Should GA surplus rebate (HB 112) apply to TY 2024 or TY 2023?

[PavelMakarchuk]

Observation

In our TY 2024 comparison, we see 30 non-S-Corp GA state tax mismatches where PE computes higher state tax than TAXSIM by ~$250/$375/$500. This appears to correspond to the Georgia surplus tax rebate.

What we found

After pulling the latest feenberg/taxsim source, the updated 11ga.for applies the HB 112 rebate via:

elseif(in(law,2023,2024)) then
   rebate=rebatef(rebate,law,2023,2024)

With the default extnd(27)=0, rebatef applies the rebate in the paid year (2024).

The question

HB 112 (O.C.G.A. § 48-7-20.3) defines the refund as:

"a one-time refund amount equal to the lesser of: (A) The qualified taxpayer's 2023 individual income tax liability... or (B) An amount, which is based on such taxpayer's filing status for the 2023 taxable year"

Since the rebate amount is determined entirely by TY 2023 liability and filing status, we've been modeling it as reducing TY 2023's effective tax burden rather than TY 2024's. The same pattern applies to HB 1302 (§ 48-7-20.1, TY 2020 liability) and HB 162 (§ 48-7-20.2, TY 2021 liability).

Is there a standard convention for which year these one-time surplus refunds should be assigned to? We could see arguments for either the liability year or the payout year. Curious how others handle this.

References

• HB 1302 (§ 48-7-20.1): enrolled text
• HB 162 (§ 48-7-20.2): enrolled text
• HB 112 (§ 48-7-20.3): enrolled text

[feenberg]

The rebate function is called from CO DE GA HI ID IL IN ME MA MT NM NY RI SC VA. We should treat all the rebates in the same way.

In every case I prefer assigning the rebate by default to the year paid, but I understand the argument for assigning it to the year it depends on (with the year paid income acting as a proxy for income in the year eligible). That is what extnd(27) allows for.

You can set extnd(27) by adding the variables

opt1=27 
opt2=1

to the csv file. There is also option 30, which if set to 1 will set option 27 to 1, and also change some other calculations to be closer to PE conventions. See https://taxsim.nber.org/taxsimtest/options.html

My argument for using the year paid, is that the most common use of the calculator is to calculate after-tax income, which implies the year paid. It is also the relevant year for revenue scoring. Lastly, the year eligible is long before the year announced, so the rebate can't influence behavior in the year eligible.

I can change my default, but I think it better that PE change its default.