> That's why this study was done by physicists who tend to not have that particular issue with not completely understanding the system and interactions ... That's the important takeaway: a simple model is unexpectedly able to do that.
Yes, I know. That's why my criticism was directed at the Tech Review article that totally misinterprets the paper, and indirectly at all of the comments in this thread that make the same mistake. The paper does not describe how gas-pump prices change; it does not describe how any actual market actually works. It does not identify time to react to price signals as the key to well-functioning markets; that is an artifact of the highly stylized model employed.
As I wrote in the first line of my comment, the take-away is that cartel-like behavior is possible without direct coordination, not that it is actually happening, ever has happened, or ever will happen. The rest of my comment was meant to point out why we can't draw those conclusions.
EDIT: And, by the way, there are lots of other ways to show that cartel-like behavior is possible without concerted price-fixing. Several authors have shown this in a game-theoretic framework, and there is a whole literature on indirect communication among economic competitors that has arisen out of law-and-economics analysis of anti-trust regulation.
I view the mention by Technology Review and the paper of gas-pump price changes as just giving an example and context that most readers will understand or at least believe they understand. If the assumption is that the cartel-like behavior is coordinated, then that a model that is clearly not coordinated exhibits similar qualitative behavior should give pause to that assumption.
I agree that it's not an actual explanation or a model of any actual market. How both the article and paper are worded is a bit ambiguous on that point most likely for purposes of attracting a bit of attention and grant money. But if you look closely, there's no outright claim in either.
As for seemingly ignoring existing literature in another field, it's kind of par for the course. A lot of "discoveries" made early on in complexity research on sandpiles were already well-documented in civil engineering.
Yes, I know. That's why my criticism was directed at the Tech Review article that totally misinterprets the paper, and indirectly at all of the comments in this thread that make the same mistake. The paper does not describe how gas-pump prices change; it does not describe how any actual market actually works. It does not identify time to react to price signals as the key to well-functioning markets; that is an artifact of the highly stylized model employed.
As I wrote in the first line of my comment, the take-away is that cartel-like behavior is possible without direct coordination, not that it is actually happening, ever has happened, or ever will happen. The rest of my comment was meant to point out why we can't draw those conclusions.
EDIT: And, by the way, there are lots of other ways to show that cartel-like behavior is possible without concerted price-fixing. Several authors have shown this in a game-theoretic framework, and there is a whole literature on indirect communication among economic competitors that has arisen out of law-and-economics analysis of anti-trust regulation.