促进区域协调发展，是贯彻新发展理念的重要内容。当前，在技术进步、结构转型等因素作用下，我国区域间发展不平衡呈现不断加速的趋势。作为城市创新体系的重要组成部分，创业孵化为初创企业提供物理空间、技术支持、创业辅导、投融资等孵育服务，提高了初创企业的存活率和绩效，不仅是城市扶持科技型中小企业发展、推动创新创业的重要载体，同时也是吸引创新型企业落地、优化产业布局的重要工具。然而，既有关于创业孵化空间配置及其对初创企业区位选择影响的研究还存在较大的缺口，无法对相关政策的制定提供有力的理论支撑，不利于创业孵化引导初创企业空间配置作用的发挥。在促进区域协调发展背景下，本文基于群分视角，研究创业孵化的空间配置及其对异质性初创企业区位选择的影响，不仅丰富和拓展了创业孵化的相关研究，而且对促进区域间产业分工协作、构建大中小城市协调发展格局、推动创孵行业高质量发展具有较强的借鉴意义。
本文由7章内容构成。第1章为绪论。第2章为文献综述。第3章是创业孵化空间配置影响初创企业区位选择的理论分析。第4章对创业孵化的空间配置进行实证研究，分析集聚效应和群分效应对创业孵化空间配置的影响。第5章进一步围绕创业孵化空间配置对初创企业区位选择的影响进行实证研究。第6章研究区域补贴对创业孵化空间配置，进而对异质性初创企业区位选择的影响。第7章为主要结论、政策启示与研究局限。本文研究的主要结论是：
第一，创业孵化的空间配置受集聚效应和群分效应影响，在集聚效应和群分效应作用下，创孵平台孵育效率随城市规模的上升而提高。本文研究发现，大城市创孵平台的效率溢价，90.12%来源于群分效应，集聚效应的贡献率为9.88%。这意味着：大城市中创孵平台较高的效率，主要是因为创孵平台群分导致大城市中创孵平台本身就具有较高的效率，集聚外部性对创孵平台效率的提升作用较小。这一研究结果表明：当前创孵平台的集聚效应偏低，创孵平台的资源整合能力需要进一步加强；同时，由于创孵平台具有较强的群分效应，因此与创孵平台相关的区域政策，应充分考虑较强的群分效应可能对政策效果造成的影响。
第二，创业孵化空间配置能够影响初创企业的区位选择，但其作用强度远弱于城市层面的集聚效应和群分效应。这意味着：通过区域政策调整创业孵化空间配置，缩小城市间创孵平台效率差距，能够引导初创企业的区位选择；但由于城市层面集聚效应和群分效应对初创企业区位选择的作用远强于创业孵化，所以应当在规模差距较小的城市间，发挥创业孵化对初创企业空间配置的引导作用。
第三，利用区域补贴调整创业孵化的空间配置，在小城市引入高效率创孵平台，将在小城市引入低生产率初创企业，高生产率初创企业仍将留在大城市。本文通过反事实分析发现，小城市对创孵平台的转移补贴，将吸引低效率创孵平台迁入小城市；而小城市位序补贴创孵平台，则能在小城市引入高效率创孵平台。但初创企业较高的集聚效应以及创孵平台对初创企业生产率较低的提升作用，造成即使小城市引入高效率创孵平台，仍然难以在小城市引入高生产率初创企业。
第四，与补贴创孵平台不同，在较高的补贴额下，小城市位序补贴初创企业能够在小城市引入高生产率初创企业，但将造成较大的生产率损失。本文通过反事实分析发现，小城市对初创企业转移补贴，将吸引低生产率企业迁入小城市；而小城市位序补贴初创企业，则能在小城市引入高生产率初创企业。但小城市较弱的集聚外部性将造成引入初创企业的生产率损失。这意味着，如果小城市利用区域补贴，同时引入高生产率初创企业和高效率创孵平台，则能够在引入高生产率初创企业的同时，一定程度上减少迁入企业由于集聚外部性降低而产生的生产率损失。
第五，对创孵平台孵育服务影响初创企业生产率的量化分析发现，观察到的高效率创孵平台中较高的初创企业生产率，95%来源于初创企业在城市层面的集聚效应和群分效应，创孵平台孵育服务的贡献率为5%。这意味着：大城市创孵平台内初创企业较高的生产率，主要来源于初创企业在城市层面的集聚效应和群分效应，创孵平台孵育服务对初创企业生产率的提升作用相对较小。这一研究结果表明：集聚效应和群分效应导致大城市初创企业本身就具有较高的生产率，因此在创孵平台绩效评价中，如果不充分考虑城市层面因素影响，或者为初创企业绩效赋予较高权重，将导致评价结果出现较大偏差。

Promoting coordinated regional development is an important element in applying the new development philosophy. However, the imbalance in development between regions shows a trend of continuous acceleration. As an important part of urban innovation system, business incubation provides physical space, technical support, entrepreneurial counseling, financing and other services for startups, which improves startups’ survival rate and performance. It is not only an important carrier for cities to support the development of startups and promote innovation and entrepreneurship, but also an important tool to attract startups. However, there are still large gaps in the existing research on the impact of business incubation on the location of startups, which cannot provide strong theoretical support for relevant policies. In the context of promoting coordinated regional development, this paper studies the influence of business incubation on the location choice of startups based on sorting theory, which not only enriches the related research on business incubation, but also has implications for building a coordinated development pattern of cities, and promoting high-quality development of business incubation.
This paper consists of 7 chapters. Chapter 1 is introduction. Chapter 2 is literature review. Chapter 3 explains the influence of agglomeration effect and sorting effect on the location choice of heterogeneous business incubators from theoretical perspective, and analyzes the influence of the incubation allocation on the location choice of startups. Chapter 4 presents an empirical study on the allocation of business incubators due to the agglomeration effect and sorting effects. Chapter 5 further conducts an empirical study around the impact of the allocation of business incubators on the location choice of startups. Chapter 6 examines the impact of regional subsidies on the allocation of business incubators and thus on the location choice of heterogeneous startups. Chapter 7 presents the main conclusions, policy implications and research limitations. The main findings of this study are:
First, incubator efficiency rises with city size, and the contribution of sorting effect is 90.12% while the contribution of agglomeratioin effect is 9.88%. This means that the higher efficiency of incubators in large cities is mainly due to the sorting effect that leads to the higher raw efficiency of the incubators in large cities, and the agglomeration effect has a smaller role in improving the efficiency of incubators. The results of this study also indicate that the agglomeratioin effect of incubators is low and the resource integration capability of the incubators needs to be further strengthened. Moreover, since the agglomeratioin has less effect on the efficiency of incubators, regional policies related to the incubators should not be oriented toward agglomeration and should take into account the possible effects of sorting.
Second, the spatial allocation of business incubation can influence the location choice of startups, but its effect is much weaker than the agglomeration effect and sorting effect. This implies that: adjusting the spatial allocation of business incubation through regional policies and narrowing the efficiency gap of business incubators between cities can guide the location choice of startups, but the role of business incubation in guiding the spatial allocation of startups is relatively limited compared with the city-level agglomeration and sorting effects. The results of this study also suggest that the role of business incubation in guiding the spatial allocation of startups can be played among cities with similar city size.
Third, small cities can hardly bring in high-productivity startups by subsidizing incubators due to the higher agglomeration effect of startups and the lower productivity-enhancing effect of business incubators. Through counterfactual analysis, this paper finds that even if small cities introduce high-efficiency business incubators, it is still difficult to compensate for the loss of profits generated by high-productivity firms moving to small cities and will introduce low-productivity startups in small cities; meanwhile. Although transfer subsidies and rank subsidies to business incubators in small cities introduce low-productivity startups, the productivity of startups introduced by the rank subsidy is slightly higher than that of the transfer subsidy because the efficiency frontier of the small-city incubators increases faster in the case of rank subsidy.
Forth, rank subsidies to startups in small cities can introduce high-productivity startups, but the productivity losses are higher than subsidise to incubators. This finding suggests that if small cities provide rank subsidies to startups to attract high-productivity startups while also providing rank subsidies to the business incubators, the introduction of high-efficiency incubators in small cities will help reduce the productivity loss of startups due to reduced agglomeration externalities.
Fifth, 95% of the higher productivity of startups in high-efficiency incubators originates from agglomeration effect and sorting effect, while the contribution of incubators is 5%. This implies that the higher productivity of incubatees in large cities is mainly due to the sorting effect and the agglomeration effect of cities, and the contribution of incubators to the productivity of startups is relatively small. The results of this study also suggest that the performance evaluation of incubators will be biased if city-level factors are not fully considered or if higher weights are assigned to the performance of incubatees.