Distributional Implications of Proposed US Greenhouse Gas Control Measures Sebastian Rausch, Gilbert E. Metcalf, John M. Reilly and Sergey Paltsev

Transcription

1 Distributional Implications of Proposed US Greenhouse Gas Control Measures Sebastian Rausch, Gilbert E. Metcalf, John M. Reilly and Sergey Paltsev Paper prepared for the UC-UI-RFF Energy Policy Symposium Distributional Aspects of Energy and Climate Policy January 20-21, 2010

2 Presentation USREP model Policy Scenarios Results Summary

3 MIT USREP Model: Overview MIT USREP (US Regional Energy Policy) model is a new multi-region, multi-sector, multi-household computable general equilibrium (CGE) model of the US economy for analyzing US energy and greenhouse gas policies Recursive dynamic model similar to the MIT EPPA (Emissions Prediction and Policy Analysis) Model (Paltsev et al, 2005) Captures heterogeneity across regions and income groups in the United States

4 MIT USREP Model: Database Base year calibration based on a new integrated statelevel economic-energy dataset for the US for the year 2006 which merges together : Economic data from IMPLAN (Social Accounting Matrices for each state) Physical energy and price data from EIA s State Energy Data System (SEDS) Population and household data from US Census Bureau and GHG inventories data from EPA (for Kyoto gases) Fossil fuel reserves data from USGS and DOE, and highresolution wind resource data from NREL Tax data from IMPLAN and the NBER TAXSIM tax simulator

5 MIT USREP: Sectoral Breakdown and Primary Input Factors Region Sectors Primary Input Factors Alaska (AK) California (CA) Florida (FL) New York (NY) New England (NENGL) South East (SEAST) North East (NEAST) South Central (SCENT) North Central (NCENT) Mountain (MOUNT) Pacific (PACIF) Non-Energy Agriculture (AGRIC) Services (SERV) Energy-Intensive (EINT) Other Industries (OTHR) Transportation (TRAN) Energy Coal (COAL) Conventional Crude Oil (OIL) Oil from Shale (OIL) Refined Oil (ROIL) Natural Gas (GAS) Electric: Fossil (ELEC) Electric: Nuclear (NUC) Electric: Hydro (HYD) Advanced Technologies Capital Labor Land Crude Oil Shale Oil Natural Gas Coal Nuclear Hydro Wind

6 Advanced Technologies Coal Gasification Biomass Liquids Biomass Electricity Intermittent Wind Wind With Gas Backup Wind With Biomass Backup Advanced Gas (Natural Gas Combined Cycle) Advanced Gas With Carbon Capture And Sequestration Advanced Coal With Carbon Capture And Sequestration Advanced Nuclear

7 MIT USREP Model : Regional Aggregation State-level dataset allows flexible regional aggregation In our analysis, we focus on 12 model regions to capture differences in electricity costs and to help focus on how different regions and states differ

8 MIT USREP Model: Income Classes Table 1. Income Classes Used in the USREP Model and Cumulative Population. Income class Annual Income (2006$) Cumulative Population for whole hhl hh10 hh15 hh25 hh30 hh50 hh75 hh100 hh150 Less than $10,000 $10,000 to $15,000 $15,000 to $25,000 $25,000 to $ $30,000 $30,000 to $50,000 $50,000 to $75,000 $75,000 to $100,000 $100,000 to $150,000 $150,000 plus US (in %) a a Based on Consumer Expenditure Survey Data for 2006.

9 MIT USREP Model: Key Features Production and consumption technologies are represented by nested constantelasticity-of-substitution (CES) functions with identical structure as in EPPA model Perfect competition in product and factor markets Vintage capital structure is similar to EPPA model and distinguishes between malleable and non-malleable capital - Malleable capital is fully mobile across industries and regions - Vintaged capital is region and industry specific Labor supply is determined by the household choice between leisure and labor and we assume that labor is fully mobile across industries in a region but it is immobile across US regions Integrated market for fossil fuel resources with regional ownership of fossil fuel resources distributed across regions in proportion to capital income Armington (1969) assumption of product heterogeneity for imports and exports among states and regions and with foreign goods Regional energy supply - Fossil fuels: based on reserves data from US Geological Service and the DOE, and resource depletion model as in EPPA model - Regional supply curves for wind based on high-resolution wind resource data from NREL and a levelized cost approach - Biomass: supply curves from Oakridge National Laboratories

10 Background & Objectives From viewpoint of policymakers, distributional effects of policies are often more important than efficiency considerations ( rectangles trump triangles ) Many climate policy provisions are designed to blunt the impact of the legislation on lower and middle income households, and to balance regional effects To date, much of the distributional analysis has been done as a side calculation or in an Input-Output framework thereby neglecting behavioral responses to relative price changes and income effects USREP analysis is general equilibrium in nature and incorporates heterogeneity

11 Policy Overview Three cap and trade proposals analyzed: Waxman-Markey (WM) Kerry-Boxer (KB) Cantwell-Collins (CANT) Focus on cap and trade allowance allocation Other aspects of these proposals not modeled

12 Policy Modeling Details All three seek an overall reduction of GHG emissions in the US to ~80% below 2005 levels by 2050 with intervening targets. Cap and trade components of the bills cover most of the economy s emissions but not necessarily all of them, with other measures directed toward uncapped sectors. Assume the national goals are met with a cap and trade system that covers all US emissions except for land use CO 2 sources (or sinks). All of these proposals including banking and limited borrowing provisions. WM and KB allow offsets; CANT does not. We allow offsets in CANT to focus on distributional differences across proposals

13 Allowance Allocation in Bills Waxman Markey Allowance Allocation

14 Allowance Allocation Incidence Allowance Allocation LDC allowance distribution Protection for lowincome households Allocation to affected industries Technology funding Treatment in Model Lump-sum transfer based on specific energy consumption Lump-sum transfer to households with income less than $30,000 Lump-sum transfer based on capital income Allocated to regions based on energy use; within regions allocated to households on a lump-sum basis See Table 3 in paper for more allocation detail

15 Allowance Allocation in Bills II KB Scenario WM CANT Scenario

16 Scenarios Scenario WM_LS KB_LS Description Waxman-Markey. A fraction of the allowance value has to be retained to satisfy revenueneutrality. Kerry-Boxer. A fraction of the allowance value has to be retained to satisfy revenue-neutrality. CANT_LS WM_TAX Cantwell-Collins. A fraction of the allowance value has to be retained to satisfy revenueneutrality. Waxman-Markey. Full allowance value is allocated. Revenue-neutrality is achieved by increasing marginal personal income taxes for each region and income class by an equal amount (in percentage points). CANT_TAX Cantwell. Full allowance value is allocated. Revenue-neutrality is achieved by increasing marginal personal income taxes for each region and income class by an equal amount (in percentage points). KB_TAX Kerry-Boxer. Full allowance value is allocated. Revenue-neutrality is achieved by increasing marginal personal income taxes for each region and income class by an equal amount (in percentage points). Note: All scenarios assume the medium offset case from the Appendix C of Paltsev, et al. (2009) with identical assumptions about supply and costs of offsets. This corresponds to a 203 bmt cumulative emissions target for

17 Aggregate Emissions: Reference: 298 bmt Policy: 203 bmt Emissions

18 Aggregate Impacts

19 Distributional Impacts Over Income WM_LS

20 CANT_LS

21 Treatment of Revenue Neutrality Lump-Sum WM Modeled Here Tax

22 Welfare Impacts by Region WM_LS

23 Welfare Impacts by Region CANT_LS

24 Incidence Drivers Burden a combination of spending side impacts and income side impacts Counterfactual analysis to decompose different burden forces assume identical consumption shares across income groups (isolates income side impacts) assume identical labor and capital income shares across income groups (isolates spending side impacts)

25 Fixed Expenditure Shares Observed Income and Expenditure Shares Fixed Factor Income Shares WM_LS

26 Summary: Efficiency CANT less costly than KB or WM CANT distributes less to lower income households than WM or KB Lump-sum distributions to low-income households increase energy demand (income effect) Lump-sum distributions to higher income households are disproportionately saved Classic equity-efficiency trade-off Note: this occurs in the lump-sum scenario

27 Summary: Distribution WM and KB appear to overcompensate in early years low-income households South Central, Texas, Florida CANT more distributionally neutral across income groups and regions Differences dissipate over time All three policies modestly progressive in short and long-run Distributional impacts driven more by variation in income sources than variation in spending patterns