縣市政府發放生育津貼，一方面鼓勵本地居民「增產報國」解決少子化問題，但另一方面也可能吸引外來民衆遷入，造成「逐補助而居」現象與縣市間生育率的消長。本文利用新近發展的中介分析合成控制法（mediation analysis synthetic control）分解桃園巾於升格後開辦的史上增幅最大一次性生育補助對總生育率的直接與間接效果。結果顯示桃園市擴大補助首年的直接效果爲0.128，僅佔總效果42.7%且不顯著，而其餘0.171總生育率增加則應歸因於大量戶口遷入帶來的間接效果。然而若檢視同年亦大幅提高生育補助的彰化縣，可發現「逐補助而居」現象並未發生，其直接效果反倒顯著提升0.187總生育率。

This paper employs a recently developed mediation analysis synthetic control method to decompose the direct and indirect effects of Taoyuan City’s baby bonus, the most substantial county-level incentive in Taiwan, on the local total fertility rate (TFR) through domestic migration. Our empirical findings show that the direct effect of Taoyuan City’s expanded subsidy increased the TFR by 0.128, which is insignificant and accounts for only 42.7% of the total effect. In contrast, the remaining 0.171 rise in the TFR should be attributed to the indirect effect of a large population influx. Interestingly, in the case of Changhua County, which also raised its maternity allowance in the same year, such “subsidy seeking” did not occur. The direct effect in Changhua County instead significantly bolstered local fertility by 0.187.