A Look at What the Next President’s Chief Economic Advisor Wrote About Tax Policy
Since Christina Romer is gonna be Obama’s Chair of the Council of Economic Advisors, its worth revisiting her work. Now, I’m not an academic, so I don’t read all that many papers. But I remember reading the March 2007 Romer and Romer paper as it made quite a spash at the time. (There’s a Nov 2008 update of the paper up now, but a quick skim indicates it doesn’t seem to address the problem I have with the paper which I’ll be getting to in a moment.)
The paper finds that at least some tax increases are highly contractionary. Now, when I read a paper with an empirical component, I pay extra attention to the data used in the analysis. Especially if the results contradict my intuition, and even more so if the results don’t match what I myself have found. Sometimes, I learn I was wrong. Sometimes I learn I wasn’t. Either way, looking at the data teaches me something important.
Now, I hesitate to criticize, because Romer and Romer are very accomplished individuals, and I am, quite frankly, a nobody who has gotten into a lot of trouble a few times for questioning my betters. But sometimes my curiosity gets the better of me, and in this instance, the first sign of trouble, to me, is the fact that the paper doesn’t actually provide the data – instead, it makes reference to a narrative record. And this graph (insert figure please!) has a few data points that just don’t match my recollection. If you want to play along at home, I suggest you start with the most recent change in the figure. Anyone remember the great tax hike of 2005? In the graph, it is bigger than the tax increase in 1993 which most of us do remember, and its a lot more recent, so it should be right there, in the forefront of your brain, right? Well, to be honest, I don’t remember the great tax hike of 2005 either.
So its off to see the
wizard narrative record which hopefully will clear up the mystery of the great tax hike of 2005. And Christina Romer’s website at Berkeley does have a November 2008 version of the “narrative record” up, so I took me a gander. It seems a large part of the effect of a law, according to this narrative record, is what the President and the Congress told us it would be around the time the law was passed or the policy was implemented. Well, I got news for the Romers – by that standard, the national debt has been plummeting since GW took office. And as I recall, the magic of the Laffer curve was supposed to make Reagan look fiscally responsible too.
But back to the great tax hike of 2005. 2005Q1 seems to be, according to Romer and Romer, an effect of the 2003 Jobs and Growth Tax Relief Reconciliation Act of 2003. And while they assign that act a big reduction in liabilities in 2003 (i.e., a tax cut), they assign it an increase in liabilities in 2005. (i.e., A tax hike. Read the thing yourself – its on page 88.) With all due respect, this, to me, is gobbledygook. One can assign all sorts of things to all sorts of things, but here’s the way I would approach 2005: there were no changes to the marginal income tax rates or the capital gains rates between 2003 and the present, so there were no marginal tax rate changes in 2005. I believe the estate tax fell a bit as a result of the earlier in the decade GW Hearts Paris Hilton and Himself Act of 2001… and will continue falling until it hits zero in 2010. But as far as I know, the only “major” new piece of legislation to hit in 2005 was the “American Jobs Creation Act of 2004.” That was a tax cut, pure and simple – the more closely the title of a law hews to “Mom and Apple Pie” the less taxes someone is gonna be paying and faster debt is gonna pile up on future generations’ tab.
Going back a bit further, the narrative points to a tax hike in 1982. Big sucker too! (Check out the graph again.) Once again, the Romer and Romer narrative doesn’t fit with how I remember things. I remember the bottom marginal income tax rate dropping from 14% to 12% from 81 to 82, and again to 11% in 83. As I remember things, the top marginal rate going from 69% in 81 to 50% in 82, and the top cap gains tax dropping from 35.6% to 27.3% to 27% from 81 to 82 to 83. In terms of GDP, tax collections also fell substantially, going from 19.6% in 1981 to 19.1% in 1982 to 17.5% in 1983 and even a smidge further down in 1984.
Now, Romer and Romer go on to tell us a lot about this tax increase (E.g., “The motivation for this tax increase was deficit reduction and fairness. The fiscal action was clearly not taken because policymakers felt the economy was overheating.”) but to me it reads like so much verisimilitude. Whether there were bones thrown to the concept of fiscal responsibility, and regardless of whatever our esteemed leaders were peddling when they sold this thing to the public, the simple fact of the matter remains: 1982 and 1983 were tax cut years. Romer and Romer can tell me taxes were hiked all they want, but I’m also free to go around saying the Loch Ness Monster was stealing lollipops from children in Central Park at the time. The fact is, nobody remembers either episode because neither one happened. (And to forestall one line of BS I expect to otherwise see in comments: reducing an anticipated tax cut of X% to Y%, where Y = X – epsilon, does not make for a tax hike.)
Anyway, I can go on, but I think its obvious that either I’ve completely misread these two papers, or there’s something seriously wrong with the data that went the work. I would also add that with the “right” data I can prove the earth is flat, not to mention that the Loch Ness Monster has a thing for lollipops. Furthermore, Christina Romer will be advising the next President (preview: tax cuts are the shiznit) so let’s hope we can chalk this up to a major misunderstanding on my part. But I’m kinda worried the problem isn’t that I’m displaying some latent feeble-mindedness – Greg Mankiw thinks Christina Romer is “an excellent choice.”