Ask Senate Minority Leader Chuck Schumer if he has a favorable or unfavorable view of political polls, and you’re likely to find it depends on whether he likes what the poll has to say.  

Case in point: Schumer’s comments at a presser last week crowing over President Donald Trump’s sinking poll numbers. “The polls this week show Trump has the lowest 100-day approval rating since they started polling 80 years ago. The lowest.”

But when a reporter then brought up the fact that a CNN poll showed the minority leader’s approval numbers at 17 percent were the lowest of any congressional leader, Schumer sputtered, “Polls come and go. Our party is united.”

Schumer is far from alone in cherry-picking polls to bestow an imprimatur of credibility.

The Trump team has done a pretty good job of panning recent public polling, most of which has shown the president’s approval numbers have slid significantly, especially when it comes to his handling of the economy and inflation. 

They’ve argued forcefully that the president’s slippage in the polls is the result of bad samples mixed with media bias. Rigorous sampling is an important element of any good poll — but just claiming that polls should reflect a prior election result, and not taking into account the demographics of the survey, is questionable at best. And it is not a convincing defense of Trump’s decline in the polls.

The electorate isn’t static. It evolves.

There has been a lot of media speculation about how Trump’s poll numbers will affect his overall favorability and next year’s congressional elections. They wonder how people can still support Trump when they disapprove of the job he’s doing, particularly on the economy.

I would argue that voters are both smarter and more complicated than most politicians and the media think, and they can have conflicting beliefs on policy and politicians. 

Bill Clinton is a good example. In the 1998 exit poll, 20 percent of the electorate said they both approved of the job he was doing as president and also had an unfavorable view of him. Obviously, an unusual combination. 

However, for these voters, Clinton’s job approval proved more important, and they voted for the Democratic congressional candidate by a margin of 62 percent to 35 percent. As a result, this key voter group helped Democrats gain seats in a midterm election where Republicans should have gained ground, if history was any indicator. At the time, some Republicans were predicting the GOP would pick up 20 to 30 seats. 

Likewise, today a voter can be unhappy with Trump’s focus on tariffs rather than inflation, for example, but still support him and would vote for him again over then-Vice President Kamala Harris, as several recent media surveys have shown. That fact should be encouraging to Republicans, but it’s clear many Trump voters are frustrated and confused by the president’s tariff policies.

In the last Winning the Issues survey (April 8-10), Trump voters believed, 44 percent to 37 percent, that tariffs make inflation worse by increasing the cost of products. Clearly, they think the tariff policy negatively affects what they want to see happen on inflation. The question for the Trump team is how they are going to address this — but first, they need to recognize that the problem exists.

Neither Schumer nor Trump have cornered the market when it comes to the business of discounting unfavorable polls. Trashing polls one doesn’t like and ballyhooing those one does has become a cottage industry in recent years, whether it’s political or policy polling or survey research done by associations, foundations, lobbyists or business. 

For some, it all boils down to this. If the poll is good for your side, it’s a good poll. If it isn’t, it isn’t.

But there are other nuances to survey research worth considering as the polling wars continue. 

I’ve worked with all kinds of clients over the years, both political and policy oriented, and I’ve found that most fall into one of two categories, what I call the validators and the listeners. 

Validators want survey data that supports whatever narrative, candidate or policy initiative they’re promoting. Most aren’t interested in understanding the public’s views, but rather see polls as a way to validate their point of view on a particular policy or in a political battle — and of course, to raise money. They want polls constructed to deliver the results they want and ignore other conflicting factors. Unfortunately, there are plenty of pollsters willing to accommodate them. 

Ironically, for those who use polls to make their case on Capitol Hill, arriving with a “perfect poll” isn’t likely to get them a win. Most members, Republican or Democratic and their staffs, are great discounters when it comes to assessing the credibility of polls, public or private. They aren’t going to be snowed by polls crafted to prove a point.

When President Joe Biden tried to sell his American Rescue Plan and later the Inflation Reduction Act, his team liked to argue that his policies were wildly popular, despite public polls to the contrary. As proof, they pointed to internal polling that tested specific elements of the bills and purportedly supported their claims. It didn’t work.

Trump has put his own focus on “80/20” issues while he and his team have been in denial mode when it comes to a slew of recent negative media polls. While some may be questionable, many are not and should be listened to. But rather than listen to what voters are saying, they have instead called out the negative polls as biased or bogus.  

The better alternative to validation is to see polling as listening to the electorate to understand their priorities. Over the past decade, survey research has gone through a revolution of its own to meet the challenges of new technologies, from cellphone usage to internet polls and now AI. What hasn’t changed, however, is the ability of polls to deliver the kind of honest assessment and insight needed to develop strategies that address voter concerns — if we are willing to listen. 

David Winston is the president of The Winston Group and a longtime adviser to congressional Republicans. He previously served as the director of planning for Speaker Newt Gingrich. He advises Fortune 100 companies, foundations and nonprofit organizations on strategic planning and public policy issues, as well as serving as an election analyst for CBS News.