Song Rui: Data Processing of Global Urban Competitiveness Report

In view of the above data collecting channels, and the challenges and complexity in the collection, the following methods are employed for data processing:

For data directly available: unified processing

       For some indices, e.g., population and area, first-hand data are available in every city. However, these data might have been collected according to different standards. In such cases, we would first study the indices and standards of United Nations Statistical Division (UNSD), World Bank World Development Indices, OECD Database and other international organizations. Then we would determine an approach for the conversion of data of each country and set up the most proper, comparable and widely used statistical standards for data processing. Eventually, we were able to build a uniform database to cover the 500 international cities. With regard to population, for example, some cities only provide domiciliary population, some provide permanent population, and others include temporary population in their statistics. In our study, they are all converted into permanent population. For another example, the “area” might be land area only for some cities, and the sums of land and water areas for others. In our study, adjustments are made so that the area means land area only. Similar situations exist for many other indices, e.g., adult literacy rate, the proportion of people with higher education and crime rate, which are all adjusted with consistent standards.

Index data that can be calculated indirectly, or for which alternatives are available

       Data that can be calculated or for which alternatives are available are processed through the following approaches:

       Direct calculation of variables

       When some variable data are not directly available, we will calculate in accordance with strict logical relationship from two or more other relevant variable data. This involves three aspects. One is the reversible calculation between the equalizing value index and the total amount index. For example, a city’s GDP, GDP per capita, GDP per square kilometer as well as the labor productivity can be reversibly calculat­ed through such intervening variables as the city’s area, population and employed population. The second is the calculation of the variable static data and the dynamic data. For example, a city’s GDP growth rate can be calculated through the chronological data of its GDP. The third is the calculation between the index absolute value and proportion, such as the reversible calculation among number of the labor force, employed population and the unemployment rate. And also, the urban population above the college level can be calculated by calculating its proportion in the city’s whole population. Additionally, the proportion of foreign-born citizens and the proportion of for­eign tourists in the urban population, etc. can also be calculated in this way. The direct variable calculation method has been extensively used in our research. Due to its conformity to the strict logical relationship between the variables, the calculated variables are undoubtedly accurate on the condition that the existing variables are known to be accurate.

       Calculation of variables based on other relevant variables

       If some data cannot be obtained direct­ly, then they can be calculated according to their quantitative relations with the relevant varia­bles collected. For example, if we cannot obtain the accurate GDP information on a city, but can obtain its accurate GVA data, then we can calculate the country’s or the city’s GDP in ac­cordance with its similar quantitative relationship with its GVA. This method has mainly been adopted in GDP data processing in the British cities, as well as some other European cities.

       Estimation of variables

       Since this is a method of estimation, the data obtained in this way are less accurate than those obtained by the above two methods. It is the calculation of the city’s variables with other relevant knowledge or experiences on the basis of the relevant variables collected. This method has been widely used, albeit not often. That is, it can almost be applied in the data processing of all the index systems, but only a few cities adopt it in their data pro­cessing. For example, as the GDP data of some cities in South America and Africa are hard to obtain, we can only refer to the GDP data of its country or other cities in its country, or even in other countries, and then estimate the GDP data of this city on the basis of the relevant infor­mation or sometimes the researcher’s experience. Other examples can be found in the data of various index systems of several cities.

       Substitution of variables

       A city is a component of its superior administrative region, so the relevant variable data of its superior administrative region is either the same as (such as some policies or systems of a country or a region, which are also applied to the cities under it), or very relevant to its own variable data. This method has been widely used in our research, such as getting credit, effec­tive exchange rate, difference of deposit and loan, the national technical infrastructure in the hard environment index, the ratio of local revenue to the national revenue, and all indices of eco­nomic liberalization, market supervision and all of the indices in the tax burden in the soft environment index, as well as the cost from terrorism in the living environment. All of these index data are constructed on the national level. Besides, there are some cities, whose other index data are substituted with data of the region or the province in which they are located.

 

Estimation of variables based on comparisons

       Estimations are made in accordance with government data and the positions of particular cities in their respective countries, as well as the performance of similar cities. For some indices, particularly those released by research and consultancy institutions, first-hand data are available in most cities. For example, Mercer World Cost of Living City Rankings, the number of management and hi-tech professionals in every 1,000 people in accordance with World Knowledge Competitiveness Report, World Bank World Development Indices’ carbon dioxide emission, wastewater treatment rate and particles, and indices of the Chinese travel organization CTRIP (www.Ctrip.com) about shopping, dining, lodging and entertainment cover most cities. However, the data for some of the cities are not available. In such cases, relevant data about these cities are compared with other cities to get estimated data for the indices.

Data that are not directly available and cannot be calculated and no alternatives are available

       For some indices, e.g., those concerning the distribution of industrial links, urban functions, city management and competitiveness of enterprises, no objective data are directly available, nor is calculation on the basis of other relevant data viable. In such cases we would substitute the data of typical samples as data of these indices.  For example, of we cannot compare average profit of all corporations in the 150 cities, we compare profit of the largest corporation in each industry for each city.

 

Scoring of alternatives

       Scoring of alternatives is a method for obtaining and scoring of alternatives for particular indices, for which no direct or indirect data are available. These alternatives must be easily accessible and able to reflect the indices to a high proximity. Scoring of alternatives proves to be an effective solution to complicated situations where key data are not directly available. By properly selecting the alternatives, it could reflect the original variables truthfully.

(1) Scope of application

As an important research method, in this research report, the substitution scoring method is mainly used to the design, research and analysis of the substitute indices for the urban industry structure. Due to the historical as well as the realistic factors, cities around the world have different development lev­els and complicated industry structures as well as industry distribution. Therefore, from the point of view of statistical analysis, it is very hard to obtain adequate data support to carry out a comprehensive research on them. Through the analysis, we can conclude that the urban industry structure will be ultimately reflected by the distribution and aggregation of the enterprises in different industries. Therefore, in this re­search report, the substitution scoring method has been adopted during the design of the index system of the urban industry structure, i.e. to design the indices that can approximately reflect the status quo of the city's industry structure in accordance with the urban distribution of the multinational corporations in different industries. The subentry competitiveness indices in the index system of the world urban competitiveness that used the substitution scoring method include: the controlling ability of international economy in the perform­ance index system, the indices in the industry structure system which reflect the status quo of the service industry's competitiveness such as the number of manufacturing multinational corporation headquarters, the number of multinational wholesale and retail corporations, the number of multinational commerce serv­ice corporations and the number of multinational advertising and media corporations, etc, the indices which reflect the competitiveness status quo of the financial industry such as the multinational financial corporation headquarter distribution and the multinational financial corporation branch distribution controlling ability of international economy and the controlling ability of international financial economy branch, etc. as well as the indices which reflect the competitiveness status quo of the high-tech industry such as the number of multinational software service corporation headquarters and the number of multinational high-tech corporation headquarters, etc.

(2) Scoring Criteria and Principles

Another important aspect of the substitution scoring method is to score the substitutes according to certain criteria. During its concrete application in the index design, and in accordance with the global network configuration and distribution characteristics of the multinational corporations around the world, the following scoring criteria will be observed: 1) the city where the multinational corporations' global headquarters congregate (five points); 2) the city where the multinational corporations' regional headquarters congregate (four points); 3) the city where the multinational corporations' national headquarters congregate (three points); 4) the city where the multinational corporations' branches congregate (two points); 5) the city where the multinational corporations' agencies (i.e. the small-scale branches with limited functions) congregate (one point). The above five items make a basic scoring criterion, while during the concrete operation, due to the unclear information provided by corporations or the differ­ent configurations of multinational corporations' global network, it is very hard to judge directly the scores of the multinational corporations' branches. In such a case, we make the subsidiary judgment mainly from two aspects: one is to search online and decide the status of the multinational corporation's branches according to the relevant information collected in this way; and the other is to make the judgment according to the number and scale of the distribution of the multinational corporations’ branches in different cities. Generally speaking, in the same country, the city is superior to other cities in the global network of the corporation if it has the most or the largest branches of a multinational corporation; moreover, the function of the branches located in it are also superior to that of the corporation's branches in other cities. On the basis of combining these two aspects, if it is still not possible to make the judgment of a city with the obtained information, then it will be given two points. After the scoring of the distribution status of the chosen multinational corporations in the same industry one by one, the marks of the substitute indices will be calculated by an equal-weight accumulation.

(3) The Sampling of the Multinational Corporations in Different Industries.

The aim of this research is to design the substitute indices. In order to better reflect the fundamental state of the urban industry structure so as to make a judgment of its industrial competitiveness, we have chosen the representative multinational corporations in such industries as general manufacturing, commer­cial service, trade, retail service, finance, high-tech, etc. for the analysis. In order to make the analysis results comparable, we have made the multinational corporation sampling in accordance with the rankings in each industry of the Forbes Global 2000.

—— From“Global Urban Competitiveness Report（2007-2008）”，Pengfei  Ni with Peter Karl Kresl