The Week In Polls--April 27--A Look Inside The UNH Granite State Poll

            As promised, I tore myself away from a baseball book (the 1941 season in this case--perhaps the greatest year ever--Williams bats .406, DiMaggio's 56 game streak, a great NL pennant race) last night and spent a good deal of time analyzing the internals of the UNH Granite State poll released this week.

            Unlike the ostrich-like Republican leader in the New Hampshire House, I would stress that Republicans choose to pooh-pooh these results at their own peril.  In fact, even though the sample is slightly skewed to Democrats, the news is better for Republicans than last week’s Dartmouth poll which was heavily skewed for Republicans.

            Obama is probably not going to win New Hampshire by the nine point margin the poll indicates, but there is every reason to believe that’s the lead he enjoys at this point.  Even though Romney leads by 1.5 points in the Dartmouth poll, Obama has a seven point lead with independents (undeclared) in that one, and make no mistake, the undeclared group will turn out in big numbers this year and will determine winners at all levels.

            Andy Smith’s results do not indicate that it’s going to be a big year for Democrats.  Far from it, they tend to confirm what I’ve been saying all along, that it’s going to be very even in the state come November.  Of course, that’s bad news for Republicans only because they have such a big advantage now.

            Even in the House would mean 200-200 or a loss of 98 Republican seats from November 2010.  Even in the Senate would mean 12-12 or a loss of seven seats.

            Smith’s data shows the generic race for State Senate is dead even (37-37) and that Democrats enjoy a slight advantage for House races (37-34).  However, he extrapolates that data to mean that Republicans would still hold control of the House 212-188 and of the Senate 14-10.

            Although that seems counterintuitive, it’s really not because Democrats tend to win their areas (Durham, Hanover, Keene, Concord, Plymouth) by huge percentages while Republican victories overall (with exceptions like Bedford of course) are by narrower margins.  Thus, Democrats could get higher percentages overall while picking up fewer seats (the same reason that Al Gore won the popular vote by half a million in 2000 but lost the electoral vote to George W. Bush).  In other words, Republican votes tend to be more “efficiently” spread out.

            Rather than take 298-102 and 19-7 as a starting point, it might then be more reasonable to take 212-188 and 14-10 as starting points as of today.  Obviously, a small change in places like Hudson and Merrimack could then effect a significant number of House seats (block voting still prevails despite the demise of the old-fashioned straight ticket mark), negatively for Republicans.  However, Democratic seats are far “safer” at the outset.

            Similarly in the Senate, getting beyond ten will be easy for Democrats with a small change (such as the news that John Gallus is not running in district one).

            The surprising thing out of the UNH poll, at least for me, is that there are really no surprises.   It’s about what I expected.

            Here are a few numbers (favorables are always listed first; don’t be surprised by the large number of undecided; the margins are more indicative at this time as are the year-to-year changes).

NH Legislative Approval

All—40-42 (much better for Republicans than the Dartmouth poll)

A Year Ago—42-38

Republicans—56-27

Democrats—28-58

Independents—41-38

Speaker Bill O’Brien Approval

All—15-27 (negative 12)

Year Ago—16-25 (negative 9)

Republicans—24-16

Democrats—7-40

Independents—18-19

Senate President Peter Bragdon Approval

All—8-14 (negative 6)

A Year Ago—13-11 (plus two)

Republicans—9-7

Democrats—6-22

Independents—11-11

Governor Lynch Approval

All—72-17 (plus 55)

Republicans—59-28 (plus 31)

Democrats—87-7 (plus 80)

Independents—64-19 (plus 45)

President Obama Approval

All—50-47 (plus 3)

Republicans--11-85 (negative 74)

Democrats—90-9 (plus 81)

Independents—36-59 (negative 23)

Note how Obama is significantly underwater with Independents.  This is not good news for the Presidents and is clearly an area in which only the over sampling of Democrats produced the overall positive number.

The UNH poll breaks each question down by age, region of the state, income level, religious participation, and even such matters as whether the person reads the Union Leader or the Boston Globe.  While there are significant differences in some areas (O’Brien does well with people who have lived here less than ten years, terribly with longer term residents for example), such data is more for wonks than the average reader, so I’ll skip it here—except to note geographic locations. 

Since State Rep and Senate predications are based in large part on area, let’s look at the State Rep numbers by region.  This provides evidence for what I’ve been saying all along, that Manchester Republicans (yes, I am one of them) could very well be nearly completely wiped out.  Note that the Mass border is the only area favorable to Republicans.

State Rep Preference (R/D in all cases)

North Country—31-39 (+8D)

Central/Lakes—32-32 (Tie)

Connecticut Valley—34-41 (+7D)

Mass Border—44-30 (+14R)

Seacoast—30-38 (+8D)

Manchester Area—27-48! (+21D)

             The entire poll, more than you’ll probably ever want to know, is available at the UNH survey center web site.  I had previously reported on the gubernatorial match-ups, rather stunning with Democrats up slightly, but it's way too early to go into much more detail.