The Importance of Electoral Votes
Given the comments on my post on political demographics, I thought I'd spend a bit of time reviewing the electoral college and our current population.

All fifty states plus Washington, DC have a total of 538 electoral votes. 270 electoral votes are needed to win the presidency (~50.2%). Every ten years the US government has a very detailed census that is used, amongst other things, to apportion these electoral votes. Thus politicians argue about what statistical methods are used. Despite this, the Census department does a remarkable job performing a difficult task. The last major census was performed in 2000 and these current population figures will be used in determining the electoral votes for each state in November's elections.

I've created a table showing how many electoral votes each state has as well as how this has changed since the last election. The green highlights are for states that now have more influence than they did in the 2000 election. Likewise, the red highlights are for states that now have less influence than they did in the 2000 election.

 
1991-2000 Electoral Votes
2001-2010 Electoral Votes
State
Votes
Percent of Total
Votes
Percent of Total
Change from Last Census
California 54 10.0% 55 10.2% 0.2%
Texas 32 5.9% 34 6.3% 0.4%
New York 33 6.1% 31 5.8% -0.4%
Florida 25 4.6% 27 5.0% 0.4%
Illinois 22 4.1% 21 3.9% -0.2%
Pennsylvania 23 4.3% 21 3.9% -0.4%
Ohio 21 3.9% 20 3.7% -0.2%
Michigan 18 3.3% 17 3.2% -0.2%
Georgia 13 2.4% 15 2.8% 0.4%
New Jersey 15 2.8% 15 2.8% 0.0%
North Carolina 14 2.6% 15 2.8% 0.2%
Virginia 13 2.4% 13 2.4% 0.0%
Massachusetts 12 2.2% 12 2.2% 0.0%
Indiana 12 2.2% 11 2.0% -0.2%
Missouri 11 2.0% 11 2.0% 0.0%
Tennessee 11 2.0% 11 2.0% 0.0%
Washington 11 2.0% 11 2.0% 0.0%
Arizona 8 1.5% 10 1.9% 0.4%
Maryland 10 1.9% 10 1.9% 0.0%
Minnesota 10 1.9% 10 1.9% 0.0%
Wisconsin 11 2.0% 10 1.9% -0.2%
Alabama 9 1.7% 9 1.7% 0.0%
Colorado 8 1.5% 9 1.7% 0.2%
Louisiana 9 1.7% 9 1.7% 0.0%
Kentucky 8 1.5% 8 1.5% 0.0%
South Carolina 8 1.5% 8 1.5% 0.0%
Connecticut 8 1.5% 7 1.3% -0.2%
Iowa 7 1.3% 7 1.3% 0.0%
Oklahoma 8 1.5% 7 1.3% -0.2%
Oregon 7 1.3% 7 1.3% 0.0%
Arkansas 6 1.1% 6 1.1% 0.0%
Kansas 6 1.1% 6 1.1% 0.0%
Mississippi 7 1.3% 6 1.1% -0.2%
Nebraska 5 0.9% 5 0.9% 0.0%
Nevada 4 0.7% 5 0.9% 0.2%
New Mexico 5 0.9% 5 0.9% 0.0%
Utah 5 0.9% 5 0.9% 0.0%
West Virginia 5 0.9% 5 0.9% 0.0%
Hawaii 4 0.7% 4 0.7% 0.0%
Idaho 4 0.7% 4 0.7% 0.0%
Maine 4 0.7% 4 0.7% 0.0%
New Hampshire 4 0.7% 4 0.7% 0.0%
Rhode Island 4 0.7% 4 0.7% 0.0%
Alaska 3 0.6% 3 0.6% 0.0%
Delaware 3 0.6% 3 0.6% 0.0%
D.C. 3 0.6% 3 0.6% 0.0%
Montana 3 0.6% 3 0.6% 0.0%
North Dakota 3 0.6% 3 0.6% 0.0%
South Dakota 3 0.6% 3 0.6% 0.0%
Vermont 3 0.6% 3 0.6% 0.0%
Wyoming 3 0.6% 3 0.6% 0.0%
Total 538 100.0% 538 100.0% 0.0%

You can see a simpler version of this table, in alphabetical order, courtesy of the US government.

I hear many Republicans complain about how much influence the Democrats in California and New York have on electing the president. I hear many Democrats complain about how much influence Texas and Florida have in electing the president. As an Independent, I'm very happy that the same party is not strong in all four states. Those four states alone represent 27.3 percent of all electoral votes (or 54.4% of the 270 votes needed to win). So it is probably a good thing for the other 46 states that these four do not all have the same perspective.

This chart also shows why Kerry's choice of a potential vice president is so important. If he can pick someone who can help him swing some key states (like Ohio or Pennsylvania), it will have much more of an impact vs. picking someone who can carry the Dakotas.

 
 
Comments

"Thus politicians argue about what statistical methods are used."

NO Statistical method is used. The Democratic Party keeps arguing we should use a statistical method, ie take a sample and then extrapolate what the entire population is from that sample. The constitution says the census shall be "an actual enumeration" ie you count them. (No sampling involved. You do your best to count every one and in any case, the number that you count is the number you use, not some mathematical function that uses that count (among other inputs) to arrive at an answer.

There is a big difference between those two. A statistical method (take a sample and calculate the whole population from the sample) is open to all sorts of shenannigans. Counting is harder to manipulate.

So far we have managed to hold the line on that revision of the Constitution.

Posted by: Drew | 04/29/2004 - 01:54 PM

Do you have at your finger tips from looking all this up what the total change in population was from 1990 to 2000? IE how much did the US population increase?

Posted by: Drew | 04/29/2004 - 01:55 PM

IBM

Does anyone in the crowd no the origin of International Business Machines (IBM)? As a clue, the reason that IBM exists is part of this topic AQ brought up.

Posted by: Drew | 04/29/2004 - 01:56 PM

California

Are you surprised that the % increase for California relative to other states is so low? We have HUGE immigration from latin america and from asia.

For the first time we had nearly equal exodus from California, though. People seem to now be using California as a way-station. They alight here when first coming to the US but are now spreading out among the other states after a few years/a generation.

Our recent fiscal/energy/government problems have also made California less a desirable place to stay, urging people to move to other states.

More on that last when I get the California Direct Democracy piece done.

Posted by: Drew | 04/29/2004 - 01:59 PM

In 1896 Herman Hollerith, statistician at the US Census Bureau, formed the Tabulating Machine Co., after constructing his Punch Card and Tabulating Machine in 1890. In 1911 this Company merged with Computing Scale of America and International Time Recording Co. The new name was The Computing-Tabulating-Recording Co., or C-T-R.

In 1915 Thomas J. Watson became the charismatic leader of the company. His slogan was: Think. . In 1924 the Company's name was changed into International Business Machines Corp. or IBM.

Posted by: khobrah | 04/29/2004 - 02:05 PM

Statistical Methods
Perhaps I should have gone into more depth here, but I thought the post was long enough. Yes, as a direct result of the aforementioned political arguing over statistical methods, the Supreme Court has prohibited the Census Bureau from using statistical sampling for the purposes of congressional apportionment purposes (which presumably also prohibits this for the electoral college, but I'm not completely positive on this). The Constitution also requires a count, but I believe the Census Bureau had ignored this for a while until the Supreme Court weighed in (I'm also not 100% sure on this either).

However, just as an experienced sergeant once told me after an officer was gone, sometimes those in charge don't need to know all the details. Those collecting all the survey information do not always get a direct count. If someone is not home, they are supposed to ask a neighbor who may or may not know the right answer. Sometimes people make things up. However, this is still a vast improvement over allowing people to use statistical techniques that would be used to give some states more representation based on how many people may have been missed by the census.

In addition, statistical techniques are used all the time in the census for other uses. For example, 1 out of 6 census forms is the long form. The census uses statistical techniques to take this data and apply it to the entire population. As you might expect, there is much disagreement about how this is done. And since this type of statistical manipulation does not impact elections ("merely" impacting policy decisions on welfare, etc.), it is perfectly legal to do so. And there are legitimate reasons for doing so – most of the controversy is about how it is done, not why it is done.

Posted by: Don Quixote | 04/29/2004 - 02:24 PM

IBM link - Kudos to Khobrah - the king of trivia.

Posted by: Don Quixote | 04/29/2004 - 02:25 PM

Do you have at your finger tips from looking all this up what the total change in population was from 1990 to 2000? IE how much did the US population increase?

In 1990, the US population was 249 million. In 2000, it was about 281 million. Now, it is about 293 million (as all good marketers know – you should always know how many potential customers you may have).

If you want specific details (such as the numbers on California), click here.

Are you surprised that the % increase for California relative to other states is so low?
Actually I'm not surprised, but you are talking to someone who studies demographics and trends. In the last few years of the 90s, I believe California may have actually lost people, but I don't have time to confirm this with the census department. I think this was short-term fluctuation vs. a long-term trend. I do believe there is a long-term trend of people leaving California, but as you say, this is about washed out by the number of immigrants (and new births). So my prediction is for California to continue to grow, but relatively slowly when you consider those fleeing the state.

Posted by: Don Quixote | 04/29/2004 - 02:41 PM

Wow! I'm a "king"

At most sites I'm just a joker!

Posted by: khobrah | 04/29/2004 - 03:01 PM

Khobrah has it!

The US had a census every even decade from 1790. But as the population grew, it took longer and longer to total all the numbers.

I think the 1890 census was the last done "manually" and it took so long (about 9 years) that it was pretty evident the next census would take MORE than 10 years to total. IE they would have the data for the next census before counting all the individual returns from the prior census. So they had to automate it and along came T.J. Watson.

Khobrah wins . . .

A handful of old programs punched out on IBM cards. (My basement has all kinds of stuff. Never can tell when the punch card reader will be a useful peripheral again)

Posted by: Drew | 04/29/2004 - 04:14 PM

Probably showing my age, but my first programming was done in ForTran on puched cards.

Posted by: khobrah | 04/29/2004 - 06:44 PM

Young Pup! I did my first programming in ALGOL. They didn't have that namby-pamby FORTRAN until I was in Grad School. Why I remember walking 3 miles to the comp center . . .

At this point they wheel Gramps away, but he continues to mumble through his dentures all the way back to the Old People's Home.

Posted by: Drew | 04/29/2004 - 06:52 PM

I've never used punch cards...but I saw both the cards and a reader once.

Think how easy programmer pranks would have been back then. Just grab your co-worker's card stack while he is at lunch and take one...or add one...or move one out of order.

Nowadays, such michief requires hacking and all that.

Posted by: King of Fools | 04/30/2004 - 07:17 AM

Drew - I have no doubt that you were programming before me, but just to keep things straight, ALGOL was an offshoot of FORTRAN 1

check out :

http://people.mandrakesoft.com/~prigaux/language-study/diagram-light.png

Posted by: khobrah | 04/30/2004 - 07:53 AM

King, no need for that, I once watched a guy trip over a cable and drop a box of 200+ cards all over the floor! Unfortunately, this was in an academic institute and we didn't have a fancy card sorter, so they all had to be manually re-sorted. Wheee!

Posted by: Khobrah | 04/30/2004 - 07:56 AM

Suspicion Confirmed:

The diagram Khobrah mentioned shows Fortran I (1956), and LISP (1958) to be among the oldest languages. But look: there is Cobol (1960), as early as Algol (1960).

My _wife_ learned Cobol as her first language (well, after English). Guess how old that makes her! *G*

Posted by: Drew | 04/30/2004 - 09:06 AM

In order of when they were learned:

Atari BASIC
FORTRAN
COBOL
RISC Assembler
C
CISC Assembler
ADA
XBase (dBase 3+, dBase IV, FoxPro)
C++
VB
HTML
SQL Query
VBA
Java
PHP
Perl

and probably some I've forgotten (LOGO, anyone).

Nowadays, I only use 4 or 5 with any regularity.

Posted by: khobrah | 04/30/2004 - 09:26 AM

I'd agree that Fortran is older than Algol. But I'd disagree that Algol is an "offshoot" of Fortran. The syntax just does not bear much resemblance. Frankly, Fortran was awful (what I saw of it) - Hollerith codes - Format statements - pah!
Algol was elegant - (if rather full of quote marks) and was the father of most of what followed - Pascal, Coral etc. Not much of Fortran seems to have survived?

Errrp! - 'scuse me - I am so far outside my knowledge base I'm just looking forward to being "shot down in flames" ! 8) Hardware Engineer Me!! 8)

I first started writing in Algol back in the late 70's when I was at school. We had an old Elliott 903 computer and I did my O'Level using it. Teletypes and paper tape - punched cards are before my time! A few years later the Atari 400 and BBC home computers came on the market. I now own that same Elliott 903 - one of only a few left in the world (that I know of).
It's the size of two chest freezers (no jokes please!)
http://www.austinfs.fsnet.co.uk/machines/elliott_903.html
(no I am not Jim Austin!)
You can also find an Elliott Emulator on the web if you want!

Anyway - I see no serious linkages that indicate the Algol is an offshot of Fortran. If anything - it seems Algol gave birth to most of modern structured progamming and Fortran kind of died out?

Eeeeaaaaarccgggghh!!! BLAM!!!
(me being shot down in flames!)

Lock and load chaps!

Posted by: Cassivellaunus | 04/30/2004 - 05:36 PM

Fortran is considered to be the first "structured" language (as you pointed out - those "Write using Format" couplets were an absolute bitch for syntax). As such, it is possible to make a case for it being the parent for most modern programming languages. Fortran is still in use, the lastest official spec is Fortran90 (which replaced Fortran77)
There are Fortran compilers for the PC, IIRC MicroFocus makes a pretty decent one.
As for emulators, there emulators for everything from Atari 400/800/1200 and Commodore 64 to any Arcade ROM (using MAME) and Cray, Elliot, Dec and many more.

The MAME emulator, BTW is absolutely superb!

Posted by: khobrah | 04/30/2004 - 08:34 PM

Census and Immigration:

US Census data is available at:
US Census Bureau

US Immigration data is available at:

US Citizenship and Immigration Service


In brief, the US Population is about 280 million. It grew by about 30-32 million in the last 10 years (since the last census). Call that 1%/year. Oops, Khobrah is here. Call that 1.18%/year (approximately).

Of that growth, about 1/3 was immigration, 2/3 excess births here (births in exces of deaths). We have about 1 million (in 2002 1,063,732) new citizens per year from immigration. Also remember that the birth rates of more recent immigrants tend to be higher than birthrates of families that have been here for many generations, so the fraction of our population that are either immigrants or descendents of the recent immigrant groups are high.

Mexico (219,380 immigrants) is the most common country of origin. The next in order are:

India 71k
China 61K
Phillipines 51K
Vietnam 34K
El Salvador 31K
Cuba 28K
Bosnia/Her 25K (first European Country)

By continent:
North America 404K (more than 1/2 from Mexico)
Asia 342K
South America 74K
Europe 174K
Africa 60K
Oceania 5K

The point of this is that although the US used to have a "came here from Europe" Culture we are changing and moving towards a "came here from Latin America or Asia" culture.

Posted by: Drew | 05/03/2004 - 04:30 PM

Nice looking URLs there Drew :-)

BTW, if any lurker wants to use Drew's numbers, make sure you notice they are from 2000 (at least the census numbers - I don't look at the immigration numbers on a regular basis).

Our current (2004) population is about 193 million (the census link Drew created has a nifty population clock on the main page).

Posted by: Don Quixote | 05/03/2004 - 05:33 PM

The Admiral meant to type "about 293 million" not 193 million. I guess those opposable thumbs got in the way again (for those who don't know me in real life, I am a very highly evolved rodent, not a primate).

Yes, the figures I gave are from the 2000 census.

Posted by: Drew | 05/03/2004 - 05:40 PM

Oh yeah, and the Immigration numbers are from 2002. That is the last date I could find quickly for which the real data, as opposed to an extrapolated estimate, are available. Those web sites I pointed too have all sorts of other data including estimated extrapolations into the future.

(Like how many 65 year olds and how many 85+ year olds there are).

Posted by: Drew | 05/03/2004 - 05:43 PM

The Admiral meant to type "about 293 million" not 193 million.

Yep, thanks. One day I'll see what that "Preview" button does...

Posted by: Don Quixote | 05/03/2004 - 07:39 PM
 
 
Send this Post
Email this entry to:


Your email address:


Message (optional):