文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 自然科学 > Economic Growth and the Expansion of Urban Land in China
侵权投诉

Economic Growth and the Expansion of Urban Land in China

Economic Growth and the Expansion of Urban Land in China

816 XIANGZHENG DENG ET AL.

2. Data

2.1 Creating the Sample

One of the most onerous tasks in preparing the data was to create a set of county-level observations that were consistent over the time of the study. The problem of consistency of county-level units over time arises because of changes in jurisdictional areas in China’s Shanghai and Chongqing; provincial capitals; and other large cities), the districts within the city’s administrative region were combined into a single, sample-period-consistent obser-vational unit. In this way, we ended up with a sample that includes 2348 observational units (excluding Taiwan, Hong Kong and Macao) at the county level that are consistent in size and jurisdictional coverage over time. In the rest administrative regions. The boundaries of some counties change over time. In other cases, towns/townships in a single county are divided into two groups and made into two counties between the mid 1980s and late 1990s. Occasionally the city core of a county is removed from the jurisdiction of the original county government and becomes an inde-pendent county-level administrative unit.Because of these changes, the number of counties rose over the study period. For exam-ple, in 1988 China had 2156 administrative units at the county level, whereas in 2000 the number had expanded to 2733.3 The organi-sational shifts of county-level administrative units are problematic for this study since data within each county observational unit need to be comparable over time.

In order to overcome this problem, we use the geo-coding system of the National Fundamental Geographical Information System (NFGIS, 2000) and a 1995 adminis-trative map of China from the Scientific Data Centre of the Chinese Academy of Sciences, which included a consistent geo-coding sys-tem with that of NFGIS. Using these tools, if two counties had been subject to border shifts (for example, one county ceded jurisdictional rights to another), we combined them into a single unit for the entire sample period. In cases in which the city core of a county had been removed from the jurisdiction of the original county-level government, we reag-gregated the municipal administrative zone

back into the county proper. In the case of large metropolitan areas (i.e. China’s four pro-vincial-level municipalities—Beijing, Tianjin, of the paper, even though the observations will include municipality districts, cities and other administrative units that are larger (and more complex) than counties, for brevity we call all observations county sampling units (or simply counties).

Because spatial analysis cannot be carried out effectively when there are missing data, we had to drop several provinces from our analysis. The basic problem of missing values in our spatial econometric approach is that, if just one county has a missing value, then some or all of the spatially weighted variables can-not be created. Ignoring this problem results in an errors-in-variables bias (Cressie, 1993). It is less of a problem in ordinary least squares (or traditional regression) estimation since most programmes drop the observations for which a variable has a missing value.

In the case of our analysis, the problem arises due to missing observations on the economic variables. After looking in as many published sources as possible (both national and provincial yearbooks and statistical com-pendia), the economic data were not complete for a number of counties in Xinjiang, Tibet, Qinghai and Gansu. In the case of the rest of the provinces, we were able to create a com-plete series of county-level data for 1995 and 2000 for all counties. After eliminating the counties from Xinjiang, Tibet, Qinghai and Gansu from the analysis, we are left with a final sample size that includes 2063 counties.4

2.2 Land Use Data

One of the strengths of our study is the nature and quality of the data that we use

上一页第4页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email