I’ve done more analysis of proportional allocation, and it can, perhaps, be made to work, but it’s hokey. Let’s start with the raw popular vote (courtesy of ABC News’ website):

State	Electoral Votes	Bush	Gore	Nader	Buchanan
Alabama	9	940153	694734	17983	6303
Alaska	3	134829	63437	22601	4194
Arizona	8	673657	609937	40301	10903
Arkansas	6	469918	417949	13205	10936
California	54	4054756	5254500	372543	39897
Colorado	8	883858	738378	91461	10479
Connecticut	8	544704	789532	59950	4382
Delaware	3	137081	180638	8288	775
District of Columbia	3	17020	162004	9925	0
Florida	25	2910203	2909872	96844	17358
Georgia	13	1404058	1101101	0	10868
Hawaii	4	137785	205209	21609	1071
Idaho	4	336299	138354	0	7687
Illinois	22	2012154	2566489	103028	16060
Indiana	12	1231713	888201	0	17173
Iowa	7	628716	633969	27898	6942
Kansas	6	614419	391026	35583	7239
Kentucky	8	869946	637518	23125	4181
Louisiana	9	924670	789837	20817	14478
Maine	4	283988	315466	37757	4315
Maryland	10	770911	1093344	51078	4067
Massachusetts	12	876106	1610175	173758	11086
Michigan	18	1936288	2140755	81089	0
Minnesota	10	1110192	1168091	126579	22256
Mississippi	7	548634	399306	7890	2233
Missouri	11	1189521	1110826	38488	9806
Montana	3	239755	137264	24487	5735
Nebraska	5	408719	215616	22975	3431
Nevada	4	301539	279949	15004	4747
New Hampshire	4	273135	265853	22156	2603
New Jersey	15	1247791	1729021	92333	6868
New Mexico	5	258795	267501	19911	1347
New York	33	2218620	3742101	222075	33202
North Carolina	14	1607238	1236721	0	8971
North Dakota	3	175547	95723	9524	7330
Ohio	21	2294049	2117555	114474	25980
Oklahoma	8	744335	474326	0	9014
Oregon	7	667803	671178	68175	6490
Pennsylvania	23	2264309	2465412	102453	16879
Rhode Island	4	132212	252844	24115	2250
South Carolina	8	804826	578143	21008	3540
South Dakota	3	190515	118750	0	3314
Tennessee	11	1056480	977789	19694	4218
Texas	32	3796249	2427785	137706	12423
Utah	5	512161	201732	35661	9277
Vermont	3	119273	148166	19810	2182
Virginia	13	1426951	1216925	58864	5578
Washington	11	794567	883794	69578	4953
West Virginia	5	329708	291088	10440	3101
Wisconsin	11	1235991	1242115	92925	11206
Wyoming	3	147674	60421	0	2724
Total	538	48889821	49108420	2685168	442052

Now, if we allocate the electoral votes in each state in strict proportion to the popular vote in the state, we get this result:

State	Bush	Gore	Nader	Buchanan
Alabama	5.1	3.769	0.098	0.034
Alaska	1.797	0.846	0.301	0.056
Arizona	4.038	3.656	0.242	0.065
Arkansas	3.092	2.75	0.087	0.072
California	22.522	29.187	2.069	0.222
Colorado	4.101	3.426	0.424	0.049
Connecticut	3.116	4.516	0.343	0.025
Delaware	1.258	1.658	0.076	0.007
District of Columbia	0.27	2.572	0.158	0
Florida	12.26	12.259	0.408	0.073
Georgia	7.255	5.689	0	0.056
Hawaii	1.507	2.245	0.236	0.012
Idaho	2.789	1.147	0	0.064
Illinois	9.423	12.019	0.482	0.075
Indiana	6.916	4.987	0	0.096
Iowa	3.392	3.42	0.151	0.037
Kansas	3.517	2.238	0.204	0.041
Kentucky	4.535	3.323	0.121	0.022
Louisiana	4.756	4.062	0.107	0.074
Maine	1.771	1.967	0.235	0.027
Maryland	4.016	5.696	0.266	0.021
Massachusetts	3.936	7.234	0.781	0.05
Michigan	8.382	9.267	0.351	0
Minnesota	4.574	4.813	0.522	0.092
Mississippi	4.009	2.917	0.058	0.016
Missouri	5.571	5.203	0.18	0.046
Montana	1.766	1.011	0.18	0.042
Nebraska	3.14	1.657	0.177	0.026
Nevada	1.995	1.852	0.099	0.031
New Hampshire	1.938	1.886	0.157	0.018
New Jersey	6.085	8.431	0.45	0.033
New Mexico	2.363	2.443	0.182	0.012
New York	11.778	19.866	1.179	0.176
North Carolina	7.887	6.069	0	0.044
North Dakota	1.828	0.997	0.099	0.076
Ohio	10.583	9.769	0.528	0.12
Oklahoma	4.85	3.091	0	0.059
Oregon	3.307	3.323	0.338	0.032
Pennsylvania	10.74	11.694	0.486	0.08
Rhode Island	1.285	2.458	0.234	0.022
South Carolina	4.574	3.286	0.119	0.02
South Dakota	1.828	1.14	0	0.032
Tennessee	5.646	5.226	0.105	0.023
Texas	19.058	12.188	0.691	0.062
Utah	3.375	1.329	0.235	0.061
Vermont	1.236	1.536	0.205	0.023
Virginia	6.849	5.841	0.283	0.027
Washington	4.986	5.546	0.437	0.031
West Virginia	2.599	2.294	0.082	0.024
Wisconsin	5.265	5.291	0.396	0.048
Wyoming	2.101	0.86	0	0.039
Total	260.965	259.95	14.562	2.493

We can already see one oddity creeping in — though Gore has the most popular votes in the nation, Bush has the most electoral votes in this scheme, but not enough to elect him.

But fractional electoral votes seem hokey to me, so let’s simply round every candidate’s electoral votes to the nearest integer. In some cases, this results in an under- or over-allocation of electoral votes. And in all cases, Buchanan’s allocation is zero, so we’ll drop him from the rest of this discussion.

State	Bush	Gore	Nader	Difference
Alabama	5	4	0	0
Alaska	2	1	0	0
Arizona	4	4	0	0
Arkansas	3	3	0	0
California	23	29	2	0
Colorado	4	3	0	1
Connecticut	3	5	0	0
Delaware	1	2	0	0
District of Columbia	0	3	0	0
Florida	12	12	0	1
Georgia	7	6	0	0
Hawaii	2	2	0	0
Idaho	3	1	0	0
Illinois	9	12	0	1
Indiana	7	5	0	0
Iowa	3	3	0	1
Kansas	4	2	0	0
Kentucky	5	3	0	0
Louisiana	5	4	0	0
Maine	2	2	0	0
Maryland	4	6	0	0
Massachusetts	4	7	1	0
Michigan	8	9	0	1
Minnesota	5	5	1	-1
Mississippi	4	3	0	0
Missouri	6	5	0	0
Montana	2	1	0	0
Nebraska	3	2	0	0
Nevada	2	2	0	0
New Hampshire	2	2	0	0
New Jersey	6	8	0	1
New Mexico	2	2	0	1
New York	12	20	1	0
North Carolina	8	6	0	0
North Dakota	2	1	0	0
Ohio	11	10	1	-1
Oklahoma	5	3	0	0
Oregon	3	3	0	1
Pennsylvania	11	12	0	0
Rhode Island	1	2	0	1
South Carolina	5	3	0	0
South Dakota	2	1	0	0
Tennessee	6	5	0	0
Texas	19	12	1	0
Utah	3	1	0	1
Vermont	1	2	0	0
Virginia	7	6	0	0
Washington	5	6	0	0
West Virginia	3	2	0	0
Wisconsin	5	5	0	1
Wyoming	2	1	0	0
Total	263	259	7	9

Bush’s lead has increased, but we’re not done. Now, let’s take the under- or over-allocation in each state and assign it to the leader in that state.

State	Bush	Gore	Nader
Alabama	5	4	0
Alaska	2	1	0
Arizona	4	4	0
Arkansas	3	3	0
California	23	29	2
Colorado	4	4	0
Connecticut	3	5	0
Delaware	1	2	0
District of Columbia	0	3	0
Florida	12	13	0
Georgia	7	6	0
Hawaii	2	2	0
Idaho	3	1	0
Illinois	9	13	0
Indiana	7	5	0
Iowa	3	4	0
Kansas	4	2	0
Kentucky	5	3	0
Louisiana	5	4	0
Maine	2	2	0
Maryland	4	6	0
Massachusetts	4	7	1
Michigan	8	10	0
Minnesota	5	4	1
Mississippi	4	3	0
Missouri	6	5	0
Montana	2	1	0
Nebraska	3	2	0
Nevada	2	2	0
New Hampshire	2	2	0
New Jersey	6	9	0
New Mexico	2	3	0
New York	12	20	1
North Carolina	8	6	0
North Dakota	2	1	0
Ohio	11	9	1
Oklahoma	5	3	0
Oregon	3	4	0
Pennsylvania	11	12	0
Rhode Island	1	3	0
South Carolina	5	3	0
South Dakota	2	1	0
Tennessee	6	5	0
Texas	19	12	1
Utah	3	2	0
Vermont	1	2	0
Virginia	7	6	0
Washington	5	6	0
West Virginia	3	2	0
Wisconsin	5	6	0
Wyoming	2	1	0
Total	263	268	7

Now, Gore is back in the lead, but still short of a majority. So let’s allocate Nader’s votes to the leader in each state.

State	Bush	Gore
Alabama	5	4
Alaska	2	1
Arizona	4	4
Arkansas	3	3
California	23	31
Colorado	4	4
Connecticut	3	5
Delaware	1	2
District of Columbia	0	3
Florida	12	13
Georgia	7	6
Hawaii	2	2
Idaho	3	1
Illinois	9	13
Indiana	7	5
Iowa	3	4
Kansas	4	2
Kentucky	5	3
Louisiana	5	4
Maine	2	2
Maryland	4	6
Massachusetts	4	8
Michigan	8	10
Minnesota	5	5
Mississippi	4	3
Missouri	6	5
Montana	2	1
Nebraska	3	2
Nevada	2	2
New Hampshire	2	2
New Jersey	6	9
New Mexico	2	3
New York	12	21
North Carolina	8	6
North Dakota	2	1
Ohio	11	10
Oklahoma	5	3
Oregon	3	4
Pennsylvania	11	12
Rhode Island	1	3
South Carolina	5	3
South Dakota	2	1
Tennessee	6	5
Texas	19	13
Utah	3	2
Vermont	1	2
Virginia	7	6
Washington	5	6
West Virginia	3	2
Wisconsin	5	6
Wyoming	2	1
Total	263	275

And we see that Gore would be elected, which is the result I personally would like — but by this point, we’ve made so many arbitrary decisions to get integral electoral votes and a final result that I suspect going to the House would be a better answer.

In the wake of this week’s never-ending recount, there have been a number of proposals to change the way the President is elected.

One of those, direct election, would complete the transformation of the United States from a republic to a democracy (and it would obviously have ended up with a Gore win) — but it would need a Constitutional amendment, and, as the Curmudgeon pointed out on Wednesday, there are enough small states which would significantly lose power to make it unlikely that the necessary number of states would ratify such an amendment.

The other proposals retain the Electoral College. One of those would allocate each state’s electoral votes in proportion to the popular vote in that state, and the other would give one electoral vote to the winner of each Congressional district and two more to the winner of the state as a whole. Both of those preserve, to some extent, the power of the smaller states, but they avoid the huge amplification effect we’ve seen this year.

The proportional allocation proposal is very subject to choices in rounding rules — assuming each candidate must be given an integral number of electoral votes, you can still get very interesting results. For example, look at Minnesota’s totals:

Gore:    1110192  48.13%
Bush:    1168091  45.74% 
Nader:    126579   5.22%
Buchanan:  22256   0.92%

Minnesota has 10 electoral votes; should Nader get one of those votes? If so, should it be taken from Gore or Bush?

Allocating the votes on a district-by-district basis would avoid those particular errors. I can’t find district-by-district returns anywhere to do an analysis, though.

A Modest Proposal

I did some more analysis of what would happen if electoral votes were allocated in proportion to the popular vote in each state; it’s too long to put inline here, but feel free to read the details. In short, though, getting to a clear result requires a number of steps, and it’s not clear that it’d work any better than today’s system.

And that raises another question

If the election does go to the House, each state gets one vote. The Republicans control the House (both in the old and new Congresses), but which party controls more state delegations? I can’t find the data in a compact, crunchable form anywhere…does anyone reading this know?

Wine of the Night

When Jeffrey had his Walkathon last month, we bid on and won a French Wine and Chocolate basket. Tonight, after watching Duck Soup, we decided it was time to try it. The chocolate (Rumpelmayer Angelina 72%) was ok, but the wine was a real winner! It was a Chateau Mossé 1995 Vignes des Causses, Appellation Rivesaltes Contrôlée, and I highly recommend it with chocolate (it would probably go well with cheese, too, much like a port).

Since none of us liked the French chocolate much, we decided to see if any of the other chocolates would go well with the wine. My favorite was a Leysieffer Halbherbe Schokolade 55%, and Diane liked that one, too. We also tried three flavors of Droste pastilles (milk, bittersweet, and orange — I didn’t particularly like the bittersweet but the other two were OK, and Pete thought the orange pastille was the best chocolate we had), an 85% Leysieffer (which, though it was chocolate of the year in 1998, didn’t do much for any of us), Callebaut Extra Fine Milk Chocolate (Jeffrey’s favorite (no, he didn’t get the wine)), a Hershey bar (Debbie’s favorite), Dove Promises, and Double Chocolate Frangos. Not a bad way to spend an evening, and I can still type, even though we did kill the wine.

The online version of the story also omits the photo showing us and the remains of our lunch (I wanted to clear the tray, but the photographer said “no”), but through the miracle of modern technology, I can repair that omission.

And, of course, the online version will vanish in a day or so, while the paper version will live on in landfills for centuries to come. So I’ve captured the online version here, complete with original typos.

The New Yorker did better on accuracy when they did a story on “where e-mail goes” — their fact checkers called several times, once to verify that you could, indeed, see mountains when sitting in the IBM Almaden cafeteria. On the other hand, they didn’t like my actual title (Senior Technical Staff Member) and had me make up something slightly crisper (and I’ve forgotten what I made up), so you can’t rely on them 100%, either!

Free Ice Cream

We had a Webvan delivery this morning — as always, they were on time and courteous. Unlike previous deliveries, though, this time the ice cream was very soft; the driver said it had been in the freezer compartment on the truck, but “maybe there was no dry ice”. I called to complain, and the customer service rep immediately credited our account for the price of the ice cream and said she’d check with the warehouse to make sure they’re putting dry ice on the trucks. I’d really rather have had the ice cream arrive solidly frozen, though — tonight, we’ll have to do a quality assurance test on it.

When I went to bed last night, it looked like Bush had won, so when I picked up the paper a few minutes ago and the headline said “Too Close to Call”, I thought they’d gone to print before 11pm Pacific and didn’t pay any attention. But I decided to check the Web anyway, and was surprised when CNN said that there was going to be a recount in Florida.

It’s too late to do anything to affect the outcome, but I can still hope.

Media Lunch

Pete, Debbie, Diane and I had lunch at Falafel’s Drive In, a San Jose institution; while we were finishing up, the people at the next table introduced themselves as a reporter and photographer from the San Jose Mercury News and wanted to know if they could ask us a few questions about our reaction to the election and take some photos.

We were game, and wound up talking with them for about 15 minutes, describing how we were following the news last night on TV and Internet, and how I tried to get news during dinner using my cellphone and Sprint’s “Wireless Web” but failed (and I tried again while we were talking and got similarly useless stories), and how Jeffrey was so interested he wouldn’t go to bed until after 11. I don’t know if any of our conversation will actually make it to the paper; I guess we’ll find out tomorrow, probably before we know who got elected!

The reporter did like Pete’s quote: “It felt like we had flipped a coin and it landed on its edge.”

How many votes does usability cost?

The Curmudgeon has a truly informative interactive graph showing just how far out of balance the Buchanan vote in Palm Beach County was compared to the rest of the state. And Dan Bricklin’s discussion of Ballot Usability in Florida is worth reading, and the South Florida Sun Sentinel has a long article from the scene of the crime (I guess that’s a bit judgmental on my part, isn’t it?). And the Jerusalem Post says
“in the end, it came down to bubbe and zayde“.

Non-political commentary site of the day

Microsith. May the farce be with you, and thanks for the pointer, Doc.

Finally, it’s Election Day. Soon, I’ll be able to stop tossing the hit pieces from the various candidates into the recycling bin and get on with the usual work of the season: tossing holiday catalogs from various companies into the recycling bin.

Vote early, vote often, vote Gore!

Diane and I just voted — the precinct is very conveniently located at Jeffrey’s school, so we usually vote at 8:30am, right after walking him over to school. This morning, there was a longer line to vote than I’ve ever seen, and the election worker who signed me in said that it was a very heavy turnout so far. I’m always surprised that I’m not asked to show ID to vote, just to sign, though the people working the precinct here have been doing it for years and years and know most of the people in the neighborhood by sight if not necessarily by name.

The house directly across the street from the school has an enormous Bush sign in its yard; I wonder how far from the actual polling place electioneering is prohibited? The answer: Any “electioneering” must be conducted from a minimum of 100 feet from the place where people are voting. (EC &167;18370), as quoted on the California Secretary of State’s website. I’m pretty sure that this sign is at least that far away, since it’s on the house, not at the front of the yard, and all the way across the street — my GPS isn’t accurate enough to measure it that precisely, I’m afraid.

How the hell am I supposed to decide?

I got a note yesterday from “Paid Prescriptions, LLC,” the manager of my prescription drug benefit program, telling me that they’d suggested to my doctor that I be switched from Nasarel to Nasacort, a “preferred medication”, and that he had agreed to the switch. They also mention that “Your decision to change to the preferred medication is voluntary and does not determine benefit coverage.”

How am I supposed to know whether to accept this change or not? If the doctor had prescribed Nasacort in the first place, I would’ve been just as happy as I am with Nasarel, I’m sure, but now I have to wonder if the substitution is really a good idea or not.

Al?
James? Any comments? (Not necessarily on this particular substitution, but on the general case.)

The Wireless Web loses the election

We went out for dinner tonight, and naturally we were all interested in what election news was available. So I used my Sprint phone to check the CNN.com “Wireless Web” site and was horribly disappointed — the news was old and hard to read; in fact, they didn’t even bother to give a total electoral vote number in the stories they ran. The New York Times’ site was worse — it still had the morning’s stories. So much for modern technology. I’m glad to be back home with a connection to the real Web, though I’m not happy with some of the news I’m reading (especially Florida).

We seem to have a political junkie in the house

Jeffrey is having another sleepless night — he’s been watching TV all night (as have we all), and just came out of his room to tell us to let him know if they announce who’s going to become President.

I’m afraid he, and all of us, are going to be waiting a long time. I just hope the result isn’t what I’m afraid it’s going to be, with another Bush league presidency looming.

The people have spoken

Or at least the media has decided that the people have spoken. We survived Reagan. We survived Bush the first. We should be able to survive Bush the second. I hope.