TCJA extension reform

[bodiyang]

@martinholmer @jdebacker

Issue found from newly added ext.json reform file, related to PR 2734

CBO has published a new document, estimation of the revenue change from TCJA extension reform

Please see table 2 (in page 5)

I have tried to run the simulation of TCJA extension reform in Tax-Calculator (by using ext.json file) and compare the result with CBO estimation. There are differences noticed.

This attached screenshot is the result from Tax-Calculator

Revenue change from TCJA extension (ext.json reform) for 2025 - 2033 calculated by Tax-Calculator (by PUF) is around 200-250 billion per year, while CBO estimation is around 350-450 billion per year.

[bodiyang]

This CBO breakdown analysis is also helpful: the supplemental data

[jdebacker]

Page 4 of the CBO document on the budget under alternative assumptions, says:

According to JCT’s estimates, if the expiring individual income tax provisions of the 2017 tax act were extended, deficits would be larger than those in CBO’s baseline, on net, by $2.5 trillion over the 2024–2033 period, excluding debt-service costs (see Table 2). Most of the effects would occur after 2026. Debt-service costs would add $278 billion to those deficits.

Using Tax-Calculator, I find the revenue effect from 2024-2033 to be $2.0 trillion. This is about 20% less than the JCT estimate, but I don't know how off model items and/or behavior may impact these numbers.

It appears CBOs much larger cost estimate include the significant increase in net interests costs that would result with extending the TCJA cuts. Those are not accounted for in Tax-Calculator, which just focuses on direct impacts on tax revenues.

[bodiyang]

The new May 2023 publication page 4 According to JCT’s estimates, if the expiring individual income tax provisions of the 2017 tax act were extended, deficits would be larger than those in CBO’s baseline, on net, by $2.5 trillion over the 2024–2033 period, excluding debt-service costs (see Table 2).

The May 2024 publication page 4 According to JCT’s estimates, if the expiring individual income tax provisions of the 2017 tax act were extended, primary deficits over the 2025–2034 period would be $3.3 trillion larger, on net, than those in CBO’s baseline projections (see Table 2).

[bodiyang]

@jdebacker Thanks Jason

I just another check. Actually, the net interest outlays / Debt-service costs (May 2023 report & May 2024 report) is separately calculated, apart from the revenue change.

Please see the two rows here in May 2024 report

The the calculation of revenue change for the year 2024-2033 Tax-Calc: $2.0 trillion CBO May 2024 report: $ 2.8 trillion CBO May 2023 report: $ 2.5 trillion

Then for the year 2034, Tax-Calc: $273 billion CBO 2024 report: $434 billion

@martinholmer

Note: Tax-Calc's calculation by year

[martinholmer]

Neither @bodiyang nor @jdebacker disclosed in issue #2748 which input dataset they used: is it puf.csv or the new tmd.csv? And they were unclear about what "revenue change" means: is that just income taxes or is it the sum of payroll taxes plus income taxes?

[bodiyang]

@martinholmer thanks Martin

1. The result is based on puf.csv data
2. Revenue is just the income tax revenue

(I think payroll tax actually have no change from the ext.json reform)

[bodiyang]

I have completed some simulations of the reforms on each single parameter change, in reference to this CBO spreadsheet

Please check this excel as a reference TCJA_item_ext.xlsx

Looking at the result, these reforms are close to CBO's estimation: Individual income rates & brackets, Standard Deduction, CTC, Itemized Deduction

While, the following reforms are far away from CBO's estimation: Business Income Deduction, AMT, personal exemption

[bodiyang]

I have completed some simulations of the reforms on each single parameter change, in reference to this CBO spreadsheet

Please check this excel as a reference TCJA_item_ext.xlsx

Looking at the result, these reforms are close to CBO's estimation: Individual income rates & brackets, Standard Deduction, CTC, Itemized Deduction

While, the following reforms are far away from CBO's estimation: Business Income Deduction, AMT, personal exemption

Please ignore this comment, this is the wrong simulation.

Please check my next comment

[bodiyang]

I have completed some simulations of the reforms on each single parameter change, in reference to this CBO spreadsheet

Please check this excel as a reference update comparison.xlsx

Looks like apart from QBID, most per parameter change reform looks good.