Do Increases in Energy Efficiency Reduce Energy Consumption: Engineering vs' Economic Analysis

1
Do Increases in Energy Efficiency Reduce Energy
Consumption Engineering vs. Economic Analysis

• By Will Larson
• The George Washington University
• February 27, 2008

I would also like to acknowledge invaluable help
and guidance from Professor Anthony Yezer
2
Plan for Talk

• Intuitive examples
• The Literature
• What economic theory says
• What empirical analysis says
• Natural experiment efficiency shift
• Implications for policy and future research

3
Intuitive Examples

• Steam engines, Stanley Jevons (1865)
• Efficiency led first to drop in coal use
• Over time, steam engines were used in more
  applications
• Eventually expanded use of steam resulted in
  increased use of coal

4
Intuitive Examples

• Computers
• Large, energy inefficient processors gave off
  substantial heat and required large, cumbersome
  cooling systems
• Microprocessors lowered energy consumption both
  in the computer and for cooling
• Then the desktop replaced the central computing
  facility, followed by laptops, ipods, cell
  phones, blackberries

5
Intuitive Counter Example
6
Literature- the beginning of the modern debate

• Hirst and Carney (1978)
• Oak Ridge Labs study-classic engineering (static)
  forecast
• Assume constant utilization
• Brookes (1979) and Khazzoom (1980)
• Postulate that rebound effects could dominate
  initial energy efficiency effect

7
Theoretical Demand Response to an Input
Efficiency Improvement
8
Literature the current debate

• Micro Studies of the Rebound Effect
• Khazzoom (1986) Rebound effect 65 for home
  heating
• Haas and Biermayr (2000) Rebound effect 30 for
  home heating
• Macro Studies of Efficiency Effects
• Brookes (2000)
• Energy efficiency leads to growth growth leads
  to higher energy consumption
• Saunders (2000)
• As energy efficiency increases, substitution
  occurs among inputs towards energy

9
Literature- the current debate

• Hertwich (2005) identifies 5 effects
• Substitution effect
• Income effect
• Secondary effects
• Reduction in marginal cost for goods requiring
  energy inputs may themselves be inputs in the
  production in other goods
• Market-Clearing price/quantity adjustments
• Transformational effects
• Changing preferences, institutions, organization
  of production

10
Literature the current debate

• Hertwich (2005)
• Substitution and income effects
• Between 5 and 65
• Secondary effects, market-Clearing price/quantity
  adjustments, and transformational effects
• No consensus in the literature
• We do know there is a rebound, but we cannot say
  that it is greater than 100 with any certainty

11
Economic Theory Household Gets Utility from his
Produced by Energy
Subject to
Rewriting E,
Where
And
And
12
Empirical Analysis household energy consumption
demand estimation

• Dubin and McFadden (1984)
• United States
• Appliance demand and energy consumption
• Nesbakken (2001)
• Norway
• Housing consumption endogenous
• Selectivity bias correction in energy technology
  choice
• Reiss and White (2005)
• California
• Uses RECS

13
Natural Experiment mandated energy efficient
household appliances

• California implemented appliance regulations in
  1976
• Federal standards implemented in the 1987
  National Appliance Energy Conservation Act
  (NAECA) taking effect between 1988 and 1994.
  Non-binding in California due to state laws
  already in effect.
• Estimate a household energy use equation
• Compare consumption in states with no standards
  to California from 1993-2001 to look for
  convergence in energy use.

14
Natural Experiment Possible Rebound Effects of
Federal Regulation
15
Natural Experiment Household Energy Use Equations

• Variable Definitions

16
Natural Experiment Household Energy Use Equation
Specifications

• 1
• 2
• 3

17
Natural Experiment Null Hypotheses

• If energy efficiency reduces energy consumption,
  then
• The regulation fixed effect should be positive
• conditional energy consumption initially above
  California
• The time coefficient should be negative or zero
• Old appliances still being phased out in
  California
• the time X regulation coefficient should be
  negative
• Energy inefficient appliance phase-out in
  regulation-constrained regions

18
Data Used

• EIAs Residential Energy Consumption Survey
  (RECS)
• Survey taken every 3-4 years
• Probabilistic sample of households across the
  U.S.
• Contains data on household characteristics,
  dwelling characteristics, appliances, energy
  consumption, energy prices

19
(No Transcript)
20
Results Pooled Cross Sections
21
Results Pooled Cross Sections
22
Results Pooled Cross Sections

• Consumption has fallen in NAECA-bound USA sample
  relative to California
• But why only 20-40? 8 years should be enough
  for most appliance turnover.

23
Results Pooled Cross Sections

• Disaggregating by region

24
Conclusions

• Energy efficiency regulations have caused some
  convergence in household energy consumption
  (10-30 depending on region).
• 75 of difference remains unexplained.
• Possible explanations
• Californians have a preference for energy
  efficiency
• Other unobserved attributes captured by fixed
  effects.

25
Conclusions

• People are wrong to say that the efficiency
  advantage California holds is due to efficiency
  regulations.
• Evidence of convergence means rebound effect is
  less than 100
• Analysis takes into account both direct and
  indirect effects within households
• Further research is needed
• 2005 RECS
• Add commuting to model (variable not in RECS)
• Other indirect effects outside of household