2025 Chinese Retirement Preparedness Index Research Questionnaire

This study includes a randomized information intervention embedded at the end of the national “China Residents Retirement Readiness Survey 2025.” The intervention aims to test whether providing personalized tax benefit information affects individuals’ willingness to open and contribute to a personal pension account.

Respondents who had not yet opened a personal pension account were randomly divided into two groups. The treatment group received a short message informing them of the potential annual tax savings if they contributed the maximum amount (12,000 RMB) to a personal pension account. The message stated, for example, that “if you open a personal pension account and contribute 12,000 RMB per year, you can reduce your personal income tax by approximately [AAA] RMB per year,” where the number [AAA] was calculated based on each respondent’s reported annual after-tax income bracket.

The control group received no such message. Both groups were then asked the same follow-up questions about their future intentions to open a pension account and planned annual contribution amounts.

This randomized double-blind design allows causal estimation of how tax-saving reminders influence participation decisions and contribution levels in China’s new personal pension system.

At the end of the “China Residents Retirement Readiness Survey 2025,” we embedded a randomized information experiment to test how personalized tax-saving reminders affect individual decisions to open and contribute to personal pension accounts.

All respondents who had not yet opened a personal pension account (QS01 = 2) were randomly assigned into two groups of approximately equal size.

Treatment group: respondents were shown a personalized message (QSa) stating that if they opened a personal pension account and contributed the maximum annual amount of 12,000 RMB, they could reduce their personal income tax by approximately AAA RMB per year. The amount AAA was calculated automatically based on the respondent’s self-reported after-tax annual income (QM01), following China’s marginal income tax brackets. Specifically:
< 60,000 RMB → 360 RMB reduction
60,000–92,041 RMB → 360 RMB reduction
92,041–189,480 RMB → 1,200 RMB reduction
189,481–322,680 RMB → 2,400 RMB reduction
322,681–402,960 RMB → 3,000 RMB reduction
402,961–609,960 RMB → 3,600 RMB reduction
609,961–854,760 RMB → 4,200 RMB reduction
854,760 RMB → 5,400 RMB reduction.

After reading the information, participants were asked again about their intention to open a personal pension account (QSa1a) and how much they planned to contribute in the next year (QSa2a).

Control group: respondents did not receive the tax-saving message and were asked the same follow-up questions (QSa1b, QSa2b) without any additional information.
Both respondents and enumerators were blind to treatment status during fieldwork (double-blind design). Randomization occurred automatically within the survey program.

The primary outcomes are:

Intention to open a personal pension account (binary);
Planned annual contribution level (categorical from 1,000 RMB to 12,000 RMB);
Changes in responses relative to pre-intervention baseline answers (QS08 and QS09).
The intervention allows us to estimate the causal effect of salient tax benefit information on retirement savings behavior and assess heterogeneity by income, education, and financial literacy.

2025-09-03

2025-09-30

retirement behaviors; individual pension accounts

This study employs a randomized controlled design embedded in the China Residents Retirement Readiness Survey 2025. The experiment aims to identify the causal impact of providing tax-saving information on individuals’ intentions to open and contribute to personal pension accounts under China’s new private pension system.

Respondents who had not yet opened a personal pension account were randomly assigned to either a treatment group or a control group. The treatment group received an informational message stating that if they opened a personal pension account and contributed the maximum allowed amount (12,000 RMB per year), they could save a specific amount of personal income tax annually, calculated based on their reported after-tax income level. The control group did not receive this message.

Both groups then answered identical follow-up questions about their likelihood of opening an account and planned annual contribution levels. Randomization was automated within the survey software, and both enumerators and respondents were blind to group assignment (double-blind design).

The primary outcomes are changes in:

Intention to open a personal pension account (binary); and
Planned annual contribution amount (categorical).
The study design allows estimation of how highlighting fiscal incentives influences retirement savings behavior in a nationally representative Chinese sample.

Randomization was conducted automatically by a computer within the online survey program. Respondents who had not yet opened a personal pension account were randomly assigned to one of two groups: half were shown the informational message (treatment group) and half were not (control group). The random assignment was implemented using the survey software’s built-in randomization function, ensuring that neither enumerators nor respondents were aware of group allocation (double-blind design).

Random assignment was implemented at the individual level within the online survey software. Each eligible respondent (those who had not yet opened a personal pension account) was independently and automatically assigned to either the treatment group (shown the tax-saving message) or the control group (no message).

There is only one level of randomization, and no clustering by household, region, or enumerator.

Yes

4,000 individual respondents (each individual is a randomization unit; no higher-level clustering).

4,000 observations at the individual respondent level (2,000 treatment, 2,000 control; no clustering).

Control group: 2,000 individual respondents (no tax-saving information shown)
Treatment group: 2,000 individual respondents (shown personalized tax-saving message about potential annual income tax reduction when contributing 12,000 RMB to a personal pension account)
Total planned sample: 4,000 individuals randomized at the individual level.