From Mark Thoma, via Brad DeLong, comes what will henceforth be my absolutely favorite example of twisting data to fit your theories. Observe the following graph of corporate tax rates vs. revenue in units of GDP:
Pretty straightforward, really. As you raise taxes, the government collects more revenue. Norway seems to collect more than its fair share, which might be interesting to dig into, but the trend seems clear. But there’s something nagging at the back of your mind — aren’t there people out there in the world who believe that raising taxes actually decreases revenue past some certain not-very-high tax rate? “Supply-side economists,” or something like that? People who exert a wildly disproportionate influence on U.S. tax policy? What would they make of such a graph?
Yes, Virginia, there is such a thing as supply-side economics, and you can find its practitioners in such out-of-the way places as the American Enterprise Institute and the editorial pages of the Wall Street Journal. Here is how such people view these data:
No, I am not being unfair. I did not draw the “Laffer Curve” on top of those data in order to embarrass the WSJ or AEI. They did it themselves; the second graph is how the plot was actually published by the Journal, while the first one was Mark Thoma’s subsequent reality-based-community version of the plot. As Kevin Drum says, it’s “like those people who find an outline of the Virgin Mary in a potato chip.”
Among other features, we note with amusement that the plotted curve implies that tax revenues hit zero at a corporate tax rate of about 33%, and become dramatically negative thereafter. As of this writing, it is unclear what advanced statistical software package was used to fit the Laffer Curve to the data; the smart money seems to be on MS Paint.
The best part is that the laffer curve doesn’t even apply to the income as percent of GDP graph. I’d be curious to see the microeconomic reasoning behind such a graph, if there is any. I guess the difference between econ and physics is that this is that this is how econ treats cranks.
Back when I took Economic (circa 1987) my economics 1 professor explained the Laffer Curve like this. Some assistant professor somewhere was having a few martinis for lunch one day. He had this idea that as the tax rate goes up people will more of an incentive to cheat so the effective tax rate (the rate that is actually collected) might climb slower. He then got the idea that perhaps if the tax rate got high enough not only would the effective tax rate decline so would the over all revenue generated. So he drew the curve on his napkin. After a few more martinis and decide that we were on the downward side of the curve.
IIRC, Norway’s corporate tax revenues come from excise taxes on the giant pool of North Sea oil. If you take out Norway’s oil boom-related revenues, it comes right down into the middle of the scatter plot.
Ironically, the supposedly optimum value at Norway’s point looks like about 28%, which I think is what Laffer originally thought was optimum. The Laffer curve has to be right ultimately of course, since at very high levels tax rates would have a destructive effect. That of course does not excuse this exercise in deceptive graph manipulation (go to RealClimate to see about similar travesties by global warming skeptics, and of course alternate history creationists do this too.) It is so typical of that right-wing crowd. If you are an honest conservative type, please work for your tradition’s reform or get into a third party, don’t defend hacks like the WSJ propagandists. (Well, what do you expect of people who complain about those who get something for nothing, but whose favorite way of making money is to buy and resell financial instruments – and then they want to pay a lower tax rate on that than those who actually produced new utility value.)
That’s fantastic.
I’d love to see the chi^2 on that…
Pretty straightforward, really. As you raise taxes, the government collects more revenue.
I have an office building. I raise the rents, I get more revenue. Pretty straightforward, really.
Lies, damn lies, and statistics?
Ok that’s pretty silly. For that matter picking one particular type of tax and graphing it against all revenue seems whacked. Not the least because those countries have varying amounts of local corporate taxes and other types of taxes such as VAT.
However AgnosticOracle seems to have taken away precisely the wrong part of his economics lessons. The key observation was that if taxes are 0% then the government gets no revenue, and if taxes are 100% then government gets no revenue (at least barring men with guns coming to force you to go to work). Yet clearly revenue is >0 for tax rates in between 0% and 100%. So there *must* be a maximum somewhere between there.
Seldom does one get to enjoy the spectacle of raw data fitted so precisely to theoretical curves. Spooky! And the results explain so much! (They explain that these guys are hilariously blind to the deficiencies of their ideological constructs.)
Graphs are such useful tools, but you’d think that they would try harder to achieve some level of plausibility. I monkeyed around with deceptive graphing and came up with a graph that plots oscillation as a perfectly straight line. [Link]
I’d have to second that, collin. But there is no doubt whatsoever that the chi^2 on the second graph is vastly, vastly higher than that in the first. You don’t draw a graph that goes above nearly all of the data points, and vastly below the rest and end up with a decent chi^2.
The curve you criticize is a deliberate and cynical manipulation of the data, so you respond with your own deliberate and cynical manipulation of the data.
Regardless of where the points actually lay, if you insist on drawing a line, and you require that the line goes through 0,0 – then it is inevitable that the line will indicate a positive correlation between corporate tax rate and government revenue.
The curve fitting you criticize is indeed flawed and manipulative, but so is your line-fitting, because by insisting on a line, you make a positive correlation inevitable regardless of the actual data.
I think it is pretty obvious that government revenue will decrease above a certain tax rate, because after a point taxation will remove the incentives for doing business, and businesses will therefore move. The interesting question is where this point is, and neither the curve you criticize, nor your own line-fitting, answer this question.
“I think it is pretty obvious that government revenue will decrease above a certain tax rate, because after a point taxation will remove the incentives for doing business, and businesses will therefore move.”
No, no, you don’t understand where Sean C. is coming from. In the left-wing fantasyland, the purpose of business is not to make money, but to MERCILESSLY EXPLOIT THE DOWNTRODDEN WORKING CLASS, AND ALL MINORITIES. So business will go on, no matter how high taxes are, as long as there is some poor person to be exploited.
The hilarious thing here, of course, is to object to the WSJ follies on “scientific” grounds, when [for example] welfarism is an experiment that has been tried thousands of times with results varying from pathetic to disastrous. I thought scientific people were supposed to discard theories when they fail experimentally….
I agree with Brad. Norway is an outlier so should be taken out. Probably the curve changes a LOT after that country is dropped.
Also, the regression should include quadratic terms, and as Ian said, should not be restricted to go through the intercept.
Here’s a different criticism that I think might call the meaning of the graph into greater doubt that we had thought.
Simplistically speaking, the tax revenue per GDP should equal the corporate profit per GDP times the corporate tax rate. Governments tax corporate profits, not corporate income. The Wall Street Journal article then makes a big deal about interpreting the chart in terms of corporate flight, and a lot of the discussion and comments about it invoke various incentives the tax rate may have to increase or decrease tax revenues.
The big problem I see is that the interpretation of the curve that we’ve seen so far relies on a hidden assumption: that all of the economies represented by data points on the chart are at the same average level of market development. In the absence of barriers to competition, a young industry with much profit should see that profit decline down to zero as new corporations enter into that industry with time. Or for another example, a nation that just abolished price controls for a consumer good might see a temporary spike in that industry’s profits before new suppliers can begin increasing the supply again.
So we might want to decide which of these nations are in the throes of a deregulatory binge before we draw any deep conclusions.
As a side note:
We have 7 years of data:
http://www.oecd.org/dataoecd/26/56/33717459.xls
So including this extra information should give a better approximation.
As a physicist I can confidently say the problem with economics is you have plenty of data, you just don’t really have a science that can do anything sensible with it.
The USA Is Becoming a Police State
http://home.comcast.net/~plutarch/PoliceState.html
Hi Sean,
This is great 🙂 thanks for once again confirming my prejudices about economists… (sorry! really trying hard to get rid of them, but mostly unsuccessful)
I just can’t understand complaints about tax rates based on nothing but the percentage, neglecting what one gets for it. The question is not how high the total tax rate is but what the money is used for, and whether this is what the public has democratically decided. If the people in country X want more (public transport? social security? health insurance? unemployment insurance?) to be shared responsibility, the thing to do so is to use taxes. That’s got nothing to do with ‘fairness’ – what the common sensus regards as fair and what not is exactly what the political system should find out, this might very well be different in Norway than in the US.
The problem with the present tax system is that the actual relation between taxes and what they are used for is too weak. In many cases some taxes are raised to use the money elsewhere. Needless to say, this doesn’t make any sense and results in people not willing to pay taxes because they don’t see what it’s good for. Bottomline is, most people come to the conclusion the lower the tax rate the better. As simple as wrong.
Best,
B.
B,
(sorry! really trying hard to get rid of them, but mostly unsuccessful)
I will gladly join you in your efforts to get rid of economists! (If there are none, you can’t have prejudices about a non-existent thing, can you?)
Best,
-Arun
Did anyone bother to look up the fact that this line is not being fit to the data?
It is merely a curve which indicates ‘good’ ratios are below the curve and ‘bad’ ratios are outiside the curve.
Richard Feynman said that if had to life live over again, he would have become an Economist. I think perhaps he would reconsider after seeing the way that some physicists treat economics with immature academic snobbery.
Of course Sean is right and people try and fiddle the data for political reasons, but those of you who just want to piss on economics because it makes you feel better… get a life! Economics is not physical science, people are not atoms, and any modelling that you do in a social science will neccesarily be very approximate because the complexity of real human interactions precludes an exact formal understanding. The question is, have the models of economics helped understand important phenomena, and the answer is undoubtedly yes. You cannot avoid problems just because they are difficult, so give the social scientists a break, they are playing by a different set of rules.
Take out Norway and UAE, both of which are highly anomalous, and the left hand straight line is scarcely more supportable than the idiocy on the right. That in itself is significant; it says corporate taxes appear not to be strongly corrleated with the ability of the government to raise money.
Wow! The premise that raising taxes forever leads to forever increasing federal revenues is ridiculous and absurd. These things have time dependance and are not independant of the state of the economy.
The Laffer curve is the perfectly obvious (and 100% accepted) observation that zero percent tax has no federal revenue, and 100% taxrate implies zero incentive to work –> another minimum. Therefore in any economy there is at time t, a global maximum.
Whats not accepted is exactly where that maximum is, and presumably it also could change with time.
Besides, things are much more complicated than that. During the Reagan administration, taxes were cut and federal revenues hit an all time high. Conversely, Clinton raised taxes, and revenues also hit an all time high.
My money is on photoshop. That looks suspiciously like a Bezier curve, which you can’t draw in paint but can draw quite easily with photoshop. Anyway, it seems like he may have assumed a parametrization of the laffer curve based on the mode and then selected the mode using order statistics (ie max). This isn’t completely arbitrary but its still extremely dishonest.