中文摘要：

通过构建由技术创新、广告强度和市场力量三个方程组成的联立模型,以中国工业37个细分行业2006年数据为样本,从产业层面对市场力量和技术创新之间的内生性问题进行实证检验.结果表明：市场力量与技术创新存在双向因果关系,市场竞争有利于企业创新活动的开展,而创新能力的增强能够加剧行业竞争程度;广告强度与技术创新之间也是相互影响的,广告支出多的企业倾向于开发新产品,创新能力的提升则会带动广告投入以显示产品差异性;产权结构对技术创新存在显著的负效应,国有企业更缺乏技术创新动力,且行业规模越大创新活动越频繁.根据市场力量方程的参数估计,我国多数工业行业仍处于＂贝恩标准＂下的低集中型寡占市场,行业资源有待于进一步优化配置.与单方程回归结果的对比显示,联立方程模型能够更加准确地反映市场力量与技术创新之间的互动机制,避免实证结论的偏差.

英文摘要：

By constructing the model composed of three simultaneous equations including technological innovation, advertising intensity and market power, and using the data from 37 segments of Chinese industrial sectors in 2006 as sampies, an endogeneity empirical test is made between market power and technological innovation from the level of industry. The results show that there is a bilateral causality relationship between market power and technological innovation, which means that market competition is conducive to business innovation activities, and enforcement of innovative ability can enhance the degree of the industrial competitive intensity. Moreover, the intensity of advertising and technological innovation are also interactive. Enterprises with more advertising expenditure have more intention to develop new products, while the enhancement of innovative ability will drive the investment into advertising which can show products differentiation. Property rghit structure has a significant negative effect on technological innovation, which means the state-owned enterprises are more lack of motivation to innovate, and the larger industry is, the more frequent innovation activities will be carried out. According to the parameters estimated by market power equation, the majority of Chinese industries are still in low-concentrated oligopoly market based on the ＂Behn standard＂, and industry resources need to be further optimally allocated. In addition, compared with the regression results of single-equation, simultaneous equation model could more accurately reflect the interactive mechanism between market power and technological innovation, so as to avoid deviations of empirical conclusions.