By Alina Malkova

When governments expand pension systems, families reduce intergenerational transfers. This paper tests the reverse: do families substitute back toward informal insurance when governments renege on pension promises? I study Russia’s 2014 funded pension freeze, which eliminated future contributions for workers born after 1967, using RLMS panel data (2010–2023) and a cohort difference-in-differences design. Affected households increased financial transfers to adult children by 3.5 percentage points under a conservative age×year fixed effects specification and by 6.5 percentage points in the baseline—a 17–30 percent increase relative to pre-reform rates. To address the core identification challenge that treated households are mechanically younger, I exploit wage-based variation within the treated cohort: because the freeze diverted 6 percent of wages to frozen accounts, higher-wage workers lost more pension wealth and show monotonically larger transfer responses (4.1 percentage points per standard deviation, p = 0.013). A placebo test on control households—who were unaffected by the freeze and faced identical macroeconomic conditions—yields a precisely estimated zero wage dose-response (−1.0 percentage points, p = 0.50), ruling out general income effects as an alternative explanation. A falsification test reinforces this interpretation: transfers to children rise significantly (7.4 percentage points, p = 0.005) while transfers to parents do not (2.8 percentage points, p = 0.12). The response is entirely on the extensive margin—households initiate new transfer relationships rather than scaling up existing ones—and is concentrated among small households and rural areas. These patterns are consistent with parents investing in non-resident children as a source of old-age support. Collectively, the findings provide evidence that even in middle-income settings with well-established state pensions, families retain the capacity to rebuild informal insurance networks when formal systems fail.

Source SSRN