Transcript An Evaluation of Tax Credits for Residential

An Evaluation of Tax Credits for
Residential Energy Efficiency
Presented by Molly Sherlock
(paper co-authored with Andre R. Neveu)
Washington and Lee University
January 23, 2015
Forthcoming in Eastern Economic Journal
Policy Context
• “Energy Paradox”
• Why don’t consumers adopt cost-saving energy technologies?
• In 2005, Congress took action to improve energy
efficiency for residential heating and cooling
 Goal of this paper
 Evaluate tax incentives for residential energy efficiency
 Big picture question
 Are tax incentives the best policy lever for achieving stated goals?
2
Relevant Legislation
• Energy Policy Act of 2005
• Nonbusiness energy property credit (Internal Revenue Code (IRC)
§ 25C)
• 10% credit for building envelope components (e.g., windows, roofs,
insulation) and certain heating and cooling equipment
• Limited to $500 in 2006 and 2007
• Residential energy-efficient property credit (IRC § 25D)
• 30% credit for solar panels, geothermal, other on-site generation
• American Recovery and Reinvestment Act of 2009
• Expanded § 25C for 2009 and 2010 – 30% credit up to $1,500
• Removed property-specific caps
• Section 25C credit now part of the “tax extenders”
3
Claims of Residential Energy Credits,
2006 - 2011
4
Evaluating Tax Policy
• Standard Economic Framework
• Equity
• Efficiency
• Simplicity / Tax Administration
5
Tax Data
• Internal Revenue Service (IRS), Statistics of Income
(SOI), Public Use File (PUF)
• Annual sample of federal tax returns
• 2006 SOI PUF
• 145,858 records representing 138.4 million individual tax returns
• Using the 2006, 2007, and 2008 SOI PUF files
• Isolate 2006 and 2007 tax year returns; clean data
• 279,536 observations representing 276.0 million individual tax
returns
6
Distribution of Residential Energy Credit Claims
2006 – 2007
7
Higher-Income Taxpayers Disproportionately
Claim Residential Energy Credits
8
Policy Questions
• Are lower-income taxpayers receiving less in tax
credits for similar levels of energy-efficiency
spending?
• Credit is nonrefundable; cannot be carried forward
• Do other tax expenditures “crowd-out” credits for
energy efficiency?
9
Determining Residential Energy Credit Amount
Tax Liability
(before credits)
• Report tax liability on IRS Form 5695
Subtract “Higher
Ranking” Credits
• Child and dependent care credit;
credit for elderly & disabled;
retirement savings credit; education
credits; foreign tax credit
Remaining Tax
Liability
• Maximum
residential energy
credit amount
10
Some Observations
• Lower income-taxpayers are
more likely to “zero out” tax
liability before residential
energy credits can be claimed
• Taxpayers that “zero out” tax
liability with energy credits
claim lower credit amount, on
average
• Changing where the energy
credit appears on IRS Form
1040 could affect the credit’s
cost
Share of Taxpayers that
"Zero Out" Tax Liability
before Residential Energy
Credits
6%
5.2%
5%
4%
3%
2%
1%
0.3%
0%
< $50,000
> $50,000
11
Determining Factors Associated with Tax Credit Claims:
Related Literature
• Consumer energy tax credits
• Residential energy efficiency
• Metcalf and Hassett (1995); Dubin and Henson (1988)
• Hybrid vehicles
• Diamond (2008, 2009)
• Adoption of “green” technologies
• Dastrup et al. (2012); Kahn and Vaughn (2009); Sexton and
Sexton (2011)
12
Factors Related to Tax Credit Claims
Mean Values for Full and Restricted Samples
Full Sample
Energy Credit Claimed
Energy Credit Value
Income
Itemizing Deductions
Itemized with Real Estate or
Interest Deduction
Interest Deduction
Married
Head of Household
Married Filing Separately
Social Security
Child Tax Credit
Average Monthly Electricity Bill
Republican State
Average January Temp
Average July Temp
Hybrid State Ranking
% BA Degree
3.1%
$220
$71,164
35.9%
32.6%
29.6%
38.2%
15.0%
1.8%
15.5%
18.4%
Sample w/ State
Identifiers
3.0%
$217
$51,734
34.1%
30.7%
28.0%
36.6%
15.5%
1.8%
15.4%
19.1%
$88.33
52.0%
33.9
74.8
28.4
27.9%
13
Motivations for Empirical Method
• Evaluate tax credit claims along both the extensive
margin (i.e., whether to claim) and intensive margin (i.e.,
how much to claim)
• Sample selection problem: credit amount is observed only
for those who are able to claim the credit
14
Empirical Method
• Two-step model
• Selection equation models probability of claiming credit
• Outcome equation looks at factors explaining the amount claimed
• Identifying the selection effect – exclusion restriction
• Homeownership is likely associated with claiming a credit or not; is not
likely to determine the amount of the credit that is claimed
• Two-step “Heckman correction” model to evaluate tax
policy
• Eissa and Hoynes (2004); Eissa et al. (2008); Newsome et al.
(2001)
15
Empirical Method
• First step: probit specification
• Variables in x
• Income, Tax Liability, and Non-Energy Credits
• Dummy variables
•
•
•
•
Likely homeowner
Non-single filing status
Social security recipient; child tax credit
Year 2007
• State-level variables
•
•
•
•
Average monthly bill
January / July average temperature
Voted Republican; % BA degree
Hybrid car ranking
• Regression coefficients estimated using maximum likelihood
16
Interpreting Results – Marginal Effects
• Average marginal effects (AME)
• Marginal effects at the mean (MEM) calculated but not
reported
17
Selection Equation: Probit Estimates & Marginal Effects
18
Results
• The probability of claiming residential energy credits
• Increases with…
• ln(tax liability)
• non-single filing status; having social security income; claiming the child
tax credit
• homeownership
• Colder winters increase the probability of claiming energy tax
credits; hotter summers don’t matter
19
Outcome Equation
20
Results
• The amount claimed in energy credits
• Increases with…
• Income and tax liability
• State’s average monthly electricity bill
• Are taxpayers near DC claiming more?
• Compared to California, taxpayers in DC, MD, and DE claim
significantly more in tax credits ($ value)
21
Our Findings and Contributions
• The tax credit, as designed, is vertically inequitable
• Climate and electricity costs matter
• Taxpayers in colder states are more likely to claim the credit
• Taxpayers in states with higher electricity costs claim higher credit
amounts
• A first look at current tax credits for residential energy
efficiency
• Novel application of the sample selection model
22
Shortcomings & Policy Implications
• Study doesn’t say whether tax credits cause additional
investment
• Don’t have data on claims of credits for specific types of property
• Given vertical equity concerns, policy options to consider
• Make credits refundable
• Allow credits to be carried forward
23
The Broader Context
• Why subsidize residential energy efficiency?
• Do consumers “underinvest” in efficiency?
• Is there a market failure?
• Other non-tax policy options
• Consumer labels
• Rebates
• Efficiency standards
24
Questions?