妻子看到广告，丈夫采取行动：妻子接触铁路公共卫生广告与丈夫采取预防行为的关系

While public health interventions have traditionally targeted decision-makers themselves, this paper shows that targeting their household members may be more effective in the context of Japan's rubella vaccination policy. Japan introduced a program which subsidizes preventive behaviors from antibody testing through vaccination targeting middle-aged men. We overcome a key limitation in measuring public health advertising effects by combining a natural experiment based on a Tokyo railway advertisement campaign with individual-level data linking exposure opportunity, viewing the advertisement, and offline preventive behavior. We find no statistically significant effect of eligible men's own use of the train lines displaying the campaign on antibody testing take-up. In contrast, their spouses' use increased antibody testing take-up by 1.4 percentage points in the short run and 4.54 percentage points in the long run. Instrumental variable estimates show that the direct effect of viewing the advertisement was not statistically significant, suggesting that spousal persuasion plays a key role. These results imply that expanding the targeting of interventions to third parties can be an effective policy option.

This is the Japanese version of the English Discussion Paper (26-E-052).