Polls:

I’d be happy to see Tom Daschle be defeated in his Senate race, but this item doesn’t much excite me:

Daschle Slipping in SD Polls

In February, an Argus Leader/KELO-TV poll showed Daschle defeating Thune 50%-43%. A new poll just taken by the Rapid City Journal/KOTA/KSFY shows that Daschle is backsliding to 48%-43%. Note, Thune hasn’t run one ad and just entered the race a few months ago! Daschle, on the other hand, has been running TV ads since June of 2003–$6 million worth! To spend $6 million in a small media market and slip in the polls indicates that Daschle has a lot of problems. No wonder Daschle freaked out over Tim Giago entering the race as a third party candidate.

Repeat after me: “I will not think that statistically insignificant changes in poll results are statistically significant — even if I really, really like them.”

     (If you want the details, the 50-43 poll claims “a margin of error of plus or minus 3.5 percentage points,” and I suspect the 48-43 has a comparable margin of error. And, yes, I realize that one can argue that even when the 50-43 to 48-43 change is seen as no change at all, this might still be bad for Daschle. But it’s not accurate, I think, to call this “slipping” or “backsliding.”)

UPDATE: Glen Whitman takes a different view; but while his point may be sound in some situations, I stick by my analysis in this case.