Fiscal Decentralization and Environmental Infrastructure in China

Under the tax system preceding the 1994 reform, provinces had different marginal revenue retention rates. After the 1994 reform, all provinces received the same, fixed share of VAT. Cities still have different VAT revenue shares after the reform.

For much of our analysis, we use the 1995 VAT sharing rate, although the tax reform in China was initiated in 1994. Wong (1997) writes that the 1994 tax reform was implemented only a few months after it was approved. She documents that neither taxpayers nor local tax officials were “prepared” for the transition. Moreover, through 1994, cities and counties were in doubt as to whether the rules of the contract system (the pre-1994 system) would govern the new tax sharing system.

We conclude from this reading that 1994 VAT sharing rates may be unreliable, since they involved a period of transition and, at best, reflect a system in place for only part of the year.

One saying, “qitong yiping,” states that, in order to attract investment, local governments must build seven forms of infrastructure: electricity, roads, water, telecommunications, cable, leveled ground, and waste treatment. We note that sewage is just one form of infrastructures that could be interesting to cities. Theoretically, each form of infrastructure should be attractive at the margin. We lack data at the city level to study the behavior of these other forms.

However, the price of water is considered a sensitive political subject in China. Cities cannot arbitrarily raise the price of water to fund the construction of new sewage treatment plants; price rises in cities are usually carefully coordinated with the central government and phased in over an extended period.

We found some occasional errors in the data of sewage treatment. This issue is discussed in “Sewage treatment data” in the Appendix in detail.

More infrastructure spending always increases the size of the tax base, since the city becomes more attractive to both businesses and consumers. Early spending can generate high returns, since important facilities such as clean water and electricity are essential to development. Later spending generates comparatively diminished marginal returns, since the highest return opportunities have already been selected.

More infrastructure always costs more to purchase and maintain. The initial units of infrastructure are comparatively cheap, while later ones are comparatively expensive.

We note that the dependent variable, change in sewage treatment capacity, is a stock variable, while the primary independent variable, a city’s VAT share, is a flow variable. Conceptually, the city’s VAT share should also affect a flow, which is the investment in sewage treatment during each year. Hence, we are answering the question of how fiscal incentives affect the aggregate set of investments in sewage treatment infrastructure.

Prior to 2002, China’s Environmental Yearbooks were not published, and the sewage treatment capacity of its cities was not known.

Sewage treatment fees are collected as a portion of the water consumption fee. These fees fund the construction and operation of treatment plants. Sewage treatment fees are widely regarded as inadequate to pay for the operating costs of sewage treatment. They fall far short of paying for the construction of sewage treatment plants (Lee 2010).

This table excludes direct-controlled municipalities.