News for hackers. Stuff that matters.
Re:Obama's Data by patio11 on Wednesday February 17, 2010 @12:51AM (#1130956)
This is the landing page for a political campaign. It has a graph on it. That makes it a landing page for a political campaign... with a graph on it.

Here's another page with a graph in it, from the same data source:

http://data.bls.gov/PDQ/servlet/SurveyOutputServlet?data_too...

Pesky things, those graphs.

Re:Obama's Data by isamuel on Wednesday February 17, 2010 @01:24AM (#1131004)
What is this new graph supposed to illustrate? I understand that by calling it "pesky" you're speaking in a sort of code, which is meant to signify factual victory on a disputed point.

But you're right: the charts show exactly the same thing. A rocketing unemployment rate beginning in early 2008, the rate of increase of which slowed by early 2009, and which has now leveled off. Both charts accurately show that there were huge job losses in one period, followed by a slowing and (for now) arrest of those losses.

The one you linked to is concededly uglier (pesky things, those ugly graphs) but otherwise it's unclear to me what it's supposed to show.

Re:Obama's Data by patio11 on Wednesday February 17, 2010 @01:43AM (#1131029)
What is this new graph supposed to illustrate?

An old hobbyhorse of mine: the perception of data is directly linked to how it is presented. (See also http://news.ycombinator.com/item?id=802439 although I think I've done it maybe three times on HN.)

Are you isamuel like "Ian who has known me for a decade" isamuel? If so, you already know this. If not, trust me on this: I have strong political beliefs, generally do not communicate them in code, and have almost zero desire to discuss them on HN. (Though if Obama is A/B testing that page I am so there for that discussion.)

Re:Obama's Data by mbreese on Wednesday February 17, 2010 @01:06AM (#1130983)
That's actually not the same data. The original graph is number of jobs gained/lost. Your link is the unemployment rate. In order for you to directly compare the two, you'd need to actually look at the derivative of the unemployment graph.

That's not to mention that the unemployment numbers aren't directly derived from the number of jobs gained/lost. You can start generating jobs and still have unemployment stagnant for a variety of reasons (population and workforce growth, for example)

So, while this is the landing page for a political campaign, it is an honest visualization of the data. Pesky things, those data.

Re:Obama's Data by ssp on Wednesday February 17, 2010 @10:15PM (#1133310)
There is a joke about politicians that you can tell how bad the economy really is by the order of derivative they have to use to get positive numbers. Apparently order 2 is sufficient.
Re:Obama's Data by deltaqueue on Wednesday February 17, 2010 @12:52AM (#1130957)
Re:Obama's Data by mbreese on Wednesday February 17, 2010 @01:32AM (#1131012)
How is deficit spending related to job gain/loss? One school of economics (I can't remember which), says that in a recession, the state is the only player big enough and with a long enough time horizon to take on debt to increase demand in the economy.

So, if you're comparing job gain/loss to deficit spending, then... spending more money correlates to a slowing in the rate of job loss.

Congratulations, you just demonstrated the opposite of the point I think you were trying to make.

And if you look at the graph, it doesn't make sense, at least in this context. One series is Total budget deficit, the other is increase in national debt. So, it looks more like a graph trying to show the changes in the national debt that aren't related to the budget deficit. That is to say, off-budget expenses (wars, etc...).

Re:Obama's Data by Subgun on Wednesday February 17, 2010 @12:55AM (#1130963)
Bravo