Working Paper: NBER ID: w26697
Authors: Hunghao Chang; Chad Meyerhoefer
Abstract: Anecdotal reports and small-scale studies suggest that elections are stressful, and might lead to a deterioration in voters’ mental well-being. Nonetheless, researchers have yet to establish whether elections actually make people sick, and if so, why. By applying a regression discontinuity design to administrative health care claims from Taiwan, we determine that elections increased health care use and expense only during legally specified campaign periods by as much as 19%. Overall, the treatment cost of illness caused by elections exceeded publicly reported levels of campaign expenditure, and accounted for 2% of total national health care costs during the campaign period.
Keywords: elections; health care; stress; public health
JEL Codes: H51; I18; P16
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| health care utilization (I11) | health care costs (I11) |
| stress from elections (K16) | physical health impacts (I12) |
| elections (K16) | health care utilization (I11) |
| elections (K16) | health care expenditure (H51) |
| elections (K16) | acute respiratory infections (I12) |
| elections (K16) | gastrointestinal conditions (I12) |
| elections (K16) | injuries (J28) |
| stress from elections (K16) | mental health service utilization (I11) |