Imprisonment Rates Vary Wildly by Race

by Steve Sailer

UPI, June 14, 2001


LOS ANGELES, June 14, 2001 (UPI) -- The Justice Department announced Wednesday that in 2000 violent crime was down a striking 46 percent from its peak in 1994, which was during the crack epidemic. This result was according to the annual National Crime Victimization Survey of 160,000 Americans.

But has this tremendous reduction in violence been accomplished at too high of a price in terms of the greatly increased numbers of imprisoned blacks and Hispanics?

The National Center on Institutions and Alternatives (, a liberal think tank advocating less imprisonment, has released a new report, "Masking the Divide," that argues that, "The overuse of incarceration is causing severe and potentially irreparable divisions in society." It wants society to "turn the criminal justice system off its racist path and begin to repair the damage it is causing."

The NCIA said, "During the twelve years we examined (1985 to 1997), the U.S. prisoner population more than doubled from 502,376 to 1,240,962. Nationally, non-whites accounted for 70 percent of this growth in state and federal prisons."

The number of blacks and Hispanics behind bars grew 180 percent from 1985 to 1997 compared with only 102 percent for Anglo whites.

In "Masking the Divide: How Officially Reported Prison Statistics Distort the Racial and Ethnic Realities of Prison Growth," the NCIA broke out Hispanics as a separate category from non-Hispanic whites, which is an improvement on standard government criminological statistics that often ignore Hispanic ethnicity. Until now, comparisons of imprisonment rates across states and races had been confused by law enforcement agencies' tendency to lump Hispanics in with Anglo whites.

"Because prison statistics don't separate out Hispanic/Latinos from other racial groups, we believe the scale of the racial divide in American prisons is masked," stated the report's author, Barry Holman. "Without distinguishing between Hispanic/Latinos, whites and African Americans, the number of white prisoners is significantly overstated."

That's because, according to the NCIA, Hispanics have a 3.7-times-higher rate of imprisonment that non-Hispanic whites, so including Hispanics in the white total made the black-white gap in imprisonment look less enormous than it really is. (In reality, blacks are imprisoned 9.1 times more than Anglo whites.) The NCIA used a variety of statistical techniques to estimate the Hispanic prison population.

Nationwide in 1997, non-Hispanic whites comprised 34.8 percent of the prisoners, African-Americans 46.9 percent, Hispanics 16.0 percent, and others 2.3 percent. Overall, the study found that 2.6 percent of the African-American adult population was imprisoned in 1997, compared to 1.1 percent of Hispanics, and 0.3 percent of non-Hispanic whites. The report does not break out imprisonment rates for Asian-Americans, but most experts believe Asians tend to be imprisoned the least of all major groups.

The NCIA argued that this high rate of imprisonment of blacks and Hispanics is unfair: "Whites seem to go to jail in smaller numbers than their share of serious crimes would indicate. During the 1990s, whites committed 56 percent of violent crimes and 62 percent of felonies in the United States, according to Justice Department statistics."

Iain Murray, who specializes in the analysis of crime statistics at the Statistical Assessment Service, a Washington D.C., non-profit, non-partisan public policy organization, was skeptical. "These people are comparing apples and pears," he said.

He noted that government's crime statistics have the same problem as its imprisonment statistics -- they lump together Anglo whites and Hispanics. So, when the NCIA improves the quality of the prison statistics but not the crime statistics, it artificially exaggerates the perception that Anglo whites are being treated more leniently.

Further, the more serious the crime (and thus the longer the prison term), the greater the black-white gap. For example, in 1990 during the heart of the crack years, blacks committed 11.4 times more homicides per capita than whites, according to Justice Department statistics. (And that number under-estimated the black-to-Anglo white ratio because the white category included many Hispanics.) By 1999, after a huge increase in the number of blacks in prison, the black-to-white murder ratio was down to only 6.6 to 1.

The NCIA argued, "What is largely driving the expanding prison population is the "war on drugs ... America's prisons are full to overflowing with half of all prisoners confined for non-violent offenses and half of these for drug offenses." It noted, "Admissions to prison for drug offenses increased a whopping 1,040 percent between 1986 and 1996."

Of course, one should not assume that someone convicted of a non-violent offense is necessarily a non-violent individual. Al Capone, for example, was sent to Alcatraz for tax fraud rather than the numerous murders he oversaw.

Indeed, imprisoning huge numbers of people for crack-related offenses was followed by a remarkable decline in violent crime in the 1990s. Some statistical analysts, such as University of Chicago economist Steven Levitt, argue that the growth of imprisonment accounts for half of the drop in crime.

Nonetheless, the huge growth in the prison population is a serious national issue. The NCIA has helped make possible a more realistic debate by providing more accurate statistics.

Interestingly, crunching the data in the report's appendices sheds light on a number of fascinating topics that did not particularly interest the report's sponsors. (Imprisonment Data by race by state is collected here. And, Ethan Herdrick's data maps showing crime rates by state by ethnicity are here, and here.

For example, "The Sopranos" television drama has revived New Jersey's reputation as a hotbed of white criminals. Yet, to the extent that a tendency to be law-abiding can be estimated from imprisonment rates, that much-maligned state appears in fact to have the second most law-abiding non-Hispanic white people in America. According to a new report that breaks down imprisonment rates by race and ethnicity, white New Jerseyites trail only the notoriously nice white folks of Minnesota in staying out of prison.

Some findings confirm common sense -- for example, whites in fast-living Nevada are more than twice as likely to be in prison as whites in the mostly Mormon neighboring state of Utah.

In contrast, some of the data undermine common myths. Besides polishing the tarnished image of New Jersey's whites, the numbers also reveal the surprising news that politically liberal states, not conservative ones, are likely to have the largest gap between the imprisonment rates of blacks and whites.

These ratios varied significantly from state to state. While one might expect that the highest proportion of black-to-white imprisonment would occur in politically conservative states, the opposite was true. It was in Democratic-leaning states where blacks had the highest rates of imprisonment relative to whites.

For instance, the racial gap in the highly liberal, black-dominated District of Columbia was found to be off the charts. In D.C., a black person is 56 times more likely than a white person to be in prison. The next-largest racial disparities were found in liberal mainstays Minnesota (a 31-times higher rate of blacks being in prison) and Wisconsin (22 times higher), followed by New Jersey, Iowa, Connecticut, Pennsylvania, Rhode Island, and Illinois. All of these states voted for Democratic presidential candidate Al Gore in 2000.

Of course, it's not uncommon for regions that are highly liberal in terms of national politics to vote for conservatives in local elections. For example, the Democratic Party's liberal bastions of New York City and Los Angeles each elected law-and-order Republican mayors in the mid-1990s, following the crack epidemic crime wave that began in the late 1980s.

The smallest difference in the black-to-white imprisonment rate was found in liberal Hawaii (only 2.9 to 1). This may have something to do with many members of Hawaii's small African-American community being active or retired members of the U.S. armed forces.

After Hawaii, though, the next 10 states closest to black-white racial equality in imprisonment rates were all Southern or Western states that voted for George W. Bush. For example, highly conservative Mississippi and South Carolina each imprisoned blacks only six times more often than whites per person, compared to the national average of nine times more often.

Eighteen of the 20 states with the least disparity between blacks and whites voted for Bush in 2000. These below-average racial ratios are driven in part by the tendency of whites in Republican states to get themselves thrown in prison more often than whites in Democratic states. The highest white imprisonment rates tend to be in old frontier states of the Wild West.

The most often locked up whites are in Alaska, followed by Oklahoma, Nevada, Arizona and Texas.

For example, to focus on two large states, whites in highly Republican Texas are 3.4 times as likely to be in the penitentiary compared to whites in highly Democratic New York. Does that mean white Texans are much more criminally inclined than white New Yorkers? Or is the criminal justice system that much harsher in Texas? Or are both true to some degree?

These are difficult questions. For many years, states differed significantly in length of prison sentences handed out, but in recent years federal minimum sentencing guidelines have reduced states' discretionary power. So, the evidence suggests that white Texans really do tend to be less law-abiding than white New Yorkers.

The same general pattern of liberal states having the highest ratio of minority-to-white imprisonment rates are found for Hispanics as well as African-Americans. Hispanics are 12 times more likely than whites to be in prison in Connecticut and Pennsylvania, 10 times more likely in New York, and nine times more likely in Massachusetts and Minnesota. All of these states voted for Gore.

Overall, though, class and race differences within the overall Hispanic population appear to be a better predictor of a state's Hispanic-to-white imprisonment ratio than the state's presidential voting tendencies.

Among states with a large Hispanic population, the closest to Hispanic vs. white equality in imprisonment rates is found in Florida. In the Sunshine state, a Hispanic is only 1.2 times more likely to be in custody than a white. That may be because Florida is home to a large middle-class, primarily white Cuban population.

At the other end of the scale are those Northeastern states, such as Connecticut and New York, where there is a large underclass Hispanic population of Puerto Ricans and Dominicans. These Hispanic nationalities tend to be a mixture of white and black.

In between Florida and the Northeast in terms of racial equality fall Western states whose Hispanic populations are largely working class Mexicans and Central Americans. These ethnic groups tend to be comprised of admixtures of whites and New World Indians.

Western states where whites tend to get themselves thrown in prison at above average rates tend to have fairly small differences in imprisonment between their whites and their Hispanics. In Nevada -- where whites, not surprisingly, wind up in cells almost twice as frequently as the national white average -- the Hispanic-to-white imprisonment ratio is only 1.5 to one. In Texas, it's 2.2- to-1 and in California, another state with a high rate of whites in custody, the ratio is only 2.4-to-1.

On the other hand, in Utah, where a majority of whites are Mormons, the Hispanic to white imprisonment proportion is a substantial 4.4-to-1. This Hispanic-to-white ratio is also above 4-to-1 in Colorado and Washington, where relatively few whites get sent to jail.

Across all races, the fewest prisoners per capita were found in the cold-weather, mostly white states of Minnesota, North Dakota and Maine. Cold weather tends to discourage street crime.

In contrast, the mostly black District of Columbia had by far the highest overall imprisonment rate. The next-most-locked-up citizenry was found in Texas, a state with sizable fractions of both African-Americans and Hispanics, along with a white population that is far above average in finding itself behind bars.

Not even frigid weather could keep Alaskans from winding up in prison almost as often as Texans.

The seemingly mild little state of Delaware came next for highest imprisonment rate, followed by spicy Louisiana, then Oklahoma, South Carolina, Nevada, Mississippi, Arizona, and supposedly starchy Connecticut.

Imprisonment Data by race by state is collected here.

Ethan Herdrick's data maps showing crime rates by state by ethnicity are here, and here.

Steve Sailer ( is a columnist for


Subscribe to The American Conservative
(because I don't post my magazine reviews online until long after the films have come and gone) home page

iSteve film reviews

email me


Steve Sailer's homepage 

iSteve film reviews

email me