透過您的圖書館登入
IP:3.90.242.249
  • 學位論文

我國所得分配模型之探討

An Exploration of the Income Distribution Model for People in Taiwan

指導教授 : 李孟峰
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


近30年來我國的所得分配,如同許多國家,有: 受僱人員報酬(薪資)比重下降、不均度持續上升現象。 然而,一般很少注意前者屬要素(功能性)所得分配的長期趨勢變化,而後者有關家庭所得分配,幾乎為大眾關注及研究的焦點。本研究針對前述兩種所得分配進行時間趨勢之探討。 本研究針對功能性所得分配中之組成成份,即家庭所得總額、受僱人員報酬、產業主所得,應用共整合、誤差修正模型分析方法,探究其長期均衡現象。研究結果顯示,經檢定存在共整合關係,顯示我國所得來源的長期供應端是穩定的。 至於家庭所得分配,將民國68至99年之年資料經適合度檢定,結果發現以兩參數之Gamma機率密度函數最適。其型態參數α與尺度參數β之估計值隨不均度持續上升,分別呈嚴格遞減、嚴格遞增反向變動。本研究嘗試利用此兩參數的估計值多變量時間序列模型,以建立家庭所得分配的模型;配適向量自我迴歸(VAR)模型,得到VAR(2) 符合資料特性。接著,續加入Gini變數後,以三變數進行建模,並採Granger因果關係檢定,以其估計式對Gini變數作預測。又依Gamma分配函數的意義,驗證型態參數α與平均每戶人數、就業人數,及型態參數與型態參數之積αβ與可支配所得之相關性均極高,相關係數皆達0.95%以上。另選擇年齡因子作為家庭特性分解變數,長期觀察戶長年齡45~54歲及35~44歲之吉尼係數的變化,其組內不均度呈和緩升高,所得水準呈穩定成長。

並列摘要


The distribution of income in Taiwan over the last three decades has a common trend, like many countries, the fall in the wage share and greater earning inequality. However﹐little attention was paid to the former, a factor(functional)income which is a long-run trend. Most of the concern and researches focused on the latter, the result of household income distribution. This study tries to model the time trend for both distributions mentioned above. This study applied cointegration and vector error correction model (VECM) procedures to establish a long run equilibrium among the components of functional income, current receipts, compensation of employees and entrepreneurial income. The result shows that current receipts, compensation of employees and entrepreneurial income are cointegrated. This shows that the sources of long-run supply end in Taiwan are stable﹒ With regard to household annual income distribution, the Gamma density with shape parameter α and scale parameter β offers a good fit to the distribution of annual income in Taiwan over the period 1979–2010 by the goodness of fit test. Note that again, the shape prameter α is decreasing in income inequality; and the scale parameter β is increasing in income inequality. This study first established a multi-variable time series model with (αt, βt) to establish a time series household income distribution model. According to the data, a VAR (2) model was obtained by fitting a vector autoregression (VAR) model. Second, a three-variable model was established by joining the Gini variable, and Granger causality test was adopted to found an estimator of Gini variables. According to the meaning of two parameters of Gamma density function, the relevant correlation coefficients between α and the average household size, α and employment person, the product of two pameters αβ and disposable income are over 0.95% respectively. Decomposition household income by age, the Gini coefficient within the household head aged 45 to 54 and 35 to 44 showed a moderate degree of inequality and increased steadily growing income levels.

參考文獻


34. Yannis Dafermos, Christos Papatheodorou(2011), “Functional and personal income distribution in a stock-flow consistent model”, the 15th Conference of the Research Network Macroeconomics and Macroeconomic Policies (FMM), ‘From crisis to growth? The challenge of imbalances, debt, and limited sources’, Berlin, Preliminary draft.
1. A. B. Atkinson(2009), “Factor shares: the principal problem of political economy?”, Oxford Review of Economic Policy, Volume 25, No. 1, 3–16.
2. A. B. Z. Salem and T. D. Mount(1974), “A Convenient Descriptive Model of Income Distribution: The Gamma Density”, Econometrica, Vol. 42, No. 6, 1115-1127.
4. Amita Majumde (1990), “Distribution of Personal Income: Development of A New Model and Its Application to U.S.Income Data”, Journal of Applied Econometrics, Vol. 5, 189-196 .
5. Anand Banerjee, Victor M. Yakovenko, T. Di Matteo (2006), “A study of the personal income distribution in Australia”, Physica A 370﹐54–59.

延伸閱讀