勞動者的工作狀態可區分為就業、失業和非勞動時等三種，在動態流動分析的架構下，勞工可在這三種狀態中，相互轉換以達其效用水準的極大，故共有六種不同的轉換率。行政院主計處每年5月份所做「人力運用調查」係屬橫斷面資料，因每年均重新抽樣調查，故有學者稱此為「重複的橫斷面資料」（Repeated Cross-Section, RCS）。本研究根據Moffitt（1993）的研究架構，使用RCS資料，運用馬可夫模型（Markov Model），分析我國勞工選擇工作狀態的行為，找出那些因素是影響狀態間轉換率的重要因子。主要的實證結果如下：（1）男性較有可能從非勞動力進入就業和失業狀態；（2）已婚者傾向由失業進入就業狀態，或傾向留在非勞動力狀態；（3）學歷愈高者，容易從失業狀態進入就業狀態，不輕易離職成為失業者或非勞動力；（4）年齡愈大，愈不易從就業狀態轉換至非勞動力狀態，也不易由失業轉至就業狀態；（5）估計工資率愈高者，傾向變換工作，從就業進入失業狀態。另一方面，愈容易從非勞動力狀態轉至就業狀態，但有性別差異；（6）愈晚出生者，由失業進入就業狀態的機率愈低，也不易由非勞動力變成就業者。

　　　　　In this study, there are three states to be considered, i.e., employment (E), unemployment (U), and non-Labor force (N). Workers can choose either of the three states depending on which state maximizes his/her utility. Workers flow among the three states implying that there exist six transition rates of interest. The repeated cross-section (RCS) data, which contain information from independently drawn sets of cross-sections of a population at two or more points in time, will be used to study which factors affect worker flow among the three states under the framework developed by Moffitt (1993). Evidence is found that males have a higher probability of changing their state from N to both E and U, that married persons tend to enter state E from U and to stay at state N, that elder workers seldom alter their status from E to N, that workers with higher estimated wage rates are more likely to change their state from E to U and from N to E, and that younger cohorts tend to remain in states U and N.