Hacker News Reader

This is way too kind. Once upon a time there was a school of important, economists who were literally Austrian -- Menger, Bohm-Bawerk, Hayek -- who had certain commonalities. Mises and Rothbard turned it into an ideological movement that deliberately did not engage with anyone else's ideas. A healthy intellectual movement doesn't need to be convinced to pay attention to empirical evidence.

Fair point. Although in my defense I don't come to HN looking for or to start ideological fights :)

>A healthy intellectual movement doesn't need to be convinced to pay attention to empirical evidence.

You misunderstand the reasoning behind the "rejection" of empiricism by Mises and those that followed him.

Austrian theory as formulated by Mises is based on a priori truths about human nature and cannot be falsifiable by evidence. This may sound outlandish when you first hear it especially when one comes from an engineering or natural science background, but there is a very well developed epistemological reasoning for this position that is often overlooked by critics.

The economy is composed of humans who act towards their own personal goals, whether that is to provide for their family, or become a billionare tech founder. This is a highly complex system with an incredible amount of variables (in the case of individual humans, the ordinal scale of preferences is quite literally infinite!). As such, there are no constants that can be inferred with the same level of reliability that we may have in physics like the rate of gravity on planet Earth.

Thus, empirical economic models have a much, much greater margin of error than what we have in the natural sciences, and it is therefore untenable to formulate economic theory based on inherently unreliable foundations.

Instead, as the economy is composed of humans acting and not objects acting, the Austrians look inwards and through a Kantian-esque process of introspection formulate the a priori action axiom ("Human action is purposeful behaviour.") and build their theory using logic based on that axiom - thereby creating a system of knowledge that is unfalsifiable through evidence and is always true.

Ironically, mainstream economics has taken the "purposeful behavior" idea very seriously. Models almost entirely consist of agents acting purposefully. If agents are smart and prices are set by supply and demand, then money doesn't have real effects and a pure quantity theory of money holds. If agents are dumb, then you get mispricing but markets still clear and you don't get unemployment or factories closing. If prices are sticky, then it doesn't matter if agents are smart or dumb -- you get recessions that last unnaturally long unless either the government or the central bank steps in. So New Keynesianism is completely compatible with "human action is purposeful behavior".

But still, the rest of us prefer to test our theories by looking at evidence.

Instead, as the economy is composed of humans acting and not objects acting, the Austrians look inwards and through a Kantian-esque process of introspection formulate the a priori action axiom ("Human action is purposeful behaviour.") and build their theory using logic based on that axiom - thereby creating a system of knowledge that is unfalsifiable through evidence and is always true.

Weird, Marxist-Leninists say the same thing (which is why I'm not an ML).

How do we know whether this system of knowledge actually describe our world?

If I understand the grandparent's point, we have to accept it as an axiom.

Which leads to the question: if "because I assume so" is evidence enough for making statements, how do you sort of the conflicting predictions of one theory based on "I assume so" and another based on as much evidence?

To give an example, let's assume that central bankers can make everyone better off by picking the right inflation rate. Therefore it is good to centrally plan the price and quantity of money. The grandparent's comment "let's assume"-based theory concluded that it is bad to centrally plan the price and quantity of money. Which advice should we follow?

I suggest we try an experiment. Appoint an austrian (say, Alan Greenspan) to run the US federal reserve. See if we get improved stability, or the largest recession since the Great Depression.

Sorry, empirical evidence should not inform our thinking because it is too shaky. Please either assume one of those two outcomes is correct.

Come on, no one in a position of government authority would do anything that stupid.

>However, the school of thought's aversion to econometrics makes their policy advice self-limiting compared to mainstream/mainline economic theory.

This I agree with, I do think there is tremendous value in using statistics as a way of analysing trends and reinforcing theoretical arguments with empirical data. I'd disagree strongly on the point that Austrian theory is not ignored by mainstream economists though as it seems like many of the underlying assumptions that justifies the Federal Reserve are rather flimsy when examined critically from an a priori Austrian perspective.

>As for your specific points, that the large central financial institutions are, according to Austrian theory, causing recessions, even Austrian RBC doesn't completely agree.[1] Economic busts will happen with or without central banks -- being able to bail out key financial platforms is a good thing if it prevents the unnecessary misery of millions or billions. As every economics student should learn in the second half of economics 101, the free market does not always perfectly allocate resources, either immediately or over time (see discussions on externalities).

Thanks for pointing this out, maybe I wasn't too clear in my post. I elaborate a bit more on this in a comment below, but didn't touch on the neccessity of having an institution able to bail out financial platforms during economic downturns.

I do think that this boils down to the question if we can ethically and legally justify fractional-reserve banking, as bailouts would not be necessary if banks were mandated to keep a 100% reserve on demand deposits. I'm personally a bit split on this issue and so are many other Austrians, with purists adhereing to the Rothbardian view that only full-reserve banking is ethically defensible while others prefer leaving it to the free market to decide. I believe that in the latter case private insurance companies would be more than suitable to provide services to bail out banks who overextend credit, and would also serve as a natural check on said overextension as an insurance company would not be willing to take the risk to insure an irresponsible bank with a tendency to not make careful judgements.

If anything, having a central institution that will bail out banks regardless of what happens is far more dangerous as it introduces a significant moral hazard whereby banks can take significant amount of risk (within applicable regulations, Basel III made it a bit more difficult after 2008) and know that they will be bailed out regardless of what happens. I can't say that this leads to a more stable economy than it would were we to have a full-reserve banking regime, or even a free market where banks choose their own reserve ratios.

> If anything, having a central institution that will bail out banks regardless of what happens is far more dangerous as it introduces a significant moral hazard whereby banks can take significant amount of risk (within applicable regulations, Basel III made it a bit more difficult after 2008) and know that they will be bailed out regardless of what happens. I can't say that this leads to a more stable economy than it would were we to have a full-reserve banking regime, or even a free market where banks choose their own reserve ratios.

Having worked in banking for the better part of the last decade, I agree with your points here. Basel III, CECL, CCAR have forced banks to be extremely conservative their lending, with a de facto scaling the conservativeness in the retail lending standard to their relative size. This being said, the banks that are designated as GSIB at least presently keep an awareness in mind of their sizes. (I worry that won't translate to bank executives over the next decade as those standards relax / roll back).

I can definitely see a situation where standards get relaxed or GSIBs opportunistically find loopholes and get us back to an overleveraged situation like 2008. While the repeal of Glass-Steagall opened up wonderful innovations in finance it also exposes retail clients to substantial risk -- and I don't think that has been substantially remediated.

If you're so inclined, can you comment on what you think Ron Paul was getting at when he used the phrase "Honest Money"? I'd like to hear what an Economist has to say about this.

He uses the term to describe using gold and silver as currency instead of notes issued by a central bank. The ideological point is that people are fallible (or in his view manipulate) and some seemingly independent metric is superior.

This ignores the empirical, historical evidence of the inflationary/deflationary boom/bust whipsaw of the gold standard eras (19th century in particular) which were hugely destructive and ruinous.

More distressingly it also ignores the fact that money is a signaling mechanism and when the amount in circulation is out of relation to economic activity the economy breaks down. Yes, humans have a hard time figuring out what the level of economic activity is, but I am quite sure that gold strikes are not correlated with the production costs of iPhones and t shirts.

I've actually not heard his use of the term. Can you provide a reference to see the context?

https://www.amazon.com/Pillars-Prosperity-Markets-Private-Pr...

https://www.lewrockwell.com/1970/01/ron-paul/bring-back-hone...

https://youtu.be/IF5pwDCpz4s?t=450

Why is econometrics important? I went to UChicago during 2007 and econometrics was a major bane of all Econ majors. There was something so anti-pragmatic about reducing irrational economic actors into mere equations.

Meanwhile, did all of that sophisticated math improve the economy or predict the recession or help us out during recovery? It's nice to have hidden mathematical knowledge but is it actually providing benefits or is it just arcana?

Econometrics (aka stats for economists) is how we test models on real data. It seems to often be confused with the idea of using math in econ however ("reducing irrational economic actors into mere equations" seems to me to be economic theory, not something specific about econometrics).

This has two purposes:

- we would like to know which of our models are true, or at least good enough, and which ones are not.

- we would like to know the actual magnitude of things. For instance, if we increase tax rate by 1%, how large will the taxable income reaction be? Decrease by 0%? .25%? 1%? 10%? That's useful information for policy making.

Is this a joke? Econometrics is literally how you test models, quantify changes in the economy, etc... Versus just hand-waving and hoping your assumptions work.

And the reason economics doesn't always help the economy is because politicians almost never listen or shills are paid to advance policy that's detrimental to the economy.

> It's nice to have hidden mathematical knowledge but is it actually providing benefits or is it just arcana?

It depends on the context. Plenty of areas of econometrics have substantial mathurbation, as do all quantitative fields. But the toolkit is not without use.

(1) One common critique (which you leverage here) says that econometrics isn't useful because it doesn't predict business cycles. Macroeconomics divorced from microeconomic foundations[0] is a difficult domain to apply reasonable econometrics -- insufficient data to describe the system completely.

(2) Individuals often present as irrational but populations tend to behave rationally -- meaning that descriptive information or predictive information can be ascertained.[1] This is the central point of ML/predictive analytics applied to human systems (e.g. marketing), which is where the value-add of ML can come from.

(3) Remember the utility of econometrics is to measure economics -- that might be something like a value impact assessment, describing the output of a randomized control trial (for which the Nobel prize[2] was just awarded[3]), and similar. Economics is much bigger than only the macroeconomy.

Notes:

[0] The Lucas critique: https://en.wikipedia.org/wiki/Lucas_critique, Theory before business-cycle measurement: http://faculty.wcas.northwestern.edu/~lchrist/papers/qr1042....

[1] To describe or predict: https://www.stat.berkeley.edu/~aldous/157/Papers/shmueli.pdf

[2] I am aware the prize is not a real Nobel, the name is The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel

[3] List of Economics Nobel Prize winners: https://en.wikipedia.org/wiki/List_of_Nobel_Memorial_Prize_l...