I am certain the question has been extensively analyzed by high-powered economists armed with all the econometric and statistical tools. Someone might have the time to search out those papers and extract the analysis . . . but I have to go to dinner, so won't be me.
Quote:
Originally Posted by Superman
Some of the graphs I see above are tantalizing. At a glance, I thought they support the notion that lower tax rates cause higher revenues. Seat-of-the-pants first glance. But then I noticed there are plenty of instances where higher tax rates caused (in the following period or two or three) higher tax revenues and visa versa. So....I think they don't answer the query.
To really answer the query would be fairly complex, but most definitely do-able. A correlation/regression analysis would identify the most powerful predictors, and would quantify their ability to predict revenues. Multicollinearity could be backed out. Sure, the tables of data to be used as variables would need to be cleaned up and of course here is the best opportunity for shenanigans. But it is nearly impossible that some economist, or group of economists, have done this analysis. Independently. Out of honest curiosity. If they have, and I think they have.....then we would know how tax rates affect revenues, and how forceful the relationship is. And we would see other relationships also.
I would find this interesting. And quite frankly, if there is a variable out there that does a good job of predicting revenues, and if that variable were as simple as verifying that tax rates have a strong negative correlation to tax revenues beginning in the following year (or current year for that matter), then this is a no-brainer for Congress. Any congress member, liberal or conservative.
Don't tell me that Democrats know this and their agenda is to hurt American families more than raise revenues. It would mean you are an idiot.
If this connection is verifiable, which I believe it is NOT.....but if it were, then we could easily get Congress to act.
|
I will say that I've played around with some simple regressions on the data - turns out the sample size is small (we're only talking a couple dozen years), there are evidently many factors, and some of the coefficients you get (well, that I got) turn out to be statistically not significant. It is easy enough to find positive coefficients with some statistical significance when regressing tax rev gro on same-year effec tax rate. But finding statistically significant coefficients, pos or neg, for tax rev gro on prior-year effec tax rate - I could not. But, again, this was just playing around - time's probably better spent finding the papers written by the actual economists.