* Corrected on July 30, 2021 *
Key Points
PWBM estimates that, over the 2022-2031 budget window, President Biden’s FY2022 budget proposals would require $5.9 trillion in new spending and would raise $3.9 trillion in new revenues on a conventional basis (without accounting for macroeconomic feedback).
Including macroeconomic feedback effects, we project a 0.2 percent increase in public debt in 2031 but a 7.3 percent decrease in public debt in 2050, relative to current law.
Despite the long-run decrease in debt and increased productivity from new public investments, we project that the capital stock would fall by 3.6 percent in 2050 and GDP would fall by 1.1 percent in 2050, relative to current law.
President Biden’s FY2022 Budget Proposal: Budgetary and Economic Effects
President Biden’s Fiscal Year 2022 budget proposal adds many details to previously announced policies and combines them into one proposal. PWBM examined several of these policies in our previous analyses of the American Families Plan and American Jobs Plan. Building on our 2020 analysis of the Biden campaign platform, this brief provides comprehensive and updated analysis of the President’s budget proposals, linking various policy proposals together under PWBM’s integrated modeling framework that also incorporates potential interactions between these policies.
Table 1 presents top-line conventional revenue estimates for the President’s budget proposal. In total, the President’s budget would raise $3.92 trillion in new revenues on a conventional basis over the 2022-2031 budget window.
| Category | Provision | Revenue Raised, 2022 - 2031 |
|---|---|---|
| American Jobs Plan | Corporate rate to 28 percent | 892 |
| Revise GILTI, disallow deductions attributable to exempt income, and limit inversions | 728 | |
| Repeal FDII deduction | 260 | |
| 15% minimum book tax | 184 | |
| Replace BEAT with the SHIELD Rule | 390 | |
| Eliminate fossil fuel preferences | 33 | |
| Reform taxation of foreign fossil fuel income | 73 | |
| American Families Plan | Top rate to 39.6 percent | 111 |
| Capital gains provisions, tax carried interest as ordinary income | 376 | |
| Repeal deferral of gain from like-kind exchanges | 41 | |
| Rationalize Net Investment Income and SECA | 189 | |
| Make permanent excess business loss limitation of noncorporate taxpayers | 162 | |
| Tax compliance and administration (Revenue) | 480 | |
| Total | 3,919 | |
Most revenue provisions and their accompanying estimates have not changed since our previous analyses of the President’s American Families Plan and American Jobs Plan. Notable exceptions are the addition of “changes to the taxation of foreign fossil fuel income,” which we estimate would raise $73 billion over the budget window as well as replacing Base Erosion Anti-Abuse Tax (BEAT) with Stopping Harmful Inversions and Ending Low-Tax Developments (SHIELD), which raises $390 billion over the budget window. PWBM will release a more detailed analysis of SHIELD and the rest of the President’s international tax proposals in the near future.
Table 2 presents top-line spending numbers for the President’s budget proposal, by broad spending category. In total, we estimate that the President’s budget would increase spending by $5.89 trillion over the 2022-2031 budget window, including $393 billion in discretionary spending not included in the American Families Plan or American Jobs Plan.
| Spending Category | New Spending, 2022 - 2031 | |
|---|---|---|
| American Jobs Plan | Infrastructure | 1,031 |
| R&D and Manufacturing* | 1,133 | |
| Workforce Development | 87 | |
| Housing | 234 | |
| Long-Term Care | 400 | |
| American Families Plan | Education | 854 |
| Individual Tax Credits | 853 | |
| Child Care | 550 | |
| Paid Leave, Nutrition, Maternal Health | 273 | |
| Additional IRS Funding Outlays | 80 | |
| Other | Discretionary Spending | 393 |
| Total | 5,888 | |
*Note: Includes additional research spending from FDII repeal revenues (see appendix table)
For most spending provisions, the President has proposed a specific dollar amount of new outlays, for example, in spending $400 billion on long-term care. For some provisions of the American Families Plan, however, the President’s budget also lays out specific goals for the spending, such as guaranteeing “2 years of universal high-quality pre-school for every 3- and 4-year old in America, and adding 2 years of free community college.” For those provisions, we provide our own conventional cost estimates.
Specifically we estimate that, over the budget window, universal pre-K would cost $382 billion while tuition-free community college would cost $299 billion.1 Additionally, we estimate that the individual tax credit expansions in the American Families Plan would cost $853 billion over the budget window, compared to the President’s budget estimate of $798 billion. Line-by-line spending and revenue estimates are available in the appendix table below.
In the PWBM model, the spending provisions in the President’s budget have one of three macroeconomic effects:
- Public investments in infrastructure and R&D boost aggregate productivity;
- Investments in education and childcare boost the labor productivity of the specific recipients of the spending;
- Transfers to specific agents change labor supply decisions via income and substitution effects.
Investments in public capital such as transit infrastructure boost the productivity of private capital and labor. However, these investments have a build-out delay, consistent with a 2016 Congressional Budget Office report, and the investments eventually depreciate over time. Our analysis of public investments follows the same approach as in our recent analysis of the bipartisan Senate infrastructure deal.2 For the current analysis, we assume that these investments all occur in the 2022-2031 budget window, with no additional spending beyond 2031.
Budget provisions that allocate spending toward education, childcare, and paid leave boost labor productivity. Universal preschool increases future productivity but gains to GDP do not show up until those cohorts enter the labor force beginning in 2039. Consistent with our previous work on COVID-19 related school closures, education proposals boost productivity to the specific demographic groups receiving the benefit. Productivity boosts are, for example, only applied to new public preschool enrollees who would not have been enrolled in a formal preschool program under current law. Some households switch from paying for private preschool to the public option, so those households face lower costs today, yet the future labor productivity of children in those households remains the same. Education provisions of the President’s budget (such as two years of tuition-free community college) are modeled as permanent, lasting beyond 2031. New childcare spending increases parental labor force participation in some households, but these households tend to be a small portion of lower-income and lower-productivity households, so aggregate productivity effects are small.
Transfer payments that substitute for existing spending (for instance, preschool and eldercare) increase the household budget and can disincentivize work via income effects. On the other hand, transfer payments that allow a switch from domestic production (e.g., childcare provided by a working-age member of the household) to market production increase output, but only if there are scaling and efficiency effects. That is, if the new, government-paid caregiver was working at another job comparable to the job taken by the parent who is now able to work in the market, then no net benefit has been realized to the economy as a whole—only a switch in jobs.
Spending earmarked toward affordable housing and home- and community-based care primarily has transfer effects. Affordable housing spending generally benefits households with below-median incomes, while caregiving provisions—mainly implemented through Medicaid—generally benefit older and Medicaid-eligible households. Transfers to working-age households let those individuals work less without consuming less, reducing overall labor supply, thereby creating a downward pull on GDP.
Table 3 presents estimated effects of the President’s budget proposal on various macroeconomic aggregates.
| Year | GDP | Capital Stock | Hourly Wage | Hours Worked | Government Debt |
|---|---|---|---|---|---|
| 2031 | -1.0 | -2.8 | -0.6 | -0.4 | 0.2 |
| 2040 | -1.0 | -3.4 | -0.7 | -0.4 | -4.0 |
| 2050 | -1.1 | -3.6 | -0.7 | -0.4 | -7.3 |
The revenue raised from increasing taxes on high-income households and reducing the tax gap reduces government debt by 7.3 percent in 2050 relative to baseline. However, higher taxes on capital gains and dividends facing the wealthiest Americans would lower their after-tax return on equity investment and therefore disincentive saving. The increase in the top individual tax rate also discourages labor by high-productivity households. Higher corporate taxes, higher taxation on foreign-sourced income, and reduction in deductions further lower the after-tax return on equity investment.
Debt reduction increases the portion of savings allocated into private capital, as discussed in our previous explainer on capital crowd out. And the current regime of accelerated depreciation means that, in isolation, increasing the corporate tax is projected to increase investment as in PWBM’s recent corporate tax estimates. However, as discussed in our previous AJP analysis, the President’s proposed minimum book tax reduces the value of depreciation deductions, disincentivizing investment. The net effect is a reduction in the capital stock of 3.6 percent relative to baseline in 2050. Lower capital reduces the wage rate businesses are willing to pay by 0.7 percent in 2050. Hours worked, therefore, decline by 0.4 percent in 2050 as workers choose to work less given lower wages. Overall, despite the boost to productivity from public investments, GDP is 1.1 percent lower than baseline in 2050.
The table below expands on Tables 1 and 2 above, showing line-by-line spending and revenue estimates for the full President’s budget proposal.
| Category | Provision | Additional Spending (+) or Revenue (-), 2022 - 2031 |
|---|---|---|
| AJP Spending | Transportation infrastructure | 596 |
| Broadband, electrical, water | 309 | |
| Housing - Clean energy accelerator | 27 | |
| Other Housing | 185 | |
| School buildings | 87 | |
| Federal hospitals/buildings | 27 | |
| Long-term care | 400 | |
| Domestic manufacturing | 299 | |
| Workforce development | 87 | |
| R&D spending | 180 | |
| Additional R&D from FDII repeal* | 260 | |
| AJP Tax Credits | Clean energy tax breaks | 367 |
| Housing tax credits | 49 | |
| Federally subsidized state and local bonds for infrastructure | 12 | |
| AJP Revenue Raisers | Corporate rate to 28 percent | -892 |
| Revise GILTI, disallow deductions attributable to exempt income, and limit inversions | -728 | |
| Repeal FDII deduction | -260 | |
| 15% minimum book tax | -184 | |
| Eliminate fossil fuel preferences | -33 | |
| Replace BEAT with the SHIELD Rule | -390 | |
| Reform taxation of foreign fossil fuel income | -73 | |
| AFP Spending | Free universal pre-K for 3- and 4-year olds** | 382 |
| Tuition-free 2-year community college** | 299 | |
| Increase Pell grants and other education spending | 173 | |
| New child care program** | 550 | |
| Paid family and medical leave | 225 | |
| Nutrition and maternal health | 48 | |
| Increase IRS funding (outlays) | 80 | |
| AFP Tax Credits | Extend ARP CTC expansion + make CTC fully refundable | 439 |
| Extend ARP ACA premium tax credit expansions | 197 | |
| Make ARP EITC expansion permanent | 125 | |
| Make ARP CDCTC changes permanent | 92 | |
| AFP Revenue Raisers | Top rate to 39.6 percent | -111 |
| Capital gains provisions, tax carried interest as ordinary income | -376 | |
| Repeal deferral of gain from like-kind exchanges | -41 | |
| Rationalize Net Investment Income and SECA | -189 | |
| Make permanent excess business loss limitation of noncorporate taxpayers | -162 | |
| Tax compliance and administration (Revenue) | -480 | |
| Other | Discretionary Spending | 393 |
| Total | 1,969 | |
* Note: The President's budget presents a net zero number for revenue raised by FDII repeal, allocating all such revenue towards additional R&D spending. We therefore assume that the additional outlays are equal to our estimate of revenues raised from FDII repeal.
** Note: For these spending provisions, we estimate the additional spending required to meet the budget proposal's stated goals, rather than taking as given dollar amounts specified in the proposal
Federal programs that include required state and local government spending also require extra care in modeling. State-level spending financed by federal grants maps in a straightforward manner. The state budget functions as a pass-through for federal spending and therefore, that spending shows up on the federal budget. If, instead, the state government finances the program by borrowing on its own, then total government debt (federal plus state) should rise and affect the macroeconomy, but an economic model focused solely on the federal budget misses this effect. PWBM includes any net increase3 in the state share of federal programs in the federal government’s budget. This approach favorably assumes that states can borrow at the same interest rate as the federal government but this approach does not allow for a “free lunch” that would occur if the funds necessary for state spending was not accounted for at all.
Our previous analysis of the President’s budget included $637 billion in spending on universal preschool over the ten-year budget window. That analysis assumed that the federal government pays for all public preschool programs, including those already in place.
Our previous analysis of the President’s budget did not properly account for the existing state spending, which should have been excluded from the unified budget. Our corrected approach replaces the total program cost with the cost of the federal expansion of public preschool. We estimate this cost at $382 billion. Table 3-A reports our previous calculations for the sake of continuity while Table 3-B reports our updated calculations.
| Year | GDP | Capital Stock | Hourly Wage | Hours Worked | Government Debt |
|---|---|---|---|---|---|
| 2031 | -1.2 | -3.0 | -0.8 | -0.4 | 1.1 |
| 2040 | -1.3 | -3.8 | -0.9 | -0.4 | -2.4 |
| 2050 | -1.5 | -4.2 | -0.9 | -0.6 | -5.1 |
| Year | GDP | Capital Stock | Hourly Wage | Hours Worked | Government Debt |
|---|---|---|---|---|---|
| 2031 | -1.0 | -2.8 | -0.6 | -0.4 | 0.2 |
| 2040 | -1.0 | -3.4 | -0.7 | -0.4 | -4.0 |
| 2050 | -1.1 | -3.6 | -0.7 | -0.4 | -7.3 |
This analysis was conducted by Alexander Arnon, Marcos Dinerstein, Maddison Erbabian, John Ricco, and Youran Wu under the direction of Efraim Berkovich and Richard Prisinzano. Prepared for the website by Mariko Paulson.
-
These estimates represent the full cost of providing universal pre-K and tuition-free community college. They are therefore not directly comparable to the President’s budget estimates for those programs, which assume that states will cover 25 percent of the total costs. In our previous analysis of the American Jobs Plan, we assumed that state governments would cover 25 percent of the cost of universal pre-K. ↩
-
This framework is discussed in detail in a previous PWBM infrastructure explainer. ↩
-
Some programs, such as federal grants, may allow states to substitute spending within state budgets. We estimate net increases in state budgets in order to properly account for actual unified budget effects. ↩