Transcript Indiana Biomass Energy Working Group Aviation Biofuel Development

Historical & Contemporary
Consumer Valuations of Fuel Economy
& Other Measures of Vehicle Utility
Richard A. Simmons & Wallace E. Tyner
Purdue University
USAEE 2015
Concurrent Session 40: Transportation Fuels
Pittsburgh, PA
October 28, 2015
Outline

Introduction and Context

Historical trends in vehicle utility and prices

Hedonic price modeling

Insights and implications of 2014 Model Year results

Conclusions
2

Vehicle Technologies
 Evolutionary efficiency improvements
 New vehicle architectures
 Alternative fuels
 Modal Shifts

Economic Practicability
 OEM & consumer attractiveness
 Oil & energy prices
 GDP impacts
 Social Cost of Carbon?

Regulatory Constraints
 CAFE standards (MPG, CO2eq/mile)
 Related: RFS (ethanol & advanced biofuels)
3

Financial attractiveness and economic practicability for consumers

No loss of consumer choice in vehicle attributes or overall utility

OEM capability to deliver technologies with sufficient overall value

Feasibility of achieving regulatory requirements

Effectiveness in banking fuel and energy savings

Lifecycle environmental impacts
4
30
50
25
40
20
30
15
20
10
0
10
5
Share of MY13 Vehicles Compliant w/Future Year
CAFE Standard (%)
Fuel Economy (EPA Combined City/Hwy MPG)
60
0
Actual Fleet (1st Y-axis)
Avg Fed Std (1st Y-axis)
Min Fed Std (1st Y-axis)
MY13 Vehicles %Compliant
MY13 ICE % Compliant
MY13 HEV %Compliant
MY13 PHEV %Compliant
MY13 Other %Compliant
 Success toward compliance, 2011-2015
 Challenge to maintain fuel economy trajectory at affordable prices
Data source: U.S. Environmental Protection Agency. “Light-Duty Automotive
Technology, Carbon Dioxide Emissions, and Fuel Economy Trends. 2013
5











Performance (power, torque, acceleration, and ride metrics)
Aesthetic value, luxury, comfort and styling
Capacity (passenger & cargo space)
Safety (crash worthiness, ratings)
Fuel economy
Environmental impacts
Standard/Optional equipment
Consumer ratings, personal experience
Cost to operate, maintain and repair
Quality, warranty and reliability
Price & Resale value
Challenge:
Down-select to a set of
objective, representative attributes?
6
General dynamic force balance:
𝐹𝑡𝑟𝑎𝑐𝑡𝑖𝑣𝑒 = 𝐹𝑟𝑜𝑙𝑙𝑖𝑛𝑔_𝑟𝑒𝑠𝑖𝑠𝑡𝑎𝑛𝑐𝑒 + 𝐹𝑎𝑒𝑟𝑜_𝑑𝑟𝑎𝑔 + 𝐹ℎ𝑖𝑙𝑙_𝑐𝑙𝑖𝑚𝑏 + 𝐹𝑎𝑐𝑐𝑒𝑙𝑒𝑟𝑎𝑡𝑖𝑜𝑛
2
𝐹𝑡𝑟𝑎𝑐𝑡𝑖𝑣𝑒 = 𝑚𝑣𝑒ℎ ∙ 𝑔(𝐶0 + 𝐶1 𝑣𝑣𝑒ℎ ) + 0.5𝜌𝑎𝑖𝑟 𝐶𝐷𝑟𝑎𝑔 𝐴𝐹 𝑣𝑣𝑒ℎ
+ 𝑚𝑣𝑒ℎ 𝑔 ∙ 𝑠𝑖𝑛𝜃𝑔𝑟𝑎𝑑𝑒 + 𝑚𝑣𝑒ℎ ∙ 𝑎𝑣𝑒ℎ
𝑊𝑡𝑟𝑎𝑐𝑡𝑖𝑣𝑒 = 𝑃 𝑡 =
𝐹𝑡𝑟𝑎𝑐𝑡𝑖𝑣𝑒 ∙ 𝑣𝑣𝑒ℎ 𝑑𝑡
Vehicle powertrain power balance:
𝑊𝑡𝑟𝑎𝑐𝑡𝑖𝑣𝑒 = 𝑊𝑒𝑛𝑔𝑖𝑛𝑒_𝑜𝑢𝑡 ∙ 𝜂𝑡𝑟𝑎𝑛𝑠 = 𝜇𝑓𝑢𝑒𝑙 ∙ 𝑚𝑓𝑢𝑒𝑙 ∙ 𝜂𝑡ℎ_𝑒𝑛𝑔𝑖𝑛𝑒 ∙ 𝜂𝑡𝑟𝑎𝑛𝑠
𝑚𝑓𝑢𝑒𝑙 =
2
2
𝑚𝑣𝑒ℎ ∙ 𝑔(𝐶0 + 𝐶1 𝑣𝑣𝑒ℎ
) + 0.5𝜌𝑎𝑖𝑟 𝐶𝐷𝑟𝑎𝑔 𝐴𝐹 𝑣𝑣𝑒ℎ
+ 𝑚𝑣𝑒ℎ 𝑔 ∙ 𝑠𝑖𝑛𝜃𝑔𝑟𝑎𝑑𝑒 + 𝑚𝑣𝑒ℎ ∙ 𝑎𝑣𝑒ℎ ∙ 𝑣𝑣𝑒ℎ 𝑑𝑡
𝜇𝑓𝑢𝑒𝑙 ∙ 𝜂𝑡ℎ_𝑒𝑛𝑔𝑖𝑛𝑒 ∙ 𝜂𝑡𝑟𝑎𝑛𝑠
What minimum number of objective attributes can lead to a reasonably accurate characterization?
Consider these 3: fuel
economy, vehicle mass, acceleration time
7
Utility function attributes:








Objective
Representative
Tangible (i.e., “direct” consumer measures)
Continuous rather than categorical
Readily grasped by OEMs and consumers
Historically tracked
Normalized to base year
Can be combined “orthogonally” (i.e., minimize collinearity)
Proposed utility function:
𝑈𝑡𝑖𝑙𝑖𝑡𝑦 = 𝑓𝑛 𝐹𝐶, 𝐴𝐶𝐶𝐸𝐿, 𝐶𝑊𝑇
Time horizon:
1978 – 2014
Market:
U.S. passenger cars (300 million +)
8
2.00
2.00
1.80
1.80
ACCEL
1.40
1.60
FC
1.40
1.20
ACCEL
FC
VOL
1.00
CWT
0.80
1.20
1.00
0.80
CWT
0.60
0.60
0.40
0.40
CAFE
2012+0.20
0.20
0.00
1975
CAFE
1978-1990
1980
1985
1996
1997
Normalized Value
1.60
1990
ACCEL
1995
FC
2000
VOL
2005
2010
0.00
2015
CWT
Note: Data are normalized to baseline values in the most recent reference year, 2014.
Simmons, R.A. and Tyner, W.E., “Fuel economy and vehicle attribute valuation
trends via historical and contemporary hedonic pricing analysis”
Submitted to the J. of Transportation Research, Part D. In review (Sept 2015).
9
2.00
2.00
1.80
1.80
ACCEL
1.60
1.40
FC
1.40
1.20
ACCEL
FC
VOL
1.00
0.80
0.80
CWT
0.60
0.60
0.40
0.20
0.00
1975
1.20
1.00
CWT
CAFE
1978-1990
1980
1985
0.40
CAFE
2012+0.20
1996
1997
Normalized Value
1.60
1995
1990
ACCEL
FC
2000
VOL
2005
2010
0.00
2015
CWT
KEY POINTS:
 Objective measures of utility have steadily improved over a 37 year period;
 Fuel Consumption (FC) and Acceleration (ACCEL) trends are:
correlated to & influenced by regulations
Simmons, R.A. and Tyner, W.E., “Fuel economy and vehicle attribute valuation
trends via historical and contemporary hedonic pricing analysis”
Submitted to the J. of Transportation Research, Part D. In review (Sept 2015).
10
Key periods and highlights:
FC
1978 through 1996:


FC
FC dominant due to regulations
CWT reduced for quick compliance

0.390
0.733
FC

ACCEL
ACCEL
CWT
0.728
CWT
1
-0.910
-0.624
FC
FC:ACCEL aligned with physical principles
CWT a proxy for many dimensions of utility
1
1
2014:

1
-0.307
FC
ACCEL dominant
CWT increases simultaneously
CWT
1
ACCEL
CWT
1997 through 2014:

ACCEL
FC
ACCEL
CWT
ACCEL
1
CWT
1
-0.832
0.644
1
-0.667
 Substantial utility gains enabled by
simultaneous technological innovation
1
11
30,000
1.20
25,000
1.00
20,000
0.80
Utility_1
15,000
0.60
10,000
Utility_1:
Utility_1(Yeari) ≈ 0.0174*(Yeari) - 34.043
R² = 0.976
Average annual growth rate≈ 2.7%
5,000
0
1975
1985
1995
Real Price_1
2005
0.40
Normalized Utility_1
Real Price (2014$)
Real Price_1
0.20
0.00
2015
Utility_1
𝑈𝑡𝑖𝑙𝑖𝑡𝑦_1 ≡ 𝑓𝑛(𝐹𝐶 −1 ∙ 𝐴𝐶𝐶𝐸𝐿−1 ∙ 𝐶𝑊𝑇)
12
30,000
1.20
25,000
1.00
20,000
0.80
Utility_1
15,000
0.60
10,000
Utility_1:
Utility_1(Yeari) ≈ 0.0174*(Yeari) - 34.043
R² = 0.976
Average annual growth rate≈ 2.7%
5,000
0
1975
1985
1995
Real Price_1
2005
0.40
Normalized Utility_1
Real Price (2014$)
Real Price_1
0.20
0.00
2015
Utility_1
KEY POINTS:
 Real price of autos in 2014 is approximately the same as it was in 1986;
 However, consumer metrics for utility have increased substantially.
𝑈𝑡𝑖𝑙𝑖𝑡𝑦_1 ≡ 𝑓𝑛(𝐹𝐶 −1 ∙ 𝐴𝐶𝐶𝐸𝐿−1 ∙ 𝐶𝑊𝑇)
13
A hedonic pricing analysis is performed to estimate the disaggregated contribution
of each attribute to overall utility, proxied by the response variable, price.
The general model follows the form:
𝑛
ln 𝑃𝑖 = 𝛽0 +
𝑚
𝛽𝑗 ∙ ln 𝑋𝑖𝑗 +
𝑗=1
𝛽𝑗 ∙ 𝑌𝑖𝑗 + 𝜖𝑖
𝑗=𝑛+1
Pi represents the purchase price for vehicle i, inflated using the new vehicle index,
b0 represents the intercept,
bj are the regressor coefficients representing elasticities of price with respect to a
set of up to n continuous variables, Xij and m-n dummy variables, Yij.
ei represents the residual error between the predicted and actual values.
14
We make simplifying assumptions:
 Considering 3 primary attributes (FC, ACCEL and CWT)
 Neglect additional categorical, continuous or dummy variables
The simplified model now has the form:
ln(𝑃𝑖 ) = 𝛽0 + 𝛽1 ∙ ln(𝐹𝐶𝑖 ) + 𝛽2 ∙ ln(𝐴𝐶𝐶𝐸𝐿𝑖 ) + 𝛽3 ∙ ln(𝐶𝑊𝑇𝑖 ) + 𝜖𝑖
Pi represents the purchase price for vehicle i, inflated using the new vehicle index;
b0 represents the intercept;
bj are the regressor coefficients representing elasticities of price with respect to a
set of 3 continuous variables, Xij;
ei represents the residual error between the predicted and actual values.
  The log-log form yields bj in terms of % change in the dependent variable for
a given change in the independent variable, holding others constant.
15
Price elasticities suggest Willingness To Pay (WTP) for key attributes
Time Period
1978-1996
1997-2014
1978-2014
2014
Response Variable
Estimator
Observations
ln(Real Price2)
ln(Real Price2)
ln(Real Price2)
ln(MSRP)
WLS
WLS
WLS
WLS
19
18
37
814
0.992
0.963
0.943
0.722
Param. Est.
(Std. Error)
Param. Est.
(Std. Error)
Param. Est.
(Std. Error)
Param. Est.
(Std. Error)
9.97 ***
0.59
R2
Attribute
Intercept
Coeff.
b0
0.41
(2.054)
ln(FC)
b1
-1.95 ***
(0.163)
ln(ACCEL)
b2
-0.05
(0.073)
ln(CWT)
b3
1.95 ***
(0.314)
The ‘dominant two’
by time period:
FC
CWT
-3.53
(3.175)
-0.08
(0.091)
-0.41 *
(0.220)
2.02 ***
(0.385)
ACCEL
CWT
(2.180)
-0.93 ***
(0.158)
-0.65 ***
(0.114)
0.50 *
(0.302)
FC
ACCEL
(0.527)
-0.26 ***
(0.049)
-0.70 ***
(0.051)
1.59 ***
(0.063)
ACCEL
CWT
Simmons, R.A. and Tyner, W.E., “Fuel economy and vehicle attribute valuation
trends via historical and contemporary hedonic pricing analysis”
Submitted to the J. of Transportation Research, Part D. In review (Sept 2015).
16
Willingness To Pay (WTP) for key attributes by time period
Price Elasticity
FUEL CONSUMPTION
2.00
1.50
1.00
0.50
0.00
-0.50
-1.00
-1.50
-2.00
ACCEL
WEIGHT
WEIGHT
FC ACCEL


A denotes 3 parameter model;
B denotes 7 parameter model (adds trim & drive)
Simmons, R.A. and Tyner, W.E., “Fuel economy and vehicle attribute valuation
trends via historical and contemporary hedonic pricing analysis”
Submitted to the J. of Transportation Research, Part D. In review (Sept 2015).
17

Real weighted average prices for autos have remained roughly
constant since 1986.

Simultaneous progress achieved on several “counter-acting”
technological dimensions.

Simple equal-weight index suggests performance has increased
about 2.7% per year while prices remained relatively constant.

Hedonic pricing methodologies quantify the relative WTP for vehicle
attribute weightings during times of regulatory constraint.

Authors’ methods accommodate sales-weighting and blocking by
footprint; will be useful in projecting future fuel economy responses.

Consumers of 2014 MY cars appear to value acceleration
performance between two and three times as much as fuel economy.
18
1
9
20
1.20
25,000
1.00
20,000
0.80
Utility_1
15,000
0.60
10,000
0.40
𝑈𝑡𝑖𝑙𝑖𝑡𝑦_1 ≡ 𝑓𝑛(𝐹𝐶 −1 ∙ 𝐴𝐶𝐶𝐸𝐿−1 ∙ 𝐶𝑊𝑇)
5,000
0.20
0
0.00
1975
1980
1985
1990
1995
Real Price_2

Normalized Utility_1
Price ($)
30,000
2000
2005
2010
2015
Utility_1
Real Price_2 is inflated using the new vehicle price index.
21
ICCT



Efficiency
 Thermochemical, electrical
 Weight, friction and drag reductions
Advanced vehicle technology
 Internal combustion engine
 Advanced powertrain
 New architectures
Alternative fuels
 Biofuels, Natural gas
Transit systems
 Intelligent routing
 Modal shifts
Clean Vehicle Technology Projections
Passenger LDV Sales (millions)

180
160
140
120
100
80
60
40
20
0
IEA
EV,FCV
PHEV
Hybrid
CNG/LNG
Diesel
Gasoline
Source: International Council on Clean Transportation, 2010, M. Kromer and C. Evans
Source: International Energy Agency (IEA), Energy Technology Perspectives (ETP), 2012.
22



OEM product offerings &
incentives
Consumer response
 Driving behavior (VMT)
 Transit options
 Benefit-Cost
Economy-wide
 Oil & energy prices
 GDP impacts
 Social Cost of Carbon
(SCC)
Source: RA Simmons, G Shaver, W Tyner, SV Garimella. A benefit-cost assessment of new vehicle
technologies and fuel economy in the U.S. market. 2015. Journal of Applied Energy,
23