As a New York City resident, I rarely get to do meaningful on-the-ground political organizing. Our Democratic margins are so wide here, there’s no real point. It makes me miss being out on the campaign trail.

Which is why I spent Saturday canvassing the sub- and ex-urbs of Bucks County, Pennsylvania for Barack Obama. It was a beautiful spring day in a well manicured part of the country, full of flowering dogwoods and magnolia trees. But more invigorating than the weather was being face-to-face with swing voters in one of PA’s most hard-fought battlegrounds. Three interesting lessons:

1. There’s upside to a long fight. Republicans in Bucks County have had a majority in voter registration for as long as anyone can remember … until 2008. This year, due to the protracted Obama-Clinton fight, Bucks has had a surge of Democratic registration. The Blues now outnumber the Reds by 27,000 voters, a significant margin. This, I think, is the hidden gem of our stretched primary — new D voters flocking to the primary, who will stay D for the general.

2. Age is it. A voter’s age was the most predictive measure of whether a voter would choose Obama or Clinton. The voters we met this weekend in the “young” areas — subdevelopments filled with strollers, dogs and minivans — leaned heavily towards Obama. Those in the “older” areas leaned Clinton, despite being more wealthy. There’s been a lot of talk recently about race and gender splitting the electorate. This weekend reinforced for me the primacy of age.

3. Persuasion ID makes for bad data. Human beings hate to say no. At least half of the voters I spoke with told me they were still undecided. For some, that’s probably true. But I KNOW that many of them were Clinton supporters who simply didn’t want to disappoint the nice young man with the Obama sticker smiling on their doorstep. Thus, when I marked down the voter’s preference and reported it to the campaign, I know I was handing back imprecise data, at best. It makes me question the value of “persuasion ID” — the practice of trying to identify a voters preference at the same time as you try to persuade them to vote for your candidate. In a GOTV environment, its pretty crucial to not pull out your opponent’s voters. If your tactics are giving such shady data, you are wasting critical resources … or worse, helping your opponent. A better tactic might have been IDing through repeated phone poll attempts, and then a wider knock-and-drag strategy. It’s not like the Obama folks didn’t have the money to spend on live phone calls!

Overall, a great experience. Days like Saturday make me wish I lived in a swing state. Especially one with such good weather!