Spatial Gini coefficient and locality entropy are used to estimate the aggregation degree and distribution of knowledge-intensive service industry and its sub-sectors in 41 prefectures in Yangtze River Delta from 2004 to 2018, and spatial Dubin model is applied to verify its impacts on the local economy and its spatial overflowing effect on the vicinity. Spatial Gini coefficient reveals that aggregation degree of commercial service sector is the largest. The aggregation index is at 0.04－0.06, generally in a U-shaped trend. Locality entropy of less than one-third cities is larger than 1 with Shanghai as center, Hangzhou, Nanjing and Hefei as sub-centers, while other cities are of low aggregation degree. Spatial Dubin modelling results show a remarkable spatial overflow effect on economy, varying with sectors. Financial service sector aggregation has constraints on the local economy, but little on the vicinity,while information technology and commercial service sector do not have a significant influence on the local economy, but have siphoning effect on the vicinity. Scientific servicing has a little influence. Generally, knowledge-intensive servicing industry aggregation largely promotes the local economy and vicinity in Yangtze River Delta,varying with sectors and regions. This paper presents suggestions on promoting a sound, balanced knowledge-intensive service industry to avoid industrial simplification, boosting it in the underdeveloped areas, optimizing its spatial aggregation and diminishing the gap between the core cities and remote cities.

Knowledge intensive business servicing plays a leading role in promoting economic growth and in establishing regional integrated network spatial pattern of “one-core-five-circles-four-zones” in Long Triangle. This paper uses attraction model and social network model to establish KIBS spatial correlation network of Long Triangle in 2011 and 2016, and applies network density, network centrality, concentration sub-group, core-rim structure to study the development level variance and structural evolution in Long Triangle. KIBS connection path in Long Triangle is in form of network with direction to the center. The industrial cluster does not produce good co-integration, but of potential. The traditional administrative barrier has been broken down, with a rising province-across cooperation. This paper presents suggestions on developing KIS in Long Triangle city group.