By now, it’s a familiar pattern.
The U.S. Justice Department swoops into a community and investigates the hiring practices of the local police force. The government finds evidence of discrimination and sues the community, charging violations of federal civil rights law.
Local police hiring usually stops because the test is in litigation. There are court battles that sometimes take years. And in the end, the local officials usually pledge to use a new test, one that is endorsed by the Justice Department. That scenario has been played out in more than 60 state and law enforcement jurisdictions nationally – including the Nassau and Suffolk police departments – as public employers struggle to find a way to diversify their police forces and comply with federal law. But now, the tests endorsed by the government to remedy the problem are themselves coming under increasing fire.
Some argue that test standards have been lowered significantly in an effort to hire more minorities. Others maintain the tests still discriminate against minorities. And after all the battling, Long Island’s police departments, like so many across the country, remain overwhelmingly white. In cities from Las Vegas to Torrance, Calif., the tests – which combine biographical questions with the traditional cognitive ones – are the subject of lawsuits and debates over whether they can really measure what makes a good cop. In Las Vegas, for example, police officials last year sought a new hiring test after a handful of officers who were hired off a test similar to Suffolk’s were later accused of criminal activity.
In Washington, D.C., a congressional subcommittee has been looking into whether the Justice Department should even be imposing a particular kind of exam on local communities. One critic of the tests who testified before that committee, Dr. Linda Gottfredson, a sociologist from the University of Delaware, said the Justice Department "has been spending federal tax dollars to strong-arm police departments into quota-hiring, in the process lowering their hiring standards and costing their jurisdictions millions of dollars."
On Long Island, Suffolk County recently jettisoned the hiring test endorsed by the Justice Department after a cheating scandal erupted involving the biographical questions. Nassau County’s test, first given in 1994 and the forerunner of the new personality-oriented exams supported by the government, has been both the subject of intense debate among test designers and the target of a so-far unsuccessful legal challenge initiated by a group of 68 white applicants who didn’t score well.
Meanwhile, test scores in police academies in both Nassau and Suffolk have fallen between 4 and 5 percent, and the failure rate has risen since the introduction of the exams supported by the government. The heart of the issue is the contention by the Justice Department and many experts that minorities don’t do as well on traditional cognitive exams, which measures reading comprehension, math skills and reasoning. The reasons are related to cultural, educational and socio-economic differences, experts say. "Our experience is that a number of local government employers, especially in public safety . . . use a written exam which has severe disparate impact upon minorities," said John Gadzichowski, special litigation counsel to the Justice Department in charge of the Long Island cases.
Given that contention, the challenge has been to construct a test that can gauge who would make a good police officer without discriminating against minorities. The government has responded with tests that add components measuring personality and biographical information. "These are jobs where predicting performance is not all about the ability to read and do math. It’s how you react, how hard you work and how steady you are," said David Rose, a former Justice Department attorney. But critics say the new tests may sacrifice high academic standards. Gottfredson said the Nassau test doesn’t truly measure cognitive skills necessary to be a good officer. "You don’t want some airhead making decisions about your life and investigating cases.
Policing requires considerable judgment," she said. Still others say it will take more than a hiring test to overcome the factors that lead to the differences in the way white and minority candidates score. "I think the answer has more to do with more recruitment in minority neighborhoods because there’s no magic bullet out there to resolve the social injustices that are there before you give the test," said Nassau Police Benevolent Association President Gary DelaRaba. Zina Leftenant, president of the Nassau Guardians Association, a group of minority police officers, said the department and minority cops have to do a better job of recruiting minorities and guiding them through the test and screening process. "We have to step out and educate our own people about the test and how to prepare for it," Leftenant said. The ultimate goal of the new tests – to diversify police forces – has remained elusive even 21 years after the Justice Department sued Nassau and 15 years after it sued Suffolk.
In May, 1989, the first year such statistics were kept, the Suffolk police department was 93.60 percent white, 2.22 percent black and 3.93 percent Hispanic. In July, 1998, the Suffolk force is 92.37 percent white, 2.17 percent black and 5.05 percent Hispanic. In Nassau, the department was 97.65 percent white in 1977, when the Justice Department first filed suit, 1.49 percent black and 0.87 percent Hispanic. In 1998, it is 91.94 percent white, 3.66 percent black and 3.85 percent Hispanic. "It’s improving over what it has been. But it has some more to go," Gadzichowski said.
But critics argue that those gains have come at a price. Since the new tests were introduced, academic performance in Nassau and Suffolk’s police academies has gone down. The average score of Nassau recruits hired off the 1987 test was 90.5 percent, compared with an average of 84.5 percent when the two classes that took the new 1994 test are combined.
When comparing the number of Nassau recruits who left before graduation, there were five resignations and one failed student in the class from the 1987 tests, compared with 15 resignations and four terminations in the combined 1994 classes. In Suffolk, when comparing averages of classes between 1980 and 1988 with classes since 1989, the score average has come down from 90.1 percent to 87.1 percent.
The rate of those leaving, either resigning or being terminated, went from 7.2 percent to 9.7 percent. Gadzichowski said he was not concerned about the slip in academic performance because that is not the only indicator of a good police officer. "The test is not necessarily designed to pick who does well in the academy but, more importantly, who will do well on the job after the academy," he said.
The new test replaced a standard cognitive test that was the subject of more than a decade of litigation. The suit ultimately led to an agreement whereby the county and the federal government jointly appointed a team to design a new "multi-dimensional" hiring test. "Their mission was to put together a selection test that broke new ground," said William Pauley III, the attorney who has represented Nassau.
But critics say that the exam, first given in 1994, was changed solely to ensure minorities would do well on it, and that the scoring of test results was manipulated so minorities would come out on top. "We think determining discrimination by the numbers is methodologically flawed, and we think degrading or changing the exam so you achieve a certain number of minorities, at a cost of testing for the skills needed to do a particular job, is a mistake," said Martin Kaufman, an attorney with the Atlantic Legal Foundation, a conservative law firm, which represents the white Nassau applicants suing in federal court.
The suit was dismissed and is now on appeal. The chairman of the team of experts that designed Nassau’s test, David Jones, denied such manipulation. He said the questions were designed to reduce adverse impact on minorities, but were thrown out only if they weren’t valid – meaning that the questions were shown, in a separate study, to be inaccurate in predicting who would be a good police officer. Suffolk officials avoided litigation by agreeing in 1986 to use a test approved by the Justice Department and designed by a Virginia-based test company. That test also included "biodata" questions and personality measures.
But Frank Erwin, the test designer, said it has a much stronger cognitive section than Nassau’s and is based on a more comprehensive validity study. But after using the government-backed test three times since 1988, Suffolk officials are now looking for a new test after the cheating scandal erupted and the federal government backed away from the exam. Twenty-two officers have received suspensions for giving false answers to key autobiographical questions. The Suffolk district attorney’s office has also filed criminal charges against a police sergeant accused of stealing test questions and departmental charges were brought against a deputy inspector accused of improperly coaching job applicants in 1988. Erwin gave basically the same version of the exam three times since 1988, a practice that Gadzichowski said may have led to the cheating problems.
But Erwin said it is not uncommon to give the same test in a situation where tests are given four years apart and generally to different populations. He said the security problem was not related to the contents of the test itself. Gadzichowski sent the county a letter last April in which he directed officials to give the government 20 days notice if they intended to hire from the latest test list. He said, based on the results, that the exam had "serious adverse impact" on minorities. Of the top 2,299 scorers, Gadzichowski said 2,008 are white, while just 54 are black and 166 are Hispanic. Suffolk police union officials and the Suffolk County Guardians, which represents the department’s black officers, also have been critical of the Erwin test. Suffolk Police Benevolent Association President Jeff Frayler said last June that the test had a "pre-determined profile" in which college-educated whites were favored.
But Erwin, of the Washington-based firm of Richardson, Bellows and Henry, said the Justice Department backed away from his test because "it wasn’t giving them the numbers they want," referring to numbers of minorities scoring well. Controversy has followed the Erwin test, known as the RBH exam, to other places where it had been used at the behest of the Justice Department. In Las Vegas, police officials chose another test last year after several instances in which a handful of officers hired off the RBH test list were accused of criminal activity, including murder.
The department switched to a new test and added oral interviews and a video test. Las Vegas officials won’t comment on whether they blame the test for the bad apples, saying only it wasn’t helping them diversify the force. "We had some issues about using the RBH test and also felt it had adverse impact on minorities," said Katey LaVelle, director of employment and classification for the Las Vegas Police Department. Gadzichowski and Erwin both say that the Erwin test is only one component of a lengthy psychological and physicial screening process and background check and that it is impossible to say the test alone brought in the corrupt officers. At the same time, in New Jersey, the NAACP Legal Defense Fund is challenging the use of the RBH test by state troopers, charging it has too much adverse impact on minorities. New Jersey agreed to use the RBH test in 1992, after protracted litigation with the Justice Department. But NAACP attorney David Thronson said that since the RBH test was introduced, the number of minorities on the force has actually dropped.
One community that fought the Justice Department and won is Torrance, Calif., where a federal judge wrote that the federal government has not convincingly argued that the city’s hiring test, a standard cognitive exam, was unrelated to job performance. U.S. District Court Judge Mariana R. Pfaelzer in her decision Sept. 18, 1996, wrote, "The methodology for designing such standardized tests has been well-known and well-established within the psychological profession for years, likely since the 1920s. The tests utilized by defendant . . . are reliable and valid predictors of performance in those jobs." In fact, Pfaelzer even granted a motion by the city that the government pay $ 1.8 million in legal costs. "Most municipalities wouldn’t stand up and fight," said Wayne Flick, the attorney for Torrance. "They buckled under and said, ‘OK, we won’t use reading and writing in hiring police officers. "