Long-Term Actuarial Balance

Definition

The Long-Term Actuarial Balance (LTAB) is the immediate, permanent change in the Old-Age, Survivors, and Disability Insurance (OASDI) payroll tax rate required to equalize the present value of all non-interest revenues and all costs (benefits plus administrative expenses) over a 75-year projection horizon, leaving terminal reserves equal to one year of benefit outflows. If LTAB $= -2.2$ percentage points, the current $12.4\%$ combined payroll tax rate would need to be raised by $2.2$ pp immediately and permanently to achieve 75-year actuarial balance. The LTAB is the primary summary solvency statistic reported in each year's Annual Report of the Board of Trustees and the central metric of most Social Security reform debates.

Key Ideas

The balanced-budget tax trajectory (Lee and Skinner 1999): Under different mortality scenarios, the pay-as-you-go payroll tax rate required to maintain yearly solvency rises substantially by 2070: Social Security Administration (SSA) middle (LE=81) → 20%; Lee-Carter point (LE=87) → 24%; LE=90 → 27%; LE=100 → 32%. Starting from 12% in 2000, every additional year of life expectancy (LE) above the SSA baseline adds roughly 0.8–1.0 pp to the required tax rate by 2070. This illustrates how sensitive the pay-as-you-go cost is to the mortality assumption, and why the SSA vs. Lee-Carter gap (~3 pp payroll) carries enormous long-run fiscal implications.
1996 benchmark values: SSA intermediate projection LTAB $= -2.2$ pp. Lee and Tuljapurkar (1998) stochastic mean $= -3.3$ pp. When the Lee-Tuljapurkar model is constrained to SSA's mortality assumptions, result is $-2.3$ pp — validating the calibration. The Lee-Tuljapurkar $95\%$ confidence interval (CI) runs from $-0.2$ to $-6.5$ pp (width $6.3$ pp), similar in width to SSA's $+0.5$ to $-5.7$ pp high-low range ( $6.1$ pp) but centered $\approx 1.1$ pp lower.
Dominant uncertainty source: Fertility is the largest driver of 75-year LTAB uncertainty in Lee and Tuljapurkar's stochastic framework, followed by productivity growth, interest rates, and then mortality (last). SSA's scenario sensitivity analysis ranks these in the exact reverse order (mortality first, fertility last). The reversal occurs partly because SSA varies one factor at a time while Lee-Tuljapurkar allow joint variation.
Misleading properties of the metric:
- The 75-year horizon and terminal-reserve convention create an artificial cutoff: a $2.2$ pp immediate tax increase would still leave the system in rapid deficit expansion post-2070.
- "Immediate and permanent" is a stylized construct — in practice, taxes and benefits would be adjusted gradually in response to actual developments.
- As a stock (present-value) measure, the LTAB can improve even as near-term deficits grow, if long-run projections become slightly more favorable.
- The LTAB understates fiscal risk by focusing on the mean: Lee-Tuljapurkar find a $22\%$ probability that even an immediate $+4$ pp tax increase fails to prevent exhaustion.
Payroll tax alternative: The pay-as-you-go tax rate required to maintain a one-year buffer reserve rises from a median of $12.4\%$ (1994 level) to $\approx 21\%$ by 2070, with the $97.5$ th percentile reaching $34\%$ — implying a $2.5\%$ chance OASDI alone requires a $34\%+$ payroll tax rate.
Why fertility dominates (Tuljapurkar and Boe 1999): Mortality forecasting has a clear historical trend — the Lee-Carter $k(t)$ declines linearly — with bounded variance around a well-identified single factor. Fertility forecasting inherits two compounding uncertainty sources: (1) autoregressive moving average (ARMA) short-run variance from volatile historical total fertility rate (TFR) shocks ( $95\%$ intervals reach one-child width in 6 years), and (2) structural uncertainty in $F^*$ , the long-run average TFR, which operates on $30$ – $50$ -year generational timescales not probed by available data and must be set subjectively. Historical validation shows that anchoring $F^*$ to recent replacement-level ( $2.1$ ) — as SSA and Census Bureau do — can systematically fail to capture baby-boom-scale swings. Over 75-year horizons, this irreducible structural uncertainty in $F^*$ dominates, making fertility — not mortality — the largest source of LTAB uncertainty. SSA's own sensitivity analysis ranks the factors in reverse order (mortality first, fertility last) because SSA varies one factor at a time with narrow fertility scenarios, ignoring joint variation and structural $F^*$ uncertainty. See Stochastic Fertility Forecasting.
Infinite horizon vs. 75-year AB — the sustainability gap (Lee, Miller, and Anderson 2004): The 75-year actuarial balance (AB₇₅) is not a sustainability measure. Raising the payroll tax by the 2004 AB₇₅ of 1.92% leaves the median Trust Fund ratio going negative by 2074 and plummeting to −82 times annual costs by 2200. Using a 500-year stochastic projection, Lee et al. estimate the true infinite horizon imbalance at −5.15% of payroll, compared to the 2004 Trustees Report's −3.5%. The difference is almost entirely attributable to different (higher) mortality improvement assumptions. The 95% probability interval spans −1.3% to −10.5% (range ~~9%), about 40% wider than the 75-year interval (~~6.5%). Simple approximations from the 75-year projection: Flat Fund Ratio Tax (4.36%) underestimates because costs grow faster than revenues; "Unstable" measure (5.21%) closely approximates the 500-year central estimate. Stochastic immigration adds negligible uncertainty to the OASDI solvency distribution. See Timothy Miller and Michael Anderson.
2003 benchmark values and methodology sensitivity (Wilmoth 2005): Trustees Report (TR) 2003 LTAB $= -1.92$ pp (vs. $-2.2$ pp in 1996, driven partly by different demographic assumptions). The 2003 Technical Panel (TP) identified five methodological decision points at which Office of the Chief Actuary (OACT) consistently chose the longevity-pessimistic option — see SSA Mortality Forecasting. The Panel's recommended changes had nearly offsetting LTAB effects: migration adjustment $-0.25$ pp (more workers: better solvency); mortality adjustment $+0.24$ pp (more longevity: more benefit years); net $\approx 0$ . Life expectancy benchmarks at 2070: TR2003 $e_0 = 82.8$ yr; TP2003 $= 84.4$ yr; TP1999 $= 85.2$ yr; Lee-Tuljapurkar stochastic median $\approx 86+$ yr. The LTAB's sensitivity to mortality assumptions is thus real but modest compared to its sensitivity to fertility and productivity (see Lee-Tuljapurkar uncertainty decomposition above).
Dependency ratio channel (Tuljapurkar, Li and Boe 2000): Each 1-year difference in life expectancy at birth corresponds to $>5\%$ difference in the old-age dependency ratio (65+/20–64). Tuljapurkar, Li, and Boe (TLB)'s stochastic median $e_0$ forecasts for 2050 exceed official central projections by $1.3$ yr (UK) to $8.0$ yr (Japan), implying dependency ratios $6\%$ (UK) to $40\%$ (Japan) higher than official by 2050 — with divergence accelerating beyond 2050 as all official scenarios assume eventual deceleration while the stochastic model does not. For the US specifically, the stochastic median $e_0$ in 2050 is $82.9$ vs. official $80.5$ , a $2.5$ -year gap implying $>12\%$ higher old-age burden than projected. See SSA Mortality Forecasting.

How It Works

$\text{LTAB} = \frac{\text{PV(Non-interest Revenue)} - \text{PV(Costs)}}{\text{PV(Taxable Payroll)}}$

Present values are computed over 75 years at the Trust Fund interest rate. Revenue excludes interest income on Trust Fund assets; costs include benefit payments plus administrative expenses. The denominator converts the surplus/deficit into payroll-tax-equivalent units (percentage points of taxable payroll). A terminal reserve of one year's benefits is added to the revenue side as a target solvency condition. A negative LTAB equals the required immediate and permanent payroll tax increase.

Why It Matters

The LTAB is quoted in every Social Security reform debate. Understanding its limitations — horizon sensitivity, the mean/distribution distinction, and the post-horizon cliff — is necessary for interpreting policy proposals accurately.
The stochastic distribution of the LTAB shows that $-2.2$ pp is the mean of a wide distribution, not a point estimate. Policy set at the mean leaves substantial probability of outcomes requiring $5+$ pp tax increases or equivalent benefit reductions.
The finding that fertility dominates long-run LTAB uncertainty suggests that immigration, pronatalist policy, or labor-force participation reforms targeting working-age populations may have larger long-run OASDI solvency implications than mortality-targeted interventions — a reordering of policy priorities relative to SSA's own framing.

Open Questions

How does the 75-year LTAB compare to SSA's infinite-horizon open-group deficit measure? Partially answered by Lee, Miller, and Anderson (2004): the Trustees' 2004 infinite horizon figure (−3.5%) understates the true stochastic central estimate (−5.15%) primarily due to mortality assumptions; the 75-year AB of 1.92% is decisively not a sustainability measure. Raising taxes by that amount leaves the system deeply insolvent in the median within 75 years.
Should the terminal reserve convention be modified to account for the rapidly accelerating demographic shift expected after 2070?
Is the LTAB still the best communication tool for legislators, or do probability-of-exhaustion metrics (as produced by stochastic simulation) convey more actionable information?
What is the cost of delay? Answered by Lee, Miller, and Anderson (2004): at a residual discount rate of 1.4%/year, a 1% payroll tax increase now is worth 2× a 1% increase in 50 years and ~70× a 1% increase in 300 years.