Introduction

The long-term shift from defined benefit to defined contribution retirement plans has increased individual responsibility for retirement security. This paper examines how trust in financial institutions and government programmes, alongside financial literacy, shapes retirement behaviours among older Americans. Using a specially designed module in the 2020 Health and Retirement Study (HRS), the authors introduce new domain-specific trust measures and document their associations with retirement account ownership, stockholding, and net household wealth.

Datasets

The study uses a module fielded to a randomly selected 10% of the 2020 HRS, a nationally representative survey of adults aged 50+ in the US, with a resulting sample size of 1,286 (p.7). The module measured trust in mutual funds, financial advisors, banks, insurance companies, Social Security, and Medicare/Medicaid on a 0–10 scale. The “Big Three” financial literacy questions were also included, alongside controls for age, sex, marital status, income, education, race/ethnicity, cognitive ability, number of children, and willingness to take risks.

Methods

Multivariate regressions examine trust, financial literacy, and three financial behaviours: retirement account ownership, stockholding, and total net household wealth. Logit regressions report average marginal effects for the two binary outcomes; median regression is used for the continuous wealth variable.

Analysis of detailed trust measures

Trust in mutual funds and financial advisors shows a generally positive and significant association with all three retirement behaviours. A one-unit increase in trust in mutual funds and financial advisors increases the likelihood of having a retirement account by 2 and 1.5 percentage points respectively (p.10). Trust in Medicare/Medicaid is negatively associated with retirement account ownership and stockholding: a one-unit increase reduces each likelihood by 19 and 17 percentage points respectively (p.10).

Financial literacy is positively associated with all three outcomes: a one-unit increase raises the probability of having a retirement account by 6 to 6.5 percentage points, the probability of investing in stocks by 7 percentage points, and corresponds to $30,300–$31,100 more net household wealth per additional correct answer (p.11).

Aggregate trust measures: Trust in financial institutions and government programmes

A one-unit increase in Trust in Financial Institutions is associated with a 6% higher likelihood of having a retirement account, a 7.5% higher likelihood of holding stocks, and a 3.2% increase in household net wealth (around $17,800) (p.12). Conversely, a one-unit increase in Trust in Government Programmes corresponds to a 4% and 6% lower likelihood of having a retirement account and holding stocks respectively, and a 1.6% reduction in net household wealth (about $8,900) (p.12). These findings indicate a “crowding out” effect. A one-unit increase in financial literacy corresponds to a $34,600 (or 6%) increase in net household wealth (p.13).

Heterogeneity by race and ethnicity

Blacks and Hispanics exhibit lower Trust in People, Trust in Financial Institutions, and financial literacy compared to Whites, yet report greater Trust in Government Programmes (p.14). Trust in Financial Institutions is significant for Whites — a one-unit increase raises the likelihood of having a retirement account by 7%, holding stocks by 8%, and net wealth by $25,400 (3%) (p.14) — but is not significant for Black and Hispanic respondents. Trust in Government Programmes is negatively associated with stockholding and wealth accumulation for minorities (p.15).

Financial literacy effects are economically meaningful across all groups: a one-unit improvement is associated with a 5.3 percentage point (48%) increase in the probability of having a retirement account for minorities, versus 8.3 percentage points (17%) for Whites (p.15). These findings underscore the importance of enhancing financial literacy in minority communities as a means of mitigating retirement savings disparities.

Robustness analysis

Multiple checks confirm the stability of results, including additional control variables (religion, life satisfaction, depression, self-rated memory), alternative financial literacy metrics, indicators for social security and Medicare/Medicaid receipt, and Principal Component Analysis of trust variables. Results remain broadly consistent across all specifications.

Discussion and policy implications

Trust and financial literacy operate through distinct channels and both matter for retirement preparedness. Policies that enhance transparency, reduce frictions in interacting with financial institutions, or build confidence in private retirement products may increase the effectiveness of financial education. Regulatory measures such as consumer protection, financial stability oversight, and improved transparency in financial advisory services could also foster greater trust in financial institutions.

Tailored trust-building policies aligned with the trust factors relevant to specific population sub-groups are recommended. Providing financial education, particularly among diverse racial and ethnic groups, is highlighted as a broadly effective strategy for strengthening retirement preparedness.