[[alternative]]Problems in Social Assistance: Policy Goals, the Definition and Measurement of Poverty

[[abstract]]本文討論社會救助政策中的三大要素:目的預設、貧窮定義與測量三者的適宜性。在文中我們分別檢視目前社會救助政策在這三方面應再思慮的地方，並提出可能的替代方式。本文認為目前的目的預設雖不必完全廢除，但就臺灣社會而言，應再放寬，並以社會對應做為貧窮的定義。貧窮者可包含二類:「基本需求不滿足者」及「社會期望生活水準不滿足者」。而貧窮門檻可以訂在後者，以前者當做是貧戶間差異需求的分野。未滿足基本需求者，政府應無條件給予協助，而未滿足社會期望生活水準者，政府可以依專業工作人員的判斷來提供適當的服務。如此的設計，有助於我們把以往所謂「近貧人口」的需求考慮進社會救助人群之中。且使社會救助有更積極的功能。至於貧窮的測量則應整合社會共識法、行為法與預算法，才得以反映本文所採之貧窮定義。 In this paper, we examine the problems involved in three important components of the public assistance policy: the goals of Social assistance policy, the definition of poverty and the measurement of poverty. It was found that there was room for improvement in each of these three areas. While it is unnecessary to totally abandon policy guidelines, certain aspects of these goals should be relaxed to meet the situation in Taiwan today. In the area of the definition of poverty, we concur with the idea of using "social coping" as the foundation. The designation to be "poor" includes two categories. One is where an individual fails to meet "basic needs," and the other is where an individual cannot reach their "living standard of social expectancy." It is our position that the "poverty threshold" should extend beyond meeting "basic needs" and should also be based on reaching the "living standard of social expectancy." We futher hold the position that the government should provide welfare services to all persons failing to meet their basic needs. However, for those who can not reach their "living standard of social expectancy." government could provide appropriate services based on professional criteria. These suggested change should improve Taiwan's public assistance policy, make it more active, and help it to be able to provide needed services to those "near poor" who are otherwise excluded from the benefits of public assistance. Finally, in the area of measuring poverty, we suggest integrating the "social consensus approaches," the "behavioral approach," and the "budget standard methods