科技企业孵化器何以激活区域创新——风险投资与孵化基金的中介作用

完善创新生态系统、提升区域创新能力是我国科技企业孵化器建设的重要目标。基于2013-2018年中国（内地）30个省份面板数据，采用面板数据模型实证检验科技企业孵化器是否促进了区域创新能力提升，并基于中介效应模型探讨风险投资和孵化基金在其中的间接作用。结果发现：①科技企业孵化器建设显著提升了区域创新水平，但主要增加的是实用新型和外观设计专利申请授权数总量，对发明专利申请授权数并没有显著促进作用。上述结果在剔除直辖市样本、采用随机效应模型及空间计量模型的稳健性检验后依然成立；②通过中介效应模型检验发现，区域风险投资和孵化基金集聚效应是科技企业孵化器影响区域创新水平的主要机制；③科技企业孵化器对区域创新的影响在不同区域间差异较大，在东部地区的创新激励效应更加显著，而且政策工具强度对科技企业孵化器与区域创新水平的关系具有正向调节作用。     

公共危机中的政府协作治理能力*——中国省级新冠肺炎疫情防控的文本分析

协作治理，是应对公共危机形成的一种新的治理形式。本文使用文本分析法考察了2020年2月末各省区市政府在本次疫情应对政策中协作治理能力的四种关键机制：促进式顶层领导、开放式参与路径、明确的协作职责、透明信息披露的运用情形及运用效果。研究结果显示，30个省区市都在运用协作治理机制抗击疫情、恢复经济。四种机制的运用存在差异，整体上沿海地区运用协作治理机制更充分。进一步的分析显示政府协作治理机制的运用显著促进了各省份疫情中第三产业的经济恢复水平，原因在于其改善了省级层面经济恢复的市场环境。本文的研究结论表明，强大的政府协作治理机制有助于社会快速有效应对公共危机，也是优化市场环境的政府治理新途径。本文为政府协作治理的机制构成与实际效果提供了重要补充，在公共危机应对的背景下发展了政府治理的理论内容，为公共危机情境下政府职能转变思路提供了参考。     

实现2035年发展目标的潜在增长率*

党的十九届五中全会提出了到2035年人均GDP达到中等发达国家水平的远景目标，因此测算和回答能否和如何如期实现该目标，对于我国实现第二个百年奋斗目标和坚持“四个自信”具有重要的意义。为此，本文根据跨越和陷入“中等收入陷阱”经济体的发展经验，对2021—2035年我国潜在增长率变化进行了测算。一是参照跨越和陷入“中等收入陷阱”经济体在我国相同发展阶段时各主要生产要素的变化，模拟设定我国未来各主要生产要素的增长率；二是通过运用附加人力资本的增长核算模型测算基准、乐观和悲观三种不同情境下未来我国经济的潜在增长率，验证我国2035年发展目标实现的可能性；三是依据主要要素对潜在增长率的贡献度，提出我国如期实现2035年发展目标的相应政策建议。     

协调度视角下长江流域产业转移效率研究

实现区域经济协调发展是当前我国在块集聚与点极化双重压力下面临的现实难题，长江经济带尤其是欠发达省份承接发达地区产业转移是解决该难题的必然路径选择。产业转移活动因区域主客体特征变量不同，呈现出典型的多发式转移和门槛转移特征，使得区域政策制定缺乏科学指导，陷入盲目拼政策红利的状态。结合效率模型及考虑熵权的经济社会发展协同度分析，测算了2004-2016长江经济带各省（市）年综合产业转移效率，并考察了门槛效应值及变化趋势。结果表明：在考察期内，长江经济带各省（市）产业转移效率提升较为明显，但不少省（市）存在不协调的阶段性平衡关系，且这种关系还未显露经济总量驱动下的协同发展能力。     

Analysis of territorial compatibility for Seveso-type sites using different risk assessment methods and GIS technique

The increasing size of chemical plants and rapid growth in residential areas has led to many incompatible land-use scenarios in the last 100 years. In this context, the authors assume that assessment and planning have their significant role in preventing the juxtaposition of hazards and population, and that, in this field, there is a broad recognition of the need for legislative and policy consistency across the European Union. The paper presents a comparative case study in which the Romanian land-use planning (LUP) criteria and the risk-based quantitative approach for a chemical plant are applied. Accident scenarios involving chlorine and propylene arecomprehensively analyzed using consequence and risk modelling software and GIS technique for the territorial compatibility assessment. The objective of the paper is threefold. Firstly, it presents an overview about current risk analysis methods; secondly, the authors advance an understanding of risk assessment practices used in several countries for the prevention and control of major industrial accidents involving dangerous substances and, also, for LUP. Thirdly, a method targeting an improved risk assessment framework for LUP, encompassing Romania’s determinants is outlined. The results obtained using the two different approaches indicate significant differences regarding the possibly affected areas and territorial compatibility. Furthermore, based on the findings, the paper ends with a set of recommendations that can be transformed into the foundation for future enactments of new safety standards that cover risk assessment for LUP. Consequently, the present study aims to become a frame of reference for decision-makers towards more sustainable and updated risk assessment practices in the field of industrial activities.     

非互惠的零关税待遇是否促进了受惠国的出口多样化——基于中美对非洲贸易政策的比较研究

非互惠的优惠贸易待遇是世界贸易组织和国际社会支持发展中国家尤其是最不发达国家发展的重要工具。为了评估该政策的效果,本文以中美对非洲实施的零关税待遇为例,采用2001—2017年HS8位贸易数据从整体、行业和区域层面评估了两种政策对受惠国出口多样化产生的影响。结果显示,美国零关税待遇分为AGOA-GSP和AGOA-服装两个条款,中国零关税待遇(FOCAC)在遵循WTO规则、受惠商品范围、受惠国家标准、原产地标准及政策有效期方面均优于AGOA-GSP,但在原产地标准灵活度和政策优惠力度方面低于AGOA-服装。这些特征导致三类政策对受惠国出口多样化的影响产生显著的差异。整体上,FOCAC和AGOA-服装对受惠国出口多样化均产生了显著的积极影响,而AGOA-GSP的影响不显著;行业层面,FOCAC对受惠国制造业、矿业和农业的出口多样化均产生了显著的促进作用,而AGOA的两个条款对受惠国三个细分行业出口多样化的影响均不显著;区域层面,三类政策对受惠国出口多样化的影响仅在部分区域发挥效果。     

A novel model to manage air cargo disruptions caused by global catastrophes such as Covid-19

The outbreak of the COVID-19 pandemic has drastically disrupted the air cargo industry. This disruption has taken many directions, one of which is the demand imbalance which occurs due to the sudden change in the cargo capacity, as well as demand. Therefore, the random change leads to excessive demand in some routes (hot-selling routes), while some other routes suffer from a big shortage of demand (underutilized routes). Routes are substitutable when there are several adjacent airports in the Origin & Destination (O&D) market. In this market, demand imbalance between substitutable routes occurs because of the above reasons. To tackle the demand imbalance problem, a novel model is introduced to estimate the quantity combinations which maintains the balance between underutilized and hot-selling routes. This model is a variant of the classic Cournot model which captures different quantity scenarios in the form of the best response for each route compared to the other. We then cultivate the model by integrating the Puppet Cournot game with the quantity discount policy. The quantity discount policy is an incentive which motivates the freight forwarders to increase their orders in the underutilized routes. After conducting numerical experiments, the results reveal that the profit can increase up to 25% by using the quantity discount. However, the quantity discount model is only applicable when the profit increase in the hot-selling route is greater than the profit decrease in the underutilized route.     

Deep learning for modeling the collection rate for third-party buyers

This study evaluates a wide range of machine learning techniques such as deep learning, boosting, and support vector regression to predict the collection rate of more than 65,000 defaulted consumer credits from the telecommunications sector that were bought by a German third-party company. Weighted performance measures were defined based on the value of exposure at default for comparing collection rate models. The approach proposed in this paper is useful for a third-party company in managing the risk of a portfolio of defaulted credit that it purchases. The main finding is that one of the machine learning models we investigate, the deep learning model, performs significantly better out-of-sample than all other methods that can be used by an acquirer of defaulted credits based on weighted-performance measures. By using unweighted performance measures, deep learning and boosting perform similarly. Moreover, we find that using a training set with a larger proportion of the dataset does not improve prediction accuracy significantly when deep learning is used. The general conclusion is that deep learning is a potentially performance-enhancing tool for credit risk management.     

Direct intervention or indirect support? The effects of cooperative control measures on farmers’ implementation of quality and safety standards

In recent years, labor-intensive agricultural products from developing countries have exhibited inferior performance in international trade due to quality and safety incidents, among which pesticide residue is a major issue. Aiming to improve food quality and safety in the context of cooperatives, we introduced three categories of control measures: outcome control, process control and social control. Based on the Pre-Harvest Interval Standard (PHIS), we selected three indices, farmers’ implementation rate of PHIS, absolute distance to PHIS and relative distance to PHIS, to evaluate appropriateness of pesticides use, reduction of pesticide residue and safety improvements of agricultural products. By using random sampling survey data, we empirically analyzed the marginal effects of control measures and their combinations on food quality and safety standards. The empirical results show that implementing process control, namely, unified production standards or supply of unified agricultural inputs, can comprehensively improve farmers’ implementation rate of PHIS, absolute distance and relative distance to PHIS respectively by 34.9%, 3.2 days and 46.0% on average. While the effects of outcome control (safety inspection) and social control (bonus-penalty incentive or training) are restricted to other measures. Therefore, we suggest cooperatives should take farmers’ features, implementation conditions and the effects of control measures into consideration in order to make a sustainable management plan for improving food quality and safety and enhancing competitiveness in international markets.     

A note on the long rate in factor models of the term structure

In this paper, we consider factor models of the term structure based on a Brownian filtration. We show that the existence of a nondeterministic long rate in a factor model of the term structure implies, as a consequence of the Dybvig–Ingersoll–Ross theorem, that the model has an equivalent representation in which one of the state variables is nondecreasing. For two‐dimensional factor models, we prove moreover that if the long rate is nondeterministic, the yield curve flattens out, and the factor process is asymptotically nondeterministic, then the term structure is unbounded. Finally, we provide an explicit example of a three‐dimensional affine factor model with a nondeterministic yet finite long rate in which the volatility of the factor process does not vanish over time.