Here is my attempt at estimating the county baselines to see the absolute minimum we can pull in the counties to be able to break 50% statewide in California. Based on trends that I’ve pointed out many times over, close races in California are likely to get rarer and rarer, so I decided that now would be my best chance at showing what a close race in California could look like. Third parties usually pull between 3 and 5% of the vote, so I gave 4% to the third parties.
| County |
2008 # voters |
% of 2008 vote |
2008% Pres. |
Needed to reach 50% |
| State |
13,743,177 |
100.0% |
61/37 |
50/46 |
| Los Angeles |
3,368,057 |
24.5% |
69/29 |
58/38 |
| San Diego |
1,245,947 |
9.1% |
54/44 |
43/53 |
| Orange |
1,167,657 |
8.5% |
50/48 |
57/39 |
| Santa Clara |
678,033 |
4.9% |
69/29 |
58/38 |
| Riverside |
657,005 |
4.8% |
50/48 |
39/57 |
| Alameda |
628,545 |
4.6% |
79/19 |
68/28 |
| San Bernardino |
616,320 |
4.5% |
52/46 |
41/55 |
| Sacramento |
546,660 |
4.0% |
58/39 |
47/48 |
| Contra Costa |
456,876 |
3.3% |
68/30 |
57/39 |
| San Francisco |
388,112 |
2.8% |
84/14 |
73/23 |
| Ventura |
343,690 |
2.5% |
55/43 |
44/52 |
| San Mateo |
307,350 |
2.2% |
73/25 |
62/34 |
| Fresno |
275,554 |
2.0% |
50/48 |
39/57 |
| Kern |
235,854 |
1.7% |
40/58 |
29/67 |
| Sonoma |
231,817 |
1.7% |
74/24 |
63/33 |
| San Joaquin |
212,214 |
1.5% |
54/44 |
43/53 |
| Santa Barbara |
176,562 |
1.3% |
60/37 |
49/46 |
| Placer |
175,215 |
1.3% |
43/54 |
32/63 |
| Stanislaus |
162,941 |
1.2% |
50/48 |
39/57 |
| Solano |
162,638 |
1.2% |
63/35 |
52/44 |
| Marin |
141,321 |
1.0% |
78/20 |
67/29 |
| Monterey |
131,381 |
1.0% |
68/30 |
57/39 |
| Santa Cruz |
128,555 |
0.9% |
77/20 |
66/29 |
| Tulare |
106,551 |
0.8% |
41/57 |
30/66 |
| Butte |
99,392 |
0.7% |
50/47 |
39/56 |
| El Dorado |
93,890 |
0.7% |
44/54 |
33/63 |
| Shasta |
81,378 |
0.6% |
36/62 |
25/71 |
| Yolo |
80,674 |
0.6% |
67/31 |
56/40 |
| San Luis Obispo |
68,424 |
0.5% |
51/46 |
40/55 |
| Merced |
64,688 |
0.5% |
53/45 |
42/54 |
| Humboldt |
64,358 |
0.5% |
62/34 |
51/43 |
| Napa |
60,366 |
0.4% |
65/33 |
54/42 |
| Nevada |
56,177 |
0.4% |
51/46 |
40/55 |
| Madera |
43,032 |
0.3% |
42/56 |
31/65 |
| Mendocino |
40,580 |
0.3% |
69/27 |
58/36 |
| Imperial |
39,823 |
0.3% |
62/36 |
51/45 |
| Kings |
35,775 |
0.3% |
42/56 |
31/65 |
| Sutter |
33,337 |
0.2% |
41/57 |
30/66 |
| Tuolumne |
27,499 |
0.2% |
42/55 |
31/64 |
| Lake |
25,863 |
0.2% |
58/39 |
47/48 |
| Tehama |
24,803 |
0.2% |
36/60 |
25/69 |
| Calaveras |
23,588 |
0.2% |
42/55 |
31/64 |
| Siskiyou |
21,723 |
0.2% |
43/53 |
32/62 |
| Yuba |
21,681 |
0.2% |
41/56 |
30/65 |
| San Benito |
19,982 |
0.1% |
60/38 |
49/47 |
| Amador |
19,006 |
0.1% |
41/56 |
30/65 |
| Lassen |
11,510 |
0.1% |
31/65 |
20/74 |
| Plumas |
11,169 |
0.1% |
43/55 |
32/64 |
| Glenn |
10,053 |
0.1% |
38/60 |
27/69 |
| Mariposa |
9,756 |
0.1% |
42/55 |
31/64 |
| Del Norte |
9,684 |
0.1% |
45/52 |
34/61 |
| Inyo |
8,625 |
0.1% |
44/53 |
33/62 |
| Colusa |
6,532 |
0.0% |
40/58 |
29/67 |
| Trinity |
6,482 |
0.0% |
50/46 |
39/55 |
| Mono |
5,621 |
0.0% |
56/42 |
45/51 |
| Modoc |
4,505 |
0.0% |
30/67 |
19/76 |
| Sierra |
2,012 |
0.0% |
37/58 |
26/67 |
| Alpine |
697 |
0.0% |
61/36 |
50/45 |
Here’s a visual aid, comparing the expected results in a close race in California with the 2008 results:
Hypothetical close race

2008

This depends on a uniform swing, which while not unreasonable as an approximation, makes me wonder how we can do better. For instance, a heavily McCain county has fewer Democratic voters to swing, and so would probably have a smaller swing in a 50/50 election. This is most obvious in a hypothetical boundary situation: if there was a 95/5 McCain county in California, the baseline would be negative votes for the Democrat, a clear impossibility. It can similarly be reasoned that a close-to-boundary county would not also have a uniform swing: a 88/12 McCain county is not going to be as likely to vote 99% Republican as a 50/50 McCain district is to vote 61%.
the pale red counties blue, it looks like the 2004 election results
here’s one recent example. And another (this time pretty close to your map).