PDA

View Full Version : Correlations can be fun!


]LoL[Harm
04-20-2004, 12:58 PM
Correlations are known to be poor at finding direct cause/effect relationships. But they do open you up to trying to find out why things are happening the way they are. The following correlation is by no means a true indicator of the conclusion the writer comes to even though the writer states it as such, but it is an interesting correlation.

It does beg the question of, just where are my federal taxes going?

...That means that instead of funding much-needed highway and
transit projects here at home, a good chunk of the tax money paid by Georgia motorists ends up building projects in places such as Alaska, which sucks up an astounding $7 for every $1 it pays in federal gasoline taxes. The new transportation bill pending in Congress, for example, contains up to $2 billion to build a bridge that would stretch from Anchorage across a bay to an undeveloped port area several miles away.


That absurdly expensive bridge to nowhere is testament to the power of two men: U.S. Sen. Ted Stevens, an Alaska Republican who chairs the Senate Appropriations Committee, and U.S. Rep. Don Young, an Alaska Republican who chairs the House Transportation and Infrastructure Committee. Neither of the two is at all apologetic about the bill; in fact, Young even bragged to his home-state media that he "stuffed it like a turkey."

That's an outrage, of course. But once you start looking into it, this business of who gets what from the federal government turns up all kinds of interesting nuggets.

According to the nonpartisan Tax Foundation, for example, 33 states got more federal spending than they paid in federal taxes in 2002. That includes Georgia, but barely. For every dollar we paid the federal government, we got back $1.01 in federal expenditures. In other words, while our congressmen may have a valid complaint about transportation funding, overall we're treated pretty fairly.

New Mexico, with its disproportionately large Native American population, was the biggest of the "welfare states," collecting a tidy $2.34 in expenditures for every dollar it paid in federal taxes, followed closely by Alaska and Mississippi. The three states hardest hit were New Jersey, Connecticut and New Hampshire, which collected roughly 66 cents for each dollar that they paid in taxes.

Things really get interesting, though, when you start comparing tax contributions to political philosophy. For example, conservatives get a lot of mileage out of depicting themselves as rugged individualists, while liberals are supposedly eager to live off the government. Well, eight of the 10 biggest welfare states -- states that wrung a lot more out of Uncle Sam than they paid in federal taxes -- voted for George W. Bush in 2000. Conversely, of the 16 "donor states" -- states that paid a lot more in federal taxes than they got in return -- 12 voted for Democrat Al Gore. In other words, the conservatives are right: The federal government does indeed act as an income redistribution system. But it takes money away from generally Democratic states and pays it to generally Republican states. The same is true at the congressional level. In 2000, according to an Associated Press computer analysis, districts that elected supposedly self-reliant, anti-government Republicans got an average of $621 million more in federal expenditures than districts that elected the supposedly government-reliant Democrats...
Full article here: http://www.commondreams.org/views04/0419-07.htm

Allison
04-20-2004, 01:10 PM
That is interesting. I will ponder now.

Sparky
04-20-2004, 01:22 PM
States with lower population tend to be conservative/republican. Lower populations means it's easier for a senator to get re-elected. More terms in D.C. means more seniority in Senate / House. More seniority means being appointed to awesomeness committees like appropriations.

On the flip side states with a higher population tend to be liberal/democratic. Higher population means it's harder to get re-elected. Less terms in D.C. means crappy committees.

That's why i'm moving to Montana when I run for Senate.

Allison
04-20-2004, 01:40 PM
That is interesting, as well. I will ponder more.

Rooster
04-20-2004, 02:09 PM
lol @ Alli

Interesting reading, and MAN does that guy have a lot of time on his hands (I love stats like that, but one must be careful not to read too much into it).

Tammarion
04-20-2004, 02:18 PM
I had a prof who one pointed out why using correlations can be so misleading by comparing the number of Bars in a given town, with the number of Churchs. :)

Eiru
04-20-2004, 03:08 PM
Haha, Tamm! That example reminds me of some Seattle lore. One neighbourhood here, Ballard, had a folk legend that when the neighbourhood was founded, an unwritten agreement between everyone concerned meant that for every bar built, a church had to be build. There's nothing on record officially about this, but there is a 1 to 1 ratio between bars and churches in the neighbourhood.

Sparky
04-20-2004, 03:33 PM
Ice cream consumption has a positive correlation with criminal activities. :X :X

.
.
.
.


of course the more likely answer for increased criminal activity in june/july was increase in temperature at night which also accounted for increase in ice cream assumption :P

Allison
04-20-2004, 04:15 PM
Ban ice cream!!!

Rooster
04-20-2004, 05:29 PM
Ban warm weather!

Eiru
04-20-2004, 06:19 PM
Banana!

]LoL[Harm
04-20-2004, 08:54 PM
And I bet more Republicans bought more ice cream therefore they ARE FUNDING CRIME!

Allison
04-20-2004, 09:03 PM
:rolly:

Hammer
04-21-2004, 12:02 AM
Seems pretty obvious to me. Republicans control both houses of congress. To the victor go the spoils. Republican congressmen are just as good as spending other peoples money as their colleagues across the aisle.

I wonder what the stats looked like before 92?

Boom
04-21-2004, 01:52 PM
Its like those anti-drug commercials where they say that 20% (or whatever) of people in car accidents tested positive for marijuana. What does that prove? Maybe 20% of people in car accidents would test positive for ice-cream! Or maybe 20% of people in car accidents like country music.

My favorite is the one showing the little girl playing by the pool and about to fall in. Then the voice over says "Tell her parents you weren't watching her because you were smoking marijuana." What kind of crazy logic is that? The girl isn't gonna drown cuz of the pot, she is gonna drown because of the not being watched. The voice over could just as well say "Tell her parents you weren't watching her because you were doing your homework." BAN HOMEWORK! Or "Tell her parents you weren't watching her because you were eating ice cream." See, once again ice cream is the culprit. Or how about "Tell her parents you weren't watching her because you were watching misleading anti-drug commercials."

spyder913
04-21-2004, 02:01 PM
'Tell her parents you weren't watching her because you were inside making out with your same sex lover!'

Boom
04-21-2004, 02:05 PM
Tell your parents you weren't watching her because you were reading the Bible.

]LoL[Harm
04-21-2004, 02:06 PM
I think we should just twist it around and use it to attack telemarketers!

'Tell her parents you weren't watching her because you were inside trying to hang up on a telemarketer!'

spyder913
04-21-2004, 08:36 PM
You can use the same sentance to blame anything.

'Tell her parents you weren't watching her because her other brother ran into the street'

Eiru
04-21-2004, 09:20 PM
'Tell her parents you weren't watching her because you were already face down after raiding the booze cabinet'