本文分三部份探討型 II 糖尿病對就業之影響。首先以外生 probit 模型觀察，在不同勞動力定義下，其迴歸結果是否會因不同定義而受到影響。預期若將因健康因素無法繼續就業之樣本刪除，會使糖尿病對就業的負向較不易被觀察到。結果顯示，對比不同勞動力定義下，以刪除其它無法就業的樣本並保留因健康因素等樣本所估計出來之結果較適當。保留因健康因素無法繼續就業之樣本對於估計糖尿病對就業的負向效果有顯著影響。若按傳統嚴格勞動力之定義，糖尿病對就業的負向效果反而較難被觀測到。刪除因健康因素無法繼續就業的樣本會低估糖尿病對就業之負向效果。

    其次，就內生二元 probit 模型而言，本研究是以近親是否有糖尿病為工具變數。研究發現: 當刪除其它無法就業的樣本並只保留因健康因素和傳統最嚴格就業定義的兩種樣本，也可觀察出原始樣本所未估計出糖尿病對就業的負向效果。但內生性檢定支持外生假設， 相關迴歸結果仍須以外生模型所估計的結果為主。

    再者，本文將工作以就業時數區分為部份工時就業以及全職就業，運用帶工具變數三元 probit 模型進行估計。

This study mainly investigates the relation between typeII diabetes and employment under various definitions of labor force, and it is separated into three parts. First, it estimates the impact of diabetes on employment using the probit model under exogenous assumption. By comparing various definitions of labor force, it finds that the data which keeps the persons who were unable to work for health
reasons is more appropriate, and its coeffeicient estimate is negative and significant. In contrast, it is difficult in finding the expected impact when following the
traditional, rigorous definition of labor force. Ignoring those persons who were unable to work for health reasons would underestimate the negative impact of diabetes.

The second part of this study is to take the potential endogenous problems of type II diabetes into account by using the bivariate probit models with instrumental
variables. The instrumental variable is the variable indicating whether there is at least one close relative suffering from the diabetes or not. We find similar outcomes in both the exogenous probit models and the bivariate probit models with instrumental variables. It finds the expected impact in two definitions of labor force. One is to keep those persons who were unable to work for health reasons. The other is to keep those
persons who were unable to work for health reasons and those who were laid off. However, the endogeity test supports the exogenous assumption. 

Moreover, it considers two types of employment, namely the part-time workers and the full-time workers. Using the trivariate probit model with an instrumental variable, the results show that the negative impact of type II diabetes is negative and significant for the part-time workers.

目錄
第一章 緒論.............................................1
第二章 文獻回顧.........................................3
第三章 資料來源及變數定義................................5
第一節 資料來源.........................................5
第二節 資料合併及樣本篩選................................6
第三節 變數定義與敘述統計................................7
第四章 實證模型設定.....................................24
第一節 糖尿病對就業市場的衝擊............................24
第五章 實證結果分析.....................................27
第一節 外生模型迴歸結果..................................27
第二節 內生模型估計結果..................................29
第三節 糖尿病對部份工時及全職就業之影響...................30
第六章 結論.............................................76
參考文獻................................................77


表目錄
表 1. 文獻就業定義比較....................................4
表 2. 變數定義表 ........................................12
表 3. 敘述統計表-全樣本...................................13
表 4. 敘述統計表-男性樣本.................................14
表 5. 敘述統計表-女性樣本.................................15
表 6. 各資料型態下糖尿病樣本與非糖尿病樣本之就業情形 .......16
表 7. 有無型 I,型 II 糖尿病對就業之敘述統計差異 ...........18
表 8. 外生模型-糖尿病對就業(原始樣本) .....................32
表 9. 外生模型-糖尿病對就業(保留因裁員無法就業) ............34
表 10 外生模型-糖尿病對就業(保留因裁員以及因健康因素無法就業) ........................................................36
表 11 外生模型-糖尿病對就業(保留因健康因素無法就業) ........38
表 12 外生模型-糖尿病對就業(僅刪除退休樣本) ...............40
表 13 外生模型-糖尿病對就業(僅保留退休樣本) ...............42
表 14 外生模型-糖尿病對就業(最嚴格樣本) ...................44
表 15 外生模型-型 I、型 II 糖尿病對就業(原始樣本) .........46
表 16 外生模型-型 I、型 II 糖尿病對就業(保留因裁員無法就業) 48
表 17 外生模型-型 I、型 II 糖尿病對就業(保留 4,5) .........50
表 18 外生模型-型 I、型 II 糖尿病對就業(保留因健康因素無法就業) ........................................................52
表 19 外生模型-型 I、型 II 糖尿病對就業(僅刪除退休樣本) ....54
表 20 外生模型-型 I、型 II 糖尿病對就業(僅保留退休樣本) ....56
表 21 外生模型外生模型-型 I、型 II 糖尿病對就業(最嚴格樣本) 58
表 22.外生-各類型糖尿病模型邊際效果........................60
表 23 內生模型-型 II 糖尿病對就業(原始樣本)................62
表 24 內生模型-型 II 糖尿病對就業(保留因裁員以及因健康因素無法就業)...................................................64
表 25.內生模型-型 II 糖尿病對就業(保留因健康因素無法就業)...66
表 26.內生模型-型 II 糖尿病對就業(最嚴格情況下)............68
表 27.內生模型工具變數檢定................................70
表 28.內生模型邊際效果....................................71
表 29.三元估計-型 I 型 II 糖尿病對全職及部分工時(原始樣本)..72
表 30.三元估計-型 I 型 II 糖尿病對全職及部分工時(保留 4,5)..73
表 31.三元估計-工具變數檢定...............................74
表 32.內生三條-邊際效果...................................75

參考文獻
林東瓏, 胡登淵.(2015).糖尿病對台灣中老年人就業的衝擊-體重控制的緩和效
果.經濟論文叢刊,43(3),359-405.
Bastida, E., & Pagán, J.A. (2002). The impact of diabetes on adult employment and
earnings of Mexican Americans: findings from a community based study. Health
economics,11(5), 403-413.
Cappellari, L., & Jenkins, S. P. (2003). Multivariate probit regression using simulated
maximum likehood. The stata journal,3(3),278-294.
Fuchs, V. R. (2004). Reflections in the socio-economic correlates of health. Journal of
health economics, 23(4), 653-661.
Harris, A. (2009). “Diabetes, Cardiovascular Disease and Labour Force Participation
in Australia: An Endogenous Multivariate Probit Analysis of Clinical Prevalence
Data.” Economic Record, 85(271),472-484.
Kahn, M. E. (1998). Health and labor market performance: the case of diabetes.
Journal of Labor Economics, 16(4), 878-899.
Knapp, L.G., & Seaks, T. G. (1998). A Hausman test for a dummy variable in probit.
Applied Economics Letters, 5(5), 321-323.
Kraut, A., Walld, R., Tate, R., & Mustard, C. (2001). The impact of diabetes on
employment and income in Manitiba, Canada. Diabetes Care, 24(1), 64-68.
Latif, E. (2009). The impact of diabetes on employment in Canada. Health
economics,18(5)577-589.
Lavigne, J. E., Phelps, C. E., Mushlin, A., & Lednar, W. M. (2003). Reductions in
individual work productivity associated with type 2 diabetes mellitus.
Pharmacoeeconomics, 21(15), 1123-1134.
78
Lin, S. J. (2011). Estimating the impact of diabetes on employment in Taiwan.
Economics Bulletin,31(4),3089-3102.
Loureiro, M.L., Sanz-de-Galdeano, A., & Vuri, D. (2010). Smoking habits: like father,
like son, like mother, like daughter?. Oxford Bullentin of Economics and
Statistics, 72(6), 717-743.
Minor, T. (2011). The effect of diabetes on female labor force decision: new evidence
from the National Health Interview Survey. Health economics,
20(12),1468-1486.
Ng, Y. C., Jacobs, P., & Johnson, J. A. (2001). Productivity loss associated with
diabetes in the US. Diabetes care, 24(2), 257-261.
Sammelson, P.A. and Nordhaus, W.D.(2010),Economics ,19th edition, New York:
McGraw-Hill Inc
Samuelson, P. A., Nordhaus, W. D.,Economics. (19th ed.).America:
McGraw-Hill/Irwin.
Seuring, T., Goryakin, Y., & Suuhrcke, M. (2015). The impact of diabetes on
employment in Mexico. Economics & Human Biology, 18,85-100.
Shelton Brown, H., Pagan, J A., & Bastida, E(2005). The impact of diabetes on
employment: genetic IVs in a bivariate probit. Health economics,14(5),537-544
Tunceli,K.,Bradley,C.J.,Nerenz,D.,Williams,L.K.,Pladevall, M.,& Lafata, J.E.
(2005).The impact of diabetes on employment and work productivity. Diabetes
care,28(11),2662-2667.
Vijan, S., Hayward, R. A., & Langa, K. M. (2004). The impact of diabetes on
workforce participation: results from a national household sample. Health
services research, 39(6p1), 1653-1670.
Zhang, X., Zhao, X., & Harris, A. (2009). Chronic diseases and labour force
participation of Australia. Journal of health economics, 28(1), 91-108