运筹与管理 ›› 2025, Vol. 34 ›› Issue (5): 156-163.DOI: 10.12005/orms.2025.0157

• 应用研究 • 上一篇    下一篇

创新价值链视角下建筑业上市公司技术创新效率研究——基于超效率网络SBM模型和BMA方法的实证分析

程敏, 易小凤, 王方亮   

  1. 上海大学 管理学院,上海 200444
  • 收稿日期:2023-01-28 发布日期:2025-08-26
  • 通讯作者: 程敏(1977-),女湖北荆门人,博士教授研究方向:建筑经济与管理。
  • 作者简介:程敏(1977-),女,湖北荆门人,博士,教授,研究方向:建筑经济与管理。
  • 基金资助:
    国家社会科学基金资助项目(22BJY232)

Research on Technological Innovation Efficiency of ListedConstruction Companies from Perspective of Innovation Value Chain:An Empirical Analysis Based on Super-efficiency Network SBM Modeland Bayesian Model Averaging

CHENG Min, YI Xiaofeng, WANG Fangliang   

  1. School of Management, Shanghai University, Shanghai 200444, China
  • Received:2023-01-28 Published:2025-08-26

摘要: 为了解建筑业上市公司技术创新效率及其影响因素,基于创新价值链视角,将超效率网络SBM模型和DEA窗口分析法相结合对2016—2020年我国48家建筑业上市公司的技术创新效率进行测度,采用贝叶斯模型平均方法分析其影响因素。研究结果表明:(1)研究期内各年48家企业技术创新效率均值介于0.50~0.54之间,技术创新效率有待提升;(2)根据技术研发效率和成果转化效率将样本企业分为四类,8家企业属于高效集约型、12家企业属于低研发高转化型、9家企业为高研发低转化型、19家企业为粗放低效型;(3)成立年限、成长能力和盈利能力对建筑业上市公司技术创新效率有显著的正向影响,企业规模、研发财力资源投入强度、政府扶持以及研发人力资源投入强度对其有显著的负向影响。最后依据研究结果提出了效率改善的建议。

关键词: 创新价值链, 技术创新效率, 建筑业, 超效率网络SBM模型, 贝叶斯模型平均

Abstract: As the development of China’s construction industry transforms from investment-driven to innovation-driven, construction companies have paid more and more attention to technological innovation and have invested a lot of capital and manpower in it. However, the input efficiency of innovation resources needs to be clarified. Technological innovation is a multi-stage process including the input and output of resources and the transformation of achievements. The innovation value chain (IVC) embodies the transfer and value addition of various resources across multiple stages. Studying the technological innovation efficiency (TIE) from the perspective of the IVC helps to understand the utilization efficiency of innovation resources at each stage.
In this study, 48 listed construction companies (LCCs) are selected as research samples for empirical analysis. Firstly, from the perspective of the IVC, the technological innovation process of LCCs is divided into two stages: technology research and development (R&D) and achievement transformation. The efficiency evaluation index system of the two stages is constructed respectively. The number of R&D personnel, R&D cost, and net fixed assets are selected as the inputs of the technology R&D stage. The number of patent applications is chosen as the output of the technology R&D stage as well as the input of the achievement transformation stage. The number of non-R&D personnel and operating costs are chosen as the additional inputs of the achievement transformation stage, and the net profit and operating income are taken as the outputs of the achievement transformation stage. Secondly, the efficiency measurement model is constructed by combining the super-efficiency network slacks-based measurement (SBM) model and the data envelopment analysis (DEA) window analysis method and is used to measure the TIE of China’s LCCs from 2016 to 2020. Finally, 17 factors influencing TIE are selected from four aspects including internal corporate governance, corporate operation, resource allocation, and external environment. The impact of these factors on LCC’s TIE is analyzed by using the Bayesian model averaging (BMA) method.
The main research results are as follows. (1)In terms of the overall efficiency, the average value of TIE of the sample companies in the study period is between 0.50 and 0.54, and the overall efficiency still needs to be improved. (2)In terms of the sub-stage efficiency, the average value of the transformation efficiency of sample companies is higher than that of the R&D efficiency, which indicates that R&D efficiency constrains the improvement of the TIE. Besides, there are 8 companies with high R&D and transformation efficiency, 12 companies with low R&D efficiency and high transformation efficiency, 19 companies with low R&D and transformation efficiency, and 9 companies with high R&D efficiency and low transformation efficiency. LCCs should improve the efficiency of both phases from an IVC perspective to improve the overall TIE. (3)Seven key factors affecting the TIE of LCCs are obtained based on the BMA method. Among them, years of establishment, growth ability and profitability have significantly positive effects on the TIE of LCCs during the study period, while company size, R&D cost intensity, government support, and R&D staff intensity do not.
The efficiency measurement index system and evaluation model of LCCs in this study are constructed from the perspective of the IVC, which can reasonably evaluate the TIE of LCCs, find the weak parts of technological innovation, and provide a basis for managers to understand the current situation of technological innovation and optimize technological innovation strategies of companies. The research method can also be used to measure the TIE of companies in other industries.

Key words: innovation value chain, technological innovation efficiency, construction industry, super-efficiency network SBM model, Bayesian model averaging (BMA)

中图分类号: