腐败、反腐败和中国县际收入差距

中国经济的快速增长和地区间收入差距紧密相联。虽然目前大部分研究文献都用经济因素来解释这一问题,但文章认为,地方政府的反腐败力度也是影响地方收入水平的重要因素。通过利用中国县级横截面数据以及采用普通最小二乘法,文章发现,反腐败力度越大的县的收入水平越高。利用最近发展起来的夏普里值分解法来量化各个解释变量对收入差距的贡献率,文章还发现,反腐败是解释中国县际收入差距的一个重要因素。     

政府公共支出结构对腐败的影响效应分析

政府公共支出结构是影响腐败的主要经济因素之一。本文建立腐败的客观衡量指标,采用固定效应模型,定量分析政府公共支出结构及政府支出规模、市场化程度、教育水平、工资水平以及经济发展水平等对腐败的影响,研究结果表明:减少寻租性支出、提高生产性支出对腐败有着显著的抑制作用,同时控制政府支出规模、提高私有化程度对反腐败也有显著正效应;提高人均收入水平和经济增长率对腐败存在着预期的负效应,但在统计上并不显著;教育水平和政府公务员相对工资的提高不利于反腐败,这与大多数实证分析的结果相反。     

‘Sweet deals’: State-owned enterprises,corruption and repeated contracts in public procurement

We show how a price analysis of stable relations between customer and supplier in the public procurement of homogeneous goods can help differentiate opportunistic from honest behaviour among economic agents. We consider two types of stable relations: repeated procurements and connections based on the state ownership of suppliers. On the basis of a large dataset on the procurement of granulated sugar in Russia from 2011 to 2013, we find that for private suppliers, prices of repeated contracts were lower compared to one-time deals when procured through more transparent procedures and higher when procured through non-transparent procedures. For non-transparent procedures, we observe significant overpricing of contracts with state-owned suppliers compared to private suppliers, especially in the case of repeated contracts, whereas for competitive e-auctions, there is only a small difference between the contract prices of state-owned suppliers and private suppliers.     

New perspectives on corruption contagion

Is corruption within one country affected by corruption within another? Few studies have examined this question in detail due to the difficulty of measuring corruption and paucity of consistent data over an adequate time span. I use a cross-country panel data-set spanning 1995–2014 to examine how domestic corruption reacts to the culture of corruption amongst a country's regional neighbors. I find evidence that a reduction in regional corruption can actually lead to a worsening of corruption within a country, and vice versa. If in an open economy, regional graft lowers the level of income that a rent-seeking government can tax, a reduction in regional corruption can increase the marginal benefit of imposing a more extractive domestic policy by increasing the pool of exploitable funds. My results suggest that corruption will be an enduring institution in a more interconnected world.     

A controlled field experiment on corruption

This paper reports on a controlled field experiment on corruption designed to address two important issues: the experimenter's scrutiny and the unobservability of corruption. In the experiment, a grader is offered a bribe along with a demand for a better grade. We find that graders respond more favorably to bigger bribes, while the effect of higher wages is ambiguous: it lowers the bribe's acceptance, but it fosters reciprocation. Monitoring and punishment can deter corruption, but we cannot reject that it may also crowd-out intrinsic motivations for honesty when intensified. Finally, our results suggest several micro-determinants of corruption including age, ability, religiosity, but not gender.     

CORRUPTION'S DIRECT EFFECTS ON PER‐CAPITA INCOME GROWTH: A META‐ANALYSIS

Corruption is a symptom of weak institutional quality and could have potentially adverse effects on economic growth. However, heterogeneity in reported findings makes it difficult to synthesize the evidence base with a view to test competing hypotheses and/or support evidence‐based policy and practice. To address this issue, we have extracted 327 estimates of corruption's direct effect on per‐capita GDP growth from 29 primary studies, following a peer‐reviewed and pre‐published systematic review protocol. Precision‐effect and funnel asymmetry tests indicate that corruption has a negative effect on per‐capita GDP growth after controlling for publication selection bias and within‐study dependence. However, multivariate meta‐regression analysis results indicate that the overall effect is not robust to inclusion of moderating variables through a general‐to‐specific procedure for model specification. We report that the marginal effect of corruption on per‐capita GDP growth is more adverse when the primary study estimates relate to long‐run growth, are based on low‐income‐country data only, and extracted from journal papers. The effect is less adverse in studies that use the International Country Risk Guide corruption perceptions index and in those reporting estimates from two‐stage least‐squares estimations.     

The Effect of Fiscal Decentralization on Corruption: A Non‐linear Hypothesis

This study advances the literature on the effect of decentralization on corruption by proposing a theoretical motivation for the hypothesis that the corruption‐minimizing decentralization arrangement prescribes intermediate decentralization degrees: a ‘middle’ decision‐making power of local governors may mitigate the tradeoff between the rent‐seeking incentive for local politicians and the effectiveness of their voters’ monitoring. The estimation of a non‐linear empirical model strongly confirms that a decentralization degree between 15% and 21% minimizes corruption even through different estimation procedures, introduction of control variables and the use of internal and external IV and of alternative decentralization and corruption measurements.