Editorial: Stimulating Incumbency

| | Comments (2)


Just got to mention that I really could not resist writing about an election. Politics is by far my favorite sport as no other so effectively combines the mindless, underhanded brutality of undergound street fighting with the shameless theatrics of professional wrestling. That said: 

The editorial I read was by George Will, entitled Stimulating Incumbency. This link is to my hometown's newspaper's website, but you need a password to get in. However, I think he's syndicated, so you can probably find it somewhere else.

His article is discussing the stimulus packages and their relationship to the upcoming midterm elections. It wasn't quite as witty as others I've read, but it did offer much more structured, and better supported argument than others I've read.

The main idea that it points out is how often arguments tend to hinge on statistics that are completely inquantifiable. The number of jobs that can be considered to have been "created" by the stimulus packages tends to come out as a negative number, since unemployment kept increasing after them. For this reason the statistic that tends to be used, is the number "saved." However, this number, while considered an estimate, is pretty immpossible to determine. As a statistic, it is emensely valuable, as arguing against an estimate is simply fighting shadows. Will's article is effective because he doesn't argue that this statistic is necessarily wrong, but simply points out that no one really knows.

He also suggests that the third stimulus will likely come pretty close to the election, and will mostly be intended to save jobs, within congress. Will is certainly accurate that it'll take some pretty deft manuvering to keep even the slimmest of majorities, let alone attain the  filibuster-proof one they'd always been dreaming of. Polls cited in another article, suggest that only half of the voters who came out for Obama will vote in the midterm, versus 66% of McCain voters. This sort of backlash against the party holding the presidency is pretty typical (Think 2006). While the uninformed of the victorious party rest on their laurels, the uninformed of the other party actually bother to vote in midterm elections. Beyond that, Democrats tend to need help from the young, yet far too busy to vote in boring midterm elections unless they're getting extra credit group, while the Republicans have the advantage of holding on to much of the old, and have nothing better to do group (many of whom have a significant intrest in health care).

Okay, I ended up off topic, but I'm fine with that. (At least I didn't say anything to outrageous, as I tend to when I start writing/talking/thinking about politics...I try to save that for my other blog, which no one reads.) I guess all that really matters is that this next election should be a bit more entertaining than the last, as they wont waste all their coverage on presidential candidates, which became pointless about 4 months prior to the election, as the conclusion got really obvious. Anyway, for now I'll give 100:1 odds to anyone betting the Dems will get that 3/5s (You can count Lieberman, but no other Independants). I'll figure the rest out later.



Aja Hannah said:

I don't understand all of what you said (I'm not versed in politics at all really), but it was interesting/eye-opening what you summed up about stats. A few teachers have said to me before that you can't trust stats especially when you can't verify them (which seems to be the case here). There are also those stats that fly around like "5/4 people don't know how to read a stat" or "98% of statistics are made up on the spot" so it's good to be skeptical and, in this case, to see how words like "created" become "saved" so as to offer a better looking statistic.

DavidWilbanks Author Profile Page said:

Its definately necessary to verify stats...I typically don't use anything in a paper unless I can get my hands on the raw data, rather than sections of it cited elsewhere. That also gives me the ability to calculate my own percentages, ratios, etc. Far too frequently, people calculate various percentages, etc. until they find one that is useful. MADD is a good example of this, as they tend to cite the reduced number of alcohol related accidents in people in the 16-20 group after they got the drinking age raised, while ignoring the nearly equal increase among the 21-25 set. They are just one of thousands of examples though.
The other misuse of statistics that is FAR more prevalent is the flawed logic that suggests that correlation = causation. People CONSTANTLY make claims suggesting relationships between phenomenons that data doesn't support. They seem to think that an increase in one thing, accompanied by an increase in another means that one causes the other. Without a control group, this is simply immpossible to assert with any accuracy. The sad thing is that in many cases it isn't just manipulative interest groups, and fanatics doing this, but otherwise intelligent people.
As far as statistics being made up, the 98% statistic seems, well, made up on the spot (which further illustrates its point). I would say that when they're accurate, they are extremely useful. However, scepticism is important, because while numbers don't lie, people using them are often liars, biased or simply just idiots.

Leave a comment

Type the characters you see in the picture above.