Princeton Election Consortium

A first draft of electoral history. Since 2004

Final Prediction: Kerry 311 EV, Bush 227 EV

November 1st, 2004, 12:00pm by Sam Wang


Here are my final calculations and predictions. Extended discussion and supporting links can be found in the November 1, 8:00AM update. I have a brochure for your use to follow the returns on Tuesday night and test my assumptions (and yours). Here it is.

The map below shows Kerry and Bush’s win probabilities, individually by state. The closer a state is to a tossup, the closer it will be to white. The map uses undecided and turnout assumptions; maps that do not make these assumptions can be seen by clicking in the box above. Note that unlike the map, the median projection takes compound events into account. Thus the map EV total is often not the same as the median EV total. An explanation is here. Click on the map for an interactive pop-up (thanks to Drew Thaler). For difficult browsers here is a static map and Ravi’s simpler calculator.

State-by-state electoral map

Overview: The basic calculation derives from polls only. Using statistical methods of meta-analysis, I use polls to calculate a starting point, referred to as “decided voters only.” This result is an uncorrected snapshot of where the polls stand. In addition to this, I estimate the effects of last-minute undecided/uncommitted voter decisions and differential turnout. My estimates are supported by evidence, but are by no means certain. Results based on uncertain assumptions are clearly labeled. To let you try your own assumptions, a table of medians with different bias values is given here.

The basic decided voter result. The median of Kerry 252 EV, Bush 286 EV among decided voters was calculated from 168 polls taken in 23 battleground states, and stepping probabilistically through all possible outcomes. Most of these polls were completed between October 25 and November 1. The EV estimate carries a large amount of uncertainty: the 95% confidence interval is ±39 EV. Thus, if only decided voters counted, the nominal Kerry win probability would be 20%, or 4-1 in favor of Bush.

Decided voters only (% Kerry win probability): AR 6, AZ 1, CO 1, FL 24, HI 67, IA 32, ME 100, MI 91, MN 82, MO 1, NC 0, NJ 99, NV 9, NH 93, NM 9, OH 30, OR 99, PA 85, TN 0, VA 0, WA 100, WV 0, WI 43.

Rank order of states: States currently in play in the 20-80% probability range, indicating a near-tie, are in bold. Turnout and how the undecideds break will shift which states are at a near-tie, but the order, from most Democratic to most Republican, should stay about the same.

Decided voters only: Democratic <- ME/WA/NJ/OR(95-100%) / NH / MI / PA / MN / HI / WI / IA / OH / FL / NV / NM / AR / CO/MO/AZ/TN/VA/WV/NC(0-5%) -> Republican

With undecided voters assigned. Undecided voters typically end up voting against the incumbent. In previous presidential races this has given a 2.5 ±2.0% advantage to the challenger. I currently estimate that 3.0% of voters are undecided. A 3-1 Kerry-Bush split gives a +1.5% net advantage to Kerry. This leads to a median EV estimate: counting undecided voters, Kerry 283 EV, Bush 255 EV, and a nominal Kerry win probability of 73%, or 3-1 in Kerry’s favor.

Turnout estimates and other corrections. The principal factor not measured by polls is turnout. Pollsters ask respondents questions to determine if they are likely to vote. However, this cannot capture efforts by voter turnout organizations. In addition, newly registered voters have no track record. Finally, telephone polls may not accurately sample the voting population. I estimate that these factors sum to an advantage in battleground states of 2 to 3% for Kerry. As I am sure you are all aware, this number cannot be known with certainty. With that caveat in mind, I use +1.5% as my turnout figure (see discussion at top of page). Combined with the +1.5% undecided allocation this makes a +3.0% bias as plugged into the MATLAB script. This leads to my final prediction. Predicted electoral outcome (11/1/2004 11:00pm EST): Kerry 311 EV, Bush 227 EV, nominal Kerry win probability 98%.

Note that all of these probabilites are conditional on the turnout and undecided voter assumptions being correct. The true probability is obtained by multiplying by a measure that is a function of whether my assumptions are accurate. The chance that I am wrong makes the true probability substantially lower than 100%! As Niels Bohr (and Yogi Berra) said, “Prediction is hard, especially of the future.” Just for the record, my gut estimate of the likelihood of a Kerry win is about 6-1 in favor.

Another way to look at it (suggested by Abe Fisher) is to turn the undecided+turnout question around: in order to get the Kerry win probability above 90%, the sum of these two factors must be 2.2%. Note that to get to exact even odds, the needed amount is the Meta-Margin, listed today as 0.8%.

Back to predictions: Based on the probabilites below, of the 23 states modeled, Kerry’s expectation value of states is approximately 14-15 of them, for a total of 23-24 states plus the District of Columbia.

Prediction, undecideds assigned, plus turnout (% Kerry win probability): AR 47, AZ 5, CO 20, FL 79, HI 97, IA 85, ME 100, MI 100, MN 99, MO 19, NC 0, NJ 100, NV 55, NH 100, NM 55, OH 83, OR 100, PA 99, TN 3, VA 7, WA 100, WV 5, WI 91.

The popular vote. To estimate the popular vote I use two approaches: (a) one based on presidential preference polls and (b) one based on Bush’s job approval numbers. In 16 national polls the medians (±SEM) are Bush 48.0 ±0.4%, Kerry 47.0 ±0.4%. Assuming 2.0% for Nader/other, the fraction of undeclared voters (“undecideds”) is 3.0%. Assuming Cook’s incumbent rule that undecideds split 3:1 for the challenger (2.25% and 0.75%), this gives a net of 1.5 ±1.2% to Kerry. This predicts a national popular vote (not corrected for turnout) of Kerry 49.3 ±0.9%, Bush 48.7 ±0.9%, Nader/other 2%. The second measure uses job approval ratings. In ten polls taken since mid-October the median ±SEM is 49.0 ±0.9%. Based on historical trends, this places an upper bound on Bush’s share of the popular vote. Thus, both approaches indicate that Bush’s popular vote share will be 49% or less.

I use the turnout factor to make a final estimate. Predicted popular outcome: Kerry 50.0%, Bush 48.0%, Nader/other 2%. National polls come from davidwissing.com, RealClearPolitics, and yougov.com. Job approval numbers come from pollingreport.com.

An electoral tie. A 269-269 EV tie would throw the election into the House and Senate, which would most likely lead to the re-election of Bush and Cheney. However, this would be an emotionally divisive event. The probability of an electoral tie is: Decided voters only, 4.1% (24-to-1 against). With undecideds, 3.1% (31-to-1 against). Final prediction with turnout included, 0.4% (273-to-1 against).

The power of your vote (the jerseyvotes calculation). Previously I have discussed where you are most effective in your door-to-door activism. My unit is the jerseyvote, which is the power of a New Jersey voter to influence the national election. Among decided voters only, the current value of a single vote in the top states is (measured in jerseyvotes): Hawaii 11,900, Iowa 7,500, Wisconsin 7,400, Florida 7,200, Nevada 7,100, New Mexico 5,800, Ohio 5,600. Counting undecideds, the top states still appear and Arkansas joins the list. Other values of relevance are (decided voters) New Hampshire 2,200 and Pennsylvania 2,500. As you can see, a jerseyvote’s value to American politics is what the Reichsmark’s was to the Weimar German economy.

Key states. These statistics summarize polls completed between October 25 and 31.

In Florida (14 polls), Bush leads in 8 polls, Kerry leads in 4 polls, and two polls are tied. Bush’s average (±SEM) margin is 1.4 ±0.9% in polls. I predict a Kerry victory by 2%. Polls close at 6:00-8:00PM Eastern time.

In Ohio (17 polls), Bush leads in 12 polls, Kerry leads in 4 polls, and one poll is tied. Bush’s average (±SEM) margin is 1.1 ±0.7% in polls. I predict a Kerry victory by 2%. Polls close at 7:30PM Eastern time.

In Pennsylvania (14 polls), Kerry leads by 2.1 ±0.7% in polls. I predict a Kerry victory by 5%. Polls close at 8:00PM Eastern time. General poll closing times are diagrammed here.

Bias analysis: The potential effects of differential turnout, splitting undecided voters, or systematic polling bias are as follows. The baseline from which bias is defined is decided voters only. Decisions by undecided voters and get-out-the-vote activities on Election Day will be major determinants of how large this bias effect is.

4 points towards Kerry: Kerry 325 EV, Bush 213 EV, Kerry win 99.9%.
3 points towards Kerry: Kerry 311 EV, Bush 227 EV, Kerry win 98%.
2 points towards Kerry: Kerry 291 EV, Bush 247 EV, Kerry win 86%.
1 points towards Kerry: Kerry 273 EV, Bush 265 EV, Kerry win 56%.
no swing (decideds only, flat turnout): Kerry 252 EV, Bush 286 EV, Kerry win 20%.
1 points towards Bush: Kerry 238 EV, Bush 300 EV, Bush win 97%.
2 points towards Bush: Kerry 217 EV, Bush 321 EV, Bush win 99.8%.
3 points towards Bush: Kerry 203 EV, Bush 335 EV, Bush win 99.99%.
4 points towards Bush: Kerry 186 EV, Bush 352 EV, Bush win 100%.

Tags: 2004 Election

No Comments so far ↓

Like gas stations in rural Texas after 10 pm, comments are closed.