墊腳石還是陷阱？台灣非典型工作者薪資與職務流動情形

Abstract

非典型勞動，是指一種工作時數與雇用期間皆不固定的彈性化勞動形式。過去對於工作品質的研究證實，相較於典型工作，非典型工作待遇大多較為低劣。而關於非典型勞動市場的分析則主要有兩種觀點：第一種為勞動市場整合的觀點。認為非典型工作除了增加組織人力配置的機動性，更可以減少雇主篩選成本；同時員工也得以透過更為開放的勞動市場、與更有效率的整體經濟獲得益處，最後整合進入勞動市場，產生「雇主─雇員」雙贏的局面，因此這種觀點也被稱為墊腳石（stepping stone）假說。第二種回應勞動市場區隔的觀點，則被稱為陷阱（entrapment）假說，這種看法認為，非典型工作安排其實是強化勞動市場區隔，個人不僅無法找到理想工作，同時增加了不穩定雇用的風險，造成非典型就業的連鎖效應（前一份是非典型工作，下一份就越可能還是非典型工作）。為了釐清何種研究觀點較適用於解釋台灣的非典型就業情形，本文利用97 至100 年「人力運用擬追蹤調查資料庫」作為資料來源，區分出不同類型的非典型工作者。針對97 至100 年台灣就業比例變化、就業者特性、收入與工時、就業者轉職意願等四方面，以描述性統計及多元名目邏輯斯迴歸（multi-nominal logistic regression）說明目前典型與非典型就業概況，以及台灣勞動力身分變化情形。本次研究結果回應了勞動市場區隔理論的看法：比起典型工作，非典型工作在工作待遇以及流動機會上都是較差的工作類型。即便非典型工作者本身希望可以找到全時正職工作，次年仍有可能只能同樣從事非典型工作型態。尤其是臨時或人力派遣工作者更是如此。整體而言，本次迴歸分析結果相對上支持陷阱說，同時也證實了非典型工作的異質性。

Nonstandard work, who are also called contingent work, atypical work,peripheral work, and vagrant work, refers to any job in which an individual does not have an explicit or implicit contract for long-term employment or one in which the minimum hours worked can vary in a nonsystematic manner. Since nonstandard work is a complex idea, there is a debate on whether these kinds of contracts make labor market more segmented. Generally speaking, the existing literatures provide two scenarios of the consequences of nonstandard work arrangement: the stepping-stone hypothesis and the entrapment hypothesis. The stepping-stone hypothesis claims that nonstandard work provides a stepping stone for employees to attend labor market and that there is no negative consequence on their career. On the other hand, the entrapment hypothesis, which is derived from labor market segmentation theory, assumes that nonstandard work has long-lasting negative consequences on job mobility because it makes individual workers “trapped” in the secondary labor market segment. To investigate which hypothesis fits the labor market of Taiwan better, I use “Manpower Utilization Quasi-Longitudinal Survey” to examine the wage inequality and job mobility between standard and nonstandard workers. As a result, the findings from multi-nominal logistic regression model provide some support for the labor market segmentation theory and the entrapment hypothesis.