Search This Blog

Monday, March 01, 2004

Data Rebuttal Rebuttal

Jim's rebuttal to Where's the Data? is in italics, my comments are in normal typeface.
Bret, here's a different perspective on the statistics (which I came across in 2 minutes of searching on the Internet).
Unfortunately, it looks like you only spent two minutes and didn't bother reading the article.
 I have no idea who the author is, but his numbers seem just as good as yours.
His numbers are the same as mine, too bad he completely ignores his own numbers when he writes the text.  I'll lead you through it if you need me to, but if you take a closer look you'll see that his numbers don't provide much support for his text, and his numbers don't in anyway refute my assertion that tax rate cuts increase growth and those numbers actually support my assertion that "the vast majority of times a country raises taxes, its economic growth slows from the baseline rate." The one thing I will concede is that I meant to say "tax rates" where I said "taxes" in my post.
 He concludes:
"Nobel laureate Robert Lucas, one of the world's most famous conservative economists, has spent over a decade looking for the secret to economic growth, and has not found it.
I don't think Robert Lucas spends a lot of time looking for "secrets" to economic growth.  Like most of the University of Chicago economists, he's very data driven. Lucas also professes to know a lot about economic growth. Here is an excerpt from a Business Week article regarding Lucas:
Considering the economy's parlous state, the timing seems strange for a Nobel laureate in economics to proclaim the triumph of macroeconomic policymaking in managing the business cycle. Yet that's exactly what University of Chicago economist Robert Lucas did in an address ("Macroeconomic Priorities") at the annual meeting of the American Economics Assn. in Washington, D.C. in early January. And Lucas, a pioneering theorist and intellectual leader of the classic school of economic thought, is always worth paying close attention to, especially when he concludes that the future of economic policymaking lies in focusing on the supply side of the economy.
That would be growth that Lucas is trying to increase with those policies.
Nobel-bound Paul Krugman, one of the world's most famous liberal economists, admits that the mystery of growth is "deep and poorly understood."
Krugman may well be clueless, no argument from me here.
People who claim that tax rates affect growth are not serious economists; more often they are journalists, radio-talk show hosts, politicians and other types of snake oil salesmen with easy solutions to complex problems. You can dismiss their bumper sticker slogans with perfect confidence."
Hmmm. If tax rates don't affect growth, why not just raise them to 100% and go back to communism?  I'd be quite surprised if even 1% of economists believed that there is no relationship whatsoever between tax rates and growth.  I find the ad hominem is usually a good clue that I can safely ignore the author.
My point is not to argue about the data. In fact, I suspect a good statistician can pretty much prove anything by drawing selectively from the huge pool of existing economic data (with the possible exception that communism worked out well).
No, only bad statisticians prove things that aren't supported by the data.  Good statisticians are able to find meaningful correlations that further understanding of the world.
I was reading a presentation this morning about how good ideas get transmitted in a corporation. In fact, presentation of charts and data have very little impact in driving improvements. Story-telling is where the real leverage comes from. This may be good news for Harvard grads, and not so good news for MIT grads. It also may explain why people from MIT are often considered nerds who are so attuned to the data that they are out of touch with the stories of "real" life. I certainly can acknowledge that this has been an occasional fault of mine.
Are you saying that we would do better by living in a fantasy world of stories rather than trying to understand real causation and correlation?  I'm not saying that story-telling isn't an effective way to leverage an idea.  But wouldn't it be better if the idea that's being leveraged at least had some foundation in reality?   Perhaps a team of Harvard and MIT would be best?

In conclusion, I don't consider Jim's post a very convincing rebuttal.

No comments: