Tuesday, February 23, 2016

Krugman v. Mankiw on Rubio’s Tax Cut – Show Me the Model

Greg Mankiw of Team Republican tries to counter an attack from Paul Krugman on Rubio’s tax cut for the rich, which may come as breathing spell from the flap over that “analysis” by Gerald Friedman (my two cents on that flap in a bit). Greg says Rubio is proposing the David Bradford plan:
Good tax policy should be pro-growth, simple, and fair. An income tax, unlike a consumption tax, penalizes saving, which undermines economic growth and introduces complexity. An income tax is often thought to be fairer than a consumption tax, however, because it taxes saving, which is disproportionately done by higher-income individuals….The reduction in capital accumulation reduces labor productivity and lowers real wages throughout the economy, depressing the standard of living of future generations. Some studies have found that a switch to consumption taxation would increase the size of the U.S. economy by as much as 9 percent in the long run, although other studies estimate smaller gains.
Bradford does cite these studies and we will note one later. Can we take this back to the debate over the 1981 tax cut and note why I am so big on Show Me the Model? Voodoo economics was coined by George H. W. Bush as he did not believe Art Laffer had a real model of how large tax cuts could lead to a miracle growth. After all, standard economics tells us that tax cuts lead to less national savings, higher real interest rates, and crowding-out of investment. All of which happened. Laffer’s excuse for a model was a cocktail napkin which does not cut it. For what is worth, the Congressional Budget Office does model out potential output, which had been growing at a 3.5% clip since the end of World War II until 1980 but slowed to a 3% clip for the Reagan-Bush era. As Peter Dorman noted, Gerald Friedman failed to model out potential output and its projected growth path under the proposals of Bernie Sanders:
He never presented his model. The appendix to his report jumps immediately to parameter estimates, but there is no list of all parameters nor a formal model displaying how they relate to one another. I take it that the implicit model calculates GDP growth from spending projections subject to a multiplier, and that this translates into labor demand with productivity as a residual. The microeconomic results are determined by macro outcomes plus additional sector-specific factors. There does not appear to be a simultaneous relationship between macro and micro (especially labor market) outcomes, which is a cause for concern.
His defenders seem to miss the point with respect to the microeconomic or potential output side. Yves Smith wrote:
Friedman did this using a completely standard model. So the real issue is about the assumptions
A standard multiplier model without any consideration of potential output might be fine for forecasting a depressed economy over the next few years but not for a decade. In some ways, this defense from Jamie Galbraith was worse:
There are not many ambitious experiments in economic policy with which to compare it, so let's go back to the Reagan years. What was the actual average real growth rate in 1983, 1984, and 1985, following the enactment of the Reagan tax cuts in 1981? Just under 5.4 percent. That's a point of history, like it or not.
Of course the CBO model would note we had a GDP gap back then near 8% with potential growing at 3% per year. CBO now says the gap is around 2.8% and it is growing at only 2% per year. But is CBO the only model? Of course not. Menzie Chinn had an excellent discussion:
Note: I do not know what the output gap actually used in the Friedman study, as it is not reported…. One thing that should be remembered is that the trend line extrapolated from 1984-2007 implies that the output gap as of 2015Q4 is … -18%. A graphical comparison which highlights the implausibility of the -18% output gap is shown below… I want to stress that estimating potential GDP and the output gap is a difficult task
Indeed it is difficult but using trend lines is how Lawrence Kudlow does this for the National Review. Let’s not go there. Sandwichman wants to ditch NAIRU (I agree) and Mark Thoma has a must read discussion. All of this is fine but I still say Show Me the Model and as one does, please note the latest from Brad DeLong:
These are principal causes of "hysteresis". I do not believe that the output gap is the zero that the Federal Reserve currently thinks it is. But it is very unlikely to be anywhere near the 12% of GDP needed to support 4%/year real growth through demand along over the next two presidential terms. We could bend the potential growth curve upward slowly and gradually through policies that boosted investment and boosted the rate of innovation. But it would be very difficult indeed to make up all the potential output-growth ground that we have failed to gain during the past decade of the years that the locust hath eaten
Team Republican will likely have their models so permit to reach back to one of the models mentioned by Bradford in discussing something akin to the Rubio proposal which was an excellent paper entitled Simulating U.S. Tax Reform:
This paper uses a new large-scale dynamic simulation model to compare the equity, efficiency, and macroeconomic effects of five alternative to the current U.S. federal income tax. These reforms are a proportional income tax, a proportional consumption tax, a flat tax, a flat tax with transition relief, and a progressive variant of the flat tax called the 'X tax.' The model incorporates intragenerational heterogeneity and kinked budget constraints. It predicts major macroeconomic gains (including an 11 percent increase in long-run output) from replacing the federal tax system with a proportional consumption tax. Future middle- and upper-income classes gain from this policy, but initial older generations are hurt by the policy's implicit capital levy. Poor members of current and future generations also lose.
This estimated 11% gain happens only very slowly over time. Tax cuts do not pay for themselves and the proposition that everyone gains is not true. Maybe Greg Mankiw wants to pretend otherwise and maybe he has in mind a different model. But like Gerald Friedman – he has not laid out a real model.


Peter said...

Romer and Krugman should have e-mailed Friedman and told him "we are about to go public with our concerns over your paper. Here are our concerns. Please let us know if you want to alter your paper before we go public."

rosserjb@jmu.edu said...

Generally pretty good discussion here, pgl. I will note one minor caveat, where Laffer was right about the 1981 tax cuts. It looks like revenue from the top 2% of the income distribution did rise after the cuts, which reduced their top marginal rate from 70% to 50%. That is indeed in the range of where the Laffer curve's hump is thought to lie, so not totally unexpected.

For all its having been drawn on a napkin, the curve is not per se theoretically wrong. Indeed at 100% tax rates we expect revenues to get very low, if not necessarily zero (in parts of Russia in the late 90s rates actually exceeded 100% for some parties; they paid bribes to get out of paying some it, and some revenues were collected). Indeed, since I have brought up Russia, with its totally wacko tax system in the late 90s, one of the few reasonable things Putin did when he first came in during 2000 was to introduce a flat tax at 20%, which did indeed raise more revenues than the monstrosity it replaced, which had led to a widestpread barter economy.

However, today, there is no question that tax rates in the US are well below the Laffer hump, wherever that is exactly. Whether tax cuts lower or increase the growth rate (depends on a lot of things), lower tax rates will unequivocally lower tax revenues at this time anywhere in the US (for state disasters arising from listening to Laffer at this time, see Kansas, Wisconsin, and some pathetic others).

ProGrowthLiberal said...

Interesting couple of comments. On Russia, the transition from Yeltsin to Putin meant at least a few corporate governance rules which were absent in the 1990's. Plus the recovery of world oil prices helped.

I can recall a small model from James Tobin that put the top of Laffer's cocktail napkin curve at an 82% tax rate. Other models differ but not my much. Even Bernie is talking at most a 52% top marginal rate.