1. The Diffusion of Energy Efficiency in Building
Nils Kok Marquise McGraw John M. Quigley
Maastricht University UC Berkeley UC Berkeley
AEA Meetings, Denver
January 2011
3. The “energy paradox,” revisited
Increasing number of buildings certified as efficient
Energy consumption and building technology are closely related
30 percent, 40 percent, 70 percent, …
Durability of real capital: existing structures continue to have impact
Slow diffusion of more efficient technology
Measures are profitable: CFLs, HVACs, …
“Energy paradox” (Jaffe and Stavins, 1994)
High discount rates (Hausman, 1979)
Lower returns? (Metcalf and Hassett, 1999)
Substantial increases in commercial buildings labeled as energy-
efficient or “green”
4. Energy-efficiency labels and property markets
Energy Star (EPA) and LEED (USGBC)
EPA’s Energy Star for Commercial Buildings (1995)
Efficiency in energy use in (top quarter relative to CBECS)
Standardized for building use (occupancy, hours) and climate
Certified by professional engineer
Based on real energy consumption (at least one year of bills)
USGBC’s Leadership in Energy and Environmental Design (1999)
Scoring systems based on 6 components of “sustainability”
Energy efficiency is just one component
Various systems and versions (e.g., NC, EB, O&M, ...)
Based on design stage (and now verified after construction)
5. Program growth: Energy Star and LEED
48 MSAs, 1995 – 2010
Dominant forces in the commercial and institutional market
2010: 2010:
10 percent of buildings 5 percent of buildings
30 percent of stock 10 percent of stock
Size effect (Snyder, et al., 2003) Registered: 27,000 buildings
(6b sq.ft.)
6. Labels reflect building technology
Energy paradox in commercial building?
Labels verify hard-to-observe energy efficiency technology
Comparable to role of patents in production technology (Keller, 2004)
Certified buildings have lower resource consumption
Energy Star: 35 percent less energy consumption, on average
LEED: efficiency of new construction unclear, existing certified buildings
on par with Energy Star requirements.
Are measures profitable?
Investments costs include: consultancy services, incremental cost of
construction, design, equipment and materials
Evidence on returns to investments
Increased rents and asset values (Fuerst and MacAllister, 2011)
Capitalization of incremental energy savings into asset values
(Eichholtz, et al., 2010)
Building technology (i.e., labels) should diffuse quickly across markets
7. Diffusion of certified space
Substantial differences in timing and growth across MSAs
New York New York
Los Angeles Los Angeles
8. The diffusion of energy efficiency in building (I)
Determinants of timing and growth
“What determines the geographic dispersion in the timing and
growth of energy efficient technology in office buildings?”
1. Variations across markets in expected cost savings
Climatic conditions (Degree days; NWS)
Adverse climatic conditions increase expected economic payoff
Energy prices (Cents/kWh; utility data EIA)
Higher prices increase expected payoff from improvements
Lower energy consumption in more expensive areas
Property market conditions (Stock, vacancy, rents, prices; CBRE-EA)
New construction depends on stage of property cycle
Green “premium” varies with market conditions
9. Variations in the expected cost savings
Simple correlations, 2010 cross-section
10. The diffusion of energy efficiency in building (II)
Determinants of timing and growth
2. Local economic conditions that affect appropriability of gains
Income (Average wages and salaries; BEA)
Ancillary benefits of “green” buildings
“Green” as a luxury good (Roe, et al., 2001); “warm glow”
Size of service sector (Fraction of people employed in service sector; BLS)
Demand for office space
Size of government (Fraction of people employed by government; BLS)
“Green” procurement policies
Building professionals (LEED APs, architecture grads; GBCI, NAAB)
Overcome information barriers (Hall, 2003)
3. Building-specific characteristics that influence expected profitability
Building size (Average building size, CBRE-EA)
Spread fixed costs over larger base (Snyder, et al., 2003)
12. The diffusion of energy efficiency in building (III)
Determinants of timing and growth
4. Institutional characteristics
Political ideology (Vote for Reagan ‘84, Bush ’92; CQ Press)
Political ideology may influence tenant and investor choices
Regulation and incentives (LEED public policies; USGBC)
Government policies may stimulate innovations (Lanjouw and Mody,
1996; Jaffe and Palmer, 1997)
Some cities have included LEED in building codes for new
construction and renovations
Numerous LEED-specific incentives: “fast-tracking” permits,
subsidies, tax credits
14. Dynamic models
Levels, first differences and Arellano-Bond
We model the dynamic relationship between the diffusion of energy
efficiency over time and across geographic markets as:
(1) Fraction it = α + βX it−2 + εit
Where X it−2 is a vector of income, prices and economic conditions
We use a two-year lag to account for real time necessary to complete
€ renovations and new construction
Serial correlation addressed by estimating AR(1) using FGLS
€
(2) ΔFractionit = α + βΔX it −2 + ε it
First differences to control for time-invariant unobserved heterogeneity
Alternatively, we estimate (2) following Arellano-Bond (1991),
€ instrumenting all covariates by their own lagged values
17. Conclusions and implications
Economic conditions important for energy efficiency diffusion
Built environment important in reducing resource consumption
Much attention to the “energy paradox” in building sector
Diffusion of energy efficiency and “green” technologies in commercial
property sector widespread and rapid
30 percent of all commercial office space certified by Energy Star
11 percent of all commercial office space certified by LEED
Considerable variation in adoption of energy efficiency technologies
Diffusion has been more rapid in areas with higher incomes and sound
property market fundamentals (low vacancy rates, high rents and prices)
This has important implications for underperforming markets (e.g.,
Dallas, Detroit, and Tampa); these markets will lag behind in energy
efficiency improvements
18. Conclusions and implications (II)
Energy paradox less important for commercial buildings
Technology seems to diffuse faster in larger properties
Improving energy efficiency of smaller buildings may need additional
“nudge”
Diffusion of energy efficient technology more responsive to energy
prices than “green” technology
Lends additional support for efficiency of investment decisions in
commercial property sector (as opposed to residential sector)
Diffusion of “green” technology is facilitated by human capital (i.e.,
LEED APs) and governmental policies
The environmental implications of this innovation remains unclear