Technical Notes for CYNICONOMICS Debt Projections

This post describes the methods and inputs that I used in two prior posts: “Previewing the Bad News That’s Likely to Complicate the Debt Ceiling Battle” and “The Chart That Every Taxpayer Deserves to See.”

For years from 2013 through 2023, the debt projections shown in the prior posts are derived from the CBO’s latest 10-year projections, which were published on May 14. Beyond 2023, the projections rely on the CBO’s past methodology for its 75-year projections, as described in its “2012 Long-Term Budget Outlook,” as well as other sources including the last two years’ Social Security and Medicare Trustees’ Reports and materials from the Center for Medicare and Medicaid Services.

Here are the key methods and inputs in greater detail:

2013 through 2023

  • The “baseline” and “alternative” scenarios are derived from a CYNICONOMICS model that takes its key inputs from the CBO’s “Updated Budget Projections: Fiscal Years 2013 to 2023” and produces debt projections that are all but identical to the CBO’s figures.
  • The “trust fund debt counts” chart segment includes all government-held debt, including the small amounts held in government accounts other than trust funds. It’s calculated as the difference between the CBO’s “gross federal debt” category and its “debt held by the public” category.

2024 through 2088

  • Annual increments to Social Security and Medicare costs (as a % of GDP) are derived from assumptions used in the CBO’s 2012 Long-term Economic Outlook and recent changes in official projections for Social Security and Medicare, based on materials sourced from the Social Security Administration and the Center for Medicare and Medicaid Services (including trustees’ and other actuarial reports).
  • Annual increments to revenue and all other non-interest costs (as a % of GDP) are derived from assumptions used in the CBO’s 2012 Long-term Economic Outlook.
  • The “trust fund debt counts” chart segment is derived partly from government projections for the Social Security and Medicare trust funds (which leave the funds depleted by 2033 for Social Security and 2026 for Medicare Hospital Insurance). Government-held debt other than the IOUs held by the Social Security and Medicare trust funds is assumed to grow in line with GDP.
  • Annual interest costs are calculated from the debt levels projected within the CYNICONOMICS model and the long-term interest rate assumptions used in the CBO’s 2012 Long-term Economic Outlook, reflecting different assumptions for “publicly held” and trust fund debt.
  • Nominal GDP is projected forward at 4.4% per year, according to the CBO’s 2012 “long-run assumptions” of 2.2% for real GDP and 2.2% for the GDP deflator.
  • CPI inflation is projected forward at 2.5% per year, also based on the CBO’s 2012 “long-run assumptions.”

All years

  • The “recessions really exist” chart segment is created by the same model used for the “baseline” and “alternative” scenarios but with annual non-interest outlays increased by 0.7% of GDP per year (with interest costs adjusted accordingly). Background for this assumption can be found in “Word Matching the ‘Deadly Sins’: #2,” which offered estimates for various categories of budget slippage, including supplemental appropriations, automatic stabilizers, bailouts and proactive fiscal stimulus in recessions. While the CBO appeared to reflect some slippage for supplemental appropriations in its latest projections (by carrying forward this year’s amounts), its projections don’t account for the other categories of budget slippage. These categories were estimated to subtract 0.7% of GDP from average budget balances between 1981 and 2010, and I carried this amount of annual slippage forward in my “recessions really exist” projections.
  • Population projections are taken from the 2013 Social Security Trustees’ Report.
  • The ratio of taxpayers to total population is held constant at 35.9990%, which is the figure as of June 15, 2013.
Bookmark and Share
This entry was posted in Uncategorized. Bookmark the permalink.