Assorted links

1. The Laffer curve, by James Surowiecki

2. What have we learned from market design? — Alvin Roth tells us, ungated version here

3. Felix Salmon summarizes my talk, on the economics of blogging, an excellent post; further commentary here

4. How not to bore others in conversation, hat tip to Ben Casnocha

5. How to build your long-term memory


The article on long-term memory points out that it is a waste of time to study what you already know ("over-learning"), and you should instead periodically review what you've studied.

The simplest way to achieve this is spaced repetition, and there are any number of software packages which implement it, some of them free. Based on a quick numerical self-evaluation of how well you remember any particular item (say, clicking on a button from 0 to 5), the interval for when to reschedule it for the next review is adjusted to be shorter or longer. Over time, all items to be remembered get reviewed at individually appropriate intervals, not too soon and not too late. Simple but quite effective. Done properly, you never forget the items under study.

Surowiecki sounds pretty sure of himself. Can someone familiar with the relevant studies provide a ruling?

I didn't find any false or questionable claims in the Surowiecki piece. Personally I might have stressed more the work on incentive effects by Feldstein, Summers, and others but I think the piece is right on the mark.

One thing I have been wondering about laffer/supply side recently is the untold factor of time. Nobody talks about how much time we have to regain that money. For example, there are arguments against "Super-Kyoto" (the one that would be needed to actually work) saying that yes, we can pay x% of gdp now, or allow the economy to grow, such that the final payment would be less painful. (sorry I don't have a link right now)
You could say that the time periods needed are politically inexpedient and therefore, it is irrelevant from the perspective of politicians, but this is different from saying we are absolulely on one side of the curve.

While I agree with Surowiecki that the Laffer Curve, like most economic theories that happen to play nice politics (an alternative example would be externalities), the logic of the theory is used without evidence to serve political talking points. However recent history and politics should not be held against the theory itself. Historically, it is likely that we have resided on the backward-bending portion of the Laffer curve. Especially during the 40's and 50's as income tax rates reached the low 90's in percentage (See Page 10 of the pdf at ). If we've been there once, we could go there again and the Laffer Curve should stand as an argument against that trip.

I am a bit of a supply sider i.e. You can over-tax and over-regulate the economy. If people and companies are spending a great deal of resources on avoiding or meeting government intrusions how can that not reduce potential GNP.


Ok Idid a little calc in excel with very steady rates, no other changes.

I figured the initial drop in revenue was 5%. The economy/tax revenues grew at 3%, and the cost to finance govt debt was 5%. Under those assumptions the tax cuts would have to cause the economy to grow an extra 0.43% or more annually for me to say they increase the PV of govt revenue.

nameless with the spreadsheet, wouldn't the needed ADDED growth rate be 5.25% (e.g., total of an unlikely 8.25 growth)? After all, you now need 5% more money to get back to where you start.

Politicians running for office couldn't tell you where the point is that government revenue is at the apex. And which sounds more appealing to a voter?

A) Government revenue 10% lower due to tax levels above maximization, economy slows.
B) Govrenment revenue 10% lower due to tax levels below maximization, economy accelerates.

In the long run, however, you have to show why economic growth would be faster under higher taxes than lower taxes. In any given year, we can say that we are above/below the optimal tax rate, but does anyone seriously believe that if we reduced taxes from 35% of GDP to 15% of GDP for a decade, that government revenues wouldn't be much higher in 2018 when rates went back to 35%, then they would if we hold taxes level for the next 10 years?

Federal Receipts: $517.1 billion
Federal Outlays: $590.9 bilion

Federal Receipts: $909.3 billion
Federal Outlays: $1064.5 billion

CAGR Receipts: 7.31%
CAGR Outlays: 7.64%

Even if revenues are higher in the long-run, it's impossible to cut the deficit if you spend more than you collect.

Unfortunately, politicians of all stripes have gotten into the habit of using the term "growth" in an imprecise way. Liberal democrats, for example, will argue that an increase in spending on a certain program from $100 to $105 is a "spending cut" because 105 is less than what was proposed in some earlier budget.

Romer and Romer conclude that tax cuts spur economic growth, but not enough to completely eliminate the impact of the tax cut. Suppose you start with an economy of size $100 and the tax rate is .20 (so tax revenues are $20), and you reduce the tax rate to .19. You do not see (say Romers) a 5% reduction in tax revenues. The economy grows to 103 (responding to lower tax rate), and tax revenues only fall by about 2% to 19.57.

Therefore, Bush is absolutely wrong to say "our tax cuts have reduced the deficit" if he means "reduced the deficit to a level that is lower than it would have been without the tax cut". But Bush is absolutely right to say "our tax cuts have reduced the deficit" if he means "reduced the deficit below what was predicted by static budget analysis at the time of the tax cut". If you hate Bush, you will argue (with some merit) that this second interpretation is a slippery one; but then you have to be equally critical of those who use the term "budget cut" to refer to lower growth in expenditures.

My problem with the way the Laffer Curve is framed is that people on the right wind up as advocates for increasing government revenues.

People on the right side of the economic spectrum ought to be arguing for cuts in tax rates in order to cut tax revenues. Which brings up the 1000 pound gorilla that gets ignored during talk of the Laffer Curve: government spending. It needs to be cut so that personal spending can go up, skipping the bureaucrat middleman in the process which takes out an unhealthy chunk and forces us to consume things when if left to individuals they would have spent it quite differently.

Not to mention the deadweight loss caused by any form of taxation, especially higher taxation.

That said, the Laffer Curve is clearly not bunk at high rates. The US raised more tax revenue as a percent of GDP in 2000 under a 40% top tax rate than it did in 1980 under a 70% top tax rate. Clearly we were on the wrong side of the curve in 1980, something that those who clamor for more government spending ought to keep in mind.

In 1960 with the top marginal income tax rate over 90%, it is damn near obvious that cutting that rate would yield greater tax revenues. Who wouldn't get on board for a free lunch like that. (Funny though, it took 20+ years.)

In 1980 the top rate was something like 60% and cutting it could also have resulted in greater tax revenues, but it is not obvious that it would. Then again, if you cut all marginal rates, the loss of revenue from taxpayers faced with lower marginal rates might swamp increased revenue from those at the top rate. Yet even here the tax-cut advocate might claim his tax-cut would increase revenues if he thought the economy would grow for reasons unconnected to the tax cut. That's cheating, but perhaps effective politics.

But now with the high marginal rate at 35% it is hard to claim that tax cuts won't cut revenue. Tax-cut advocates should be arguing that we will be better off if tax cuts let individuals spend their own money than if the government gets to spend it. And I think that was tax-cutters have been saying for hundreds of years, but it was downplayed when the easier Laffer argument worked.

Surowiecki, Jonathan Chait, Paul Krugman are all making hay by claiming dishonest and flakey supply-siders are the intellectual backbone of tax cut advocates. Is that empirically true? Are there no better advocates on our side? I don't think so, but to take the offensive we need to distance ourselves from the Laffer argument. It was valid and important up thru the early 1980's perhaps, but we need to get back to more traditional arguments now.


1) the government does not borrow to cut taxes
2) the government does not attempt to capture the new revenue immediately. i.e., a cut of x dollars persists for T years.
3a) assume that the tax-cut is targeted at the top bracket, which is to say that most of the 'returned' taxes feed investment activity not consumption
3b) government spending is predominately consumption not investment
4) private sector investment yields 10%
5) the marginal tax-rate is 30%

Then tax-cuts do pay for themselves in ~30 years time.

x*T < 0.3 \sum_{t=1}^{T} (exp(0.1*t)-1) This is the power of compound interest. BTW, this has nothing to do with the Laffer curve which is concerned with direct (nonpayment) and indirect (inactivity) tax evasion. The Laffer curve tells us that the loss (LHS) is over-estimated. Tuning these parameters in various ways will create a net surplus from tax-cuts within five years time.

Comments for this post are closed