The State of the Race: 2004 vs 2000
by CA Pol Junkie
We've seen alot of bad polls out there recently, but here is a good one. ARG has polled all 50 states, calling enough people to get 600 responses from every state. That's 30,000 interviews over the phone - definitely not a cheap poll to run. ARG has a good reputation, and the results look highly credible:
National: Bush 47, Kerry 46
Here are the states in order from best to worst with Kerry's margin in the poll:
Strong/Safe Kerry: DC+67, MA+37, RI+28, CT+15, NY+12, CA+11, HI+10, VT+10
Lean Kerry: DE+9, MD+9, MI+8, NJ+8, WA+7, IL+6, NM+5
Slight Kerry: ME+4, MN+2, OR+2, PA+1, FL+1
Dead Even: WI, WV
Slight Bush: CO-1, NV-2, NH-2, IA-2, OH-2, AR-3
Lean Bush: NC-5, AZ-6, MO-6, VA-6, TN-7, LA-8, MS-9
Strong/Safe Bush: GA-11, SC-12, AL-14, IN-15, OK-17, KY-18, SD-19, KS-22, TX-22, AK-27, MT-28, ID-29, ND-29, NE-31, WY-36, UT-37
To give us some perspective, I googled up results of ARG's 50-state poll released September 21, 2000. At just about this time four years ago, Bush caught and then passed Gore in the national polls. At the time this ARG poll was conducted, however, Gore was leading nationally by 0-8 points (depending on the poll), so figure an overall bias in Gore's favor of about 4 points. Here were the results for key states:
Strong/Safe Gore: DE+16, ME+15, IL+10
Lean Gore: WA+8, IA+7, LA+7, KY+6, OR+6
Slight Gore: WI+4, FL+3, MO+3, PA+3, AR+1
Slight Bush: NC-1, CO-2, OH-2, NV-3
Lean Bush: NH-5
That lets us identify trends over the last four years and gives us some expectation of what would happen in a close national outcome on November 2nd. Both Bush and Kerry are following more or less the same electoral strategy as in 2000: fight like hell in the Midwest and Florida and ignore the rest. Following are the trends from 2000 to 2004 for key swing states, corrected for the 5 point difference in national performance between the polls. Given a repeat of the essentially tied national outcome of 2000, we can predict how Kerry will do relative to Gore and thus on which side purple states will fall:
State / 2000-2004 Trend / 2000 Outcome / 2004 Outcome
Illinois: Trend +1 / 2000 Result +12 / Projected: +13
Delaware: Trend -2 / 2000 Result +13 / Projected: +11
Washington: Trend +4 / 2000 Result +6 / Projected: +10
Pennsylvania: Trend +4 / 2000 Result +4 / Projected: +8
New Hampshire: Trend +9 / 2000 Result -2 / Projected: +7
Florida: Trend +4 / 2000 Result -0 / Projected: +4
Ohio: Trend +6 / 2000 Result -4 / Projected: +2
Oregon: Trend +1 / 2000 Result +1 / Projected: +2
Nevada: Trend +5 / 2000 Result -4 / Projected: +1
Wisconsin: Trend +1 / 2000 Result 0 / Projected: +1
Colorado: Trend +7 / 2000 Result -8 / Projected: -1
Maine: Trend -6 / 2000 Result +5 / Projected: -1
Arkansas: Trend +2 / 2000 Result -5 / Projected: -3
Iowa: Trend -4 / 2000 Result +0 / Projected: -4
Missouri: Trend -3 / 2000 Result -3 / Projected: -6
North Carolina: Trend +2 / 2000 Result -13 / Projected: -11
Louisiana: Trend -10 / 2000 Result -8 / Projected: -18
Kentucky: Trend -18 / 2000 Result -15 / Projected: -33
OK, I guess Kerry won't lose Kentucky by 33 points, but the method shows where Kerry is currently over- and under-performing relative to Gore. Since Gore got 260 electoral votes in 2000 (adjusted for the reapportionment), under this analysis Kerry would add NH, FL, OH, and NV while losing ME and IA for a net gain of 45, giving Kerry 305 and the White House. It looks like he could lose the popular vote by a point or so, giving up NV and WI, and still win. While it would be sweet justice to beat Bush the same way he "beat" Gore, let's hope Kerry cleans Bush's clock in the debates so the only suspense on November 2 is whether or not South Carolina goes blue.