Financing home energy retrofits in France - Louis-Gaëtan Giraudet, Cyril Bourgeois, Philippe Quirion

Financing home energy retrofits in France

 Louis-Gaëtan Giraudet, Cyril Bourgeois, Philippe Quirion

 Rome – 18 October 2019

Motivation

• Energy efficiency key to meeting Paris goals
 – France scores second in EE (ACEEE 2018 Scorecard)
 – Has set multiple targets in residential buildings

• Adequacy between targets and instruments? A
 comprehensive assessment
 – Broad: multiple subsidies, taxes, building codes
 – Deep: economic, environmental, distributional performance

• Methodological approach
 – Res-IRF model: highly detailed depiction of barriers to EE
 – This exercise: careful treatment of policy interactions

 2

Stratégie nationale bas-carbone mandates: Supporting policies:

1. Reduction of energy use by 20% in 2030 1. Income tax credit
 and 50% in 2050 compared to 2012 2. Zero-interest loans
2. Yearly renovation of 500,000 dwellings 3. Reduced VAT
3. Elimination of EPC labels F et G by 2025 4. Carbon tax
4. Performance label B or higher 5. White certificates
 widespread by 2050 6. Building codes
5. Fuel poverty alleviation by 15% in 2020 + others

 Effectiveness to targets?
 Policy efficiency and distributional impacts?

 3

Res-IRF TECHNICAL PARAMETERS
Giraudet et al., En J, 2011
Giraudet et al., En Econ, 2012 Renovation and construction costs
Branger et al., Env Mod Soft, 2015 Demolition rates

 INPUTS OUTPUTS
 Population
 Renovation and constr.
 +0.3% p.a.
 (ext./int. margins)
 Household income
 +1.2% p.a. Resulting consumption
 for elec, ngas, oil, wood
 Fuel prices
 ~ +1.5% p.a. Heating comfort

 Main extension
 in version 3.0 Landlord-tenant dilemma
 (based on Barriers to decision-making in collective housing
 Phébus survey) Non-energy costs
 Credit constraints
 BEHAVIORAL PARAMETERS 4

Policy parameterization
 Reference variant Tighter variant
CITE 17% ad valorem subsidy, uniform rate Restricted to high performance
EPTZ ~9% ad valorem subsidy, restricted to HP Higher rate ~23%
CEE Non-uniform subsidy, equivalent to an Subsidy and tax components x3
 average ad valorem rate 5%
 + energy tax
Taxe C Carbon tax, myopically expectated Perfect expectation
TVA r Subsidy, VAT rate of 5,5% instead of 10%
RT 2020 Building code mandating BEPOS level in 2020

4 scenarios
• All policies (TP) ~ reference

• No policy (ZP)
• All policies in their tighter variant (TP+) counterfactuals
• All policies, no land./ten. dilemma (TP sans DPL)
 5

Target 1: Energy use

 Feasible…with tight policies maintained until 2050!
 2/3 are autonomous improvements (energy prices, building codes, etc.)
 6

Target 2: Yearly renovations

 Ext. margin : +115,000

 Int. margin: increase expenditures

 Easily reached – at odds with Hulot’s resignation statement ?!?!
 Note the definition: renovation = upgrade by at least one EPC label
 Estimate in line with Ademe’s latest TREMI survey (2018)
 7

Targets 3 & 4: Dwelling stock

 -75% in 2025. 50% to 70% in 2050
Target met in 2040 if landlord-tenant
 dilemma is overcome.
 8

Objectif 5: Fuel poverty

 Energy-to-income ratio: heating conventional expend. >10% income
 Natural decline, despite structural increase ~0.6% p.a. (=0.3%+1.5%-1.2%)
 Carbon tax has a retarding effect, subsidies accelerating
 9

Summary

Target Fulft Comment
 • Non-specific to the residential sector
1 Reduction of energy use by
 20% in 2030 and 50% in 2050 ? • Requires tight policies maintained until
 2050
 • Progress largely autonomous
 • Largely fulfilled in private housing
2 Yearly renovation of 500,000
 dwellings, incl. 120,000 in vx • Largely missed in social housing
 social housing • The definition matters!
 • Important progress, -75% en 2025
3 Elimination of labels F and G
 by 2025 x • Target fulfilled in 2040 if landlord-tenant
 dilemma overcome
 • 50% to 70% at best with tight policies
4 Label B or higher widespread
 by 2050 x
 • Fulfilled only with tightest policies
5 Fuel poverty alleviation by
 15% in 2020 ?

 10

Simulations vs. Observations, 2016

 EPTZ over-estimated by one order of magnitude!
 Unaccounted for barriers on the demand and supply sides?
 11

Long-term costs

 12

Policy effectiveness
Considering all possible interactions among policies:

  Carbon tax plays on investment + utilization
  CITE is the most effective of all subsidies
 13

Leverage, 2015

 1

 Subsidies have leverage ≥ 1, declining over time
 Interactions are mostly over-additive, due to model non-linearities
 14

CITE variants

Leverage increases when…
 Ad valorem rate is reduced
 Eligibility is restricted to the most significant upgrades
 Eligibility is restricted to the first two income quintiles
 15

Conclusion

• Key insights
 – Target fulfillment requires tight policies, extended to
 rented dwellings and maintained until 2050
 – If budget constraints were to bind, restricting
 eligibility to low-income households would be a nice
 opportunity to reconcile efficiency and equity
 – The 500,000 target needs to be properly defined!

• Contribution
 – Unique integrated assessment framework
 – Simulation/observation gap reveals barriers to EPTZ
 – Original approach to addressing policy interactions

 16

Renovation costs

€/m² Etiquette finale
 F E D C B A

Etiquette G 76 136 201 271 351 442 Decreasing returns
initiale +
 F 63 130 204 287 382
 Economies of scale
 E 70 146 232 331
 D 79 169 271
 C 93 199
 B 110

 17

Income Distribution Across Dwellings

Source: Phébus survey

 Income category Single-family Multi-family Social housing
 housing housing
 C1 (lowest income) 15% 37% 4%
 C2 10% 25% 4%
 C3 7% 15% 4%
 C4 5% 7% 4%
 C5 (highest income) 4% 5% 4%
 Wgd avg discount rate 8% 17% 4%
 18

Landlord/Tenant Distribution in Stock & Flow

 Type of dwelling Renovation rate
 Owner-occupied Single-family 4,7%
 Multi-family 3,6%
 Rented Single-family 2,0%
 Multi-family 1,8%
 Social housing Single-family 1,5%
 Multi-family 2,0%
 19

Intensity of Utilization
 Source: Cayla and Osso (2013)
Intensity of utilization

 Energy efficiency, Carbon tax
 Income

 Theoretical budget share dedicated to heating

 = ~ 
 
 20

Elasticités-prix de la demande d’énergie

 Court terme: -0,2

 Estimé sur la courbe
d’intensité d’utilisation

 Long terme: -0,4

 Estimé sur 160 scénarios
d’évolution du prix des énergies

 21

Simulations rétrospectives
Sur la base de 1920 simulations,
erreur moyenne de 3,7% par rapport aux données CEREN

 22

Variables d’entrée

• Population: +0,3%/an (projection INSEE)
• Revenu des ménages: +1,2%/an (poursuite tendance)
• Prix des énergies: scénario ADEME/DGEC/CE, ~+1,5%/an

 23

Profil de la taxe carbone
Scénarios TC TC+
Mode d’anticipation myope parfaite

 +6%/an +4%/an

 Contenus carbone (gCO2/kWh PCI) décroît de 1%/an à partir de 2021,
 reflétant une pénétration de gaz
 Gaz naturel 206
 décarboné de 10% en 2030, 18% en
 Fioul domestique 271 2040 et 26% en 2050
 24

Barèmes des aides
 F E D C B A
G 11% 10% 8% 7% 6% 5% Taux de subvention des
F 8% 6% 5% 4% 3% CEE à 4€/MWh cumac
E 5% 4% 3% 2%
D 3% 2% 2%
C 2% 1%
B 1%

 F E D C B A
 G 0 0 1 1 1 1 Ciblage de l’EPTZ
 F 0 1 1 1 1
 E 1 1 1 1
 D 0 1 1
 C 1 1
 B 1 25

Scénario d’obligation CEE

 26

Evolutions par énergies

 27

Rénovations par segments

 28

Rénovations énergétiques

 29

Dépenses de rénovation

 30

Intensité d’utilisation

 31

Coût-efficacité

 32

Certificats d’économies d’énergie

 33

Certificats d’économies d’énergie

 34