Objective: This study provides the first harmonized, micro‑level comparative analysis of pension reform effects on elderly labor force participation in three paradigmatic welfare regimes: Germany (coordinated market economy, conservative‑corporatist welfare), the United States (liberal market economy, liberal welfare), and Brazil (dualistic economy, conservative‑informal welfare). We test the institutional mediation hypothesis: that reform effects are systematically shaped by labor market structures, social protection arrangements, and production regimes.

Methods: We employ harmonized microdata from IPUMS-CPS (USA) and IPUMS-International (Germany, Brazil) , comprising 495,000 person‑year observations spanning 2000–2023. A difference‑in‑differences (DiD) framework identifies causal effects by comparing reform‑affected cohorts/age groups to unaffected controls, with country‑specific treatment definitions: USA (birth year ≥1938, post‑2000), Germany (birth year ≥1947, post‑2012), and Brazil (age ≥65 men/62 women, post‑2020). Triple‑differences (DDD) models formally test cross‑country heterogeneity. Extensive heterogeneity analyses examine differential effects by education, occupation, gender, and formal/informal sector status. Event studies, placebo tests, and multiple robustness checks validate identification.

Findings: Pension reforms significantly increased elderly labor force participation in all three countries, but the magnitude varies predictably with institutional context. Germany exhibits the largest increase (+11.8 percentage points, p<0.001), concentrated among both college‑educated and non‑college workers, with manual workers responding as strongly as non‑manual workers. The United States shows a moderate increase (+5.2 pp, p<0.001), but gains are highly polarized: college graduates respond twice as strongly as non‑college workers (+7.1 pp vs. +4.1 pp), and non‑manual workers significantly outperform manual workers (+6.8 pp vs. +3.2 pp). Brazil demonstrates a modest aggregate increase (+4.1 pp, p<0.001), entirely driven by formal sector workers (+6.7 pp, p<0.001); informal workers exhibit no significant response. Triple‑differences confirm that Germany’s effect is significantly larger than the USA’s (DDD = +6.6 pp, p<0.01), while Brazil’s aggregate effect does not differ statistically from the USA; however, Brazil’s formal sector effect significantly exceeds the USA’s when disaggregated.

Secondary analyses reveal that: (a) Germany’s reform reduced the retirement hazard by 22% (HR=0.78, p<0.001), increasing median retirement age by 1.4 years; (b) Brazil’s reform increased formal elderly employment by 4.8 percentage points with no detectable informality spillover; (c) disability claiming rose in Germany and the USA, suggesting benefit substitution among workers unable to meet higher retirement ages. Decomposition analysis attributes approximately two‑thirds of the observed LFP increase to behavioral responses (reform effects) and one‑third to compositional improvements (rising education).