| 摘要: | 本研究蒐集2013 年至2022 年台灣19 個縣市消防機關之統計及預算書資料，利用最小平方法進行多元迴歸分析，探討地方政府預算對消防救災效能之影響。由於預算為政府機關達施政目標之有限資源，然而過去文獻鮮少對消防機關施政效能及預算做深入探討，既有文獻探討期間較早、且評估之效能指標僅火災損失及發生頻率，本研究提供較多樣之效能指標且為近期資訊，供各界評估衡量決策時之參考依據。
本研究主要探討預算類型為地方政府預算、消防機關預算、人事費用預算、設備及投資預算四項；效能指標分別為消防員傷亡率、民眾傷亡率、救出人數比例及火災次數比例四項。實證結果顯示：首先，增加地方政府預算可降低下一年度消防員傷亡率，其餘預算類型與消防員傷亡率未達顯著關係。其次，地方政府預算、消防局預算、人事費用預算或設備及投資預算均未能顯著影響民眾傷亡率。再者，增加地方政府預算、消防局預算或人事費用預算皆能提高下一年度火災救出人數比例。最後，地方政府預算、消防局預算或人事費用預算與火災次數比例呈正向顯著關係，與本研究假說不符。;This study examines the impact of local government budgets on emergency response effectiveness by utilizing the rdinary Least Squares (OLS) method for multiple regression analysis. The research collects statistical data and budget records from fire departments across 19 counties and cities in Taiwan, covering the period from 2013 to 2022. Given that budgets are limited resources for government agencies to achieve their policy goals, previous literature has rarely delved into the effectiveness and budgets of fire departments, focusing primarily on financial fire losses and fire occurrence rates. This study offers a more omprehensive set of performance indicators and timely data, providing a valuable reference for decision-making.
The study investigates four types of budgets: local government budgets, fire department budgets, personnel expenses, and equipment and investment budgets. The performance indicators include firefighter casualty rates, civilian casualty rates, rescue rates, and fire occurrence rates. The empirical findings indicate that an increase in local government budgets can reduce firefighter casualty rates in the next year, while other budget types show no
significant regression results with firefighter casualties. Moreover, none of the budget types significantly affect civilian casualty rates. However, increasing local government, fire department, or personnel expenses budgets can improve rescue rates in the subsequent year. Lastly, local government, fire department, and personnel expenses budgets positively correlate with fire occurrence rates, contradicting the study′s hypothesis. |
