Stochastic infinite horizon forecasts for US social security finances

Publication date: 28 Oct 2005 | Publication type: National Institute Economic Review | External Author(s): Ronald Lee & Michael Anderson, University of California, Berkeley | Journal: National Institute Economic Review No. 194

Summarised from the National Institute Economic Review, number 194, October 2005. To order a the full version of this article or a subscription, please contact Sage Publications by telephone: +44 (0) 20 7324 8701, email: <a href="mailto:subscriptions@sagepub.co.uk">subscriptions@sagepub.co.uk</a> or online at <a href="http://ner.sagepub.com">http://ner.sagepub.com</a>.

Population ageing poses a major threat to the longÐterm finances of public pension programs around the world. In 2003, Social Security Administration actuaries estimated that a tax increase equal to 1.92% of total payroll would keep the Trust Fund in balance over a 75Ðyear horizon. However, forecasts of PAYGO pension finances are misleadingly optimisticÐ after the 75Ðyear period, the system is running growing deficits.

Anderson's forecasts of the Social Security Trust Fund use an "infinite horizon" to predict fund imbalances. The paper constructs estimates using a horizon of 500 years, as the present value of fund imbalances beyond this horizon is negligible. It predicts a fund imbalance of Ð5.15% of payroll, compared to SSA's infinite horizon estimate of Ð3.5% presented in the 2004 Trustees Report, and the standard 75Ðyear imbalance of Ð1.92%.

Demographic and fiscal forecasts show serious errors only a few years into the forecast period, so scepticism about infinite horizon forecasts is certainly warranted. The paper addresses this by constructing probabilistic forecasts that reflect estimates of their uncertainty. These are based on probabilistic models of the amount of randomness in key demographic and economic variables. They show that there is considerable content to infinite horizon forecasts, in part because outcomes in the distant future are heavily discounted.

The resulting predictions can be expressed probabilistically: there is an estimated 95% chance that the infiniteÐhorizon Trust Fund imbalance will fall between Ð10.5% and Ð1.3% of payroll. The cost of waiting to increase taxes is significant. A 1% increase in the payroll tax today has an effect that is more than twice as big as an increase delayed for 50 years, and five times as powerful as an increase made in 100 years, relative to the scales of the economies. Put differently, it would take a percentage point increase in payroll taxes twice as large if postponed for 50 years, and five times as large if postponed for a century.