Boston — Carl Klump says he could have solved four murders and one bombing by now. He has also helped prevent a teen-age girl from committing suicide. He's heard more confessions than many a police interrogator or priest. But he works for businesses.
Mr. Klump is an expert on honesty.
As head of the Stanton Corporation in Chicago, he administers "honesty tests" for food chains, clothing stores, banks, and other industries. For $6 to $12 per test, he'll decide if a job applicant is a "good risk," or likely to pilfer company money or materials.
The pencil-and-paper test Klump developed quizzes job candidates about their attitudes toward theft and toward punishment for those who steal. Some of the answers he receives are remarkably frank. In one sample of 159 test-takers, the amount of money and merchandise subjects confessed to stealing from previous employers totaled $3,663.
Buoyed by the success of these tests, Klump is designing new exams which he says will predict an applicant's rate of absenteeism and level of productivity.
Klump, a criminologist and polygraph specialist, developed the original Stanton Survey in 1964. Like a growing number of reputable honesty exams, the Stanton Survey is a series of "yes and no" and essay questions. Applicants fill out an hour-long, confidential form which is sent directly to the Stanton Corporation for evaluation. The answers are compared with attitudes and behavior patterns Klump has identified as common among typically honest and dishonest people.
The tests rate job applicants as low-risk (almost never steals, turns down the opportunity to steal); marginal-risk (takes "nickel- and-dime" things if the opportunity arises); or high-risk (will steal "anything that's not nailed down," creates opportunities to steal, and encourages others to do so).
In a random sample of test results, 53 percent of the applicants were rated "low- risk," 27 percent were "marginal-risk," and 20 percent were rated "high-risk."
Klump says he is used to skepticism. Company officials, although increasingly in need of ways to avert theft, usually doubt the effectiveness of the preemployment tests. They often hire people despite a "high-risk" rating, only to catch them in theft later.
But in the case of the teen-age job applicant who wrote she ought to "take some pills and end it," Klump explained the problem to the employer, who decided to hire the girl and arranged for her to see a counselor. "She's now a valued employee," he says.
It may seem presumptuous -- even dangerous -- to suggest that Klump and a growing number of other "honesty analysts" can assess the honesty of people they've never seen, merely by reading between the lines of their test answers.
"I was extremely suspicious of the test at first," says Dr. Homer Reed, a neuropsychologist at Tuffs New England Medical Center. In his effort to discredit the test, Dr. Reed eventually validated it, and for a number of years acted as a consultant to the Stanton Corporation. His investigation concluded that the test "does not result in unfair rejections" of honest applicants.
To be labeled "high-risk," Klump says, applicants must "fail" three aspects of the test: They must have a high numerical score (indicating greater likelihood of dishonesty); they must admit previous delinquent behavior; and, in the essay portions, they must use key phrases which Klump has found to be used consistently by dishonest people. For example, he says the phrase "I am what I am" is almost neverm used by an honest person, but suggests an attempt to justify dishonesty and an unwillingness to correct it.
"The premise of the test is . . . that past dishonest acts are indicative of future dishonest acts," says the security manager (who asked not to be named) of a nationwide retail clothing chain which has used the Stanton test for several years. He says the spread in scores between honest and dishonest people is usually so obvious that it precludes erroneous "labeling." But Dr. Reed concedes that the test doesn't necessarily account for the person who reforms drastically from past habits.
The questions are probing, giving applicants the chance to talk philosophically about their goals and their lives. A few have confessed to serious crimes, -- but Klump says he only releases information that relates to an applicant's honesty rating.
Samples of test questions cannot be published here, since that could skew future results. But all of the questions meet states' standards of fair employment practices.
So who is the dishonest person?
That's hard to say.
"We expected that blacks would be found to be more dishonest just because of socioeconomic factors, but the results went against anything we expected to find ," Dr. Reed says. "There's absolutely no difference" between the of blacks and whites tested. Another tried- and-true honesty test, the Reid Report, developed by criminologist John E. Reid, has a similar format and corroborating results.
The only exception is South Africa, where attitudes of blacks are so vastly different that Klump had to devise a separate test for them. "It's not because they're black," he says. "It's just a cultural thing."
Ongoing tests in the Philippines, Australia, South Africa, and Canada have yielded data similar to tests in the United States, showing that employees abroad are no more or less honest than American employees.
Both the Reid and Stanton surveys indicate slightly more honest scores for women and increasingly better scores for higher age groups -- which analysts attribute to maturity and the desire not to endanger one's job.