Expected turnout in Ohio

According to today's Columbus Dispatch, nearly 5.8 million people are expected to vote in Ohio this week.  This is approximately 73 percent of nearly 8 million voters registered in the state.

The 5.8 million voters will exceed the state record turnout (set in 1992) by about 750,000, and will be about a million more than the number from four years ago.  The 73% turnout will fall short of 1992's 77% turnout.  More than a million new voters have been registered in Ohio this year.

Voting patterns by age...

The front page of yesterday's USA Today included a graphic on percentage of registered voters and adult citizens who voted four years ago.  The figures were...

Registered Voters
Age 18-34, 77% voted
Age 35-54, 88% voted
Age 55-74, 90% voted
Age 75 and up, 85% voted

Adult Citizens
Age 18-34, 44% voted
Age 35-54, 63% voted
Age 55-74, 71% voted
Age 75 and up, 67% voted

So, according to these figures, the percentages within each age group who voted, were registered but did not vote, and were not registered were approximately...

Age 18-34:  44% voted, 13% registered but did not vote, 43% not registered
Age 35-54:  63%, 9%, and 28%, respectively
Age 55-74:  71%, 8%, and 21%, respectively
Age 75 and up:  67%, 12%, and 21%, respectively

One can only hope that the figures for the youngest demographic change this year.  So far, there is every reason to think that they will.  Certainly the "not registered" figure should be down from four years ago.

Battleground simulation

For the purpose of this analysis, Kerry started out with 231 EV, which includes New Jersey, Hawaii, Washington, Michigan, Oregon, Pennsylvania, New Hampshire, Maine (at large), and Maine (District 1).  Bush started out with 216 EV, which includes Missouri, Tennessee, Louisiana, Arizona, North Carolina, Virginia, Colorado, and West Vieginia.  Obviously, some of these states could shift, but I disregarded this possibility in the interest of keeping the analysis to a manageable size.  A total of 91 EV are at stake in the remaining states.

The following probabilities of a Kerry victory were assigned to each remaining state...

Florida (27 EV), 0.50
Ohio (20 EV), 0.70
Minnesota (10 EV), 0.80
Wisconsin (10 EV), 0.50
Iowa (7 EV), 0.50
Arkansas (6 EV), 0.40
New Mexico (5 EV), 0.70
Nevada (5 EV), 0.20
Maine District 2 (1 EV), 0.60

Kerry wins with at least 39 of the 91 EV at stake, while Bush wins with at least 53.  I considered the dreaded 269-269 tie as a win for Bush.

After all the numbers were run - and no, I'm not going to list out 512 possible scenarios - Kerry has a 74% probability of winning the election, compared to 26% for Bush.  It was not possible for Bush to win without both Florida and Ohio.  Kerry could win without either state, by holding Minnesota, Wisconsin, Iowa, Arkansas, and at least two of New Mexico, Nevada, and Maine District 2.

Be still my beating heart...

Kerry has taken a 2-point lead according to the Rasmussen tracking poll.  Today's figures show Kerry at 48.4, Bush at 46.4, Other at 1.7, and Not Sure at 3.5.

This is Kerry's first lead in two months.  Just three days ago, Bush led 49.1 to 45.9.


Yeah, I know, still too soon to break out the champagne.  But isn't it strange that to get good news from Rasmussen and bad news from Zogby?

A simplistic analysis of the five major battlegrounds

FWIW, this is my first diary on MyDD.

For the purposes of this analysis, I considered only Florida, Ohio, Minnesota, Wisconsin, and Iowa.  Kerry started with 237 electoral votes, which included the swing states of Michigan, Pennsylvania, Oregon, New Hampshire, and Maine.  (The potential loss of one EV in Maine had no impact on this analysis.)  Bush held all other states, for a total of 227 electoral votes.  So, Kerry wins the election with at least 33 of the 74 electoral votes at stake in the five states.

Next, I assigned probabilities of a Kerry win to each state.  Currently, Florida, Iowa, and Wisconsin look like true tossups, so I assigned a probability of 0.5.  Kerry appears to be doing well in Ohio now, so I assigned a probability of 0.7.  Finally, I assigned a probability of 0.8 to Minnesota, both for current polls and past history.  There are few more reliably liberal states than Minnesota.  The states were considered statistically independent.

I have no intention of listing all 32 possible combinations of outcomes (and their probabilities), so here is a summary of results...

Kerry gains at least 33 electoral votes in 19 of 32 scenarios.  Considering the probability of a win in each state, the overall probability that Kerry gains at least 33 electoral votes is 72%.  The expected mean total is 44 votes out of 74.

No one state of the five guaranteed a Kerry victory.  Kerry won every scenario that included wins in Florida and at least one of the other four states.  Kerry also won every scenario in which he lost Florida, won Ohio, and won at least two of the remaining three states.


Advertise Blogads