An otherwise perfectly polite campaign manager challenged me rather forcefully last week over our use of polls.

It turns out she couldn't understand why we conducted them and questioned our seeming bias in doing so.

"It's OK when you report the news," she said. "But it's not OK when you make the news."

I'd never heard this line of reasoning before. Political polls are by no means perfect, but they are a perfectly accepted form of knowing where voters stand. Public opinion polls are part of our landscape. They tell us whether President Bush is doing well (he's not) or whether we should still be in Iraq (not much longer) or how the economy is doing (pretty well, in Hawai'i's case). Our own Hawai'i Poll asked voters about important races for U.S. Senate, the U.S. House and governor as well as how they felt about Native Hawaiian issues and local problems like traffic and affordable housing.

The animated campaign strategist I ran across believed it was better for us to take the polls conducted by her candidate's polling company and report those as news but refrain from conducting our own. That's a more peculiar position to take. Can we really trust a poll by somebody who has a vested interest in the outcome?

We have long debated at The Advertiser the import and meaning of polling. More and more people have caller ID and probably are not going to pick up the phone if they know a pollster is at the other end. Since cell-phone numbers are not called and more people than ever have them, how credible are the results? And even if the numbers are credible, are we somehow influencing the outcome in the same way that early precinct results on the East Coast in presidential races tend to make voters farther west skip the contest altogether?

Polls are easy to second-guess because you can always measure the results against reality.

Here's what we reported a month before the 2004 presidential election:

"President Bush and Sen. John Kerry are deadlocked among likely voters in Hawai'i, a surprising boost for the Republican president in a state that many Democrats had considered safe for Kerry. The Hawai'i Poll ... had Bush at 43.3 percent and Kerry at 42.6 percent. The margin of error was 4 percentage points. A large number of voters, 12 percent, said they were still undecided, giving supporters of both candidates hope during the final days of the campaign."

When all was said and done, Kerry finished with about 54 percent and Bush 45 percent.

Likewise, our October 2004 poll for the Honolulu mayoral race showed Duke Bainum leading Mufi Hannemann 48.3 percent to 39.9 percent, with 11.7 percent undecided.

We all know how that one turned out.

The point isn't that polls are flawed. They're just a snapshot in time reflecting what people think at the moment they are being questioned. Views and events can change quickly from the time polls are developed, results are tabulated and votes are cast.

I would argue that our four days of poll results were fascinating. Hawai'i's views on Iraq, Ed Case, Dan Akaka, Linda Lingle, Mufi Hannemann, the economy and so on are reflective of something voters are feeling, at least on the day the polling is conducted. It's fun for people to debate what the numbers mean or don't mean, and we intend before the September primary to get another peek into voters' minds.

To respond to the campaign manager's complaint, we're not making the news because we have no vested interest in how voters respond. It's more likely that candidates with their own polls are framing the questions with intended results in mind.

Several years ago, we commissioned an excellent series called "The Vanishing Voter" that detailed over many installments the problems Hawai'i historically had getting people to the polls. We got some good response, but the overall reaction was a collective shrug. It seems that nobody had a good answer about how we might get people in this state to care about voting.

One way for The Advertiser to keep them interested is to make sure the races and the issues are fresh in the public's mind. That's why we spend so much effort developing comprehensive voters' guides and devote so much space in print and online to issues-oriented pieces. Part and parcel of that coverage is knowing how voters feel about the issues.

Polls are a useful way to find whatever resembles a political pulse in a state that was dead last in voter turnout the past two presidential elections.

Mark Platte is vice president/editor of The Honolulu Advertiser. Reach him at mplatte@honoluluadvertiser.com or 525-8080.