The Capitalization of Energy Efficiency in the Residential Housing Market
Energy & Cities Conference
Boston, October 22 2012
Nils KokMaastricht University
Matthew E. KahnUCLA
Energy consumption and the built environmentSubstantial environmental externalities
1960196219641966196819701972197419761978198019821984198619881990199219941996199820002002200420062008
0%
10%
20%
30%
40%
50%
60%
70%
80%
38%
40%
42%
44%
46%
48%
50%
52%
Energy Consumption in Buildings (LHS) Of Which Commercial (RHS)
Total energy expenditures $166bln in 2010About 50% goes to space conditioning
How to reduce energy consumption in the residential housing market?
0. Raise energy prices
1. Stricter building codes and subsidizing retrofits– Works, but mostly for new construction, and effects are small
Building codes are effective at saving energy (Jacobsen and Kotchen, in press)– Fiscal tight-belting constrains subsidies
2. Stimulating market efficiency through energy labels– Energy efficiency gap (Jaffe and Stavins, 1994)– Investments in energy efficiency may lead to:
• Save on current resources, insure against future price increases• Higher transaction prices
3. Change occupant behavior, for example through nudges (Alcott, 2011)
Related literatureHow energy literate are private consumers?
• Current policies to reduce energy consumption assume rational decision-making by informed investors
• That seems to hold for sophisticated investors in commercial property…– Labels have financial implications (Eichholtz et al., 2010, Fuerst and
McAllister, 2011, etc.)– Efficient capitalization of energy bill (Eichholtz et al., 2012)
…but not necessarily for private consumers– Residential “energy literacy” is low (Brounen et al., 2012)
• Solar is capitalized into home prices (Dastrup et al., 2012)
• Labeling programs in Europe seem to have the desired effect– Mandatory disclosure of EU energy label– Price differentiation based on energy label (Brounen and Kok, 2011)
The (in)famous EU energy label
This paperWhat is impact of “green” labels on the housing market?
• Recent growth in labels attesting to energy efficiency and other “green” attributes in housing market
• What are the market implications of “green” certification for residential dwellings?
(1)
– Rijt is the home’s sales price commanded by dwelling i in cluster j in quarter t
– Xi is a vector of hedonic characteristics
Size, vintage (since renovation), swimming pool, etc.– Zip code fixed effects to control for cross-area differences in local public
goods, interacted with year/month indicators to capture price dynamics– Standard errors clustered at the zip code level
Model specification (II)Model expanded with interaction terms
• Is the willingness to pay affected by climate, energy prices? But also: role of ideology and competition?
• Recover heterogeneous effects of green home labels:
(2)
– N is an interaction term that reflects: Local climatic conditions Local electricity prices Consumer ideology Green density
• Caveat– Green homes are mostly production homes, not high-end custom homes,
but…we have no information about the developer – possibility of bundling valuable amenities with green attributes (appliances, etc.)
“Green” homesEnergy Star (EPA), LEED (USGBC), and GreenPoint Rated
• Green labels for homes: reflection of steady state efficiency– EPAs Energy Star for Homes (1995)
Asset rating (i.e., does not account for actual performance) For new construction only Changed in 2006 and 2012 Certified by professional engineer
– USGBCs LEED for Homes (2005) Scoring systems based on 6 components of “sustainability” Energy efficiency is just one component Based on design stage (and now verified after construction)
– GreenPoint Rated Comparable to LEED for Homes Primarily marketed in California Also for existing homes
• Diffusion of green home labels substantially lags the commercial sector
Random example of a LEED certified home
DataDataQuick’s transaction database
• DataQuick database of single-family home transactions in California (2007 – 2012)– 4,231 green homes transacted (out of 10,348 green homes)– 1.6mln control homes
• Standard hedonics (vintage, size, swimming pool, view, …)– Flag for “distressed” sale (foreclosure or mortgage delinquency)
• Environmental ideology proxied by Prius registration share at zip code level (R.L. Polk)– Model (2) includes country-fixed effects, Census variables and distance to
CBD and closest train station
• Local climatic conditions measured by annual cooling degree days during year of sale at nearest weather station (NOAA)
• Consumer electricity prices measured at zip code level
The geography of green homesClustering in areas with significant construction activity
Descriptive statisticsSignificant fraction of “distressed” sales
Results High explanatory power of models – significant green premium
• Green homes sell at a premium of 12%, ceteris paribus– Commercial office buildings: 16% (Eichholtz et al., 2010)– Label “A” homes in Europe: 10% (Brounen and Kok, 2011)
ResultsDecomposition of “green” labels
• LEED and GreenPoint Rated insignificant– Small sample bias (Eichholtz et al., 2010)– Limited brand recognition
More robustness checksRecently constructed homes
• Analysis restricted to dwellings constructed < 5 years
• Green homes sell at a premium of 8.7%, ceteris paribus– Coefficient on LEED comparable, but insignificant
Heterogeneity in capitalization of green labelsWeather and ideology matter, price and competition do not
• Distinguish effects of energy-savings aspect of rating from other, intangible effects of label itself
Discussion of resultsThe costs and benefits of green homes
Ceteris paribus, green homes have higher selling prices by 9%
The average non-green home in the sample would be worth $34,800 more if it were converted to green
What about relative input costs? Anecdotal evidence shows cost is $10,000 higher (at most), to construct
a dwelling that is 35 percent more efficient than code
What about the value of energy savings? 30 percent savings on a typical $200/month energy bill translate in a
simple payback period of 48 years for the green increment
Other features seem to add value Unobservables – savings on resources other than energy, but also:
advanced ventilation systems, higher comfort, better IEQ Some homeowners attribute non-financial utility to a green label
(comparable to heterogeneity in solar premium)
Conclusions and implicationsEvidence comparable to Europe and US commercial market
Durable building stock creates significant environmental externality
Certification can help market transformation through transparency Private consumers are generally “energy illiterate”
Homes in California that have a green label transact at a premium Premium seems to outweigh costs Developers will reap the opportunity and drive down the marginal effect
Labeling programs by government and non-profit institutes seem to be effective and are incorporated by market participants Cost of the labeling programs is small Does this directly affect energy consumption (and emissions)? More aggressive policies through mandatory disclosure (e.g., Seattle)? Standards versus labels
Limitations and next steps
We cannot disentangle (modeled) energy savings required to obtain label from unobserved effects of label itself – are labels necessary? Incorporate information on thermal efficiency of non-rated homes Incorporate realized energy consumption in green homes and
conventional homes
Repeat sales would allow for further studying the value persistence, and filtering out the effect of developer quality
Focus on owner-occupied single-family dwellings only – the rental housing market offers an interesting experiment on the value of “green” labels (with landlords often facing principal-agent problem)
Green home diffusion – how do owners sort into green homes? Co-ordination device?