Sam Passey

Becoming successful is not luck. It’s math. If your probability of success is 1/100 and you try 100 times, you have a 100% chance of success.

Becoming successful is not luck. It’s math. If your probability of success is 1/100 and you try 100 times, you have a 100% chance of success.

I will never get tired of pointing out that Democrats have attempted many times in both congress and the Supreme Court to enact a national ban on partisan gerrymandering, literally all Republicans need to do on this is say "yes."

This might be unacceptable to say to a lot of modern applied economists, but I don’t really care about model misspecification too much. In my view, all models are misspecified, but some mechanisms are useful. I think economic theory is useful because it disciplines mechanisms, and structural models are useful because they provide clear illustrations about stories concerning mechanisms and their relative sizes. I do not believe for a second the size of an estimate in a structural econometric model as being true. I just care about if it helps to discipline and illustrate and validate the economic foundation & narrative. I write a microfounded model because it gives an economic story about the assumptions needed to estimate and identify credibly using it instead of having to just rely on statistical assumptions. I think many actually seek to get effective sizes using structural models, which I view as nothing but a nice secondary effect at best. I care about validating stories so one can construct more future work whether it be using bigger data & more model-free methods, richer structural models, or more complicated theories. This is why I like a lot of Acemoglu’s macroeconomic work. He writes a rich theoretical model then independently tests the mechanisms using the best reduced form methods, using economics to justify the assumptions. I think that is one of the best ways to do empirical economics.

This might be unacceptable to say to a lot of modern applied economists, but I don’t really care about model misspecification too much. In my view, all models are misspecified, but some mechanisms are useful. I think economic theory is useful because it disciplines mechanisms, and structural models are useful because they provide clear illustrations about stories concerning mechanisms and their relative sizes. I do not believe for a second the size of an estimate in a structural econometric model as being true. I just care about if it helps to discipline and illustrate and validate the economic foundation & narrative. I write a microfounded model because it gives an economic story about the assumptions needed to estimate and identify credibly using it instead of having to just rely on statistical assumptions. I think many actually seek to get effective sizes using structural models, which I view as nothing but a nice secondary effect at best. I care about validating stories so one can construct more future work whether it be using bigger data & more model-free methods, richer structural models, or more complicated theories. This is why I like a lot of Acemoglu’s macroeconomic work. He writes a rich theoretical model then independently tests the mechanisms using the best reduced form methods, using economics to justify the assumptions. I think that is one of the best ways to do empirical economics.

🚨 JD VANCE JUST NOW: "I am married to the daughter of immigrants from India, and I love my in-laws. They're great people and have been great contributors to the USA." "You can believe that there's a lot of H-1B fraud, while also believing people have come to the US who have enriched this country." "One of the responsibilities we MUST expect of citizens...you have to think about the best interests of the country, NOT the country you came from beforehand."