What are its key elements? Isn’t any structural model doing a counterfactual doing so in partial equilibrium?
Partial equilibrium

The way linear regression coefficients are interpreted is ceteris paribus. Meaning, they are interpreted as how much x changes y in partial equilibrium holding all else constant.
General equilibrium effects (e.g a change in x which changes w which changes v which changes x more is not considered)

The way linear regression coefficients are interpreted is ceteris paribus. Meaning, they are interpreted as how much x changes y in partial equilibrium holding all else constant.
General equilibrium effects (e.g a change in x which changes w which changes v which changes x more is not considered)
Thanks. So it’s basically a distinction between what my analysis holds fixed versus moving for a given exercise e.g., counterfactual policy analysis?

See undergraduate Econ curriculum.
The way linear regression coefficients are interpreted is ceteris paribus. Meaning, they are interpreted as how much x changes y in partial equilibrium holding all else constant.
General equilibrium effects (e.g a change in x which changes w which changes v which changes x more is not considered)
Thanks. So it’s basically a distinction between what my analysis holds fixed versus moving for a given exercise e.g., counterfactual policy analysis?

The way linear regression coefficients are interpreted is ceteris paribus. Meaning, they are interpreted as how much x changes y in partial equilibrium holding all else constant.
Linear regressions are statistical models not economic models.
Probably models are economic models. See http://fitelson.org/woodward/haavelmo.pdf or Wikipedia for Haavelmo’s Nobel prize.
The interpretation of a “loglog” regression coefficient is that it is an elasticity. As you know, for cross price elasticity how much a %change quantity in x has a %change quantity change in y assumes ceteris paribus.
It is true that empirical discussions of general vs partial equilibrium sounds different than in economic theory (one market clearing vs all markets clearing together). But they are really the same. If you hear an applied economist saying “general equilibrium effects” or “spillover effects” they mean not assuming ceteris paribus.

The way linear regression coefficients are interpreted is ceteris paribus. Meaning, they are interpreted as how much x changes y in partial equilibrium holding all else constant.
Wait, no. There are no equilibrium effects of any kind, not even partial equilibrium effects, in most regression analyses.

The way linear regression coefficients are interpreted is ceteris paribus. Meaning, they are interpreted as how much x changes y in partial equilibrium holding all else constant.
Wait, no. There are no equilibrium effects of any kind, not even partial equilibrium effects, in most regression analyses.
Usually one assumes stationary distributions which is an equilibrium assumption.

The way linear regression coefficients are interpreted is ceteris paribus. Meaning, they are interpreted as how much x changes y in partial equilibrium holding all else constant.
Linear regressions are statistical models not economic models.
Probably models are economic models. See http://fitelson.org/woodward/haavelmo.pdf or Wikipedia for Haavelmo’s Nobel prize.
The interpretation of a “loglog” regression coefficient is that it is an elasticity. As you know, for cross price elasticity how much a %change quantity in x has a %change quantity change in y assumes ceteris paribus.
It is true that empirical discussions of general vs partial equilibrium sounds different than in economic theory (one market clearing vs all markets clearing together). But they are really the same. If you hear an applied economist saying “general equilibrium effects” or “spillover effects” they mean not assuming ceteris paribus.NO, there may not exist any representation of tastes, tecnology, and equilibrium conditions that correspond to a regression that a researcher runs.