GPPI - Thesis - Annotated Bib...

Beck, Nathaniel and Jonathan Katz. 1995. "What to Do (and Not to Do) with Time-Series Cross-Section Data." The American Political Science Review 89 (3): 634-647.

Describes challenges of using OLS with TSCS data
Trashes Parks GLS method of analyzing TSCS data for overestimating confidence in regression coefficient estimates (i.e. se's are too small)
NB - analysis limited to TSCS data where T>N (Parks only works for these datasets)
Endorses panel-corrected standard errors based on Monte Carlo experiments

Bernstein, Mark et al. 2003. State-Level Changes in Energy Intensity and Their National Implications. Santa Monica, CA: RAND.

Examines changes in energy intensity by states in the previous two decades
Attributed changes to measurable factors and identified portion potentially due to unobserved factors--including state policies
Looks at residential, commercial, industrial, and transportation sectors from 1977-1999
Factors examined to explain changes in energy intensity include: energy prices, composition of state output, capacity utilization, capital investment and new construction, population and demographics, climate, technological innovation, energy policies
RAND regression model includes state fixed effects and an aggregate time trend
Energy intensity defined as energy use per dollar of GSP
Theoretical models use log of energy intensity as the dependent variable and include both state and year fixed effects
Study addresses sectors separately and in aggregate
Considered the periods 1977-1988 and 1988-1999 separately because of the much greater energy price stability seen in the latter
In the residential sector, the authors found the effects of the following variables to be statistically significant (p<0.05): heating-degree days; cooling-degree days; electricity prices; natural gas prices; disposable income per capita; employment per capita; household size
In the commercial sector, the authors found the effects of the following variables to be statistically significant (p<0.05): heating-degree days, cooling-degree days; energy prices; employment per GSP from Services; Commercial GSP from Retail Trade; Commercial GSP from F.I.R.E. and Legal; Commercial GSP from Health
In the industrial sector, the authors found the effects of the following variables to be statistically significant (p<0.05): energy prices; percent of industrial GSP from Petroleum, Paper, and Metallurgy; Percent of Industrial GASP from Food, Textile, and Lumber; Percent of Industrial GSP from Mining; Percent of Industrial GSP from Agriculture; Deviation from Equilibrium GSP

Bernstein, Mark et al. 2002. "The Public Benefit of Energy Efficiency to the State of Washington." Santa Monica, CA: RAND.

Report addresses benefits to state of Washington from energy efficiency focusing on:

effects on gross state product of energy efficiency improvements in the commercial and industrial sectors
effects on air emissions
effects on households, especially low-income households

ceteris paribus, authors find that improvements in energy efficiency are associated with increases in GSP
energy efficiency improvements may offer more significant benefits to low-income households since energy expenditures are a higher fraction of their income and since such households tend to be the least energy efficient to begin with

Blumstein, Carl, Charles Goldman and Galen Barbose. 2005. "Who Should Administer Energy-Efficiency Programs?" Energy Policy 33 (8): 1053-1067.

Examines choices in several states regarding what type of entity to entrust with administering energy efficiency programs and what scope of responsibility to give that entity. Discusses Pacific Northwest, CA, NY, VT, and CT.

Boyd, Gale and Joseph Roop. 2004. "A Note on the Fisher Ideal Index Decomposition for Structural Change in Energy Intensity." The Energy Journal 25 (1): 87-101.

Authors propose the chain weighted Fisher Ideal Index as a formula for decomposing aggregate trends in energy intensity into its component changes in economic activity and energy efficiency
it does not matter whether the underlying measures of activity can be summed to equal an aggregate measure of activity (e.g. GSP components vs. floor space or passenger miles). Rather, what matters is that intensity, which is analogous to prices in the index number theory, is well defined, i.e. is a "good" measure of the sector level energy intensity
Authors apply their Fischer Ideal index approach to the manufacturing sector and found that for 1983-1998, the rate of aggregate electricity intensity change was -0.3%, about half of which was sectoral shift and half was real intensity
Authors find that for electricity there was little net change in the eighties, but both structure and intensity factors lead to a decline in electricity intensity in the nineties. The authors suggest that is might indicate a reversal of a historical shift toward electrification [Note: a shift away from electricity would lead to a lower electricity intensity/efficiency without actually requiring improvements in energy intensity/efficiency]

Brennan, Timothy J. 1998. "Demand-Side Management Programs Under Retail Electricity Competition." Resources for the Future (RFF) Discussion Paper 99-02.

Brennan explains why deregulation and the introduction of competition to electricity markets threatens the viability of DSM programs but also weakens the rationale for having DSM at all
Reviews arguments for DSM programs: consumers cannot tell how much energy they will save from certain investments and agency problems (e.g. landlords/tenants, home builders/home buyers)
Brennan supposes that competitive electricity markets might lead to creative marketing by firms who will bundle DSM with electricity provision
Competition will not eliminate the discrepancy between market prices and true marginal costs of electricity. As such, DSM could still be warranted as a means of getting consumers to choose the level of energy efficiency that they would have selected had they faced the true social marginal cost of electricity
Brennan points out that as a "second-best" tool compared to energy taxes, DSM leads to an overconsumption of the services produced with electricity

Bressand, Florian et al. 2007. "Wasted Energy: How the US Can Reach Its Energy Productivity Potential." McKinsey Global Institute.

McK estimated the reductions in energy intensity possible with existing technologies and energy-efficiency investments with IRRs of 10% or more
US has lowest energy productivity of any developed economy
US residential energy demand is the largest single energy end-use sector world-wide
Authors estimate potential to reduce energy demand by 21% compared to base case by 2020
Largest opportunities in residential sector (heating and cooling, CFLs)
Interesting chart on high hurdle rates reportedly required for commercial sector energy efficiency technology adoption (p. 16)
NB - important potential independent vars for thesis regression:

some states have recently adopted energy efficiency resource standards (EERS)

some states allow ESCOs to aggregate and bid on demand reduction opportunities as alternatives to new power generation

Chamberlin, John and Patricia Herman. 1996. "How Much DSM Is Really There? A Market Perspective." Energy Policy 24 (4): 323-330.

Authors consider the future of DSM in a deregulated electricity market and conclude that DSM spending will shrink but that there will be a significant place for DSM in competitive retail markets
Reviews developments in energy services market

DeCanio, Stephen J. 1998. "The Efficiency Paradox: Bureaucratic and Organizational Barriers to Profitable Energy-Saving Investments." Energy Policy. 26 (5): 441-454.

Author investigates the "energy paradox"--the observation that firms do not undertake energy-efficiency investments that appear to have positive net present values
DeCanio performs regression analyses using EPA data from its Green Lights program--a voluntary effort by firms to report on savings from efficient lighting investments
Finds that paybacks and IRRs are neither uniform across firms nor close to the risk-adjusted discount rate that would constitute the appropriate "hurdle rate" for investments under conventional investment analysis
Data suggest that Green Lights projects are far more profitable than any plausible risk-adjusted cost of capital for comparable projects
DeCanio concludes from his regression results that transaction costs cannot explain "energy paradox"

Eldridge, Maggie et al. 2007. "The State Energy Efficiency Scorecard for 2006." American Council for an Energy-Efficient Economy. Report Number E075.

Authors rate states on each of eight energy efficiency policy areas:

spending on utility and PBF energy efficiency programs
energy efficiency resource standards (EERS)
combined heat and power (CHP)
building energy codes
transportation policy
appliance and equipment efficiency standards
tax incentives
state lead by examples and R&D

authors claim that data on energy efficiency spending rather than savings have been more accurately tracked now that more programs are run by non-utilities
utility spending on energy efficiency declined rapidly as states restructured their electricity markets since utilities were worried about cost recovery
ACEEE claims that, after reaching a low point in the late 1990s, energy efficiency spending increased as states realized that restructured markets did not contribute to energy efficiency and under-investment in infrastructure had led to reliability concerns (e.g. 2003 Northeast blackout)
in the 1980s and 1990s, utilities ran almost all energy efficiency programs. Following restructuring, many states enacted PBFs and created new organizations or contracted third parties to administer energy efficiency programs. In many states with PBFs, however, the funding is still directed to utilities to administer energy efficiency programs
Note on the energy efficiency spending data used for the ACEEE report: looks like they used EIA-861 plus supplemental data from individual states as needed where those states had non-utility energy efficiency programs
spot check of ACEEE data showed that their state ee spending does not match my first-cut at the EIA data (off by 100% in some cases)
Twelve states have EERS or similar policies: TX, HI, NV, CT, CA, VT, CO, PA, WA, MN, IL, NJ. Some of these states (HI, PA, NV) have Sustainable Energy Portfolio Standards that allow for both renewable and energy efficiency

Report table details EERS

TX created EERS in 1999 that took effect in 2003
VT had energy efficiency mandates for Efficiency Vermont starting in 2000
CA goals started in 2004
CT legislation in 2005 --> too late for my study
CO standard takes effect in 2006
NV RPS amended in 2005 to include energy efficiency
WA 2006 ballot initiative
IL not yet implemented
MN in 2006
PA ? unclear when it started
NJ program still under development

Building energy codes

As of 2006, 39 states and DC had mandatory building energy codes
codes are typically adopted at the state level and enforced at the local level by county or municipal officials
CA attributes 25% of its electricity savings to its Title 24 building standard

Appliance and equipment efficiency standards

Rationales

principal-agent problem

"split-incentive"
"panic purchases" - immediately needed replacements purchased by plumbers (for example)

Information-cost barriers

DOE has never used its authority to add new products to the federal efficiency standard list, so states have had to take the initiative
Between 2002 and 2005, 11 states adopted product energy efficiency standards; although, many of these standards have been pre-empted by the Energy Policy Act (EPAct) of 2005

State tax incentives

Tax incentives include: direct income tax credits, reduced sales tax on qualified products, and income tax deductions
Tax incentives first offered at the state and federal level in the 1970s
Rationales for tax incentives:

first cost issues
risk aversion
low visibility in the market (due to low market share)
Low importance of energy expenditures for many consumers and firms

Five states (MD, NV, NY, MT, OR) offer tax incentives for green buildings
Three states (AZ, LA, OK) have tax incentives for energy-efficient new homes
Four states (CA, ID, MT, NY) offer tax incentives for home weatherization
Four states (CA, CT, MA, OR,) offer tax incentives for energy-efficiency equipment purchases

Eto, Joseph. 1996. "The Past, Present, and Future of US Utility Demand-Side Management Programs." Enrest Orlando Lawrence Berkeley National Laboratory (LBNL). http://eetd.lbl.gov/EA/EMP/reports/39931.pdf

Eto defines 7 types of DSM program: (1) general information; (2) technical information, including energy audits; (3) loans and direct payments; (4) direct or free installation of energy efficiency improvements; (5) third-party performance contracting; (6) load control and shifting; (7) innovative tariffs (e.g. interruptible, time-of-use and real-time pricing)
Author provides a history of US DSM programs
Eto explains three types of regulation used to remove utilities' disincentive to do DSM: (1) percentage adder on DSM expenditures; (2) bonus payments for energy or capacity saved; (3) percentage of net resource value of DSM program

Eto, Joseph et al. 2000. "Where Did the Money Go? The Cost and Performance of the Largest Commercial Sector DSM Programs." The Energy Journal. 21 (2): 23-49.

Authors examined 40 commercial sector DSM programs in 1992
Total resource cost of saved energy includes:

Measure costs (cost of purchasing and installing energy-efficient product, borne by either utility, customer, or both)
Non-measure costs (costs associated with operating a program--e.g. advertising, evaluation, administrative overhead--borne by utility)
Eto et al include shareholder incentives (i.e. bonuses, rate-return adjustments, shared-savings, combinations of these) as a non-measure cost

Authors found inconsistent reporting of consumer-borne measure costs by utilities
Excluding shareholder incentives decreases the simple average for total resource cost of saved energy by about 7%
Authors review methodologies employed by utilities to calculate energy savings (tracking database methods, billing analyses, end-use metering)
In contrast to free-rider effect, authors discuss participant spillover (energy savings actions taken by participants in addition to those supported by the DSM program) and non-participant spillover (energy savings actions taken by non-participants as a result of the DSM program)
Authors found that total resource costs were 3.2 cents/kWh (44% utility-borne measure costs, 31% customer-borne measure costs, and 25% non-measure costs)
Least expensive programs were those that were: very large (energy savings or # of customers), older, targeted to normal (not early) replacement of equipment
Most expensive programs were: small (in terms of energy savings), newer programs, direct-install programs
Authors defend utilities' estimates of energy savings

all energy savings estimates included some degree of post-program savings verification
classified energy savings estimates into three types:

tracking databases (i.e. engineering estimates - database of which customers made what energy efficiency investments)
billing analyses

statistically adjusted engineering estimate (SAE) - regresses estimated engineering savings on actual customer bills

end-use metering

ex post verification measures include

verification of installation
hours of operation of energy efficiency measures

authors compare average negawatt costs for utilities that used different energy savings measurement methodologies and did not find a statistically significant difference

Feltes, Dan. 2003. "The Effects of Public Energy Efficiency and Energy Demand-Side Management Program Expenditures on State Electricity Energy Efficiency." GPPI Practicum.

Provides overview of electricity deregulation and trends in spending on utility DSM programs and state-run PBFs
Justifies use of energy intensity as dependent variable by citing relevant literature
Reviews estimates of costs and benefits of energy efficiency policies available in literature
Time-series panel study of 48 continental states & DC for years 1991-2000
Dependent variable is ln(Mhr/GSP)
Model with fixed effects for years found the following variables to be significant at the 0.99-level: manufacturing percent of State Gross Product, population, ln(state DSM expenditures), electricity price, and net metering. Year variables were all significant for 1995 onward
Feltes's independent variables were: population (millions); manufacturing % of GSP; price of electricity (cents/kWh); ln(DSM exenditures); ln(LIHEAP expenditures); ln(Low-Income Weatherization expenditures); net metering rule (dummy); industrial retail deregulation (dummy); residential retail deregulation (dummy); year dummies; heat day index; cooling day index; ln(energy efficiency expenditures) (subset of states);

Gehring, Kay. 2002. "Can Yesterday's Demand-Side Management Lessons Become Tomorrow's Market Solutions?" The Electricity Journal 15 (5): 63-69.

Gehring describes continuing role of DSM even following deregulation
Describes some problems with traditional DSM program evaluation methods--e.g. lower than anticipated energy prices and regulatory disallowances of DSM costs
Gehring highlights free-rider problem with both energy efficient appliance rebate programs and time-of-use rates

Gillingham, Kenneth, Richard Newell, and Karen Palmer. 2004. "Retrospective Examination of Demand-Side Energy Efficiency Policies." Resources for the Future. Discussion Paper.

Authors survey demand-side energy efficiency programs with the exclusion of building codes and CAFE
Authors focus on appliance standards, financial incentives, information and voluntary programs, and the management of government energy use
Authors review the literature on program costs and energy savings as well as estimating environmental benefits (including CO2 emissions avoided)

Golove, William and Joseph Eto. 1996. "Market Barriers to Energy Efficiency: a Critical Reappraisal of the Rationale for Public Policies to Promote Energy Efficiency." Lawrence Berkeley National Laboratory.

Authors review theoretical evidence for and against the "efficiency gap"
Golove and Eto review the implications of transaction cost economics for energy efficiency policy
Paper discusses seven market barriers: misplaced incentives, lack of access to financing, flaws in market structure, mis-pricing imposed by regulation, decisions influenced by custom, lack of information, and "gold plating"
Authors review findings from the "gray literature" (i.e. nonacademic literature) on real-world experience with energy efficiency investments in two areas: building design and commercial HVAC and lighting

Greening, Lorna et al. 2000. "Energy Efficiency and Consumption--the Rebound Effect--A Survey." Energy Policy 28: 389-401.

Authors summarize findings from a review of 75 empirical studies of the rebound effect
Outline four sources of technology efficiency gains

relative change in factor input prices
"autonomous" or naturally occurring and not linked to any specific cause
regulation-induced
result of government supported R&D

Authors define four categories of market responses to changes in fuel efficiency:

direct rebound effects
secondary fuel use effects
market-clearing price and quantity adjustments or economy-wide effects
transformational effects

wrt direct rebound effects, authors' review of literature leads them to conclude that increases in efficiency will be partially offset by increases in consumption but will nonetheless result in a net reduction in energy consumption
Authors review empirical studies of residential space heating, space cooling, and transportation

Space heating take-back of between 10 and 30% of energy savings
Space cooling take-back of between 0-50% of energy savings

More limited results for residential refrigeration, lighting, and water heating

Water heating: 10-40% take-back
Lighting: 5-20% take-back

Direct measurements of take-back effect for industrial and commercial firms are extremely limited

2% for process fuels in industrial facilities
30%+ for lighting

Authors summarize: "for the energy sectors of the US economy for which studies are available, we can conclude that estimates of the rebound are very low to moderate. The upper estimates, however, indicate a rebound effect that is not insignificant. Even these upper bound estimates, though, indicate that most or all of any reductions in energy use or carbon emissions are not lost to changes in behavior."

Hassett, Kevin and Gilbert Metcalf. 1995. "Energy Tax Credits and Residential Conservation Investment: Evidence from Panel Data." Journal of Public Economics 57: 201-17.

Authors seek to explain the "energy paradox" by modeling households' decision to make residential energy efficiency investments as an irreversible investment in the face of uncertainty
Authors find that the option value of postponing an energy efficiency investment helps explain low-rate of uptake of such investments even in the face of tax credits for energy efficiency investments
Empirical investigation (tax return panel data) shows that tax credits do have a statistically significant effect on energy efficiency decisions
Nonetheless, uncertainty and its effect are such that investment in energy efficiency is so low that the effect of tax credits is negligible

Haugland, Torleif. 1996. "Social Benefits of Financial Investment Support in Energy Conservation Policy." The Energy Journal 17 (2): 79-102.

Author estimate the difference between the NPV of energy efficiency investments from a Norwegian program in the early 1990's and the Net Social Benefit (NSB) of the program
Notes that Norwegian studies have shown general information campaigns to be ineffective
Program evaluation included estimate of free-rider effect from survey of enterprises (paper includes numeric estimates of free-riding based on responses to survey questions)
Author concludes that, from a NSB perspective, the efficiency gap is not real (at least not in large part)

Horowitz, Marvin. 2006. "A National Energy Efficiency Data Center: Removing the Curse of Invisibility." Presented at Modeling Workshop, Energy & Economic Policy Models: A Reexamination of Some Fundamental Issues. The University of California and The American Council for an Energy Efficient Economy. November 16th and 17th, 2006. www.aceee.org/conf/06modeling/horowitz.ppt.

Author reviews lack of data on energy efficiency
What data are available?

Of course, Form EIA-861
At the local level, most of the data are specific, related to local energy efficiency program, participants, or technologies
At the regional and federal level and lab level, the databases are spotty, and/or non-specific, and/or unreliable
Not-for-profit data collection is hit-or-miss depending on project funding, i.e., ACEEE, RFF, ASE
Nadel, Geller, et al. did some DSM program cataloging in the 1980’s
Many failed attempts in the 1990s, e.g., EPRI, DEEP
Rosenstock for EEI in 2005 a recent electric utility catalog
Kushler and York state scorecard for the past few years

Horowitz, Marvin. 2005. "Econometric Models for Distinguishing Between Market-Driven and Publicly-Funded Energy Efficiency." Proceeding of the European Council for an Energy Efficient Economy Summer Study, Mandelieu, France, May 2005.

Horowitz explains how public energy efficiency programs are a second-best policy after energy taxes for equalizing the social and private costs of energy consumption
Horowitz notes that with public energy efficiency programs consumers do not appear to be paying higher energy prices as they would under a system of energy taxes, but they are in fact at some point paying higher prices or taxes in order to fund the energy efficiency programs

Horowitz, Marvin. 2004. "Electricity Intensity in the Commercial Sector: Market and Public Program Effects." The Energy Journal 25 (2): 115-137.

Horowitz tries to separate the effects of DSM and market transformation public programs on energy intensity in the commercial sector
DSM said to include: energy audits, technical training, financial incentives or some combination thereof to those utility customers who purchase or install approved energy efficiency products and services
Market transformation said to begin in 1991 with EPA Green Lights program and its successor Energy Star program; author also includes some of DOE's R&D spending on energy efficiency
In addition to federal market transformation programs, state public benefits charges fund market transformation programs including those of regional organizations--Northeast Energy Efficiency Partnership (NEEP); the Midwest Energy Efficiency Alliance (MEAA); Southwest Energy Efficiency Project (SWEEP); and the Northwest Energy Efficiency Alliance (NEEA)--and state agencies in CA, NJ, NY, TX, VT, and WI.
Horowitz pays attention to specification bias and measurement error in econometric analysis of energy efficiency trends (p. 119)
Author uses DSM program data from Form EIA-861 noting that between 1989 and 1997 only utilities with retail sales or resales of more than 120,000 MWh were asked to report DSM data where this limit was raised to 150,000 MWh from 1998 on
Overcoming shortcomings in Form EIA-861 data:

Author apportioned DSM values to states based on proportion of utilities sales that were in each state
Pre-1992 figures were disaggregated based on proportions in 1992
Horowitz found an apparent underreporting of DSM savings in later years and used IVs to correct for measurement error

Author uses amount of electronic ballast shipments (for high efficiency commercial lighting) he attributes to market transformation programs as a proxy variable for market transformation program activity
Regression model includes lagged value of dependent variable
Author employs WLS after testing for heteroscedasticity
Author tests for endogeneity of electricity price variable and uses price in final model (not instrument)
Horowitz includes as independent variables (among others):

natural gas price since it is the closest substitute for commercial sector energy

absolute level of GSP since one might presume some economies of scale

Horowitz finds 54% realization rate of reported DSM energy savings--i.e. utilities were overstating their actual net energy savings from DSM programs by almost double

Jaffe, Adam and Robert Stavins. 1994. "Energy-Efficiency Investments and Public Policy." The Energy Journal 15 (2): 43-65.

Authors examine factors that determine the rate of adoption of energy-conserving technologies and what types of public policy can accelerate their adoption
Authors identify "nonmarket-failure" causes, which do not warrant government intervention and include: private information costs, high discount rates, and heterogeneity among potential adopters
Market-failure causes, which do justify government intervention, include: information problems, principal-agent slippage, and unobserved costs

Joskow, Paul and Donald Marron. 1992. "What Does a Negawatt Really Cost? Evidence from Utility Conservation Programs." The Energy Journal 13 (4): 41-74.

Authors examined data reported by 10 utilities on their conservation programs to calculate the life-cycle cost of a unit of energy savings
Authors estimate that utilities underestimate the life-cycle cost of a "negawatt" by at least a factor of two

more accurate measurement of actual energy savings likely to increase cost estimate by 50%
fully including all utility admin costs likely to increase cost estimate by 10% to 20%
including customer costs and accounting for free riders expected to increase costs by 25% to 50%

Authors found that utilities actually faced a much higher cost per "negawatt" than that alleged by proponents of energy efficiency such as Lovins (of the Rocky Mountain Institute) and EPRI because of:

overly optimistic assumptions in technical assessments of energy efficiency cost and savings
administrative costs for programs
diversity of customers' energy use patterns
imperfect information and the attraction of customers for whom energy efficiency investments are not cost effective

80% of energy savings came from industrial and commercial customers
Authors identified sources of bias in utilities' cost estimates

Utilities do not report all relevant costs

many types of administrative costs, including measurement and evaluation and overhead are not uniformly tracked or reported

Utilities rely on engineering estimates of energy savings rather than ex post audits

most utilities in authors' sample report energy savings based upon ex ante engineering estimates
energy savings estimates can differ from actual savings because of diversity in customer utilization patterns, changes in these patterns over time, the limited information a utility has about baseline electricity usage and subsequent changes, differences between participants and the "typical customer," and changes in behavior induced by conservation
CA required ex post evaluation beginning in 1994

Utilities fail to adequately adjust for free riders

many utilities simply ignore free riders
utilities fail to recognize that free riders may have utilization characteristics different from the customer population as whole such. In fact, free riders are likely to be those customers for whom energy efficiency is most beneficial. This would exacerbate the error introduced by not controlling estimates for free riders

Utilities assume excessive life times for energy efficiency investments

utilities use a wide range of assumptions for the investment lifetimes of what would appear to be nearly identical efficiency investments

customer transaction costs are real costs

time spent shopping, with energy auditors, and dealing with those who install the measures
value of lost business and inconvenience incurred while measures are being installed

cost of scrapping equipment often ignored

Leadership Group of the National Action Plan for Energy Efficiency. 2006. National Action Plan for Energy Efficiency

Benefits of energy efficiency

lower cost than supplying new generation only from new power plants

could also be used to avoid or delay investments in distribution or transmission systems

modular and quick to deploy
energy savings
environmental benefits
economic development (I think these are bullshit...)

employment impact of energy savings spent on other goods/services instead of energy
construction and installation jobs
increased property values

energy security

Barriers to energy efficiency

market barriers

"split-incentive" (builders vs. homeowners/tenants)
transaction cost

customer barriers

lack of information on energy efficiency opportunities
lack of awareness of programs
lack of funding for energy efficiency investments

utility, state, and regional planning barriers

energy efficiency not treated equally as supply-side resource

energy efficiency program barriers

impediments to optimal implementation of energy efficiency programs (e.g. lack of knowledge about best practices)

Energy efficiency programs have been operating in at least some areas since the late 1980s
Policy models for energy efficiency delivery (report includes table with examples by state of each model)

Systems Benefit Charge (SBC) Model

fixed amount per kWh collected by distribution or integrated utility
programs administered by utility, state agency, or third party

Integrated Resource Plan (IRP) Model

energy efficiency is treated on an equivalent basis as other supply-options in resource planning
cost recovery allowed through base rates or a separate charge

Request for Proposal (RFP) Model

utility or independent system operator (ISO) issues competitive solicitation to acquire energy efficiency from a third party provider to meed demand--typically used to save energy at peak demand
recently approved settlement at FERC allows energy efficiency along with load response and distributed generation to participate in the Independent System Operator New England (ISO-NE) Forward Capacity Market (2005, 2006)--so likely too late a development to include in my study

States also engage in energy efficiency RD&D (report includes table with examples of RD&D spending for some states)

Levine, et al. 1995. "Energy Efficiency Policy and Market Failures." Annual Review of Energy and the Environment 20: 535-555.

Several categories of hidden costs may affect purchase decisions related to energy efficiency:

reduced level of energy service
irreducible private costs (inconvenience of installing efficient equipment)
time and effort needed to learn about, search for, or develop confidence in the performance of new technologies

Authors discuss case studies of failures to adopt cost-effective energy-efficiency measures as prima facie evidence of market failures
until the mid- to late-1980s, utility DSM was advocated by environmentalists and energy efficiency proponents. Starting in 1989, several PUCs changed rate-of-return rules to allow utilities to profit from DSM thus making some utilities into DSM advocates. The main opposition to DSM came from industry groups (especially in the Northeast) who were concerned about price increases and cross-subsidization
Authors claim that utility DSM programs are good public policy because they open up large markets for better energy technology. DSM programs are sort of like large-scale demonstration programs for energy-efficient technology (this is the market transformation argument)

Loughran, David and Jonathon Kulick. 2004. "Demand-Side Management and Energy Efficiency in the United States." The Energy Journal 25 (1) 19-43.

Examines panel data set covering 324 utilities' reported DSM spending (from EIA Form-861, including imputed values for missing data) over period 1989-1999 to gauge the effect of energy efficiency program spending (a subset of DSM) by utilities on actual electricity sales
Finds significantly smaller effects on electricity demand from DSM programs (and thus a higher price per "negawatt") than reported by the utilities themselves
Reviews DSM evaluation literature and highlights potential reasons for bias in utilities' estimates of DSM energy efficiency gains
Study uses log of electricity sales as its dependent variable and controls for state and year fixed effects as well as number of customers and economic output
Regression model includes lagged variables for DSM expenditures (two lagged variables)

Lovins, Amory. 1996. "Negawatts: Twelve Transitions, Eight Improvements and One Distraction." Energy Policy 24 (4): 331-343.

Lovins argues for why there is an enormous efficiency gap--one that he thinks is largely due to perverse incentives in building design and a lack of integrated system engineering/design
Lovins claims that whole-system engineering can yield large energy efficiency gains but also even larger savings from right-sized equipment (in light of improved efficiency) and other benefits (e.g. improved worker productivity from optimized office building lighting and heating/cooling)
Perverse incentives - e.g. building designers are rewarded for spending money rather than saving on equipment and energy costs

Lovins, Amory. 2005. "More Profit with Less Carbon." Scientific American 293 (3): 74-83.

Lovins claims that energy efficiency offers the cheapest way to reduce carbon emissions
Lovins reviews barriers to energy efficiency that include the way that energy efficiency opportunities exist in many small areas rather than a few large ones so that people have trouble recognizing and taking advantage of them
Discusses importance of whole system optimization in design--e.g. optimizing the combination of insulation and heating/cooling equipment rather than adding the cost-effective insulation given the heating/cooling equipment
Notes importance of rewarding desired outcomes in the case of paying architects and builders for saving money on energy consumption rather than rewarding expenditures on materials

Metcalf, Gilbert. 2006. "Energy Conservation in the United States: Understanding Its Role in Climate Policy." National Bureau of Economic Research (NBER) Working Paper 12272. May 2006.

Uses state-level data on energy consumption from 1970 to 2003 to examine the effect of income and prices on energy consumption
Attempts to decompose the changes in energy intensity due to energy efficiency (defined as energy per unit of sector-specific economic activity) and changes in the composition of economic activity using a Fisher ideal index number methodology
Metcalf claims his paper is the only analysis of energy intensity changes at the state-level that uses a perfect decomposition method
Attributes between 2/3 and 3/4 of decline in energy intensity since 1970 to improvements in energy efficiency
Describes the arguments for and against various energy efficiency policies--e.g. the efficiency losses from applying Procrustean standards to heterogeneous consumers/firms and the behavioral responses (e.g. free-riding, rebound effect) that diminish energy efficiency policies' impact
Highlights that energy intensity has been declining consistently since 1960 but that the rate was higher in the 1970's and 1980's following the oil crises. In addition, the energy intensity of states has taken on a greater variance over the years with some states reducing intensity greatly and others not at all
Lays out reasons why his study is superior to the 2003 RAND book that also addressed state-level energy intensity
Uses average weighted price of energy in the state based on fuel uses as computed by EIA

Metcalf, Gilbert. 2006. "Federal Tax Policy Towards Energy." National Bureau of Economic Research (NBER) Working Paper 12568. Oct 2006.

Metcalf reviews both supply- and demand-side tax policy provisions of the Energy Policy Act of 2005
Reviews four arguments underpinning federal intervention in energy markets: externalities, national security, market failures and barriers in energy conservation, and rent expropriation
Describes "energy paradox" claim--i.e. that consumers require excessive rates of return before making energy efficiency investments
Lists tax provisions related to energy efficiency

Metcalf, Gilbert. 2008. "An Empirical Analysis of Energy Intensity and Its Determinants at the State Level." The Energy Journal 29 (3). Forthcoming.

Efficiency refers to the reduced energy use per unit of economic activity within a particular sector (e.g. industrial sector) while activity refers to the changing mix of economic activity (shift from energy intensive economic activity towards non-energy intensive economic activity) holding efficiency constant
To carry out this analysis, I analyze a data set on energy consumption at the state-level data between 1970 and 2001. I undertake an econometric analysis of the drivers of changes in energy intensity, efficiency, and activity at the state level. I find that rising per capita income and higher energy prices contribute to declines in energy intensity, primarily through improvements in energy efficiency.
Boyd and Roop (2004) first used a Fisher Ideal index as the basis for an exact decomposition of changes in energy intensity
into changes in energy efficiency and economic activity
Because of the lagged dependent variable, standard fixed effect regression procedures will produce biased estimates. I report estimates for the intensity index regression using the Arellano and Bond (1991) estimator

Nadel, Steven. 1992. "Utility Demand-Side Management Experience and Potential - A Critical Review." Annual Review of Energy and the Environment 17: 507-535.

Cited by Eto (1996) as the source of his list of DSM components:

general information for to foster awareness of energy saving opportunities
technical information to identify specific energy saving opportunities--e.g. energy audits
financial assistance (including loans and rebates) to promote investments in energy efficiency
direct or free installation of energy-efficient products
performance contracting
load control and load shifting
innovative tariffs (e.g. time-of-day and real-time pricing)

Nadel, Steven and Marty Kushler. 2000. "Public Benefit Funds: A Key Strategy for Advancing Energy Efficiency." The Electricity Journal 13 (8): 74-84.

Authors claim utilities began cutting DSM spending in anticipation of deregulation
In addition to PBFs, deregulated states rely on distribution company rates and energy service providers to promote energy efficiency; however, the authors do not see states using these widely

National Academy of Sciences. 1991. Policy Implications of Greenhouse Warming. Washington, DC: National Academy Press.

recommends energy efficiency as "no regret" option for combating climate change citing
includes negative costs for energy efficiency
[above info from reference to this report in Joskow and Marron (1992)]

Parfomak, Paul W. and Lester B. Lave. 1996. "How Many Kilowatts Are in a Negawatt? Verifying 'Ex Post' Estimates of Utility Conservation Impacts at the Regional Level." The Energy Journal 17 (4): 59-88.

utilities face economic incentives to overestimate the impact of their conservation efforts; nonetheless, the authors found no evidence of this in their econometric analysis
specifically, regulated utilities that can recover costs of DSM have an incentive to invest in programs with high ex ante conservation impacts but low ex post effects
factors that make it difficult to trust utilities reported DSM energy savings:

the difficulty of separating utilities' impacts from other forces (e.g. free-ridership, market transformation, take-back and price responses)
Overestimation:

utilities' incentive to overestimate conservation impacts

technical phenomena: improper installation or operation of equipment, inaccurate equipment operating profiles, mistreatment of equipment interactions
free-ridership
take-back

Underestimation

market transformation - indirect effects of conservation programs in changing the market availability of efficient equipment and affecting public attitudes toward conservation

study looked at reported conservation savings from a sample of 39 utilities in the Northeast and CA for the commercial and industrial sectors from program inception through 1993

unit of observation was utility service area
model used first differences approach to predict changes in electricity demand rather than absolute level
variables included beyond mine were fuel oil and coal prices
authors calculate electricity price the same way that I do, and they note that it is an imprecise proxy for actual electricity pricing since it includes miscellaneous charges not directly related to electricity consumption (i.e. it's not the true marginal price)
authors multiplied cooling degree days by a linear time variable to reflect the long-term increase in AC equipment penetration
reported conservation savings were an independent variable so the coefficient on this term is the % of reported savings actually realized, and the estimated coefficient for all utilities is -0.994
authors conclude that there is no systematic bias in utilities' reporting of DSM impacts
limitations: small sample, commercial and industrial only, no data on expenditures
authors did not trust the EIA-861 data after spot checking it and so went directly to the utilities and PUCs

authors report that utilities have improved their estimates of net conservation by using customer surveys, market studies, statistical analyses of billing records, and other techniques

Schleich, Joachim and Edelgard Gruber. 2006. "Beyond Case Studies: Barriers to Energy Efficiency in Commerce and Services Sectors." Energy Economics 30 (2): 449-464.

Authors look at 19 subsectors in the German commercial and services sectors and conduct Logit regressions to identify factors that affect the likelihood of adopting energy efficiency measures
Data for regressions come from a survey of ~2800 companies and public institutions
Authors review barriers to energy efficiency as found in the literature:

information and other transaction costs
bounded rationality (use of routine or rules of thumb)
capital constraints (particularly in public sector where entities cannot borrow)
uncertainty and risk (stochastic energy prices and option value of postponing ee investments)
investor/user dilemma

bias towards projects with short-payback despite inferior NPV because of short tenure of agents
lack of department accountability for energy costs
separate, nonfungible budets for capital and operating costs

Survey involved asking interviewees if they had adopted any of a range of energy saving measures, and interviewees were asked to rate the importance of potential barriers to energy efficiency
Logit models based on dichotomous dependent variable equal to one for entities that had implemented at least 50% of the feasible energy saving measures based on the survey data
Authors found that barriers varied considerably across subsectors with two or three variables usually statistically significant but without any clear pattern
Rented independent variable was significant in more than half of the models which the authors take to be support for the investor/user dilemma

Sola, Antonio Vanderley Herrero and Antonio Augusto de Paula Xavier. 2007. "Organizational and Human Factors as Barriers to Energy Efficiency in Electrical Motors Systems in Industry." Energy Policy 35 (11): 5784-5794.

Authors assessed the correlation between 27 organization human factors (OHF) and estimated electric energy losses from motor systems in ten industries in the Brazilian state of Parana where data on OHFs came from a survey
Paper not likely to be a useful citation

Sutherland, Ronald. 2000. "'No Cost' Efforts to Reduce Carbon Emissions in the US: An Economic Perspective." The Energy Journal. 21 (3): 89-112.

Author contrasts mainstream economic cost-benefit analyses of energy efficiency investments with the "free lunch" arguments proposed by adherents to the "energy conservation paradigm"
Sutherland criticizes energy efficiency proponents for failing to differentiate between actual market failures and other impediments to investments in energy efficiency
Some market barriers touted by energy efficiency proponents are really just the costs of market adjustments, which do not represent actual economic inefficiencies
Sutherland criticizes the use of below-market discount rates when estimating the benefits from energy saved
Author notes that empirical evidence shows that higher income households demonstrate lower discount rates
Author argues that "high discount rates do not reflect irrational decisions and market inefficiencies. Instead, high discount rates are efficient and rational when investments are risky, require sunk costs and can be delayed and considered at a later date"
A rebate provides a larger benefit to free riders than to new adopters. Rebates to free riders are merely a redistribution of wealth.
Tax credits to promote energy efficiency may have adverse equity effects since most participants in residential energy conservation programs are from higher income classes. However, the cost of such programs is paid for by all classes, and energy costs are a larger share of the incomes of lower income classes.

Tonn, Bruce and Jean Peretz. 2007. "State-Level Benefits of Energy Efficiency." Energy Policy 35: 3665-3674.

Reviews market barriers to energy efficiency (from an engineer's perspective...like other reports from national labs)
Authors boil down examples of market barriers to:

awarness
cost
capabilities
transaction issues

authors review some state-level ee programs including those of CA, NY, WI
paper offers examples of the economic development benefits of energy efficiency

Train, Kenneth. 1985. "Discount Rates in Consumers' Energy-Related Decisions: A Review of the Literature." Energy 10(12): 1243-1253.

Material below cited in Sutherland (2000)
Train reviews a large number of empirical studies of observed discount rates for household investments. He presents a range of estimated discount rates: for refrigerators, 39-100%; for air conditioning, 3.2-29%; for automobiles, 2-41%, with most of the estimates clustered at the higher end of the range.

Vine, Edward et al. 2003. "Public Policy Analysis of Energy Efficiency and Load Management in Changing Electricity Businesses." Energy Policy 31: 405-430.

Examines impacts of generic electric industry models (e.g. vertically integrated monopoly, unbundled monopoly, unbundled full competition) on energy efficiency and load management

Commercialization - utilities have an incentive to improve energy efficiency only up until the customer's meter
Privatization - when ownership of a utility shifts from the public to private sector, the utility will likely use a higher discount rate when assessing projects thus rejecting some energy efficiency investments that a public utility would not
The "wires" businesses may have incentives to promote energy efficiency to avoid bottlenecks, but they may lack customer relationships
Competition - if end-users see only short-term generations costs, they are less likely to engage in energy efficiency investments since short-term costs are uncertain. Moreover, electricity providers have little incentive to invest in energy efficiency if doing so means raising rates for non-captive customers

Includes extensive lists of policy and program barriers as well as how they are addressed by various mechanisms (i.e. control, funding, support, and market mechanisms)

Control mechanisms include: mandatory sourcing of energy efficiency, integrated resource planning (IRP)
Funding mechanisms include: PBFs
Support mechanisms include: developing the ESCO industry, voluntary agreements
Market mechanisms include: taxes on energy, tax credits for energy efficiency investments, demand-side bidding in wholesale markets

Interesting factors to consider for thesis: wholesale markets' allowance of demand-side resource bidding

Wing, Ian Sue. 2007. "Explaining the Declining Energy Intensity of the U.S. Economy." In Press. Resource and Energy Economics.

Econometric study of energy intensity from 1958-2000
Author finds that changes in the sectoral composition of the economy was the main driver behind energy intensity
Among the intra-industry forces, disembodied exogenous technical progress is the predominant energy-saving influence
Until 1973, most of the reduction in energy intensity was due to changes in the sectoral composition of the economy whereas industries were actually increasing their energy intensities.
After the first OPEC oil shock, the effects of structural changes diminished and instra-industry efficiency increased
The influence of technical change is very slightly energy-using prior to 1980 and energy-saving thereafter
Declining energy intensity within industries is largely due to energy demand from vehicle stocks

Wirl, Franz. 2000. "Lessons from Utility Conservation Programs." The Energy Journal 21 (1): 87-108.

Wirl uses a theoretical, economic framework to explain why DSM programs will face significant challenges in delivering actual energy savings due to principal-agent slippage (driven by consumers' private information) and the regulatory incentives utilities face which lead them to select inefficient DSM programs

Wirl, Franz. 1999. "Conservation Incentives for Consumers." Journal of Regulatory Economics 15 (1): 23-40.

using theoretical approach, Wirl identifies the optimal design for energy conservation incentives in light of customers' strategic reactions and utilities' imperfect information
Wirl criticizes much of the literature on DSM for overlooking the strategic behavior on the part of consumers that private information induces
Wirl concludes that, given his assumptions, the optimal DSM program will actually subsidize the most efficient consumers--since it is costly to masquerade as high efficiency
Wirl assumes the existence of two market failures commonly used to justify DSM programs--namely, the observation of too short payback times (i.e. too high discount rates) and prices for electricity that are below the marginal cost of production