Population distribution in the Republic of Kosovo: a comparative analysis on urban population and its classification based on
administrative and non-administrative criteria
Idriz Shala, Burim Limolli (Kosovo Agency of Statistics, Kosovo)
Statistical data for urban and rural areas are of some considerable importance for the central government and for local authorities while planning and managing services for local communities. For instance, the allocation of health and social care funding, housing, roads, water and sewerage and the provision and maintenance of schools have all distinctive aspects in urban and rural areas. Employment for urban and rural population has different features as well.
Recently, in many countries, including the Balkan countries, this distinction has become unclear and the principal difference between urban and rural areas in terms of the circumstances of living tends to be a matter of the degree of concentration of population. Indeed, rapid urbanisation processes have greatly raised the need for actual information related to different sizes of urban areas, and to the need to define standards for data comparability.
Like in all countries conducting a population census, in Kosovo the census data was disseminated following the administrative structure of the countries based on census legislation. In most of the Western Balkan countries, the law classifies the administrative units as urban or rural. However, like in many other countries, their cities and other urban areas are usually enlarging their size faster than the capacity of the law to revise such definitions which are needed to make urban boundaries consistent with the actual size of urban and non-urban areas. Therefore, the breakdown of census data by urban areas, both at national and regional level, is underreported. As a consequence, census results show significant differences in terms of urban/rural breakdown if different criteria for data classification are applied. Moreover, taking into account that the 2011 censuses have been used to update the sample frame for household surveys, also their results are affected by the definitions used for urban and rural population.
The main objective of this paper is to analyse the distribution of the urban resident population of Kosovo, as obtained from their 2011 population and housing censuses, according to administrative criteria and on the basis of a new approach for data classification as well, and to compare the differences at national and regional levels. Selected census variables are tabulated by urban/rural modalities using administrative and non-administrative criteria, including the 1 km² grid-based typology recently adopted by the European Union.
Official grid-based statistics : regional statistics in Andalusia (Spain)
Iria Enrique Regueira (Institute of Statistics and Cartography of Andalusia (IECA), Spain)
The Institute of Statistics and Cartography of Andalusia (IECA), official body member of the Regional Government of Andalusia, is the result of the merge in 2011 of the Statistical Office of Andalusia and the Institute of Cartography of Andalusia. The IECA has worked ever since on the integration and synergies of these two areas: Statistics & Geography.
Namely, IECA is engaged in a long-term project to integrate the grid of cells sized 250_250 m as a standard in statistics production procedures and data dissemination. Hereunder we share some of the milestones, lessons learnt and future goals of this project.
In 2013 “A population grid for Andalusia. Year 2013” was published, a first hybrid approach to population georeferentiation (top-down & bottom-up). A second edition of “A population grid for Andalusia Year 2013” was released in 2014 improving georeferentiation, thanks to 2011 building census data, and achieving 97.2% bottom-up georreferentiation of the population settled in Andalusia. This edition built entirely from a bottom-up approach provided us with a benchmark to review and validate the initial hybrid approach. Lessons learnt from 2013 grid editions and improvements in GIS infrastructure integration at IECA lead our current work on population grids. A shared PostGis Infrastructure is being built in order to share and consult updated unique data (population register, cadastre, census…) and initial linkage results for 2014 population register attained georeferentiation of 94% of populated buildings, standing for 95% of the population settled in Andalusia on 1st of January 2014.
Furthermore, since 2013 first edition, population grid data has been required and proved to be highly valuable for administrative planning & evaluation, namely regional educational planning or infrastructure planning. Additionally, grid standard has also been used to elaborate derived indicators such as Bicycle lane accessibility or Smoothed Standardized Mortality ratio by grid Cells, recently published by IECA.
Finally, IECA has also initiated georeferentiation of establishments, in order to provide a supplementary view of disaggregated activity and employment in Andalusia.
Capturing the Synergy of Geospatial and Statistics : A Singapore’s Perspective
Ng Siau Yong, Angelinie Winarto (Singapore Land Authority, Singapore)
Globally, it is acknowledged that the integration of statistical and geospatial information aid to improve the applicability of evidence-based decision-making.
Statistics are used to understand and make decisions on complex economic, social, security and environmental issues. Spatial information is crucial to derive patterns not readily apparent to the observer by using Geospatial Information System and Technology (GIST). However, to get a holistic view of the economy, society, security and environment, integrating spatial information with statistical information and processing the data by spatial analysis methodology is essential.
Through the spatial-analysis approach, and methods of coordination, harmonisation and collaboration between statistical and geospatial establishments, the observer is able to make sound decisions by finding clarity from complexity, to tackle some of the world’s most pressing challenges (i.e. transportation, security, healthcare and housing).
In 2012, the Secretary General of the UN Economic and Social Council stated: “The work on global geospatial information management over the past two to three years has confirmed that one of the key challenges is better integration of geospatial and statistical information as a basis for sound and evidence-based decision-making”.
Some Department of Statistics in the world use GIST for the purpose of census and demographic mapping and for updating census and demographic maps. There is, however, an increased acknowledgement that spatial analysis is important not just for national statistics, it provides a structure for collecting, collating, processing, storing, aggregating, generalising, disseminating and analysing operational and business data.
Tobler’s First Law of Geography states that “everything is related to everything else, but near things are more related than distant things”. We should be mindful of the power of location and statisticians are aware that the harmonisation of statistics, location and boundary delineation could affect sampling and therefore their analysis and results. By linking people, business, economy and environment to a location, the outcome is a better understanding of the social and economic issues of that location.
Today, there is an increased demand for geospatial analysis of the socio-economic data by various users. By integrating geospatial and statistical data and processing the data via spatial analysis, it facilitated the building of complex multidimensional location-based information resources. This could potentially generate unique visualisation, insightful spatial analysis and valuable predictive modelling results in the form of a map. The integration of statistical and geospatial data lead to cost savings, greater credibility, increased accuracy and better decision-making by the various users.
This presentation aims to share Singapore’s perspective on the postulations above. Experience in working with various public agencies in integrating geospatial and statistics for service delivery and policy decisions will be presented. Some Asian examples of the use of geospatial and statistics will be discussed.
Enhancing reliability of soil sealing indicators by use of geostatistical modeling
Patrick Sillard (Observation and Statistics Service - Ministry of environment, France)
Soil sealing (imperviousness) indicators are among most important indicators used to follow the extension of soil consumption. They are useful to monitor the pressure of human activity put on soil and at the end on biodiversity. Various sources are available to compute Soil sealing indicators: CORINE Land Cover and especially the High resolution layers, LUCAS survey, highly accurate geographical databases such as the property tax geographical database. All these data are acquired with different methods of observation and at different working scales. They lead, without further analysis, to very different rates of imperviousness. For example the rate of imperviousness estimated from the CORINE land cover geographical database is 50% less than the one we may compute from LUCAS.
The goal of the paper is to develop a statistical model of the imperviousness seen as a continuous spatial stochastic process that makes it possible to conciliate the various observations. As an application, the High resolution layers of CORINE and the LUCAS Survey are compared on the French territory. In particular, we study the consequence of the specific structure of the process (autocorrelation, long memory) on the imperviousness indicators and their associated variance.
A Room with a View or Rear Window ? The Demand for Housing Attributes in Paris
Mathilde Poulhes (Ministère de l'Environnement, de l'Energie et de la Mer, France)
This paper estimates buyers’ preferences for dwelling attributes and neighbourhood characteristics. The collected data allows for the simultaneous consideration of a wide range of intrinsic characteristics, such as building type, surface, floor, etc., and environmental characteristics, including noise, crime, school quality, distance to jobs, etc. The marginal willingness to pay is identified from transaction data under the hypotheses of the hedonic model described by Rosen (1974). Estimation is achieved by using flexible semi-parametric methods. Characteristics explain more than 90% of the variance of dwelling prices, showing a positive marginal willingness to pay for job accessibility and school quality and a weaker but significant negative marginal willingness to pay for a higher crime rate in the area. By contrast, noise level or public transport accessibility have less influence on housing prices. These results are robust to the inclusion of census tract fixed-effects, which also drastically reduces the spatial correlation of the residual prices from 0.10 (significant at 1%) to an insignificant correlation of 0.001.