Professional Documents
Culture Documents
Agenda
A d
Why?
Theory and Literature Review
Research Design, Data and Methods
Findings
Implications
Future Extensions
1
Importance of Building Energy Codes
Buildings use over 40% of all primary energy in the
country.
In 2004/2005
residential buildings
g accounted for almost 22% of p
primary
y
energy use
commercial buildings represented the balance of 18% of
primary
p y energy
gy use ((US DOE, 2007).)
Codes reflect (non) “Lost Opportunities”
Energy efficiency measures have to be installed during
construction or else they wont be
50+ year life of buildings
Long life of capital equipment means its not cost effective
to replace before end of equipment’s
equipment s life
2
What are Building Energy Codes?
Energy
gy codes are standards and p
practices that
builders must follow when constructing or
renovating a building
Examples:
Wall insulation requirements
Windows: glazed or double or triple paned
Heating Ventilation and AC equipment
Roofing materials
Reduce overall building energy use by installing
more efficient equipment, design, processes
3
Effects of Policy Interventions
California
has led
the nation
in energy
efficiency
Title
Titl 24
building
codes are
California
specific
p
4
Extreme Variations in Code Adoption
Much different
picture today
than in 2000
when 14 states
didn’t have
codes
5
Explaining Energy Code Adoption
Higher
g energy
gy p
prices?
Political capacity?
Political ideology?
Climate (energy use)?
6
Theoretical Explanations
Environmental protection: “states w/liberal public opinion, strong
environmental
i t l iinterest
t t groups, lib
liberall llegislatures,
i l t and
d
professionalized legislatures are the most committed to
environmental protection” (Hays, Elser, Hays 1996, pg. 41).
Kuznets Curve: increased wealth leads to more demands for
environmental protection.
Expected utility: states with higher electricity prices, and that use
more energy,
energy are more likely to adopt energy saving policiespolicies.
Spatial: states with efficient neighbors are more/less likely to
adopt.
Environmental federalism of Oates (2001) vs diffusion of Berry
and Berry (1990).
Political capacity: large enforcement component to energy codes
means states with low capacity will not enact because they know
they will not be able to enforce.
7
Research Design
Dependent variables: Estimated vintage of building energy code:
Residential (1992
(1992-2005):
2005): No statewide code =0
=0, oldest code = 1
1,
newest = 7
Commercial (2000-2005): No statewide code =0, oldest code = 1,
newest = 4
E l
Explanatory
t variables
i bl ((expected
t d sign):
i ) source
Log of population (+): US Census
Log of Gross State Product per capita (+): US Bur. Econ. Analysis
Climate: NOAA
Heat degree days: # of days requiring heat (+)
Cool degree days: # of days requiring cooling (+)
Political ideology % voting for republican in previous presidential
election (-):
( ): US elections commission
3 year average of Relative Political Capacity (RPC) 3 year average (+):
Calculated from US census, BEA data (see next slide)
8
RPC Variable
1992-2005 (we have available updated 2006 data )
50 US States
Two stage equation
First stage: predict tax effort as percent of gross state product using the
following functional form
10
Results
1 2 3 4
OLS Ordinal Logit OLS Ordinal Logit
res_code res_code comm_code comm_code
? Ln Population 0.616*** 0.605*** -0.134*** -0.264
(0.0738) (0.0997) (0.0457) (0.183)
! Lagged Ln GSP/capita 1.106*** 0.0477 1.018*** 2.252***
(0.424) (0.452) (0.379) (0.868)
%VVoting
ti Republican
R bli -0.0460***
0 0460*** -0.0345***
0 0345*** 0 0267***
0.0267*** 0 0789***
0.0789***
(0.00723) (0.0113) (0.00578) (0.0207)
Relative Political Capacity 0.454 -1.055 4.751*** 11.02***
(1.282) (1.510) (0.689) (1.819)
Lagged Electricity Prices 0.107*** 0.103** 0.0704** 0.253***
! (0.0302) (0.0436) (0.0284) (0.0938)
Cool Degree Days -0.00223*** -0.00191*** -0.000403*** -0.00115***
? Heat Degree Days
(0.000121)
-0.000595***
(0.000261)
-0.000567***
(5.52e-05)
-0.000219***
(0.000430)
-0.000625***
(7.54e-05) (9.73e-05) (3.25e-05) (0.000220)
0 south 0.632*** 0.137 -0.0553 -0.108
((0.124)) ((0.256)) ((0.157)) ((0.411))
0 coastal 0.299 0.349 0.312** 0.396
(0.199) (0.228) (0.149) (0.420)
! Residential Energy Code 0.202*** 0.428***
(0.0147) (0.0571)
Constant 0.867 -6.869***
(1.964) (1.920)
Observations 575 575 288 288
R-squared 0.281 0.096 0.468 0.24
Standard errors in parenthes 11
*** p<0.01, ** p<0.05, * p<0.1
Interpretation of Findings
Population: mixed results
Positively related to residential code adoption
Weakly negative relationship with commercial code adoption
Perhaps population density a better measure?
Differential impact of politics in each code type
More republican presidential votes predicts weaker residential codes
codes, stronger
commercial codes
An explanation of why SC codes weren’t adopted in 1994; “Opposition was based on the
need for local governments to raise taxes or fees to implement the codes.” (DOE, 2008)
More democrats predicts weaker commercial codes: union opposition?
Political capacity: matters for commercial codes
codes, but not residential
Enforcement for commercial buildings more concentrated
Kuznets curve: Wealthy states more likely to adopt
Although lower income states have greater relative need
Spend more on energy as % of total income
Expected utility: higher retail prices predict more recent code adoption
Climate (HDD & CDD) negatively predicts adoption
Surprising. Added the variable for coastal states but this didn’t account for it
Perhapsp an interaction term between these two variables?
Residential code adoption predicts commercial code adoption
12
Predicted Probabilities
Ideal type #1
Poor, interior state, with low electricity prices
(coal/hydro) and no residential code
Predicted probability of no commercial energy code
= 98%
Ideal type #2
Rich, coastal state, with high electricity prices
(gas/nukes) with residential code
Predicted probability of no commercial energy code
= 1%
Cumulative predicted probability of having a recent
(3 or 4 out of 4 possible) commercial energy code =
82% 13
Implications for a Carbon Constrained
W ld
World
States with growing populations are more likely to adopt
codes in the future
Rural areas are being left further behind?
Higher energy prices due to a carbon tax is likely to improve
building energy codes
Clear differences between building types regarding ideology
Consensus will need to be made on adoption
RPC predicts commercial code adoption—but not
residential.
Expectations of a lack of enforcement might be deferring
residential
id ti l adoption?
d ti ?
Poor states need incentives/regulations to adopt statewide
codes
G tti energy code
Getting d adopted
d t d forf one building
b ildi type
t is
i likely
lik l
to lead to adoption of the other type 14
Future Extensions
Extend commercial code data back to 1992
Include educational attainment measure (?)
Spatial variable for contiguous state code
adoption
Implementation
p measures
Whether executive agencies or legislatures are
responsible for code cycle improvements
Suggestions and comments are welcome
THANKS!!!
15
Residential Energy Codes
Coding by code and year
Codes age relative to new model energy codes
BCAP data from 2000-2007
US DOE data uses to estimate 1992-1999
residential code
YEAR 2004/2006
2003/2000 98 1995 1993/92 1990 1989/88 1985/83 1979/81 1975/76 1973 none
1992 7 6 5 4 3 2 1 0
1993 7 6 5 4 3 2 1 0
1994 7 6 5 4 3 2 1 0
1995 7 6 5 4 3 2 1 0
1996 7 6 5 4 3 2 1 0
1997 7 6 5 4 3 2 1 0
1998 7 6 5 4 3 2 1 0
1999 7 6 5 4 3 2 1 0
2000 7 6 5 4 3 2 1 0
2001 7 6 5 4 3 2 1 0
2002 7 6 5 4 3 2 1 0
2003 7 6 5 4 3 2 1 0
2004 7 6 5 4 3 2 1 0
2005 7 6 5 4 3 2 1
16
0
2006 7 6 5 4 3 2 1 0
2007 7 6 5 4 3 2 1 0
Commercial Energy Code
Coding by code and
year
Codes age
g relative to
new model energy YEAR
1992
2004
commercial code
2001 1999 1989 none
4 0
codes 1993
1994
1995
4
4
4
0
0
0
1996 4 0
1997 4 0
1998 4 3 0
1999 4 3 0
2000 4 3 2 0
2001 4 3 2 0
2002 4 3 2 0
2003 4 3 2 0
2004 4 3 2 1 0
2005 4 3 2 1 0
2006 4 3 2 1 0
2007 4 3 2 1 0
17
Correlations
. corr comm_code res_code lnpop ln_gsp_cap repub threerpc1 rtl_elec cdd65 hdd65 south
(obs=288)
( )
comm_c~e res_code lnpop ln_gsp~p repub threer~1 rtl_elec cdd65 hdd65 south
comm_code 1.0000
res_code 0.5537 1.0000
lnpop 0.1717 0.2864 1.0000
ln_gsp_cap 0.1926 0.2630 0.1895 1.0000
repub -0.2047
0 2047 -0.3514
0 3514 -0.2632
0 2632 -0.3938
0 3938 1 0000
1.0000
threerpc1 0.2859 0.0248 0.1498 -0.2805 -0.1197 1.0000
rtl_elec 0.3047 0.3069 0.1083 0.3522 -0.6367 0.1111 1.0000
cdd65 -0.0983 -0.1725 0.3745 -0.1814 0.1865 -0.0626 -0.1381 1.0000
hdd65 -0.0604 -0.0045 -0.5120 0.1234 -0.0685 0.0502 0.0804 -0.8857 1.0000
south -0.0262 -0.0428 0.2617 -0.3412 0.2656 -0.0707 -0.2317 0.5970 -0.6459 1.0000
18
Descriptives
19