Reading Time: 5 minutes

In the weeks before the 2004 general election, after the House ethics committee admonished Rep. Tom DeLay for ethics violations, his re-election campaign tried a new tack.

“They went up with a television ad that attempted to show the other side of Tom DeLay — a soft, cuddly version of Tom DeLay,” pollster Rob Autry told the audience at a 2006 political consulting conference.

It fell to the Texan’s pollsters at Public Opinion Strategies, the Republican research firm of which Autry is a vice president, to tell them that the public wasn’t buying it.

“While they acknowledged that, ‘Hey, he’s probably a decent guy,’” Autry said, surveyors kept hearing “‘I’m not so sure he’s the right guy for me.’” Worse, over the two- or three-week run of the ad, “the numbers hadn’t moved at all.”

With only days before the election, the pollsters argued for a change in strategy: focus attention on DeLay’s opponent. The new ad linked his Democratic opponent Richard Morrison to John Kerry, Howard Dean and Michael Moore. On Election Day, DeLay regained his seat by a 14 percent margin.

Since its 1991 founding, Public Opinion Strategies has done more than 7,600 surveys for politicians, government agencies and corporations. Its Web site describes the firm as “strategic partners,” who “use data to make decisions” and “don’t hesitate to have an opinion, make a judgment, and then live or die by the results.” In the 2004 elections, the firm took in more than $4.5 million for its campaign services, which constitute “about half” of its work, according to the Web site.

Knowing what the numbers are, and how to move them, has led Public Opinion Strategies to the top of the list of Republican pollsters. According to the National Journal, the firm’s 2004 win-loss record was the best among major consulting firms. Its 2004 clients included the National Republican Congressional Committee (which paid more than $2.3 million for the firm’s services, according to the Center for Public Integrity’s analysis of 2003 and 2004 campaign filings), the Republican National Committee ($219,000) and George Bush ($204,000), who also employed his own in-house pollsters.

The same surveys that tell researchers that voters aren’t likely to, for instance, cuddle up to DeLay, can also reveal that voters aren’t enamored of a “liberal” supporter of “radical Michael Moore,” as DeLay’s commercial put it. Testing reactions to those sorts of negative messages has landed Public Opinion Strategies in hot water on several occasions, when the firm has been accused by opponents of crossing the line between research and spreading negative views, a practice known as “push polling.”

Surveys testing negative messages are not usually made public, but they form a basis for a campaign’s decisions about whether — and how — to attack opponents. Questions tend to be along the lines of “Would you be more or less likely to vote for this person if you knew that …?”

In 1996, 800 voters in Texas were asked for their opinions on then-Attorney General Dan Morales, who was preparing to file suit against several major tobacco companies. The survey, taken by Public Opinion Strategies on behalf of some of the tobacco companies, and later published online by Mother Jones magazine, asked several questions about voters’ opinions of Morales and the proposed lawsuit. Voters were asked to react to Morales’ support of gun control and affirmative action, among other issues — positions he later told The New York Times were mischaracterized.

In a telephone interview, Public Opinion Strategies co-founder Glen Bolger acknowledged that his company commonly tests for negative messages — but not just about the opponent.

“Most of the time we’re [also] testing stuff our candidate is likely to get hit with,” he said.

The year 1996 was big for push polling, and the practice garnered heavy media attention. According to media accounts, the Bob Dole for president campaign engaged in a push poll attack on fellow Republican contender Steve Forbes in several primary states.
As the official Dole pollsters, Public Opinion Strategies caught the flack for the incident, until the State Journal-Register of Springfield, Ill., and the Wall Street Journal brought to light a telemarketing company, Campaign Tel Ltd., which had been paid more than $1 million by the campaign to make at least 10,000 quick, anonymous calls criticizing Forbes’ flat tax plan or his position on gays in the military.

In August of that year, 31 members of the American Association of Political Consultants, professional pollsters including Public Opinion Strategies, sent a letter to the association and to media outlets outlining the differences between push polls and real polls. Public Opinion Strategies founders Bolger and Bill McInturff co-authored an article for Campaigns & Elections magazine titled “‘Push Polling’ Stinks,” echoing the AAPC’s position in distinguishing between the practice and legitimate “survey research.”

Legitimate calls, they wrote, are drawn from a small sample of voters (less than 1,000 people) to gauge public opinion. The calls last from five to 40 minutes as pollsters test a variety of messages, including negative messages about the candidate and his or her opponent. Quick hit push polls, by contrast, go out to many thousands of people in an often last-minute effort to change votes and spread negative, usually biased, information. Bolger and McInturff objected to the term “poll” for such calls, preferring “persuasion or advocacy phone calling.”

The article ended with a piece of advice for push poll victims: “If your campaign is hit with a last minute bout of misleading phone calls, you should spotlight it. But if in January or June before a November election you should receive wind of a 20-minute survey, settle down before you dial up the press. At least your opponent may have telegraphed a likely campaign strategy.”

A number of opponents of Public Opinion Strategies clients have failed to heed that advice. Based on a complaint from one recipient of a call two weeks before a 1998 election, the Florida Elections Commission investigated that state’s Republican Party on charges of push polling in connection with a Public Opinion Strategies poll.

A 1997 Florida law requires pollsters to disclose who is sponsoring the call. The law presumes that telephone campaigns of fewer than 1,000 calls lasting longer than two minutes are legitimate polls. For the poll in question, Public Opinion Strategies made 300 calls lasting four to six minutes, but included, commission staff reported, statements that “appear to have been distortions, exaggerations, misrepresentations or innuendo.”

Upon review, however, the commission found that the questions in the poll were “efforts to seek information from voters for use in tabulating the voters’ reactions to certain statements,” and cleared the party in 2000.

In 2004, the Public Opinion Strategies’ poll of 300 likely voters in Georgia for congressional candidate Lynn Westmoreland drew fire from his opponent in the Republican primary, Dylan Glenn. The poll included questions highlighting Glenn’s race (African-American) and marital status (single) and comparing them to those of Westmoreland (white and married).

Bolger says the poll asked about Glenn’s race because it seemed to be helping him. The pollsters found that “a lot of people were supporting him because they wanted to support an African-American candidate,” he said. “Conservatives want to support an African-American candidate.” Bolger says the firm was testing the electoral waters “to see how much of a boost that was giving him.”

Public Opinion Strategies keeps a denunciation of push polling on its Web site: “‘In fact, ‘push polling’ is NOT polling at all. … The differences between push-polling and survey research could not be more dramatic.” Bolger stresses that his firm is strictly a research company — it makes no telemarketing or “get out the vote” calls.

In addition to telephone surveys, the company tests messages through focus groups, Internet surveys and “mall intercepts” — asking random shoppers to rate advertising. In 2004, they launched a trial consulting division.

Public Opinion Strategies has teamed with Democratic polling firms to produce surveys for corporations and interest groups, as well as for news outlets such as National Public Radio, NBC and the Wall Street Journal.

Help support this work

Public Integrity doesn’t have paywalls and doesn’t accept advertising so that our investigative reporting can have the widest possible impact on addressing inequality in the U.S. Our work is possible thanks to support from people like you.