Hacker News new | comments | ask | show | jobs | submit login
Economics is overdue for a quantum revolution (aeon.co)
54 points by tgragnato on Jan 12, 2018 | hide | past | web | favorite | 34 comments

All models are wrong, some are less wrong than others.

Just like the standard model is just a model, economics relies on models as well. Matter of fact is that it's hard to predict what humans in aggregate are capable of. But I welcome anyone trying to come up with new models ...

Full disclosure, econ PhD student (high-freq trading area)

> All models are wrong, some are less wrong than others

A sentiment often echoed by economists. I'd be inclined to counter with the following:

All models are wrong. Some are useful, some are dangerous. Do you have a mechanism to tell the difference?

Models that bury certain uncomfortable realities in assumptions (e.g. "perfect information/competition") are useful to some people but not others.

What I'd like to see studied is any connection between funds allocated to econ research and the assumptions made in the publication output.

More study on how econ research is funded in general would help to better understand the incentives and motives of researchers.

> All models are wrong, some are less wrong than others.

All models are wrong, some are useful

I wanted to quote both, but opted not to in the end. I like both quotations as they are both true in my honest opinion.

IMHO the "less wrong" isn't very relevant, while the "useful" is the important part :)


Hey, I have a few questions regarding the Econ-PhD you're doing. (Basically questions about the work you're doing, is it worth it??...etc.) My email in my profile if you'd like to get in touch.

(By is it worth it, I mean your personal opinion on the value you are getting from your current area of research, which you've mentioned is HFT)

What is a high-frequency trading PhD?

I'm doing research into high-frequency price contagion!

What a whole load of bad and unnecessary analogies.

A "classical physics + some constraints" analogy for Economics would also do the job.

Just because some magnitudes are discrete in a system, it doesn't mean quantum mechanics is needed to describe it. The distinguishing features of quantum mechanics come from the counter-intuitive mathematical predictions (entanglement, superposition, probability distributions, etc). But these are very specific and have no known analogy for Economics.

I agree that the quantum analogy is terrible but, as a physicist, I don't believe anything like a classical theory is possible for economics. At least not for getting the kinds of results that people want out of economics.

Physics is different because there are measurable fundamental laws that apply in all systems - or at least we have been able to determine some set of laws or approximate laws at scales where we can calculate useful things with them. Finding analogous fundamental laws in economics just won't be possible. There is no version of gravity or electromagnetism in economics where indivisible (or approximately indivisible) objects have known masses or charges. The closest thing you have is game theory, which is still very far from universal - I can't enumerate all the possible options available to each individual agent in an economic system and assign an objective utility to each option for a given agent. Not even a single individual can do that for themselves.

A better physics analogy, which still falls short, is the study of emergent phenomena in complex systems. When you take large numbers of objects with well-understood properties and observe them in certain environments, large scale processes can occur across the system which cannot be easily explained by the known properties of the individual components. The sum is greater than the whole and just because you find an empirical relationship of the system in one environment doesn't mean you necessarily understand anything about its behaviour in other environments. This is what is happening in economics but we don't even understand the individual components. Assuming that game theory is the correct way to understand the dynamics of individuals, we would need a more complete theory of psychology to fully describe them, which would probably require a better understanding of neuroscience, etc.

>emergent phenomena in complex systems

Emergent phenomena in complex systems are not necessarily quantum. Weather, population dynamics, neural networks, chaotic systems can all be described with good'ol lagrangian mechanics. Other phenomena like certain phase transitions or spin glasses cannot. But what's the economic correspondence to those?

Overall I agree with you. Economics should be described with the math and physics of complex systems. Unfortunately data is poor, the math is hard, the psychology is complex. So we're not quite there yet.

Go look up representative agent models. You'll be horrified.

> Economists, it seems, think about money less than most people do: as Mervyn King, the former governor of the Bank of England, observed in 2001: ‘Most economists hold conversations in which the word “money” hardly appears at all.’

Eh yeah, that's because there are other words for money. An Economist thinks about resources, their (perceived) value, and less about cash. It's because cash somehow ends up representing those points.

That doesn't mean an Economist 'thinks/talks less about money'.

Wonderful set of video lectures on Money and Banking by Perry Mehrling, where he emphasizes exactly this topic of economists abstracting away from money and liquidity, therefore leaving abstraction leaks in their theories.




Oh, it's Aeon. A.k.a. Vice for people who've gotten bored with Vice's bourgeois pedestrianism. Do you think unconventional thoughts, like "due process is overrated"? Do you barely think at all? Then Aeon is the platform for you! It's the "your views are intriguing to me and I wish to subscribe to your newsletter" newsletter.

I'm in support of new science-based economic theories but this author seems too ignorant of both economics and quantum physics to suggest a useful, actionable alternative to the current models. Instead we get on-the-nose analogies like money being like elementary particles in quantum mechanics. How does this help economics formulate better models of how people produce, acquire, and trade goods?

I suppose the subject's inability to measure anything with any precision is sort of quantum. They could make that a theorem about the nature of economics rather than their own professional failings.

Yet another of those articles about economics.

"For one thing, if economics is about solving scarcity and making people happy by optimising prices, then it appears to be doing a rather poor job. In many countries, inequality has ballooned in recent decades, while reported happiness levels seem to have peaked some time back in the 1960s. The financial crisis didn’t make many people happy, except some bankers."

This is like blaming the wheels for the fact that you crashed your car into a wall.

Or, imagine we would allow people to vote (along with all the political circus) on how physical principles are applied in industrial production: Why are the wings on 747 so ridiculously long and the capacity limited to couple hundred people? And what fool would build a ship from steel? Every child knows steel is heavier than water and will never flow, right?

Once our societies will be ruled by technocratic dictators applying the science of economics to the whole market you could write articles like this.

> Once our societies will be ruled by technocratic dictators applying the science of economics to the whole market you could write articles like this.

I've lived in one such state (only for like 8 years, but still). It was a complete failure. They had to inroduce martial law to keep people in line, and finally it defaulted and dictators left on their own.

It didn't help that the science applied was marxism-leninism, but I'm not sure any of modern theories are that much better if you let a dictator follow them to the letter.

I'm not sure that marxism-leninism is a science, but if we're going to use the Soviet Union in the 70-80s as an example of failure, one could just as well use success of China of the 00's.

Most economic theories are tested by the scientific method, and it doesn't seem that was the case in the Soviet Union or its satellites, which I assume you are referring to.

Frankly, by the 80s it was difficult to see the Soviet system as much more than a kleptocratic gerontocracy (theft by the government controlled by the old). The methods of distribution were no more communist than those currently used in China.

> Most economic theories are tested by the scientific method

I don't think they are. Can you give me an example of a prominent theory developed in the last thirty years that is considered by the mainstream to be debunked by data?

I even agree with you that economics is pretty soft as science goes, more math but also a lot of p-hacking, but... they are argued, provide data, and are evaluated just like the other Social Sciences. See Picketty as a pop-econ example with plenty of research backing.

The idea that in order for something to be a science a prominent theory developed in the last 30 years has to be debunked by data is silly. Most of the questionable basis for economics is much more than 30 years old (1988?). I might as well ask you the same question in Physics... is Physics no longer a science?

I'm not a physicist but the discovery of the Higgs boson probably put paid to a number of competing theories [0].

Some papers probably are well-reviewed and evaluated. But as a small experiment, I picked a paper at random (Artificial Intelligence, Automation and Work [1]) from this year's AEA conference and had a look through it. Despite having a promising title, there is (to my mind) no content in there. Nothing but unjustified assumptions about 'production functions' and 'productivity'.

I don't see how a paper like that tells us anything about artificial intelligence or automation or work.

[0] https://en.wikipedia.org/wiki/Alternatives_to_the_Standard_H...

[1] https://www.aeaweb.org/conference/2018/preliminary/paper/HZr...

Well, kinda. The Higgs and Higgsless proposals mostly go back to the 60s and mid 70s, even if some papers are later. The actual discovery was on 2012 40-50 years later and the Nobel handed out.

A random short paper from even a good journal like Phys Rev Letters is unlikely to be Nobel earthbreaking either, much less have effort to debunk it.

As an undergrad I had to discover that the 20 year olf PRL used as a basis for the 2 week Lab (Thorium C x-section) actually had an incorrect analysis. All data taken was at the wrong energies so we had to spot the error, rederive the equation and values to measure it correctly before the halflife brought our counts down too low.

That PRL was never corrected and the TAs comment was, "that'll teach you to trust some random derivation you found in an article somewhere". Similar story about the incorrect Hydrogen radius expectation value integrating factor in Baym. Even text books aren't always correct.

I'd be happy to see a mainstream economic theory that hasn't been debunked by data...

True, but in our current regime in the US, it seems that both sides assiduously ignore technocrats in nearly every field. I think this is because the magnitude of our problems has been calling for large scale change in several related fields for some time.

Of course, the trouble is that the technocrats are supposed to tell the government how to do something, but we seem to have fixated on actually doing nothing while looking busy. Our government would rather not change, in spite of all external evidence that we should.

Over decades, research and a development of better economic theories should be pursued.

But if you think a better understanding of economics will solve all our ills w/r/t scarcity, income, and wealth inequality, you're way delusional.

Solving for scarcity, and the only reason to undertake the effort - the well being of a society - is a political and social exercise, not an economic one.

Quantum? Let's start with using complex adaptive systems thinking.

"So will the heterodox become the new orthodoxy, and economics go quantum?"

So did Thomas Friedman change his name?

Also, to all people in economics, or rather social sciences: stop with the dinky physics comparisons.

IMO the economy is much more like a biological system than a physical one. Systems that are studied in physics are neat and simple, with components that obey precise laws. Systems that are studied in biology are irreducibly complex, and often the laws that govern them are unknown.

First-principles thought experiments can work beautifully in astrophysics, but they're bound to fail in anatomy. So it goes with economics.

Isn't behavior economics making headway into the rethinking about economics?

> No one knows how this collapse occurs, but a conscious observer is usually assumed to be involved

No it's not. Only fringe interpretations like https://en.wikipedia.org/wiki/Von_Neumann%E2%80%93Wigner_int...

involve conciousness at all. And i'ts not strange that people don't like the idea of undefined, possibly non-existing stuff (conciousness) to be involved in physics. It's like defining laws of motion basing on "walking". Walking is what some configurations of atoms do, you don't put special cases in your theory for these things.


> A poll was conducted at a quantum mechanics conference in 2011 using 33 participants (including physicists, mathematicians, and philosophers). Researchers found that 6% of participants (2 of the 33) indicated that they believed the observer "plays a distinguished physical role (e.g., wave-function collapse by consciousness)". They also mention that "Popular accounts have sometimes suggested that the Copenhagen interpretation attributes such a role to consciousness. In our view, this is to misunderstand the Copenhagen interpretation."


Dismissing the entire field of economics requires a level of arrogance that can only be explained by gross ignorance. You really need to provide evidence to make this claim otherwise it is bizarre academic partisanship.

Economics at its core is the study of the problem of scarcity and how that is solved. If you think that's equivalent to alchemy but don't explain why, then I have nothing for you.

Additionally, simulation and games are interesting areas that do get studied by economists.

The world doesn't need more complex economics. It doesn't need more economics period.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact