Bloomberg effect

Pollster.com has the results of a 16-state survey of 2-way and 3-way matchups between Giulaini, Clinton, and Bloomberg. First, the data:

	3-way			2-way		Bloomberg Effect
State	Giuliani	Clinton	Bloomberg	Giuliani	Clinton	From G	From H
Alabama	46	39	11	53	41	-7	-2
California	40	45	10	44	49	-4	-4
Iowa	37	42	11	41	47	-4	-5
Kansas	47	36	8	53	41	-6	-5
Kentucky	42	41	10	47	44	-5	-3
Massachusetts	37	47	9	42	52	-5	-5
Minnesota	37	48	7	41	50	-4	-2
Missouri	39	44	10	47	46	-8	-2
New Mexico	41	45	8	44	50	-3	-5
New York	32	49	15	38	56	-6	-7
Ohio	41	47	8	46	49	-5	-2
Oregon	38	44	11	44	48	-6	-4
Texas	48	34	10	54	37	-6	-3
Virginia	45	40	9	48	44	-3	-4
Washington	41	42	11	47	44	-6	-2
Wisconsin	40	44	10	46	47	-6	-3

Several things to notice. First, the differences created by a Bloomberg candidacy are real. While there is plenty to be suspicious of, here are some thoughts:

1. Even without Bloomberg, Giuliani loses nearly every swing state. Giuliani would win only Missouri and Washington. Note that WV, CO, NV, AZ, PA, MI, and FL are not included in this sample.
2. Bloomberg seems to take his votes from both sides, but somewhat more from the GOP. However, in the swing states, the damage seems to be almost 2-1 against the GOP. However, these results are almost all within the margin of error.
3. It is going to take more data to figure this out.

But this is not good for the GOP with or without Bloomberg