Policy Simulation Group

Pension Projection Models

Enhancements Introduced in Current 7/30/08 Version of the PSG Models

1. Add an extra year of life before accelerated mortality starts in PENSIM runs that support SSASIM OLC mode runs. The accelerated mortality feature has been in place for several years in order to speed up OLC runs &mdash it kills off people who would otherwise live well beyond the end of the OASDI actuarial test period specified in the OLC run. In anticipation of a new model feature, it has been necessary to delay the onset of accelerated mortality by one year. This means that, for individuals after the first individual in each birth cohort sample, the stream of random numbers is different and hence the cohort sample is different. This causes a tiny bit of sampling variability in the new version of the PSG models. And this sampling variability causes the OASDI actuarial balance (when using the old AAAF_RI factor) to round to -1.94 percent rather than the prior (and 2007-TR) value of -1.95 percent. This can be corrected by using 2.31 rather than 2.28 in the run specification language statement for the GEMINI AAAF_RI factor. The RSF Toolkit's Build interview has been revised to use this slightly different factor when a request is made to emulate the non-microsimulation methods used by OCACT. But if you are using old run specification files that emulate OCACT methods under the 2007 Trustees Report intermediate-cost assumptions, then you will need to change (in each run specification file) the two instances of 2.28 to 2.31 before executing the run with the current version of the PSG models.
2. Add a pension excise tax as a policy option in the PENSIM LAW_CYR table. The revenue of this tax on employer-sponsored pension distributions is deposited in the OASI trust fund.
3. Add individual behavior parameters to PENSIM PAROLL1 table that can be used to eliminate completely pension account rollovers at job-end. This feature has been added to make it easy to simulate an unrealistic situation in which individuals accumulate pension wealth in accounts of only some types of pension plans, as is often done in other pension simulation models.
4. Add two statistics to the PENSIM .cov output file that report the number of lifetime jobs on which an individual vests in a DB (DC) pension plan on a job where that individual's earnings are not subject to social security payroll taxes.
5. Revise SSASIM documentation of MTE and TAXR_?I tables to clarify usage of XTE-related parameters (when earnings above the MTE are taxed, but not used in benefit calculations) and of ETE-related parameters (when some earnings below the MTE are excluded from both payroll taxation and from benefit calculations). The XTE-related parameters are useful when specifying a "donut-hole" reform in which the taxed earnings above the MTE are not used in benefit calculations, while the ETE-related parameters are useful when specifying a "donut-hole" reform in which earnings above the range of excluded earnings (but below the MTE) are used to calculate both payroll taxes and benefits. These changes affect only the documentation, not the logic of the PSG models.
6. Upgrade the PSG models to SQLite 3.6.0 version, which should have no visible effects on the operation of the models.
7. Add SSASIM constraint that the values of pia_frac_1 (the PIA-formula fraction that applies to AIME amounts below the first bend point) in the PIADYNR, PIADYNS, PIADYND tables must be the same. The values can vary over time, but in any given year they must be the same for the retirement event (R), the death event (S), and the disability event (D).
8. Upgrade the PSG models to GNU AWK 3.1.6 version (and newer documentation of the AWK programming language), which should have no visible effects on the operation of the models.