Greetings to Roselyn and the Data & News Society. I’m sorry that I cannot be with you in person, but I welcome this opportunity to send my image instead.
There is a new book related to my topic. Robert H. Frank of Cornell University is the author of Success and Luck. He argues that we tend to underestimate the degree to which simple good luck accounts for whatever success we have. Psychologists call this “hindsight bias.” The success of my Precision Journalism, still in print after 43 years, offers a case study.
My first lucky break looked at first like a bad break. I got so involved in undergraduate journalism at Kansas State University, that I deprived myself of a broad liberal arts education. When I graduated in 1952, I was a highly skilled writer, but I had nothing to write about. It was a minor issue at the time because the Korean War was under way, and I faced compulsory military service. So I joined the Navy and voted for Dwight D. Eisenhower for president. He had promised to end the war.
He kept that promise. I got early release from the Navy and plunged into reporting for the Topeka Daily Capital back in Kansas. And I started applying to graduate schools – not to study journalism but to fill the gaps in my education, to learn something to write about. My undergraduate grades were a problem, but my lucky break was having the support of a professor with strong ties to the University of North Carolina. He persuaded the political science department to accept me. And it made me a teaching assistant, a bold move as I had never taken a course in political science. But I was born in the Great Depression, a period of low birth rates. There was not a lot of competition in my age cohort. Another lucky break.
The use of quantitative methods in political science was a fringe specialty in those days. The only computational aid came from a mechanical device that performed multiplication and long division with a great deal of noise and shaking. Too much work, I said, and I stayed on the traditional track of studying the great books and political institutions.
My graduate work was enough to persuade the Miami Herald to hire me for its education beat, which I treated as a political story. Two issues, racial desegregation and international competition in science, made that relevant. I worked my way up to Washington as correspondent for the Herald’s sister paper, the Akron (Ohio) Beacon Journal. But the timing was poor. Ohio was a Republican state, and the election of John F. Kennedy in 1960 had swept the Democrats to power in Congress. The Ohio members were neither powerful nor interesting. Moreover, the Beacon Journal was not read in Washington, and that was a major disadvantage.
One way to nudge a stalled career in journalism is take advantage of one of the mid-career fellowships designed for journalists. The most prestigious of these was the Nieman Fellowship at Harvard University, and my bureau chief, Edwin A. Lahey, had been a member of the first Nieman class in 1938. He agreed to back me, but I had to wait a year because a member of the Beacon Journal city staff had already signaled his intention to apply. That proved to be good luck, because it gave me more time to work on a study plan, a key element in the selection process. I found my inspiration in a newly-published novel by Eugene Burdick, a political scientist. Its title was The 480, and it was based on a real-life case, Kennedy’s use of a computer simulation to test different strategies for his 1960 campaign. This was the sort of methodology that I had avoided in graduate school, but now computers were making it newly relevant. My study proposal argued that if politicians were going to use these techniques to manipulate public opinion, journalists should know enough to explain them.
It proved to be an easy sell. When I got to Harvard, I contacted the professor who taught the graduate course in quantitative methods in political science. Surely I should start at that level. I had a master’s degree! A brief conversation revealed my ignorance, and the professor steered me to an introductory course designed for sophomores. It turned out to be a good fit.
That timing was perhaps my greatest stroke of luck. Harvard faculty had written a software package called DATA-TEXT that made an IBM 7090 accessible to sophomores. Programs were written in a form that was close to plain English. “Tell it, ‘compute correlations,’ said the instructor, and it happens!” I tried, and it did happen. When the machine gave me a three-inch stack of continuous-feed paper with correlations and crosstabs, I took it to Signet House, where Nieman fellows held their weekly dinners, and unfolded it across the floor.
David Hoffman, science writer for the New York Herald Tribune, was impressed. “Package it and sell it,” he said.
It was about then that my goal began to shift from reporting about the use of these methods in politics to putting them to work for my own goal – better reporting.
Selling the idea turned out to be easy because of yet another accident of timing. I was back in the Knight Newspapers Washington bureau when racial violence broke out in Detroit. It started on a Sunday morning. On Wednesday, I was alone in the Washington office when Derick Daniels, the editor of the Detroit Free Press, called. His reporters needed help. I volunteered to go. I went home to pick up a toothbrush and a change of clothes, then caught the next available flight to Detroit.
I spent that evening riding through the burning neighborhoods with the National Guard, and listening to gunfire. Then I gathered material for a narrative on the riot and the questions about its causes. At the end of the week, when the troops were gone and the fires extinguished, the staff met in the city room to discuss what to do next. We were in intense competition with the Detroit News, and we wanted to be visibly better. I proposed a survey, using my Harvard-learned methods plus some consulting help found through my political science contacts at the University of Michigan. “Write me a memo,” said Daniels, who had been my city editor in Miami. I did, and he approved.
The survey falsified two prevailing theories of the riot. The “riff-raff” theory held that the rioters were at the bottom of the economic scale, the poorest of the poor. The survey indicated that economic status was not a predictor of riot participation. Next was the assimilation theory, which blamed the riot on the difficulty that southern blacks found in adjusting to the urban north. The survey found that riot participation was lower among those migrants.
That left rising expectations as the surviving theory. The closer one gets to a much-wanted goal, the greater the frustration at not reaching it. This theory was consistent with Detroit’s reputation as generally progressive in race relations. The report became a factor in the awarding of a Pulitzer Prize to the Detroit Free Press staff.
One of my Harvard connections was Thomas Pettigrew, a social psychologist who specialized in race relations. We had frequent conversations at the weekly social hour of the Department of Social Relations. He told the Russell Sage Foundation about my work, and that led to a friendly letter from its president inviting me to New York to explore our possible common interests. I suggested a book to explain social science research methods to journalists, and that led to a contract for a year’s support to write it. Its working title: The Application of Social and Behavioral Science Research Methods to the Practice of Journalism. It was a work for hire, not a grant, meaning that the Foundation would own and publish it, provided of course that it was judged worthwhile.
I completed the manuscript on time and under budget, and the Foundation sent it out for review, both to journalists and social scientists. They did not like it! And that proved to be lucky for me. The journalists thought it had too much social science and the social scientists could not imagine a journalist doing these things. That caused ownership to revert to me, and I found a publisher on my own. Indiana University Press published it as Precision Journalism in 1973. A fourth edition is still in print, with an international commercial press, Rowman and Littlefield.
Every step was blessed with lucky timing: finding out about computers just as higher-level languages like DATA-TEXT were becoming available, finding an immediate application in Detroit, and getting foundation support for the book.
I used to read a lot of science fiction. One of my favorites was Kurt Vonnegut Jr., whose second novel was called The Sirens of Titan. After a series of exquisitely complicated adventures involving interplanetary travel, the hero lands back on earth and steps out of his space ship to confront a crowd. “What happened to you?” he is asked.
“I was a victim of a series of accidents,” he responds. “As are we all.”
One thing certain about the future is that accidents will happen. You will see value in some accidents and trouble in others. Your goal should be to recognize the valuable accidents and apply their benefits to your reporting on the troublesome ones. I wish you much luck and success.
(This is the video speech that Prof. Phil Meyer sent us for the opening of Data & News Society Workshops 2016.)