search

PSLmodels / Behavioral-Responses

PSL module that estimates partial-equilibrium behavioral responses to tax changes simulated by Tax-Calculator
Other
4 stars 14 forks source link

Should CG income be included in the base for calculating ETI substitution effect? #36

Open MattHJensen opened 5 years ago

MattHJensen commented 5 years ago

Currently the substitution effect is calculated using taxable income (c04800) as the base. See https://github.com/PSLmodels/Behavioral-Responses/blob/master/behresp/behavior.py#L199.

If one were to apply both a non-zero substitution elasticity and a non-zero long-term capital gains elasticity, does it mean that we are double counting some behavior?

And if so, would a reasonable solution be to remove LTCG from the substitution effect by subtracting it from c04800 prior to the calculation of the substitution effect?

If one were to model short-term capital gains separately, would STCG also need to be subtracted from c04800 prior to the calculation of the substitution effect?

--

This is related to https://github.com/PSLmodels/Tax-Calculator/issues/2189, so cc @andersonfrailey @codykallen @maxghenis.

This is related to https://github.com/PSLmodels/Tax-Calculator/issues/2180, so cc @evtedeschi3 @martinholmer.

Given Dan's familiarity with the ETI literature, cc @feenberg.

martinholmer commented 5 years ago

The resolution of issue #36 will be handled after version 0.6.0 is released.

martinholmer commented 5 years ago

@MattHJensen, Can you repeat here the reason you mentioned in our conversation about why removing long-term capital gains from the taxable income base used to compute the substitution effect is desirable.

martinholmer commented 5 years ago

@MattHJensen, after our recent phone conversation, I reread issue #36. Your original comment in that issue says nothing about what I understood you to say on the phone. My understanding of the motivation is that somebody specified a reform that change the top bracket rate only for regular income, but not for long-term capital gains or for qualified dividends. And then that person was surprised to see a behavioral response (using the response function) for filing units that had only long-term capital gain or qualified dividend income. If you view that as a problem that needs to be fixed, then (to me) it suggests a couple of important things about your approach:

  1. Both long-term capital gains and qualified dividends need to be netted out of taxable income.

  2. Doing the netting out should be an option (not hardwired into the response function's logic) because many reforms change the rates on both kinds of incomes.

  3. The implications for the income effect also need to be considered.

An alternative way to handle this might be to hardwire the response function to compute the substitution effect as the sum of two terms: a response on regular taxable income (using the MTR on labor income as now) and a response (using MTRs on dividend income) on long-term capital gain plus qualified dividend income. If you could make that approach work, you could hardwire the logic and obviate a need for a confusing option. The income effect could be computed as the sum of two terms computed using the change in taxable income caused by labor-income tax effects and capital-income tax effects.