||Population issues affect the intensity of development in a country or region. Migration can also influent both economic development and resource allocation. In this study, I used Kaohsiung City as the main target to conduct the descriptive analyses with population data over the years and change of profiles. Otherwise, I tested the related variables (population migration, salary level, rental of premise, health care and transportation) of the phenomenon’s of the migration in Kaohsiung and whether if it was gathering or not by correlation and spatial autocorrelation systems. Before Japanese colonial period, the population and distribution of Kaohsiung showed high natural increasing, because of low mortality rate, and gathered in one place with primary industrial sectors. Along with transportation construction and urban planning, the population was urbanized in Yan-Cheng District.|
After World War II, many Chinese people retreated to Taiwan, which brought about a new space structure for Kaohsiung multitudes. Moreover, northern Taiwan not only retreating military and civilians, but setting up the central competent authorities were both attracting population immigration into. From 1960, industrial policies attracted many people to move to Kaohsiung. At the same time, the high natural increase rate made the population growth. Because of that, Kaohsiung gained the population from 1 million to 2 million in 18 years. Such rapid growth in population had become negative growth state because of the low natural increase rate and loss of population.
The population of Tainan City and Pingtung County are the most migration from Kaohsiung City in recent years. It presented that the major migration pattern in Taiwan is a short distance migration. I conduct Pearson Correlation Coefficient Analysis with the variables I mentioned before. The result showed that only salary level was significant. However, if I separated city and country into two variables, only country showed significant with salary level. Besides, I used spatial autocorrelation systems of Getis. I found that migration was highly correlated with spatial autocorrelation, which showed that migration would affect each other.