Original text printed in Caijing magazine
The Third Plenary Session of the 18th Central Committee to be held in Beijing has been given worldwide attention. There are numerous predictions for the launching of reform measures, involving many areas of the financial sector, financial institutions, administrative controls, factor prices, industry monopolies and income distribution. We believe that the reform of local governments is particularly noteworthy. If the behavior of local governments doesn’t change, it will either be difficult to implement these reforms, or their effect will be of less value than anticipated or intended. In the past, competition between regions, corporatization of the behavior of the Chief Executive and soft budget constraints, promoted the economic growth greatly, however it also caused a number of structural problems. For example, economic growth is highly dependent on investment, administrative intervention in the market and the surge of debt risk. The new government now uses the new urbanization as a driving force of economic growth in the next cycle of development. If the behavior of local governments cannot be changed effectively, it is likely to lead to serious economic, financial and fiscal risks.
In the past 30 years of reform, local governments have competed with each other. The heads of the political party and governments would act as chairman and general manager of the local economy, keen on attracting investment. The causes of this “corporatization of local governments” mode, praised highly by economists like Steven Cheung, are numerous, but the most important ones may be the following three: First, decentralization reform that transfers economic decision-making power from central to local governments. Benefits of incipient decentralization reform are obvious because local officials can make better decisions according to local conditions. Second, GDP growth became the most important factor of promotion for local officials. Li’an Zhou and other economists have verified this hypothesis by the data concerning tenure of provincial governors and party secretaries during the reform. The analysis done by Jing Wu and other professors who use the data of municipal directorate not only support the conclusion that quick GDP growth is conducive for the promotion of the heads of government, but also found that when the proportion of environmental investment in local GDP increased by 0.36%, the opportunities of promotion for the local secretaries would be reduced by 8.5%. These “city construction mayors” have become living examples for mayors across the nation. Third, the direct interference in economic activities by local governments can compensate for the deficiencies of an imperfect market system. For instance, enterprises can get the support of loans, investments and even marketing channels through the establishment of good relations with the government.
Although the “corporatization of local government” approach was once very successful, it is merely the product of an unfinished reform. Some successful East Asian economies have had some experience with “development-oriented government” and China is not a special case. However, those East Asian economies which successfully became high-income countries all got rid of “development-oriented government” and transformed into more complete market economies. Although the “corporatization of local government” contributed to the rapid growth, it is obvious that it also caused negative consequences. Quoting professor Qiren Zhou, after decentralization of decision-making power transferred from the central government to local governments, “Xifeng Wang is stricter than Jiamu” (reference to the characters from the famous ancient Chinese novel Dream of Red Chamber meaning that the local government became stricter than the central government once was). One of the most common problems is the efficiency of resource allocation. Because the government directly engaged in economic activities, the state-owned sector occupied a large number of investments, including bank credit. It is therefore easy to discriminate or even exclude private economic activities. Another common problem is that the government was so busy developing the economy that it didn’t have enough energy to provide public services to maintain market and economic order. The third common problem is official corruption. The government controlled a lot of resources, which encouraged rent-seeking behavior.
What made the “corporatization of local government” worse were the soft budget constraints for the officials who engaged in economic activities. The central government assessed local GDP growth but not its balance sheet. This greatly encouraged officials to sacrifice long-term interest in pursuit of GDP growth. Therefore, selling lands became a rational act. In addition, local governments do not have a separate balance sheet, so the liabilities of a local government can only be paid by the central government, which actually only creates a soft budget constraint on local government finance. For example, annual revenue of Erdos is 80 billion RMB, however, its debt has reached 300 billion RMB. Such a situation is not unique in China, and when accumulated, local government debt has become the largest risk in China’s debt crisis.
In 2011, the National Audit office published a report on the total debt of financing platforms of the local governments in China. The amount of the debt is 10.7 trillion RMB, equivalent to 27% of the GDP. However, most scholars believed that this figure underestimates significantly the debt burden of local governments. In 2013, the National Audit Office decided to once again do a thorough investigation. This is merely one of many illustrations of the saying “official data is not reliable yet private data is too much”, because specific estimates range from more than ten trillion to more than twenty trillion RMB. Currently, the total size of the Chinese government bonds account for approximately 15.5% of the GDP. This figure is far below the 100% of the US and UK and the 200% of Japan. However, China has a lot of hidden debt burden, including local debt, pension gaps and non-performing assets of state-owned commercial banks. All liabilities added together may be more than 100% of GDP.
Despite the previous, we still believe that current debt risks in China are controllable. Firstly, China’s 100% and the US/UK’s 100% are not the same concept, because the debt burden of the US/UK didn’t include pension, medical insurance and other aspects of hidden debt. Secondly, some of China’s local government debts are not net debt but expectant assets. For example, local financing platform borrowed some money from banks to build infrastructure. There should be some positive gain from the infrastructure in the future. Thirdly, if we regard the government as an independent body, its balance sheet is still healthy. Note, the assets of state-owned enterprises whose statistics are known have exceeded 100% of GDP.
To sum up, we believe that the biggest risk appears not to be the problem of stock of debt but of flow of debt. In other words, if the total debt can be controlled at the current level, then it is entirely possible for the government to gradually resolve this risk. China has some national resources, including overseas assets. The economic situation is also good and finance is relatively healthy.
Whether new urbanization advocated by the government can succeed or not, it depends in part on whether the behavior of the local government can be changed. If local governments continue past practices it is likely to regard urbanization as a new investment in fixed assets to promote GDP growth, thus become an important tool to increase opportunities of personal promotion for local leaders. However, investment needs a lot of capital and past practices were to sell land and obtain loans. If these practices continue, overinvestment problems will become more severe, and local debt will also continue to swell, and the pressure of financial institutions will continue to increase. Therefore, maybe we are not only facing local debt risk, but also risk of economic recession and financial crisis. This may have been the reason why central government emphasized on people-oriented urbanization and tried to play down the role of investment in fixed assets. In order to promote people-oriented urbanization, there must be a change of incentives and constraints on local governments, thus forcing them to change their behavior.
The current core of local government reform should be: First, removal of corporatization; Second, strict budget constraints. Government interventions in the economy and government leaders playing the role of CEO in a company were just the product of a particular stage of reform. We are now moving towards a market economy, and governments should return to their original role, namely as a maker and defender of public policy and social order, and a public service provider. To achieve this transformation, two aspects must be changed. First is to eliminate the supremacy of the GDP in the promotion of officials. Even if the central government needs to assess the local economic indicators, it should also consider some of the indicators related to the people’s livelihood, like employment, social security and pollution levels. Also, not superiors but more local residents should be allowed to evaluate an official’s employment.
Meanwhile, the government should speed up the exit of direct intervention from many economic activities. The main line of the economic reform during 30 years has been marketization, but the marketization is not complete. The product market has basically been liberalized but the distortion of the factor market is very serious. The next step should be a transition to complete market economy, especially promoting the marketization process of finance, energy, and even land and other factors. If the market is liberalized, the space for government intervention will also be reduced.
The second core of local government reform is strict budget constraints. As a local government, it is inevitable to have some economic activities, such as building infrastructure and providing social securities. However, engaging in these economic activities should be based on an independent balance sheet. In brief, whoever borrowed the money should be responsible for it. With an independent balance sheet, local officials can invest depending on how much money they have maintained in their region. Investors can also determine the financial conditions according to the financial position of funding agencies.
After having a relatively strict budget constraint institution, it is reasonable to open up sustainable financing channels for the local governments. Currently, one of the reasons for financial chaos of local governments is that after the tax reform in 1994, executive power and financial power of local governments were mismatched severely. Frequently there were situations that when the central government launched a policy the local governments had to pay the bill. It is difficult to promote urbanization if this situation does not change.
Funds can be raised for local urbanization in several ways in the future: first, to increase revenue directly controlled by local governments to eliminate the mismatches between executive power and financial power. Specific practices may include changing the sharing ratio of tax revenue between central and local governments, reducing executive responsibility of local governments, and opening up new sources of revenue, including property tax and resource tax, etc.; Second, better management of transfer payments from the central government. There is always some local construction that needs support from the central government; Third, the development of a municipal bond market. This market must be based on an independent and clear balance sheet of every region and even every project; Fourth, using various forms of private capital to support regional economic development, including raising funds for local public service enterprises through listing.
The author is a professor at the National School of Development in Peking University.
黄益平：改变地方政府行为
原文在于《财经》杂志
即将在北京召开的十八届三中全会受到了世界范围的关注。对于可能推出的改革措施有各种版本的猜测，涉及金融部门、财政体制、行政管制、要素价格、行业垄断和收入分配等多个领域。我们认为地方政府的改革特别值得关注，地方政府的行为不改变，上述改革要么难以落实，要么效果就会打折扣。过去地区之间相互竞争、行政长官行为公司化及软预算约束等做法既极大地促进了经济增长，但也造成了许多结构性的问题，比如经济增长高度依赖投资、行政干预市场和债务风险剧增等。现在新一届政府把新型城镇化作为下一轮经济增长的一个驱动力，倘若不能有效改变地方政府的上述行为，极有可能酿成重大的经济、金融与财政风险。
改革30多年间，各地方政府互相竞争，党政首长充当地方经济的董事长与总经理的角色，热衷于招商引资。这个曾经受到张五常等经济学家推崇的“地方政府公司化”模式成因很多，但最主要的可能是如下三条：第一，放权改革，把一些经济决策权力从中央移交到了地方。改革初期放权的好处很明显，因为地方官员可以更加因地制宜地决策；第二，GDP增长成为决定地方官员升迁最重要的因素。周立安等经济学家用改革期间省长和省委书记任职的数据验证了这一猜测。而吴静等教授运用地市级首长数据所做的分析，则不仅支持GDP高速增长有利于首长升迁的结论，还进一步发现一个地方的环保投资占当地GDP的比例每升高0.36％，当地书记的升迁机会便会下降8.5％。各地纷纷得到提拔的“造城市长”更成为全国市长们活生生的榜样；第三，改革初期地方政府直接介入一些经济活动可以弥补市场制度不完善的缺陷。比如企业通过与政府建立良好的关系获得信贷、投入品甚至营销渠道等方面的支持。
虽然“地方政府公司化”的做法一度十分成功，但它只是改革尚未完成的一个中间产品。东亚一些成功的经济体都有过或多或少的“发展型政府”经历，中国也并非特例。但成功进入高收入阶层的东亚经济，后来都摆脱“发展型政府”的模式而走向了比较完全的市场经济。现在这个挑战已经摆到我们面前。尽管“地方政府公司化”促成了高速增长，它所造成的负面后果也十分明显。借用周其仁教授的语言，放权之后决策权从中央到了地方，但“王熙凤比贾母管得还要紧”。一个最常见的问题是资源配置效率，政府直接从事经济活动，令国有部门占有大量的投入品，包括银行信贷，却容易歧视甚至挤出民间经济活动。另一个普遍的问题是政府忙于发展经济，便没有足够的动力与精力来提供公共服务、维持市场与经济秩序。还有一个常见的问题是官员腐败，政府控制大量的资源，容易鼓励寻租行为。
而令“地方政府公司化”这一模式雪上加霜的，是官员从事经济活动的软预算约束。中央考核地方GDP增长率，却不考核其资产负债表。这就极大地鼓励地方官员以牺牲长期利益为代价追求GDP增长，因此，卖地就成为一种理性行为。另外，因为各级地方政府没有独立的资产负债表，地方政府的负债最终只能由中央政府埋单，这实际是造就了地方政府融资的软预算约束。比如鄂尔多斯市的年度财政收入为800亿元，其负债却达到3000亿元。这样的情况在全国并非绝无仅有，累积起来便成为中国债务危机的最大风险。
2011年，国家审计署公布全国地方政府融资平台总共负债10.7万亿元，相当于当年GDP的27％。但大多数学者认为，这个数字明显低估地方政府的债务负担。2013年中，国家审计署再次决定展开摸底调查。这是又一个“官方数据不可靠、民间数据满天飞”的例证，具体的估计值从十几万亿元到二十几万亿元不等。目前，中国国债的总规模占到GDP的约15.5％，这个数字远远低于美国和英国的100％以及日本的200％。不过，中国有很多隐形的债务负担，包括地方负债、养老金缺口和国有商业银行的不良资产。所有的负债加在一起，可能已经超过GDP的100％。
即便如此，我们仍然认为目前中国的债务风险还是可控的。第一，中国的100％和英美的100％并非一个概念，因为英美的债务负担尚未考虑养老金、医疗保险等方面的隐形债务；第二，中国的一些地方政府负债并非净负债，而是有对应的资产，比如地方融资平台从银行借钱修建基础设施。这些基础设施未来应该会有回报的；第三，如果把政府作为一个独立的主体看，其资产负债表尚属健康，光是有统计的国有企业的资产就已经超过GDP的100％。
总结起来，我们认为，目前看来债务问题的最大风险还不是存量问题，而是流量问题。也就是说，如果总债务能够被控制在目前的水平，那么政府完全有可能逐步化解这个风险，一方面国家有资产，另一方面经济态势比较好，财政也比较健康。况且国家还有其他一些资源，包括海外资产。
政府倡导的新型城镇化能否成功，关键之一就看地方政府的行为能否改变。如果地方政府延续过去的做法，那就很可能把城镇化当作以新的固定资产投资推动GDP增长，从而成为增加地方首长个人升迁机会的重要工具。但投资又需要大量的资金，过去的做法是卖地及贷款。这些做法延续下去，投资过度的问题会变得更加严重，地方负债也会继续膨胀，最后，金融机构的压力也会不断加大。因此，我们要面对的可能就不仅仅是地方债务风险，可能还有经济衰退和金融危机的风险。这可能是最近中央一直强调人的城镇化，尽量淡化固定资产投资的作用。但是如何以人为核心推动城镇化，还需要改变地方政府的激励与约束，从而改变其行为。
当前地方政府改革的核心应该是：一、去公司化；二、硬化预算约束。政府干预经济，行政长官做公司CEO的工作，都是特定改革阶段的产物。我们现在走向市场经济，政府也应回归其本来的角色，即作为公共政策和秩序的制定者和维护者，作为公共服务品的提供者。实现这一转型，需要做两个方面的工作，首先要消除官员选拔中GDP至上的做法，即便要考核地方的经济指标，也应该同时考虑就业、社保和污染等涉及民生的一些指标，评价一个官员做得好还是不好，要更多地让当地居民而不是上级来评价。
与此同时，政府要加快退出对许多经济活动的直接干预。30多年经济改革的主线是市场化，但市场化并不彻底，产品市场基本已经放开，要素市场扭曲十分严重。下一步的改革就应该尽可能地完成向市场经济的过渡，特别要推动金融、能源甚至土地等生产要素的市场化进程。市场放开了，政府干预的空间也就缩小了。
地方政府改革的第二个核心是硬化预算约束。作为地方政府，总归还有不少经济活动，比如建设基础设施和提供社会保障等。但地方政府从事这些活动，都应该以一个独立的资产负债表为基础，简单地说就是谁借钱、谁负责。有了独立的资产负债表，地方官员也可以看米下锅，投资者也可以根据筹资机构的财务状况决定融资条件。
有了一个相对严格的预算约束机制之后，就应该为地方政府开辟合理、可持续的融资渠道。目前，地方政府财务混乱的一个原因是在1994年的分税制改革之后，地方政府的事权与财权严重不匹配，经常出现中央出政策、地方埋单的情形。这一状况不改变，城镇化也很难推进。
未来可以从几个方面为地方城镇化筹资：一是增加直接由地方政府控制的财政收入，消除事权与财权不匹配的现象。具体的做法可以包括改变地方、中央税收分成比例，减少地方政府的事权责任，而且开辟新的收入来源包括开征房产税和资源税等；二是规范中央政府的转移支付。全国总归有一些地区的建设需要中央支持；三是发展市政债市场，但这个市场必须以每个地区甚至每个项目独立、清晰的资产负债表为基础；四是以各种形式利用民间资本支持地区经济发展，包括通过上市等方式为地方公共服务企业融资。
作者为北大国家发展研究院经济学教授
