Leave a comment

Comments 47

garote October 23 2014, 08:43:41 UTC
Oh look it's you again, quoting the same study, but this time with no citation, and no explanation for what "highest poverty" means!

Did you make this yourself?

I'm no fan of Chris Christie, but I'm also not a fan of content-free rage-baiting clip art.

Reply


madscience October 23 2014, 17:27:26 UTC
And you still haven't figured out how to Google the phrase ad hoc ergo propter hoc.

Reply

red_pill October 23 2014, 17:52:05 UTC
the thing is, that this the main way we can work things out. now, dont get me wrong. we run lots of tests in say, phsycis or chemisty to make sure that its not fluke, but economics proves more diffcult. its a very big experement that can be diffcult reset. or more like impsoble. so we have to make infrences based upon things that happen ( ... )

Reply

madscience October 23 2014, 18:49:39 UTC
Only the shallowest analysis suggests that the short-term effects are poor. To explain it intuitively, if there's already a trend, and you make a policy change, you can't just look at the trend; you have to look at the change in the trend, and then compare it to the change in the trend in as many other places as possible. Then you use a technique called regression analysis to determine the relative weights of all the different possible causal factors you can think of. This is how we know, for example, that gun control has little or no effect on homicide rates, but stricter laws do decrease suicides while increasing other violent crimes. It's how we know that the biggest predictor of poverty in America is not race, but parents' income. It's how we know that the gender wage gap is around 2%, not 23% as the popular myth claims.

AFAIK, the claim that raising the minimum wage increases job growth has never been supported by this kind of analysis. If anyone has evidence to the contrary, they need to post a link.

Reply

red_pill October 23 2014, 17:52:32 UTC
also, i suspect you mean "post hoc" not "ad hoc"

Reply


Leave a comment

Up