Mun Ho (Harvard U. & Resources for the Future) Dale Jorgenson (Harvard University)

1. Reconciling China’s economic development with global and local environmental protection Mun Ho (Harvard U. & Resources for the Future)Dale Jorgenson (Harvard University) Mitigation of air pollution and climate Change in China Workshop Oslo, October 17-19 2004

2. A research project of Institute of Environmental Science & Engr, Tsinghua University Harvard University Center for the Environment Hao Jiming, Wang Shuxiao, Liu Bingjiang, Lu Yongqi Sue Greco, James Hammitt, Mun HO, Dale Jorgenson, Jon Levy, Chris Nielsen, ZHOU Ying http://www.ksg.harvard.edu/cbg/ptep/ http://environment.harvard.edu/activities/sponsored/chinaproject/

3. Topics • Intake Fraction method to estimate health damages from air pollution • Contrast with World Bank (1997) Clear Water Blue Skies method of Lvovsky et. al. (1997,2000) • Estimate of damages by industry; by health effects; etc. • Use of ‘green tax’ policies to reduce damages and promote tax reform and economic efficiency

4. Methodology Goal: Efficient national policies to reduce local health damages, also calculate coincident CO2 reductions Step 1: Estimate marginal damages by industry (in terms of yuan of damages per yuan of output). Step 2: Estimate marginal damages by type of fuel Step 3: Construct an economic model that traces output and inter-industry transactions, including fuel use; traces consumption and investment. Use model to simulate ‘green tax’ policy effects.

5. Table 1 Sector emissions in 1997

6. Table 2 Emissions, fuel use and output 1997

9. THE PROBLEM • Have one number for pollutant concentration which is believed to be the determinant of health effects and other environmental damage • Have emissions for each industry • Need to assign each industry’s contribution to the ambient concentration of PM and SO2 • Not feasible to model each emission source • Lvovsky et. al. (1997,2000) method • Intake Fraction (iF) method

10. Lvovsky et. al. (1997,2000) innovation: 3 γ’s derived from air dispersion models Each industry assigned to low, medium or high stack height Strength: simple to implement, comprehensive Weakness: x=TSP, EM(TSP) is only primary TSP, ignoring secondary TSP; not all Cx due to human activity; not all Cx due to local emissions;

11. Intake fractions for Electricity, Iron&Steel, Cement, • Chemicals, Transportation • (Wang Shuxiao, Liu Bingjiang, Zhou Ying et.al. ) • - fraction of emissions from a source that is eventually ingested • samples of plants from 5 cities • used short range air dispersion model (ISCLT) to calculate • concentration due to primary PM and SO2 • - used Calpuff model to long range effects, secondary SO4,NO3 • used detailed population location data to estimate exposures • scaled up to national intake fractions using national samples

12. iF calculation in Jinan city: Population, concentration, pop*conc

13. Estimated intake fractions for China industries

14. Dose-response and valuations of health effects, base parameters

15. Health damages due to outdoor air pollution, 1997

16. Marginal damages, total damages by industry

17. Marginal damages from fuels, averaged over all industries

18. National policies to reduce local and global pollution Efficient general policy is tax on emissions but not feasible - analyze 2 policies: i) tax on output proportional to health damage caused ii) tax on fuels proportional to health damage caused - model calculates both costs and benefits of pollution control : - estimate reduction in health damages - estimate change in output of each sector, change in Consumption, change in GDP growth (some cost, e.g. pollution control equipment, not explicitly considered)

19. Methodology for cost-benefit analysis - estimate a base case growth path of the economy where pollution taxes = 0 - estimate alternative case where tax(j) = MD(j); new revenues used to cut other pre-existing taxes - compare the time paths of both cases

21. Base case projection of the economy (30 year growth)

22. Effects of an output tax based on pollution damages

23. Effects of an output tax based on pollution damages

24. Effects of a fuel tax based on damages

25. Effects of a fuel tax based on damages

27. Conclusions • Magnitude of control costs of well designed policies is • much smaller than the local benefits, even ignoring • global benefits. • - Local pollution reduction would very likely involve carbon • emission reduction. • Green tax policies help tax reform goals • - The short run adjustment costs may be higher than estimated • here leading to less control efforts by China if unaided by • outside world. Hence, rich countries may play a very • important role in tipping the balance towards greater efforts • and sooner.

28. Reconciling China’s economic development with global and local environmental protection Mun Ho (Harvard U. & Resources for the Future) ho@rff.org Dale Jorgenson (Harvard University) djorgenson@harvard.edu Papers available at: www.ksg.harvard.edu/cbg/ptep/ http://environment.harvard.edu/uce/

29. Economic activity, pollution and health: A general • equilibrium framework • Emissions = f( output, fuel use, control technology) • Concentration = f(Emissions, meteorology, geography) • Health effects = f(Concentration, population; DR) • Value of Damage = f( Health effects, valuation(income)) • Costs of control = f( level of reduction, control tech, ...) • f(output, prices, income, consumption, investment) = 0 • Costs of control versus Value of Damage reduction

30. Lvovsky-Hughes simple method for all sectors: EM : emissions x=PM10,SO2 f=coal, oil, gas QI : sector output j = sector 1,2,... AF : fuel input c=high, medium, low height C : concentration

31. LH Method, continued: DR : dose-response HE : health effect h=Mortality,CB,.... POP: population V : valuation in ¥

32. Damages by sector. Marginal damage per unit emissions in sector j: Marginal damage per yuan(1997) of output: Marginal total damage of sector j: Sector j's share of total marginal damages:

33. iF method to estimate sector damages: Health effects due to sector j: Marginal total damages due to j: