Full CBO Score of Revised Shuster FAA Bill Still Can’t Give A Full Picture
September 25, 2017
This morning, the nonpartisan Congressional Budget Office released a full, 24-page cost estimate of the revised version of the House’s Federal Aviation Administration reauthorization bill (H.R. 2997) that was posted on the website of the House Rules Committee six weeks ago. The “Rules Print” of the bill contained the excise tax changes from the chairman of the Ways and Means Committee and discretionary spending cap changes from the chairman of the Budget Committee, as well as other changes – which meant that the cost estimate of the version of the bill approved by the Transportation and Infrastructure Committee on June 27 was no longer remotely accurate.
If one extrapolates CBO’s methodology, the cumulative 10-year deficit impact of the bill could drop from the $99 billion deficit increase projected last month to a more manageable $11 billion. But inconsistencies in budget scorekeeping laws and rules don’t allow CBO to connect the dots.
CBO issued a basic “PAYGO score” of the Rules Print on August 16 – a simple table that added up their estimates changes in mandatory spending and in revenue collections made by the bill. But that earlier score did not take into account proposed changes in discretionary spending, nor did it have any explanatory detail. The new, full-fledged “cost estimate” of the Rules Print includes all of the above – but, because of the binary nature of the budget scoring system, CBO is unable to give a full accounting of potential discretionary spending changes.
Two scoring systems, two baselines. The FAA bill, pushed by T&I chairman Bill Shuster (R-PA), would split up the FAA and turn the air traffic control system over to a new, non-profit, non-governmental corporation. Shuster and other supporters believe that once the transition takes effect in 2021, the cash flow of the new “AANS Corporation” should not be recorded as part of the federal budget. CBO disagrees and records the corporation’s spending as as federal mandatory spending and its user fee receipts as federal revenues. But that’s not the big problem, scorekeeping-wise.
The big problem is that most FAA spending is currently classified as discretionary, not mandatory, and the budget process is not set up to deal with bills that move programs from discretionary to mandatory, or vice versa. There are two separate scorekeeping systems – one for mandatory spending and revenues, and the other for discretionary spending – and they work very differently.
The scoring system for mandatory spending and revenues (PAYGO) measures proposed changes in law versus a baseline – for spending, that means either last year’s spending level projected forward with bump-ups for inflation and (for entitlements) demographic and underlying cost changes, and for taxes, that means economic modeling for how much taxable revenue there will be in the economy and what the scheduled rates are for those years.
The new CBO cost estimate, like the earlier PAYGO score, says that the total 10-year cumulative result of the mandatory spending and revenue changes in the Rules Print of H.R. 2997, measured on their own, will be a deficit increase of $98.5 billion.
But by taking air traffic control spending away from the FAA and giving it to a corporation, the bill also relieves Congress of the need to make annual appropriations for air traffic control after 2020. Air traffic control spending is currently part of the discretionary side of the budget. Legally, “discretionary” means that the money is appropriated in annual appropriations bills and that Congress has the discretion to provide whatever funding level it wants – all the way down to zero, if necessary – without exposing the federal government to legal liability.
Practically speaking, of course, it is impossible for Congress to just stop appropriating money for air traffic control or many other ongoing programs without having a replacement funding source already lined up – but from a budget perspective, they are free to do so, because the budget process is all about at what point a legal “obligation” occurs. (If an appropriation is made to meet a previously incurred obligation, that appropriation is ipso facto mandatory, not discretionary, or else the appropriation is a ministerial function that does not score.)
Accordingly, the discretionary authorization and appropriation scoring works from a “zero base.” Current law as of this date does not have any authorized funding for air traffic control after the start of fiscal year 2018 on October 1, 2017. So the authorizations in H.R. 2997 are labeled in Table 3 of the new cost estimate as being “increases in spending subject to appropriation” because they are increases above zero.
Just to be clear and reiterate this point – for purposes of the scoring of discretionary authorizations like air traffic control, even a freeze at the level appropriated in 2017 would be considered a massive spending increase in 2018, because the expiring authorization in a zero base environment makes even one dollar of spending on air traffic control a spending increase. 1 > 0. (CBO does maintain an inflation-adjusted baseline for future discretionary appropriations, but it is not currently used in any scorekeeping/enforcement system.)
The difference between the two scoring systems means that CBO is still unable to put one comprehensive dollar amount on the FAA bill. If enacted, Congress certainly won’t pay twice each year for the same air traffic control services (which the earlier PAYGO score seemed to imply) – but the question is, will the Appropriations Committee spend that $12+ billion per year that is no longer needed for air traffic control on something else? Or will the savings be applied to deficit reduction?
This is the $90 billion question – how to put a firm dollar amount on the savings on the discretionary side of the equation to counteract the increased costs on the mandatory/revenue side of the equation. This paragraph from the new cost estimate explains CBO’s quandary:
CBO expects that shifting ATC-related spending to the AANS Corporation would not materially change the magnitude of spending related to air traffic control if the total amount of discretionary appropriations provided in the future were reduced accordingly to reflect the shift—from discretionary to mandatory—of such spending. Indeed, compared to the amounts authorized over the 2018-2020 period when the FAA would continue to operate all aspects of existing aviation programs, H.R. 2997 would authorize far less funding over the 2021-2023 period for the agency to meet its residual responsibilities under the bill. Those reduced authorizations would be far less than CBO’s baseline projections of discretionary funding for the FAA over the 2021-2023 period, which incorporate the assumption that future appropriations for the agency will remain at the 2017 level, adjusted for inflation…
(That focus on the 2021-2023 period brings up another problem: CBO is required to do PAYGO scorekeeping for changes in mandatory spending and revenues over a 10-year period, but only scores legislation that would authorize discretionary appropriations over the length of the authorization. H.R. 2997’s authorizations run out after FY 2023. So the new cost estimate is lacking on any information on the discretionary side after that.)
The new cost estimate comes close to an apples-to-apples comparison only in FY 2021. Section 643 of the Rules Print of H.R. 2997 lowers the Budget Control Act’s cap on non-defense discretionary appropriations for FY 2021 by the amount that Congress winds up appropriating for air traffic control in FY 2020 (estimated by CBO to be $11.7 billion). This is included in a memo line in Table 1 of the new cost estimate to allow a direct comparison with the other totals.
But the Budget Control Act spending caps expire after FY 2021. Everyone anticipates that they will be extended by Congress, but that hasn’t happened yet, and CBO can’t score changes to numbers that do not yet exist.
The expiration of the caps means that there is no guarantee that the Appropriations Committees won’t take the $12+ billion per year in savings from no-longer-needed air traffic control appropriations and spend it on other things starting in FY 2022.
But if one assumes cap extensions and the reduction of those cap levels by the amount of air traffic control spending, then one can use the numbers in Table 5 of the new cost estimate for FY 2022 and FY 2023 and get $24.3 billion in savings (budget authority). Then one can download the big spreadsheet on the CBO website from the June baseline and get the comparable totals for FY 2024-2027, plug in inflation-adjusted guesstimates for the expenses of the rump safety-oriented FAA still subject to appropriation, and get savings of around $51.7 billion.
The $11.7 billion in more-or-less guaranteed discretionary savings in FY 2021 from the spending cap adjustment, the $24.3 billion in reduced air traffic control appropriation needs in FY 2022-2023 in Table 5 of the cost estimate, and the extrapolation of the Table 5 methodology for FY 2024-2027 ($51.7 billion) add up to a total potential discretionary savings of $87.7 billion over the 10-year period. A discretionary savings of $87.7 billion applied against a mandatory/revenue cumulative deficit increase of $98.5 billion over the same period gives a combined deficit increase of $10.8 billion, which could conceivably be fixable by tweaking the percentage by which existing excise taxes on aviation are being reduced by the bill.
But the expiration of the Budget Control Act spending caps after 2021, and the other factors mentioned above, mean that CBO can’t make those extrapolations on its own.
The table below walks the reader through the torturous extrapolations explained above.
|Summary of Congressional Budget Office Cost Estimate of the Rules Committee Print of H.R. 2997, FAA Reauthorization|
|(Full cost estimate dated September 25, 2017. Billions of dollars. “BA” = Budget Authority.)|
|New AANS Corp. Outlays||0.0||0.0||0.0||9.3||11.6||12.7||13.6||14.1||14.5||15.0||90.7|
|New AANS Corp. Fee Receipts||0.0||0.0||0.0||-12.2||-12.6||-13.0||-13.5||-14.0||-14.5||-15.0||-94.6|
|Subtotal, AANS Corp.||0.0||0.0||0.0||-2.9||-1.0||-0.3||0.1||0.1||0.0||0.0||-3.9|
|Net Decrease in Tax Receipts||0.0||0.0||0.0||12.4||13.7||14.2||14.8||15.2||15.8||16.4||102.4|
|Deficit Increase from Mandatory Spending and Revenue Changes||0.0||0.0||0.0||+9.5||+12.7||+13.9||+14.9||+15.3||+15.8||+16.4||+98.5|
|However, a memo line in Table 1 of the full cost estimate shows how discretionary appropriations would be reduced from the bill’s reduction in the Budget Control Act discretionary spending cap in FY 2021:|
|FY21 Cap Reduction – BA||-11.7||-11.7|
|FY21 Cap Reduction – Outlays||-6.1||-3.5||-1.2||-0.3||-0.3||-11.5|
|CBO can’t assume savings from any Budget Control Act cap reductions after 2021 – because there are no spending caps after 2021 (yet). Table 5 of the cost estimate does compare HR 2997 discretionary authorization levels to the CBO baseline, but only through 2023 (we don’t show 2021 because it is referenced above):|
|CBO Discretionary BA Baseline||14.7||15.0|
|HR 2997 Discretionary Authorization||2.7||2.7|
|Then, one can add the FAA discretionary baseline totals for FY 2024-2027 from the spreadsheet on the CBO website and use guesstimates for the Shuster bill discretionary authorizations to compare post-2023:|
|CBO Discretionary BA Baseline||15.3||15.6||16.0||16.3|
|HR 2997 Discretionary Authorization (guesstimate)||2.8||2.8||2.9||3.0|
|If you add all the opportunities for reduced discretionary spending (the $11.7 billion in FY21 cap reduction, the $24.3 billion in FY22-23 reductions in Table 5 of the cost estimate, and the $51.7 billion in FY24-27 reductions using extrapolated CBO methodology, the total (as applied against the $98.5 billion mandatory/revenue total) looks like this:|
|10-Year Deficit Increase from Mandatory Spending and Revenue Changes vs. Baseline||+98.5|
|Scoreable (FY21) and Non-Scoreable (FY22-27) Opportunities for Reduced Discretionary BA||-87.7|