深度-未来能源研究所-收入对汽车需求的影响：来自中国新车市场的证据（英）-2021.6_39页_756kb

Summary of "The Effect of Income on Vehicle Demand: Evidence from China's New Vehicle Market"

Core Content

This working paper investigates the causal relationship between income and vehicle demand in China's new vehicle market, using city-level data from 2005 to 2017. The study highlights the importance of accurate projections of vehicle ownership and emissions for effective climate policy design, particularly in the context of China's significant contribution to global transportation emissions.

Main Objectives

• To estimate the elasticity of new car sales to income in China.
• To assess whether recent projections of vehicle ownership and emissions are accurate.
• To contribute to the literature on income and energy-consuming durables, as well as the Environmental Kuznets Curve (EKC) literature.

Key Findings

• A 1% increase in income leads to a 2.5% increase in new car sales, indicating a strong income elasticity.
• This elasticity suggests that recent projections of vehicle sales in China have understated actual sales by 36% and carbon dioxide emissions by 18 million metric tons in 2017.
• The study concludes that China's climate policies may need to be more aggressive than previously forecasted to meet its GHG reduction targets.
• Income elasticity of vehicle sales is robust to alternative functional forms and controls for socioeconomic variables.
• The income-sales relationship appears to be proportional across cities, despite varying levels of public transportation infrastructure and urban development.

Methodology

• The researchers use city-level data on new car sales and income from 2005 to 2017.
• To address endogeneity and measurement error, they employ an instrumental variables (IV) approach.
• The instrument is a Bartik-style variable, which is the interaction of a city's education employment in 2004 with China's high-technology exports.
• The instrument is justified by:
  • High-technology exports being a key driver of China's economic growth.
  • Cities with high initial education employment having skilled workers who can produce high-technology exports.
• The exclusion restriction is supported by the lack of correlation between 2004 education employment and subsequent shocks to other drivers of vehicle ownership, such as public transportation quality.

Data and Variables

• Data Source: New vehicle registrations and socioeconomic data from the China City Statistical Yearbook.
• Variables Included:
  • New car sales (in thousand units).
  • New car expenditure (in billion RMB).
  • Average car price (in thousand RMB).
  • Income per capita (in thousand RMB).
  • Built-up area (in square km).
  • Paved road area (in square km).
  • Population (in thousand people).
  • Number of buses and taxis.
  • Total retail revenue (in billion RMB).
  • Education employment share (from 2004).
  • National high-tech export (in billion RMB).

Income Growth and Convergence

• From 2005 to 2017, income and vehicle sales grew rapidly, with annual growth rates of 9.6% and 20%, respectively.
• Income levels across cities converged over the period, with the lowest quintile cities experiencing 28% average annual growth.
• The income-sales relationship was consistent across cities, suggesting that the effect of income on vehicle demand is proportional.

Implications for Climate Policy

• The study suggests that current projections of vehicle sales and emissions in China are likely too conservative.
• This implies that climate policies need to be more aggressive to meet the country's emissions reduction goals under the Paris Agreement.
• The findings could also affect global climate policy analysis, as China is a major contributor to transportation emissions.

Contribution to Literature

• This is the first study to examine the effect of income on vehicle demand at the city level.
• It contributes to the EKC literature by showing that income affects pollution through micro-level household behavior, rather than just macro-level data.
• It also contributes to the literature on income and energy-consuming durables by providing empirical evidence that income has a strong and consistent effect on vehicle demand, unlike the mixed findings on household appliance ownership.

Limitations and Considerations

• The study uses registration data as a proxy for sales, which may have measurement issues due to delays in registration.
• However, the researchers note that registration lags are minimal (at most one month), and the data are aggregated to the annual level, reducing the impact of such lags.
• The study controls for policy changes such as tax reductions and fuel economy standards, which may influence vehicle prices and demand.

Conclusion

The paper provides evidence of a strong and proportional relationship between income and new vehicle sales in China. This relationship has important implications for climate policy, as it suggests that current projections of emissions and vehicle ownership are significantly understated. The study calls for more accurate forecasts and more aggressive climate policies to meet China's environmental targets.