The Sunday before Election Day 2006, a Richmond Times-Dispatch headline screamed that polling showed the Marriage Amendment campaign had tightened. The poll said that the amendment defining marriage as between one man and one woman was supported by a slim 49-45 percent margin. That was the closest poll we had ever seen on the issue. Two days later, the amendment passed by a 14 point margin, 57-43 percent. How could the T-D poll have been so wrong just two days prior to the vote?

Polls taken over the years on the definition of marriage have wavered more than Tim Kaine on gay adoption (remember, running for governor in 2005 he opposed homosexual couples adopting, but now he's in favor of it). For example, Gallup polling on the issue of homosexual marriage went from 46 percent support in 2007, down to 40 percent in 2008 and 2009, but back up to 44 percent in 2010. So it doesn't surprise me at all that a Washington Post media poll of 1,000 people has found that, according to the Post, "Virginians are closely split on gay marriage" — and that the rest of the state's mainstream media ran with it.

But are they really?

The truth is that polls have been overwhelmingly disconnected from reality when it comes to the issue of homosexual marriage. One need look only as far as the 31 states that have had the issue put to the voters, and in every case the traditional definition of marriage has won, including California.

The longer I am involved in politics, the more dismissive I have become of most media polling. Many experts believe that, particularly on the issues of abortion and homosexuality, a lot of people tell a pollster what they think the pollster wants to hear. On the issue of same-sex marriage, while a few media polls indicate that people support it, in the 31 states where it has gone to the ballot the people have overwhelmingly rejected it. One might tell their neighbor they are open to homosexual marriage, but when the reality is in front of them in the voting booth, traditional marriage still resonates instinctively, intuitively, justly . . . morally.

Social issues such as abortion and homosexuality have dynamics at play that I don't think can be measured with simple media polling. Asking 1,000 people a simple question doesn't generally get to the core of complex issues. It makes for interesting editorial page fodder, but I doubt too many people take it seriously, except for the so-called "progressives" who will no doubt champion the media poll and bring the issue before the next General Assembly. I suspect some will even attempt to make it a campaign issue (funny, I thought it was all about the economy).

But I also find it interesting that the same "progressives" reject professional (not media) polling that shows an overwhelming number of Virginians support school choice. You see, polling can work both ways, which is why no one should base their beliefs or agenda on it. Sure, professionally done, in depth issue polling can provide insight, but hastily done media polls done over a weekend for the mainstream media isn't something I want to base any policy decision on. I certainly wouldn't want to base the future of our children on it.