducing subsequent criminal behavior.
Effectiveness will not always mean total abstinence. Drug dependence is characteristically a chronic, relapsing disorder, and more than one episode of treatment typically is required. The better the quality of programs, staff and services, the more likely that an episode of treatment will have a lasting, powerful effect. Money spent to improve and expand treatment is money well spent—not only for those treated, but for everyone whose lives are affected by them.
October 21, 1990
Lawrence S. Lewin, president of Lewin/ICF, a consulting firm in Washington, D.C., chaired a committee of the Institute of Medicine that studied drug treatment programs. Dean R. Gerstein, now with the National Opinion Research Center, was the committee's study director.
* * *
Making Sense of Statistics in the Courtroom
Stephen E. Fienberg and Miron L. Straf
Some of society's most important court cases are decided on the basis of statistics, often because no substantial direct evidence is available. Yet, too often, statistical analyses are confusing to non-experts. Juries and judges struggle to decide whether to believe the statistician for the plaintiff or for the defense, grappling with such concepts as regression analyses and statistical significance.
One way to ease their task would be to set better guidelines for people who provide the analyses and inferences, mostly statisticians, economists and other social scientists. Many experts now called on to testify in court are caught in a dilemma: They are almost always hired by attorneys for a particular side who want them to testify in a certain way, yet they are expected to maintain a professional commitment to a complete, accurate and balanced inquiry that will guide the court to a wise decision.
This puts the squeeze on experts whose testimony may be crucial to the outcome. Statistical evidence has been central to such celebrated disputes as whether Agent Orange harmed Vietnam veterans and whether housing discrimination in some cities was so pervasive that busing was needed to desegregate the schools. It also has been an essential part of such important cases as these:
A federal appellate court relied on statistics in overturning an order of the Consumer Product Safety Commission (CPSC) that banned the use of ureaformaldehyde foam insulation in residences and schools. The court said the CPSC used data incorrectly and produced an inaccurate model of the foam's cancer risk to humans.
A nurse in Maryland was acquitted of murder after a jury rejected the argument that she fatally injected several
patients in a hospital. The state had relied exclusively on statistical models in charging that it was almost impossible for so many patients to die while the nurse was on duty unless she had a hand in their deaths.
Published opinions in federal courts suggest that this use of statistical evidence is expanding dramatically. Statistical methods and analyses are seen as invaluable by many lawyers who use them to present succinct summaries of complex data, to provide a reliable basis for predictions, to prepare quantitative estimates of damages and to clarify complex relations.
Yet this trend also has its costs—and not only the direct costs of hiring expert witnesses or having longer trials. There also is the danger that an increasing dependence on complex statistical procedures could lead some Americans to feel that justice is being turned over to technocrats.
As the use of statistics in judicial proceedings grows, the proper role and conduct of statistical experts need to be clarified. A committee of the National Research Council, which we helped lead, recently suggested some guidelines.
Most simply, we said expert witnesses should never present opinions they do not believe. If they state an alternative viewpoint hypothetically, they should make clear what they are doing. Also, even though a court may not require them to volunteer information, statistical expert witnesses should do so anyway if they believe it is necessary to maintain honesty and accuracy.
Statisticians testifying as experts should be free to carry out their analyses in a scientific manner with access to the necessary data and their colleagues. And if they work under a contingency-fee arrangement in which they are paid only if their side wins the case, this fact should be a matter of court record.
Judges, meanwhile, could make better use of statistics through greater use of court-appointed statistical experts.
Establishing these and similar guidelines would be of great benefit to all of us who might one day find ourselves in-
volved in a lawsuit with expert testimony. As our courts depend on statistics more and more, it is essential that we have confidence in the experts who stand behind them.
February 14, 1989
Stephen E. Fienberg, dean of humanities and social sciences at Carnegie Mellon University, chaired a committee of the National Research Council that studied the role of statistical evidence in the courts. Miron L. Straf, director of the Research Council's Committee on National Statistics, was the study director.
* * *
The Economics of Reality
Richard H. Thaler
Human beings make mistakes. While this statement may not seem controversial, it flies in the face of much of traditional economic theory. The theory does not match the reality of human behavior, and that is a dilemma at a time when our nation faces economic problems ranging from the federal budget deficit to the trade gap.
Regardless of their political persuasion, most economists build their models on the assumption that investors, managers, workers and consumers are highly rational. To model complex situations, economists assume that people act as if they are able to make economic forecasts with the same sophistication as a computer running an econometric program. This premise of economic rationality has proved successful in some applications in the past, but it is increasingly belied by experimental evidence from those who study the psychology of decision-making, and it is inconsistent with the way economists talk about the behavior of their spouses, colleagues and government leaders.
As a recent National Research Council report on the social and behavioral sciences discussed, people are best thought of as having ''bounded rationality,'' to use the term coined by Nobel laureate Herbert Simon. Humans can only process limited amounts of information, and their memories are unreliable. To cope with their limitations, people adopt simple rules of thumb to solve complex problems. These rules work well in most situations—but not always.
Suppose, for example, you were asked to guess the ratio of homicides to suicides in the United States. If you are like most people, you would judge homicides as more frequent. In fact, there are more suicides. Psychologists Daniel Kahneman and Amos Tversky use this example to illustrate that people often judge frequency by ease of recall. When some events are more heavily publicized than others, they are thought to be more frequent, even if they are not.
Other examples come from news events:
In the spring of 1986, there was a flurry of terrorist activity in Europe, and several travelers were killed by a bomb placed in the Rome airport. By summer, millions of Americans had canceled their plans for European vacations and took driving vacations in North America instead. Yet the risk of being the victim of a terrorist bomb in Europe was almost certainly lower than the risk of driving on American highways.
After his poor performance in the Iowa caucuses, many reporters declared George Bush's presidential candidacy dead. Oops.
Both of these are examples of overreaction. In making predictions, people often give too much weight to recent events and too little to long-run tendencies.
The stock market crash last fall provides another example of people making conclusions and acting in ways that defy standard economic models. On Monday, October 19, the Standard and Poor's stock index fell 20 percent. This fall in stock prices, the largest percentage drop in New York Stock Exchange history, followed a fall of 5 percent the previous Friday. A two-day price drop of this magnitude would be
predicted by statistical models only once in every trillion years or so.
Following the Monday crash, prices rose 5 percent on Tuesday and 9 percent on Wednesday, the exchange's two largest postwar increases. The following Monday, the market fell 8 percent. It is hard to imagine a model in which such enormous daily price changes could be judged rational in the absence of any important economic news. Rather, investors seemed to be reacting emotionally to the price movements themselves.
Recent research on the stock market has also uncovered numerous "anomalies," facts which are difficult to explain within a totally rational framework. For example, stock prices tend to rise on Fridays and before holidays and fall on Mondays. It seems possible that this pattern is produced by the mood of market participants. Indeed, contrary to most Fridays, prices actually tend to fall on Fridays that occur on the 13th day of the month!
These and other examples all illustrate the basic point that people are human, not economic robots. Economists and the people who use their models—public officials, business leaders and others—must begin to acknowledge and take account of this human side of economic decision-making. A growing body of research is becoming available on how the vagaries of human nature affect economic decision-making, and it needs to be applied in making business decisions, passing laws and determining economic policy.
While traditional economic models have made numerous important contributions to our economic understanding, they can be enriched by incorporating human elements of behavior. Stubbornly clinging to them as they now exist is itself irrational.
October 11, 1988
Richard H. Thaler, the H.J. Louis Professor of Economics at Cornell University's Johnson Graduate School of Management, served on a National Research Council team that studied recent developments in the behavioral and social sciences.
* * *