Abstract：At present, the spatial spillover of innovation output is mostly carried out at the provincial and municipal scales, but the research at county scale is less. In this paper, representing innovation output by patent licensing, the standard deviation, coefficient of variation, spatial autocorrelation and spatial econometric model were employed to study the differences of innovation output in counties of Jiangsu Province from 1986 to 2014. The results show as follows: (1) Before 2006, the absolute difference innovative output among counties of Jiangsu Province increased slowly, then rapidly widening, increased from 268.42 in 2006 to 7045.6 in 2012. The relative difference reached its highest of 2.91 in 2000, after that it began to decline. The relative difference fluctuated from time to times, but it showed downward trend overall. (2) The innovation output of Nanjing City showed a weakening trend during the Ninth Five-Year Plan. Innovation output in Suzhou City, Kunshan City, Wuxi City and South of Jiangsu Province were increasing rapidly. (3) The innovation output in counties of south Jiangsu, middle Jiangsu and north Jiangsu declined in gradient, and it showed the spatial pattern of high in north and low in south. (4) Most of the counties in north Jiangsu showed the“low-low”gathering and had a tendency to spread to middle Jiangsu. The counties of the“high-high”gathering were few and mainly concentrated in south Jiangsu. (5) Moran's I of the innovation output in the counties of Jiangsu province was 0.477, normal statistic Z-value was 4.794, and spatial interaction terms of coefficient was 0.609, so it shows that spatial interaction is the important reason for innovation output difference. (6) Regional innovation environment has a significant impact on the county's innovative output, But the role of government policy is relatively small, according to the regression coefficients which were 0.529 and 0.110, respectively. (7) In specific factors, there was no difference of regression coefficients between economic basis, R&D investment, R&D personnel, traditional infrastructure, communications infrastructure and FDI, which were 0.092, 0.09, 0.091, 0.089, 0.094 and 0.089, respectively. However, regression coefficient of science and technology policy was very small, which was only 0.012. To a certain extent, this study reveals the mechanism for the formation of innovation output difference, which has a revelation to explore the difference of county innovation output.
Keyword：innovation output; county; differential evolution; spatial autocorrelation; Spatial Econometric analysis;
【source】Arid Land Geography2017年01期
