Scott Sumner fails to read third sentence of his own blog post

I just don't get this.

The third sentence (or actually part of the second sentence) of Scott Sumner's post is:

Keynesian: Fiscal austerity is contractionary at the zero bound regardless of whether you have an independent central bank.

Note: "at the zero bound" and "regardless of  ... independent central bank". Sumner recognizes the Keynesian model ...

Then he says that this graph shows fiscal contraction has no effect in countries (some at the ZLB, some not ... who cares!) with exclusively independent central banks (forget the regardless above, that didn't matter) after throwing out the countries without independent central banks:

What???!!!

Basically Scott Sumner undid his sentence:

Keynesian: Fiscal austerity is contractionary at the zero bound regardless of whether you have an independent central bank.

You could even take out the "austerity" in that sentence as we see in the first point below. This is not what Keynesian's say at all. This is the same garbage analysis I've seen before from Sadowski and Sumner:

1. Iceland is a garbage data point. It was never at the ZLB (at 6% back in 2014, today it's at 5.75%) and the fiscal consolidation treated as austerity is simply disingenuous (it happens after the finance minister declared victory over the recession!). It's just intellectual malpractice of the worst kind. You can go into the gory details here.

2. Why eliminate countries without independent monetary policy? Austerity at the ZLB doesn't care if you control your monetary policy or not. That assumes the market monetarist model in order to prove the market monetarist model. Those points should not be removed. Note Scott's sentence! It is contractionary at the ZLB regardless of whether you have an independent central bank!

As I say here:

Sumner does cite approvingly of a purported "takedown" of Krugman's austerity graph, but that "takedown" assumes the market monetarist model in order to throw out data (basically, all the liquidity trap countries) that make up the bulk of the correlation.

Let's just throw out all the bulk of the liquidity trap countries engaging in austerity! Lo and behold, what's left over says austerity isn't contractionary.

3. You can't include countries that aren't at the ZLB in order to say something about austerity at the ZLB. South Korea (2% in 2014, 1.5% today), Australia (it was at 2.5% in 2014, 2% today), and New Zealand (3.5% back in 2014, 3.25% today) weren't at the ZLB in 2014. So not only have we thrown out all the liquidity trap countries, but we get to add in a bunch that aren't in a liquidity trap! Sumner does recognize the zero lower bound -- remember his sentence quoted at the top of this post. But somehow it only selectively applies to data that he wants to include.

Overall, this is Sumner's blog post where we replace "austerity isn't contractionary at the ZLB" with "all swans are black"

All swans are black. So let's throw all the white swans out of the data set and put in a bunch of black ducks. Look: all birds are black!

Really???!!!

I used to think Sumner was the best advocate of the monetarist position. I mean he even pulled Matthew Yglesias over to the dark side. But this is really disappointing. You can't assume your model in order to include or leave out data points that favor or disfavor your model. You just can't do that. It's wrong.

Does this really pass for economic research? It makes me sad.

No wonder there hasn't been any uptake of the information transfer model. Economists don't know what good research looks like.

3/4 of a knife, 3/4 of a fork and 3/4 of a spoon

Cesar Hidalgo is looking into information theory with his new book. I haven't read it; however I have read Diane Coyle's review (H/T Mike Norman) and it seems like it might be a great source of analogies for this blog. This one immediately struck me:

Hidalgo makes the same point as the final chapter of [Diane Coyle's] GDP book, that in adding things up in terms of their monetary value we are not capturing the value of diversity: three spoons are not as valuable as a knife, fork and spoon.

That is actually the exact point of this post here. We'll almost exactly. If we have M dollars to spend on knives, forks and spoons, (assuming they equally cost one Euro [1] WOLOG) then all consumption possibilities (blue points) are located under the budget constraint hyperplane in the first (left side) figure:

The maximum entropy point is given by the black dot in the second (right side) figure. Now for three items, that point is actually at {M/4, M/4, M/4} so if you had 4 €, you'd have 1 fork, 1 knife and 1 spoon at the maximum entropy point (and 1 € left over).

The entropy maximum is related to equilibrium in the information transfer model  as well as effectively ideal markets. Under certain (ideal) conditions, it recovers all of the properties of maximizing utility.

I would disagree with the analogy of crystallizing imagination Hidalgo uses -- that is a lower entropy state for one thing. The key idea we want to capture is Jaynes' dither. We want to make the world safe for people to move about the space of possibilities -- to try and fail.

PS This also maximizes entropy.

Update +15 minutes:

Since the maximum entropy state (3/4, 3/4, 3/4) with budget constraint M = 3 is not realizable given quantized knives, forks and spoons, you'd actually have a combination of the states (1, 1, 1), (1, 1, 0), (1, 0, 1) and (0, 1, 1) realized among a quarter of the population each.

Footnotes:

[1] I'm using Euros (€) here because the dollar symbol messes with mathjax.

Growth and the business cycle in the information transfer model

There was some good discussion in comments on the previous post; I thought I'd summarize some aspects of the information transfer model (ITM) brought up by John Handley (on the components of the model involved in the growth rate) and LAL (on the business cycle). The overall model shows that without the recessions and the ITM trend, the fluctuations in RGDP growth do not have unit root -- which means they are essentially random fluctuations with zero mean. This means we can look at RGDP growth as the sum of three components:

These three components are pictured above:

Noise (in gray). Random fluctuations around the ideal growth path. It may be entirely measurement error or fundamental fluctuations because the economy is (in physics terms) a mesoscopic system with the system size N > 1, but not N >> 1.  Note that this is simulated with a normal distribution with standard deviation of 2 (percentage points) and mean of zero.

Non-ideal information transfer (recessions, dashed blue). This is the component that is least understood in terms of the model itself. The ITM allows non-ideal information transfer, but doesn't tell us what it looks like. The piece I put here is a guess (it's a step function on the leading edge and a Fermi distribution on the trailing edge). Although it is uncertain at this point how much of these recessions are real shock and how much are human emotion. More on that here. 

Ideal information transfer (information equilibrium, solid blue). This is the trend of GDP that falls out of the partition function approach (or just treating it as a line in NGDP-M0 space). RGDP slowly falls because as an economy grows, a given dollar is more and more likely to be found facilitating a transaction in a low growth market. I just simulated this with a slowly falling function here.

You can sum them up to produce these pictures of the growth rate and the level:

With this interpretation there really isn't a thing that I'd call a "business cycle". Going back to the sandpile analogy we have a randomly fluctuating flow of sand with an average rate. The height of the sand pile grows, but more slowly over time. Every once in awhile there is an avalanche and the height drops by more than usual. There is a cycle in the sense that the growth rate becomes negative and then positive again when avalanches occur. But instead of referring to the "sandpile height cycle", I'd personally just refer to "avalanches". By analogy, there isn't a "business cycle", but rather occasional "adverse shocks". Those shocks may be entirely random (although I have a speculative theory they are not in the sense that you can predict an increasing likelihood of shock ... and interest rates look like a good indicator too).

...

As a postscript, John Handley did have one question about the moving average that was the subject of the last post. The Great Recession looks like it permanently lowers the average growth rate in the graphs in that post. However that is just an artefact of using a 10-year window when the Great Recession was less than 10 years in the past. In the simulation above, I use a 5 year window to show that the effect of the lower average goes away when the shock leaves your averaging window (black line):

I also show the average rate using Cochrane's original method (gray). It comes out a bit higher as it did in the previous post.

Mathiness is next to growthiness (the 4% solution)

4% average growth has not happened since the 1970s.

I stole Sandwichman's excellent title from EconoSpeak.

John Cochrane put up a graph today trying to persuade us that Jeb Bush's goal of 4% real GDP growth is possible contra several other economists. I tried to figure out what he did because it couldn't have been a simple moving average (orange curve). His graph went until 2015 so it would have to be a backward looking average of some kind. But then both a backward looking moving average (green dashed) and a backward looking integral average (blue, they're basically on top of each other) don't quite make it up to 4% since the 1970s.

I did eventually figure it out (the red curve), but it involves the same mistake that I took Scott Sumner to task for back awhile ago. Cochrane took the change in RGDP from a point t - 10 to t and divided by RGDP at t. That gives the slope of the secant from the start point to the end point.

It may be called the mean value theorem, but that doesn't mean that the secant gives an average value. All it does is say there is at least one point that has that slope -- that 4% growth happened sometime during that 10 year period.

These diagrams might make this clearer. The first one (on the left) shows an RGDP function that gives 4% growth between the endpoints 10 years apart. You can see there is a point just after year 8 that has a tangent with the same slope. In the second graph (on the right) you can see that this function has nearly zero growth across the entire 10 year period. Not many people would consider that averaging 4% RGDP growth (it actually averages around 3.5% growth during that entire period per the averaging methods above). Cochrane's method over-weights the high growth spikes.

The method other economists are using to say growth has been less than 4% is the better method. Of course, you can choose to use the mean value theorem method if you want to overweight growth spikes to make growth to look higher ... and you're into mathiness.

The definition, origin and purpose of money

How is that for a bold title? Well, you're in for a (surprisingly short) strange ride from abstract mathematics to ancient history.

I have made the argument several times (first here, most recently here) that the most likely allocation against a n dimensional budget constraint with n >> 1, even when allowing states that don't saturate it, is actually at the budget constraint (M) because the location of the centroid of the n-dimensional polytope gets closer to the n − 1 dimensional budget constraint hyperplane as n → ∞. Here is a picture:

The hyperplane is the blue triangle (with maximum values of Cᵢ = M at the corners), and there are random points uniformly distributed along axes C₁ - C₃. You can see how the centroid (black) is a bit closer to the hyperplane than you'd expect for the 2D triangle that appears in the edge-on view of the budget constraint hyperplane. 

Let's imagine that budget constraint represents the total amount of money in the economy at a given time being used in transactions for various goods, services, investments, etc C₁, C₂, C₃, ... Cn.

What does the argument above imply? It means that (at any given time) money, if there isn't some coordinating factor, will most likely be completely allocated towards goods, services, investment, etc and that the difference between the information content in the money allocation [the information required to e.g. store everyone's bank balance] and the information content of the allocation of all goods and services [the information required to store the list of which goods belong to whom] I(N) − I(M) will be minimized so that if  I(N) ≥ I(M)

α I(N) = I(M)

for some  0 < α ≤ 1, and the information transfer equation (in the system N→M) becomes:

α (N/dN) log kn = (M/dM) log km

N/dN = k' M/dM 

with k' = (1/α) (log km/log kn). That is to say we've recovered an effective version of the original information transfer equation with a modified information transfer index k' but without non-ideal information transfer where N/dN  ≥  k M/dM. Maximum entropy results in the ideal information transfer condition I(N) = I(M) we've just assumed in the past (or taken to be a first order approximation).

Because of the increased number of identical states when goods are measured in terms of money, money helps saturate the entropy bound and therefore the budget constraint hyperplane. Combined with this post on how money can be introduced to mediate information equilibrium between two quantities leaving you with a theory that only requires one of the quantities and money when the information equilibrium equation holds ... we have a pretty complete theory of what money is and does.

Money is a thing that mediates transactions and has high information entropy

It maximizes information entropy when it has no intrinsic purpose other than mediating transactions -- i.e. if it is one of the commodities (or goods, or services, or investments, etc) C₁, C₂, C₃, ... Cn -- it will more likely line up along that dimension, resulting in C₁ + C₂ + C₃ + ... + Cn < M.

Note that we will fail to have ideal information transfer if the dimension is low (there will be larger fluctuations away from saturation) or the allocation of money becomes coordinated (e.g. panic and race towards one of the commodities Cᵢ). So a large, diverse economy that is totally random [1] is an ideal information transfer system -- an effectively ideal market.

...

The above definition leads to an interesting theory of the evolution of money. If gold or other metals were valuable and had intrinsic purposes besides mediating transactions (like being made into things), it is unlikely they would lead directly to money. Instead, the theory above suggests we should start out with the tokens of Mesopotamia that were likely used to keep track of transactions (see e.g. here):

Accounting tokens (?) of Mesopotamia. Image from here.

These are already intrinsically worthless except in exchange -- meeting the first definition of money above. One of these would have marginally greater information entropy than the others (and it likely wouldn't be the least valuable one or most valuable one), leading it to be taken in lieu of other tokens at various (market) rates, and eventually leading to that one becoming the precursor to money.

Now here's some wild speculation: what if we ended up with coins (flattish round things) in the West and Middle East because that was the shape of the Mesopotamian token with the highest information entropy? Like that one at the top left [2] ... eventually cast in metal because it needed to be durable (high information entropy means it's traded a lot), not because of the value of the metal (although there could have been some mixing of the two paradigms).

Update 6/17/2015:

Added "effective" in the narrative above. The result isn't an ideal market where I(N) = I(M) but rather an effectively ideal market where I(N) = α I(M) with α being some constant less than one.

Update 7/14/2015:

Here is some more evidence for my wild speculation: in China, cowrie shells were used as an early currency and subsequently cast in bronze/copper. See here. Picture below ...

Footnotes

[1] This suggests that news coverage of markets (the WSJ, CNBC, Bloomberg, etc) actually make markets less ideal as they can lead to coordinated behavior.

[2] I don't really mean it has to be the one in the picture. But there seem to have been a lot of similar-looking disc shaped ones that dealt with clothing, sheep, wool, etc. Even to the point of where a whole sheep probably had some interest rate relative to the wool of one sheep, leading to 5 wool = one sheep and the invention of exact change. The industrial revolution can be seen as revolving around clothing, why not the invention of money?

Economics really needs a framework

Representation of the functional renormalization group flow ... zzzz. If anyone wanted to know where this idea came from this is was probably the source.

There is nothing quite like the warm glow you get when you read discussions of technical subjects on the internet ... at least if you're a physicist. People attribute magical powers to you and think you comprehend subjects outside of your training and experience; the way you, physicist, approach a technical subject is the prototype for any approach to any subject anywhere. I know for a fact that it sometimes frequently goes to one's head. Anytime I'm feeling that way I go back and read one of these posts ([A], [B], [C]).

But if I'm ever feeling down, I will probably try to read Dierdre McCloskey's review of Piketty's Capital in the Twenty-First Century. Why? Because of Noah Smith's review of the first three pages of McCloskey's review.

Noah does a pretty good job of calling out the ridiculousness of what McCloskey says, but I thought I'd defend? ... nah, respond to Noah's characterization a bit more charitably and talk about where the "modern physicist as model of research" paradigm falls down.

My charitable impression is that McCloskey knows one or more experimental physicists at the University of Chicago. And I don't blame her. Experimental physicists are more human preferable as friends than theorists. Of my friends from graduate school that I keep in touch with, all are experimental physicists. And I was a theorist! I'm sure there are sociological reasons for this (experimentalists work in cooperative groups while us theorists are competitive loners; and I'm pretty sure the part of the brain that understands RG equations and BRST quanitization is the part of the brain most people use for social interactions).

Because McCloskey only knows an experimental physicist or two, she says things like this:

Piketty gives a fine example of how to [be a scientific economist]. He does not get entangled as so many economists do in the sole empirical tool they are taught, namely, regression analysis on someone else’s “data” .... Therefore he does not commit one of the two sins of modern economics, the use of meaningless “tests” of statistical significance (he occasionally refers to “statistically insignificant” relations between, say, tax rates and growth rates, but I am hoping he doesn’t suppose that a large coefficient imprecisely measured so far as sampling is concerned is “insignificant” because R. A. Fisher in 1925 said it was). Piketty constructs or uses statistics of aggregate capital and of inequality and then plots them out for ... inspection, which is what physicists, for example, also do in dealing with their experiments and observations. Nor does he commit the other sin, which is to waste scientific time on existence theorems. Physicists, again, don’t. If we economists are going to persist in physics envy let’s at least learn what physicists actually do.

As Noah points out, when physicists analyze data they tend to do it exactly as Fisher lays it out. Physicists didn't announce the Higgs discovery until it was significant at the 5-sigma level ... physicists cutoffs are generally higher (but not always) because we can run experiments. And for the theorists, it is almost always is someone else's data.

Noah also points out the many papers with "existence theorem" in the title, but I thought he missed the greatest example: the unsolved problem of the existence of quantum Yang-Mills theory and the mass gap. But I also think we should be more charitable to McCloskey here because there are different types of physicists.

Overall, McCloskey's description is an appropriate one for an experimental physicist. A phenomenologist (a theorist that connects theory to experiments) plots out lines and does check significance -- ruling models out. A theorist sometimes does prove existence theorems, although sometimes that kind of thing is reserved for mathematical physicists.

That is to say there are four jobs (not always clear-cut) in modern physics that essentially differ by how closely they deal with data:

• Experimental physicists [arXiv: hep-ex, nucl-ex, etc]: create data
• Phenomenological physicists [arXiv: hep-ph]: connect theory to data
• Theoretical physicists [arXiv: hep-th, nucl-th, etc]: make theory about data
• Mathematical physicists [arXiv: math-ph]: analyze theory

I was sort of a phenomenologist, but my papers are all in nucl-th (they haven't broken out nucl-ph yet). Overall, here's an example of how this works together using an accelerator experiment from JLab that tested some theories (one of which I put together):

This graph from Steffen Strauch shows the results of some polarization transfer experiments where they shot polarized high energy electrons at Helium nuclei and looked at the polarization of the protons (Hydrogen nuclei) that got knocked out at different amounts of energy transferred (Q²). The data came from the experimentalists. The calculations came from some phenomenologists and theorists (the orange dashed curves are my theoretical nuclear model [CQS] plus a piece that accounts for the scattering phenomenology in the experiment [RDWIA]). My model was the chiral-quark-soliton [CQS] model -- and the topological properties of the soliton solution in my model came from mathematical physicists working on quantum field theory.

And I think this is where we can let McCloskey slide with her comment on existence theorems. It's true she is wrong saying physicists don't prove existence theorems, but she is right that economists shouldn't waste time on them. The reason is that economics doesn't appear to have a mathematical framework for mathematical economists to work with. It has a bunch of models and a couple of general rules (marginalism and optimization) but those are not frameworks.

Optimization is not a framework itself. Even given something to optimize (e.g. utility) doesn't make optimization a framework. In physics you optimize "action" (i.e. the principle of least action ... action is energy times time), energy (e.g. finding the lowest energy configuration), entropy (maximizing it in equilibrium), etc but these happen inside a framework of (classical or quantum) Lagrangian dynamics, Hamiltonian systems or statistical mechanics. These are actually all related to each other and form the basic framework of physics. And that framework is what mathematical physicists work on. Without a framework, there shouldn't be proofs of existence theorems.

If economists are out there proving existence theorems it seems to be a bit like natural philosophers proving existence theorems about physics in the 1600s before its first framework (from Newton). It's really important to note that the first framework practically invents the branch of mathematics that Newtonian mathematical physicists would work on! Calculus got a big leg up from Newton creating a practical application for it; it's possible that the entire branch of mathematics that will be associated with the first economic framework doesn't exist yet! I imagine mathematical economists today are proving the equivalent of number theory theorems associated with Aristotle's crystal spheres. It would be quite a coincidence if the mathematics of space and geometry of manifolds stemming from calculus could be adopted wholesale for use in economics. (My opinion, and the raison d'être for this blog, is that the eventual economic framework should come out of information theory.)

In any case, economists should be modeling themselves after physicists from the 1600s! It would be a weird mish-mash of philosophy, religion, invented mathematics, astronomy, and astrology with long rambling treatises by people desperately trying to get at something but not exactly sure what.

That is to say they should keep doing exactly what economists already do. Zing!

Don't take this quasi-defense of something McCloskey said as an endorsement of anything else. I agree with Noah:

But my main problem with Ms. McCloskey is not the poorly executed flowery baroque writing style, or even the reminder that plenty of people mistake flowery baroque writing for good writing. It's that McCloskey frequently makes declarations that are, to put it politely, in contradiction of the facts. She says these things with utmost confidence but without evidence or support, making it clear that the fact that she has said them is evidence enough. She argues from authority, and the authority is always herself.

That is to say she is an economist. Zing!

In the end, Noah is puzzled as to why John Cochrane thinks the essay is excellent. I'm not. McCloskey bashes Piketty:

... Startling evidence of Piketty’s miseducation occurs as early as page 6.... "If the supply of any good is insufficient, and its price is too high, then demand for that good should decrease, which would lead to a decline in its price.” The [emphasized words] clearly mix up movement along a demand curve with movement of the entire curve, a first-term error at university. The correct analysis ...is that if the price is “too high” it is not the whole demand curve that “restores equilibrium” ...

Wait: doesn't it depend on whether supply or demand adjusts faster to the "disequilibrium"? In the information transfer model, it does. (Generally, both adjust together.) Piketty is saying demand adjusts faster. McCloskey is saying supply adjusts faster. Maybe I am wrong. It would really be nice to have that framework right about now.

Cochrane basically has praised McCloskey's essay because she comes out defending capitalism -- but not before re-branding it "trade-based betterment". We can't forget there is a lot of politics pervading a field that isn't understood very well. And let's not also forget Newton was a neurotic weirdo into alchemy and a crypto-goldbug as the Master of the Mint in England.

Euler's theorem and non-ideal information transfer

Leonhard Euler and Claude Shannon.

Dietrich Vollrath has a handy framework for looking at the issues I looked at here about Romer's discussion of Euler's theorem and rival inputs:

1.  Output is constant returns to scale in rival inputs

2.  Non-rival inputs receive some portion of output

3.  Rival inputs receive output equal to their marginal product 

Pick two. 

Romer’s argument is that (1) and (2) are true. (1) he asserts through replication arguments ...

In my post I argued that (1) is generally not true because replication arguments assume replication of non-interacting things, indistinguishable things or effectively infinite things.

Now, I'd like to show in the information transfer framework that you don't have to pick two (Euler's theorem doesn't tell you anything) -- even if we let (1) be true. Let's start with a production function $Y(R, N)$ where $Y$ is output, $R$ are the rival inputs and $N$ are the non-rival inputs. Therefore we can say:

$$
\text{(a) }\; \frac{\partial Y}{\partial R} \leq \alpha \; \frac{Y}{R}
$$

$$
\text{(b) }\; \frac{\partial Y}{\partial N} \leq \beta \; \frac{Y}{N}
$$

$$
\text{(c) }\; Y(R, N) \leq c \left( \frac{R}{R_{0}} \right)^{\alpha} \left( \frac{N}{N_{0}} \right)^{\beta}
$$

Turning equation (a) around, we can say:

$$
\text{(d) }\; \alpha Y \geq R \frac{\partial Y}{\partial R}
$$

So now we can talk about Vollrath's framework in terms of our symbols:

1.  $\alpha = 1$
2.  If  $Y = p + w R$ and $p > 0$ then $w < \partial Y/ \partial R$
3.  If $w = \partial Y/ \partial R$, then $Y = w R$

If we take the equality sign in equation (d), then yes, (1), (2) and (3) cannot be simultaneously true (only two can be true at the same time). But if we take the greater than sign in equation (d) we can say:

$$
Y = p + R \frac{\partial Y}{\partial R} = p + w R  >  R \frac{\partial Y}{\partial R} = w R
$$

where $w = \partial Y/ \partial R$, $p>0$ and $\alpha = 1$. All three of (1), (2) and (3) are simultaneously true!

Where this comes from in the information transfer framework is the idea that wage (or price) of $R$ doesn't necessarily reflect the final value of $R$ -- i.e. $V(R) \geq w R$. Maybe there is a behavioral economics reason for this (the endowment effect comes to mind). Maybe there are other reasons (e.g. the computer cluster in this post). The information transfer framework doesn't explain what is happening when the information $I(Y) \geq I(R)$ ... it just allows it to happen.

And a lot of the time you can get away with $I(Y) \approx I(R)$, but invoking Euler's theorem as a theoretical constraint requires ideal information transfer $I(Y) = I(R)$ even if you force $\alpha = 1$.

Everything is awesome

I mentioned in a footnote to the last post that it's fun to replace adjectives modifying a central bank or monetary policy and policy instruments with various conjugations of the word awesome. Here's an example with a post from Matt Yglesias when he was back at Slate:

Is [awesomeness] hard? 

One way for monetary policy to be [awesome] at the lower bound is to make a promise about the future. You say that not only will short-term interest rates be [awesome] until full employment returns, but a bit beyond that to the point where you get a little [extra awesomeness]. One issue with this, per Paul Krugman, is that a central banker attempting to [be awesome] may face a [awesomeness] problem. In Krugman's memorable phrase, you have to [awesomely] promise to be [awesome] in the future.

This is a funny line and a sound underlying model, but as I've said before I don't think it's worth worrying too much about [awesomeness] in practice. The fact of the matter is that these [awesomeness] problems arise all the time in life, and resolving them is pretty routine. 

Consider a boss overseeing a team working on a project that's about 5-6 weeks away from conclusion. Unfortunately, the client really wants it done by Thanksgiving. The boss says to his team, "guys if we can buckle down and get this done by Thanksgiving then we can all take the whole week between Christmas and New Year's off." On its face, there's [an awesomeness] problem here. By December if the project is already complete, the boss has no incentive to [be awesome]. And such things do happen. People lie. Workers get screwed. Such is life. 

But at the same time, [being awesome] about the future is pretty routine. Part of navigating through life competently is avoiding a reputation for being [not awesome], and most of us seem to manage to pull it off. 

And the Fed is in some ways in a more [awesome] position. To summon [awesomeness], it's not necessary to persuade everyone that the Fed's promised [awesomeness] will come to pass. It simply has to be the case that on the whole people revise their expectations about the [likelihood of future awesomeness]. If Bernanke makes a promise [to be awesome in the future], some people will [think he's awesome] and nobody's going to revise their expectations downward to offset that. So the strategy should [be awesome]. What's true is that if the central bank plays the [being awesome] card once and then [isn't awesome], they'll have a very hard time [being awesome] a second time. But that's an excellent reason to not [not be awesome], and that itself is part of what makes the promise [awesome].

Awesome.

You forgot to use my model: the non-case of Icelandic austerity

Somewhere along the way to the West Fjords from Reykjavik in Feb 2013

Scott Sumner has two posts up in a row now that quote others (Britmouse and Mark Sadowski) wherein everyone involved is invoking what is usually called straw man argument: proving an argument is wrong by inventing a different easier argument no one has made to prove wrong.

In the first post, Scott does attempt to deal with the problem of making a straw man argument in comments -- saying his post only shows the Keynesian interpretation isn't obvious given the data. But that's another straw man: no one is saying any interpretation is prima facie obvious. Everyone from Diane Coyle to John Cochrane to Janet Yellen to Paul Krugman knows that all economic data requires a model to interpret it. Therefore "obviousness" (or lack thereof) is model dependent. I attempted to show what obviousness looks like in a completely non-economic model (namely my background in physics) in this post.

In science, if you're trying to show some model is incorrect you have to bend over backwards to not only understand the model you're attacking but also bend over backward show how that model can be consistent with observations. That is to say obviousness or lack thereof is always a straw man argument.

You should actually assume any data is obvious in any given model! And that brings me to my key point: if you are trying to prove a model wrong with data, you need to find out how any given data is obvious given the model [1]. If you are trying to prove me wrong, don't forget to use my model ... and use it correctly [2].

This is where Britmouse and Sadowski fail. Britmouse doesn't account for the possibilities of different counterfactuals -- that's why I called the graph a derp Rorschach test: you come away with your prior because everyone's prior is consistent with any data given the right counterfactual.

Sadowski's post is more nuanced. He fails to correctly set up the antecedent. I'll forego quibbles with the specific measures of government spending data [3] and cede that Iceland did engage in fiscal consolidation between 2009 and 2014. I'll even cede that fiscal austerity was offset by monetary policy. I'll even emphasize it with an indented block ... in bold type!

Iceland engaged in fiscal consolidation at some point between 2009 and 2014 that was offset by monetary policy.

My question is this: who cares? This is the basic economic consensus of how fiscal and monetary policy work. This totally misses the point of the austerity debate. For that, you need to address several points:

1. Was Iceland at the zero lower bound from 2009 to 2014?
2. Was Iceland in a liquidity trap from 2009 to 2014?
3. Was the austerity implemented when monetary policy couldn't offset it?

For that, let's get some data from FRED. Sadowski and Sumner have argued that the EU wasn't at the zero lower bound because nominal rates were above zero. Now this doesn't matter in the information transfer model, but I'll try and address their concerns by looking at real rates. The idea of a liquidity trap is that nominal rates and inflation are too low to push real interest rates down far enough to spur private borrowing. In a sense, you can't get the real rate to where the Taylor rule says it should be. Inflation can't credibly be pushed higher to bring real rates lower because everyone thinks the central bank will just take any base increase back.

I also had to use year-over-year inflation because there doesn't seem to be any seasonally adjusted price level data from European countries [4], so there's more correlated noise in this data than I'd like. But I think it still shows rather well that Sadowski isn't addressing the austerity debate.

So here is the real interest rate data for Iceland, the EU and the US:

1. Was Iceland at the zero lower bound from 2009 to 2014? No -- either in terms of nominal interest rates or real interest rates. The massive deflation devaluation allowed real interest rates to reach well below the values for the US and the EU.
2. Was Iceland in a liquidity trap from 2009 to 2014? No -- Iceland was never in a liquidity trap. It was able to push real interest rates to large negative values and now real interest rates are positive. 
3. Was the austerity implemented when monetary policy couldn't offset it? According to the liquidity trap model, Iceland seems to have never been in a liquidity trap and could likely have always offset fiscal policy.

So Iceland and Sadowski's post is basically irrelevant to the austerity debate. But there's a bit more.

The thing is that Sadowski's time period from 2009 to 2014 is a bit disingenuous. Austerity from 2009 to 2012 could ostensibly be called "austerity" in the sense that the term is used: cutting government spending during a downturn in order to spur business investment because government borrowing won't crowd out private borrowing. The finance minister of Iceland declared victory over the recession in 2012 (as Sadowski points out). And we can look at the table that Sadowski points to himself:

Iceland (cyclically adjusted general government balance)

From Table A4 [pdf] 

2006    4.7  

2007    3.1 

2008   −4.6 

2009   −6.9 <--- start austerity (Sadowski)

2010   −4.9 

2011   −1.8 

2012    0.4 <--- Finance minister of Iceland 

2013    2.3         declares victory 

2014    6.2 <--- end austerity (Sadowski)

Since the finance minister declared victory over the recession in 2012 (which makes sense in terms of the interest rate trend in the graph above -- it's where it crosses through zero -- as well as the table), it is disingenuous to declare any of the fiscal consolidation from 2012 to 2014 "austerity". Austerity is "we must cut back because the economy isn't recovering" not "we must cut back because the economy has recovered". In Sadowski's graph that Sumner reproduces, this moves Iceland from a 13.1 percentage point shift to a 5.1 percentage point shift leaving it just to the left of the Netherlands and Denmark (ceding that the other amounts are valid). Note that puts Iceland to the right of any of the countries in the news about austerity being wrong-headed: Greece, Spain, Ireland, Portugal and the UK.

When Sadowski says this:

This is because it includes the increase in spending attributable to rising interest payments on the national debt.

You should immediately know he's talking rubbish. Rising interest payments is a key indicator that the government should start reigning in spending -- reigning in spending because of rising interest payments is not "austerity".

Sadowski's refusal to look at what the Keynesian model says for Iceland not only led him to make an argument that is tangential to the austerity debate, but to make a disingenuous use of data that hurts the credibility of the monetarist model. It's going to be hard for me to look at an argument from Sadowski and not say: "OK, where is he trying to trick me this time?"

It's always good advice: try and understand things from the perspective of someone who would disagree with you [5]. As everyone disagrees with me, I've become somewhat of an expert!

Before you say that I don't get the monetarist argument note that 1) I actually ceded the primary argument of Sadowski's post: Iceland offset monetary fiscal policy, 2) the information transfer model cannot be distinguished from a monetarist model for an information transfer index κ ~ 0.5 -- even I believe in monetary offset in certain cases! and 3) I can see how the liquidity trap doesn't make sense given a determined central bank in the monetarist model [6].

Footnotes:

[1] This is not the only way to disprove a model. You can show the model predicts something that doesn't happen (falsify the model). You can show the model isn't falsifiable (market monetarism isn't falsifiable). You can also show a different model is much better at explaining data. If that different model is itself falsifiable, it can be used to disprove an unfalsifiable model. [This is how e.g. science works against religious theories of the physical world.]

[2] There is a corollary to this: if you say your model is consistent with the data, I can't come back and say you forgot to use my model which shows it isn't. I am showing the hypothesis H

H(data | my model) = True

not 

H(data | my model ∪ your model) = True

[3] This doesn't look like austerity to me, but then that should be obvious given my model.

[4] FRED: Can we get a toggle to apply a seasonal adjustment to any data? You can add a disclaimer about it with all kinds of caveats. Yes, it's numerically complicated -- maybe just a toggle to switch between NSA and SA data if it's available on each graph?

[5] It's probably not good advice when trying to get internet traffic -- that seems to flow from straw man attacks and derp.

[6] I just don't like "no true Scotsman" arguments in my theories. Whether a central bank is "determined" or "credible" is rhetorically the same as a question of whether a central bank is a true Scotsman. I personally like to replace terms like "credible" and "determined" or even "competent" with "awesome" whenever they are modifying the noun "central bank". Here's a good example:

On theoretical, practical, and historical grounds, policy [awesomness] is simply not a problem for fiat money central banks.

Everything is awesome. 

Counterfactuals and the second derivative

Scott Sumner put up on his blog what I called a derp Rorschach test on teh twitters -- since no counterfactual was presented in the picture, you can invent whatever counterfactual you want to confirm your priors. However, I'd like to point out that it is far from obvious that you can draw Sumner's preferred conclusion for a simple reason: the second derivative.

See, the curvature of total employment in the graph on Sumner's blog is far sharper at the beginning of the dip than at the end -- something consistent with an inertial picture of the economy where "forces" act on employment much like forces act on objects. In that case, we can have an acceleration that is more than normal for awhile (normal is just trend growth) and then becomes less than normal afterwards. Stimulus and austerity. It looks something like this:

Of course, this is perfectly consistent with the picture Sumner shows (I added the line from the graph above):

Basically, the idea that this picture "obviously" shows anything is just confirmation bias. We need to know your counterfactual! And I can show you the counterfactual for the model above ... it looks like this:

Sumner can't show you a counterfactual. In fact, no market monetarist can ever show a counterfactual -- no market monetarist is the market. What the market would have done in an alternate world is completely unobservable. That's probably why Britmouse and Sumner think the graph shows something obvious. Whatever happened is the only thing that could have been observed to have happened!




Update:

One thing I wanted to note, but forgot, is that ordinary automatic stabilizers have roughly the same effect of "stimulus + austerity". When a recession hits, a lot of money goes out as unemployment insurance -- but it slows over time.