國立高雄大學圖資館 |

Language: English

Back

台灣教育報酬率估計—以九年國教為工具變數之應用 = Estimating...

國立高雄大學應用經濟學系碩士班

台灣教育報酬率估計—以九年國教為工具變數之應用 = Estimating Returns to Schooling in Taiwan—Using Instrument Variable Based on Nine—Year Education Program

Record Type:	Language materials, printed : monographic
Paralel Title:	Estimating Returns to Schooling in Taiwan—Using Instrument Variable Based on Nine—Year Education Program
Author:	陳增達,
Secondary Intellectual Responsibility:	國立高雄大學
Place of Publication:	[高雄市]
Published:	撰者;
Year of Publication:	2010[民99]
Description:	65面圖，表 : 30公分;
Subject:	教育報酬率
Subject:	Returns to Schooling
Online resource:	http://handle.ncl.edu.tw/11296/ndltd/93307340062014393158
Summary:	傳統的OLS估計平均教育報酬率，因教育選擇變數與誤差項中個人能力沒有相互獨立，且在資料上我們也缺乏個人家庭背景與能力之代理變數，來控制個人能力的差異，而導致估計結果產生偏誤。本文試圖解決此內生性問題，並期望估計出較準確的平均教育報酬率。本文運用普通最小平方法與兩階段最小平方法估計平均教育報酬率，運用與九年國教相關的工具變數，分別為九年國教前後世代的虛擬變數以及各縣市國中學校數、班級數與學生人數的密度，本研究使用「人力運用調查」1978-2008年選取1948-1952年(九年國教實施前之世代)與1956-1960年(九年國教實施後之世代)的出生世代作為本文的研究對象。研究結果顯示OLS估計結果約為3.1%，九年國教後之世代之教育報酬率3.15%大於九年國教前之世代的2.94%。使用九年國教前後世代的虛擬變數為工具變數估計之平均教育報酬率為5.3%。若使用台灣各縣市國中班級數的密度作為工具變數，則估計之教育報酬率約為4.2%。我們使用工具變數法估計之平均教育報酬率均高於OLS所估計之結果。 Traditional OLS estimates of returns to schooling are subject to ability bias, due to the presence of unobservable innate abilities which are correlated with schooling and also correlated with wages. It is difficult to obtain family background variables or proxy variables to control innate ability. The 2SLS method is used in this study to overcome the endogeneity bias. Instrumental variables applied include the binary compulsory schooling dummy and the ratios of the number of schools, classes or pupils in junior high level to the number of children aged 12-14. The data comes from the 1978-2008 Manpower Utilization Survey. A subsample including both of 1948-1952 cohorts (pre-policy cohort) and 1956-1960 cohorts (post-policy cohort) mainly are used to estimate the return to schooling. The OLS estimate of returns to schooling is 3.1% for the whole sample. Using the subsample, the OLS estimate of the returns to schooling is 3.1%. In addition, the OLS estimate of the return to schooling is higher for post-policy cohort (3.15%) than pre-policy cohort (2.94%). After correcting the endogeneity bias, the 2SLS estimates of the returns to schooling is 5.3% and 4.2% based on binary compulsory schooling dummy and the ratio of the number of classes in junior high level to the number of children aged 12-14, respectively.

台灣教育報酬率估計—以九年國教為工具變數之應用 = Estimating Returns to Schooling in Taiwan—Using Instrument Variable Based on Nine—Year Education Program
陳, 增達

台灣教育報酬率估計—以九年國教為工具變數之應用 = Estimating Returns to Schooling in Taiwan—Using Instrument Variable Based on Nine—Year Education Program / 陳增達撰 - [高雄市] : 撰者, 2010[民99]. - 65面 ; 圖，表 ; 30公分.
參考書目：面.
教育報酬率Returns to Schooling

台灣教育報酬率估計—以九年國教為工具變數之應用 = Estimating Returns to Schooling in Taiwan—Using Instrument Variable Based on Nine—Year Education Program
LDR:03731nam a2200277 4500 001 273062
005 20170214092352.0
009 273062
010 0 $b 平裝
010 0 $b 精裝
100 $a 20170214y2010 k y0chiy09 e
101 0 $a chi $d chi $d eng
102 $a tw
105 $a ak am 000yy
200 1 $a 台灣教育報酬率估計—以九年國教為工具變數之應用 $d Estimating Returns to Schooling in Taiwan—Using Instrument Variable Based on Nine—Year Education Program $f 陳增達撰
210 $a [高雄市] $c 撰者 $d 2010[民99]
215 0 $a 65面 $c 圖，表 $d 30公分
314 $a 指導教授：許聖章博士
320 $a 參考書目：面
328 $a 碩士論文--國立高雄大學應用經濟學系碩士班
330 $a 傳統的OLS估計平均教育報酬率，因教育選擇變數與誤差項中個人能力沒有相互獨立，且在資料上我們也缺乏個人家庭背景與能力之代理變數，來控制個人能力的差異，而導致估計結果產生偏誤。本文試圖解決此內生性問題，並期望估計出較準確的平均教育報酬率。本文運用普通最小平方法與兩階段最小平方法估計平均教育報酬率，運用與九年國教相關的工具變數，分別為九年國教前後世代的虛擬變數以及各縣市國中學校數、班級數與學生人數的密度，本研究使用「人力運用調查」1978-2008年選取1948-1952年(九年國教實施前之世代)與1956-1960年(九年國教實施後之世代)的出生世代作為本文的研究對象。研究結果顯示OLS估計結果約為3.1%，九年國教後之世代之教育報酬率3.15%大於九年國教前之世代的2.94%。使用九年國教前後世代的虛擬變數為工具變數估計之平均教育報酬率為5.3%。若使用台灣各縣市國中班級數的密度作為工具變數，則估計之教育報酬率約為4.2%。我們使用工具變數法估計之平均教育報酬率均高於OLS所估計之結果。 Traditional OLS estimates of returns to schooling are subject to ability bias, due to the presence of unobservable innate abilities which are correlated with schooling and also correlated with wages. It is difficult to obtain family background variables or proxy variables to control innate ability. The 2SLS method is used in this study to overcome the endogeneity bias. Instrumental variables applied include the binary compulsory schooling dummy and the ratios of the number of schools, classes or pupils in junior high level to the number of children aged 12-14. The data comes from the 1978-2008 Manpower Utilization Survey. A subsample including both of 1948-1952 cohorts (pre-policy cohort) and 1956-1960 cohorts (post-policy cohort) mainly are used to estimate the return to schooling. The OLS estimate of returns to schooling is 3.1% for the whole sample. Using the subsample, the OLS estimate of the returns to schooling is 3.1%. In addition, the OLS estimate of the return to schooling is higher for post-policy cohort (3.15%) than pre-policy cohort (2.94%). After correcting the endogeneity bias, the 2SLS estimates of the returns to schooling is 5.3% and 4.2% based on binary compulsory schooling dummy and the ratio of the number of classes in junior high level to the number of children aged 12-14, respectively.
510 1 $a Estimating Returns to Schooling in Taiwan—Using Instrument Variable Based on Nine—Year Education Program
610 0 $a 教育報酬率 $a 內生性 $a 工具變數
610 1 $a Returns to Schooling $a Endogeneity $a Instrument Variable
681 $a 008M/0019 $b 301206 7543 $v 2007年版
700 1 $a 陳 $b 增達 $4 撰 $3 483202
712 0 2 $a 國立高雄大學 $b 應用經濟學系碩士班 $3 353959
801 0 $a tw $b 國立高雄大學 $c 20101103 $g CCR
856 7 $z 電子資源 $2 http $u http://handle.ncl.edu.tw/11296/ndltd/93307340062014393158