The Florida Department of State's Divisions of Elections results page has been updated yet again. Sixty-three counties (out of 67 total) are listed on the County Reporting page as having updated since my last published data collection at 11pm on Nov. 10, 2004. As with all prior days, the "County Reporting" page still has a "Y" in the "Final Report" and "Absentees" columns for all counties. No counties are listed as having updated recently on the "Provisionals" page. My data tables for this most recent update can be found here.
Summing across all counties, voter turnout increased by a net total of 16,716, while 16,781 presidential votes were added (table 1). As in my previous post, no counties are reporting more votes for president than voters voting, though Osceola County still has exactly the same number of votes for president and voters voting (table 2; they last updated 11/13/2004 10:34). Counties that added or removed votes for president again appeared to add them to or remove them from a mix of candidates (table 3).
For my prior posts on Florida county election data, see here, here, and here.
Saturday, November 13, 2004
Recounts in NH and OH
Some third party candidates are asking for recounts in both New Hampshire and Ohio (and no, not because they think they should have won).
Other organizations are also popping up to help fund recounts. One linked to on the DemocraticUnderground forums (post) is the Help America Recount Fund, a new 527 group.
- The Nader/Camejo campaign is officially starting a recount in 11 New Hampshire counties (Nader/Camejo press release). According to a recent Washington Post article, it sounds like Nader has met the requirements for a recount, so it may be happening soon (link to an earlier article). Nader is accepting donations to help pay the required fees.
- Cobb and Badnarik, the Green and Libertarian candidates for president, are both calling for an official recount in Ohio. Both parties are looking for donations to pay the $110,000 fee by Monday ($150,000 including additional expenses; Green donation page, Libertarian donation page). The recount request was announced on the Cobb/LaMarche website (the announcement was at the top on the page without a permalink, and can also be found here). Cobb also released a statement on November 5 explaining more of his reasons.
Other organizations are also popping up to help fund recounts. One linked to on the DemocraticUnderground forums (post) is the Help America Recount Fund, a new 527 group.
Friday, November 12, 2004
Insects as food
Here in the United States we don't consider eating insects, other than as stunts on reality shows or in novelty candies like Chocolate Crickets (sold by Fluker's Farms). However, in many parts of the world, insects have formed a major portion of people's diets for years, and it's only through the westernization of these cultures that they've started to think that insects don't make proper food (DeFoliart 1999).
Thus it's refreshing to see that a recent UN study suggests that insects could be used as food sources to counter food-supply problems in some areas. In fact, the article points out, "Contrary to what many may think, caterpillars are not considered an emergency food, but are an integral part of diet in many regions according to seasonal availability. They are consumed as a delicacy." (via BoingBoing)
If you'd like to learn more about the worldwide consumption of insects, and the negative effects western culture is having on it, I highly recommend DeFoliart's (1999) review article. This article's description of the delightful taste of winged termites (better than a good steak, according to some) has made me want to try them ever since reading it.
DeFoliart, G. R. 1999. Insects as food: why the western attitude is important. Annual Review of Entomology 44: 21-50.
Thus it's refreshing to see that a recent UN study suggests that insects could be used as food sources to counter food-supply problems in some areas. In fact, the article points out, "Contrary to what many may think, caterpillars are not considered an emergency food, but are an integral part of diet in many regions according to seasonal availability. They are consumed as a delicacy." (via BoingBoing)
If you'd like to learn more about the worldwide consumption of insects, and the negative effects western culture is having on it, I highly recommend DeFoliart's (1999) review article. This article's description of the delightful taste of winged termites (better than a good steak, according to some) has made me want to try them ever since reading it.
DeFoliart, G. R. 1999. Insects as food: why the western attitude is important. Annual Review of Entomology 44: 21-50.
More election links
As anyone who's been reading progressive boards on the net knows, there's been a lot of speculation flying around about whether there were widespread voting errors or fraud in this past election; I thought I'd provide some links for people who want to read more about the topic.
While many of these articles are official media reports, some are from individual bloggers, academics, or independent writers; I've organized them roughly by type (though a few are hard to identify).
Articles from major US media sources:
While many of these articles are official media reports, some are from individual bloggers, academics, or independent writers; I've organized them roughly by type (though a few are hard to identify).
Articles from major US media sources:
- Vote counting errors found in New York - A democratic candidate for state senate from Yonkers initially was reported to have lost by 1,674 votes, but after recounting the margin shurnk to only 128 votes. "Jonathan Rosen, director of the state Democrats' campaign committee, said the recount had revealed "a highly unusual error rate."" (WNBC - article 1, article 2)
- Sarpy County (Nebraska) found evidence of overvoting in 32 of 80 precincts, for a total of around 10,000 excess votes (first WOWT article). After a recount, 3,000 "phantom votes" were found (second WOWT article).
- Warren County (Ohio) Defends Lockdown Decision - In Warren County, officials prevented media and other people from overseeing the vote counting (reported here), citing security concerns. A followup article in the Cincinati Enquirer reports that FBI and Homeland Security officials say that they never informed the county of any specific threats.
- Exit Polls to Protect the Vote - A New York Times article written before the election discussing the use of exit polling in fraud detection.
- Florida E-Vote Fraud? Unlikely - An article discussing Kathy Dopp's dataset, citing academic sources who say it doesn't prove fraud, by Wired News.
- Worst Voter Error Is Apathy Toward Irregularities - a column in the Washington Post.
- See my earlier post for a few additional articles.
- The Unexplained Exit Poll Discrepancy - a well-written draft academic paper (PDF) by Steven Freeman, discussing the general accuracy of exit polling, as well as detailing and statistically analyzing discrepancies between exit polls and actual results in Florida, Ohio, and Pennsylvania. This is probably the most thoughtful exit poll analysis I've read.
- Odds of Bush gaining by 4 percent in all exit polling states 1 in 50,000 - The Blue Lemur
- BlackBoxVoting.org reports that they have acquired "two Ciber certification reports [for electronic voting machines] indicating that security and tamperability was not tested and that several state elections directors, a secretary of state, and computer consultant Dr. Britain Williams signed off on the report anyway, certifying it." (posted Nov 7th on their homepage; I can't find a permalink)
- Foreign monitors 'barred' from US polls - an article in the Independent Online
- An analysis of New Hampshire voting data - Invisible Ida
- An analysis of New Hampshire results using Benford's Law - lesspress.com
- Kerry Won - An analysis of data from Ohio - TomPaine.com
- Twelve ways Bush is now stealing the Ohio vote - written before the election, by The Free Press
- Problems with Ohio vote tallying early in the election - Oliverwillis.com
- Errors plague voting process in Ohio, PA. - an article on Vindy.com
- On The Front Lines in Florida - an essay discussing voting problems in Broward County, FL.
Thursday, November 11, 2004
Happy Birthday!
Our baby mice turned one month old today! So, for a change of pace from my recent posts, I thought I'd post up some cute baby mouse pictures. These pictures were all taken last Friday; I haven't had time to get them posted due to all my recent election-related posting (I can't help it if I'm easily distracted).
The first thing you'll notice about the babies is how big they've gotten:
Narrow Stripey (tentatively renamed Meryl)
If you're wondering why the hand in that picture looks all shiny and wet, that's because we've developed a super-secret new method for getting the babies to walk onto our hands: feed them ice cream.
Five babies around a hand with some melted vanilla ice cream in it.
Of course, the problem with feeding them ice cream to get them into your hands is that sometimes you get, well, a handful:
Five baby mice in two hands. From left to right: Vash (the mouse formerly known as Runt), Narrow Stripey (tentatively renamed Meryl), Wide Stripey, Athos (aka Ace, in front), and D'Artagnan (aka Deuce).
What's especially fun is to compare the above pictures to when they were only six days old (link to picture), when they could easily fit into one hand.
The babies have also developed a keen sense of adventure, and as soon as they crawl onto anyone they try to go everywhere. D'Artagnan recently climbed up my ear to get onto the top of my head, Genie likes to crawl down the backs of people's shirts, and most of them seem to love crawling into shirt sleeves:
And, as a final picture, here's Rem, the proud mother of six who has done all the work to get these babies up to their one month birthday. I think she deserves a big treat tonight.
Rem
The first thing you'll notice about the babies is how big they've gotten:
Narrow Stripey (tentatively renamed Meryl)
If you're wondering why the hand in that picture looks all shiny and wet, that's because we've developed a super-secret new method for getting the babies to walk onto our hands: feed them ice cream.
Five babies around a hand with some melted vanilla ice cream in it.
Of course, the problem with feeding them ice cream to get them into your hands is that sometimes you get, well, a handful:
Five baby mice in two hands. From left to right: Vash (the mouse formerly known as Runt), Narrow Stripey (tentatively renamed Meryl), Wide Stripey, Athos (aka Ace, in front), and D'Artagnan (aka Deuce).
What's especially fun is to compare the above pictures to when they were only six days old (link to picture), when they could easily fit into one hand.
The babies have also developed a keen sense of adventure, and as soon as they crawl onto anyone they try to go everywhere. D'Artagnan recently climbed up my ear to get onto the top of my head, Genie likes to crawl down the backs of people's shirts, and most of them seem to love crawling into shirt sleeves:
And, as a final picture, here's Rem, the proud mother of six who has done all the work to get these babies up to their one month birthday. I think she deserves a big treat tonight.
Rem
10,000
I just passed the 10,000 visit mark here at Rhosgobel. Thanks to everyone who has visited, left feedback, and just generally helped to make blogging one of the things I look forward to doing every day.
Wednesday, November 10, 2004
More Florida vote changes
This past Sunday I reported that there were more votes for president in some counties of Florida than there were voters voting. On Monday a number of the counties updated their data, and I posted about that as well (here).
Four counties updated their data between my Monday and Tuesday checks (Nov. 9, 2004 ~1am to Nov. 9, 2004 ~10pm), but no counties appeared to update their data between my Tuesday and Wednesday checks (Nov. 9 ~10pm to Nov. 10 ~11pm). There were no changes in voter turnout data or presidential votes cast between my Tuesday and Wednesday data collection. All data were, as usual, obtained from the Florida Department of State's Divisions of Elections results page. As with all prior days, the "County Reporting" page still has a "Y" in the "Final Report" and "Absentees" columns for all counties.
See table 1 for a summary of changes made by the four counties which updated on Tuesday. Gadsden County, which is listed as having an update at 11/9/2004 10:34, had no changes for either turnout or presidential votes cast.
Both counties that had excess votes recorded as of 11/9/04 1am (Osceola and Duval) now have no excess votes reported (table 2). Osceola County's voter turnout numbers were increased to precisely match the number of presidential votes cast; it is the only county to have such a matchup. The one county whose total number of presidential votes cast was changed had the votes distributed between a number of candidates (table 3).
In order to keep the data tables relatively short I am only posting data for the counties that have altered totals since 11/9/04 1am. If anyone needs the full dataset I can provide it upon request.
Table 1: Change in voter turnout and presidential votes cast from data obtained ~1am 11/9/04 to data obtained ~10pm 11/9/04. All changes were calculated by subtracting the 1am value from the 10pm value.
Table 2: Tabulated data of total voter turnout and total votes cast for president in the Florida counties that updated their data on November 9, 2004 (for the November 2, 2004 election). Data was obtained at ~10pm November 9, 2004 from the Florida state site linked to above. Note that there are no longer any excess votes recorded.
Table 3: Comparison of vote totals for each candidate in the one Florida county whose presidential vote totals changed from 11/9/04 1am to 11/9/04 10pm. Numbers listed are the gain in votes the candidate received from 1am to 10pm.
Four counties updated their data between my Monday and Tuesday checks (Nov. 9, 2004 ~1am to Nov. 9, 2004 ~10pm), but no counties appeared to update their data between my Tuesday and Wednesday checks (Nov. 9 ~10pm to Nov. 10 ~11pm). There were no changes in voter turnout data or presidential votes cast between my Tuesday and Wednesday data collection. All data were, as usual, obtained from the Florida Department of State's Divisions of Elections results page. As with all prior days, the "County Reporting" page still has a "Y" in the "Final Report" and "Absentees" columns for all counties.
See table 1 for a summary of changes made by the four counties which updated on Tuesday. Gadsden County, which is listed as having an update at 11/9/2004 10:34, had no changes for either turnout or presidential votes cast.
Both counties that had excess votes recorded as of 11/9/04 1am (Osceola and Duval) now have no excess votes reported (table 2). Osceola County's voter turnout numbers were increased to precisely match the number of presidential votes cast; it is the only county to have such a matchup. The one county whose total number of presidential votes cast was changed had the votes distributed between a number of candidates (table 3).
In order to keep the data tables relatively short I am only posting data for the counties that have altered totals since 11/9/04 1am. If anyone needs the full dataset I can provide it upon request.
Table 1: Change in voter turnout and presidential votes cast from data obtained ~1am 11/9/04 to data obtained ~10pm 11/9/04. All changes were calculated by subtracting the 1am value from the 10pm value.
| County | Change in turnout (11/9/04 10pm minus 11/9/04 1am) | Change in presidential votes cast (11/9/04 10pm minus 11/9/04 1am) |
| Duval | 1305 | 0 |
| Gadsden | 0 | 0 |
| Osceola | 261 | 0 |
| St. Lucie | 0 | 114 |
Table 2: Tabulated data of total voter turnout and total votes cast for president in the Florida counties that updated their data on November 9, 2004 (for the November 2, 2004 election). Data was obtained at ~10pm November 9, 2004 from the Florida state site linked to above. Note that there are no longer any excess votes recorded.
| County | Turnout | Presidential votes | T-v | (T-v)/T | Excess votes |
| Duval | 380562 | 379614 | 948 | 0.25% | |
| Gadsden | 21094 | 20984 | 110 | 0.52% | |
| Osceola | 82178 | 82178 | 0 | 0.00% | |
| St. Lucie | 100374 | 100027 | 347 | 0.35% | |
| Sum | 0 |
Table 3: Comparison of vote totals for each candidate in the one Florida county whose presidential vote totals changed from 11/9/04 1am to 11/9/04 10pm. Numbers listed are the gain in votes the candidate received from 1am to 10pm.
| County | Bush / Cheney | Kerry / Edwards | Peroutka / Baldwin | Badnarik / Campagna | Cobb / LaMarche | Harris / Trowe | Brown / Herbert | Nader / Camejo |
| St. Lucie | 56 | 56 | 0 | 1 | 0 | 0 | 0 | 1 |
Tuesday, November 09, 2004
Update: Florida election turnout vs. presidential votes
The Florida Department of State's Divisions of Elections results page was updated Monday for a number of counties, with turnout totals altered for 42 counties (63% of all counties) and vote totals altered for 5 counties (7.5% of all counties), compared with the data I summarized in my previous post (table 1).
The changes have largely removed the excess votes from the counties that had them, though two counties still have more votes for president recorded than voters voting in the election (Duval and Osceola Counties, for a total of 618 excess votes; table 2). Duval County's last update is recorded as 11/4/04 13:11, while Osceola County's last update is now listed as 11/8/04 09:45. Orange County, one of the counties with extra votes on 11/7/04, now only has 0.09% more voters than votes for president, a value much lower than that found in all the other counties.
Overall a total of 271,137 net voters were added across all counties, while 4,493 votes were added for president. The five counties that added presidential votes each appeared to add them to a mix of candidates (table 3).
I can't find any explanation on the Florida site for why the turnout and vote totals changed, other than that the counties are listed as having updated their data on the county reporting page. The pages are all still titled "1st set of unofficial returns (pursuant to s. 102.141, F.S.)", the same as they were on Sunday, and all counties still report "Y" in the "Final Report" and "Absentees" column. The provisionals page does not list any updates having occured after 11/5/04.
Data tables:
Table 1: Change in voter turnout and presidential votes cast from data obtained ~6pm 11/7/04 to data obtained ~1:15am 11/9/04. All changes were calculated by subtracting the Nov. 7, 2004 value from the Nov. 9, 2004 value.
Table 2: Tabulated data of total voter turnout and total votes cast for president in all Florida counties for the November 2, 2004 election. Note that a negative number in the "turnout minus votes" column indicates that there were more votes cast for president than voters voting in the county. Data was obtained at ~1:15am November 9, 2004 from the Florida state site linked to above.
Table 3: Comparison of vote totals for each candidate in the five Flordia counties whose presidential vote totals chagned from 11/7/04 to 11/9/04. Numbers listed are the gain in votes the candidate received from 11/7/04 to 11/9/04.
The changes have largely removed the excess votes from the counties that had them, though two counties still have more votes for president recorded than voters voting in the election (Duval and Osceola Counties, for a total of 618 excess votes; table 2). Duval County's last update is recorded as 11/4/04 13:11, while Osceola County's last update is now listed as 11/8/04 09:45. Orange County, one of the counties with extra votes on 11/7/04, now only has 0.09% more voters than votes for president, a value much lower than that found in all the other counties.
Overall a total of 271,137 net voters were added across all counties, while 4,493 votes were added for president. The five counties that added presidential votes each appeared to add them to a mix of candidates (table 3).
I can't find any explanation on the Florida site for why the turnout and vote totals changed, other than that the counties are listed as having updated their data on the county reporting page. The pages are all still titled "1st set of unofficial returns (pursuant to s. 102.141, F.S.)", the same as they were on Sunday, and all counties still report "Y" in the "Final Report" and "Absentees" column. The provisionals page does not list any updates having occured after 11/5/04.
Data tables:
Table 1: Change in voter turnout and presidential votes cast from data obtained ~6pm 11/7/04 to data obtained ~1:15am 11/9/04. All changes were calculated by subtracting the Nov. 7, 2004 value from the Nov. 9, 2004 value.
| County | Change in turnout | Change in presidential votes | County | Change in turnout | Change in presidential votes | |
| Alachua | 284 | 279 | Lee | 0 | 0 | |
| Baker | -11 | 0 | Leon | 344 | 0 | |
| Bay | 108 | 0 | Levy | 0 | 0 | |
| Bradford | 0 | 0 | Liberty | 0 | 0 | |
| Brevard | 0 | 0 | Madison | 0 | 0 | |
| Broward | 1436 | 0 | Manatee | 71 | 0 | |
| Calhoun | 0 | 0 | Marion | 0 | 0 | |
| Charlotte | 0 | 0 | Martin | 0 | 0 | |
| Citrus | 5 | 0 | Miami-Dade | 56169 | 0 | |
| Clay | 86 | 0 | Monroe | 0 | 2 | |
| Collier | 1490 | 0 | Nassau | 8 | 0 | |
| Columbia | 0 | 0 | Okaloosa | 423 | 0 | |
| DeSoto | 2 | 0 | Okeechobee | 4 | 0 | |
| Dixie | 0 | 0 | Orange | 1991 | 0 | |
| Duval | 0 | 0 | Osceola | 18328 | 0 | |
| Escambia | -17624 | 0 | Palm Beach | 95279 | 0 | |
| Flagler | 20 | 0 | Pasco | 5 | 0 | |
| Franklin | 0 | 0 | Pinellas | 213 | 0 | |
| Gadsden | 0 | 0 | Polk | 146 | 0 | |
| Gilchrist | 5 | 0 | Putnam | 4 | 0 | |
| Glades | 758 | 0 | Santa Rosa | 39 | 0 | |
| Gulf | 0 | 0 | Sarasota | 349 | 0 | |
| Hamilton | 15 | 0 | Seminole | 359 | 0 | |
| Hardee | -4 | 0 | St. Johns | 0 | 0 | |
| Hendry | 3 | 0 | St. Lucie | 0 | 0 | |
| Hernando | 79308 | 3247 | Sumter | 2 | 0 | |
| Highlands | 7875 | 0 | Suwannee | 17 | 0 | |
| Hillsborough | 2481 | 941 | Taylor | 0 | 0 | |
| Holmes | 0 | 0 | Union | 0 | 0 | |
| Indian River | 292 | 0 | Volusia | 20046 | 24 | |
| Jackson | 48 | 0 | Wakulla | 0 | 0 | |
| Jefferson | 0 | 0 | Walton | 36 | 0 | |
| Lafayette | 0 | 0 | Washington | 2 | 0 | |
| Lake | 725 | 0 | Sum | 271137 | 4493 |
Table 2: Tabulated data of total voter turnout and total votes cast for president in all Florida counties for the November 2, 2004 election. Note that a negative number in the "turnout minus votes" column indicates that there were more votes cast for president than voters voting in the county. Data was obtained at ~1:15am November 9, 2004 from the Florida state site linked to above.
| County | Turnout | Presidential votes | T-v | (T-v)/T | Excess Votes |
| Alachua | 111571 | 111301 | 270 | 0.24% | |
| Baker | 10048 | 9955 | 93 | 0.93% | |
| Bay | 75208 | 74998 | 210 | 0.28% | |
| Bradford | 10913 | 10851 | 62 | 0.57% | |
| Brevard | 265764 | 265075 | 689 | 0.26% | |
| Broward | 707202 | 704376 | 2826 | 0.40% | |
| Calhoun | 6006 | 5961 | 45 | 0.75% | |
| Charlotte | 80140 | 79730 | 410 | 0.51% | |
| Citrus | 69696 | 69462 | 234 | 0.34% | |
| Clay | 81473 | 81230 | 243 | 0.30% | |
| Collier | 128899 | 128352 | 547 | 0.42% | |
| Columbia | 25075 | 24984 | 91 | 0.36% | |
| DeSoto | 9534 | 9495 | 39 | 0.41% | |
| Dixie | 6472 | 6440 | 32 | 0.49% | |
| Duval | 379257 | 379614 | -357 | -0.09% | 357 |
| Escambia | 143559 | 142990 | 569 | 0.40% | |
| Flagler | 38552 | 38475 | 77 | 0.20% | |
| Franklin | 5973 | 5930 | 43 | 0.72% | |
| Gadsden | 21094 | 20984 | 110 | 0.52% | |
| Gilchrist | 7047 | 7012 | 35 | 0.50% | |
| Glades | 4204 | 4188 | 16 | 0.38% | |
| Gulf | 7299 | 7256 | 43 | 0.59% | |
| Hamilton | 5131 | 5079 | 52 | 1.01% | |
| Hardee | 7281 | 7246 | 35 | 0.48% | |
| Hendry | 9813 | 9774 | 39 | 0.40% | |
| Hernando | 79308 | 79079 | 229 | 0.29% | |
| Highlands | 41871 | 41491 | 380 | 0.91% | |
| Hillsborough | 464253 | 462461 | 1792 | 0.39% | |
| Holmes | 8349 | 8298 | 51 | 0.61% | |
| Indian River | 61613 | 61321 | 292 | 0.47% | |
| Jackson | 19898 | 19797 | 101 | 0.51% | |
| Jefferson | 7501 | 7477 | 24 | 0.32% | |
| Lafayette | 3352 | 3325 | 27 | 0.81% | |
| Lake | 124476 | 123938 | 538 | 0.43% | |
| Lee | 242434 | 241433 | 1001 | 0.41% | |
| Leon | 136573 | 136314 | 259 | 0.19% | |
| Levy | 16742 | 16649 | 93 | 0.56% | |
| Liberty | 3051 | 3021 | 30 | 0.98% | |
| Madison | 8342 | 8307 | 35 | 0.42% | |
| Manatee | 143901 | 143539 | 362 | 0.25% | |
| Marion | 140321 | 139644 | 677 | 0.48% | |
| Martin | 72709 | 72430 | 279 | 0.38% | |
| Miami-Dade | 772743 | 768553 | 4190 | 0.54% | |
| Monroe | 39629 | 39527 | 102 | 0.26% | |
| Nassau | 32748 | 32664 | 84 | 0.26% | |
| Okaloosa | 89908 | 89707 | 201 | 0.22% | |
| Okeechobee | 12247 | 12197 | 50 | 0.41% | |
| Orange | 388095 | 387752 | 343 | 0.09% | |
| Osceola | 81917 | 82178 | -261 | -0.32% | 261 |
| Palm Beach | 547340 | 542835 | 4505 | 0.82% | |
| Pasco | 191859 | 190866 | 993 | 0.52% | |
| Pinellas | 457426 | 455203 | 2223 | 0.49% | |
| Polk | 211346 | 210777 | 569 | 0.27% | |
| Putnam | 31063 | 30964 | 99 | 0.32% | |
| Santa Rosa | 67368 | 67213 | 155 | 0.23% | |
| Sarasota | 196291 | 195530 | 761 | 0.39% | |
| Seminole | 186537 | 186115 | 422 | 0.23% | |
| St. Johns | 86603 | 86262 | 341 | 0.39% | |
| St. Lucie | 100374 | 99913 | 461 | 0.46% | |
| Sumter | 32000 | 31837 | 163 | 0.51% | |
| Suwannee | 15878 | 15801 | 77 | 0.48% | |
| Taylor | 8613 | 8580 | 33 | 0.38% | |
| Union | 4714 | 4675 | 39 | 0.83% | |
| Volusia | 229098 | 228382 | 716 | 0.31% | |
| Wakulla | 11820 | 11763 | 57 | 0.48% | |
| Walton | 24063 | 23974 | 89 | 0.37% | |
| Washington | 10452 | 10365 | 87 | 0.83% | |
| Sum | 7622037 | 7592915 | 29122 | 618 |
Table 3: Comparison of vote totals for each candidate in the five Flordia counties whose presidential vote totals chagned from 11/7/04 to 11/9/04. Numbers listed are the gain in votes the candidate received from 11/7/04 to 11/9/04.
| County | Bush / Cheney | Kerry / Edwards | Peroutka / Baldwin | Badnarik / Campagna | Cobb / LaMarche | Harris / Trowe | Brown / Herbert | Nader / Camejo |
| Alachua | 137 | 139 | 0 | 1 | 0 | 0 | 0 | 2 |
| Hernando | 1710 | 1515 | 9 | 4 | 0 | 0 | 0 | 9 |
| Hillsborough | 544 | 391 | 0 | 3 | 0 | 0 | 0 | 3 |
| Monroe | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 |
| Volusia | 8 | 16 | 0 | 0 | 0 | 0 | 0 | 0 |
Monday, November 08, 2004
Florida 1996 vs. 2004 presidential election data
My SO and I found Kathy Dopp's analysis of Florida's county election data a few days ago. Ms. Dopp's dataset compares political party registration numbers to vote counts in each Florida county, and finds a difference between counties using optical scan and electronic voting machines. I was thinking of trying to summarize the topic here, but Pharyngula, Preposterous Universe, and CommonDreams have all summarized it nicely, so I'll let them do the talking.
Ms. Dopp has also analyzed data from the 2000 presidential election, but we decided to go back and analyze the 1996 presidential election data to see if the optical scan counties showed the same Republican voting trends back then. We obtained the 2004 election data from Ms. Dopp's page, and obtained the 1996 election results from the Florida Department of State's Division of Elections Election Results page (data obtained ~Nov. 6, 2004). We did not obtain registration data for 1996.
In our analysis we calculated the fraction of total votes cast for the Democratic and Republican candidates in both 1996 and 2004 within every Florida county. We then compared the 2004 to 1996 voting percentages for each party and each county (subtracting the 1996 vote percentages from the 2004 vote percentages) to generate a percent change in votes for each party's candidate. Averaging across all counties, Kerry got a 3.4% smaller fraction of the vote in 2004 than Clinton in 1996, while Bush got a 14.9% larger fraction of the vote in 2004 than Dole in 1996, meaning that there was a total "shift" of 18.3% to the Republican candidate. A possible confounding factor was the presence in 1996 of Perot as a significant third party candidate (receiving 9.1% of the Florida vote).
As Ms. Kopp did, we also grouped the data by the style of voting machines used in 2004. We found that counties that used optical scan balloting in 2004 had an average Republican "shift" of 20.4%, while counties that used electronic voting in 2004 had an average Republican "shift" of 11.3%.
Here are a few selected examples:
Tables:
I have included the summary data below, and have posted the full 1996 data table separately here.
Table 1: Percent of total votes cast in the 1996 and 2004 presidential elections for the Democratic candidate (Clinton in 1996, Kerry in 2004) or Republican candidate (Dole in 1996, Bush in 2004). The shift in voting preferences between the two elections is tabulated for each party separately, which is then used to calcluate the overall "shift" towards the Republican candidate from 1996 to 2004. Voting method data are from the 2004 election; "e" is electronic voting, and "op" is optical scan.
Ms. Dopp has also analyzed data from the 2000 presidential election, but we decided to go back and analyze the 1996 presidential election data to see if the optical scan counties showed the same Republican voting trends back then. We obtained the 2004 election data from Ms. Dopp's page, and obtained the 1996 election results from the Florida Department of State's Division of Elections Election Results page (data obtained ~Nov. 6, 2004). We did not obtain registration data for 1996.
In our analysis we calculated the fraction of total votes cast for the Democratic and Republican candidates in both 1996 and 2004 within every Florida county. We then compared the 2004 to 1996 voting percentages for each party and each county (subtracting the 1996 vote percentages from the 2004 vote percentages) to generate a percent change in votes for each party's candidate. Averaging across all counties, Kerry got a 3.4% smaller fraction of the vote in 2004 than Clinton in 1996, while Bush got a 14.9% larger fraction of the vote in 2004 than Dole in 1996, meaning that there was a total "shift" of 18.3% to the Republican candidate. A possible confounding factor was the presence in 1996 of Perot as a significant third party candidate (receiving 9.1% of the Florida vote).
As Ms. Kopp did, we also grouped the data by the style of voting machines used in 2004. We found that counties that used optical scan balloting in 2004 had an average Republican "shift" of 20.4%, while counties that used electronic voting in 2004 had an average Republican "shift" of 11.3%.
Here are a few selected examples:
- Dixie County voted 45.6% Democratic and 36.8% Republican in 1996, compared to 30.4% Democratic and 68.8% Republican in 2004, for a total "shift" of 47.2%.
- Holmes County voted 34.0% Democratic and 47.8% Republican in 1996, compared to 21.8% Democratic and 77.2% Republican in 2004, for a total "shift" of 41.6%.
- Putnam County voted 47.8% Democratic and 38.9% Republican in 1996, compared to 40.1% Democratic and 59.1% Republican in 2004, for a total "shift" of 27.9%.
Tables:
I have included the summary data below, and have posted the full 1996 data table separately here.
Table 1: Percent of total votes cast in the 1996 and 2004 presidential elections for the Democratic candidate (Clinton in 1996, Kerry in 2004) or Republican candidate (Dole in 1996, Bush in 2004). The shift in voting preferences between the two elections is tabulated for each party separately, which is then used to calcluate the overall "shift" towards the Republican candidate from 1996 to 2004. Voting method data are from the 2004 election; "e" is electronic voting, and "op" is optical scan.
| County | Dem 96 | Rep 96 | Dem 04 | Rep 04 | Voting method in 2004 | Dem 04 minus Dem 96 | Rep 04 minus Rep 96 | Republican "shift" from '96 to '04 |
| Broward | 63.5% | 28.3% | 64.3% | 34.5% | e | 0.8% | 6.2% | 5.4% |
| Charlotte | 43.0% | 44.2% | 42.9% | 55.7% | e | -0.1% | 11.5% | 11.6% |
| Collier | 32.0% | 58.7% | 34.1% | 65.0% | e | 2.1% | 6.3% | 4.1% |
| Dade | 57.3% | 37.9% | 53.7% | 45.8% | e | -3.6% | 7.9% | 11.5% |
| Hillsborough | 46.8% | 44.3% | 46.3% | 53.0% | e | -0.5% | 8.7% | 9.2% |
| Indian River | 37.2% | 51.7% | 39.0% | 60.2% | e | 1.8% | 8.5% | 6.7% |
| Lake | 40.3% | 47.5% | 38.9% | 60.0% | e | -1.3% | 12.5% | 13.9% |
| Lee | 39.6% | 48.7% | 39.8% | 59.0% | e | 0.2% | 10.3% | 10.1% |
| Martin | 38.2% | 52.2% | 41.7% | 57.1% | e | 3.5% | 4.9% | 1.4% |
| Nassau | 34.4% | 57.3% | 26.2% | 72.7% | e | -8.2% | 15.3% | 23.5% |
| Palm Beach | 58.1% | 33.7% | 60.8% | 38.5% | e | 2.8% | 4.9% | 2.1% |
| Pasco | 49.8% | 36.2% | 44.4% | 54.1% | e | -5.4% | 17.8% | 23.3% |
| Pinellas | 49.1% | 40.4% | 49.5% | 49.6% | e | 0.4% | 9.2% | 8.8% |
| Sarasota | 42.7% | 46.5% | 45.2% | 53.5% | e | 2.5% | 7.1% | 4.6% |
| Sumter | 45.6% | 38.7% | 36.4% | 62.2% | e | -9.2% | 23.5% | 32.6% |
| Average for electronic voting counties | -1.0% | 10.3% | 11.3% | |||||
| Alachua | 53.9% | 34.0% | 56.2% | 42.9% | op | 2.3% | 8.9% | 6.7% |
| Baker | 34.3% | 55.5% | 21.9% | 77.7% | op | -12.4% | 22.2% | 34.6% |
| Bay | 33.0% | 54.9% | 28.1% | 71.2% | op | -4.9% | 16.3% | 21.2% |
| Bradford | 40.7% | 49.0% | 29.9% | 69.6% | op | -10.8% | 20.6% | 31.4% |
| Brevard | 41.2% | 45.1% | 41.6% | 57.7% | op | 0.3% | 12.6% | 12.2% |
| Calhoun | 43.1% | 41.3% | 35.5% | 63.4% | op | -7.6% | 22.1% | 29.8% |
| Citrus | 44.5% | 40.6% | 42.1% | 56.9% | op | -2.3% | 16.3% | 18.6% |
| Clay | 28.2% | 64.5% | 23.3% | 76.2% | op | -4.9% | 11.7% | 16.6% |
| Columbia | 41.0% | 46.5% | 32.1% | 67.1% | op | -8.8% | 20.6% | 29.4% |
| Desoto | 43.0% | 43.7% | 41.2% | 58.0% | op | -1.8% | 14.3% | 16.1% |
| Dixie | 45.6% | 36.8% | 30.4% | 68.8% | op | -15.2% | 32.0% | 47.2% |
| Duval | 44.2% | 50.0% | 41.7% | 57.7% | op | -2.5% | 7.8% | 10.3% |
| Escambia | 35.1% | 56.5% | 33.7% | 65.3% | op | -1.3% | 8.8% | 10.1% |
| Flagler | 47.7% | 41.0% | 48.3% | 51.0% | op | 0.5% | 10.0% | 9.5% |
| Franklin | 45.9% | 34.2% | 40.5% | 58.5% | op | -5.4% | 24.3% | 29.7% |
| Gadsden | 66.3% | 26.9% | 69.7% | 29.8% | op | 3.5% | 2.9% | -0.6% |
| Gilchrist | 41.4% | 40.4% | 28.8% | 70.4% | op | -12.6% | 29.9% | 42.6% |
| Glades | 44.6% | 39.7% | 41.8% | 57.7% | op | -2.8% | 18.1% | 20.9% |
| Gulf | 41.4% | 40.5% | 33.0% | 66.1% | op | -8.4% | 25.6% | 34.0% |
| Hamilton | 47.2% | 41.4% | 44.5% | 55.0% | op | -2.8% | 13.6% | 16.4% |
| Hardee | 39.0% | 47.2% | 29.6% | 69.7% | op | -9.3% | 22.5% | 31.8% |
| Hendry | 43.6% | 43.3% | 40.5% | 58.9% | op | -3.1% | 15.6% | 18.7% |
| Hernando | 49.1% | 38.0% | 46.2% | 52.9% | op | -3.0% | 15.0% | 17.9% |
| Highlands | 42.3% | 46.3% | 38.5% | 60.8% | op | -3.7% | 14.5% | 18.2% |
| Holmes | 34.0% | 47.8% | 21.8% | 77.2% | op | -12.2% | 29.5% | 41.6% |
| Jackson | 43.0% | 46.3% | 38.1% | 61.2% | op | -4.9% | 14.9% | 19.7% |
| Jefferson | 52.9% | 38.5% | 55.3% | 44.1% | op | 2.4% | 5.6% | 3.2% |
| Lafayette | 35.7% | 50.2% | 25.4% | 74.0% | op | -10.3% | 23.8% | 34.1% |
| Leon | 54.6% | 37.0% | 62.0% | 37.3% | op | 7.4% | 0.3% | -7.1% |
| Levy | 44.6% | 38.9% | 36.5% | 62.5% | op | -8.2% | 23.7% | 31.8% |
| Liberty | 40.1% | 42.2% | 35.4% | 63.8% | op | -4.7% | 21.6% | 26.3% |
| Madison | 50.0% | 39.3% | 48.7% | 50.5% | op | -1.2% | 11.2% | 12.4% |
| Manatee | 43.2% | 45.5% | 42.7% | 56.6% | op | -0.6% | 11.1% | 11.7% |
| Marion | 41.1% | 45.9% | 41.0% | 58.2% | op | -0.1% | 12.3% | 12.4% |
| Monroe | 46.9% | 37.0% | 49.7% | 49.2% | op | 2.8% | 12.2% | 9.4% |
| Okaloosa | 26.1% | 64.5% | 21.6% | 77.6% | op | -4.5% | 13.1% | 17.6% |
| Okeechobee | 48.6% | 34.4% | 42.3% | 57.2% | op | -6.3% | 22.9% | 29.2% |
| Orange | 45.7% | 45.9% | 49.8% | 49.6% | op | 4.1% | 3.8% | -0.4% |
| Osceola | 47.0% | 39.4% | 47.8% | 51.7% | op | 0.7% | 12.3% | 11.6% |
| Polk | 44.4% | 45.3% | 40.8% | 58.6% | op | -3.7% | 13.4% | 17.0% |
| Putnam | 47.8% | 38.9% | 40.1% | 59.1% | op | -7.7% | 20.2% | 27.9% |
| Santa Rosa | 25.8% | 62.0% | 21.8% | 77.3% | op | -4.0% | 15.3% | 19.4% |
| Seminole | 39.2% | 52.0% | 41.3% | 58.1% | op | 2.1% | 6.1% | 3.9% |
| St. Johns | 34.4% | 56.3% | 30.6% | 68.6% | op | -3.8% | 12.3% | 16.2% |
| St. Lucie | 48.9% | 39.1% | 52.4% | 47.0% | op | 3.4% | 7.9% | 4.5% |
| Suwannee | 36.9% | 47.3% | 28.6% | 70.6% | op | -8.3% | 23.3% | 31.6% |
| Taylor | 44.8% | 39.9% | 35.5% | 63.7% | op | -9.3% | 23.8% | 33.1% |
| Union | 40.1% | 47.3% | 26.8% | 72.6% | op | -13.3% | 25.4% | 38.7% |
| Volusia | 49.3% | 39.4% | 51.3% | 48.1% | op | 2.0% | 8.7% | 6.7% |
| Wakulla | 42.6% | 40.9% | 41.6% | 57.6% | op | -1.0% | 16.7% | 17.7% |
| Walton | 34.4% | 49.7% | 25.9% | 73.2% | op | -8.5% | 23.5% | 32.0% |
| Washington | 38.1% | 44.8% | 28.1% | 71.1% | op | -10.0% | 26.3% | 36.3% |
| Average for optical scan counties | -4.1% | 16.3% | 20.4% | |||||
| Averages for all counties: | -3.4% | 14.9% | 18.3% |
Sunday, November 07, 2004
Election turnout vs. presidential votes, pt. 2
A few hours ago I posted about a discrepancy between the number of voters voting and the number of votes cast in a number of Florida counties. In order to expand on that analysis my SO and I set out to acquire data from a number of other states to see how widespread the problem was.
After analyzing Florida and Oregon we attempted to acquire data from several other populous states (ranked here). We have now successfully analyzed data from Florida, Michigan, Ohio, Oregon, and Texas (Oregon and Florida were analyzed in my previous post). We were not able to analyze data from the following states for various reasons:
Texas: None of Texas' 254 counties reported more votes cast for president than voters voting. Data from Texas Secretary of State Elections Division results page, all precincts listed as having reported, accessed ~9:30pm today.
Ohio: None of Ohio's 88 counties reported more votes cast for president than voters voting (data here and here, the election night page lists 100% of the precincts as having reported, accessed ~9:30pm today).
Michigan: Four of Michigan's 83 counties (4.8%) recorded more votes for president than voters voting in the election, for a total of 1,582 excess votes in the four counties (Jackson, Muskegon, St. Clair, and Wexford Counties). Data from the Michigan Secretary of State's unofficial results page (data here and here, no data for % reporting precincts, accessed ~10:15pm today).
Due to the length of the dataset tables I have posted them as separate pages:
After analyzing Florida and Oregon we attempted to acquire data from several other populous states (ranked here). We have now successfully analyzed data from Florida, Michigan, Ohio, Oregon, and Texas (Oregon and Florida were analyzed in my previous post). We were not able to analyze data from the following states for various reasons:
- California - data formatted inconveniently for analysis
- Georgia - couldn't find turnout
- Illinois - no statewide county summary
- New Jersey - couldn't find turnout
- New York - no results posted until after election certified on Dec. 6
- Pennsylvania - couldn't find turnout
Texas: None of Texas' 254 counties reported more votes cast for president than voters voting. Data from Texas Secretary of State Elections Division results page, all precincts listed as having reported, accessed ~9:30pm today.
Ohio: None of Ohio's 88 counties reported more votes cast for president than voters voting (data here and here, the election night page lists 100% of the precincts as having reported, accessed ~9:30pm today).
Michigan: Four of Michigan's 83 counties (4.8%) recorded more votes for president than voters voting in the election, for a total of 1,582 excess votes in the four counties (Jackson, Muskegon, St. Clair, and Wexford Counties). Data from the Michigan Secretary of State's unofficial results page (data here and here, no data for % reporting precincts, accessed ~10:15pm today).
Due to the length of the dataset tables I have posted them as separate pages:
Florida: election turnout vs. presidential votes
[Update 11/9/04: The Florida Department of State Division of Elections modified the data on their website on 11/8/04; I've posted about the new data here. I haven't found an explanation for the changes.]
The number of votes cast in a county for a specific race should never exceed the number of voters who voted. However, in comparing the numbers published on the Florida Department of State's website, it appears that in some counties there were more votes cast for president than there were voters voting (I first saw this mentioned here).
My SO and I obtained data for the 2004 election in Florida from the "1st set of unofficial returns" on the Florida Department of State Division of Elections election results page (accessed today ~6pm). The results pages report that
But are these just normal election day hiccups? One way to test this is to look at another state and see if it has the same trend in its numbers. We first attempted to look at California, but its data was formatted inconveniently for copying to Excel, so we switched to looking at Oregon. We obtained data from the Oregon Secretary of State's Elections Division website today at ~6:30 pm (specific pages here and here). We chose these states haphazardly (CA because we live here, OR because it's next door).
We analyzed Oregon's data in exactly the same way as we did for Florida, and none of Oregon's 36 counties reported more votes for president than there were voters voting.
[Update: We've now analyzed data from Texas, Ohio, and Michigan - see this post for the analysis.]
Here are the data tables:
Table 1: Tabulated data of total voter turnout and total votes cast for president, in all Florida counties in the November 2, 2004 election. Note that a negative number in the "turnout minus votes" column indicates that there were more votes cast for president than voters voting in the county.
Table 2: Tabulated data of total voter turnout and total votes cast for president in all Oregon counties in the November 2, 2004 election.
The number of votes cast in a county for a specific race should never exceed the number of voters who voted. However, in comparing the numbers published on the Florida Department of State's website, it appears that in some counties there were more votes cast for president than there were voters voting (I first saw this mentioned here).
My SO and I obtained data for the 2004 election in Florida from the "1st set of unofficial returns" on the Florida Department of State Division of Elections election results page (accessed today ~6pm). The results pages report that
- For the presidential election, the "% report" is 100.0% for all counties
- For the turnout data "Y" (presumably "yes") for all counties in the "Final Report" and "Absentees" column.
But are these just normal election day hiccups? One way to test this is to look at another state and see if it has the same trend in its numbers. We first attempted to look at California, but its data was formatted inconveniently for copying to Excel, so we switched to looking at Oregon. We obtained data from the Oregon Secretary of State's Elections Division website today at ~6:30 pm (specific pages here and here). We chose these states haphazardly (CA because we live here, OR because it's next door).
We analyzed Oregon's data in exactly the same way as we did for Florida, and none of Oregon's 36 counties reported more votes for president than there were voters voting.
[Update: We've now analyzed data from Texas, Ohio, and Michigan - see this post for the analysis.]
Here are the data tables:
Table 1: Tabulated data of total voter turnout and total votes cast for president, in all Florida counties in the November 2, 2004 election. Note that a negative number in the "turnout minus votes" column indicates that there were more votes cast for president than voters voting in the county.
| County | Turnout | Presidential votes cast | Turnout minus votes | (T-v)/T | Excess votes |
| Alachua | 111,287 | 111,022 | 265 | 0.2% | |
| Baker | 10,059 | 9,955 | 104 | 1.0% | |
| Bay | 75,100 | 74,998 | 102 | 0.1% | |
| Bradford | 10,913 | 10,851 | 62 | 0.6% | |
| Brevard | 265,764 | 265,075 | 689 | 0.3% | |
| Broward | 705,766 | 704,376 | 1,390 | 0.2% | |
| Calhoun | 6,006 | 5,961 | 45 | 0.7% | |
| Charlotte | 80,140 | 79,730 | 410 | 0.5% | |
| Citrus | 69,691 | 69,462 | 229 | 0.3% | |
| Clay | 81,387 | 81,230 | 157 | 0.2% | |
| Collier | 127,409 | 128,352 | -943 | -0.7% | 943 |
| Columbia | 25,075 | 24,984 | 91 | 0.4% | |
| DeSoto | 9,532 | 9,495 | 37 | 0.4% | |
| Dixie | 6,472 | 6,440 | 32 | 0.5% | |
| Duval | 379,257 | 379,614 | -357 | -0.1% | 357 |
| Escambia | 161,183 | 142,990 | 18,193 | 11.3% | |
| Flagler | 38,532 | 38,475 | 57 | 0.1% | |
| Franklin | 5,973 | 5,930 | 43 | 0.7% | |
| Gadsden | 21,094 | 20,984 | 110 | 0.5% | |
| Gilchrist | 7,042 | 7,012 | 30 | 0.4% | |
| Glades | 3,446 | 4,188 | -742 | -21.5% | 742 |
| Gulf | 7,299 | 7,256 | 43 | 0.6% | |
| Hamilton | 5,116 | 5,079 | 37 | 0.7% | |
| Hardee | 7,285 | 7,246 | 39 | 0.5% | |
| Hendry | 9,810 | 9,774 | 36 | 0.4% | |
| Hernando | 0 | 75,832 | -75,832 | 75832 | |
| Highlands | 33,996 | 41,491 | -7,495 | -22.0% | 7495 |
| Hillsborough | 461,772 | 461,520 | 252 | 0.1% | |
| Holmes | 8,349 | 8,298 | 51 | 0.6% | |
| Indian River | 61,321 | 61,321 | 0 | 0.0% | |
| Jackson | 19,850 | 19,797 | 53 | 0.3% | |
| Jefferson | 7,501 | 7,477 | 24 | 0.3% | |
| Lafayette | 3,352 | 3,325 | 27 | 0.8% | |
| Lake | 123,751 | 123,938 | -187 | -0.2% | 187 |
| Lee | 242,434 | 241,433 | 1,001 | 0.4% | |
| Leon | 136,229 | 136,314 | -85 | -0.1% | 85 |
| Levy | 16,742 | 16,649 | 93 | 0.6% | |
| Liberty | 3,051 | 3,021 | 30 | 1.0% | |
| Madison | 8,342 | 8,307 | 35 | 0.4% | |
| Manatee | 143,830 | 143,539 | 291 | 0.2% | |
| Marion | 140,321 | 139,644 | 677 | 0.5% | |
| Martin | 72,709 | 72,430 | 279 | 0.4% | |
| Miami-Dade | 716,574 | 768,553 | -51,979 | -7.3% | 51979 |
| Monroe | 39,629 | 39,525 | 104 | 0.3% | |
| Nassau | 32,740 | 32,664 | 76 | 0.2% | |
| Okaloosa | 89,485 | 89,707 | -222 | -0.2% | 222 |
| Okeechobee | 12,243 | 12,197 | 46 | 0.4% | |
| Orange | 386,104 | 387,752 | -1,648 | -0.4% | 1648 |
| Osceola | 63,589 | 82,178 | -18,589 | -29.2% | 18589 |
| Palm Beach | 452,061 | 542,835 | -90,774 | -20.1% | 90774 |
| Pasco | 191,854 | 190,866 | 988 | 0.5% | |
| Pinellas | 457,213 | 455,203 | 2,010 | 0.4% | |
| Polk | 211,200 | 210,777 | 423 | 0.2% | |
| Putnam | 31,059 | 30,964 | 95 | 0.3% | |
| Santa Rosa | 67,329 | 67,213 | 116 | 0.2% | |
| Sarasota | 195,942 | 195,530 | 412 | 0.2% | |
| Seminole | 186,178 | 186,115 | 63 | 0.0% | |
| St. Johns | 86,603 | 86,262 | 341 | 0.4% | |
| St. Lucie | 100,374 | 99,913 | 461 | 0.5% | |
| Sumter | 31,998 | 31,837 | 161 | 0.5% | |
| Suwannee | 15,861 | 15,801 | 60 | 0.4% | |
| Taylor | 8,613 | 8,580 | 33 | 0.4% | |
| Union | 4,714 | 4,675 | 39 | 0.8% | |
| Volusia | 209,052 | 228,358 | -19,306 | -9.2% | 19306 |
| Wakulla | 11,820 | 11,763 | 57 | 0.5% | |
| Walton | 24,027 | 23,974 | 53 | 0.2% | |
| Washington | 10,450 | 10,365 | 85 | 0.8% | |
| Sum | 7,350,900 | 7,588,422 | -237,522 | 268,159 |
Table 2: Tabulated data of total voter turnout and total votes cast for president in all Oregon counties in the November 2, 2004 election.
| County | Turnout | Presidential votes cast | Turnout - votes | (T-v)/T | Excess Votes |
| Baker | 8,990 | 8,904 | 86 | 1.0% | |
| Benton | 46,135 | 45,842 | 293 | 0.6% | |
| Clackamas | 183,139 | 181,768 | 1,371 | 0.7% | |
| Clatsop | 19,329 | 19,188 | 141 | 0.7% | |
| Columbia | 24,978 | 24,795 | 183 | 0.7% | |
| Coos | 33,501 | 33,276 | 225 | 0.7% | |
| Crook | 10,106 | 10,028 | 78 | 0.8% | |
| Curry | 11,423 | 11,336 | 87 | 0.8% | |
| Deschutes | 74,295 | 73,809 | 486 | 0.7% | |
| Douglas | 54,945 | 54,358 | 587 | 1.1% | |
| Gilliam | 1,160 | 1,136 | 24 | 2.1% | |
| Grant | 4,234 | 3,157 | 1,077 | 25.4% | |
| Harney | 3,735 | 3,702 | 33 | 0.9% | |
| Hood River | 9,823 | 9,760 | 63 | 0.6% | |
| Jackson | 96,611 | 95,824 | 787 | 0.8% | |
| Jefferson | 8,124 | 8,034 | 90 | 1.1% | |
| Josephine | 42,017 | 41,631 | 386 | 0.9% | |
| Klamath | 31,314 | 31,078 | 236 | 0.8% | |
| Lake | 3,989 | 3,881 | 108 | 2.7% | |
| Lane | 180,353 | 174,856 | 5,497 | 3.0% | |
| Lincoln | 24,555 | 24,317 | 238 | 1.0% | |
| Linn | 51,153 | 50,657 | 496 | 1.0% | |
| Malheur | 10,893 | 10,773 | 120 | 1.1% | |
| Marion | 116,702 | 115,929 | 773 | 0.7% | |
| Morrow | 4,173 | 4,110 | 63 | 1.5% | |
| Multnomah | 345,957 | 343,290 | 2,667 | 0.8% | |
| Polk | 35,118 | 34,802 | 316 | 0.9% | |
| Sherman | 1,122 | 1,099 | 23 | 2.0% | |
| Tillamook | 13,746 | 13,605 | 141 | 1.0% | |
| Umatilla | 26,145 | 25,890 | 255 | 1.0% | |
| Union | 13,442 | 13,284 | 158 | 1.2% | |
| Wallowa | 4,569 | 4,500 | 69 | 1.5% | |
| Wasco | 11,866 | 11,685 | 181 | 1.5% | |
| Washington | 224,850 | 223,276 | 1,574 | 0.7% | |
| Wheeler | 913 | 876 | 37 | 4.1% | |
| Yamhill | 40,794 | 40,417 | 377 | 0.9% | |
| Sum | 1,774,199 | 1,754,873 | 19,326 | 0 |
Chiang Mai Curry (Thai red chicken curry)
My SO and I made a favorite Thai curry last weekend, so I thought I'd post it up today as my end-of-the-week recipe blogging post. The curry has a coconut milk base and is packed full of flavorful ingredients (curry paste, ginger, garlic, fish sauce, turmeric, etc.) I brought some leftovers to work for lunch, and three separate people walking past my door asked what the delicious smells were. This recipe is modified from Bhumichitr, 1988.
5 tablespoons oil
4 medium garlic cloves, pressed (using a garlic press) or finely minced
3 1/2 tablespoons Thai red curry paste
1 can coconut milk (13.5 oz)
1 pound boneless skinless chicken, cut into thin strips (we use thighs)
3 inch ginger piece, peeled and grated
1/2 cup chicken stock (or water)
5 tablespoons fish sauce
4 teaspoons sugar
2 teaspoons ground turmeric
3 tablespoons lemon juice
8 cloves pickled garlic, pressed or minced (or 8 cloves fresh garlic pressed or minced and mixed with a bit of rice vinegar)
0) Start some rice cooking. Get all the above ingredients ready and lined up on the counter before you start cooking.
1) Heat the oil until it's very hot, then add the garlic and fry until golden brown (a few seconds).
2) Add the curry paste, mix in, and cook for a few seconds.
3) When the electric stove burner you're using starts to shoot out sparks, shriek and transfer the pot to a different burner (yes, this actually happened when we were making the recipe last weekend).
4) Add the coconut milk, stir, and cook until it starts to reduce (bubbling for a bit).
5) If you're not using real pickled garlic, press 8 cloves of the fresh garlic, mix with a bit of rice vinegar, and add this mixture while the coconut milk is reducing. If using real pickled garlic, add it later.
6) Add the chicken, stir well, and cook until it's nearly done (a few minutes).
7) Add, stirring after each addition: ginger, stock, fish sauce, sugar, turmeric, and lemon juice.
8) Cook for a minute or two longer to make sure everything's cooked.
9) If you are using real pickled garlic, add it now.
10) Serve.
We've also added various vegetables to the dish; last weekend we added a half-pound of frozen green beans a few minutes after adding the chicken, and they added a nice contrast, though if we add them again we plan to defrost them first, since they halted the curry's simmering.
This dish is extremely flavorful, and depending on the curry paste you use will likely be quite spicy as well; be sure to have lots of rice on hand.
We made this recipe with red curry paste that we bought in a 4 oz can at an Asian market. We used "Maesri" brand; its contents were 40% chili, 20% garlic, 20% onion, 6% lemon grass, 6% salt, 5% spice, 3% galangal, and it was made in Thailand. I've found it available online at importfood.com, though I have no idea if this is a good curry paste or not (it's just the only brand we've used). We've also used that manufacturer's yellow curry paste, and it's been just as good (and a bit less spicy).
Bhumichitr, Vatcharin, 1988. The Taste of Thailand. MacMillan, NY. Pp. 143-144.
[Update Jan 2006: Want to make a Thai curry but don't have the exact ingredients specified above? Take a look at our flexible Thai curry recipe for more ideas.]
5 tablespoons oil
4 medium garlic cloves, pressed (using a garlic press) or finely minced
3 1/2 tablespoons Thai red curry paste
1 can coconut milk (13.5 oz)
1 pound boneless skinless chicken, cut into thin strips (we use thighs)
3 inch ginger piece, peeled and grated
1/2 cup chicken stock (or water)
5 tablespoons fish sauce
4 teaspoons sugar
2 teaspoons ground turmeric
3 tablespoons lemon juice
8 cloves pickled garlic, pressed or minced (or 8 cloves fresh garlic pressed or minced and mixed with a bit of rice vinegar)
0) Start some rice cooking. Get all the above ingredients ready and lined up on the counter before you start cooking.
1) Heat the oil until it's very hot, then add the garlic and fry until golden brown (a few seconds).
2) Add the curry paste, mix in, and cook for a few seconds.
3) When the electric stove burner you're using starts to shoot out sparks, shriek and transfer the pot to a different burner (yes, this actually happened when we were making the recipe last weekend).
4) Add the coconut milk, stir, and cook until it starts to reduce (bubbling for a bit).
5) If you're not using real pickled garlic, press 8 cloves of the fresh garlic, mix with a bit of rice vinegar, and add this mixture while the coconut milk is reducing. If using real pickled garlic, add it later.
6) Add the chicken, stir well, and cook until it's nearly done (a few minutes).
7) Add, stirring after each addition: ginger, stock, fish sauce, sugar, turmeric, and lemon juice.
8) Cook for a minute or two longer to make sure everything's cooked.
9) If you are using real pickled garlic, add it now.
10) Serve.
We've also added various vegetables to the dish; last weekend we added a half-pound of frozen green beans a few minutes after adding the chicken, and they added a nice contrast, though if we add them again we plan to defrost them first, since they halted the curry's simmering.
This dish is extremely flavorful, and depending on the curry paste you use will likely be quite spicy as well; be sure to have lots of rice on hand.
We made this recipe with red curry paste that we bought in a 4 oz can at an Asian market. We used "Maesri" brand; its contents were 40% chili, 20% garlic, 20% onion, 6% lemon grass, 6% salt, 5% spice, 3% galangal, and it was made in Thailand. I've found it available online at importfood.com, though I have no idea if this is a good curry paste or not (it's just the only brand we've used). We've also used that manufacturer's yellow curry paste, and it's been just as good (and a bit less spicy).
Bhumichitr, Vatcharin, 1988. The Taste of Thailand. MacMillan, NY. Pp. 143-144.
[Update Jan 2006: Want to make a Thai curry but don't have the exact ingredients specified above? Take a look at our flexible Thai curry recipe for more ideas.]
Saturday, November 06, 2004
New mouse cages
My SO and I bought two more cages for our mice today, bringing our total number of cages to four. Three of the cages are CritterTrail, and the fourth is a Habitrail. All four of the cages are connected by tubes (the manufacturers' tubes interconnect), so our seven mice now have plenty of room to run around in. Here's the new setup:
Our new mouse cage setup. The two new cages are on the right, while the two cages we bought two weeks ago are on the left.
To hold the new setup we assembled two of our IKEA modular shelf systems in our living room, which will also give us much-needed storage space. We simply attached the two new cages to the two old cages (while the mice were still in them), and let the mice discover the new cages on their own. It didn't take long before Rem, Deuce, and Ace all found the new cages, and were happily exploring.
Ace (or maybe Deuce) looking into the new cage.
The mice all displayed nearly identical exploring behaviors at first. They crawled through the tube leading into a new cage, sniffed the new cage's interior (and maybe put down a paw or two), then turned around and rushed back to the familiar cage. Then they turned right around and walked back through the tube and into the new cage without looking back. It didn't take the mice long to learn how to get around the entire setup.
Ace exploring one of the shelves of our new CritterTrail Three.
Right now the mice are running all over the tall cage, and are having a blast. The new red wheel is a huge hit (they now have three wheels total), and they can't seem to get enough of climbing on the bars.
Our new mouse cage setup. The two new cages are on the right, while the two cages we bought two weeks ago are on the left.
To hold the new setup we assembled two of our IKEA modular shelf systems in our living room, which will also give us much-needed storage space. We simply attached the two new cages to the two old cages (while the mice were still in them), and let the mice discover the new cages on their own. It didn't take long before Rem, Deuce, and Ace all found the new cages, and were happily exploring.
Ace (or maybe Deuce) looking into the new cage.
The mice all displayed nearly identical exploring behaviors at first. They crawled through the tube leading into a new cage, sniffed the new cage's interior (and maybe put down a paw or two), then turned around and rushed back to the familiar cage. Then they turned right around and walked back through the tube and into the new cage without looking back. It didn't take the mice long to learn how to get around the entire setup.
Ace exploring one of the shelves of our new CritterTrail Three.
Right now the mice are running all over the tall cage, and are having a blast. The new red wheel is a huge hit (they now have three wheels total), and they can't seem to get enough of climbing on the bars.
Friday, November 05, 2004
Voting Problems
If you read any of the news immediately after the election, it sounded as though voting went off without a hitch. However, recent reports indicate that a number of problems did occur:
I've read some arguments that since Bush's lead is large enough that most errors wouldn't have changed the outcome, we should not even bother looking for errors. I find this illogical. First, we do not know the magnitude of any potential errors at this point; no major discrepancies have been proven to have occurred, but there also hasn't been much time to look for them. Second, even if correcting errors in this election would not change the outcome, it's still important to discover and correct the underlying causes so future elections will be more accurate.
- An Ohio electronic voting error gave Bush an extra 3,893 votes in Franklin County - Yahoo news and CNN
- A south Florida machine lost 70,000 votes for a ballot measure - reported in two Miami Herald articles (from the blog Blue Lemur)
- Craven County (North Carolina) had problems collecting precinct data, double counting ~11,000 votes, and then compounded the problem with human error before it was fixed - The Sun Journal
- LaPorte County (Indiana) officials discovered that their vote totals only listed 300 registered voters in each precinct, forcing them to stop the count. They're still trying to solve the problem according to the article - The News-Dispatch
- Palm Beach County (Florida) logged 88,000 more votes than voters - reported on the website Washington Dispatch; the county has since revised their numbers.
- Approximately 4,000 early votes were irretrievably lost in Carteret County (North Carolina) due to problems with an electronic voting system - The Sun Journal
I've read some arguments that since Bush's lead is large enough that most errors wouldn't have changed the outcome, we should not even bother looking for errors. I find this illogical. First, we do not know the magnitude of any potential errors at this point; no major discrepancies have been proven to have occurred, but there also hasn't been much time to look for them. Second, even if correcting errors in this election would not change the outcome, it's still important to discover and correct the underlying causes so future elections will be more accurate.
Thursday, November 04, 2004
The 15th Tangled Bank
I will hosting the next Tangled Bank here at Rhosgobel on Wednesday November 17th. If you want your latest and greatest science-related post included, just send your link to me at rhosgobel2@comcast.net (see the Tangled Bank page for more information).
The "toss out the state" game
Prashant Mullick just left a comment linking to a post describing why he feels Bush does indeed have a "Broad Nationwide Victory." (I wrote on this topic a few hours ago.) The crux of his argument is this:
I have a few words in response to this:
1) Why throw out CA and NY? They are, after all, the first and third most populous states in the nation. Are their votes somehow invalid or unimportant?
2) You can calculate similar differences in number of votes by throwing out a couple of conservative states. In Texas (the second most populous state) the margin of victory for Bush was 1.68 million votes, which makes up nearly half his margin of victory right there.
In fact, if we throw out
"A-HA!" you say, "but you threw out five states, not just two." But California's and New York's populations sum to 54.6 million, while the total population of the five conservative-voting states I threw out is 43.8 million (link to data). Thus, in throwing out these states I've actually thrown out fewer total people, so if anything I've been more fair than Prashant.
OK, clearly this is ridiculous. We could play the "toss out the state" game all night. The fact of the matter still remains: only about 20% of this nation actually voted for Bush on Tuesday, while another 20% or so voted for his opponent (and ~60% didn't vote for either one).
(state election vote counts from CNN at approximately 1am today)
"John Kerry victory margin:
New York: 1.19 million votes
California: 1.02 million votes
Combined total of NY and CA: 2.21 million votes
George Bush victory margin:
Nationwide: 3.56 million votes
This means that in the 48 states other than NY and CA that voted in this election, George Bush had a mandate of 5.77 million votes. That is pretty substantial!"
I have a few words in response to this:
1) Why throw out CA and NY? They are, after all, the first and third most populous states in the nation. Are their votes somehow invalid or unimportant?
2) You can calculate similar differences in number of votes by throwing out a couple of conservative states. In Texas (the second most populous state) the margin of victory for Bush was 1.68 million votes, which makes up nearly half his margin of victory right there.
In fact, if we throw out
- Texas – 1.68 million vote margin for Bush
- Georgia – 0.56 million vote margin for Bush
- Indiana – 0.51 million vote margin for Bush
- Alabama – 0.48 million vote margin for Bush
- Utah – 0.38 million vote margin for Bush
"A-HA!" you say, "but you threw out five states, not just two." But California's and New York's populations sum to 54.6 million, while the total population of the five conservative-voting states I threw out is 43.8 million (link to data). Thus, in throwing out these states I've actually thrown out fewer total people, so if anything I've been more fair than Prashant.
OK, clearly this is ridiculous. We could play the "toss out the state" game all night. The fact of the matter still remains: only about 20% of this nation actually voted for Bush on Tuesday, while another 20% or so voted for his opponent (and ~60% didn't vote for either one).
(state election vote counts from CNN at approximately 1am today)
Wednesday, November 03, 2004
"A broad, nationwide victory"
CNN reports the vice president as saying, "The result is now clear: a record voter turnout and a broad, nationwide victory." He's trying to argue that their administration now has a strong mandate from the country for their second term. That's just bull.
Let's look at the numbers. Based on the most recent CNN estimates, Bush/Cheney received 59,108,395 votes, while Kerry/Edwards received 55,554,114 votes (Nader received 395,871). Now, I'm no mathematician, but even I can realize that the margin of victory here was only 3.09%. To look at it another way, Mr. Cheney, there are at least 55,949,985 voters who did not agree with you, and who in fact wanted to send you home come January.
Let's also look at the total proportion of the population that voted. Yes, voter turnout was at a "record high", with CNN's best estimate being 119.8 million total voters. They report that this is approximately 60% of eligible voters.
However, America is not composed solely of eligible voters, so let's look at the total population, which the CIA World Factbook estimated to be 293,027,571 in July of 2004 (has the CIA ever heard of significant digits?). Thus, using CNN's estimate of total votes cast, only 40% of the US population voted. Of those 40%, only 51.3% voted for Bush, meaning that Bush only has evidence for the support of a whopping 21% of the nation's population (or ~30% of all eligible voters). This is not a broad mandate.
Admittedly these are not atypical numbers for a presidential election (margin of victory, turnout, and % of population all fall within the range of historical data; see Dave Leip's Atlas of Presidential Elections), but I still find it heartening that only one in five Americans actually voted for Bush.
(all 2004 voting numbers from CNN at ~8pm PST)
Let's look at the numbers. Based on the most recent CNN estimates, Bush/Cheney received 59,108,395 votes, while Kerry/Edwards received 55,554,114 votes (Nader received 395,871). Now, I'm no mathematician, but even I can realize that the margin of victory here was only 3.09%. To look at it another way, Mr. Cheney, there are at least 55,949,985 voters who did not agree with you, and who in fact wanted to send you home come January.
Let's also look at the total proportion of the population that voted. Yes, voter turnout was at a "record high", with CNN's best estimate being 119.8 million total voters. They report that this is approximately 60% of eligible voters.
However, America is not composed solely of eligible voters, so let's look at the total population, which the CIA World Factbook estimated to be 293,027,571 in July of 2004 (has the CIA ever heard of significant digits?). Thus, using CNN's estimate of total votes cast, only 40% of the US population voted. Of those 40%, only 51.3% voted for Bush, meaning that Bush only has evidence for the support of a whopping 21% of the nation's population (or ~30% of all eligible voters). This is not a broad mandate.
Admittedly these are not atypical numbers for a presidential election (margin of victory, turnout, and % of population all fall within the range of historical data; see Dave Leip's Atlas of Presidential Elections), but I still find it heartening that only one in five Americans actually voted for Bush.
(all 2004 voting numbers from CNN at ~8pm PST)
Despair
My hope for a positive outcome faded last night as I realized that the popular vote totals, and even apparently the electoral vote totals, were adding up for Bush. There is still hope, and like Pharyngula I have not given it up yet (and I hope that Kerry does not give up either, though reports indicate he may have), but I am distraught.
I am left wondering how more than half of my voting countrymen and women could have voted for an administration that is so obviously flawed, when there was a viable option that was so much better. Kerry was going to bring sanity and respect back to an office that has had so little of it these past four years, and I am disturbed that a majority of people in this country decided to vote against that.
Ever since 2000 I have consoled myself by hoping (and believing) that Bush would be a one term president, that we would see the error of our ways in 2004. I was wrong, and now I have nothing left to console me for the next four years.
I despair for our future.
I am left wondering how more than half of my voting countrymen and women could have voted for an administration that is so obviously flawed, when there was a viable option that was so much better. Kerry was going to bring sanity and respect back to an office that has had so little of it these past four years, and I am disturbed that a majority of people in this country decided to vote against that.
Ever since 2000 I have consoled myself by hoping (and believing) that Bush would be a one term president, that we would see the error of our ways in 2004. I was wrong, and now I have nothing left to console me for the next four years.
I despair for our future.
Tuesday, November 02, 2004
Anxiety
My SO and I both voted today, and we were able to use the paper ballot option without issue. They didn't ask us to put the ballots in provisional envelopes (unlike some areas); we just put them straight in the ballot box.
My students were pretty wired about the election today in lab, sharing regular updates about electoral vote counting (but largely ignoring our 16 state ballot initiatives). I was thankfully far too busy to keep up with results during the day (I only got my lab slides finished a few minutes before lab), but now that I'm home I'm anxiously refreshing the election results pages.
Here's hoping that Pharyngula gets to make those root beer floats ...
My students were pretty wired about the election today in lab, sharing regular updates about electoral vote counting (but largely ignoring our 16 state ballot initiatives). I was thankfully far too busy to keep up with results during the day (I only got my lab slides finished a few minutes before lab), but now that I'm home I'm anxiously refreshing the election results pages.
Here's hoping that Pharyngula gets to make those root beer floats ...
Goering Quote
Snopes has the full version of Hermann Goering's often-quoted description of how easy it is to lead a populace to war. They also have background on how the quote was obtained, which I'd never read before. The full version is below, along with a question asked by the psychologist who recorded the quote.
"Why, of course, the people don't want war," Goering shrugged. "Why would some poor slob on a farm want to risk his life in a war when the best that he can get out of it is to come back to his farm in one piece. Naturally, the common people don't want war; neither in Russia nor in England nor in America, nor for that matter in Germany. That is understood. But, after all, it is the leaders of the country who determine the policy and it is always a simple matter to drag the people along, whether it is a democracy or a fascist dictatorship or a Parliament or a Communist dictatorship."
"There is one difference," I [Gustave Gilbert, the psychologist] pointed out. "In a democracy the people have some say in the matter through their elected representatives, and in the United States only Congress can declare wars."
"Oh, that is all well and good, but, voice or no voice, the people can always be brought to the bidding of the leaders. That is easy. All you have to do is tell them they are being attacked and denounce the pacifists for lack of patriotism and exposing the country to danger. It works the same way in any country."
Subscribe to:
Posts (Atom)
Posts
