Libertarian Party candidate Michael Badnarik's campaign announces its intention to spend $20,000 on polling; skeptics wonder why a sure-to-lose campaign should bother. Thomas Knapp at Rational Review gives some reasons:
In a number of states, Bush…Kerry…are so closely matched that those states' electoral votes may turn on a few thousand, or even a few hundred, votes. That's what happened in Florida in 2000. In New Mexico that year, Bush lost the state by 300 votes or so—and Libertarian candidate Harry Browne polled more than 2,000. You do the math.
It is entirely possible that Michael Badnarik's popular vote totals will determine the outcome of this presidential election … and that's where the polling comes in.
Polling will allow the campaign to determine not only which states are statistically tied between Bush and Kerry, but which of those states are potentially good places for Badnarik to spend time, money and effort. Kerry and Bush are close in Ohio. Is Badnarik doing well enough to possibly "bridge the gap" and force the state in one direction or another? If so, he should be focusing some attention there. If not, then perhaps Arizona, Oregon or Missouri are better places for him to stump between now and November.
Likewise, polling will allow Badnarik to find out what issues he can best appeal to the voters on…The tool for getting that information is: polling.
The use of polling data to direct the campaign's efforts isn't a waste of money. It is a necessary prerequisite to spending the campaign's money wisely, so as to maximize the campaign's impact.
It still seems more likely to me that a Bush loss that the media could "blame" on Badnarik, based on the ridiculous premise that all his votes "really belonged" to Bush, will cripple the LP than propel it to greater plateaus. But it could very likely mean a more libertarian GOP down the line–which in our first-past-the-post system will probably be ultimately more helpful for libertarian ideas in electoral politics than the LP has managed to be so far.