assessing the strength and effectiveness of renewable electricity feed-in tariffs
DESCRIPTION
Presented at U.S. Association for Energy Economics conference in Washington, DC in October 2011.TRANSCRIPT
1
Assessing the Strength and Effectiveness of Renewable Electricity Feed-in Tariffs
Joe Indvik, ICF International
Steffen Jenner, Harvard University
Felix Groba, DIW Berlin
USAEE/IAEE 2011 North American Conference:"Redefining the Energy Economy: Changing Roles of Industry, Government and Research"
Background
Renewable electricity (RES-E) is rapidly expanding in magnitude and geographic scope
Literature generally claims that government incentives are justified by...
Climate and pollution externalitiesBarriers to entryEnergy security concerns
2
3
RES-E Policy Levers
Price Quantity
InvestmentInvestment subsidies
Tax credits
Low interest/ soft loans
Tendering systems for investment grants
Generation Feed-in tariffsRenewable portfolio standards (RPS)
Tendering systems for long term contracts
4
Price-based RES-E production incentive Funded by state budget and/or electricity price
increase Helps renewables achieve grid parity
Everything you need to know about FIT’s in 60 seconds
RES-E Generator Grid
Electricity Price
State budget
Tariff
Contract
€
5
Years of RES-E policy enactment in Europe:
Feed-in tariff
Quota
BE
CZ BG
HU EE IE
IT DK GR FR LT NL MT RO BG
DE IT LU ES AT PT GB SE SI SK CY
1990 1992 1993 1994 1998 2001 2002 2003 2004 2005 2006
6
FIT Policies and RES-E Capacity
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 20080
5
10
15
20
25
0
2000
4000
6000
8000
10000
12000
14000
FIT policies enacted
Annual RES-E capacity added*
* Solar PV and onshore wind
Correlation = 0.87 Causation?
Polic
ies
Megaw
atts
7
Have feed-in tariffs significantly increased onshore wind power and
solar PV deployment in Europe?
8
The Traditional Approach
Capacity Added = β1(Policy Dummy) + β2(Some Controls)
Inevitably, β1 is positive and highly significant.
So the policy is effective!
Except for...
Two Problems
1
Policy Heterogeneity“Not all FIT’s are created equal.”
Omitted Variables Bias“What you don’t see can hurt you.”
2
Linear OLS pooled cross-section regression:
9
Problem 1: Omitted Variables Bias
Establishing Causality
PolicyCapacity Growth
Political Environment
Natural Resources
Socio-Economics
Electricity Prices
Other Policies
Region Transmission
UnobservedState Traits
Broader Trends
Bias
10
11
Our Model
ln(Added Capacityist) = β0 + β1SFITist + β2INCRQMTSHAREst + βxZist + βyWist + μs + uist
Incremental Share
Measure of quota stringency developed by Yin and Powers (2009)
Policy Controls
Suite of binary policy control variables for other RES-E policies
Socio-Economic Controls
Suite of socioeconomic controls
Country Fixed Effects
Controls for country characteristics that do not
change over time
Added Capacity
Additional RES-E nameplate generation
capacity added each year
for energy technology i, in country s, in year t.
FIT Strength
Our new measure of the generation incentive
provided by a FIT
12
Problem 2: Policy Heterogeneity
13
1/0
Binary Variable: The king of renewable energy policy
analysis thus far.
Duration
Magnitude
Electricity price Risk and uncertainty
Binary variables do not accurately represent the true production incentive created by a policy
Buy what does it neglect?
Production cost
14
SFIT: A more nuanced approach
Contract DurationTariff Amount
FIT contract length (years)
Size of FIT contract established in year t
(Eurocents/kWh)
Electricity Price
Wholesale market price of electricity (Eurocents/kWh)
Capacity Lifetime
Lifetime of PV or wind capacity installed in year t
(years)
Generation Cost
Average lifetime cost of electricity production
(Eurocents/kWh)
for energy technology i, in country s, in year t.
15
SFIT: A more nuanced approach
Expected profit over the lifetime of capacity installed under a FIT
contract
Expected generation cost over the lifetime
of capacity
= ROI
for energy technology i, in country s, in year t.
Results of Cross-Sectional RegressionsDependent Variable: Added RES-E Capacity (ln)
Solar Photovoltaic Onshore Wind(1) (2) (3) (4)Binary FIT 0.654***(0.184) 1.011***(0.215)SFIT 1.025***(0.128) 0.412***(0.151)Binary Tax or Grant -0.109(0.186) 0.179(0.167) 0.179(0.325) -0.305(0.337)Binary Tendering Scheme -0.567**(0.239) 0.131(0.210) 0.235(0.399) 0.138(0.409)INCRQMTSHARE, ln -8.402**(3.978) -1.079(3.051) 5.154(4.745) -3.121(4.329)GDP per capita, ln 0.990**(0.450) -0.165(0.341) 3.672***(0.376) 3.847***(0.377)Area, ln 0.509***(0.101) 0.387***(0.071) 1.086***(0.094) 1.129***(0.088)Net import ratio, ln -0.314*(0.186) 0.018(0.167) 0.005(0.245) 0.002(0.262)Energy cons. per capita, ln 0.076(0.429) 0.305(0.373) -2.011***(0.510) -1.780***(0.509)Nuclear share, ln -0.322(0.524) -0.008(0.444) -0.728(0.795) -1.224(0.759)Oil share, ln -20.501(15.250) -19.261*(10.868) -22.747*(11.842) -12.115(11.626)Natural gas share, ln 1.160(1.111) 1.259(0.878) 1.760*(1.067) 1.020(1.024)Coal share, ln 0.755(0.672) 0.671(0.459) 2.614***(0.592) 2.957***(0.599)EU 2001 binary -0.121(0.226) 0.114(0.175) -0.177(0.302) -0.144(0.307)N 253 253 264 264R2 0.328 0.575 0.665 0.654
Policy Variables
Socio-Economic
Controls
Fuel Mix Variables
Feed-in tariffs appear to drive RES-E development.
Cannot be interpreted as causal because of OVB
*** <1% significance, ** <5% significance, * <10% significance
How do the results change when we control for fixed country characteristics?
17
Results of Fixed-Effects RegressionsDependent Variable: Added RES-E Capacity (ln)
Solar Photovoltaic Onshore Wind(1) (2) (3) (4)Binary FIT 0.068(0.197) 0.758***(0.280) SFIT 0.743***(0.106) 0.262*(0.156)Binary Tax or Grant -0.327(0.380) -0.411(0.342) 0.052(0.531) 0.037(0.541)Binary Tendering Scheme 0.052(0.286) -0.047(0.258) -0.946**(0.406) -1.090***(0.407)INCRQMTSHARE, ln 4.600(5.584) 1.544(5.062) -3.500(7.864) -5.754(7.928)GDP per capita, ln 0.689(0.699) -0.073(0.630) 3.187***(0.912) 2.626**(1.130)Area, ln (dropped) (dropped) (dropped) (dropped)Net import ratio, ln -0.145(0.252) -0.019(0.229) -0.117(0.350) -0.152(0.353)Energy cons. per capita, ln -1.038(1.590) -1.550(1.427) -0.809(2.137) 0.937(2.142)Nuclear share, ln -1.929(1.534) -2.517*(1.386) -0.281(2.147) 0.355(2.163)Oil share, ln 98.175***(32.774) 76.960***(29.643) 11.882(46.330) 13.754(46.867)Natural gas share, ln 4.235***(1.142) 2.391**(1.060) 2.162(1.621) 1.257(1.614)Coal share, ln -10.249***(2.477) -6.480***(2.288) 3.427(3.386) 3.518(3.511)EU 2001 binary -0.064(0.192) 0.080(0.174) -0.212(0.267) -0.220(0.270)N Yes Yes Yes YesR2 253 253 264 264*** <1% significance, ** <5% significance, * <10% significance
Coefficients on FIT variables are universally lower
Unobserved country characteristics positively bias the
pooled cross-section results
18
Results of Fixed-Effects RegressionsDependent Variable: Added RES-E Capacity (ln)
Solar Photovoltaic Onshore Wind(1) (2) (3) (4)Binary FIT 0.068(0.197) 0.758***(0.280) SFIT 0.743***(0.106) 0.262*(0.156)Binary Tax or Grant -0.327(0.380) -0.411(0.342) 0.052(0.531) 0.037(0.541)Binary Tendering Scheme 0.052(0.286) -0.047(0.258) -0.946**(0.406) -1.090***(0.407)INCRQMTSHARE, ln 4.600(5.584) 1.544(5.062) -3.500(7.864) -5.754(7.928)GDP per capita, ln 0.689(0.699) -0.073(0.630) 3.187***(0.912) 2.626**(1.130)Area, ln (dropped) (dropped) (dropped) (dropped)Net import ratio, ln -0.145(0.252) -0.019(0.229) -0.117(0.350) -0.152(0.353)Energy cons. per capita, ln -1.038(1.590) -1.550(1.427) -0.809(2.137) 0.937(2.142)Nuclear share, ln -1.929(1.534) -2.517*(1.386) -0.281(2.147) 0.355(2.163)Oil share, ln 98.175***(32.774) 76.960***(29.643) 11.882(46.330) 13.754(46.867)Natural gas share, ln 4.235***(1.142) 2.391**(1.060) 2.162(1.621) 1.257(1.614)Coal share, ln -10.249***(2.477) -6.480***(2.288) 3.427(3.386) 3.518(3.511)EU 2001 binary -0.064(0.192) 0.080(0.174) -0.212(0.267) -0.220(0.270)N Yes Yes Yes YesR2 253 253 264 264*** <1% significance, ** <5% significance, * <10% significance
For a 10 percentage point increase in ROI provided by a FIT, countries will install• 7.4% more solar PV capacity per year• 2.6% more onshore wind capacity per year
Even when innate country traits are controlled for, FIT policies
have driven RES-E development since 1998
19
Results of Fixed-Effects RegressionsDependent Variable: Added RES-E Capacity (ln)
Solar Photovoltaic Onshore Wind(1) (2) (3) (4)Binary FIT 0.068(0.197) 0.758***(0.280) SFIT 0.743***(0.106) 0.262*(0.156)Binary Tax or Grant -0.327(0.380) -0.411(0.342) 0.052(0.531) 0.037(0.541)Binary Tendering Scheme 0.052(0.286) -0.047(0.258) -0.946**(0.406) -1.090***(0.407)INCRQMTSHARE, ln 4.600(5.584) 1.544(5.062) -3.500(7.864) -5.754(7.928)GDP per capita, ln 0.689(0.699) -0.073(0.630) 3.187***(0.912) 2.626**(1.130)Area, ln (dropped) (dropped) (dropped) (dropped)Net import ratio, ln -0.145(0.252) -0.019(0.229) -0.117(0.350) -0.152(0.353)Energy cons. per capita, ln -1.038(1.590) -1.550(1.427) -0.809(2.137) 0.937(2.142)Nuclear share, ln -1.929(1.534) -2.517*(1.386) -0.281(2.147) 0.355(2.163)Oil share, ln 98.175***(32.774) 76.960***(29.643) 11.882(46.330) 13.754(46.867)Natural gas share, ln 4.235***(1.142) 2.391**(1.060) 2.162(1.621) 1.257(1.614)Coal share, ln -10.249***(2.477) -6.480***(2.288) 3.427(3.386) 3.518(3.511)EU 2001 binary -0.064(0.192) 0.080(0.174) -0.212(0.267) -0.220(0.270)N Yes Yes Yes YesR2 253 253 264 264*** <1% significance, ** <5% significance, * <10% significance
No statistically significant relationship between FIT enactment and solar PV
development once country characteristics are controlled for
Highly significant when SFIT is used instead of binary
Binary variables obscure the true relationship between FIT policies
and solar PV development
20
If you take one thing away from this paper, let it be...
FIT Variable
Fixed Effects?
Model 1: Cross-sectional Approach
Model 2: Fixed Effects Approach
Model 3: Nuanced Approach
Do FITs work?
Binary Binary SFIT
Yes
YesVariesToo Well
No Yes
Overstates effectiveness
Understates effectiveness
Just right
Nuanced indicators and smart controls are key for accuracy and consistency in energy policy analysis
21
Conclusion
Feed-in tariffs have driven solar PV and onshore wind power development in Europe since 1998.
Controlling for policy design elements and country characteristics is crucial.
Policy design matters more than the enactment of a policy alone!
22
Thank you! Questions?
Joe Indvik, ICF [email protected]
515-230-4665
Steffen Jenner, Harvard [email protected]
857-756-0361
Felix Groba, DIW [email protected]
+49-30-89789-681
Data Sources
Capacity: Eurostat and the UN Energy Statistics Database
Policy: GreenX (University of Vienna) and supplemental sources
Cost: GreenX (University of Vienna)• 2006 – 2009 actual• 2010 – 2020 projected• 1998 – 2005 linearly extrapolated