DATA TABLE TO ACCOMPANY “Count WHOSE Vote”

COLUMN #		1	2	3	4	5	6	7	8	9	10	11
Date	state	%dem	% clinton dems	% obama dem	ttl votes	ttl clinton	ttl obama	ttl clinton - obama	ttl dems	ttl clinton dems	ttl obama dems	ttl dems clinton - obama
8-Jan	NH	54%	45%	34%	287322	112,251	104,772	7479	155154	69819	52752	17067
26-Jan	SC	73%	28%	57%	532227	141,128	295,091	-153963	388526	108787	221460	-112672
2-Feb	AL	82%	40%	58%	543074	226,454	302,684	-76230	445321	178128	258286	-80158
2-Feb	AZ	78%	52%	42%	446098	228,158	191,681	36477	347956	180937	146142	34796
2-Feb	AR	78%	74%	25%	309506	217,313	80,774	136539	241415	178647	60354	118293
2-Feb	CA	80%	57%	38%	4434533	2,306,361	1,890,026	416335	3547626	2022147	1348098	674049
2-Feb	CN	80%	50%	48%	353515	164,831	179,349	-14518	282812	141406	135750	5656
2-Feb	DE	83%	42%	54%	96341	40,751	51,124	-10373	79963	33584	43180	-9596
2-Feb	GE	77%	32%	67%	1054831	328,129	700,366	-372237	812220	259910	544187	-284277
2-Feb	IL	79%	36%	63%	2016316	662,845	1,301,954	-639109	1592890	573440	1003520	-430080
2-Feb	MA	65%	58%	41%	1254537	704,591	511,887	192704	815449	472960	334334	138626
2-Feb	MO	73%	50%	47%	823280	395,287	405,284	-9997	600994	300497	282467	18030
2-Feb	NJ	78%	56%	42%	1119768	602,576	492,186	110390	873419	489115	366836	122279
2-Feb	NM	87%	51%	47%	149098	73,105	71,396	1709	129715	66155	60966	5189
2-Feb	NY	87%	60%	37%	1748833	1,003,623	697,914	305709	1521485	912891	562949	349941
2-Feb	OK	78%	58%	30%	417096	228,425	130,087	98338	325335	188694	97600	91094
2-Feb	TN	76%	57%	39%	618723	332,599	250,730	81869	470229	268031	183389	84641
2-Feb	UT	63%	47%	49%	124307	48,719	70,373	-21654	78313	36807	38374	-1566
9-Feb	LA	83%	38%	57%	384348	136,959	220,588	-83629	319009	121223	181835	-60612
12-Feb	MD	84%	40%	59%	777675	285,440	464,474	-179034	653247	261299	385416	-124117
12-Feb	VA	70%	38%	62%	979712	347,252	623,141	-275889	685798	260603	425195	-164592
	totals	78%	50%	47%	18471140	8586797	9035881	-449084	14366877	7125083	6733091	391992
12-Feb	DC	100%	24%	75%	113746	27,326	85,534	-58208	113746	27299	85310	-58010
	totals	78%	49%	47%	18584886	8614123	9121415	-507292	14480623	7152382	6818401	333981
29-Jan	FL	79%	52%	35%	1725264	857,208	569,041	288167	1362959	708738	477035	231703
	totals	78%	50%	46%	20310150	9471331	9690456	-219125	15843581	7861120	7295436	565684
15-Jan	MI	79%	60%	36%	593837	328,151	237,762	90389	469131	281479	168887	112591
	totals	78%	50%	46%	20903987	9799482	9928218	-128736	16312713	8142599	7464323	678276
Super Tuesday		78%	52%	45%	15509856	7563767	7327815	235952	12165143	6303351	5466434	836917
Through Super Tuesday		78%	51%	45%	16329405	7817146	7727678	89468	12708823	6481957	5740645	741312
Thru Sup Tues +FL+MI		78%	51%	44%	18648506	9002505	8534481	468024	14540912	7472175	6386568	1085606

Notes

“% dems” (Column 1), “% Clinton dems” (Column 2), and “% Obama dems”(Column 3) are based on the answer to the CNN exit poll under “Vote by Party ID”. The three choices were “Democratic”, “Independent”, and “Republican”. In most cases, this data can be found on the third page of each states exit polling data, which can be accessed from the CNN Election Central homepage (http://www.cnn.com/ELECTION/2008/ ) Because no exit poll data was available for the District of Columbia, the table assumes that all voters in the Democratic primary would have answered “Democratic” to the “Party ID” question.

“Ttl Votes” (Ttl = total) (Column 4), “ttl Clinton Votes” (Column 5) , and “ttl Obama votes” (Column 6) are based primarily on the data found that can be accessed from the Washington Post Election Guide Results page (http://politics.nytimes.com/election-guide/2008/results/index.html ) because their results included all candidates, while the CNN data does not. In certain cases, however, CNN had more complete numbers for Obama, Clinton and occasionally other candidates– in those cases, the CNN numbers were used, an “ttl Votes” was adjusted accordingly.

Michigan vote totals were assigned assuming that all uncommitted voters would have voted for Obama.

“Ttl Clinton-Obama” (Column 7) is the difference in the raw vote totals between Clinton and Obama. (Column 5 minus Column 6) A positive number indicates a state where Clinton had more votes, a negative number indicates that Obama received more votes.

“Ttl Dems”(Column 8) is based on the percentage of all voters who identified themselves as “Democratic” to the “Party ID” question (% dems) (Column 1 times Column 4)

“Ttl Clinton Dems” (Column 9) and Ttl Obama Dems”(Column 10) is based on the respective percentages of Clinton and Obama voters among those who identified themselves as “Democratic”. (Column 2 times Column 8, and Column 3 times Column 8, respectively)

“Ttl Dems Clinton – Obama” (Column 11) is the difference in the extrapolated Democratic vote totals between Clinton and Obama. (Column 9 minus Column 10) A positive number indicates a state where Clinton had more votes, a negative number indicates that Obama received more votes

In the total rows, the whole numbers are the sum of the votes in that category. “% Dems” is the sum of the “ttl dems” divided by sum of “ttl votes”. “% Clinton Dems” and “% Obama Dems” is the sum of “ttl Clinton Dems” and “ttl Obama dems” respectively, divided by the sum of “ttl dems”.

Caveat – the numbers for Democratic voters are, by necessity, inexact, insofar as they are based on whole number percentages. But because the rounding errors should balance out, the projected numbers themselves represent what should be a close approximation of the actual voters – assuming that the exit polling data is reasonably accurate.

Finally, the totals here difference from those put out by MSNBC at http://firstread.msnbc.msn.com/archive/2008/02/13/661146.aspx for three reasons

1) I assigned all the “uncommitted” votes from Michigan to Obama. MSNBC assigned them to “others.” It should be noted that a separate exit poll question asked if all the candidates were on the ballot, who would they have voted for. 46% said Clinton (vs 55% who did vote for Clinton) and 35% said Obama (vs 40% who voted “uncommitted). One can assume that Obama’s overall totals would have decreased slightly less than Clintons if these totals had been used. I chose to use 55%-40% because there was exit polling data relevant to party affiliation for those numbers.

2) MSNBC included numbers from the caucuses in its totals. However, those numbers reflect delegates to countywide (or similar) caucuses, rather than actual participation. Moreover, the method of determining the number of delegates assigned per local caucus varies widely between states – and sometimes, between counties in the same state. Finally, caucuses are in no way a reliable indicator of overall support for a candidate -- rather, they are a reflection of the effectiveness of a candidate’s “ground game”. (e.g. Obama received 79% of delegates of county caucuses in Idaho. It is inconceivable that, in a Democratic primary election in Idaho, he would receive 79% of the votes cast.) For all these reasons, only data from states that held primary elections are included in this data.

3) MSNBC’s numbers are based on less complete returns than the data that I used.