Does the Form of Retirement Wealth Matter for Spending? Evidence from HRS/CAMS

This study examines whether holding retirement wealth as annuitized income streams or liquid account-based balances is associated with household consumption among older retirees. Using the RAND Health and Retirement Study Longitudinal survey merged with the RAND Consumption and Activities Survey, the analysis focuses on retired and partly retired respondents aged 65 and older with household consumption data and retirement resources. In weighted pooled regressions with respondent-clustered standard errors, both wealth forms are positively associated with consumption, but the association is significantly larger for liquid account wealth. Holding total measured resources constant, a higher annuitized share is associated with lower consumption. However, person fixed-effects models reduce both wealth associations to statistical insignificance, suggesting that the finding primarily reflects differences across households rather than changes within households over time. The findings offer measured support for a liquidity-and-accessibility view of retirement spending while cautioning that cross-sectional wealth-composition patterns should not be interpreted as causal. Implications for retirement-income and decumulation planning are discussed.