\documentclass[english,german]{scrartcl}
\usepackage[latin1]{inputenc}
\usepackage{geometry}
\geometry{verbose,a4paper,tmargin=1.75cm,bmargin=2cm,lmargin=2cm,rmargin=2cm,headheight=0.8cm,headsep=0.5cm,footskip=0.5cm}
\usepackage{babel}
\usepackage{url}
\urlstyle{same}
\newcommand\email{\begingroup \urlstyle{same}\Url}\usepackage{graphicx}
\usepackage{units}
\usepackage{tabularx}
\usepackage{wasysym} % enable use of \permil instead of \textperthousand
\usepackage{marvosym}
\usepackage{arabtex}
\usepackage{hyphenat}
\usepackage{booktabs}
\usepackage{mdwlist}
\renewcommand{\thefootnote}{\relax}
\usepackage{index}
\newindex{key}{kdx}{knd}{Keywords}
\newindex{author}{idx}{ind}{Authors}
\usepackage{fancyhdr}
\pagestyle{fancy}
\renewcommand{\headrulewidth}{0pt}
\fancyfoot{}
\usepackage{ifpdf}
\ifpdf
%\usepackage[activate]{pdfcprot}
%\usepackage[owner=Tapioka, print=true, copy=true, edit=false, annotate=false,set]{pdfcrypt}
%encryption: not supported any more (since pdftex 1.10)
\usepackage[pdftex]{thumbpdf}
\usepackage[pdftex]{hyperref}
\hypersetup{pdftitle={Development of Operational Poverty Indicators in Northern Viet Nam},pdfauthor={Thi Tuyet Van Dinh, Manfred Zeller},pdfsubject={Tropentag 2010: Abstract},pdfkeywords={Operatonal indicators, poverty, poverty assessment tools, Viet Nam},pdfpagemode=None,colorlinks=true}
\else
\usepackage{ae}
\usepackage{textcomp}
\fi
\begin{document}
\ifpdf
\parbox[b]{2.5cm}{\resizebox*{2.5cm}{!}{\includegraphics{dtt2010.jpg}}}
\else
\parbox[b]{2.5cm}{\resizebox*{2.5cm}{!}{\includegraphics{dtt2010.eps}}}
\fi
\hfill
\parbox[b]{13.4cm}{\centering \large{\textbf{Tropentag, September 14-16, 2010, Zurich}}\\[1ex] \Large{``World Food System  ---\\A Contribution from Europe''\\[2ex]}}
\hspace{0.1cm}
\begin{center}
\Large{\textbf{Development of Operational Poverty Indicators in Northern Viet Nam\footnote{\textbf{Contact Address:} Thi Tuyet Van Dinh, University of Hohenheim, Dept. of Agricultural Economics and Social Sciences in the Tropics and Subtropics, Schloß Osthof-Süd, 70599~Stuttgart, Germany, \mbox{e-mail}: \email{dinhthi@uni-hohenheim.de}}\\[0.8ex]}}
\normalsize{\textsc{Thi Tuyet Van Dinh, Manfred Zeller}}
\end{center}
\begin{itemize*}
\item[]{\small{\textit{University of Hohenheim, Dept. of Agricultural Economics and Social Sciences in the Tropics and Subtropics, Germany}}}
\end{itemize*}
\index[author]{Dinh, Thi Tuyet Van}
\index[author]{Zeller, Manfred}
\begin{center}
\vspace{1.1cm}
\textbf{Abstract}
\begin{abstract}
\normalsize{
To annually update poverty status of households in Viet Nam, local authorities define poverty status of households based on how knowledgeable neighbours are about each others, neglecting the income benchmarks developed by the Ministry of Labour, Invalids and Social Affairs (MOLISA) in 2005. This method, thus, is subject to manipulation which can be affected by political, administrative, and budgetary reasons. Therefore, more precise and practicable tools for identifying the poor are needed. 



The purpose of this paper is to compare and discuss the accuracy and practicability of three alternative tools, compared to the MOLISA tool. Two alternative tools are the IRIS and the Schreiner tools, using nationally representative samples in Viet Nam. The other tool has been developed by the authors, using household expenditure data of 300 households in northern Viet Nam. We investigates two research objectives: identifying alternative poverty indicators; testing predictive accuracies of the newly developed tool together with the three other tools. Our hypothesis is that tools calibrated for the nation will perform worse than tools specifically calibrated for certain smaller regions within a country.



A questionnaire contained many potential poverty indicators related to demography factors, education, housing characteristics, assets, and other facets of poverty. Four models were used, namely Ordinary Least Square, Quantile, Linear Probability Model, Probit, applying the MAXR procedure to yield best 10 variables in correctly predicting the poverty status of households.



Results show that the quantile regressions yield highest accurate performance among regressions regardless of specifying with all variables or with only non-monetary variables. Best derived poverty indicators are reliable and cost"=effective to measure poverty, allowing for higher accuracy performances. Provided, this is, that the new tools do not take into account the mentioned manipulation. The developed tools will be tested across time and space in a newly selected 200"=household sample together with the MOLISA and the other two tools to seek for a best local-specific tool in defining the poor in northern Viet Nam.







}

\vspace{0.5cm}
\end{abstract}
\end{center}
\noindent \textbf{Keywords:} Operatonal indicators, poverty, poverty assessment tools, Viet Nam
\index[key]{Operatonal indicators}
\index[key]{Poverty}
\index[key]{Poverty!assessment tools}
\index[key]{Viet Nam}
\end{document}