Source: blue whale OLS Lab
Author: Wang Yi, Nanjing University
The original title "breaking the" high-level political connection ": the anti-corruption movement in China's first-class land market"
Source: Ting Chen, James Kai sing Kung, "funding the 'princes': the campaign against corruption in China's primary land market", the Quarterly Journal of economics, volume 134, issue 1, February 2019, pages 185 – 226
1、 Introduction
Corruption is a very important and realistic problem in the current development of our country. Typical corruption is that businessmen bribe powerful government officials in exchange for cheap national rare resources or preferential policy treatment. However, the corruption studied in this paper is provided by local officials, that is to say, by offering price discounts to some companies with "high-level political connections" - their investors or owners have some connection with members of the Politburo - in exchange for possible promotion expectations. The universality of this nepotism "deal" has greatly undermined the efficiency of competition, so the first major task of the new leaders is to launch anti-corruption actions to eradicate corruption. This paper uses the data of more than one million government land transactions as a single seller from 2004 to 2016, and uses the spatial matching method to carry out empirical research on the following two main lines: (1) how much profit do local officials and companies with "high-level political connections" get from their transactions in the primary land market? (2) What is the effect of the anti-corruption campaign of the state leaders on this transaction?
2、 Background
Anti corruption Campaign
One
Shortly after the new leader of our country took office, he launched a comprehensive fight against corruption. Compared with the past, this campaign is more intense and lasting. Its main characteristics include taking the land market as the main anti-corruption target, strengthening the central discipline supervision committee, replacing and appointing the provincial Party secretaries of 14 provinces, etc. In addition, the biggest breakthrough is to insist on "tiger" and "fly" fighting together, such as arresting Zhou Yongkang and making plans. As a result, the country's leaders sent a strong signal to these family members at the top of the political spectrum - that no corruption will be tolerated.
China's nepotism: a conceptual framework
Two
How does China's nepotism work? It can be clearly shown in the figure below.
Local officials have control over scarce and state-owned resources. They offer price discounts to companies that have connections with members of the Politburo's standing committee or Politburo's standing committee, and can earn money (outright corruption) or promotion expectations from each other.
So why does China's nepotism work like this? It is based on two institutional features.
One is the decentralization of the control power of state-owned assets in a large scale. Taking land as an example, in 1998, the state passed a bill to grant legal ownership of land within a geographical jurisdiction to local governments, which puts local officials in a special position. They can implement price discrimination by offering substantial discounts to purchasers in exchange for certain benefits. The question is who do they offer discounts to?
The other is the "superior" policy, that is, the appointment and promotion within the party and the state are decided by senior personnel. The members of the Political Bureau in a special position can exert influence on the promotion of provincial leaders, especially the Secretary of the provincial Party committee. In order to reach the peak of political career, provincial leaders have great motivation to provide any form of preferential policies to companies with high-level political connections.
Corruption in China's primary land market
Three
Why is the land market ideal for studying the interest transaction between "high-level political association" companies and local officials? First, the local government is the only owner and supplier of land, which is equivalent to the monopolist in the market; second, the local financial system in China has changed from mainly enterprise tax to mainly land tax, which leads to the local governments at and below the prefecture level increasingly rely on the income from the sale of land use rights.
3、 Data structure
Companies with "high-level political connections"
One
(1) Define "high level political connections"
We regard "children and family members of elite party members who served between 1997 and 2016" as objects with "high-level political connections". In the sample, we found 64 members of the Political Bureau from the 15th to 18th National Party Congress.
(2) Identify objects with "high-level political connections"
Among the 64 members of the Politburo, 48 were selected. Through rich online media resources, 134 "high-level political related" objects were found. Among them, about 58.21% of the children or in laws of the members of the Political Bureau in the sample belong to the private sector (45.52%). The closeness of these relationships gives us confidence in the reliability of the data.
(3) Identify companies with "high-level political connections"
A total of 3530 companies with "high-level political connections" were found. Most of them are engaged in real estate, finance, leasing and business services, indicating that the primary land market is an area where rents can be contested.
Companies that match land deals with "high-level political connections"
Two
Find the data set on the official website of the Bureau of land and resources, including all the land transaction data on China's first-class land market from 2004 to 2016 (each specific transaction details are disclosed, including size, location, total payment, date, name of the buyer and the seller, specific transaction method, two land use codes, land quality (subjective evaluation by officials, 20 point system), three purchase Company code). Match the list of "high-level political association" objects obtained above with the land transaction data, and find 19812 pieces of land purchased by "high-level political association" companies in this period. The geographical distribution is shown in Figure III.
Spatial matching samples for recognition
Three
Assuming that the same size of land sold in the same region and the same year is highly comparable in quality and price, in order to ensure similar comparison, a land transaction sample purchased by "no high-level political association" company in the same year and within the same distance is generated. For example, within 500m radius (3 blocks), within 1500m radius (8 blocks). A total of 359539 plots within 1500m radius and 207564 plots within 500m radius purchased by "no high-level political association" company can be selected.
4、 Price discount implemented by "senior political affiliated" companies
The regression model of estimated price discount is as follows:
Priceickst=β0+β1PrincelingPurchaseikjt+γXi+Tcst+vickst
Dependent variable priceickst: the logarithm of the price of land I sold by the local government C to the company K in t-month-year;
Explanation variable princelingpurchaseikjt: virtual variable, if company K is related to "high level political association" object J, the value is 1, otherwise the value is 0;
Control variables: including logarithm of land sales area, land quality, transaction method, company size and company ownership, etc.
The model uses high-dimensional city year use fixed effect, monthly fixed effect and three industry fixed effect, and clusters the standard deviation according to province and company to eliminate individual differences under different categories.
The regression results are shown in table V:
(1) In the whole sample, it indicates that the influence of "high-level political association" company purchase on land price is significant at the 0.1% significance level;
(2) And (3) to a certain extent, we deal with the problem of choice bias: the purchase of high-level political related companies also has a significant impact on land prices. For example, within a radius of 500m, the price discount of "high-level political related" companies is 57% (calculated according to the coefficient of -0.844).
To ensure that the results are not disturbed by outliers, figure IV shows that in most cases "high-level political affiliated" companies pay less.
Considering that in addition to the purchase of "high-level political association" companies, there are also additional factors that influence the power of "high-level political association" objects or whether they can continue to influence personnel appointment after retirement, the princelingpurchaseikjt and the virtual variables pscmjp and retired JPT are interactive respectively. (4) To (6) indicates that the contact between the "senior political association" company and the Standing Committee of the Political Bureau increased the price discount by 17.6% (exp (- 0.62) - exp (- 0.62-0.396) = 0.176) to 20.7% (exp (- 0.545) - exp (- 0.545-0.442) = 0.207). (7) To (9) indicates that the price discount has not decreased after the members of the Politburo retire, indicating that retirement does not affect the continuity of power.
In order to ensure that the estimation is not affected by the selection bias of selectivity or false reports, two robustness tests are carried out. Two small but accurate sub samples were used for the test. One is different business groups associated with Zhou Yongkang and his son, the other is land purchase of a listed company with "high-level political connection" in the primary land market, and it is found that the estimated value of discount is almost the same as our previous estimate (table AII).
5、 The effect of business preference on Promotion
In order to study the influence of the purchase of "high-level political related" companies on the promotion of officials, the ordered probit model was established based on the provincial panel data from 2004 to 2016
Turnoverit=?0+?1PrincelingPurchaseit+?2FactionalTiesjt+?3GDPGrowthit+κXit+ωWj+λi+Tt+φj+τijt.
The dependent variable turnoverit is the political turnover of provincial officials measured by year: promotion = 3, lateral transfer or retention = 2, retirement = 1, suspension = 0;
The independent variable princelingpurchaseit is a virtual variable (same as above);
In order to avoid other confusing effects, two variables of "economic performance" (GDP or tax growth rate as proxy variable) and "sectarian tiesjt" are controlled;
Other control variables include the personal characteristics of provincial officials, such as age, age square and years of education.
In addition, in order to prevent the virtual variable princelingpurchasept from lacking the discount range information provided by the local government, we can use princelingdiscountspt (the difference between the transaction price of "high-level political affiliated" companies and the corresponding "no high-level political affiliated" companies within 500m) and areaoflandpurchasedit (the number of land sold to "high-level political affiliated" companies) to replace it To build a more accurate measurement method.
The test results are shown in table VIII:
It was found that the promotion possibility of provincial Party committee secretaries who offered discounts to "high-level political related" companies increased by 23.4%, and the promotion possibility increased with the size of price discount and the increase of the number (area) of land purchased by "high-level political related" companies. But for governors, they must be promoted by raising the economic level or GDP growth within their jurisdiction, and only provincial Party secretaries can benefit from speculation and trading.
One possible assumption is that provinces offer discounts to attract more investment and achieve higher growth rates, leading to the promotion of provincial officials not because they offer incentives, but because they can promote faster economic growth. Select the proxy variables related to investment growth, table AIII shows that GDP growth is insignificant in the reasons for the promotion of party secretary, so reject the alternative hypothesis of growth promoting promotion.
For municipal officials, as the land custodian in the primary market, the provincial Party Secretary also needs the cooperation of municipal officials when selling cheap land to "high-level political affiliated" companies. The estimated result is similar to that of the provincial level (omitted).
6、 The effect of the anti-corruption campaign of state leaders
To study the effect of the anti-corruption campaign of Chinese leaders, three indicators need to be established, namely, "transactions after 2012", "Central Inspection" and "Xi appointed officials"