by Mike Kimel
Cross posted at the Presimetrics blog.
In Response to Bryan Caplan et. al., Part 1 – An Empirical Look at the US Economy From the Perspective of Economic Schools of Thought
One charge that often comes up from libertarian and conservative economists is that there is no data to support the positions typically adopted by those can be said to subscribe, at least in broad strokes, to the general outline of the economy as described by Keynes. As an example, Bryan Caplan wrote this
Keynesians have been a smug bunch from their earliest days. Here’s how Keynes once replied to Hayek:
Thus those who are sufficiently steeped in the old point of view simply cannot bring themselves to believe that I am asking them to step into a new pair of trousers, and will insist on regarding it as nothing but an embroidered version of the old pair which they have been wearing for years.
But what is the source of Keynesians’ self-confidence? Every Keynesian I’ve ever known has a stock answer:
The empirics are on our side.But when probed, they rarely deliver any details about these empirics. It’s been three decades since Keynesians could merely point at the Phillips Curve and say, “See? See?!” And in terms of merely “fitting the data points,” Prescott infamously showed that a simple RBC model works rather well.
At risk of getting sidetracked, a couple notes about the Phillips Curve and the RBC model. The Phillips Curve, purportedly showing a trade-off between the inflation rate (of wages, at least) and unemployment was an erroneous idea glommed onto by many Keynesians because it provided an easy-to-explain way for government actions to improve the economy (i.e., when unemployment rose too much, the government could spend some money to push it down, but not too much less it spur inflation).
While its been a long time since I read Keynes, I don’t think he mentioned anything of this sort, being familiar with the German post-World War I experience and all that. To say that the Phillips curve was is integral to Keynesian views because some prominent Keynesians used it as an excuse is to say that invading Iraq and cutting taxes are a part of the Southern Baptist theology. As to the RBC model working rather well – it may be adequate for spotting things in the rearview mirror as they go by, but if it isn’t too deus ex machina to actually serve some purpose (say, anticipating events), someone please tell me what that might be.
Given so many people, many of them thoroughly incompetent, built (to use the term loosely, in many instances) on the work that Keynes did, criticizing Keynes for the opinions of some of his lesser disciples makes as much sense as criticizing Hayek because of something that Glenn Beck said in last night’s broadcast.
The basic Keynesian insight, and about the only thing that all the various flavors of Keynesian “thought” agree on, is that when the economy is in a recession, government spending can be tremendously helpful. (By contrast, most libertarians and conservative economists are against most government spending, even during periods of economic weakness.) Now, if you look at every recession since 1929, the first year for which there’s data from the National Income and Product Accounts tables, there aren’t many instances where the government cut spending during or just after a recession, but those instances have led to shorter, slower expansions, which can be summarized in this graph:
Notice that when it comes to overall growth during an economic expansion (i.e., length of expansion * speed of expansion), cutting spending produced worse outcomes than small increases in government spending, and large increases in government spending produced better results still. (More info and a link to data sources here.) That fits very well with Keynesian theory of all stripes. On the other hand, it doesn’t fit with libertarian, Austrian, or conservative economic theory.
Another thing about Figure 1… had you shown that graph to Keynes in the early 30s, he’d have had no trouble believing it. Had you showed it to Hayek, be it in the Road to Serfdom era, or decades later, he wouldn’t have. My bet is that most folks on Caplan’s side of the fence don’t that’s what happened, and have been happily theorizing based on a reality a 180 degrees different from what Figure 1 shows for a long, long time.
Now, let’s flip things a bit, and look at the most basic tenet of libertarian/conservative economic thought, namely that because of the profit motive, the private sector is relatively efficient compared to the public sector. Note that a number of people who call themselves Keynesians believe this as well, at least under certain circumstances that have been observed in the not unrecent past. An implication of this belief is that, all else being equal, jurisdictions that tax less will produce faster economic growth over a relatively long period of time.
I tested that not long ago, using data on the growth of real per capita income and the tax burden (from libertarian favorite, the Tax Foundation) for the fifty states plus the District of Columbia. At times, some states are raising tax burdens, and others are lowering them – conservatives/libertarians would expect the former states to grow slower than the latter. At other times, almost all states are lowering tax burdens, and in this case, conservatives/libertarians would expect the states that lowered tax burdens the most to produce the fastest growth rates. Finally, in those periods where most states are raising the tax burdens on their citizens, conservatives/libertarians would expect the most from those states that raised their tax burdens by the least.
Now, for every two year increment between 1977 and 2002, states were broken up into three equal sized-groups: relatively high tax burden states, relatively low tax burden states, and medium burden states. Growth rates were measured for each state for the subsequent six years, thus creating a series of rolling 8 year windows made up of a tax decision in years 1 and 2 and a growth rate in years 3 through 8. (The eight year window was selected because most successful governors seem to serve two four year terms.) Here is what the data shows:
As I noted at the time:
The table is interpreted as follows: the smallest tax cutters (i.e., those who cut taxes the least or raised them the most in any given year) produced the fastest economic growth about 42% of the time, second about 29% of the time, and third 29% of the time. By contrast, the middle group came in first in one third of all occasions. The “tax cuttingest” group came in first place a mere quarter of the time; in half of all years, it came in last place.
Many conclusions are reasonable from this. However, concluding that states that cut taxes have produced the fastest growth rate in per capita income is not one of them.
(The full post, including links to the data sources here.)
We find something similar at the Federal level. The following graph shows the growth rate in real GDP per capita for each President from 1929 (again, the first year for which data is available from the NIPA tables) to the present, color coded according to whether they increased or decreased tax burdens.
Clearly this is not the truth that you’ll hear from many on the conservative/libertarian side of the aisle. Note that FDR is only shown until 1938 to sidestep the usual arguments from folks on the right who seem to believe that he did nothing right until World War 2; 1938 is sufficiently ahead of WW2 (i.e., even before lend-lease) that his performance cannot be influenced by WW2.
Now, I’m just warming up, but this post is getting long, so I’ll save my best material for Part 2. But this is enough to say, for now, that the basic assumption that underlies Keynesian thought is supported by the available data. The basic assumption that underlies both conservative and libertarian economics is completely contradicted by simple graphs of the data.