以企业为主体的科技创新效率评价是衡量科技投入后的产出水平、资源利用效率并寻找影响科技创新效率的主要因素，为优化资源投入结构打下基础。以企业作为科技投入和产出的主体，通过数据包络分析（DEA）中CCR模型和BCC模型测算中国15个副省级城市科技创新的总体投入产出效率、纯技术效率和规模效率。通过DEA-t检验的方法萃取出显著影响效率的关键指标，找到不同副省级城市各自效率偏低的关键问题。创新与特色一是将保留所有指标的DEA效率测算模型作为基础模型，将逐一剔除单个指标后的DEA效率测算模型作为对比模型。通过t检验，依次检验基础模型与每个对比模型的总体投入效率TE是否存在显著差异。若差异显著，则对比模型中剔除的指标为显著影响效率的关键指标，由此确定了影响科技创新效率的五个关键指标。改变现有研究的数据包络分析（DEA）方法仅仅能测算效率、无法萃取影响效率的关键因素的弊端。二是实证结果表明：显著影响以企业为主体的科技创新效率的关键指标为规模以上工业企业R&D经费内部支出、财政科技经费投入、科技从业人员数、技术市场成交额、规模以上工业企业新产品产值。三是实证结果表明：不同副省级城市导致效率偏低的关键问题大相径庭。如杭州的关键问题是“财政科技经费投入”的冗余率过大，而哈尔滨的关键问题是“规模以上工业企业新产品产值”的产出不足超高。
The efficiency evaluation of technology innovation is aimed to measure the output of technology, efficiency of resource utilization centered on enterprise and find out important factors of technology innovation efficiency. The enterprise is the main part of technology input and output in this research. Data Envelopment Analysis is used to do empirical analysis of 15 sub-provincial cities in China. The pure technology efficiency and scale efficiency is measured to determine the scale is big or small. By screening the important factors that signally influence technology efficiency, the key problem of every sub-provincial city is find out. The innovations and characters: firstly, in this paper the efficiency evaluation model included all indicators is the basic model and the efficiency evaluation model without one indicator Xi is the comparison model. By t test, it is verified if the efficiency difference between basic model and comparison model is significant. If the difference is significant, the indicator Xi is the key factor. It changes the existing research can't screen the key indicators by DEA. Secondly, empirical result shows that the key indicators include "R&D expenses of industrial enterprises above designated size" "financial funds to technology" "technology professionals" "new products output of industrial enterprises above designated size" "technical market turnover". Thirdly, empirical results show that the reasons why technology efficiency is low for different sub-provincial cities are different.
[1] 中国经济网. 十八大三中全会《决定》、公报、说明(全文).[EB/OL].[2013-11-18].http://www.ce.cn/xwzx/gnsz/szyw/201311/18/t20131118_1767104.shtml.
[2] 中国行业研究网. 2014年全国政协"一号提案"正式出炉.[EB/OL].[2014-03-03].http://www.chinairn.com/news/20140303/0919103.html.
[3] 杜娟, 霍佳震. 基于数据包络分析的中国城市创新能力评价[J]. 中国管理科学, 2014, 22(6): 85-93.
[4] 晏蒙, 孟令杰. 基于DEA方法的中国工业科技创新效率分析[J]. 中国管理科学, 2015,23(S1):77-82.
[5] 李洪伟, 任娜, 陶敏, 等. 基于三阶段DEA的我国高新技术产业投入产出效率分析[J]. 中国管理科学, 2012, 20(S1): 126-131.
[6] 范斐, 杜德斌, 李恒. 区域科技资源配置效率及比较优势分析[J]. 科学学研究, 2012, 30(8): 1198-1205.
[7] Sueyoshi T, Goto M. A use of DEA-DA to measure importance of R&D expenditure in Japanese information technology industry [J]. Decision Support Systems, 2013, 54(2): 941-952.
[8] Odeck J, Brathen S. A meta-analysis of DEA and SFA studies of the technical efficiency of seaports: A comparison of fixed and random-effects regression models [J]. Transportation Research Part A: Policy and Practice, 2012, 46(10): 1574-1585.
[9] Lafarga C V, Balderrama J I L. Efficiency of Mexico's regional innovation systems: an evaluation applying data envelopment analysis (DEA)[J]. African Journal of Science Technology Innovation & Development, 2015, 7(1):36-44.
[10] 蒋仁爱, 高昌林, 冯根福. 中国研究机构的科技资源利用效率评价研究[J]. 统计研究, 2013, 30(6):19-29.
[11] 戚湧, 郭逸. 基于SFA方法的科技资源市场配置效率评价[J]. 科研管理, 2015, 36(3):84-91.
[12] 肖文, 林高榜. 政府支持、研发管理与技术创新效率—基于中国工业行业的实证分析[J]. 管理世界, 2014,(4): 71-80.
[13] Xie Xin, Liu Fangfang. Analysis on regional technology innovation efficiency evaluation and influence factors in China[J]. International Journal of Business & Management, 2015, 10(3).
[14] Afzal M N I. An empirical investigation of the National Innovation System (NIS) using Data Envelopment Analysis (DEA) and the TOBIT model[J]. International Review of Applied Economics, 2014, 28(4):1595-1598.
[15] Hu Jinli, Yang C H, Chen C P. R&D Efficiency and the national innovation system: An international comparison using the distance function approach [J]. Bulletin of Economic Research, 2014, 66(1): 55-71.
[16] European Commission. Innovation union scoreboard 2014[R]. European Commission,2014.
[17] World Economic Forum (WEF).The global competitiveness index 2013-2014: Country profile highlights .World Economic Forum, 2013
[18] 中国科技发展战略研究小组.中国区域创新能力报告2011:区域创新与战略性新兴产业发展[M].北京:科学出版社,2011.
[19] 中国国家统计局. 2012年全国科技经费投入统计公报[R]. 国家统计局,2013.
[20] 中国技术市场管理促进中心. 2013全国技术市场统计年度报告[R]. 中国技术市场管理促进中心, 2013.
[21] 马占新. 数据包络分析模型与方法[M]. 北京:科学出版社, 2010.
[22] 高惠璇.应用多元统计分析[M]. 北京:北京大学出版社, 2005.
