Thursday, August 21, 2008

New CBO projections of Social Security financing

The Congressional Budget Office today released new projections of future Social Security financing, similar to those done in the annual Social Security Transferred Report. Also see Peter Orzag's blog on the new report, plus the report's data in Excel format (something which endears CBO to the wonky set).

The headline deficit figure – 1.06% of payroll over the next 75 years – is significantly lower than the Trustees 2008 projection of 1.70% of payroll, which was itself a big reduction from the 2007 projection of 1.95%. This post will focus on the source of changes in the CBO projections and as well as how they differ from the Trustees' Report.

There are three principal differences in assumptions between the CBO projections and those from the Social Security Trustees:

  • Revenues from income taxation of Social Security benefits: In a change versus prior practices, CBO assumes that "current law" will prevail with regard to income taxes over the next 75 years. This implies a very significant increase in both overall income tax revenues and income taxes levied on Social Security benefits. As I discussed in a somewhat different context here, current law on income taxes implies much higher future revenues in large part because the income tax brackets are not indexed for the growth of real incomes. As a result, individuals will find themselves pushed into higher and higher tax brackets over time. The Trustees assume that income tax revenues will remain roughly constant relative to GDP in the future. The new CBO assumption lowers their projection of the 75-year shortfall by around 0.24% of payroll, from 1.3% to 1.06%.
    • My comment: This is the most important change, in my opinion. It is hard to say which assumption is "right" in this area. Clearly CBO is correct about what current law entails; however, the Trustees' assumption of roughly steady ratios of tax revenues to GDP are almost certainly more realistic. The income tax code has always had a bias toward higher effective tax rates, but Congress has over time altered the law and kept revenues fairly constant relative to the economy. So there is no right or wrong here and either assumption is reasonable, but readers should bear in mind what each assumption means. "Current law" doesn't equal "likely law."
  • Interest rates: CBO assumes a real interest rate of 3.0 percent above inflation, versus 2.9 percent for the Trustees. This has two related effects: First, investments in the Trust Fund are assumed to earn a higher interest rate, which raises the balance and causes the Fund to last longer. Second, and more important, the interest rate is used to discount future surpluses or deficits to present values, which are used to calculated the system's overall balance. The difference between a 3.0% and 2.9% interest rate can be significant over the long term: a deficit of $100 million (in today's dollars) 75 years from now would have a present value about 7% lower under a 3.0% discount rate than a 2.9% rate.
  • Real wage growth: CBO now assumes a 1.4% annual rate of real wage growth, versus 1.1% for the Social Security Trustees. This derives from their somewhat higher assumption for future productivity growth, of 1.9% vs. 1.7% for the Trustees. Higher real wage growth reduces the 75-year shortfall on a roughly 1-for-1 ratio. This means that a 0.3% difference in the assumed rate of wage growth would account for around 0.30 percentage points of the difference in projected 75-year deficits between CBO and the Trustees.

There are other differences in assumptions between CBO and the Trustees, plus CBO's model uses somewhat different methods than the SSA actuaries, but the above three factors account for most of the differences between the two reports.

I will have additional posts on other topics, in particular new and helpful ways in which the CBO report presents uncertainty regarding future Social Security financing.

No comments: