An Authority On Public Opinion Polling, Born Of His Respect For People

Andy Kohut was at NPR headquarters every election night for decades trying to explain what was happening with patience and humor. On the famous, never-ending election night of 2000, Kohut said, chuckling, "We're not lost ... but we don't have the foggiest clear idea about a trend."

Anne Fengyan Shi/Courtesy Pew Research Center

It is hard to imagine a presidential election cycle at NPR without the warm presence and reassuring wisdom of Andy Kohut.

The man who founded the Pew Research Center for the People & the Press, and who later served as president of the expanded Pew Research Center from 2004 to 2012, died early Tuesday after years battling leukemia. He was 73.

He had previously been president of the Gallup Organization from 1979-1989 before founding the Times Mirror Center for the People & the Press in 1990. Pew Charitable Trusts became the center's sole sponsor in 1996. He was also seen often on PBS NewsHour and many other TV programs.

But we always thought of Andy as ours.

Because for decades, he was NPR's on-air and in-house expert on the mind of America. And we called on him often — more than 500 times, in fact. We called him our public opinion consultant and analyst, which meant election night was always Andy Night at NPR. From the Age of Reagan to the Era of Obama, Andy was The Man. And he was the man to see, as soon as the data started rolling in. Because, while always disguised as a mild-mannered reporter, Andy seemed to have superpowers.

One of these was the power of always keeping his head, especially when all around him were losing theirs. On one fateful night in November 2000, when media organizations large and small were "calling" the state of Florida for Vice President Al Gore, the Democratic nominee for president, Andy was unconvinced. The numbers were not complete enough, the margin too narrow. Funny things were happening with the data in specific places. No call, said Andy.

Later that night, news people of all kinds ate humble pie and retracted the Florida call for Gore. Then the news industry as a herd called it for George W. Bush — giving him enough Electoral College votes to be president. No, Andy said throughout, it's still not clear that either candidate has won Florida.

And indeed it was not. Only a few hundred votes divided Bush and Gore in the actual tally, a difference far too small to show up in exit polling. It would be five weeks before the U.S. Supreme Court finally stopped the recount and legal warfare, in effect awarding the election to Bush.

We will also remember the afternoon in November 2004 when the first wave of exit polls seemed to show Sen. John Kerry defeating President Bush. There were rumors of gloom and concession speeches at the White House that afternoon. But Andy took a look at the first wave data and shook his head. The demographics indicated the data came from disproportionately Democratic voting groups. When all the votes were in the real result might be quite different. There should be no on-air discussion of the exit polls. And within hours it was clear Andy was right, as Bush pulled even and then pulled ahead.

But grateful as we were for Andy on election night, he was welcome on countless other nights, as well. We turned to Andy after big primaries and major presidential speeches, at the national conventions and at other times when events demanded deep insight into the mind of America.

Many of those occasions were generated by Andy and Pew's own original research, which went far beyond horse race polls in presidential years. His passion was to know what people thought and understand their thinking. He was always looking for a broader canvass, a wider world of literally global opinion. He took the Pew brand international, creating a mosaic from 270,000 survey interviews done in 57 countries.

He also worked hard to stay ahead of the social and technological changes that were making good polling harder to achieve. When cellphones began replacing landlines, especially for younger voters, Andy was devising the means and arguing for the resources to include cellphones in his surveys.

He was a fierce defender of quality polling, and an advocate for finding the resources to do it right. Cost and competition may have forced other pollsters to fall back on smaller and smaller sample sizes, or to switch to online and "robocall" polling. But Andy kept prioritizing the classic principles of scientific opinion research, having learned literally from the inventors of the technology, George Gallup and Paul Perry. Getting it right was the controlling value.

He was also an innovator in interpreting the data that poured in from his surveys. Even back to his days as the wunderkind at Gallup, he was looking for ways to get beyond the usual breakdowns of gender and age, race and occupation, education and income. The more sophisticated "typologies" he invented and began refining as far back as 1988 separated out the competing strains of liberalism and conservatism, activism and apathy. They became a source of insight for generations of journalists and political scientists, and were among the reasons Andy was awarded the American Political Science Association's Carey McWilliams Award in 2010, in recognition of his "major contribution to our understanding of political science." It was the only time the award has gone to a pollster.