Revisiting Selection in Heterogeneous Returns to College Education

再論大學教育之異質回報的選擇性

社會學家認為大學教育之異質性經濟回報是負向選擇的結果：最不可能上大學者，其回報最高。經濟學家則主張正向選擇：回報愈高者，愈可能上大學。本文說明：這兩個假設之所以看似矛盾對立，和「可觀察變數的選擇性」與「不可觀察變數的選擇性」之差別，息息相關。臺灣之實徵分析結果顯示：在九○年代初期，基於家庭背景的「負向社會選擇」與基於比較優勢原則的「正向自我選擇」可以並存，並非彼此互斥。然而，近二十年來大量擴展高等教育的結果，晚近的回報已顯著不如早期，自我選擇的分類收益有由正轉負的跡象，但負向的社會選擇依然如故。

Sociologists describe the pattern of selection in heterogeneous returns tocollege education as negative, in contrast to the positive selection proposed byeconomists. This article moves beyond such a conflicting contrast, suggestingthat the contradictions between “selection on observables” and “selection onunobservables” are at the heart of the contradictions between these two selectionhypotheses. Employing both sociological and econometric counterfactualapproaches to estimate college treatment effects, this article shows that the negativepattern of social selection based on family background characteristics andthe positive pattern of self-selection based on the principle of comparative advantageare not mutually exclusive—both patterns emerged in the early 1990s, whenTaiwan’s higher education systems were rationed with structural barriers. SinceTaiwan’s swift expansion in higher education over the last two decades, nevertheless,there have emerged signs of decline in the treatment effect for the treated,coupled with a sorting loss in the face of negative social selection.