Click for next page ( 53


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 52
Chapter 3 Energy-Related Tax Expenditures PLAN OF THE CHAPTERS ON SUBSTANTIVE PROVISIONS Earlier chapters have outlined the scope of the present report as well as the approach that the committee has taken in addressing its charge. Chapters 3 through 6 present the detailed analysis of the impacts of different tax provisions on greenhouse gas (GHG) emissions. As was described in the last two chapters, the bulk of the results are based on specific modeling for this report undertaken by four external contractors. The plan is the following. The present chapter examines the impact of the major energy-related tax expenditures. Chapter 4 reviews energy-related excise taxes. Chapter 5 analyzes the subsidies and regulations affecting biofuels, a sub- set of the energy-related tax expenditures characterized by significant subsidies and a particularly complex set of markets and regulations. Finally, Chapter 6 examines a number of broad-based tax expenditures to determine whether they may have a significant impact on greenhouse gas emissions. Each of the chapters has a parallel structure. It describes the major tax ex- penditures, describes the modeling efforts undertaken by the modeling groups, explains the results, and then presents the overall conclusions. Chapter 7 then presents an overall summary of the results along with findings and recommenda- tions. ENERGY-RELATED TAX EXPENDITURES In 2011 the 10 largest tax expenditures that directly affected the energy sector resulted in a loss of $16.9 billion in tax revenues (see Chapter 2, Table 2-1). The largest of these, the alcohol fuel credit and the biodiesel production tax credit, are discussed in Chapter 5. This chapter analyzes the impact of other en- ergy-related tax expenditures, focusing on credits for electricity production from renewable resources (the renewable energy production and investment tax cred- its) and the depletion allowance tax preference (the tax provisions that allows 53

OCR for page 52
Chapter 3 Energy-Related Tax Expenditures PLAN OF THE CHAPTERS ON SUBSTANTIVE PROVISIONS Earlier chapters have outlined the scope of the present report as well as the approach that the committee has taken in addressing its charge. Chapters 3 through 6 present the detailed analysis of the impacts of different tax provisions on greenhouse gas (GHG) emissions. As was described in the last two chapters, the bulk of the results are based on specific modeling for this report undertaken by four external contractors. The plan is the following. The present chapter examines the impact of the major energy-related tax expenditures. Chapter 4 reviews energy-related excise taxes. Chapter 5 analyzes the subsidies and regulations affecting biofuels, a sub- set of the energy-related tax expenditures characterized by significant subsidies and a particularly complex set of markets and regulations. Finally, Chapter 6 examines a number of broad-based tax expenditures to determine whether they may have a significant impact on greenhouse gas emissions. Each of the chapters has a parallel structure. It describes the major tax ex- penditures, describes the modeling efforts undertaken by the modeling groups, explains the results, and then presents the overall conclusions. Chapter 7 then presents an overall summary of the results along with findings and recommenda- tions. ENERGY-RELATED TAX EXPENDITURES In 2011 the 10 largest tax expenditures that directly affected the energy sector resulted in a loss of $16.9 billion in tax revenues (see Chapter 2, Table 2-1). The largest of these, the alcohol fuel credit and the biodiesel production tax credit, are discussed in Chapter 5. This chapter analyzes the impact of other en- ergy-related tax expenditures, focusing on credits for electricity production from renewable resources (the renewable energy production and investment tax cred- its) and the depletion allowance tax preference (the tax provisions that allows 53

OCR for page 52
54 Effects of U.S. Tax Policy on Greenhouse Gas Emissions capital costs for oil and gas wells to be recovered as a percent of revenues in- stead of costs). The major provisions are analyzed using the National Energy Modeling System for the National Academy of Sciences (NEMS-NAS) model. The NEMS-NAS model was unable to capture other provisions—the special tax rate on reserve funds set aside by firms for the decommissioning of nuclear power plants and the credit for energy-efficiency improvements to existing homes— and these are discussed qualitatively. Some introductory remarks on the impacts of the provisions will set the stage. Each of the provisions discussed in this chapter has only a small impact on the greenhouse gases emitted in the United States. This is to be expected giv- en the nature and magnitude of the tax expenditures. Energy-related tax expendi- tures alter either the supply of or the demand for various types of energy. In some cases, the magnitude of the shift in supply or demand is small; hence, so is the estimated impact on energy consumption and CO2 emissions. For example, the excess of percentage over cost depletion for natural gas lowers the cost of producing natural gas; however, the provision affects only independent produc- ers, so the impact on natural gas production is small. In the case of the renewable energy production tax credit, the credit, on a per-unit basis, is substantial: It lowers the cost of electricity production from wind by 2.3 cents per kWh, or almost 20 percent of the average retail price. However, the base to which the credit is applied (i.e., the fraction of electricity generated from wind) is small and so therefore is the impact on CO2 emissions.1 The magnitude of estimated impacts also depends on the price responsive- ness of energy consumption and production (technically, the price elasticities of supply and of demand), which depends on many factors, including the time horizon considered. Generally speaking, the price elasticity of demand for ener- gy is higher in the long run than in the short run, since users are more easily able to adjust stocks of energy-using capital (e.g., appliances) when given longer time horizons. The same is true of the elasticity of energy supply: The price elasticity of supply of electricity from a particular fuel source will, in general, increase with the length of the time horizon considered. It is also true that the estimated impact of a tax credit will, in most cases, be greater the larger the price elasticities of demand and supply are for the affected energy source. Thus, long-run impacts are likely to differ from short-run impacts. 1 Only 2.9 percent of electricity in the United States was generated from wind power in 2011.

OCR for page 52
Energy-Related Tax Expenditures 55 FINDINGS FROM PRIOR LITERATURE Several scholars and researchers have investigated the impact of the pro- duction tax credit on new installation of renewable electricity generation capaci- ty. Some of those studies (Wiser, 2007; Metcalf, 2010) found that the tax credit for production of electricity from certain renewable resources (commonly called the Production Tax Credit, or PTC) did reduce the cost of installing new renew- able generating capacity, especially for wind, but was costly to the Treasury Department (Metcalf, 2007). Those studies concluded that the PTC had in- creased the amount of installed generation capacity. One study (Price, 2002) found that the credit had no significant impact on new installations when state renewable portfolio standards were taken into account. Regardless of the find- ings on capacity, none of these studies estimated the production tax credit’s im- pact on greenhouse gas emissions. For other provisions discussed in this chapter, the existing literature was even thinner. Papers considering energy-efficiency improvement to existing homes looked at the possibility that homeowners taking the credit would have made the improvements in the absence of the incentive but did not evaluate any anticipated effect on emissions (Hirst et al., 1982; Metcalf, 2007; Jaffe and Stavins, 1994). Little has been written on the special tax rate on nuclear decom- missioning reserve funds. A few studies have considered the impact of the per- centage depletion allowance rules on levels of investment by firms in oil and gas wells (Krueger, 2009, and Metcalf, 2009). One other study estimated the impact of this subsidy on global petroleum production, while other researchers argue that global markets would largely offset any changes in U.S. production (Metcalf, 2007, and Bogdanski, 2011). None of these studies explicitly consid- ered the impact of these provisions on GHG emissions. ANALYSIS USING NEMS The committee considered several modeling approaches to estimating the impact of the energy-sector tax provisions. As was explained in Chapter 2, the primary analysis was conducted using a version of the U.S. Energy Information Administration’s (EIA) National Energy Modeling System, or NEMS, for the committee, with modifications made by a firm that maintains the model, OnLocation, Inc. This modified model is labeled the NEMS-NAS model.2 2 The version of NEMS used in this study was run by OnLocation, Inc., for the Nation- al Academies and was run without a link to a macroeconomic model. The committee omitted the macroeconomic linkage because it included business-cycle linkages that were thought inappropriate for the long-run analysis undertaken here. Some model modifica- tions were made in order to represent the NAS tax policy cases. To distinguish these re-

OCR for page 52
56 Effects of U.S. Tax Policy on Greenhouse Gas Emissions The Reference Scenario was developed starting with the Energy Infor- mation Administration’s Annual Energy Outlook (AEO) for 2011 (U.S. Energy Information Administration, 2011) as the benchmark for the U.S. energy system (see Table 3-1). As with the other modeling efforts reported in later chapters, we standardized the modeling runs by assuming that the provisions of the Internal TABLE 3-1 Assumptions Underlying NEMS Scenarios Reference High Oil Low Gas (AEO 2011) High Macro Price Prices No RPS GDP Growth 2.7% 3.2% 2.7% 2.7% 2.7% (real annual) 2035 World Oil Price $125/bbl $125/bbl $200/bbl $125/bbl $125/bbl (2011 USD) U.S. Shale Gas 50% higher Reserves Renewable Portfolio Yes Yes Yes Yes No Standards Revenue Code (IRC) as of 2011 would remain in force through 2035 in the baseline (Reference) scenario. A set of counterfactual scenarios were then run, each scenario removing a particular tax provision beginning in 2010. By com- paring the counterfactual (no-tax-preference) scenario to the reference (the base- line with-tax-preference) scenario, we were able to estimate the impact of the provision on GHG emissions and other related energy system and economic variables. Note that for these simulations, as in all partial equilibrium models in this study, the economic impacts of the government’s revenue gains or losses from changing a provision was omitted from the calculations. We do consider the impacts of recycling the gained or lost revenues in Chapter 6 when we examine the results for broad-based provisions studied with a general equilibrium model. One of the issues arising in conducting modeling calculations of the kind reported here is to understand the uncertainties associated with the results. As with other models in this report, a formal quantification of the key uncertainties in NEMS was not conducted for this study. Based on its review and understand- ing of energy modeling, the committee determined that a full uncertainty study was not feasible within the constraints and resources available. However, for the NEMS model, sensitivity analyses were conducted to determine the impact of alternative assumptions on the results. In AEO 2011, gross domestic product (GDP) is assumed to grow at an an- nual rate of 2.7 percent, and the price of oil in 2035 is assumed to be $125 per sults from EIA results, this version of NEMS is called NEMS-NAS. For more detailed results and further materials, see the online Appendix to this report.

OCR for page 52
Energy-Related Tax Expenditures 57 barrel (2011 USD). To examine how sensitive the results are to plausible alterna- tive assumptions, three additional economic scenarios were run. These examined how alternative market conditions might affect the impacts of the tax provisions relative to the Reference scenario. We label these the High-Macroeconomic- Growth scenario, the High-Oil-Price scenario, and the Low-Gas-Prices scenario.  The High-Macroeconomic-Growth scenario was run assuming a real GDP growth rate for the United States of 3.2 percent per year.  The High-Oil-Price scenario assumed a 2035 price of $200 per barrel (2011 USD).  The Low-Gas-Prices scenario assumed 50 percent higher ultimate re- covery of natural gas from shale relative to the Reference scenario. The inclusion of the High-Oil-Price and the Low-Gas-Price scenarios helped capture the impact of some of the major shifts in energy production and supply in the United States and then evaluated these changes against the influence of the tax provisions studied. For example, cumulative CO2 emissions between 2015 and 2035 in the U.S. energy sector are projected to be 141,201 MMT in the Reference scenario. The High-Macroeconomic-Growth scenario projects cumu- lative emissions of 147,675 MMT, or about 4.5 percent greater than the Refer- ence scenario. The High-Oil-Price scenario projection is 138,695 MMT, or about 6 percent lower, and the Low-Gas-Prices scenario projection is 140,616 MMT, or about 1.5 percent higher. A comparison of these changes in emissions can be seen in Tables 3-2a and b, which summarize the impacts of the selected provisions. All four scenarios assumed that nonrenewable federal tax incentives and state Renewable Portfolio Standards (RPS) remain in place. The assumptions about environmental regulations are complicated in both the modeling and in reality. NEMS-NAS assumes that the Clean Air Interstate Rule (CAIR) will be implemented; however, it assumes that two other important rules are not imple- mented (the Cross-State Air Pollution Rule, CSAPR, and the rule on Mercury and Air Toxics Standards, MATS). The legal status of these rules is somewhat different from the NEMS-NAS assumptions as of the time this report was completed. The MATS and CAIR are currently in force. The CSAPR is in a complicated legal limbo. The D.C. Circuit Court of Appeals vacated CSAPR in August 2012 and reinstated CAIR. Howev- er, in April 2013, the U.S. Environmental Protection Agency (EPA) asked the U.S. Supreme Court to review that decision. As a result, the lower court’s ruling leaves CAIR in place until the litigation is resolved. The major implication for the present study is that these regulations might have a major impact on future electrical generation from coal. This, in turn, could influence the impact of tax provisions that affect the power sector. In addition to the four aforementioned scenarios, we ran a No-RPS scenar- io. This was identical to the Reference scenario, but assumed that state Renewa-

OCR for page 52
58 Effects of U.S. Tax Policy on Greenhouse Gas Emissions ble Portfolio Standards would not remain in force over the modeling period. RPS require roughly 420 billion kWh of qualified renewable generation by 2035; hence, removing these requirements is likely to affect the impact of tax expenditures designed to promote renewable energy. When analyzing the impact of removing the production tax credit/ invest- ment tax credit (PTC/ITC), we ran all but the High-Oil-Price scenario. In ana- lyzing the effects of replacing the percentage depletion allowance with cost de- pletion, we ran all scenarios except the No-RPS scenario. NEMS-NAS RESULTS FOR THE PRODUCTION TAX CREDIT AND INVESTMENT TAX CREDIT A summary of CO2 emissions impacts from two main provisions analyzed with the NEMS-NAS model is presented in Table 3-2a. Cumulative and annual average CO2 emissions in MMT from the U.S. energy sector are provided for the period 2010 to 2035 across the Reference scenario and the additional four sensi- tivity scenarios. For comparison, the impacts of removing the renewable energy production and investment tax credits and excess of percentage over cost deple- tion provisions are also provided for each scenario. These calculations using the NEMS-NAS model indicate that the impacts of changes in these two tax provisions on CO2 emissions are small across a spec- trum of alternative market and regulatory conditions. The impact across policies and economic scenarios is between -0.1 and +0.3 percent of cumulative emis- sions from the energy sector over the period 2010-2035. Results for the special Reference No-RPS scenario are included in Table 3-2b. This is the case where the state RPS are not included in the Reference sce- nario and the PTC/ITCs are also removed. For comparison, the main Reference scenario emissions are also included. NEMS-NAS model results indicate a greater impact of removing the PTC/ITCs in the situation when there are no state RPS. For the No-RPS scenario, there is an increase of 0.5 percent in both cumulative and average annual emissions from the energy sector over the period 2010-2035. We discuss the finding in more detail in the following sections. RENEWABLE ELECTRICITY TAX CREDITS (ENERGY PRODUCTION AND INVESTMENT TAX CREDITS) Legal Description and Expected Impact At the time of our analysis, taxpayers could claim a nonrefundable credit of 2.3 cents per kWh of electricity generated from wind, biomass, and geothermal en- ergy resources and a credit of 1.1 cents per kWh for electricity generated from solar energy, small irrigation power, and municipal solid waste (trash

OCR for page 52
TABLE 3-2a Summary of CO2 Emissions Impacts Difference from Respective Reference Tax Policy Scenario Percent Difference U.S. Energy Sector Scenario No Percentage No Percentage CO2 Emissions (MMT) (Base Value) No PTC/ITC Depletion No PTC/ITC Depletion Cumulative 2010-2035 Reference Scenario 141,201 360 -37 0.3% -0.03% High Economic Growth 147,675 393 58 0.3% 0.04% High Oil Prices 138,695 Nc 286 nc 0.2% Low Gas Prices 140,616 -129 11 -0.1% 0.0% Average Annual 2010-2035 Reference Scenario 5,883 15 -1.5 0.3% -0.03% High Economic Growth 6,153 16 2.4 0.3% 0.04% High Oil Prices 5,779 Nc 12 nc 0.2% Low Gas Prices 5,859 -5.4 -0.5 -0.1% 0.0% Source: NEMS-NAS model for this study. Note: “nc” is not calculated; MMT = million metric tons of CO2. 59

OCR for page 52
60 TABLE 3-2b Summary of CO2 Emissions Impacts Difference from U.S. Energy Sector Reference Scenario Reference Reference No-RPS Scenario Percent Difference CO2 Emissions (MMT) (Base Value) No-RPS Scenario No PTC/ITC No PTC/ITC Cumulative 2010-2035 Reference Scenario 141,201 141,576 762 0.5% Average Annual 2010-2035 Reference Scenario 5,883 5,899 32 0.5% Source: NEMS-NAS model results for this study. Note: MMT = million metric tons of CO 2.

OCR for page 52
Energy-Related Tax Expenditures 61 combustion and landfill gas) for the first 10 years after a facility is built.3 This provision is commonly known as the Production Tax Credit (PTC). The Ameri- can Recovery and Reinvestment Act (ARRA) of 2009 (P.L. 111-5) temporarily made the subsidy available as a cash grant in lieu of the tax credit, thereby mak- ing it refundable, that is, available to firms with no tax liability. The ARRA also enabled firms to take a 30 percent investment tax credit (ITC) in lieu of the 10- year PTC. Roughly 75 percent of all credits have gone to wind generation facili- ties and 16 percent to biomass facilities. In 2010, the Treasury Department esti- mated that the PTC/ITC reduced government revenues by $4.2 billion, compris- ing 36 percent of the estimated aggregate energy-related tax expenditures.4 The PTC and ITC lower the cost of electricity generated from renewable resources, encouraging increased substitution of renewable resources for coal or other fuels for electricity generation. They also lower the price of electricity and thereby increase overall demand. In many states, RPS also encourage electricity generation from renewable sources, and the tax credits reduce the cost of com- plying with the RPS. The amount of GHG reduction from renewable energy generation depends on the source of the electricity. According to the EPA, the national average car- bon dioxide output rate for electricity generated in 2009 was 1.2 lb CO 2 per kWh for delivered electricity. 5 To the extent that the PTC/ITC encourages the substi- tution of electricity from wind or solar power for electricity from fossil fuels, CO2 emissions are expected to decrease. Modeling in NEMS We present the results of this scenario in detail. This will help readers un- derstand the logic of the modeling results as well as give a glimpse into the complexity of the energy system and the difficulty of accurately capturing all the forces at work. In each scenario, the baseline includes several production tax credits and investment tax credits related to renewable power generation, extending to the end of the NEMS-NAS forecast period (i.e., 2035). For the counterfactual No- 3 The legislation, IRC section 45, set the credit at 1.5 cents per kWh of electricity gen- erated from wind, biomass, and geothermalenergy resources and half that for electricity generated from solar energy, small irrigation power, and municipal solid waste (trash combustion and landfill gas). The credit is refundable and indexed to inflation. At the time of the analysis the inflation-indexed rate was 2.2 cents per kWh. At the time this report went to press, June 2013, the rate had risen to 2.3 cents per kWh. 4 Percentage computation from estimates in Analytical Perspectives Table 17-1. 5 EPA (U.S. Environmental Protection Agency). 2012. Emissions & Generation Re- source Integrated Database (eGRID) 2012 Version 1.0, Year 2009 Summary Tables [on- line]. Available: http://www.epa.gov/cleanenergy/documents/egridzips/eGRID2012V1_ 0_year09_SummaryTables.pdf [accessed June 5, 2013].

OCR for page 52
62 Effects of U.S. Tax Policy on Greenhouse Gas Emissions PTC/ITC scenario, these tax credits were removed starting in 2010.6 State re- newable portfolio standards remain unchanged except in the No-RPS scenario. In all scenarios, nonrenewable tax credits remain unchanged, including credits for advanced coal, nuclear,7 and combined heat and power.8 The ARRA enabled renewable project developers to choose a cash grant or a 30 percent ITC in lieu of a 10-year PTC. The NEMS-NAS Reference sce- nario assumes the following (values in 2009 USD, increasing with projected inflation):  2.3¢ per kWh PTC for onshore wind and geothermal;  1.1¢ per kWh PTC for landfill gas and hydroelectric facilities;  30 percent ITC per grant for biomass, offshore wind, and utility solar systems;  30 percent ITC per grant for distributed rooftop photovoltaic (PV) sys- tems and small wind turbines. In NEMS-NAS analysis, investment tax credits and cash grants are both treated as a percentage reduction in the capital cost of the technology and are therefore identical. Under current law, most of these provisions have expired or are scheduled to expire; however, under the committee’s methodology, they are extended through 2035 in the baseline analysis for each scenario. An important feature of the technological assumptions in the NEMS-NAS model is the introduction of learning by doing (see the discussion of this issue in Chapter 2).9 The “learning rate” in the NEMS-NAS model is defined as the frac- tional reduction in capital costs for every doubling of cumulative capacity. The learning rates are determined separately for each component of the system. More specifically, each new technology is broken into its major components, and each component is identified as revolutionary, evolutionary, or mature. There is a minimum linear cost reduction for each component and also a formula for cost reductions based on new capacity additions. The resulting learning factor is based on whichever is greatest. The learning rates for onshore wind power, land- fill gas, and hydropower are the same as for conventional fossil-fuel power gen- 6 Facilities claiming the PTC beginning before 2012 are assumed to receive the credit for a 10-year period in the No-PTC/ITC scenario. 7 The Energy Policy Act of 2005 provides a 20 percent investment tax credit for Inte- grated Coal Gasification Combined Cycle capacity and a 15 percent investment tax credit for other advanced coal technologies. Both of these are limited to 3 GW. There is also a production tax credit of 1.8 cents per kWh for new nuclear capacity beginning operation by 2020. 8 The Energy Improvement and Extension Act of 2008 provides for a 10 percent tax credit for combined heat and power projects, applicable to only the first 15 MW of a system smaller than 50 MW. The system must be placed in service between October 2008 and December 2016. 9 For full details and specifications, see Electricity Market Module, AEO Assumptions at http://www.eia.gov/forecasts/aeo/assumptions/pdf/electricity.pdf.

OCR for page 52
70 Effects of U.S. Tax Policy on Greenhouse Gas Emissions MMT (or 0.8 percent higher) than the Reference scenario. Thus, the RPS poli- cies have, according to the NEMS-NAS modeling, virtually the same effect in reducing GHG emissions as the PTC/ITC tax preferences. Last, we considered the effect of removing the PTC/ITC tax preferences if the RPS mandates are not in place (No-RPS and No-PTC/ITC scenario). Ac- cording to the simulation, cumulative power-sector CO2 emissions are 56,124 MMT (or 1.5 percent) above the No-RPS/Reference scenario. In other words, the increase of CO2 emissions, the result of removing the PTC/ITC tax credits, are about twice as large if RPS mandates are not in effect. The finding on the role of the RPS is important. It indicates that the regu- latory mandates constrain production and emissions. As a result, the impacts of tax policies on emissions are reduced, in this case by half, when the regulatory mandates are considered. This finding is similar in other results, particularly the impacts of the biofuels mandates analyzed in Chapter 5. While the exact magni- tudes will be sensitive to the detailed specifications, the general point about in- cluding regulations and mandates in estimating the impacts of tax policy should be emphasized. Summary The committee’s analysis of the tax provisions for renewable electricity indicates that they lower CO2 emissions. This finding confirms the first-order intuition that lowering the cost of low-carbon renewable fuels will lead to substi- tution away from high-carbon fossil fuels. The reduction in CO2 emissions associated with the PTC/ITC is, however, small, amounting to about 0.3 percent of CO2 emissions from the energy sector in the Reference scenario. If the revenue lost as a result of the PTC/ITC is divid- ed by the reduction in CO2 emissions, just under $250 in revenues are lost per ton of CO2 reduced. While this does not represent the social cost of reducing the ton of CO2 emissions (because revenue losses are not a dead-weight loss, as ex- plained in Chapter 2), the fiscal cost per ton of CO 2 reduced is high relative to other, more efficient approaches. EXCESS OF PERCENTAGE OVER COST DEPLETION Legal Description and Expected Impact The second provision described in detail is the depletion allowance. The depletion allowance permits owners of oil and gas wells to deduct the decline in the value of their reserves as oil or gas is extracted and sold. The allowance, which is a form of cost recovery for capital investments, can be calculated using either cost depletion or percentage depletion. Under cost depletion, the annual deduction is equal to the unrecovered cost of acquisition and development of the resource times the estimated proportion of the resource removed during that

OCR for page 52
Energy-Related Tax Expenditures 71 year. Under percentage depletion, taxpayers deduct a percentage of gross in- come associated with the sale of the resource. Percentage depletion for oil and gas is currently limited to U.S. production by independent companies up to a certain limit, currently set at 15 percent of costs associated with production. Per- centage depletion typically allows for total deductions that exceed the cost of capital invested to acquire and develop the resource. A percentage depletion rate of 22 percent applies to natural gas sold under fixed contracts. The excess of percentage over cost depletion affects primarily natural gas production. Since only independent producers may utilize this tax subsidy, the Joint Committee on Taxation (JCT) estimates its total impact to the Treasury at $0.5 billion for 2010. With increasing production, it reaches $1.0 billion in 2014, with a 5-year total of $4.1 billion. For natural gas development and production, the largest component of capital investment is the drilling of exploration and production wells. Any policy affecting capital investment directly, as the depletion allowance does, affects this front-end activity in the natural gas market life cycle. In the Cost-Depletion scenario, where cost depletion replaces the percentage depletion allowance, cap- ital recovery is slower, resulting in higher drilling costs, and reducing incentives to explore and develop new supply. Less investment in drilling would be ex- pected to reduce domestic production and raise the price of natural gas. Modeling in NEMS Explicit treatment of tax deductions associated with resource depletion is limited in the NEMS model. The tax treatment of depletion in the coal-mining sector is not explicitly modeled in NEMS-NAS and is therefore excluded from the Reference and Cost-Depletion scenarios. Current estimates are that the de- pletion allowance for coal is very small, so this is unlikely to have a large impact on the results.11 The treatment of depletion in the oil and gas sector is complicat- ed by the fact that the depletion allowance depends on firm size (small inde- pendent, large independent, and major producer). To capture this:  In the Reference scenario, a 15 percent depletion allowance is assumed for all onshore activity, all of which is assumed to be performed by in- dependent oil and gas producers.  All offshore gas and oil production is assumed to be owned by major companies, which are not allowed to use the percentage depletion al- lowance.  A 22 percent depletion allowance (typically reserved for natural gas sold under fixed contracts) was run as a sensitivity case. 11 The Comptroller estimates coal’s share of the percentage depletion allowance to be $29.7 million in 2006. See http://www.window.state.tx.us/specialrpt/energy/subsidies/index. php#coal.

OCR for page 52
72 Effects of U.S. Tax Policy on Greenhouse Gas Emissions MODELING RESULTS Natural Gas Production and Consumption The modeling finds that removing the percentage depletion preference re- sults in higher drilling costs, reducing incentives to explore and develop new supply. The reduced incentives to drill wells lead to a reduction in domestic nat- ural gas production of 14 Tcf over the 2010–2035 timespan in the Reference scenario, as shown in Figure 3-8. A very modest increase in natural gas imports of 0.4 Tcf is not sufficient to offset the drop in domestic production. Lower supply leads to an increase in gas prices, as shown in Figure 3-9. (The dashed black and green lines correspond to the Reference scenario and Cost-Depletion scenarios, respectively.) This, in turn, reduces natural gas con- sumption by 2 percent, on average. A sensitivity scenario was run where a cost depletion allowance of 22 per- cent was applied to the same resources that were otherwise receiving the benefit of the 15 percent allowance, further lowering exploration and production costs. While the impact is directionally consistent (higher allowance leads to lower gas prices, higher production, and consumption), the impact was minor, changing gas prices by 1.2 percent and production and consumption by less than 0.5 per- cent. FIGURE 3-8 Natural Gas Consumption Under Three Scenarios.

OCR for page 52
Energy-Related Tax Expenditures 73 FIGURE 3-9 Natural Gas Prices Under Several Scenarios. Electricity-Generation Mix The primary impact of removing the percentage depletion tax preference is to increase the cost of natural gas production and, hence, natural gas prices. All sectors reduce their natural gas consumption. However, the biggest impact oc- curs in the power sector because substitution of other fuels is easiest there. Depending on underlying market conditions, natural gas is replaced by coal, nuclear, and/or renewable generation. Of the renewable resources filling a portion of the generation gap created by lower levels of natural gas generation, biomass and wind contribute the greatest amount. Compared to the Reference scenario, wind generates 122 TWh more electricity, and biomass (co-firing) contributes an additional 63 TWh over the model time horizon. Compared to the Reference scenario, there are changes in the pattern of re- tirement of nuclear plants because gas prices are higher. End-use fossil fuel elec- tricity generation (primarily gas-fired combined heat and power) also declines. Sensitivity Analysis The impact of removing percentage depletion was examined for three al- ternative economic assumptions: high macroeconomic growth, high oil prices, and low natural gas prices. The High-Macroeconomic-Growth scenario results in higher energy con- sumption and energy prices, which magnify the impact of the tax provision. The High-Oil-Price scenario increases the impact of replacing percentage with cost

OCR for page 52
74 Effects of U.S. Tax Policy on Greenhouse Gas Emissions depletion. The Low-Natural-Gas-Prices scenario reduces the impact of the ex- cess of percentage over cost depletion. As intuition would suggest, the primary impact of the move to cost deple- tion from percentage depletion is to increase the cost of natural gas production and prices, with the High-Macroeconomic-Growth scenario showing the largest difference and the Low-Natural-Gas-Prices scenario showing the least differ- ence. All sectors reduce their natural gas consumption, with the biggest impact occurring in the electricity sector. Depending on underlying market conditions, the reduction in gas is replaced by coal, nuclear, and/or renewable generation. In the Reference and Low-Natural-Gas-Prices scenarios, there are some nuclear plants whose retirements are postponed in the Cost-Depletion scenario. In addition, end-use electricity generation from gas-fired combined heat and power plants also declines. CO2 Emissions The impact on CO2 emissions of removing the percentage depletion allow- ance is small under all four scenarios. In the Reference scenario, there is a net reduction in CO2 emissions of approximately 37 MMT over the model time horizon, summed across all sectors. This implies an average reduction of 1.5 MMT per year (see Table 3-2), or 0.03 percent of total CO2 emissions. Higher gas prices, as a result of cost depletion, discourage generation from natural gas, which is replaced primarily by more carbon-intensive coal generation and by renewable generation. There is a slight increase in renewable generation to meet the gap caused by the shift away from natural gas, but this is not sufficient to offset the greater use of coal except after 2030. Figure 3-10 shows these project- ed changes in CO2 emissions. In the other three scenarios there are small increases in CO 2 emissions when the percentage depletion allowance is removed. The Low-Natural-Gas- Prices scenario reduces the impact of the otherwise relatively higher natural gas prices. The High-Macroeconomic-Growth scenario leads to higher energy con- sumption and energy prices, which magnifies the fuel switching from natural gas to higher-carbon fuels and, hence, to slightly higher CO2 emissions. Lastly, the High-Oil-Price scenario displays the greatest increase in emissions: 0.2 percent over the 2010-2035 period. In this scenario, the shift in the power sector is most dominant due to the loss of the relatively less expensive natural gas. It should, however, be emphasized that the magnitude of changes in CO 2 emissions is in all these scenarios extremely small, especially in the Low-Natural-Gas-Prices and Reference scenarios.

OCR for page 52
Energy-Related Tax Expenditures 75 FIGURE 3-10 Changes in CO2 Emissions under the Cost Depletion Scenario. Summary The primary impact of the depletion allowance is its impact on the produc- tion of natural gas and the spillover impacts in other markets. To a first approx- imation, the depletion allowance produces no impact on greenhouse gas emis- sions. While natural gas production goes down when percentage depletion is removed, the complex substitution patterns lead to largely offsetting forces and to a minimal overall impact on CO2 and other GHG emissions. The four scenar- ios examined here have different signs in their impacts on CO 2 emissions, alt- hough each of them is tiny. We conclude that the sign of the change—whether it is positive or negative—is in reality uncertain but the size of the effect is likely to be very small. From a fiscal point of view, the oil depletion allowance was not motivated by concerns about climate change when it was enacted in 1926. From the point of view of climate change, this is not an effective subsidy for reducing emissions. CREDIT FOR ENERGY-EFFICIENCY IMPROVEMENTS TO EXISTING HOMES Legal Description Homeowners can benefit from two tax credits for adding energy- efficiency improvements to their homes. The Qualified Energy Efficiency Im- provements (IRC Section 25C) provides a 10 percent credit for the purchase of qualified energy-efficiency improvements to existing homes. The energy- efficiency home products must have been placed in service between January 1, 2011, and December 31, 2011. Under section 25C, the maximum credit for a taxpayer for all taxable years is $500, and no more than $200 of the credit may be attributable to expenditures on windows.

OCR for page 52
76 Effects of U.S. Tax Policy on Greenhouse Gas Emissions The Wind, Solar, Geothermal and Fuel Cell Tax Credit (IRC Section 25D) provides tax credits equal to 30 percent of the cost, with no cap through 2016, for construction of geothermal heat pumps, solar energy systems, solar water heaters, and small wind-energy systems and fuel cells.12 The energy-efficiency products must be placed in service before the end of 2016. The credits are valid only for improvements made to the taxpayer’s principal residence, except for qualified geothermal, solar, and wind property, which can be installed on any home used as a residence by the taxpayer. Short Description of Economic-Fiscal Impact The credit for energy-efficiency improvements to existing homes is meant to encourage the installation of energy-efficiency technologies in homes by de- creasing the costs of installation. The purpose is to lower energy use in the resi- dential sector, which should result in lower GHG emissions from energy con- sumption. The JCT estimated that these provisions would cause tax expenditures of $1.7 billion in fiscal year 2010 and expenditures of $2.9 billion from 2010– 2014. There is little solid empirical work on the impact of this provision on greenhouse gas emissions. The literature contains some theoretical and sugges- tive empirical evidence that points to positive impacts of government policies on energy-efficiency investment in homes. Theoretical models that consider various regulations, subsidies, or informational provision find that government interven- tion can help drive adoption of energy-saving technology. Both theoretical and empirical work suggests that market and behavioral failures (e.g., externalities, principal-agent issues, and informational barriers) can cause underinvestment in residential energy efficiency, and that government intervention can help. A recurring theme in the literature is that energy prices to end-use con- sumers are distorted and are likely to understate the true marginal cost of energy. This suggests energy should be priced more accurately to reflect its true cost, which may encourage energy-efficient investments. There is a literature that models empirically consumer purchases of ener- gy-consuming appliances and the associated demand for energy; however, it is difficult to apply the results of this literature to the provisions in sections 25C and 25D of the tax code. The Center for Business and Economic Research (CBER) study13 modeled estimates of the impact of these provisions and found that they led to reductions in CO2 emissions of 4.2 MMT in 2009. However, the methodology used to derive this estimate has not been sufficiently validated empirically to allow the committee to adopt this estimate. The committee be- 12 On fuel cells the credit cannot exceed $500 per 0.5 kW of installed capacity. 13 University of Nevada, Las Vegas, Center for Business and Economic Research (CBER) paper by Allaire and Brown, 2011.

OCR for page 52
Energy-Related Tax Expenditures 77 lieves that this is probably an upper bound on the impact, but the actual impact may be substantially smaller. In the absence of detailed and reliable results in the existing literature, the committee investigated the possibility of undertaking modeling along the lines of the provisions discussed elsewhere in this report. Such modeling would re- quire estimating the impact of credits on the prices to consumers of energy- consuming capital goods; then to calculate the impact of these price changes on investments made by homeowners; then to further calculate the impact of the changed energy-efficient investments on energy demand by fuel; and finally to calculate the impacts on greenhouse gas emissions. It became apparent that none of the existing models was well designed to do this analysis. In particular, the NEMS-NAS model has no policy levers that translate tax changes into prices and then into energy demands. For this reason, the committee was unable to provide estimates that it found reliable for the im- pact of this provision. Summary The committee did not find, and was unable within its time and resource budget to produce, detailed and reliable estimates of the impact of the Credit for Energy Efficiency Improvements to Existing Homes. In practical terms, the ef- fect of this section of the tax code on GHG emissions is likely to be limited. The reactions of households may be small because of (a) the volatility of the credits, which have varied in availability and in amount over time; (b) the relatively small credit limit (with the maximum being $500); and (c) the complexity of the provisions. Notwithstanding these reservations, existing research points to the poten- tial importance of incentives in this area. The combination of high potential payoff and limited research leads the committee to conclude that research on understanding the impacts of tax incentives on household energy consumption and GHG emissions should be encouraged. SPECIAL TAX RATE ON NUCLEAR DECOMMISSIONING RESERVE FUNDS Legal Description Nuclear power plant operators can elect to set aside reserve funds for the decommissioning of plants. The code provides for special tax treatment of these funds in two ways. Contributions are deductible in the year they are made and not taxed until distribution. This defers tax on those funds into the future. Once distributed, the funds, along with any gains from investments through the years, are taxed at a rate of 20 percent. Most utilities operating nuclear plants are large enough that their income would normally be taxed at 35 percent. Thus, a plant

OCR for page 52
78 Effects of U.S. Tax Policy on Greenhouse Gas Emissions operator benefits from a deferred, lowered tax on the income made from the reserved funds. Short Description of Economic-Fiscal Impact When a nuclear power plant is retired from service (decommissioned), the residual radioactivity at the facility must be reduced to a level that allows trans- fer of the property. The U.S. Nuclear Regulatory Commission (USNRC) has rules governing nuclear power plant decommissioning. These involve cleanup of radioactively contaminated plant systems and structures and removal of the ra- dioactive fuel. Before a nuclear power plant begins operations, the operator must ensure that there will be sufficient funds to cover the ultimate decommissioning of the facility. Each plant operator must report biannually to the USNRC the status of its decommissioning funds for each unit. 14 According to the USNRC, the estimated decommissioning cost for a nu- clear reactor can range from $300 to over $600 million, or between 10 and 25 percent of construction costs.15 The total cost of decommissioning a nuclear re- actor depends on the timing and sequence of the various stages of the construc- tion program, the reactor type, the location of the facility, the radioactive-waste burial costs, and the ultimate plans for spent-fuel storage. Realized decommis- sioning costs have reached over $1 billion at some plants. Given the size of decommissioning costs, a utility responsible for decom- missioning a nuclear power plant can create a reserve fund to pay for these costs. About 70 percent of current operators are authorized to accumulate de- commissioning funds over their plants’ operational lifetimes. Analysis and Summary The committee was unable to find any detailed and reliable estimates of the impact of the nuclear decommissioning tax preference on greenhouse gas emissions.16 This is a particularly difficult provision to analyze for several rea- sons. First, nuclear power plants have a very long useful lifespans (the lifetime of the plant could well be at least 60 years), and that time span extends beyond 14 Currently, the U.S. power sector operates 104 commercial nuclear power plants. Most were built in the 1970s and are scheduled for decommissioning during the next three decades. As of April 2011, there were 23 nuclear units in various stages of decom- missioning, with 10 of those completely cleaned up. 15 For more information, see http://www.nrc.gov/reading-rm/doc-collections/fact-sheets/ decommissioning.html. 16 The CBER model did calculations of the impact of removing the nuclear decommis- sioning tax credit. However, it was in a supply-and-demand framework that was static and did not have the same detailed treatment of taxes, costs, and load curves as the NEMS-NAS model. The committee therefore finds that the CBER results are an upper bound of potential results, with the more likely lower bound being zero.

OCR for page 52
Energy-Related Tax Expenditures 79 the time horizon of any of the models used for this study. Second, the provision involves complex computations of the opportunity cost of funds, the return on capital, as well as the future regulatory treatment of these costs under federal and state public utility laws and regulations. Third, the runs that were undertak- en for the committee in the NEMS-NAS model actually project that there will be no new nuclear power plants licensed in the time period of the runs. Given that there are no new plants in the base run, this number clearly cannot be reduced by removing a tax preference. Therefore, to the extent that these results are consid- ered reliable, the best estimate of the impact on nuclear power plant construction would be zero. There might, however, be an effect on extending the lifespan of existing plants. While the committee was unable to find or commission detailed and relia- ble modeling runs on the nuclear decommissioning tax preference, it finds that the most likely impact is a negligible impact on greenhouse gas emissions. SUMMARY The present chapter examines some of the important energy-related tax expenditures. Two of them have been examined in detail using the NEMS-NAS model, while the other two were examined qualitatively. Several other provi- sions were not modeled, but the current four comprise more than half of all en- ergy-related tax expenditures. One important result is that the net estimated impact of the modeled provi- sions is very small. The central estimate of the net impact of the renewable tax credits, the depletion allowance, and the nuclear decommissioning credit is about 0.1 percent of total national GHG emissions over the next quarter-century. While the central estimate is close to zero, alternative assumptions about GDP growth, natural gas costs, and oil prices make the range larger and include both positive and negative numbers. We estimate that the net total is likely to be in the range of plus or minus ½ percent of total U.S. greenhouse gas emissions over this period. A second result is that the web of interacting impacts is extremely com- plex and often leads to counterintuitive results. For example, the renewable elec- tricity tax credit not only lowers costs but also apparently decreases total elec- tricity output because of changes in the composition of generation. The depletion allowance, which is usually associated with oil, has its major effect on gas. The nuclear decommissioning credit is likely to have no effect at all be- cause the number of nuclear power plants triggered by its existence is likely to be zero given that the number of new plants in most projections is already zero. Moreover, these estimates are undertaken in a partial equilibrium framework, and, as we will see in subsequent chapters, including the reaction of other sec- tors in a general equilibrium model makes the reactions even more complex and difficult to estimate accurately. Third, the committee found that some of the provisions (such as the credit for energy-efficiency improvements to existing homes) are so complex, or in-

OCR for page 52
80 Effects of U.S. Tax Policy on Greenhouse Gas Emissions volve so many unanswered questions, that it is not possible to provide reliable estimates of their impacts on greenhouse gas emissions. For both the impact of tax credits for energy efficiency in homes and the nuclear decommissioning tax preference, the uncertainties about future tax, regulatory, financial, and behav- ioral responses are so large that the committee was unable to provide what it regarded as reliable estimates. For other provisions, the estimates cannot resolve whether the net impact is negative or positive. Finally, the energy-sector tax preferences are a good example of the rea- sons that tax specialists hold tax expenditures in low regard. Many of them were introduced in an earlier era and have outlived their original purposes. Others were introduced in order to foster important national goals, but the complexity of the energy-sector interactions actually leads to perverse impacts that are con- trary to the original purposes.