In the previous chapter, we saw that support for segregation has nearly disappeared, and that Americans’ neighborhoods, schools, workplaces and families are much less segregated than they were. But the end of segregation is not the same as equal opportunity.
In this chapter, we’ll look at how Black Americans are represented in the workplace and in government, and how their employment, income and housing compare with those of other Americans. Then we’ll look at what Americans think explains the gaps that remain -- discrimination, lack of opportunity, or the choices people make -- and what, if anything, the government should do about them.
As in the previous chapter, most of the figures show the responses of non-Black Americans, and many show Black Americans’ responses as well.[1]
Representation in the workplace has improved¶
The following figure shows the percentage of Black employees at companies that report to the EEOC[2], overall and in selected roles. It also shows the percentage of CEOs at Fortune 500 companies who are Black, based on data published by Fortune. The shaded area shows the Black share of the population, based on Census data.
Figure 1:corporate_leadership_black — Percent Black among all employees, professionals, and managers at companies that report to the EEOC, and among the CEOs of Fortune 500 companies (purple; Black CEOs serving on June 1 of each year, out of 500). Through 2013, “Officials and Managers” is one category (gray line); from 2014 it is split into mid-level and senior managers. Shaded gray: the Black share of the U.S. population (Census; Black alone or in combination from 2000; each census holds until the next).
In 1966, 8% of employees at EEOC-reporting companies were Black, so they were moderately underrepresented; in 2023 15% were Black -- slightly higher than proportional representation. Among professionals (jobs that usually require a college degree or more), the percentage rose from 1% to 10% between 1966 and 2023.
Among “Officials and Managers,” it rose from 1% to 7% between 1966 and 2013. After that, the category was split into two: mid-level managers, where representation reached 9% in 2023, and senior managers, where representation reached 4%.
Since 1999, when Franklin Raines of Fannie Mae became the first[3] Black CEO of a company on the Fortune 500 list, the number of Black CEOs on the list has increased to 11, about 2%. So Blacks are still underrepresented in corporate leadership, especially among senior managers and CEOs.
Discrimination and affirmative action¶
About discrimination in the workplace, the GSS asks:
Do you feel in any way discriminated against on your job because of your race or ethnic origin?
Figure 2:wkracism — Time series by race and ethnicity: percent of workers saying yes. Hispanic is based on primary reported ancestry.
Among White workers, the percentage saying yes is small and nearly constant at 2–3%. Among Black workers, it fell from 18% to 10% between 2002 and 2022, and among Hispanic workers, from 14% to 8%, though the estimates are uncertain for both groups. So the experience of workplace discrimination seems to be declining -- but still higher for non-White workers.
Since 1986, the ANES has asked this question about preferential hiring:
Some people say that because of past discrimination Blacks should be given preference in hiring and promotion. Others say that such preference in hiring and promotion of Blacks is wrong because it gives Blacks advantages they haven’t earned. What about your opinion -- are you for or against preferential hiring and promotion of Blacks?
The following figure shows the percentage who favor affirmative action, for non-Black and Black respondents.
Figure 3:VCF0867a — Time series by race of respondent: percent in favor of preferential hiring and promotion of Black people, ANES, non-Black and Black respondents.
Among non-Black respondents, support is low, but it has increased from 12% in the early 2000s to 30% in 2024. Among Black respondents, a majority supported affirmative action throughout: 67% in 1986, 58% in the mid-2000s, and 72% in 2024.
On a related topics, the GSS asks whether people think White workers lose out to Black workers:
What do you think the chances are these days that a White person won’t get a job or promotion while an equally or less qualified Black person gets one instead? Is this very likely, somewhat likely, or not very likely to happen these days?
The following figure shows the percentage who say very or somewhat likely, for Black and non-Black respondents.
Figure 4:discaff — Time series by race of respondent: percent saying very or somewhat likely, non-Black and Black respondents.
Among non-Black respondents, the percentage fell from 70% to 57% between 1990 and 2024. Among Black respondents, it rose from 34% in 1990 to 51% around 2008, then fell to 37% in 2024.
So belief that there is widespread discrimination against White people is declining.
Equal representation in some government branches¶
Next we’ll consider progress toward equal representation in government.
The following figure shows the number of Black members of the House and Senate in each Congress since 1869, based on records from the House Historian.
Figure 5:black_members_of_congress — Number of Black members of the House and Senate in each Congress, 1869–2025: the most serving at the same time, not counting non-voting Delegates (House Historian).
The first Black members of Congress took their seats in 1870, during the Reconstruction era: Hiram Revels of Mississippi in the Senate, and Joseph Rainey of South Carolina and Jefferson Long of Georgia in the House. The number in the House reached 7 in 1873. Blanche Bruce of Mississippi was elected to the Senate in 1874 and served until 1881.
But after Reconstruction the number fell. George White of North Carolina left the House in 1901; after that, there were no Black members of Congress until Oscar De Priest of Illinois in 1929. There was no Black senator after 1881 until Edward Brooke of Massachusetts, who took office in 1967.
The end of Reconstruction, and the following decades of disenfranchisement of Black voters, are a reminder that positive trends are not guaranteed to continue. Some go into reverse -- and take a long time to get going again.
In the House, the number of Black representatives grew from 1 in 1929 to 6 in 1965, 16 in 1975, 26 in 1991, and then 38 in 1993. After the 1990 census, states redrew congressional districts, and under the Voting Rights Act many drew those districts with Black majorities; in the 1992 election, Black candidates won most of the new seats. In the current 119th Congress, there are 62 Black members of the House and 5 Black senators.
The next figure shows the percentage of seats held by Blacks in the House, Senate, Supreme Court, and Cabinet since 1917. The shaded area shows the Black share of the population.
Figure 6:black_representation_in_government — Percent of seats held by Black Americans in the House and Senate (House Historian), on the Supreme Court, and in the Cabinet, 1917–2025, and, shaded gray, the Black share of the U.S. population (Census; Black alone or in combination from 2000; each census holds until the next). Cabinet: the most Black members serving at the same time in each administration, as a percent of cabinet and cabinet-level posts, counted as in the gender chapter’s figure (posts from CAWP).
The Black share of House seats was below 1% until 1963, passed 5% in 1987, and was over 14% in 2025, about the same as the share of the population. In the Senate, it has never been more than 5%.
Thurgood Marshall became the first Black justice of the Supreme Court in 1967; Clarence Thomas succeeded him in 1991, and with Ketanji Brown Jackson, who joined the Court in 2022, two of the nine justices are Black.
Robert Weaver became the first Black member of the Cabinet in 1966, as Secretary of Housing and Urban Development. Until 1993, there were never more than two Black members of the Cabinet at the same time. The highest percentages of cabinet-level posts were 23% in Clinton’s first term, 17% in both of Obama’s terms, and 28% under Biden. There was only one in each of Trump’s terms, both Secretaries of Housing and Urban Development.
In summary, Black Americans are proportionately represented now in the House of Representatives and Supreme Court. The composition of the President’s Cabinet varies between administrations, but in several recent cabinets Blacks have actually been overrepresented. Representation in the Senate lags far behind.
People would vote for a Black president¶
Now let’s look at two survey questions related to Black representation in government: do people think Blacks have enough influence, and would they vote for a Black presidential candidate.
Since 2008, the ANES has asked:
Would you say that Blacks have too much influence in American politics, just about the right amount [...], or too little influence [...]?
The following figure shows the percentage who said too little, for non-Black and Black respondents.
Figure 7:VCF9275 — Time series: percent saying Black people have too little influence in American politics, ANES 2008–2024. The 2024 point is from the 2024 Time Series study, not yet in the cumulative file.
Among non-Black respondents, the percentage rose from 30% to 49% between 2008 and 2024. Among Black respondents, it was 71–76% throughout. So the racial gap on this question has narrowed, but it is still wide.
From 1972 to 2010, the GSS asked:
If your party nominated an African-American for President, would you vote for him if he were qualified for the job?
Gallup has asked a similar question since 1958. The following figure shows the percentage who said they would vote for a Black nominee.
Figure 8:racpres — Time series: percent who would vote for a qualified Black presidential nominee. Line: GSS, 1972–2010. Circles: Gallup, 1958–2024.
In the GSS, the percentage who would vote for a Black candidate rose from 79% to 96% between 1972 and 2010, when the question was dropped. In Gallup surveys, it rose from 38% in 1958 to 77% in 1978 and 93% in 1997; in 2024 it was still around 92%. Among Black respondents, Gallup found nearly unanimous support from the start: between 92% and 100% from 1958 to 1971.
Most of the increase in the GSS is due to changed minds. In 38 years, support increased by 16 percentage points, of which 5 are due to generational replacement and 11 due to changed minds.
In the chapter on gender, we discussed the possibility that social desirability bias affects these results. People say they would vote for a Black nominee, but would they?
It’s hard to be sure, but we have a few data points to work with. The Democratic party has nominated a Black candidate three times; from ANES data, we can get the percentage of Democrats who voted and the percentage of those who voted for the nominee. The following table shows the results.
| Year | Democratic nominee | Voted | Voted for the nominee |
|---|---|---|---|
| 2000 | Al Gore | 74% | 86% |
| 2004 | John Kerry | 76% | 89% |
| 2008 | Barack Obama | 78% | 90% |
| 2012 | Barack Obama | 81% | 91% |
| 2016 | Hillary Clinton | 80% | 87% |
| 2020 | Joe Biden | 87% | 95% |
| 2024 | Kamala Harris | 84% | 95% |
In 2008 and 2012, the turnout for Barack Obama was comparable to other recent elections, and his share of Democratic votes was about average. For Kamala Harris, turnout and vote share are higher than average. So there is little evidence here that Democrats, at least, are unwilling to vote for a Black candidate.
Differences in outcomes¶
In the next section we’ll look at several questions that ask about reasons for racial differences in outcomes. In the GSS, these questions start with a preamble:
On the average African-Americans have worse jobs, income, and housing than White people. Do you think these differences are …
Before we get to the responses, let’s look at the facts respondents were asked to explain. Is it true that there are differences in jobs, income, and housing, and how have those differences changed over time? The answer to the first questions is yes; the answer to the second is that the gaps have narrowed but not closed.
The following figure shows unemployment rates[4] for three groups from 1976 to 2025.
Figure 9:Unemployment rate, annual average, people 16 and older. White and Black are non-Hispanic; Hispanic is any race. Computed from monthly CPS microdata (IPUMS CPS).
Unemployment rates are consistently highest for non-Hispanic Blacks, lower for Hispanics, and lowest for non-Hispanic Whites. But over this period, unemployment declined in all three groups, and the Black-White gap closed from more than 7 percentage points to less than 4.
The following figure shows median household income between 1970 and 2025, in inflation-adjusted 2025 dollars.
Figure 10:Median household income, in thousands of 2025 dollars, by race and Hispanic origin of the householder. White and Black are non-Hispanic; Hispanic is any race. Computed from CPS Annual Social and Economic Supplement microdata (IPUMS CPS).
Incomes have increased in all three groups, with most of the gain since 1990. But the gaps have also increased:
The gap between White and Black households was $26,800 in 1970 and $36,500 in 2025
The gap between White and Hispanic households was $16,600 in 1970 and $23,500 in 2025.
The next two figures show measures of housing quality[5]. The first is the percentage of people living in housing classified by the Census Bureau as moderately or severely inadequate.
Figure 11:Inadequate housing: percent of people living in units the Census Bureau classifies as moderately or severely inadequate, based on plumbing, heating, electrical, kitchen, and upkeep problems. Comparable from 1985. Race and Hispanic origin of the householder; White and Black are non-Hispanic; Hispanic is any race.
The percentages decreased in all three groups, and the gaps have narrowed. In 1985, about 23% of Blacks, 19% of Hispanics, and 6% of Whites lived in inadequate housing; in 2023, those percentages were 8%, 6%, and 4%. So the Black-White gap closed from about 17 percentage points to 4.
The following figure shows the percent of people living in crowded housing, defined as more than one person per room (counting living rooms, dining rooms, and kitchens as well as bedrooms, but not bathrooms).
Figure 12:Crowding: percent of people living in homes with more than one person per room. Race and Hispanic origin of the householder; White and Black are non-Hispanic; Hispanic is any race.
Again, the percentages have come down in all three groups, and the gaps have narrowed but not closed. In 1973, 35% of Hispanics, 28% of Blacks, and 9% of Whites lived in crowded housing; in 2023, those percentages were 13%, 5%, and 2.3%.
In summary, the premise of the GSS questions is true: Black people have higher unemployment, lower income, and worse housing than White people, on average. Now let’s see what people believe about the causes of these differences.
Explanations of racial gaps¶
The GSS and ANES ask several questions about racial differences; I’ve put them in three categories: circumstances, attributes, and behavior.
Starting with circumstances, the GSS includes two questions that ask whether differences in outcome are:
… because most African-Americans don’t have the chance for education that it takes to rise out of poverty?
… mainly due to discrimination?
The ANES asks if respondents agree that:
Generations of slavery and discrimination have created conditions that make it difficult for Blacks to work their way out of the lower class.
The following figure shows the percentage agreeing with each explanation, for Black respondents (solid lines) and non-Black respondents (dashed lines).
Figure 13:racdif3, racdif1, VCF9039 — Time series by race of respondent: percent agreeing that racial gaps are due to circumstances, numbered as in the text. Solid lines: Black respondents; dashed lines: non-Black respondents (GSS, ANES).
Black respondents are more likely to endorse explanations based on circumstances, and the gap has been consistent over time.
In both groups, support for these explanations declined before 2010 and have increased since. These increases are driven by generational differences; young adults are more likely than previous cohorts to endorse these circumstantial explanations.
Another set of questions asks about supposed racial attributes.
Both the ANES and GSS ask respondents to rate each group on a scale from “lazy” to “hard-working”. In the following figure the lines labeled (1) and (2) show:
ANES, percent rating Black people lazier than White people.
GSS, percent rating Black people lazier than White people.
The GSS also asks about intelligence; the line labeled (3) shows:
GSS, percent rating Black people less intelligent than White people.
Finally, the GSS asks whether different outcomes are
… because most African-Americans have less in-born ability to learn?
The line labeled (4) shows the percent who say yes.
The following figure shows the results for Black respondents (solid lines) and non-Black respondents (dashed lines).
Figure 14:VCF9271_rel, workblk_rel, intlblk_rel, racdif2 — Time series by race of respondent: percent attributing racial gaps to attributes of Black people, numbered as in the text. Solid lines: Black respondents; dashed lines: non-Black respondents (GSS, ANES).
Over time, non-Black respondents are less likely to say that racial differences explain differences in outcomes. The changes are substantial: 25 percentage points for the first question, 37 for the second, 33 for the third, and 21 for the last.
These changes are driven by a combination of changing minds and generational replacement. The current generation of young adults is less likely to say that racials differences explain gaps in outcomes.
Black respondents are generally less likely to endorse racial differences: in the most recent surveys, all four questions are below 20%.
The gap between Black and non-Black respondents has generally narrowed; for the third question it has closed. But for the fourth question, surprisingly, it has reversed: Black respondents are now more likely than non-Blacks to attribute outcomes to “in-born ability to learn”. The most recent surveys have small samples of Black respondents, so the size of the gap is uncertain, but it is larger than we would expect by chance.
Finally we’ll look at questions that ask whether racial gaps are due to behavior -- that is, what Black people do or should do. The GSS and ANES ask if people agree with this statement
Irish, Italians, Jewish and many other minorities overcame prejudice and worked their way up. Black people should do the same without special favors.
The following figure shows the percent that agree in the GSS (1) and the ANES (2), for Black respondents (solid lines) and non-Black respondents (dashed lines).
The line labeled (3) shows the percent in the ANES that agree:
It’s really a matter of some people not trying hard enough; if Blacks would only try harder they could be just as well off as Whites.
And the line labeled (4) shows the percent in the GSS who say that racial gaps are
… because most African-Americans just don’t have the motivation or will power to pull themselves up out of poverty?
Figure 15:wrkwayup, VCF9040, VCF9041, racdif4 — Time series by race of respondent: percent agreeing that racial gaps are due to the actions of Black people, numbered as in the text. Solid lines: Black respondents; dashed lines: non-Black respondents (GSS, ANES).
Among non-Black respondents, agreement with all four statements declined, especially after 2010. The total decline is 30 points for the first question, 26 for the second, 29 for the third, and 37 for the last.
The generational changes are consistent: each generation is less likely than the previous to endorse the idea that Black people could work their way up if they tried harder.
A substantial minority of Black respondents agreed with these statements, and agreement increased until about 2008; since then it has declined. The net declines are 16 points for the first question, 23 for the second, and 18 for the third; for the fourth question the net decline is only 2 percentage points.
For question (2), the gap between Black and non-Black respondents is almost unchanged. For questions (1) and (3), it has narrowed by about half. And for question (4), it has closed and possibly crossed over.
In summary:
To explain racial gaps in outcomes, Black respondents are more likely than non-Blacks to point to circumstances rather than racial differences -- but some of those gaps have closed.
In both groups, since about 2010, more people point to circumstances and fewer to lack of effort.
Among non-Black respondents, recent cohorts are also more likely to endorse circumstantial explanations over traits and behaviors.
Government programs and spending¶
Next we’ll look at questions related to policy -- whether and how government should help Black people. The ANES asked two questions about government action to ensure integration and fairness:
ANES, 1962–2000: Some people say that the government in Washington should see to it that White and Black children go to the same schools. Others claim this is not the government’s business.
ANES, 1964–2012: Should the government in Washington see to it that Black people get fair treatment in jobs or is this not the federal government’s business?
The following figure shows the percent in favor, for Black respondents (solid lines) and non-Black respondents (dashed lines).
Figure 16:VCF0816, VCF9037 — Time series by race of respondent: percent favoring government action to ensure equal treatment, numbered as in the text. Solid lines: Black respondents; dashed lines: non-Black respondents (ANES).
Among non-Black respondents, support for integrated schools (1) was 54% in 1962, fell to 37% during the years of court-ordered busing, then increased to 49% in 2000, when the question was last asked. Support for ensuring fair treatment in the workplace (2) rose from 43% in 1964 to 50% in 1990, then fell to 41% in 2012.
Among Black respondents, support for both was much higher, but it declined somewhat: from 88% to 79% for schools (1), and from 93% to 80% for jobs (2).
For both questions, the gap between the groups is mostly unchanged.
The ANES and GSS have also asked two questions about special efforts to help Black people:
ANES, 1970–2024: Some people feel that the government in Washington should make every effort to improve the social and economic position of Blacks. … Others feel that the government should not make any special effort to help Blacks because they should help themselves.
GSS, 1975–2024: Some people think that Black people have been discriminated against for so long that the government has a special obligation to help improve their living standards. Others believe that the government should not be giving special treatment to Black people.
The following figure shows the percent favoring government effort to help Black people.
Figure 17:VCF0830, helpblk — Time series by race of respondent: percent favoring government help for Black people, numbered as in the text. Solid lines: Black respondents; dashed lines: non-Black respondents (ANES, GSS).
In both groups, support for special effort is lower than for equal treatment.
Among Black respondents, support for special effort fell initially, but increased more recently. Among non-Black respondents, the changes are similar but smaller.
The gap between Black and non-Black respondents is smaller now than in the past, and since 2010 the trends are roughly parallel.
Looking at generational differences, recent cohorts are more likely to support government action -- but only among non-Black respondents. Among Black respondents, the cohort component of the change is flat or possibly falling.
If recent generations are more likely to think the government should help Black people, are they willing to allocate tax money to do it? The GSS includes a battery of questions about government spending, starting with this preamble:
We are faced with many problems in this country, none of which can be solved easily or inexpensively. … Are we spending too much, too little, or about the right amount on …
Then they ask about:
… improving the conditions of Black people?
… assistance to Black people?
Similarly, the ANES asks whether federal spending should be increased on
programs that assist Blacks?
Questions 1 and 2 are a pair; since 1984, each respondent gets one wording or the other. The following figure shows the percent who think the government spends too little, for Black respondents (solid lines) and non-Black respondents (dashed lines):
Figure 18:natrace, natracey, VCF9050 — Time series by race of respondent: percent saying spending on Black people should increase, numbered as in the text. Solid lines: Black respondents; dashed lines: non-Black respondents (GSS, ANES).
Among non-Black respondents, support for more spending is generally low and has never reached a majority -- the highest, “improving conditions” (1), reached 49% in 2019. Responses are more favorable to “improving conditions” than “assistance” or “programs that assist Black people”. Support for the first two has increased since 2010 -- although it might have reversed around 2020.
Among Black respondents, support for spending is much higher, and it has increased since 2010. Support for programs (3) fell from 73% to 57% between 1984 and 2002, when the question was last asked.
Among non-Black respondents, the cohort components of (1) and (2) increase consistently: each generation supports government spending more than the previous." Among Black respondents, support is highest among people born in the 1940s, and a little lower in later generations.
Too far, too fast?¶
Finally, let’s see how people feel about the changes that have happened and the people who worked to bring about those changes.
Starting in 1964, the ANES asked:
Some say that the civil rights people have been trying to push too fast. Others feel they haven’t pushed fast enough. How about you: do you think that civil rights leaders are trying to push too fast, are going too slowly, or are they moving about the right speed?
ANES dropped the question after 1992, but Pew Research asked a related question several times around 2000 and in 2025.
When it comes to Black people having equal rights with White people, do you think our country has gone too far, not gone far enough, or been about right?
The following figure shows the percentage who said “too fast” or “too far”, by race of respondent.
Figure 19:VCF0814 — Time series by race of respondent: percent of non-Black and Black respondents saying civil rights leaders are pushing too fast (ANES, lines; in 1968 the middle answer was worded differently). Circles: White and Black adults saying the country has gone too far in giving Black people equal rights with White people (Pew, 2019–2025), a different question.
Among non-Black respondents, the percentage of “too fast” dropped from 49 percentage points in 28 years. As we should expect with change as fast as that, it’s mostly changed minds, not generational replacement.
Few Black people think civil rights have moved too fast. Between 1964 and 1992 it dropped from 12% to 8%.
In the more recent Pew data, about 5% of Black respondents say the country has gone too far, compared to 20% of Whites.
What are the most important problems?¶
Since 1948, Gallup has asked “What do you think is the most important problem facing this country today?” The following figure shows the percentage who named race relations or racism[6].
Figure 20:Percent naming race relations or racism as the most important problem facing the country (Gallup). Through May 2014, values are digitized from the chart in Gallup’s 2021 article; before 2001 the question was asked irregularly, so straight segments connect polls months or years apart. From June 2014, Gallup’s published monthly readings.
Several of the peaks align with notable historical events, including:
| Date | Peak | Possible events |
|---|---|---|
| Fall 1957 | 29% | The Little Rock Crisis (federal troops escorting Black students into a desegregated school in Arkansas). |
| Sep. 1963 | 52% | The Birmingham campaign and the March on Washington. |
| Aug. 1967 | 35% | Urban riots and racial unrest, especially the Detroit Riot and Newark Riots. |
| Mid-1968 | 19% | Assassination of Martin Luther King Jr. followed by nationwide riots in more than 100 cities. |
| May 1992 | 15% | 1992 Los Angeles Riots after the acquittal of police officers filmed beating Rodney King. |
| 2014–2017 | 13–18% | Multiple events, possibly including protests after the killings of Michael Brown and Eric Garner in 2014, the Charleston church shooting in 2015, the killing of Philando Castile in 2016, and a white supremacist rally in Charlottesville in 2017. |
| June 2020 | 19% | Killing of George Floyd in May 2020, nationwide protests. |
What about the Trump Era?¶
Since his 2015 presidential campaign and continuing into his second term in office, Donald Trump has said many things that his critics interpret as a racist “dog whistle” -- that is, he has used words and phrases that could be understood as support for racist views and policies, but which are ambiguous enough that his defenders can deny that that’s what he meant.
But it is noteworthy that even Donald Trump, who seems so unfiltered on so many topics, stops shy of saying things that are overtly, unambiguously racist. Let’s contrast that restraint with a previous presidential candidate, George Wallace, who said in 1963, “I say segregation now, segregation tomorrow, segregation forever.”
At least part of the difference, we can assume, is that public support for segregation was much higher in 1963 than in 2015. George Wallace probably thought that endorsing segregation would win more votes than it would lose. And Donald Trump, who is old enough that he could have voted for George Wallace, probably realizes that’s no longer true.
Nevertheless, events since 2015 might create the impression that racism is more widespread, more acceptable, and more politically influential now. White supremacist groups became more visible, most notoriously at the 2017 Unite the Right rally in Charlottesville, where marchers carried Nazi symbols and chanted antisemitic slogans. In 2019, a gunman who expressed hatred of Hispanic immigrants killed 23 people at a Walmart in El Paso. In 2020, the killing of George Floyd by police lead to widespread protests, some accompanied by unrest and violence. And since 2025, the Trump administration’s aggressive use of Immigration and Customs Enforcement (ICE), including raids in predominantly Hispanic communities, gives many observers the impression of racial discrimination as government policy.
So we might wonder: If Barack Obama was right to say, in 2016 that now is the best time to be “young, gifted, and black” in America, would he still be right in 2026? To check, I looked at every question in this chapter and the previous one that was asked before and after 2016 and compared the results. Among non-Black respondents, the responses to all 22 questions about racial attitudes moved toward greater fairness and tolerance. Some of these changes were large, and most were due to people changing their minds, not generational replacement. Among Black respondents, 20 of 21 questions moved in the same direction; the exception is the small increase, since 2015, in the percentage who say racial gaps are due to less in-born ability.
There is no evidence of increasing racism since 2015, or reversals in long-term trends. So one conclusion of this chapter is the story of a dog that did not bark[7], or maybe a dog that did not heed the whistle. Donald Trump’s election and re-election were not the result of a widespread increase in racism; they happened despite measureable decreases in racist beliefs and attitudes.
Would you choose now?¶
Across these two chapters, almost every trend goes toward less racist attitudes, less segregation, less discrimination, and smaller gaps in outcomes.
Fewer people want segregation -- and fewer experience it -- in housing, schools, and the workplace. Many of the laws and contracts that created segregation have been repealed or invalidated, and replaced with laws that ban discrimination. Some of those laws were not popular when they were enacted, but they are supported by large majorities now.
Social segregation has decreased, too. Opposition to interracial marriage has all but disappeared and (as we’ll see in the next chapter) marriages between people from different racial and ethnic backgrounds are more common.
Since the 1970s, racial gaps in employment and housing have narrowed and real income has increased in all groups -- but the income gap is wider now. Support for affirmative action has increased, and fewer people believe that White employees are victims of reverse discrimination. Fewer Black employees report discrimination in the workplace, and representation in corporate leadership has increased -- but Blacks are still underrepresented at the top levels.
Representation in government has increased, too, reaching or exceeding equality in the Supreme Court, the House of Representatives, and several recent presidential Cabinets -- but not the Senate. More than 90% of Americans say that they would vote for a Black presidential candidate, and when that assertion was put to the test, the results were consistent with the claim.
To explain persistent gaps in outcomes, people are more likely to recognize discrimination as a cause, and less likely to point to inborn ability or lack of effort. And support has increased for government assistance for Black people.
These changes were driven by a combination of changing minds and generational replacement. Looking at generational differences, each generation has been less likely than the one before to express racist attitudes -- which suggests that the decline of racism is likely to continue.
In survey questions quoted in this book, I capitalize Black and White, which the codebooks usually do not. When the wording of a question changed over time, I quote the most recent version. Earlier versions of many of these questions said “Negro” or “Black”; some later versions say “African-American.”
Private companies with 100 or more employees report to the Equal Employment Opportunity Commission (EEOC) how many of their employees are in each job category, by sex, race, and ethnicity. We have summary data from 1966 to 2013 and more detailed data from 2014 to 2023.
Clifton Wharton Jr. became CEO of the pension fund TIAA-CREF more than a decade earlier, in 1987, but at the time it was not on the list.
Inadequate housing and crowding are computed from the American Housing Survey, which the Census Bureau has conducted every year or two since 1973.
Through May 2014, the values are digitized from a chart in Gallup’s 2021 article. From June 2014, they are Gallup’s published monthly readings.
The origin of this phrase is a Sherlock Holmes story in which the key to solving a mystery is not an event that happened but rather an event that informatively did not happen.
- Flood, S., King, M., Rodgers, R., Ruggles, S., Warren, J. R., Warren, D., Breton, E., Cooper, G., Drew, J. A. R., Richards, S., Van Riper, D., & Williams, K. C. W. (2025). IPUMS, Current Population Survey: Version 13.0. Minneapolis, MN: IPUMS. 10.18128/D030.V13.0