管理层权力与国企高管腐败——基于政府审计调节效应的研究

以2011—2017年沪深A股国有上市公司为样本，实证检验管理层权力诱发腐败以及政府审计对管理层权力与腐败之间关系进行调节的作用机制，在此基础上揭示政府审计参与国有企业腐败治理的具体路径。研究发现：管理层权力越大，国企高管腐败越严重；政府审计能够显著抑制由管理层权力诱发的腐败，即政府审计对管理层权力与国企高管腐败之间的关系有显著的负向调节作用。研究结果揭示了政府审计作用于腐败治理的渠道机制，对国有企业腐败治理和进一步发挥政府审计在深化国企改革方面的作用具有一定的参考价值。     

Budget support to the health sector—The right choice for strong institutions? Evidence from panel data

This article examines the relationship of sector budget support to the health sector and the infant mortality rate for developing countries. Project-type interventions have been widely used in developing countries in the past decades. These smaller-scale interventions often did not bring the results that the donors would have wanted, at least on a macro level. At the beginning of the millennium, forums on aid effectiveness proposed new principles to increase the effectiveness of aid. Many scholars agreed that one of the answers would be budget support. This article tries to answer whether budget support is the efficient aid modality in countries with strong institutions. In the baseline scenario, a panel data analysis is applied, which includes 113 countries between 2010 and 2018. This dynamic linear panel model is estimated by using ordinary least squares (OLS) and system generalized method of moments (GMM). Health sector aid, in general, has a significant and negative effect on the infant mortality rate in the average country. Sector budget support is insignificant in the baseline estimation and when interacted with a governance variable. In contrast, project-type interventions exhibit significant and negative effects on the outcome variable. The results indicate that sector budget support might not be the superior choice among the aid modalities in the health sector, even in countries with good governance.     

西江流域两广开发利用的博弈策略及其仿真

针对西江流域开发过程中出现的水源林面积减少、水土流失严重、移民安置、水体污染严重等由经济发展带来的负面生态文明问题,采用演化博弈方法探究西江流域两大治理主体广西政府与广东政府的策略与可能产生的改变与均衡。结果表明,传统情况下两广政府博弈系统将朝着(过度开发,不合理加速)的帕累托最劣稳定状态演化,不利于西江流域生态的绿色可持续发展；引入问责机制对广西政府“过度开发”和广东政府“不合理加速”进行惩罚,博弈系统存在9种演化情景,当且仅当满足特定条件时,系统会收敛到(适度开发,合理加速)的帕累托最优状态；数值仿真模拟下根据惩罚值高低设计9种不同情景,分析惩罚值对博弈结果的影响,发现在中央政府及上级相关部门给予两广政府的问责惩罚都较高时,博弈系统朝着帕累托最优状态演化；为促进西江流域生态经济与治理的均衡发展,在问责机制设计上,要给予两广政府满足相应条件的较高惩罚值。     

The effects of an environmental accountability system on local environmental governance and firms’ emissions

The Chinese central government introduced the environmental accountability system (EAS) policy in 2006 to encourage local governments to execute environmental policies appropriately. However, the effects of this EAS policy have not been investigated at the firm level. In this study, we first measured the environmental governance intensity of local governments by calculating the environment word portion (EWP) of the annual governments’ work reports and found that local governments increased their efforts after the EAS policy implementation. We then constructed a continuous difference-in-difference (DID) strategy to estimate the effect of the EAS on firms’ emission and found that firms significantly reduced their sulphur dioxide (SO₂) emission under the regulation intensity of local governments after the EAS policy. The reductions are mainly because of the private, non-export, labour- and capital-intensive firms. The EAS policy may encourage firms to reduce coal usage, design more products, decrease capacity investment and even withdraw from the market, thus reducing pollutant emission. Finally, we tested firms’ response in various polluting sectors and found that the EAS policy may cause local governments to fail in distinguishing low pollution sectors from high pollution sectors.     

Does agricultural official development assistance facilitate foreign direct investment in agriculture: Evidence from 63 developing countries

Although past studies have separately explored the direct impact of agriculture official development assistance (ODA) and foreign development investment (FDI) on agricultural production, the nexus between these two elements is often neglected. This article aims to understand the linkage between agricultural ODA and FDI, using data from 63 developing countries from 1991 to 2019. Poisson pseudo maximum likelihood estimations reveal that agricultural ODA considerably promotes FDI in the agriculture, fishery and forestry sectors (FDI_aff) by approximately 0.5%, while its impact on FDI in the food, beverages and tobacco industries (FDI_fbt) is overall insignificant. Geographical and ecological conditions play a decisive role in accounting for FDI in agriculture. While coastal and land-rich countries receive a significantly higher amount of FDI_aff and FDI_fbt, tropical countries are evidently more attractive destinations for FDI_fbt. The empirical analysis also shows that a peaceful social environment encourages FDI_aff. Well-established legal systems and reductions in corruption facilitate FDI_aff, whereas the impact of overall institutional quality on agricultural FDI is insignificant. Results suggest that donors prioritise agricultural initiatives with higher positive spillover effects, such as programmes supporting food crop production and agricultural research.     

财政收入竞争、地方政府行为与要素资源扭曲

本文构建了包含政府部门、工业部门、农业部门等多行为主体在内的理论模型，证明了地方政府在财政收入最大化动机下，会带来资本、劳动和土地三种资源配置的扭曲。在此基础上，选取2000-2017年中国248个地级市面板数据，实证检验了竞争性财政行为对上述三种要素资源扭曲的影响，研究发现：地方政府财政收入竞争显著提高了资本、劳动和土地三种资源的错配程度；在考虑官员更替、政治晋升、从政经历等政治因素的影响后，实证结果依然稳健；财政竞争导致的要素资源扭曲不仅存在于省内竞争上，在全国范围内同样广泛存在。异质性分析显示，东部地区地方政府财政竞争会引起劳动力资源扭曲，中西部地区地方政府竞争行为会引起资本和土地资源扭曲；2008年以前地方政府财政竞争主要引起资本扭曲，2008年以后则同时引发了资本扭曲和劳动力资源扭曲。研究结论表明，相对于政治激励，财政压力才是引发地方政府资源错配的根本原因。中央政府应当继续深化财税体制改革，逐步建立现代化的国家治理体系，科学、动态地评估官员任期内决策带来的中长期影响，有效降低地方政府财政竞争行为所带来的负面影响。     

政府审计、贫困县官员政绩压力与债务风险

以经济人假设、政府审计是治理政府官员代理问题的免疫系统为理论基础，利用2015-2018年H省贫困县政府债务数据，实证检验政府审计抑制贫困县官员政绩压力对债务风险的影响。研究发现：官员政绩压力与贫困县债务风险显著正相关，政府审计的揭示、预防功能对抑制贫困县官员政绩压力影响债务风险的作用显著，而其抵御功能不显著。鉴于此，应构建和运用政府债务“数智审计”、系统改进官员绩效考核制度和科学合理地利用债务资金的杠杆作用，以助力实施“巩固拓展脱贫攻坚成果同乡村振兴有效衔接”战略。     

Testing machine learning systems in real estate

Uncertainty about the inner workings of machine learning (ML) models holds back the application of ML-enabled systems in real estate markets. How do ML models arrive at their estimates? Given the lack of model transparency, how can practitioners guarantee that ML systems do not run afoul of the law? This article first advocates a dedicated software testing framework for applied ML systems, as commonly found in computer science. Second, it demonstrates how system testing can verify that applied ML models indeed perform as intended. Two system-testing procedures developed for ML image classifiers used in automated valuation models (AVMs) illustrate the approach.