This post was originally published on October 17, so I thought I'd trot it out for your thoughts and analysis on this Election Day, particularly as far as how things are looking in your own county.
Finally, this is your last chance to offer your final predictions for tonight's election.
Here are my final numbers:
- Governor: Beshear (D) 61 -- Fletcher (R) 39
- Attorney General: Conway (D) 62 -- Lee (R) 38
- Auditor: Luallen (D) 63 -- Greenwell (R) 37
- Treasurer: Hollenbach (D) 58 -- Wheeler (R) 42
- Secretary of State: no prediction -- too close to call
- Agriculture Commissioner: Farmer (R) 56 -- Williams (D) 42
.....................................................................................................
With 20 days until Election Day, and with the dynamics having largely played out in the governor's race, I think it's time to start making some projections on how things are likely to end-up on November 6th. (Note: for those of you who like getting into the weeds with data and numbers, you'll love this. The rest of you will probably glaze-over.)
I spent the past weeks looking at historical election data from some selected statewide races over the past decade to better understand the voters in the various counties. In particular, I pulled out statewide data from President Bush's (R) big margin in 2004, President Clinton's (D) narrow victory in 1996, Governor Fletcher's (R) 10-point win in 2003, and Jonathan Miller's (D) 12-point victory in 2003.
I picked these four races because they showed a broad range of the voters' behavior in each county over a period of a decade. For example, I picked Miller because he was the one Democrat that won handily in 2003, a year when the Republican wind blew very strongly. Miller's numbers provide a soft outer limit of how much support a Democrat could get in the various counties in such an election. Conversely, Bush's 2004 victory should provide an outer limit on the other side of the divide. Granted, counties have changed a great deal over a decade, so this is not science, but simply a rough sketch of counties have behaved in recent statewide elections.
As far as selecting counties, it turns out that if you look at the block of voters from the state's 20 largest counties -- but remove Laurel (#14) for Shelby (#28) -- not only do they account for almost exactly half of the registered voters but, as a group, they remarkably mimic both the statewide voter registration breakdown as well as the ultimate statewide results in these elections. Regardless of whether you're looking at the 1996 presidential, 2004 presidential, or 2003 gubernatorial elections, those 20 counties have been right on target.
So, let's take a look at those 20 counties. The number next to the county names is the current voter registration breakdown between just Democrats and Republicans. I omitted independent voters to get a cleaner gauge of how the counties are registered, versus how they've behaved on Election Day.
Finally, I've offered my projections on how I think the gubernatorial race will play-out in each county. While I think Beshear will outperform Miller's 2003 performance (when he won by 12) across the state, I think he will blow-out Fletcher in the Louisville media market given the Northup primary support and the polling which shows Fletcher trailing by vast margins in the region.
Using the single assumption that turnout in each of these counties will be exactly the same as the 2003 general election (by party), these projections would have Steve Beshear (D) defeating Governor Fletcher (R) by a 60.9% to 39.1% margin on Election Day, a margin that should mimic the statewide numbers if correct.
So, that's my on-the-record projections. How about yours? How do you think the various counties will vote next month? What's your projection of the final outcome?
While this is not offered as a scientific analysis of the race, I think looking at the history of some different elections provide us with enough information to make some assumptions.
Have fun.
| County (%R/%D) |
Bush v. Kerry (2004) |
Dole v. Clinton (1996) |
Fletcher v. Chandler (2003) |
Koenig v. Miller (2003) |
Fletcher v. Beshear (2007) |
| Boone (58/42) |
72/28 |
64/36 |
72/28 |
70/30 |
63/37 |
| Boyd (35/65) |
53/47 |
42/58 |
42/58 |
38/62 |
37/63 |
| Bullitt (38/62) |
68/32 |
53/47 |
59/41 |
49/51 |
45/55 |
| Campbell (49/51) |
64/36 |
58/42 |
63/37 |
60/40 |
57/43 |
| Christian (32/68) |
67/33 |
55/45 |
55/45 |
48/52 |
45/55 |
| Daviess (32/68) |
62/38 |
51/49 |
52/48 |
43/57 |
41/59 |
| Fayette (41/59) |
53/47 |
50/50 |
54/46 |
38/62 |
37/63 |
| Floyd (9/91) |
37/63 |
25/75 |
34/66 |
19/81 |
16/84 |
| Franklin (19/81) |
51/49 |
39/61 |
47/53 |
28/72 |
26/74 |
| Hardin (41/59) |
68/32 |
53/47 |
60/40 |
49/51 |
44/56 |
| Hopkins (24/76) |
66/34 |
47/53 |
57/43 |
44/56 |
42/58 |
| Jefferson (36/64) |
49/51 |
44/56 |
49/51 |
39/61 |
29/71 |
| Kenton (50/50) |
66/34 |
60/40 |
65/35 |
63/37 |
58/42 |
| McCracken (29/71) |
62/38 |
45/55 |
59/41 |
49/51 |
44/56 |
| Madison (41/59) |
62/38 |
53/47 |
59/41 |
41/59 |
42/58 |
| Oldham (58/42) |
70/30 |
63/37 |
68/32 |
58/42 |
54/46 |
| Pike (22/78) |
47/53 |
34/66 |
43/57 |
31/69 |
26/74 |
| Pulaski (71/29) |
77/23 |
69/31 |
73/27 |
63/37 |
59/41 |
| Shelby (38/62) |
67/33 |
53/47 |
60/40 |
45/55 |
44/56 |
| Warren (39/61) |
64/36 |
58/42 |
61/39 |
50/50 |
48/52 |
| Counties Total (39/61) |
58/42 |
49/51 |
55/45 |
44/56 |
39.1/60.9 |
| Statewide (39/61) |
60/40 |
49.5/50.5 |
55/45 |
43/57 |
? |
(One final note: As I did with the voter registration breakdown for the various counties, I removed Ross Perot's 8.7% of the vote in 1996 and only showed the percentage among those voting for either Clinton or Dole in 1996 to again allow for an easier read of the partisan movement in the counties for each election.)
Recent Comments