Beveridge 曲線之再探討：空間共整合模型之應用 William H. Beveridge 於1945 年首次提到Beveridge曲線的一個現象，即在經濟景 氣的循環中，當景氣處於較佳的狀態下，此時失業率(unemployment rate)會下降，而職 缺率(vacancy rate)會提高；相反的當景氣處於衰退的狀態下，此時失業率會升高，而職 缺率會下降。他發現失業率與職缺率在景氣循環的過程中，二者會有反向的關係。由 於自1990 年代開始歐美國家的高失業率且持續的居高不下，致使政府決策者及經濟學 家們著手重視Beveridge 曲線背後所隱含的經濟意涵及其政策的實用性。自1989 年的 Blanchard et al.的論文“The Beveridge Curve”開始至今，陸陸續續有許多的理論及實證 的文獻針對不同的國家或地區做研究，驗證其Beveridge曲線關係的存在性與穩定性。 本研究計畫嘗試利用中國30 個省市的區域勞動市場之panel data，期間將蒐集過去至 少25 年各省市之失業率及職缺率，來估計中國區域之Beveridge 曲線。我們將利用 Beenstock and Felsenstein (2010)所發展出之空間誤差修正模型(spatial error correction model, SpECM)，來克服區域空間相依性及區域變數間共整合的問題。此空間誤差修正 模型結合了空間計量模型與Engle and Granger (1987)的誤差修正及共整合模型 (ECM)，不但可以補捉到變數在區域間的外溢效果，亦可得到變數間短期的動態調整 過程及其長期的均衡關係，彌補了各自單一模型的缺點。再者，於第二年的計畫中， 我們將進一步對中國區域勞動市場之Beveridge 曲線隨著時間的改變而移動的現象與 造成其移動的原因進行探討，並與過去對其他國家相關研究的文獻之實證結果進行分 析與比較。 Re-estimation of Regional Beveridge Curve: The Application of Spatial Cointegration Model William H. Beveridge in 1945 first identified an empirical phenomenon in labor market: as the economy grows, the unemployment will decrease and job vacancies will increase; as the economy is in recession, the unemployment will increase and job vacancies will decrease. The negative relationship between unemployment and job vacancies is so-called “Beveridge curve”. The high unemployment and its persistence in the US and European countries have motivated policymakers and researchers to investigate the implications and applications of Beveridge curve. Since Blanchard et al. (1989), numerous literatures have provided theoretical and empirical analyses for several countries and regions to explore the existence and stability of Beveridge curve. In this research proposal, differently from the past literature, we try to analyze the Chinese regional Beveridge curve by using a panel data covering 25 years in 30 regions. We intend to utilize a spatial error correction model by Beenstock and Felsenstein (2010), which combines the methodologies of spatial econometrics and Engle and Granger’s (1987) error correction model, to catch the characteristics of spatial dependency and spatial error corrections. Furthermore, in the second year of this research, we also try to investigate the determinants of shifts of Chinese regional Beveridge curve, and compare with the results of current literature for other countries.
