Hacker News new | ask | show | jobs
How Democrats, Republicans compare (sfgate.com)
12 points by Eight 5761 days ago
6 comments

Correlation is not causation.

Sure, people who are wealthier and live in more stable societies tend to vote Democrat; but that doesn't mean that electing Democrat representatives causes the societies to be wealthier and more stable.

Working hard and being wealthy are also correlated, but that doesn't imply that you can cause people to work hard by giving them lots of money.

(Lest anyone accuse me of political bias here: If I was in the US, I would be voting Democrat, just like the author.)

Your first line is a truism that your main argument doesn't support, and your third paragraph is an appeal to "common sense" that isn't supported by data.

Ignoring those, your argument boils down to your opinion that voting has no impact on wealth or stability. That cannot be true, at least not unless you subscribe to a deterministic world view, surely?

His argument is not that voting has no impact on wealth or stability, it's that you can't deconvolve voting's impact on wealth from wealth's impact on voting.

One could imagine a scenario where a Party X state had vastly impoverished natural resources compared to the typical Party Y state. It's conceivable that typical Party Y policies would have a worse outcome than typical Party X ones in that particular state, even if states that implement Party Y policies tend to have better outcomes overall. It's furthermore conceivable that even states that implement Party Y policies could do better if they had instead implemented Party X ones.

The flaw in this article isn't political, it's statistical.

edit: anonymized parties in the second paragraph.

your argument boils down to your opinion that voting has no impact on wealth or stability

No. I'm not arguing that there is no causation. I'm arguing that the correlations noted by the author are not evidence of causation.

Not that I disagree with the conclusion of the article, but I bet you can come up with an equally impressive set of stats showing the Republican party coming out ahead. If the article wasn't so one sided it would have been stronger.
Go ahead: come up with such a set.
The 'quiz' barely convinced me: * It's Congress that has shaped public policy in the United States for most of the post-FDR period, not the administration in the White House.

* Does public policy have tangible, next-day consequences? For example, is the population suddenly healthier, now that the vaunted Health Insurance reform has passed?

What's the "interesting new phenomenon" that this political article elucidates to make it not off-topic?
Attempting to measure the performance of a political party statistically and without being influenced by the personalities involved is interesting and new. The particular approach described is horrendously flawed, but it seems possible that there is good discussion to be had in proposing a better methodology.

Flag it if you think it's off-topic, and then something mysterious and not well documented might or might not happen. Personally, I decided it merited neither an upvote nor a flag.

"Attempting to measure the performance of a political party statistically and without being influenced by the personalities involved is interesting and new."

I've been presented with these exact same arguments back in High School, which was in the early 80s. You can skew the numbers to favor Republicans by just choosing the data ranges from when the politicians actually joined office.

In over two decades of being involved in politics in some capacity I've found that the only difference between the two major parties is how they want to legislate control of my life.

Sure looks like there's a strong correlation here, right?

http://en.wikipedia.org/wiki/File:US_Debt_Trend.svg

Err, or maybe not? Even more important, the President doesn't control the budget, congress does. What's the correlation between the budget deficit and the party in congressional control?

We'd be a lot better off if political discussions were founded on substance and reason rather than on silly half-backed "statistics" used to score lame partisan points.

> What's the correlation between the budget deficit and the party in congressional control?

None. Because the factor is actually the interaction between executive and legislative. When republicans have both seats, they have marching orders and follow them to the letter leading to rubberstamping of whatever the other seat wants and budget deficit increases.

When democrats control both seats, pretty much the same historically (might have changed with the Obama administration because Obama's house shows a distinct lack of balls and dem congresspeople haven't found their spine in the 15 years they've been looking for them).

Rep+Dem and Dem+Rep theoretically lead to the same thing, which is the necessity for "bipartisanship" (== exchanging favors actually), limiting the excesses of either side (Clinton admin).

Except now you have to get the current behavioral factors:

* Democrat congresspeople haven't had a spine in 15 years at least (since Reps got congress back in 1994), when faced with the GWB administration they mostly did nothing, from half cowardice and half political calculation (the GWB admin was not getting any better, letting it sink 2 more years and shifting any and all blame to the first 6 years would ensure a sweep in 2008)

* Republicans have gone completely batshit insane, they're practicing scorched-earth politics and since ~2007 they're not just pandering to the most insane side of their base, they're pandering solely to it. This voyage started with the Southern Strategy, Reagan's rhetorics gave it a sharp acceleration, GWB and 9/11 propelled into orbit and after 2007 finally got outside of sanity's gravitational well.

From a glance, his deficit and job growth assessments are likely sensitive to definitions -- what's the first year that counts for an administration? was Ford part of the Nixon administration? etc. -- and assumptions about lag effects. (Were the 70s a hangover from the 60s boom? The 00s from the 90s? How much of a ding should Obama, GWB, Reagan get for 09-10, 01-02, 81-82? Does Vietnam belong to Kennedy, Johnson, or Nixon?)

The reasoning about state-by-state indicators and presidential-party preferences is very bogus; you couldn't turn 'red' states into 'blue' states by adopting blue-state policy preferences; for one thing, the red states can't afford them! (Some of those preferences are luxuries purchased with wealth, rather than the original causes of wealth.)

Also consider the following: richer people lean Republican, but richer states lean Democrat. People with high school diplomas are more Republican than those without... but states with higher levels of high-school graduation are more Democratic. Same with undergraduate degrees.

These counterintuitive results for the aggregates are all examples of Simpson's Paradox:

http://en.wikipedia.org/wiki/Simpson%27s_paradox