At the end of Chapter 1, we concluded that measuring index system and explanatory index system of GUCI are both composed of a number of indicators. Therefore, it is of some difficulty to test the correctness of the index systems and to analyze factors affecting global urban competitiveness of specific cities and their significance in affecting the results. In this study, GUCI is integrated by a number of non-linear weighed indicators, including GDP, economic growth rate, GDP per capita, GDP per square kilometer, productivity, employment rate, price advantage indicator, patent applications and the presence of transnational companies. Specifically, the GUCI system consists of 7 level-I indicators, including enterprise, human resource, industry structure, soft environment, hard environment and global connectivity (as well as 40 level-II and 105 level-III indicators). To test the rationality of the GUCI indicators, linear and non-linear F-tests and t-tests of the 152 indicators of level-I/II/III were conducted using the GUCI. Both tests got consistent results. Here we would focus on the linear test only. As the comprehensive competitiveness of each city is obtained through the combination of a number of indicators and there might be possible error in the process, linear and non-linear F-tests and t-tests of the 152 indicators of level-I/II/III were conducted again, using the 9 GUCI component indicators. If any one of the 152 explanatory indicators passes the correlation test of all 9 measuring indicators, that indicator is relevant to the comprehensive competitiveness of the city.
The results show that, by competitiveness as a dependent variable, only 22, or less than 15%, of the 152 explanatory indicators failed the F-test. By the 9 component elements, only 7, or less than 5%, failed the F-test. More specifically, only 1 level-II indicator and none of the level-I indicators failed the test. The t-test showed the same result. Table 3.1 provides some results of the F-test and the t-test. Essentially, it indicates that our sequencing and indicator system design are correct.
Theoretically, the performance of enterprises is supposed to have significant impact on the comprehensive competitiveness of the cities. However, as only short-term return on equity and profit growth were available, and these data were subject to the volatilities of the national and international economies and financial markets, the indicators of return on equity and enterprise performance failed the test. In the future, we would consider data of a longer time span for our study. Theoretically, labor is the foundation of economic development and the competitiveness of each city. However, the number and scale of simple labor is having less and less influence on the competitiveness of a city, as is the population of the city. Labor and population passed the integrated multi-indicator tests, but failed the tests by individual GUCI indicators. Theoretically, living environment has significant impact on the competitiveness of a city. However, it failed the test by dining, lodging, and culture and entertainment elements. The data of the indicator was sourced from an online survey. In the future, we are going to collect higher-quality data for indicators of this class. An effective exchange rate, costs incurred by terrorism, and closing of businesses are critical to the competitiveness of a city. As these three indicators are based on data from World Bank and World Economic Forum, discrepancy may occur. Although the data indicates that regional autonomy has little to do with the competitiveness of a city, we insist that the expansion of autonomy has positive significance on the competitiveness of a city. We believe that water transportation connectivity is critical to the competitiveness of a city, but failed to obtain support from the data.
——From Global Urban Competitiveness Report(2007-2008)，Pengfei Ni with Peter Karl Kresl