The United States Social Security system is expected to release its 2021 Annual Trustees Report shortly. It is widely anticipated that the report will indicate the effects of the Covid-19 pandemic on the system’s short-range and long-range financial condition and will disclose a number of changes in the various demographic and economic factors underlying the system’s financial projections; these include the range of plausible future financial outcomes as indicated by the applicable stochastic methodology. Of particular interest will be the focus on any changes adopted by the system’s actuaries for the projected mortality experience of existing and future beneficiaries. In many respects, the “Post-Pandemic New Normal” for the United States will be regarded as a bellwether for the Social Security systems of other countries, although there is evidence of considerable differences in the effects of the Covid-19 pandemic between various different countries. This topic presents multiple opportunities for research and analysis by actuaries, demographers, epidemiologists, and economists. Communicating the effects of the Social Security Finance Post-Pandemic New Normal presents a number of challenges in terms of interpreting the scientific evidence and countering misinformation, including pre-conceived notions regarding the potential impact on the solvency and sustainability of Social Security systems.
This situation also provides an opportunity for achieving greater clarity in communicating the key results in the United States Social Security Trustees Report so as to improve understanding by the media and the general public; this objective requires a particular focus on the most appropriate and meaningful metrics available in the 200-plus page report that are often overlooked or disregarded. For example, the report discusses the results of 75-year financial projections in great detail, but does not provide a comparable discussion of the 25-year and 50-year financial projections that are shown in various statistical tables in the report. Apart from disclosing results for the 75-year projection period from the current year, a useful addition would be to report results based on the remainder of the 75-year projection period from the 1983 baseline when the present financing arrangements were established. The report also discusses the financial projections on an “intermediate” basis in great detail, but does not provide a comparable discussion of alternative “high-cost” and “low-cost” projections. There is also very little discussion of the more meaningful stochastic financial projections that are available in the report, showing results at the 50th, the 2.5th, and 97.5th, and the 10th and 90th percentiles (but not the 25th and 75th); the stochastic results should replace the “high-cost” and “low-cost” projection results to provide a much fuller explanation of the array of plausible outcomes with their associated probabilities. A particularly useful metric that could be added is the semi-interquartile range (50 percent confidence interval), representing the difference between the results at the 25th and 75th percentiles, indicating the most likely range of deviation for the expected projection results.
Another topic for potential improved communication is the approximately three trillion dollar trust fund. The use of the term “trust fund” is in fact a misnomer; it is actually not a trust fund in the traditional sense applicable for funded pension systems, but is merely a financing stabilization reserve. Its purpose is simply to provide for a stable payroll insurance contribution rate (currently 6.2% for employers and employees) that exceeds the cost of benefits and administration in the earlier part of the 75-year projection period, but is less than the cost of benefits and administration in the later part of the 75-year projection period. A fuller explanation of exactly how the assets held in this trust fund and their cash flow proceeds are valued would also be helpful. The assets are held mainly as special issue treasury bonds bearing different coupon interest rates according to the year of issue and are valued at nominal par value, essentially treating the coupon rate of each bond as the effective discount rate for valuing future interest and maturity proceeds. In aggregate, this process implies a valuation discount rate that resembles the weighted average coupon rate, whereas the discounted value of a dollar of cash flow proceeds from each separate bond differs from this implied average, depending on the year of issue and the actual coupon interest rate. A more explicit method of valuing the cash flow interest and maturity proceeds would treat all cash flow proceeds at the same rate according to the year received and could reasonably involve the direct determination of an appropriate discount rate or sequence of discount rates to be applied. Moreover, a stochastic model of this asset valuation process could demonstrate the extent of variation in the actuarial value of the trust fund cash flow proceeds under a range of alternative discount rate assumptions. This is an area for further research to bring greater clarity to the question of whether the use of nominal bond values is appropriate and consistent with other elements involved in developing the financial projections.
Ken Buffin, Editor