This is going to be a very long post with independent chapters (I won’t impose by making a series of posts)
I. What is “Ricardian Equivalence”
The basic idea is that the timing of taxes has no effect on anything, and especially not on consumption/saving choices, because rational economic agents know that the state has a budget constraint which is binding and anticipate their share of paying for the national debt.
This idea is clearly crazy. Even Ricardo (who generally lived in a world of theory detached from reality) wrote that it clearly had nothing to do with reality. Sadly, modern macroeconomists are more detached from reality even than Ricardo, and it is a standard feature of standard models.
Another way of putting it is that domestic government bonds owned by domestic agents are not part of national wealth — they are money we owe each other not an true asset like physical capital. Now it is clearly true that government bonds are not net wealth. The idea that they don’t influence consumption as net wealth does is clearly crazy. There is no evidence in US data for Ricardian equivalence. The most basic implication is that, if one looks at consumption as a function of disposable personal income (personal income net of taxes) then one should be surprised that it is so low when budget deficits are high. The equivalence claim is that the variable should be disposable income minus the deficit that is personal income minus government spending. But at the very least, deficits should crowd out some private consumption. There is no evidence of any such effect in US data (long pdf warning)
I shoud just quote Paul Krugman who writes better than I do
Ricardian equivalence says that what determines consumption is the lifetime present value of after-tax income, and hence that, say, a temporary tax cut won’t stimulate spending, because people will figure that whatever they gain now will be offset by higher taxes later. It is a dubious doctrine even done right; many people are liquidity constrained, and very few people have the knowledge or inclination to estimate the impact of current government budgets on their lifetime tax liability.
II Future tax increases will and won’t follow temporary tax cuts.
The point of this post (if any) is that advocates of tax cuts switch Ricardo on and off when he is or is not convenient.
I am going to assume no Keynesian stimulus effects. It is as if I assumed that there is no unemployment or spare capacity so GDP is determined by labor supply and technology. What I actually assume is that the Fed will set interest rates to make unemployment equal to their guess of the natural rate of unemployment. This means that vulgar Keynesianism (TM) is an error. As always I cite Krugman (who is right about vulgar Keynesians even if he regrets one word in “Greenspan … is not quite God”). I am also going to assume a closed economy. This is really silly, but it’s what I am going to do. The sensible reader is invited to stop reading.
Consider a proposal to stimulate investment by cutting the tax on profits.
An argument against cutting the tax on profits is that the lead to deficits and the national debt creates the illusion of wealth (as people act as if government bonds are net wealth). This causes higher than sensible consumption and forces the monetary authority to set high interest rates to prevent over heating and inflation and in the end this crowds out investment. The counterargument is that people know that taxes will be increased in the future (and not just back to the old level but higher in order to pay interest on the debt) so there won’t be distortion of consumption savings decisions. On the other hand, the good incentive effect of lower tax rates, which cause higher investment, aren’t affected because higher taxes in the future cause no bad incentive effects because taxes are terrible now but won’t matter then.
This makes no sense at all. Future tax increases are assumed when one discusses consumption, but the incentive effects of future taxes are ignored when discussing the desired level of investment for a given interest rate. I think there is no way to make a coherent argument. The actual belief of advocates of cutting taxes on profits is that taxes and government spending should be cut. But they don’t propose that, because government spending is popular. They then argue that, even without spending cuts, taxes should be cut. This makes no sense. Sometimes they argue that debt is good because it will force lower government spending & not tax increases. There is no support for this view — the evidence such as it is, is that when Congress discovered that they USA could run huge deficits, they increased spending. People claim to dislike deficits, but they really dislike taxes. To starve the beast, one would have to insist on a balanced budget (which is also advocated by the same Republicans who are eagerly adding what they claim will only be $1.5 Trillion to the national debt).
[rant bumped down after the jump]
III Ricardo Vs Laffer
The more extreme advocates of tax cuts argue that they will pay for themselves; That the effect on growth and tax receipts is large enough that no future spending cuts or tax increases will be needed to pay for increased debt. This claim, made by Arthur Laffer when he drew a graph on a napkin for Jack Kemp, isn’t taken seriously by any wonks. It is always made using another equivocation. Supply siders argue that tax cuts will be followed by growth which will cause increased revenues. This is obviously true, so will tax increases, the economy tends to grow. The trick is to equivocate post hoc and propter hoc — between after the tax cut, GDP will grow so revenues will be higher years after the tax cut than they were immediately after the tax cut and because of the tax cut GDP will grow more than it would have without the tax cut so revenues will be higher than they would have been. This ultra clumsy rhetorical trick seems to have worked for decades, but also seems not to be working anymore.
My point (if any) is that you can’t have both Laffer voodoo and Ricardian equivalence. If the tax cut now doesn’t imply any tax increases or spending cuts in the future, it will cause higher consumption. Laffer’s claim is that tax cuts will make us much richer. Those who claim to believe in Ricardian equivalence must argue that this will cause us to consume more (other things equal). It doesn’t really matter if the effect on consumption is due to bonds mistaken for net wealth or for authentic wealth. This means that Laffer must admit that growth enhancing tax cuts will cause consumption to increase if other things are equal. Laffer may have made even more extreme claims, but supply siders now argue that the benefits of their tax cuts aren’t a jump in GDP but a higher growth rate. This means that higher consumption implies lower national saving.
Demand will equal supply (recall I assumed above that it is fixed in the short run) because interest rates will increase causing lower investment. Any increase in investment could only occur if higher interest cause lower consumption. There is absolutely no evidence that they do (and even the simplest theory doesn’t unambiguously imply that they should — in elementary models higher interest rates could cause eithe higher or lower consumption depending on paramters with parameters fitting the data definitely implying that high intereste rates cause high consumption).
To avoid predicting increased consumption (which would crowd out investment in a closed economy) advocates of tax cuts have to argue both that they will and won’t make us richer.
OK so before posting, I have to note that advocates of tax cuts have basically conceded most of what I argue here. They, and in particular the Tax Foundation, now argue that cutting the tax on profits will make us richer by attracting foreign capital. They don’t seem to have considered the fact that the foreigners will collect interest on money they send to the USA. Nor that foreigners already own shares of US firms and so will get some of the direct automatic benefits of the tax cuts.
As usual, Krugman has explained this. Click here, here and here.
IV Ricardian equivalence and the assumption that 1=0
Having more detachment for reality than Ricardo isn’t even half of the problem with many leading economists (notably the two examples I have in mind work at the University of Chicago). Many economists rely on the assumption that there is Ricardian equivalence and don’t know what the phrase means. One is Nobel memorial prize winner Robert Lucas. Krugman again (same link — just click and read — Lucas’s error is actually shocking)the math error of assuming 1=0. They argue that the effect of government spending on GDP is zero (so the multiplier is zero). They then argue that increasing government spending is a bad idea even when the economy is depressed and the safe short term interest rate is zero. Then they look at data which is consistent with a multiplier of roughly 1 and say it confirms their claim. Now Ricardian equivalance implies that the effect of a very brief increase in G (govenment consumption plus investment) on GDP is one for one. The assumpting that 1=0 causes lunatics to think this means Keynesian stimulus doesn’t work at all.
I think part of the utter insanity is based on a deeply held assumption that government investment is pure waste. If G contributes nothing to our happiness or productivity, then the correct measure of national income is NNP-G (gross national income minus depreciation of capital minus govenment consumption minus government investment).
Sometimes, Keynesia stimulus is redefined to mean spending which serves no purpose except to stimulate demand. This is based on taking jokes told by Keynes (or Krugman) out of context. It has nothing to do with Keynes’s proposals or, say, the Keynesian stimulus bill signed by Obama.
This chapter is of little value. The lessons are that you should ignore everything Robert Lucas and John Cochrane write and say. I’m sure you already do, so I put it down here after the jump.
V Lump sum taxes
[this is the rant I moved down from above the jump]
The insane argument is reconciled with economic theory with budget constraints and first order conditions and such like by assuming that the future taxes will be lump sum taxes. Those are taxes which you have to pay no matter what you do. They are not necessarily (in theory) the same for everyone — this would imply extracting money from people who have none, bleading stones and other miracles. They are necessarily (in the USA) the same for everyone. A specific lump sum tax with a name attached is a bill of attainder (unconstitutional) a taking (unconstitutional) and a violation of equality under the law (unconstitutional). The idea has no relevance to the US policy debate. It also has a central role in economic theory. An even older and more elementary fact is that people tend to revolt when the government tries to collect lump sum taxes. A poll tax proposal ended Margaret Thatcher’s career. Centuries earlier John of Gaunt attempted to impose a poll tax. Roughly 100,000 Englishmen rose up and, among other things, burned his house. Maybe it could be done, but the historical record sure suggests otherwise.
Even more, the enthusiasts for reductions of taxes on capital all seem terrified that the majority of their countrymen will notice that they have below average wealth. They fear a tipping point when the “takers” meaning the non rich, will take their wealth. They can read polls. So why do they want to inflame populist anger and provoke class conflict by grabbing even more ? It makes no sense. I honestly don’t understand their thinking even assuming they are infinitely greedy and selfish.
Yeah, but you can have neither, and it turns that’s the way the real world rolls. I have yet to see anyone actually fit tax data to anything remotely resembling Laffer’s napkin without some truly heroic fudging, nor have I seen anything believable that resembles support for Ricardian equivalence.
FWIW, there is absolutely no reason why Ricardian equivalence wouldn’t be true in a world where everyone is perfectly rational and has complete knowledge of the future. I’m not completely sure that’s enough to get a real live Laffer curve though.
Absolutely. I am very sure that there is not Ricardian equivalance and that no country is near the top of the Laffer curve.
I think you just described religious belief systems again. … you know the ones based on belief in a fictional non-human entity.
(I actually find it incredibly stupid (I mean either really, really stupid, or else just pure con artists) for economist to conjure up these fictional entities to create foundations to support their religious beliefs. What surprises me most is that seemingly academic institutions give them a pedestal and credential — much like perhaps the church does for it’s priests).
A very clear case of the empirical irrelevance of Ricardian equivalence dates from the first Reagan tax cut, a period when Ricardian equivalence was gaining influence as an idea in the economics profession. After the tax cut, the savings rate fell when it should have risen. End of story.
There are probably a handful of countries near the top of their Laffer curves, although probably not any seriously over the hump. One case where a nation may have been over and did increase revenues by both cutting and simplifying taxes was Russia when Putin first came to power. There were locations in Russia where the official tax rate for some people was over 100%. Really. They had to pay bribes to avoid paying taxes. Anyway, in one of the few good things Putin did soon after he took office initially was to impose an across the board 20% flat tax rate. Revenues did indeed increase.
When I say I have never seen the Laffer credibly fit to real world data, I mean just that. Perhaps a decade ago I had a series of posts where I did fit US tax rates (and sometimes tax burdens) v. tax revenues. I could never replicate the shape of quadratic shape of the Laffer curve with something resembling zero at both ends and an optimum in the middle. Depending on assumptions, in fact, what came our was the exact inverse of the napkin drawing.
Granted, I no longer pay as much attention to the literature, but while people talk about the Laffer curve all the time, I have yet to see anyone fit data to a cure and get something that looks at all like the shape on Laffer’s napkin. And unless you can get a curve that rises to maximum and then dips, the whole concept of being near, past, or not close to the optimal point on the Laffer curve is completely meaningless. Saying things like being near the top of the Laffer curve is like a physicist describing the effects of the aether or a chemist suggesting a bit more phlogiston will help generate more useful flame.
It doesn’t mean that there isn’t a relationship between tax rates and tax revenues. It doesn’t mean using the Laffer curve as terminology is at best useless, and at worst misleading.
A quadratic in tax rates and something good ? How about
Note this is *not* the Laffer curve. The peak of the Laffer curve would be at a higher rate. The data suggest that that the growth maximizing top rate is somewhere from 50% to 70%