Originally Posted by
Serra
I take offense at this.
Is the data as clear cut as hard science data? No, of course not -- there's a human element to the data and real experiments are generally impossible/frowned upon. But, any economist, political scientist, or other social scientist (those who actually do econometrics) know that their predictions are based on imperfect data. That's why things published in academic journals have statistical significance, confidence intervals, error terms, and things like that. It's not that they're saying "given x, y, and z, this is definitely what's going to happen," but rather "the population or sample of data makes us believe, with some degree of confidence, that given these factors, we will observe this phenomenon."
It's just people don't want to hear that stuff, don't want to hear about the statistical analysis, the methods used, the painful math, the assumptions made, etc. I'm not convinced that the majority of people in the fields really want to read all that. Most people just want to simply know, if we do X, what will happen?
Is there a lot of voodoo math involved? Unless someone can succinctly explain to me what robust standard errors are and what they actually do, you're not going to convince me there's not.
Am I saying econometric models are right all the time? No, of course not. But to just say, "lolEconomist can't predict anything because it's not a hard science" makes it seem like you don't know anything about econometrics.
And for the record, based on a cursory reading of Rothbard's wiki page, he's not one of those who actually believed in econometrics -- IE, he didn't believe in looking at actual evidence.