Conceding too much to Supply Siders
This is my usual post lamenting the fact that reasonable people allow supply side dynamic scorers to set the terms of debate. They confidently assert that tax cuts for the rich cause more rapid GDP growth and want to argue whether that means they pay for themselves . Jeb!’s economists admit that they don’t without 2 magic asterisks. But if one concedes that the effect on growth is positive and argues about the magnitude then one should concede that taxes on the rich should be cut — we don’t dislike deficits because they cause us pain by themselves but rather because they harm the economy.
There is no reason to believe that cutting the top marginal tax rate will cause higher GDP growth. The evidence points the other way.
The very excellent Kevin Drum makes my knee jerk with one sentence (with which I don’t even disagree) in this excellent post on the conservative Tax Policy Center’s analysis of Jeb!’s proposed tax cuts.
Nice catch and post. However I think you (impicitly) conceded too much when you wrote “The Tax Foundation has a very rosy view of dynamic effects, which are almost certainly far less than they estimate.” The reason is that “less” can mean smaller (closer to zero) or lower (closer to negative infinity). In principle, it is also possible to argue that Jeb’s tax cuts will cause 20% lower GDP. is this less than the effect the tax policy center asserts ? It’s larger in absolute value.
I haven’t run the numbers (I can’t) but I am confident that Jeb’s plan would cause lower growth. Partly this is because of post WWII data on top marginal tax rates (I admit on labor income) and growth in OECD countries. Atheoretic estimates, if taken literally, suggest that the growth is maximized at a top rate of over 50% (some estimates are 70%). In contrast, I know of essentially no evidence published in the peer reviewed literature that lower rates cause more rapid growth (the essentially is in regressions which consider convergence that is incude initial per capita GDP as a regressor — it’s negative coefficient is overwhelmingly statistically significant for the sample.
The Jeb team analysis (not tax policy center but Feldstein Hubbard et al) assumes additional growth from unspecified regulatory reform and also assumes unspecified spending cuts. They consider no effects of spending but costs — that is assume the spending is pure waste. In standard models, eliminating wasteful government spending causes increased growth, but the assertion that government spending is waste is completely unproven in the Jeb team analysis — it is just assumed.
This is important because if the tax cuts without the magic spending cuts cause higher deficits, the higher deficits will cause higher interest rates (The Fed is lifting off the zero lower bound soon or in a few months) which cause lower investment which causes lower growth.
Back to atheoretic data analysis — deficits are strongly negatively correlated with growth. Of course the case against supply side voodoo is very familiar to Americans who follow the news with the Clinton tax increase followed by a boom and the W Bush tax cuts followed by disaster and the 2013 tax increases followed by a (slight) increase in GDP growth.
My point is that international evidence and reasonable theory support the impression that the Tax Policy center didn’t just get the magnitude of the dynamic effect wrong but also are wrong about its sign.
I would be one hell of a poker player is I could carry a couple of asterisks around with me.
As opposed to reality, when it was made clear to me 40 years ago that I was one of the worst poker players in the history of the world.
In other words, if they really want to increase the rate of GPD growth, they should increase the top marginal tax rate.
Let’s call it what it really is. Lowering the tax rate is just a payoff to the big donors.
That “evidence” is speculative at best. The residuals in the curve fit are HUGE.
But the odds are…
… that there is no correlation whatsoever.
So, if there is no correlation, lowering the top marginal tax rate will not increase the rate of growth in the GDP. Thus you might as well raise it and use the additional money for things we need.
if we are at less than full employment, and the government spending goes to the poor and middle class who will spend most of it, then it is not wasteful even it it doesn’t lead directly to production, because it will increase aggregate demand
so they can focus on getting more production out of our government spending buck, but until we reach full employment, cutting that government spending that helps the poor and the middle class will hurt the economy
on the other hand if the government spending is going to rich, cutting that would do less harm (and then spentding it in a way that helps the poor and middle class, that could help he economy)
I’m not so sure anymore that outcomes matter at all to the American people.
The correlation is present and statistically significant. There have been many posts her and elsewhere over the last many years.
There are the always legitimate questions about correlation and causation, but there is certainly the opposite of an empirical case for lowered taxes leading to increased growth.
Critter: “[You] might as well raise [the top marginal rate] and use the additional money for things we need.”
That depends on how high you go with it.
From 1951 to 1963, the top personal income tax rate was 90% or more. The individual income tax netted, on average, 7.5% of GDP.
From 1964 to 1981, the top rate was 70%+. Those years, we netted 7.9% of GDP in personal income taxes.
From 1982 to 1986, the top rate was 50%. Those years, we netted 8.0% of GDP in personal income taxes.
1987 stands alone. The rate dropped to 38%, and income tax revenue went up to 8.2% of GDP.
Then from 1988 to 1992, the top rate was 30% or less, and we netted 7.8% of GDP.
1993-2002 saw the top rate at just under 40%, and we netted 8.3% of GDP.
From 2003 to 2012, we were again down to 33%, and we netted, on average, 7.1% of GDP.
Finally, since 2013, the top rate has been back around 40%, and we have netted 8.0% those two years.
My conclusion from this data is that our top marginal rate is right about where it should be for maximum revenue collection.
In short, Clinton and Obama raised the top rate to about the best place possible. That sweet spot seems to be around 40%, which is where we are now.
Djb: “[If] we are at less than full employment, and the government spending goes to the poor and middle class who will spend most of it, then it is not wasteful even it it doesn’t lead directly to production, because it will increase aggregate demand”
The rich also spend their money, just on different things.
However, the government should be spending more on infrastructure, which in turn would both increase employment of the manual laborers and facilitate growth. We get a triple-win there — better infrastructure, more employment, and less welfare.
Arne: “The correlation is present and statistically significant.”
With residuals like that? Uh, not so much. A high R-squared does not always mean a good fit.
warren, I thought you said there was no correlation. If that is the case, then there is no “right place”.
Also, if an increase in the top marginal income tax rate results in an increase in the rate of growth of GDP, then it is entirely possible for the percent of income tax revenue to decrease as a percent of GDP. After all, wages do not necessarily raise at the same time or rate as GDP.
Why concede anything to supply siders until they prove something?
Sorry for the confusion, Critter. The first was between the top personal income tax rate and growth of GDP. The second was between the top personal income tax rate and the percentage of GDP collected via personal income tax.
Good to see that Angry Bears haven’t forgotten Mike Kimel’s work here. (Newer AB readers might benefit by typing “Kimel” in the search bar for this website.) Kimel never fully developed theory to explain his findings, but he was on to something that was important to business tax advisers when top income tax rates were high. My take-away was: a 65-70 percent income tax rate–for income above about a million dollars today–could encourage business and overall economic growth (through multiple mechanisms), and top tax rates much below this retard business and overall economic growth. Something Kimel didn’t address, but perhaps worth considering, is whether low top rates also support bubble-creation.
“Back to atheoretic data analysis — deficits are strongly negatively correlated with growth. ”
Of course it is. That is the entire reason for progressive income taxation that has tax brackets continuously adjusted for inflation.
The idea is that when incomes are rising taxpayers pay in more taxes as they get bumped into higher brackets so taxes increase faster than incomes. And conversely when income growth is slow or downward then people are moved into lower tax brackets and taxes go down faster than income. This makes income tax extremely counter-cyclical. The negative correlation to growth is exactly what income tax structure designed to achieve.
The only problem is the correlation is not strong enough.
The payroll tax is strongly pro-cyclical. For top earners the
more you earn the smaller percentage you pay. And Capital gains tax is too flat. Had all taxes been more progressive (with progressive brackets adjusted with inflation) then the excesses of the last several decades would have been less and recovery on downturns would have been stronger. And the negative correlation of growth to deficits even more pronounced. Almost all of federal tax is based on income, but today only about 1/3 is progressively taxed. 40 years ago almost 2/3 of it was progressively taxed and the part that was progressive was more progressive than today.
Jim: “Almost all of federal tax is based on income, but today only about 1/3 is progressively taxed. 40 years ago almost 2/3 of it was progressively taxed….”
Uh, no. In 2014, 42.41% of revenue was taxed “progressively” — 34.48% of federal revenue was from individual income taxes, and 7.93% was from corporate income taxes (which are not only progressive, but are paid mainly by high-income earners who own the companies). In 1974, 46.58% was taxed “progressively” — 35.16% individual and 11.42% corporate — not even close to two-thirds.