Improve imputation of overtime income eligible for new OBBBA deduction

[martinholmer]

In the discussion of merged PR #397, @donboyd5 mentioned this issue:

3. Overtime pay premium -- see Q7 item 7A, and also Q8: In addition to the improvements I suggest above, a nice-to-have change would be to separate the overtime pay imputation into two steps: (1) estimate overtime pay using SIPP, (2) estimate the premium portion of overtime pay, which will be less, and make that the TMD variable - it probably makes sense to use what Claude believes is IRS guidance (1/3 of overtime pay -- e.g., the "half" portion if we assume overtime pay, on average, is time-and-a-half) to estimate the premium portion. Then, in the calibration step, you'd probably need to increase the scale factor from 2.6 to somewhere around 7.8, meaning in the end, we're right back to the values you have now. So this has no impact on results. However, it has the advantages that (1) it would make future changes easier and more intuitive if we get better targeting information in the future, and (2) it highlights just how much you have to adjust the data to be consistent with TPC estimates. This suggests to me that TPC may have substantially overestimated the impact of the overtime pay premium tax preference.

While all these points are sensible, at the current time there is a shortage of reliable estimates of the new overtime income deduction, which makes doing anything more that what is done in PR #397 speculative. This issue will be reassessed after better estimates of the deduction become available from JCT, IRS, or private tax-analysis groups.