Public sector employee with defined benefit pension, 403(b)/457, Social Security offset (WEP/GPO), retiree health benefits.
R-03 is the public-sector pre-retirement profile where the Windfall Elimination Provision, the Government Pension Offset, and a layered 403(b)/457 contribution surface combine to make Social Security and supplemental-savings projections wrong by default in software not built for them.
R-03 captures the public-sector pre-retiree — teachers, state and municipal employees, federal CSRS/FERS holdouts, and uniformed services civilians — where a defined-benefit pension is genuinely the dominant retirement vehicle and Social Security is partially or wholly offset. Two regulatory mechanics distinguish this archetype from every other pre-retirement profile in the catalog. The Windfall Elimination Provision (WEP) reduces the worker's own Social Security benefit when they receive a pension from non-covered employment; the Government Pension Offset (GPO) reduces spousal and survivor Social Security benefits by two-thirds of the non-covered pension. Software that applies the standard PIA formula to an R-03 household will overstate Social Security income by thousands of dollars per month. Alongside this, the supplemental savings stack is unusual: a 403(b) for K-12 and nonprofit-hospital employees, a governmental 457(b) that has no 10% early-withdrawal penalty on separation regardless of age, and in many jurisdictions a DROP (Deferred Retirement Option Plan) election that lets the employee accrue pension payments into a side account for a defined period.
Cash flow reflects the mass-market wealth tier this archetype is calibrated to. Median combined income is $53,757 — materially below R-01 and R-02 — but median investable assets of $260k and median net worth of $466k are higher than the income alone would suggest, because the pension itself is the dominant retirement asset and the household has been saving the gap inside the 403(b)/457 stack rather than as W-2 deferrals into a corporate 401(k). Homeownership sits at 56%, slightly above the income cohort norm because public-employee mortgages benefit from MOP, state housing-finance-agency programs, and employer-tied housing assistance in some jurisdictions.
What separates R-03 from R-01 is not 'has a pension' — R-01 households often have one too — but the regulatory texture: WEP and GPO are R-03-only, retiree health benefits remain a real promise in this archetype where they have largely disappeared in R-01, and the 457(b) penalty-free-on-separation rule creates planning options that don't exist in 401(k)-based corporate retirements. The R-03 corpus is deliberately sized at 25 households to give buyers a tight, recognizable reference for the public-sector edge cases their software needs to handle without the noise of higher-balance corporate testing scenarios.
Aggregated across the 25 R-03 households in the shipped v3 corpus corpus. Numbers describe the corpus, not population claims.
Eric and Melissa are the diagnostic R-03 case: a K-12 public-school employee married to a healthcare worker, both still in pre-retirement employment at ages 57 and 59, sitting essentially at the corpus net-worth median ($462,525 vs $465,921). Combined income of $67,989 is modest but the household carries $170k of accumulated retirement savings, a $170k mortgage, and only $11k of non-mortgage debt — the classic public-sector wealth shape where balance-sheet assets understate the household's true retirement position because the DB pension is not capitalized. Retirement shows off-track against a $1.25M target the explicit balance sheet cannot reach, while the $441k legacy goal is on track at $367k of progress — a household that fails a naive net-worth-to-target test but passes a properly modeled pension-plus-Social-Security income-replacement test once WEP is applied to the worker's own benefit and GPO is applied to spousal coverage.
Every R-03 household ships with — at minimum — these JSON fields populated. The full schema is documented in the data set you purchase.
Three buyer profiles draw on R-03 most heavily. Retirement-planning software teams use it to validate WEP and GPO calculations against published SSA examples — a calculation absent in corporate-pre-retiree test corpora and critical for any platform that markets to teachers, firefighters, police, or municipal employees. Benefits-administration vendors serving state pension systems and 403(b)/457 plan providers use the corpus to test contribution-limit edge cases, including the §457(b)(3) final-three-years catch-up and the §414(v) age-50 catch-up stacked across multiple plan types. Wealth platforms with public-sector advisor channels (public-sector retirement-plan recordkeepers) use R-03 to populate realistic retiree-health-benefit cost projections and DROP-vs-retire decision UX.
R-03 is calibrated as a US public-sector pre-retiree where the DB pension and Social Security are both in play (with WEP/GPO applied). Federal employees under FERS (where Social Security is fully covered alongside a smaller DB component and a TSP) are partially represented; pure CSRS-only edge cases are rare in the corpus because the cohort has largely retired. Public-sector employees who never enrolled in a DB plan and rely solely on a 401(a)/457 stack are not the intended fit — closer to P-05. Military pre-retirees with uniformed DB-pension elections belong in MV-02. Recently retired public-sector households are in RE-03 (the pension-rich early retiree) rather than R-03. Cross-border or foreign-pension-receiving households are excluded.
Income bands during v3 synthesis were anchored to BLS Occupational Employment and Wage Statistics for K-12 teachers, state and local government workers, and uniformed-services civilians in the 55–65 age cohort. Pension-prevalence and 403(b)/457 participation rates were loosely informed by EBSA Form 5500 filings for governmental plans. WEP/GPO applicability is implicit in the household structure — the corpus assumes the dominant pension is non-covered employment for Social Security purposes, which is the realistic case for most state-teacher and many municipal plans. State distribution favors CA, TX, and PA where large state pension systems concentrate the relevant testing scenarios. Per CLAUDE.md §9 the corpus is frozen and not regenerable; calibration claims here are descriptive rather than reproducible.
Corporate W-2 pre-retiree at the mass-affluent tier with a private-sector pension. No WEP/GPO, no 457(b), much larger 401(k) and NQDC stack. Use R-01 when the test surface is corporate benefits rather than WEP/GPO calculations.
Self-employed pre-retiree with no pension at all. The retirement event is a business sale, not a pension election. Different planning surface entirely.
Already-retired pension-rich household, often the natural next stage for an R-03. Use RE-03 when the pension is already in pay status and the household is inside Medicare and the active decumulation phase.
Career military officer with a uniformed-services DB pension and TSP. Similar pension-dominant profile but with VA benefits, SBP elections, and a uniformed-services Social Security treatment that is fully covered, so WEP/GPO does not apply.
R-03 is the Government/Teacher Pre-Retiree: a public-sector employee within five to ten years of retirement, with a defined-benefit pension as the primary retirement asset, a 403(b) or 457(b) supplemental savings stack, retiree health benefits, and Social Security partially or wholly offset by WEP and GPO.
The Windfall Elimination Provision modifies the Social Security benefit formula's first bend point for workers who receive a pension from non-covered employment and have fewer than 30 years of substantial covered earnings. The Government Pension Offset reduces spousal or survivor Social Security by two-thirds of the non-covered government pension. The corpus assumes the dominant pension is non-covered, which is the realistic case for most state teacher plans and many municipal plans.
Partially. FERS employees have Social Security coverage (so WEP/GPO does not generally apply), a smaller FERS basic-benefit DB component, and a TSP. R-03 leans toward state and local plans where WEP/GPO is the planning issue; for FERS-specific testing scenarios buyers typically apply additional overlays to the household structure.
R-01 is private-sector with full Social Security coverage and typically a larger 401(k) plus NQDC stack at a higher wealth tier. R-03 is public-sector with a dominant non-covered DB pension, WEP/GPO applied to Social Security, a 403(b)/457 layered stack, and retiree health benefits that mitigate the pre-Medicare bridge. The diagnostic difference is the regulatory texture, not just sector.
R-03 is tagged for six bundles — B03, B06, B14, B18, B19, and B25 — covering retirement income, healthcare and Medicare, employee benefits, decumulation, Social Security optimization, and public-sector employee planning. See the sidebar for the specific data sets that ship R-03 households.
No. The v3 corpus is frozen and not regenerable from current code (drift was confirmed on 2026-05-09). The 25 R-03 households are a fixed reference dataset; sampler improvements land in v4 with per-archetype golden fixtures in CI to prevent silent drift.
Download households matching this archetype as part of a Wealth Data Set.
Browse Data Sets