4 ms·

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

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?

duxut_staglatz 7y ago
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.
Mikeb85 7y ago
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.
tomrod 7y ago
> 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 https://en.wikipedia.org/wiki/Lucas_critique, Theory before business-cycle measurement: http://faculty.wcas.northwestern.edu/~lchrist/papers/qr1042.pdf http://faculty.wcas.northwestern.edu/~lchrist/papers/qr1042.... [1] To describe or predict: https://www.stat.berkeley.edu/~aldous/157/Papers/shmueli.pdf 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_laureates_in_Economics https://en.wikipedia.org/wiki/List_of_Nobel_Memorial_Prize_l...

duxut_staglatz 7y ago

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.

Mikeb85 7y ago

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.

tomrod 7y ago

> 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 https://en.wikipedia.org/wiki/Lucas_critique, Theory before business-cycle measurement: http://faculty.wcas.northwestern.edu/~lchrist/papers/qr1042.pdf http://faculty.wcas.northwestern.edu/~lchrist/papers/qr1042.... [1] To describe or predict: https://www.stat.berkeley.edu/~aldous/157/Papers/shmueli.pdf 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_laureates_in_Economics https://en.wikipedia.org/wiki/List_of_Nobel_Memorial_Prize_l...