ArticlePDF Available

The solar influencer next door: Predicting low income solar referrals and leads

Authors:
  • Veritone

Abstract and Figures

Increasing the adoption of solar among low-to-moderate income (LMI) households remains an important policy goal because of its promise to simultaneously reduce energy burden and support the just distribution of benefits of renewable energy. However, scaling LMI solar remains challenging due to affordability and access issues. Most existing LMI adoption has occurred under public-funded programs, highlighting the importance of increasing the cost-effectiveness of these programs at scale. We develop a new household-level data set on LMI solar lead acquisition, referrals, and adoption to understand the processes through which LMI solar uptake has occurred in California. Then, we develop models to predict two sub-mechanisms in the solar adoption process: whether an otherwise qualified lead becomes “lost” i.e. non-responsive to outreach and, for existing clients, whether they refer solar to others. For the program analyzed, participants received their solar system at no cost, which deemphasizes economic drivers of solar adoption and could differ from other program experiences. Both models substantially improved the accuracy of prediction relative to a baseline. Overall, we find that peer effects and solar economics are important to predicting referrals, and household demographic factors in lead loss prediction. Finally, we find that referrals are both the highest quality and largest source of LMI solar leads, providing a promising mechanism to expand LMI programs further.
Content may be subject to copyright.
UNCORRECTED PROOF
"G>K@R.>L>:K<A/H<B:E/<B>G<> QQQ QQQQ 
HGM>GML EBLML :O:BE:;E> :M /<B>G<>!BK><M
"G>K@R.>L>:K<A/H<B:E/<B>G<>
CHNKG:EAHF>I:@> PPP>EL>OB>K<HFEH<:M>>KLL
0A> LHE:K BGXN>G<>K G>QM =HHK -K>=B<MBG@ EHP BG<HF> LHE:K K>?>KK:EL :G=
E>:=L
>G /B@KBG LAHD />D:K "FF: 0HF>
National Renewable Energy Laboratory, 15013 Denver W Pkwy, Golden, CO 80401, USA
.0& )"&+#,
Keywords:
)HP BG<HF> LHE:K
-K>=B<MBO> FH=>EL
/HE:K K>?>KK:EL
/HE:K E>:=L
!>IEHRF>GM FH=>EL
NLMHF>K :<JNBLBMBHG <HLML
"G>K@R CNLMB<>
/0. 0
&G<K>:LBG@ MA> :=HIMBHG H? LHE:K :FHG@ EHPMHFH=>K:M> BG<HF> )*& AHNL>AHE=L K>F:BGL :G BFIHKM:GM IHEB<R
@H:E ;><:NL> H? BML IKHFBL> MH LBFNEM:G>HNLER K>=N<> >G>K@R ;NK=>G :G= LNIIHKM MA> CNLM =BLMKB;NMBHG H? ;>G>WML
H? K>G>P:;E> >G>K@R %HP>O>K L<:EBG@ )*& LHE:K K>F:BGL <A:EE>G@BG@ =N> MH :??HK=:;BEBMR :G= :<<>LL BLLN>L *HLM
>QBLMBG@ )*& :=HIMBHG A:L H<<NKK>= NG=>K IN;EB<?NG=>= IKH@K:FL AB@AEB@AMBG@ MA> BFIHKM:G<> H? BG<K>:LBG@ MA>
<HLM>??><MBO>G>LL H? MA>L> IKH@K:FL :M L<:E> 3> =>O>EHI : G>P AHNL>AHE=E>O>E =:M: L>M HG )*& LHE:K E>:= :<JNB
LBMBHG K>?>KK:EL :G= :=HIMBHG MH NG=>KLM:G= MA> IKH<>LL>L MAKHN@A PAB<A )*& LHE:K NIM:D> A:L H<<NKK>= BG :EB
?HKGB: 0A>G P> =>O>EHI FH=>EL MH IK>=B<M MPH LN;F><A:GBLFL BG MA> LHE:K :=HIMBHG IKH<>LL PA>MA>K :G HMA>K
PBL> JN:EBW>= E>:= ;><HF>L EHLMB> GHGK>LIHGLBO> MH HNMK>:<A :G= ?HK >QBLMBG@ <EB>GML PA>MA>K MA>R K>?>K LH
E:K MH HMA>KL #HK MA> IKH@K:F :G:ERS>= I:KMB<BI:GML K><>BO>= MA>BK LHE:K LRLM>F :M GH <HLM PAB<A =>>FIA:LBS>L
><HGHFB< =KBO>KL H? LHE:K :=HIMBHG :G= <HNE= =B??>K ?KHF HMA>K IKH@K:F >QI>KB>G<>L HMA FH=>EL LN;LM:GMB:EER
BFIKHO>= MA> :<<NK:<R H? IK>=B<MBHG K>E:MBO> MH : ;:L>EBG> ,O>K:EE P> WG= MA:M I>>K >??><ML :G= LHE:K ><HGHFB<L
:K> BFIHKM:GM MH IK>=B<MBG@ K>?>KK:EL :G= AHNL>AHE= =>FH@K:IAB< ?:<MHKL BG E>:= EHLL IK>=B<MBHG #BG:EER P> WG=
MA:M K>?>KK:EL :K> ;HMA MA> AB@A>LM JN:EBMR :G= E:K@>LM LHNK<> H? )*& LHE:K E>:=L IKHOB=BG@ : IKHFBLBG@ F><A:GBLF
MH >QI:G= )*& IKH@K:FL ?NKMA>K
1. Introduction
1.1. Importance of LMI solar programs
"G<HNK:@BG@ >JNBMR BG MA> :=HIMBHG H? =BLMKB;NM>= IAHMHOHEM:B<L
I:KMB<NE:KER :FHG@ EHPMHFH=>K:M> BG<HF> )*& AHNL>AHE=L BL :
@KHPBG@ IHEB<R <HG<>KG 78 )*& LHE:K :=HIMBHG BGM>KL><ML PBMA <KHLL
<NMMBG@ IHEB<R H;C><MBO>L BG<EN=BG@ K>=N<BG@ >G>K@R ;NK=>G INKLNBG@
>G>K@R CNLMB<> OB: >JNBM:;E> :<<>LL MH K>G>P:;E> >G>K@R :G= K>:EBSBG@
=><:K;HGBS:MBHG @H:EL %NG=K>=L H? @B@:P:MML $3 H? IARLB<:E IHM>G
MB:E >QBLM HG )*& KHH?L BG MA> 1GBM>= /M:M>L >JNBO:E>GM MH H??L>MMBG@ 
H? )*& :GGN:E >E><MKB<BMR <HGLNFIMBHG 78 %HP>O>K F>:GBG@?NE I:MA
P:RL MH :<AB>OBG@ L<:E>NI K>F:BG >ENLBO> 78 -KH@K:FL MH >G<HNK:@>
:=HIMBHG H? KHH?MHI LHE:K :FHG@ )*& AHNL>AHE=L A:O> MH =:M> ;>>G <:I
BM:EBGM>GLBO> PBMA : AB@A ?K:<MBHG H? MA> LRLM>F LN;LB=BS>= :EB?HKGB:
MA> G:MBHGL E:K@>LM )*& F:KD>M IKHOB=>L >OB=>G<> ?HK BG?HKFBG@ IHEB<R
=>LB@G :G= >G<HNK:@BG@ )*& =>IEHRF>GM FHK> ;KH:=ER
!>LIBM> BML ;>G>WML :=HIMBHG H? LHE:K K>F:BGL =><BLBO>ER LD>P>= MH
P:K= FB==E> :G= NII>KBG<HF> AHNL>AHE=L 78 0ABL =BLMKB;NMBHG:E
=B??>K>G<> <:NL>L <HG<>KG ;><:NL>  M:Q K>O>GN> BL K>=BLMKB;NM>= MH
AB@ABG<HF> AHNL>AHE=L =>OB:MBG@ ?KHF IKBG<BIE>L H? ?:BK M:Q:MBHG 
LHE:K :=HIMBHG ;R AB@ABG<HF> AHNL>AHE=L F:R E>:= MH K>M:BE IKB<>L BG
<K>:L>L BFI:<MBG@ IHHK>K AHNL>AHE=L PAH :EK>:=R I:R : E:K@>K LA:K> H?
MA>BK BG<HF> MHP:K= >G>K@R :G=  MABL IKH@K:F <HGMKB;NM>L MH :
PB=>GBG@ BG>JN:EBMR @:I BG MA> 1GBM>= /M:M>L 78
1.2. Challenges to LMI deployment
*NEMBIE> ;:KKB>KL <A:EE>G@> PB=>LIK>:= )*& LHE:K I>G>MK:MBHG /M:KM
BG@ PBMA MA> <E:LLB< LIEBM BG<>GMBO> IKH;E>F 78 : E:K@> ?K:<MBHG H? )*&
AHNL>AHE=L EBO> BG FNEMB?:FBER :G=HK K>GM:E ;NBE=BG@L PAB<A =BL<HNK
)*& )HPMHFH=>K:M> &G<HF> $3 $B@:P:MML //% /BG@E> #:FBER ??HK=:;E> /HE:K %HF>L IKH@K:F !  !BL:=O:GM:@>= HFFNGBMR !,& !BYNLBHG H?
&GGHO:MBHGL 0- 0A>HKR H? -E:GG>= >A:OBHK 2+ 2:EN>>EB>?+HKF 4$ "QMK>F> $K:=B>GM HHLMBG@ .* NLMHF>K .>E:MBHGLABI *:G:@>F>GM =:M:;:L> %1!
%HNLBG@ :G= 1K;:G !>O>EHIF>GM ."-)&  .HH?MHI "G>K@R -HM>GMB:E H? )HP &G<HF> HFFNGBMB>L BG F>KB<: =:M:L>M D3A DBEHP:MMAHNK ,)/ ,K=BG:KR )>:LM
/JN:K>L /%- /A:IE>R ==BMBO> "QIE:G:MBHG ),"// )H<:EER "LMBF:M>= /<:MM>KIEHM /FHHMABG@ -$" -:<BW< $:L "E><MKB< NMBEBMR -* -:KMB<NE:M> *:MM>K %%
%HNL>AHE=
HKK>LIHG=BG@ :NMAHK
E-mail addresses: >GC:FBG/B@KBG+.")@HO  /B@KBG LAHD/>D:K+.")@HO  />D:K "FF:0HF>+.")@HO " 0HF>
AMMIL=HBHK@C>KLL
.><>BO>=  IKBE  .><>BO>= BG K>OBL>= ?HKF />IM>F;>K  <<>IM>=  +HO>F;>K 
V
+HM>)HPK>LHENMBHGBF:@>LP>K>NL>=MH<K>:M>MABL-!#0A>HKB@BG:EBF:@>LPBEE;>NL>=BGMA>WG:E<HFIHLBMBHG
UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
:@>L :=HIMBHG :KKB>KL K>F:BG ?HK LBG@E>?:FBER HPG>KH<<NIB>= AHNL>
AHE=L K:G@BG@ ?KHF :??HK=:;BEBMR LMKN<MNK:E LNBM:;BEBMR ?HK LHE:K :;BEBMR
MH L><NK> WG:G<BG@ :G= BG:;BEBMR MH FHG>MBS> ;>G>WML MAKHN@A MA> M:Q
<H=> 78 >A:OBHK:E ;:KKB>KL BG<EN=> BG?HKF:MBHG :LRFF>MKR L>:K<A
<HLM :G= =BLMKNLM H? HNMLB=>KL 78 &GMKB@NBG@ER 3HELD> 78 LN@
@>LML MA:M LH<B:E BGXN>G<> BL FHK> K>E>O:GM ?HK )*& LHE:K :=HIM>KL MA:G
AB@ABG<HF> HG>L
"QBLMBG@ )*& =>IEHRF>GM A:L E:K@>ER H<<NKK>= OB: IN;EB< IKH@K:FL
>RHG= :??HK=:;BEBMR ;NLBG>LL FH=>E ;:KKB>KL :K> <>GM>K>= :KHNG= <NL
MHF>K :<JNBLBMBHG <A:EE>G@>L )*& AHNL>AHE=L :K> : GHMHKBHNLER A:K= MH
K>:<A IHINE:MBHG =N> MH MBF> L<:K<BMR E:G@N:@> BLLN>L :G= BGM>KG>M :G=
IAHG> :<<>LL 78 *:GR )*& LHE:K IKH@K:FL :K> HI>K:M>= ;R :
MABK=I:KMR BGM>KF>=B:KR HG MA> @HO>KGF>GML ;>A:E? 78 0ABL BGM>K
F>=B:KR H<<NIB>L : GHGIKH?BM KHE> :G= BL K>LIHGLB;E> ?HK @>G>K:MBG@ IKH
@K:F E>:=L LBM> :LL>LLF>GM :G= BGLM:EEBG@ LRLM>FL $.&! EM>KG:MBO>L
A:L :=FBGBLM>K>= :EB?HKGB:L LM:M>PB=> IKH@K:FL ?HK HO>K : =><:=>
PBMA G>:KER  LBG@E>?:FBER )*& LRLM>FL =>IEHR>= ?KHF  MH
 B> /BG@E>?:FBER ??HK=:;E> /HE:K %HF>L //% :G= MA> !BL:=
O:GM:@>= HFFNGBMB>L ! //% IKH@K:FL 78 +HM:;ER BG :EB
?HKGB: MA> LM:M> IKH@K:F LN;LB=B>L <HO>K MA> >GMBK> LRLM>F <HLM 78
/NKIKBLBG@ER :G= =>LIBM> MA> LRLM>F LN;LB=BS:MBHG MA> <HLM H? B=>G
MB?RBG@ :G= <HGO>KMBG@ JN:EBW>= )*& IKH@K:F I:KMB<BI:GML B> E>:=L
BL GHGMKBOB:E >LMBF:M>= MH <HLM  I>K BGLM:EE 78 HLML H? <NL
MHF>K :<JNBLBMBHG :K> I:KM H? MA> HO>K:EE IKH@K:F ;N=@>M MA>K>;R FHMB
O:MBG@ K>=N<BG@ <NLMHF>K :<JNBLBMBHG <HLML )*& LHE:K :=HIMBHG L>>FL MH
IK>L>GM : NGBJN> IKH;E>F H? AHP MH <HLM>??><MBO>ER M:K@>M MABL A:K=MH
K>:<A IHINE:MBHG !:M:=KBO>G E>:= :<JNBLBMBHG F>MAH=L FB@AM ;> NL>=
MH EHP>K <NLMHF>K :<JNBLBMBHG <HLML ;R ;>MM>K IK>=B<MBG@ AHNL>AHE=L
MA:M :K> NGEBD>ER MH I:KMB<BI:M> BG MA> IKH@K:F B> : EHLM E>:= MAHL>
MA:M FB@AM IKHOB=> K>?>KK:EL :L P>EE :L BG<HKIHK:MBG@ LI:MB:E :G= =>FH
@K:IAB< <KBM>KB: ?HK IKH@K:F I:KMB<BI:MBHG
1.3. Referrals as a key to LMI adoption
.>?>KK:EL :K> <NKK>GMER MA> EHP>LM<HLM F>MAH= H? @>G>K:MBG@ G>P LH
E:K <EB>GML O:EN:;E> >GHN@A MA:M K>?>KK>KL :K> H?M>G I:B= L>O>K:E ANG
=K>= =HEE:KL I>K LN<<>LL?NE BGLM:EE 78 .>?>KK:EL <HGLMBMNM> 
H? :EE F:KD>MK:M> LHE:K <EB>GML 78 :G= :EFHLM :EEH? MA> LHE:K BG
LM:EE>KL BGM>KOB>P>= ;R *H>SSB >M :E  78 K>IHKM>= MA>R :LD>= HK
BG<>GMBOBS>= <EB>GML MH IKHOB=> K>?>KK:EL 1GEBD> F>=B: :=O>KMBL>F>GML
K>?>KK:EL :L : E>:= LHNK<> <:GGHM LBFIER ;> F:GN?:<MNK>= :G= MANL :K>
<A:EE>G@BG@ MH PB=>ER L<:E> ?HK OHENF> HK BGMH G>P @>H@K:IAB>L
!>LIBM> MA>BK >OB=>GM BFIHKM:G<> MH MA> IKH<>LL H? L>EEBG@ LHE:K EBMME>
>QBLMBG@ K>L>:K<A A:L =BK><MER >Q:FBG>= MA> K>?>KK:E :G= EHLM E>:= IA>
GHF>G: )BM>K:MNK> MA:M L>>DL MH >QIE:BG :G= IK>=B<M LHE:K :=HIMBHG BL
>QM>GLBO> >@ 78 MAHN@A :EFHLM >Q<ENLBO>ER ?H<NL>= HG >Q:F
BGBG@ MA> :=HIMBHG =><BLBHG :G= GHM K>?>KK:EL I>KLBLM>GM >E>F>GM BG
MA> :=HIMBHG EBM>K:MNK> BL =>M>KFBGBG@ MA> O:EN> H? BG?HKF:MBHGPAB<A
MRI>L H? =:M: F:MM>K ?HK IK>=B<MBG@ >G>K@R M><AGHEH@R =><BLBHGL 7
82BLB;E>:MMKB;NM>L LN<A :L ;NBE=BG@ LBS> HK :O>K:@> G>B@A;HKAHH=
BG<HF> :K> LMK:B@AM?HKP:K= MH =>M>KFBG> ;NM =BL<HNGM I>KLHG:E =KBO>KL
BG ;>A:OBHK &GOBLB;E>:MMKB;NM>L BG<EN=BG@ :MMBMN=>L ;>EB>?L :G= LH<B:E
<N>L :K> LB@GB?B<:GM BG IK>=B<MBG@ BGM>K>LM BG LHE:K 78 ;NM P>EE
NG=>KLM:G= K>E:MBHGLABIL ;>MP>>G <HO:KB:M>L 78 :K> =B?W<NEM MH :LL>LL
PBMAHNM LNKO>RBG@ I:KMB<BI:GM H;L>KO:MBHG HK BGM>KOB>PL
,G> K>:LHG PAR K>?>KK:EL :K> >??><MBO> BG =KBOBG@ M><AGHEH@R :=HI
MBHG BL ;><:NL> MA>R =K:P HG I>>K BGXN>G<> 78 =HIMBG@ G>P
M><AGHEH@B>L BL KBLDR PBMA NG<>KM:BG K>E:MBO> ;>G>WML .>ERBG@ HG MA>
:=OB<> :G= >QI>KB>G<> H? I>>KL A>EIL MH =>KBLD: M><AGHEH@R 1GLNK
IKBLBG@ER LMN=B>L A:O> K>I>:M>=ER LAHPG MA:M I>>K BGXN>G<> ;HMA LI:
MB:E :G= LH<B:E BG<K>:L>L EBD>EBAHH= MH :=HIM LHE:K 78 *HK>
HI:JN> MH MK:=BMBHG:E =BLM:G<>;:L>= I>>K >??><ML F>:LNK>L AHP>O>K
:K> MA> G:MNK> H? MA> LH<B:E MB>L MA:M MA>R BGM>G= MH IKHQR 0K:=BMBHG:E
I>>K >??><ML F>MAH=L :K> ><HGHF>MKB< :G= =H GHM FH=>E MA> =R:= ;>
MP>>G K>?>KK>K :G= K>?>K>> 0ANL PABE> I>>K >??><ML LNK>ER =KBO> LHE:K
:=HIMBHG MH PA:M =>@K>> :K> MA>R : IKHQR ?HK MA> NG=>KERBG@ K>?>KK:E
F><A:GBLF
1.4. Framework for predicting referrals and lost leads
3> LM:KM PBMA MA> IK:<MBMBHG>KL I>KLI><MBO> ?HKFNE:M>= :L L>>DBG@
MH F:QBFBS> MA> GNF;>K H? <EB>GM BGLM:EE:MBHGL :M E>:LM <HLM LN;C><M MH
IKH@K:F =>LB@G :G= >JNBMR <HGLB=>K:MBHGL 1LBG@ MA> F>M:IAHK H? :
<EB>GM ?NGG>E MA> <EB>GM HNMK>:<A IKH<>LL BL : L>KB>L H? IKH@K>LLBO> :G=
PBGGHPBG@ LM>IL MH B @:MA>K E>:=L MA:M :K> IKHLI><MBO> LHE:K <EB>GML BB
L<K>>G MA> E>:=L ?HK MA>BK LNBM:;BEBMR ?HK IKH@K:F <KBM>KB: :G= LHE:K :=HI
MBHG BBB BGLM:EE LHE:K ?HK MA> K>F:BGBG@ E>:=L MA:M F>>M :EE L<K>>GBG@ <KB
M>KB: 0PH LN;@H:EL H? MA> IKH<>LL :K> MH BG<K>:L> MA> GNF;>K H? E>:=L
>GM>KBG@ MA> ?NGG>E >LI><B:EER MAHL> MA:M PBEE F>>M L<K>>GBG@ <KBM>KB:
:G= FBGBFBS> K>LHNK<>L G>>=>= MH L<K>>G E>:=L &=>GMB?RBG@ E>:=L EBD>ER
MH =KHI HNM H? MA> IKH<>LL =N> MH E:<D H? BGM>K>LM K>=N<>L MA> K>LHNK<>L
G>>=>= MH L<K>>G E>:=L .>?>KK:EL M:K@>M ;HMA @H:ELA:OBG@ <EB>GML @>G
>K:M> K>?>KK:EL MH MA>BK :<JN:BGM:G<>L <K>:M>L G>P E>:=L :G= E>:=L @>G
>K:M>= MAKHN@A K>?>KK:EL :K> FHK> EBD>ER MH K>LNEML BG :=HIMBHG LBG<> MA>
K>?>KK>K <:G <HGO>R IKH@K:F <KBM>KB: MH MA> K>?>K>> :G= IKHFHM> MKNLM BG
MA> LHE:K IKH@K:F
%HP>O>K MA> K>?>KK:E :G= EHLM E>:= IA>GHF>G: <:G :ELH ;> NG=>K
LMHH= PBMABG M><AGHEH@R =B??NLBHG ?K:F>PHKDL >LI><B:EER !B??NLBHG H?
&GGHO:MBHGL !,& 3BMABG MA> !,& ?K:F>PHKD LH<B:E <HGM:@BHG BL MA>
F:BG ?HK<> IKHI:@:MBG@ =B??NLBHG MAKHN@A MA> LRLM>F :G= K>?>KK:EL BML
F:GB?>LM:MBHG 0><AGHEH@R :=HIMBHG BG !,& BL :ELH =KBO>G ;R F:O>GL HK
HIBGBHG E>:=>KL MA:M :K> BGXN>GMB:E BG LIK>:=BG@ BG?HKF:MBHG :;HNM :G
BGGHO:MBHG MH I>>KL B> I>>K K>?>KK:EL !,& :ELH AB@AEB@AML MA> BFIHK
M:G<> H? trialability PA>K> G>P :=HIM>KL <:G OB<:KBHNLER MKR HNM: G>P
M><AGHEH@R ;>?HK> :=HIMBG@ BM ,G> P:R H? MKRBG@ HNM LHE:K BL ;R =BL
<NLLBG@ BM PBMA : I>>K MA:M A:L :EK>:=R :=HIM>= LHE:K I>KA:IL E>:=BG@ MH
: K>?>KK:E 0A> EHLM E>:= IA>GHF>G: K>E:M>L ;KH:=ER MH =BLBGM>K>LM BG LH
E:K :G= I>KLN:LBHG H? BML ;>G>WML %>K> ?K:F>PHKDL EBD> MA> 0A>HKR H?
-E:GG>= >A:OBHK 0- :G= 2:EN> :L>= +HKFL 2+ :K> BGLMKN<MBO>
#HK BGLM:G<> PBMABG 0- EHLM E>:=L :K> BGXN>G<>= ;R MA>BK :MMBMN=>L MH
P:K=L LHE:K :G= I>K<>BO>= LH<B:E IK>LLNK> ?HK :=HIMBG@ 78 /BFBE:KER
O:EN>L ;>EB>?L :G= GHKFL 2+ :;HNM LHE:K :K> :EE <HGC><MNK>= MH BG
XN>G<> MA> =><BLBHG MH :=HIM LHE:K 78
0A>L> ?K:F>PHKDL @NB=> MA> ?:<MHKL MA:M P> <HGC><MNK> <HNE= IK>=B<M
K>?>KK:EL :G= EHLM E>:=L -KBG<BI:EER P> >QI><M MA:M G>:K;R HK K><>GM LH
E:K BGLM:EEL PBEE >Q>KM I>>K >??><ML MA>K>;R BG<K>:LBG@ IKH;:;BEBMR H? IKH
OB=BG@ : K>?>KK:E :G= K>=N<> EBD>EBAHH= H? : EHLM E>:= ->>K >??><ML :K>
DGHPG MH H<<NK LH<B:EER :G= LI:MB:EER 78 LH MA>K> <HNE= ;> <HFI>MBG@
BGXN>G<> ?KHF ;HMA >QBLMBG@ )*& :=HIM>KL :L P>EE :L F:KD>MK:M> LHE:K
BGLM:EE:MBHGL />O>K:E LMN=B>L A:O> =>FHGLMK:M>= MA:M ><HGHFB< ;>G>WML
H? LHE:K MH MA> AHNL>AHE= :K> IHLBMBO>ER <HKK>E:M>= PBMA :=HIMBHG 7
8 =K:PBG@ HG <HG<>IML H? I>K<>BO>= ;>A:OBHK:E <HGMKHE BG 0-
:G= MA> M><AGHEH@RL K>E:MBO> :=O:GM:@> BG !,& 0AHN@A ><HGHFB< BG
XN>G<> BL EBD>ER ;ENGM>= BG MABL =:M:L>M =N> MH MA> LN;LB=BS>= LRLM>F
<HLM O:KB:;E>L LN<A :L :FHNGM H? >E><MKB< ;BEE K>=N<MBHG LBS> H? MA> -2
LRLM>F :G= M><AGB<:E LNBM:;BEBMR ?HK LHE:K LAHNE= IHLBMBO>ER BGXN>G<>
IKH@K:F I:KMB<BI:MBHG :G= IKHOB=BG@ : K>?>KK:E
/HE:K BL H?M>G ?K:F>= :L : <HNGM>KP>B@AM MH >G>K@R :G= >GOBKHGF>G
M:E <HG<>KGL :G= IKH>GOBKHGF>GM:E GHKFL :K> DGHPG MH BG<K>:L> BGM>K
>LM BG LHE:K 78 0ANL P> >QI><M MA:M >GOBKHGF>GM:E ?:<MHKL LN<A :L >Q
IHLNK> MH :BK :G= P:M>K IHEENMBHG PBEE BG<K>:L> BGM>K>LM BG LHE:K #BG:EER
LH<BH><HGHFB< BLHE:MBHG LN<A :L E:<D H? :<<>LL MH BGM>KG>M IHO>KMR HK
EBG@NBLMB< BLHE:MBHG A:O> :ELH ;>>G LAHPG MH =><K>:L> LHE:K :=HIMBHG >L
I><B:EER BG EHPBG<HF> <HFFNGBMB>L 78 ,O>K:K<ABG@ :EE P> :K> BGM>K
>LM>= BG MA> =>@K>> MH PAB<A G>B@A;HKAHH=E>O>E ?:<MHKL <:G LN<<>LL
?NEER IK>=B<M K>?>KK:EL :G= EHLM E>:=L K>E:MBO> MH BG=BOB=N:EE>O>E ?:<MHKL
UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
2. Material and methods
3> =>O>EHI : G>P AHNL>AHE=E>O>E =:M: L>M HG )*& LHE:K E>:= :<JNB
LBMBHG K>?>KK:EL :G= :=HIMBHG MH NG=>KLM:G= MA> IKH<>LL>L MAKHN@A
PAB<A )*& LHE:K A:L H<<NKK>= BG :EB?HKGB: 0A>G P> =>O>EHI @K:=B>GM
;HHLM>= .:G=HF #HK>LM FH=>EL 4$ 78 MH IK>=B<M : <EB>GML E>:= LM:
MNL :G= MA>BK EBD>EBAHH= MH K>?>K LHE:K MH LHF>HG> BG MA>BK LH<B:E G>MPHKD
>@ :<JN:BGM:G<> G>B@A;HK K>E:MBO> ?KB>G= <HPHKD>K 0A> LMN=R BL
@KHNG=>= BG MPH IK:<MB<:E FHMBO:MBHGL B PA:M <:G P> E>:KG ?KHF :EB
?HKGB:L )*& LHE:K IKH@K:F BFIE>F>GM:MBHG MH BG?HKF >QI:G=>= )*& =>
IEHRF>GM BG HMA>K @>H@K:IAB>L BB <:G IK>=B<MBO> FH=>EL ;> IK:<MB
<:EER NL>= MH BFIKHO> )*& IKH@K:F HI>K:MBHGL
2.1. Data
,NK AHNL>AHE=E>O>E =:M: L>M BL =>O>EHI>= ?KHF : <HF;BG:MBHG H?
L>O>K:E IKHIKB>M:KR :G= IN;EB< =:M: L>ML 0A> <HK> =:M: L>M BL : <NLMHF>K
K>E:MBHGLABI F:G:@>F>GM  .* =:M:;:L> <HFIKBLBG@ :EE E>:=L <EB>GML
:G= LRLM>F BGLM:EEL $.&! EM>KG:MBO>L BGM>K:<M>= PBMA ?KHF  MH
 BG :EB?HKGB: 0A> .* =:M:;:L> :ELH :EEHPL NL MH <HGLMKN<M MA>
=:M> H? <K>:MBHG :G= LHNK<> H? >:<A E>:= :G= MA>BK HNM<HF>L B> BGLM:EE
HK GHM #HK E>:=L MA:M <HGO>KM MH <EB>GML :G= IKHOB=> K>?>KK:EL MA> =:M:
;:L> IKHOB=>L =:M> EH<:MBHGL H? K>?>KK>K :G= K>?>K>> :G= HMA>K >E>F>GML
H? MA> =R:= :L>= HG MA>BK EH<:MBHGL :G= =:M>L :==BMBHG:E BG?HKF:MBHG
:;HNM MA> E>:=L :G= <EB>GML :K> :O:BE:;E> :M MA> MK:<ME>O>E BG<EN=BG@ =>
FH@K:IAB<L LH<BH><HGHFB< ?:<MHKL >GOBKHGF>GM:E BG=B<:MHKL :G=
IKHQBFBMR MH HMA>K LHE:K BGLM:EE:MBHGL
0A> =:M: L>M H??>KL : NGBJN>ER <HFIK>A>GLBO> OB>P H? EHPBG<HF>
LHE:K :=HIM>= BG :EB?HKGB: MH =:M> K>IK>L>GMBG@  H? :EE )*& LRLM>FL
BGLM:EE>= BG :EB?HKGB: MAKHN@A  BG<EN=BG@  H? :EB?HKGB: /HE:K
&GBMB:MBO>L //% IKH@K:F BGLM:EE:MBHGL :G=  H? $K>>GAHNL> $:L .>
=N<MBHG #NG= IKH@K:F BGLM:EE:MBHGL 0PH =BLMBG<M :LI><ML H? MA> =:M:L>M
:K> BML L<:E>PBMA G>:KER  E>:=L :G=  BGLM:EE:MBHGL ;R OHE
NF> BM >Q<>>=L :GR HMA>K LM:M> HK EH<:E )*& IKH@K:F BG MA> 1GBM>=
/M:M>L +>QM $.&! EM>KG:MBO>LKHE> :L MA> LHE> :=FBGBLMK:MHK H? MA>
)*& IKH@K:F ?HK MA> IK>OBHNL =><:=> :EEHPL NL MH :G:ERS> :EE BGMK:
IKH@K:F K>?>KK:EL ><:NL> MA> =:M:;:L> BG<EN=>L I>KLHG:E B=>GMBW:;E>
BG?HKF:MBHG P> ?HEEHP %NF:G /N;C><ML IKHMH<HE :IIKHO>= ;R HNK &GLMB
MNMBHG:E .>OB>P H:K= LN<A :L :GHGRFBSBG@ MA> =:M: :G= IKH<>LLBG@ MA>
=:M: HGER HG L><NK>= ?:<BEBMB>L HGL>GM ?HK =:M: <HEE><M>= ;R $.&! E
M>KG:MBO>L MH ;> NL>= BG IKH@K:F >O:EN:MBHG :G= K>L>:K<A BL I:KM H? MA>
<EB>GM :@K>>F>GM
3> =BLMBG@NBLA>= ?>:MNK>L ;:L>= HG PA>MA>K MA>R :K> ;:L>= HG IKH
IKB>M:KR BG?HKF:MBHG :G= MANL F:R ;> FHK> O:EN:;E> BG BG<K>:LBG@ IK>
=B<MBHG HK ;:L>= HG IN;EB< BG?HKF:MBHG >@ <A:K:<M>KBLMB<L H? MA> <>G
LNL MK:<M MA> BG=BOB=N:E K>LB=>L BG KHNMBG>ER NL>= BG HMA>K LMN=B>L H? LH
E:K NIM:D> 0A> ?HKF>K FB@AM ;> FHK> IK>=B<MBO> H? :G BG=BOB=N:EL ;>
A:OBHK LBG<> MA>R :K> LI><BW< MAHN@A :K> A:K=>K B? GHM BFIHLLB;E> ?HK :G
BGLM:EE>K MH =>M>KFBG> IK><HGM:<M PBMA MA> AHNL>AHE= &G <HGMK:LM MK:<M
E>O>E O:KB:;E>L :K> IN;EB< :G= A:O> : EHP>K ;NK=>G MH <HEE><M
&G MA> G>QM LN;L><MBHGL P> =>L<KB;> MA> =:M: L>M BG =>M:BE /><MBHG
 =>L<KB;>L MA> <HK> =:M: ?KHF MA> .* :G= MA> IKH<>LLBG@ BG
OHEO>= BG =>O>EHIBG@ =:M: L>ML MA:M =>M:BE E>:=L K>?>KK>KL :G= MA>BK K>
LI><MBO> :MMKB;NM>L /><MBHG  =>L<KB;>L MA> ?>:MNK>L ?KHF HMA>K =:M:
L>ML MA:M :K> :MMKB;NM>= MH MA> E>:=L :G= K>?>KK>KL .>:=>KL <:G NL> 0:;E>
?HK : LNFF:KR H? EBLM H? ?>:MNK>L :G= MA>BK LHNK<>L :G= MA>
/NIIE>F>GM:KR &G?HKF:MBHG ?HK MA> ?NEE EBLM H? ?>:MNK>L :G= MA>BK LHNK<>L
0A> IKBO:M> G:MNK> H? MA> =:M: :G= AB@A KBLD H? B=>GMBW<:MBHG IK>O>GM NL
?KHF IN;EBLABG@ MA> =:M:L>M %HP>O>K K>L>:K<A>KL BGM>K>LM>= BG :<<>LL
BG@ MA> =:M: LAHNE= <HGM:<M MA> :NMAHKL HG : <:L>;R<:L> ;:LBL
2.1.1. Household-level data from the CRM
0A> .* =:M: ?KHF $.&! :EM>KG:MBO>L <HGLBLML H? MAK>> =:M: L>ML
:;HNM AHNL>AHE=L :M =B??>K>GM LM:@>L BG MA> L:E>L IBI>EBG> 0A> Leads
=:M: L>M <HGM:BGL <HGM:<M BG?HKF:MBHG :;HNM :EE AHNL>AHE=L MA:M :K> IH
Table 1
/NFF:KR H? ?>:MNK>L NL>= BG FH=>EL <E:LLBW>= ;R MA>BK LHNK<> 0A>K> :K> 
?>:MNK>L BG MA> K>?>KK:E FH=>E :G=  BG MA> EHLM E>:= FH=>E EBLM H? ?>:MNK>L
:G= MA>BK =>L<KBIMBHGL BL :O:BE:;E> BG MA> LNIIE>F>GM:KR BG?HKF:MBHG
-KHIKB>MR =:M: ?KHF $.&!
:EM>KG:MBO>L :G= 0K:<DBG@ MA>
/NG =:M: L>M
-N;EB<ER :O:BE:;E> MK:<ME>O>E =:M: :==>= MH MA> =:M:
L>M ;:L>= HG EH<:MBHG H? AHNL>AHE=
Demographics: SolarForest features: CalEnviroScreen
criteria:
@> $BGB BG=>Q Pollution Exposure:
GGN:E BG<HF> -HINE:MBHG =>GLBMR ,SHG>
$>G=>K .:<>>MAGB<BMR -* 
%HNL>AHE= LBS> @> !B>L>E -*
&GM>KG>M <<>LL ,<<NI:MBHG !KBGDBG@ P:M>K
MAK>:ML
):G@N:@> NL>= BG AHNL>AHE= 0K:GLIHKM:MBHG ->LMB<B=>L
.:<>"MAGB<BMR 0K:O>E 0BF> 0HQB< K>E>:L>L
->K<>GM *& MAK>LAHE=
   
*&
%>:EMA BGLNK:G<> 0K:?W<
!BO>KLBMR /BFILHGL BG=>Q
System characteristics: O>K:@> >E><MKB<BMR K>M:BE K:M>
"QI><M>= :GGN:E IKH=N<MBHG
D3A
*>=B:G AHF> O:EN> Environmental
Effects:
GGN:E >E><MKB< NL:@> D3A
%>:MBG@ ?N>EL E>:GNI LBM>L
EBF:M> A>:MBG@<HHEBG@
=>@K>> =:RL =:BER LHE:K
BGLHE:MBHG
$KHNG=P:M>K
MAK>:ML
Installation base: %:S:K=HNL P:LM>
$.&! EM>KG:MBO>L <EB>GML REPLICA features: &FI:BK>= P:M>K
;H=B>L
&GM>K<HGG><M>= LRLM>FL
*:KD>MK:M>
0HM:E )*& LBG@E>?:FBER HPG>K
H<<NIB>= ;NBE=BG@ <HNGM
/HEB= P:LM>
%HNLBG@ OBGM:@>
Tracking the Sun
Database:
1MBEBMR Sensitive
populations:
Spatial bands:   
 FBE>L
)H<:E> MRI> LMAF:
AHLIBM:EBS:MBHGL
Temporal bands: PBMABG
 =:RL H? IK>L<K>>G
FHGMAL R>:K :EEMBF>
O>K:@> FHGMAER AHNL>AHE=
>E><MKB<BMR >QI>G=BMNK>L ?HK EHP
O>KR EHP FH=>K:M> BG<HF>
AHNL>AHE=L
)HP ;BKMA P>B@AM
BG?:GML
:K=BHO:L<NE:K
=BL>:L>
Referral base:
HNGM H? K>?>KK:EL IKBHK MH
BGLM:EE:MBHG PBMABG 
   FB
Socioeconomic
factors:
!BLM:G<> MH G>:K>LM K>?>KK>K "=N<:MBHG
)BG@NBLMB< BLHE:MBHG
Installation process: -HO>KMR
!:RL ?KHF IK>L<K>>G MH
BGLM:EE:MBHG
1G>FIEHRF>GM
&GLM:EE:MBHG :?M>K K>?>KK:E
K>P:K= IKH@K:F
%HNLBG@ ;NK=>G
$.&! EM>KG:MBO>L K>@BHG MH
PAB<A : <EB>GM BL :LLB@G>=
Lead source type
#>:MNK>L :K> :O:BE:;E> ?HK K>?>KK:E =:M: L>M HGER
M>GMB:E E>:=L )>:=L MA:M :K> BGM>K>LM>= :G= JN:EBW>= ?HK MA> )*& IKH
@K:F :K> LBFNEM:G>HNLER MK:<D>= BG MA> Contacts :G= Project =:M: L>ML
0A> Contact =:M: L>M <HGM:BGL NI=:M>= <HGM:<M BG?HKF:MBHG H? :EE MA>
JN:EBW>= E>:=L :G= MA> Project =:M: L>M <HGM:BGL BG?HKF:MBHG :;HNM MA>
AHNL>AHE= G><>LL:KR ?HK MA> BGLM:EE:MBHG IKH<>LL >@ LRLM>F LBS>
.* =:M:;:L>L :K> IKHG> MH BKK>@NE:KBMB>L BG<EN=BG@ FBLLBG@ =:M:
=NIEB<:M> K><HK=L :G= GHGLM:G=:K=BS>= W>E=L 3> I>K?HKF>= >QM>GLBO>
=:M: <E>:GBG@ :G= IKH<>LLBG@ BG <HGLNEM:MBHG PBMA $.&! EM>KG:MBO>L
LM:?? MH IK>I:K> MA> =:M: ?HK NL> BG FH=>EBG@ #BKLM P> F>K@>= MA> E>:=L
<HGM:<ML :G= IKHC><M =:M: L>ML MH@>MA>K ;R B=>GMB?RBG@ >:<A AHNL>AHE=
PBMA : @EH;:E NGBJN> B=>GMBW<:MBHG GNF;>K /><HG= P> LM:G=:K=BS>=
K>LIHGL>L ?KHF K>E>O:GM W>E=L BGMH F>:GBG@?NE <:M>@HKB>L MH WM MA> IKH
UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
C><ML L<HI> &G MA> <:L> H? =:M: =BL<K>I:G<R B> MPH =B??>K>GM O:EN>L
?HK MA> L:F> W>E= P> =>?:NEM MH MA> FHLM K><>GM >GMKR
0A> =:M:;:L> A:= : E:K@> GNF;>K H? FBLLBG@ O:EN>L :MMKB;NM>= MH
MPH ?:<MHKL #BKLM =:M: BL <HEE><M>= :M O:KBHNL LM:@>L H? MA> L:E>L
IBI>EBG> MA>K>?HK> MA>K> BL : <HKK>E:MBHG ;>MP>>G FBLLBG@ =:M: :G= :=
O:G<>F>GM BG MA> HNMK>:<A IBI>EBG> #HK >Q:FIE> =>FH@K:IAB< BG?HK
F:MBHG BL <HEE><M>= =NKBG@ MA> IK>L<K>>GBG@ IKH<>LL PA>K>:L KHH? LNBM
:;BEBMR :G= >E><MKB<BMR <HGLNFIMBHG <HEE><M>= :M : E:M>K LM:@> 0A> L><
HG= ?:<MHK MA:M @>G>K:M>L FBLLBG@ =:M: BL BG>?W<B>G<R BG =:M: <HEE><MBHG
IKH<>LL BG<EN=BG@ GHGK>LIHGL> MH JN>LMBHGL >KKHKL PA>G NI=:MBG@ :G=
F:BGM:BGBG@ MA> =:M:;:L> :G= >OHENMBHG H? =:M: <HEE><M>= ;R MA> BG
LM:EE>K HO>K MBF> &G MABL <:L> P> BFINM>= F>:G :G= FH=> ?HK JN:GMBM:
MBO> =:M: :G= JN:EBM:MBO><:M>@HKB<:E =:M: K>LI><MBO>ER #HK EH<:MBHG
=>I>G=>GM O:KB:;E>L LN<A :L BG<HF> :G= >E><MKB<BMR <HGLNFIMBHG P>
BFINM>= MK:<ME>O>E >LMBF:M>L ?KHF IN;EB<ER :O:BE:;E> =:M: L>ML LN<A :L
MA> >GLNL :G= 1/ !>I:KMF>GM H? %HNLBG@ :G= 1K;:G !>O>EHIF>GM
%1! 78 .>LIHG=>GML PBMAHNM EH<:MBHG =:M: E>LL MA:G  L:FIE>L
P>K> K>FHO>= ?KHF MA> =:M: L>M
,G> BFIEB<:MBHG H? =:M: <HEE><M>= OB: MA> )*& IKH@K:F BL MA:M :L :
IHM>GMB:E <EB>GM IKH@K>LL>L MAKHN@A MA> <EB>GM :<JNBLBMBHG IBI>EBG> MA>BK
.* K><HK= :<<KN>L FHK> =:M: 0A:M BL LHE:K :=HIM>KL A:O> LB@GB?B
<:GMER FHK> =:M: :MMKB;NM>= MH MA>F MA:G GHG:=HIM>KL PAB<A :??><ML
MA> ?>:MNK>L MA:M <HNE= ;> NL>= MH IK>=B<M :=HIMBHG IKHI>GLBMR ><:NL>
H? MABL :LRFF>MKR P> <AHL> MH IK>=B<M : JN:EBW>= E>:=L IKHI>GLBMR MH
;> EHLM BGLM>:= H? MA>BK IKHI>GLBMR MH BGLM:EE LHE:K #HK ;HMA MA> EHLM E>:=
:G= K>?>KK:E FH=>EL P> :K> <:K>?NE MH HGER NL> ?>:MNK>L MA:M <HNE= ;>
DGHP:;E> :M MA> MBF> H? IK>=B<MBHG :G= PA>K> MA>K> BL GH K>E:MBHGLABI BG
?>:MNK> :O:BE:;BEBMR MH MA> =>I>G=>GM O:KB:;E>
2.1.2. Relational and tract-level features
0H BFIKHO> IK>=B<MBO> :<<NK:<R H? MA> FH=>E L>O>K:E K>E>O:GM ?>:
MNK>L P>K> >G@BG>>K>= 0A>L> ?>:MNK>L :K> =>KBO>= ;:L>= HG F><A:GBLFL
MA:M A:O> ;>>G LAHPG MH HK ARIHMA>LBS>= MH BFI:<M MA> :=HIMBHG H? )*&
LHE:K
1LBG@ MA> E>:= HK K>?>KK>KL =:M> H? >G@:@>F>GM :G= EH<:MBHG P>
<HGLMKN<M L>O>K:E ?>:MNK>L MA:M =>L<KB;> LI:MB:E :G= M>FIHK:E IKHQBFBMR
MH HMA>K LHE:K BGLM:EE:MBHGL MH FH=>E I>>K >??><ML 0A>L> K>E:MBHG:E ?>:
MNK>L BG<EN=> BGLM:EE>= ;:L> O:EN>L MA:M BL <HNGML H? BGLM:EEL MA:M H<<NK
PBMABG =B??>K>GM <HF;BG:MBHGL H? LI:MB:E :G= M>FIHK:E ;:G=L /I:MB:E
;:G=L FH=>E>= P>K> K:=BB H?    :G=  FBE>L M>FIHK:E
;:G=L P>K> BGLM:EEL MA:M H<<NKK>= BG MA> IK>OBHNL  =:RL FHGMAL
R>:K :G= :EEMBF> 0A> O:EN>L P>K> <E:LLBW>= :L >BMA>K )*& LHE:K HK
F:KD>MK:M> LHE:K /I:MB:E :G= M>FIHK:E ;:G=L P>K> <:E<NE:M>= NLBG@ ?HK
>:<A )*& H;L>KO:MBHG ?KHF : <HF;BG:MBHG H? MA> .* =:M:;:L> :G=
LNIIE>F>GM>= ;R : GHGIN;EB< O>KLBHG H? MA> 0K:<DBG@ MA> /NG =:M: WE>
78
3> NL> MAK>> =:M: L>ML MH <HEE><M MK:<ME>O>E :MMKB;NM>L ?HK >:<A E>:=
:G= K>?>KK:E :E"GOBKH/<K>>G 78 ."-)&  78 :G= /HE:K#HK>LM 78
0:;E>  :E"GOBKH/<K>>G BL : F:IIBG@ MHHE MA:M A>EIL B=>GMB?R F>:
LNK>L H? <HFFNGBMRE>O>E IHEENMBHG >QIHLNK> &M BL :ELH NL>= ;R MA> LM:M>
H? :EB?HKGB: :L : <HGMKB;NMBG@ ?:<MHK BG AHNL>AHE= >EB@B;BEBMR MH I:KMB<B
I:M> BG MA> )*& IKH@K:F *>:LNK>L ?KHF MABL =:M: L>M BG=B<:M> AHP LH
<BH><HGHFB< :G= >GOBKHGF>GM:E <A:K:<M>KBLMB<L FB@AM BGXN>G<> LHE:K
:=HIMBHG 78 0A> +.") ."-)&  =:M: L>M :LL>LL>L ;NBE=BG@ LMH<D LNBM
:;BEBMR ?HK LHE:K :G= <HNGML H? LBG@E>?:FBER EHPBG<HF> HPG>K
H<<NIB>= AHNL>AHE=L :G= BL NL>= MH :LL>LL LHE:K LNBM:;BEBMR ;R MK:<M 78
#BG:EER P> NL> : LN;L>M H? ?>:MNK>L ?KHF MA> 5N >M :E /HE:K#HK>LM FH=>E
78 PAB<A A:L IK>OBHNLER ;>>G NL>= MH LN<<>LL?NEER IK>=B<M LHE:K :=HI
MBHG :M MA> MK:<M E>O>E 3A>K> MK:<ME>O>E ?>:MNK>L :K> K>IK>L>GM>= ;R
FHK> MA:G HG> O:KB:;E> P> HFBM K>=NG=:GM ?>:MNK>L
><:NL> MA> IKBF:KR BGM>G=>= INKIHL> H? MA> FH=>E BL IK>=B<MBHG
GHM >QIE:G:MBHG P> P>K> <:K>?NE MH HGER BG<EN=> ?>:MNK>L MA:M <HNE= ;>
DGHP:;E> :M MA> MBF> H? IK>=B<MBHG HK PA:M BL DGHPG :L =:M: E>:D:@>
78 /HF> >Q:FIE>L H? MA>L> BG<EN=>= JN:KM>KER WQ>= >??><ML HK : W>E=
<HKK>LIHG=BG@ MH MA> HNMK>:<A <HHK=BG:MHK :LLB@G>= MH MA> LHE:K <EB>GM
3ABE> ;HMA H? MA>L> ?>:MNK>L :G= HMA>KL PHNE= BFIKHO> MA> FH=>E MK:BG
BG@ WM MA>R >BMA>K PHNE= GHM ;> DGHP:;E> MH : ?NMNK> IK:<MBMBHG>K HK
<:GGHM ;> @>G>K:EBS>=
3> K>?HKFNE:M> MA> <E>:G>= =:M: BGMH MAK>> =:M: L>ML ?HK FH=>EBG@
0A> ,utcome data set BL =>O>EHI>= MH <HG=N<M :G >QIEHK:MHKR :G:ERLBL HG
MA> ?K>JN>G<R H? E>:=L ;R MA>BK LHNK<>L :G= MA>BK HNM<HF> 0A> Referral
data set BL NL>= MH IK>=B<M PA>MA>K :G >QBLMBG@ )*& <EB>GM PBEE K>?>K :G
HMA>K AHNL>AHE= MH :=HIM LHE:K 0A> K>?>KK:E =:M: L>M <HGM:BGL BG?HKF:
MBHG :;HNM :EE $.&! <EB>GML PAH BGLM:EE>= LHE:K BKK>LI><MBO> H? PA>MA>K
MA>R IKHOB=>= K>?>KK:E HK GHM G  #BG:EER MA> Lead loss data
set BL <HGLMKN<M>= MH IK>=B<M PA>MA>K : JN:EBW>= E>:= BL EHLM
G  0A> E>:= EHLL =:M: L>M <HGM:BGL BG?HKF:MBHG :;HNM :EE
JN:EBW>= E>:=L BKK>LI><MBO> H? PA>MA>K MA> E>:=L <HGO>KM>= HK GHM
2.2. Lost lead and referral predictive models
0PH BG=>I>G=>GM FH=>EL P>K> =>O>EHI>= MH IK>=B<M MA> HNM<HF>L H?
BGM>K>LM  E>:= EHLL PA>MA>K : JN:EBW>= E>:= =H>L GHM :=HIM LHE:K =>
LIBM> HMA>KPBL> ;>BG@ JN:EBW>= :G=  K>?>KK:EL PA>MA>K :G BGLM:EE>=
<EB>GM IKHOB=>L : K>?>KK:E 3> NL> MK>>;:L>= FH=>EL LI><B?B<:EER LMH
<A:LMB< @K:=B>GM;HHLM>= 4$ K:G=HF ?HK>LML ;><:NL> MA>R LMKBD> :
@HH= ;:E:G<> ;>MP>>G IK>=B<MBO> :<<NK:<R :G= BGM>KIK>M:;BEBMR 0A>
FH=>EL :K> BFIE>F>GM>= NLBG@ 4$HHLM 4$ EB;K:KR 78
HG<>IMN:EER K:G=HF ?HK>LML 78 HI>K:M> ;R =>O>EHIBG@ : =><BLBHG
MK>> MA:M B=>GMBW>L MA> HIMBF:E LIEBMMBG@ IHBGML ?HK >O>KR ?>:MNK> BG MA>
=:M: L>M ;:L>= HG MA> HNM<HF> H? BGM>K>LM .:G=HF ?HK>LML MK:BG FNEMBIE>
=><BLBHG MK>>L HG : K:G=HF LN;L>M H? O:KB:;E>L :O>K:@BG@ MA>BK IK>=B<
MBHGL MH RB>E= : E>LL ;B:L>= FH=>E 0A> LN<<>LLBO> MK>> ;K:G<A>L G:MBO>ER
BG<HKIHK:M> GHGEBG>:K :G= BGM>K:<MBO> O:KB:;E> >??><ML PBMAHNM >QIEB<BMER
LI><B?RBG@ MA>F MAHN@A :M : <HLM H? : =><K>:L> BG BGM>KIK>M:;BEBMR LMH
<A:LMB< @K:=B>GM ;HHLM>= K:G=HF ?HK>LM BG <HFI:KBLHG MH : K:G=HF ?HK
>LM FH=>E ;NBE=L :G >GL>F;E> H? =><BLBHG MK>>L LN<<>LLBO>ER WM MH IK>
=B<M MA> K>LB=N:EL >KKHKL H? IK>OBHNL FH=>EL MA>K>;R K>=N<BG@ HO>KWM
MBG@ 78 ):M>K P> NL> MA> MK:BG>= FH=>E MH @>G>K:M> /%- O:EN>L 78
:L : F>MAH= H? BGM>KIK>MBG@ =>@K>> :G= =BK><MBHG H? BGXN>G<> ;R MA> <H
O:KB:M>L HG MA> =>I>G=>GM O:KB:;E>L
L : IHBGM H? <HFI:KBLHG MH MA> 4$ FH=>EL P> :ELH =>O>EHI :
+:UO> :R>L FH=>E ?HK MA> E>:= EHLL :G= K>?>KK:E =:M: L>ML BFIE>F>GM>=
BG MA> :K>M EB;K:KR 78 +:UO> :R>L FH=>E <HGLB=>KL >:<A BG=>I>G
=>GM O:KB:;E> ?>:MNK> BG=BOB=N:EER :G= BL FHK> LHIABLMB<:M>= MA:G :
?K>JN>G<RP>B@AM>= K:G=HF @N>LL BG MA:M MA> IK>=B<MBHG BL ;:L>= HG MA>
<HG=BMBHG:E =>I>G=>G<> HG MA> BG=>I>G=>GM O:KB:;E> ?K>JN>G<R %HP
>O>K ;><:NL> +:UO> :R>L FH=>EL =H GHM <HGLB=>K BGM>K:<MBO> M>KFL :G=
:K> IKHG> MH HO>KWMMBG@ MA>R :K> <HGLB=>K>= BG?>KBHK MH ;HHLM>= K:G=HF
?HK>LML 0A> INKIHL> H? >LMBF:MBG@ MA> +:UO> :R>L FH=>EL BL MH NG=>K
LM:G= MA> K>E:MBO> BFIKHO>F>GM H? MA> 4$ F>MAH= MAHN@A MA>K> :K>
F:GR HMA>K FH=>EL >@ ,)/ )H@BM MA:M <HNE= ;> NL>= ;NM P>K> GHM
>O:EN:M>=
2.2.1. Data processing and feature engineering
!:M: IKH<>LLBG@ :G= ?>:MNK> >G@BG>>KBG@ BL MA> IKH<>LL PA>K> MA>
=:M: BL MK:GL?HKF>= BGMH ?>:MNK>L MA:M :K> NL>?NE ?HK =>O>EHIBG@ MA>
FH=>E 0ABL BG<EN=>L <E>:GBG@ MA> =:M:L>ML =>O>EHIBG@ G>P ?>:MNK>L
:G= MK:GL?HKFBG@ >QBLMBG@ ?>:MNK>L BGMH MA> G><>LL:KR ?HKF:M &G /><MBHG
 P> =>L<KB;> LHF> IHKMBHG H? =:M: IKH<>LLBG@ :G= ?>:MNK> >G@BG>>K
BG@ BG<EN=BG@ :==K>LLBG@ FBLLBG@ =:M: :G= =>O>EHIF>GM H? G>P ?>:
MNK>L MA:M =>L<KB;> LI:MB:E :G= M>FIHK:E IKHQBFBMR MH HMA>K LHE:K BGLM:E
E:MBHGL MH FH=>E I>>K >??><ML
#BKLMER :EE <:M>@HKB<:E =:M: :K> HG>AHM >G<H=>= ,G>AHM >G<H=BG@
BL : IKH<>LL PA>K> >:<A ?:<MHK H? : <:M>@HKB<:E O:KB:;E> BL <HGO>KM>= BGMH
: ;BG:KR O:KB:;E> :G= BL : G><>LL:KR ?HKF:M ?HK BFIE>F>GMBG@ <:M>@HKB
<:E ?>:MNK>L BG 4$ FH=>EL /><HG= ?>:MNK>L MA:M A:O> O:KB:G<> <EHL> MH
S>KH :K> K>FHO>= +>:K S>KH O:KB:G<> F>:GL MA:M MA>K> BL O>KR EBMME> BG
?HKF:MBHG BG : O:KB:;E> ;><:NL> MA>R FHLMER <HGLBLM H? : LBG@E> O:EN>
0ABK=ER P> :==K>LL FH=>E FNEMB<HEEBG>:KBMR B> FNEMBIE> AB@AER <HKK>
UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
E:M>= O:KB:;E>L IKHOB=BG@ LBFBE:K BG?HKF:MBHG MH MA> FH=>E 2:KB:;E>L
PBMA : AB@A <HKK>E:MBHG L<HK> P>K> B=>GMBW>= :G= K>FHO>= &=>GMBW<:
MBHG :G= K>FHO:E H? ?>:MNK>L PBMA G>:K S>KH O:KB:G<> :G= AB@A <HKK>E:
MBHG P:L I>K?HKF>= NLBG@ MA> :K>M I:<D:@> 78 /M:G=:K=BS:MBHG :G=
GHKF:EBS:MBHG ?HK GNF>KB<:E O:KB:;E>L P:L GHM <HG=N<M>= LBG<> MK>>
;:L>= FH=>EL :K> :@GHLMB< MH L<:EBG@ +HM> MA:M MA>L> =:M: IKH<>LLBG@
LM>IL :K> <HG=N<M>= ?HK >:<A L:FIE>= =:M:
2.2.2. Feature selection and hyperparameter tuning
3> >FIEHR L>O>K:E =:M: L<B>G<> ;>LM IK:<MB<>L ?HK =>O>EHIBG@ ;HMA
IK>=B<MBO> FH=>EL #BKLM =:M: BL K:G=HFER LIEBM ;>MP>>G MK:BG
BG@O:EB=:MBG@ MA> FH=>E  :G= MA> K>LM ?HK M>LMBG@  #>:MNK>L
:K> MA>G LN;L>M ?HK >:<A LI><BW<:MBHG B> IN;EB<HGER =:M: IKBO:M>
HGER :G= :EE ?>:MNK>L /BG<> MA> M:K@>M O:KB:;E> ?HK ;HMA MA> E>:= EHLL
:G= K>?>KK:E FH=>EL P>K> BF;:E:G<>= BG MA> MK:BGBG@ =:M: L>M MA> HO>K
K>IK>L>GM>= <E:LL B> GHGEHLM E>:=L GHGK>?>KK>KL P>K> =HPG L:F
IE>= PBMAHNM K>IE:<>F>GM MH =>O>EHI WO> <E:LL;:E:G<>= MK:BGBG@ =:M:
L>ML 3BMABG >:<A MK:BGBG@ BM>K:MBHG MA> FH=>E I>K?HKFL : K:G=HF WO>
?HE= <KHLLO:EB=:M>= @KB= L>:K<A MH MNG> FH=>E ARI>KI:K:F>M>KL #HK
>:<A LN;L>M H? ?>:MNK>L MA> L:F> ARI>KI:K:F>M>K LI:<> P:L L>:K<A>= MH
B=>GMB?R MA> ;>LM I>K?HKFBG@ FH=>E ?HK MA:M L>M H? ?>:MNK>L
&G 4$ FH=>EL ARI>KI:K:F>M>KL :K> MNG>= MH F:QBFBS> I>K?HK
F:G<> PABE> FBGBFBSBG@ HO>KWMMBG@ 0A> eta HK MA> E>:KGBG@ K:M> IK>
O>GML HO>KWMMBG@ ;R P>B@AMBG@ MA> G>P IK>=B<MBHGL F:=> ?KHF FH=>E
K>LB=N:EL Gamma LI><BW>L MA> FBGBFNF EHLL K>=N<MBHG K>JNBK>= MH
F:D> : ?NKMA>K I:KMBMBHG HG : E>:? GH=> H? MA> MK>> MA> E:K@>K gamma
MA> FHK> <HGL>KO:MBO> MA> FH=>E />O>K:E ARI>KI:K:F>M>KL <HGMKHE MA>
<A:K:<M>KBLMB<L H? MA> ;HHLM>= MK>>L maximum depth =B<M:M>L AHP F:GR
LN<<>LLBO> LIEBML H? MA> =:M: G>P MK>> ;K:G<A>L :K> I>KFBMM>= 3> NMB
EBS>= : K:G=HF @KB= L>:K<A HO>K : IK>=>WG>= LI:<> L>> 0:;E>  3>
:ELH M>LM>= MA> <HFI:K:MBO> I>K?HKF:G<> H? >QI:G=>= :G= K>=N<>= ?>:
MNK> L>ML PBMA ?H<NL HG K>=N<BG@ FH=>E <HEEBG>:KBMR
0A> WG:E FH=>E LI><BW<:MBHG BL =>M>KFBG>= ;R MA> I>K?HKF:G<> H?
MA> FH=>E HG MA> =:M: NL>= MH MK:BG BM LIEBM BGMH WO> ;:E:G<>= ?HE=L :G=
MA> M>LMBG@ AHE=HNM 0H L>E><M : E>LLHO>KWM FH=>E P> <AHL> MA:M PAB<A
FBGBFBS>= MA> @:I ;>MP>>G MA> F>:G MK:BGBG@ I>K?HKF:G<> :G= M>LMBG@
I>K?HKF:G<>
3. Results
,NK K>LNEML ?H<NL HG B NG=>KLM:G=BG@ MA> F><A:GBLFL MAKHN@A
PAB<A )*& LHE:K E>:=L HKB@BG:M> <HGO>KM MH <EB>GML :G= IKHOB=> K>?>K
K:EL :G= BB IK>=B<MBG@ PA>MA>K :G )*& <EB>GM ;><HF>L : EHLM E>:= HK
IKHOB=>L : K>?>KK:E :G= MA> ?>:MNK>L MA:M BG?HKF MA> IK>=B<MBHG 1G=>K
LM:G=BG@ MA>L> F><A:GBLFL IKHOB=>L BFIHKM:GM BGLB@AML MA:M A>EI BF
IKHO> MA> >??><MBO>G>LL H? >QBLMBG@ IKH@K:FL :G= MA> =>LB@G H? ?NMNK>
)*& IKH@K:FL 0A> =:M: L>M NL>= A>K>BG <:G ;> <HGLB=>K>= MH ;> K>IK>
L>GM:MBO> H? MA> )*& LHE:K >QI>KB>G<> BG :EB?HKGB: MAHN@A L>O>K:E
Table 2
)BLM H? ARI>KI:K:F>M>KL BG HNK FH=>E :G= MA>BK L>:K<A LI:<>
%RI>KI:K:F>M>K !>L<KBIMBHG />:K<A /I:<>
>M: )>:KGBG@ K:M> 

G9>LMBF:MHKL +NF;>K H? >LMBF:MHKL   
FBG9L:FIE>L9
LIEBM
*BGBFNF L:FIE>L BG >:<A MK>> LIEBM 
F:Q9=>IMA *:QBFNF =>IMA H? MK>>    
LN;L:FIE> #K:<MBHG H? H;L>KO:MBHGL MH ;> K:G=HFER
L:FIE>= ?HK >:<A MK>>
  
<HEL:FIE>9
;RMK>>
#K:<MBHG H? ?>:MNK>L MH NL> BG >:<A MK>>   

@:FF: *BGBFNF EHLL K>=N<MBHG K>JNBK>= MH F:D> :
G>P LIEBM
  
FBG9<ABE=9
P>B@AM
.>JNBK>= FBGBFNF LNF H? P>B@AML H? :EE
H;L>KO:MBHGL BG <ABE=
  
K>WM ->K?HKF:G<> F>:LNK> NL>= MH K>WM FH=>E .,
<:O>:ML LAHNE= ;> <HGLB=>K>= PA>G BGM>KIK>MBG@ MA> =:M: *HLM GHM:;ER
G>:KER :EE  H? <EB>GML BG MA> IKH@K:F K><>BO>= MA>BK LRLM>FL :M GH
<HLM 0ANL MA> =:M: >QIEHK>L =KBO>KL H? LHE:K :=HIMBHG BG=>I>G=>GM H?
><HGHFB< ?:<MHKL ELH MA> )*& IKH@K:F BFIHL>L @>H@K:IAB< AHF>
HPG>KLABI :G= BG<HF>;:L>= <KBM>KB: ?HK IKH@K:F >EB@B;BEBMR PAB<A
FB@AM =B??>K BG HMA>K CNKBL=B<MBHGL
3.1. LMI leads: Sources and outcomes
0A> .* =:M:;:L> <HGM:BG>=  E>:=L ?KHF IKBE  
MA> WKLM //%?NG=>= BGLM:EE:MBHG MH !><>F;>K   ;HNM 
H? MA> E>:=L G  P>K> <HGM:<M>= ;R $.&! HNM H? PAB<A  H?
MA> E>:=L G  A:= @HG> MAKHN@A MA> :<JNBLBMBHG IBI>EBG> :G=
A:= : =>?BGBM> HNM<HF> B> BGLM:EE>= NGJN:EBW>= EHLM PABE> MA> K>
F:BGBG@  H? MA> E>:=L G  P>K> GHM =>WGBMBO>ER
IKH<>LL>= >EHP P> =>L<KB;> MA> IKH<>LL MAKHN@A PAB<A MA> E>:=L :K>
@>G>K:M>= :G= MA>BK K>LI><MBO> HNM<HF>L
0A>K> P>K> L>O>K:E F><A:GBLFL MAKHN@A PAB<A )*& LHE:K E>:=L P>K>
@>G>K:M>= :G= PBMA =B??>K>GM E>O>EL H? >??><MBO>G>LL #B@  0A>L> :K>
 K>?>KK:EL ;R >QBLMBG@ <EB>GML HMA>K BG=BOB=N:EL <HFFNGBMR LNIIHKM
>KL :G= HK@:GBS:MBHGL  E>:=L <HGG><MBG@ PBMA $.&! MAKHN@A P>;
L>:K<A  :=O>KMBL>F>GM BG P>;LBM>L G>PLI:I>KL :G= HMA>K IKBGM :=L
 M:K@>M>= :=L ;R IHLM:E F:BE>KL :G= >F:BEL MH IHM>GMB:EER >EB@B;E> I:K
MB<BI:GML  <:GO:LLBG@ >EB@B;E> G>B@A;HKAHH=L =HHKMH=HHK :G= 
AHE=BG@ ;EH<D I:KMRMRI> LHE:K BGLM:EE:MBHG >O>GML ,? MA>  E>:=L
MA:M A:O> :G HNM<HF> K>?>KK:EL  <HGLMBMNM>= MA> E:K@>LM source of
leads ?HEEHP>= ;R P>; L>:K<A  !B@BM:E :G= IKBGM :=L F:LL F:BE>K
HK >F:BE <:GO:LLBG@ E>:=HKB@BG:MBHG >O>GML EBLM :<JNBLBMBHG :G=
IAHG> LHEB<BM:MBHG :G= HMA>K F>MAH=L <HEE><MBO>ER <HGLMBMNM>L : JN:KM>K
BG MHM:E 0A> K>F:BGBG@  H? MA> E>:=L =B= GHM A:O> : K><HK=>= E>:=
LHNK<>
.>?>KK:EL P>K> :ELH one H? MA> FHLM >?W<B>GM LHNK<>L H? E>:=L PA>K>
 H? :EE K>?>KK:EL E>= MH BGLM:EELMHII>= ;R =HHKMH=HHK <:GO:LLBG@
:G= E>:= HKB@BG:MBHG >O>GML  E>:=L PBMAHNM : EBLM>= LHNK<>
GHG> HK HMA>KLHNK<>L  0A>L> :K> ?HEEHP>= ;R M:K@>M>= F:BE
>F:BE  :G= IAHG> <:EE  !B@BM:E :G= IKBGM :=L  P>;
L>:K<A  :G= EBLM :<JNBLBMBHG  P>K> MA> E>:LM >?W<B>GM E>:=
LHNK<>L ,O>K:EE  H? :EE BGLM:EEL P>K> K>?>KK>=  H? MA> BGLM:EEL
Fig. 1. /HNK<>L :G= ?K>JN>G<R H? HNM<HF>L H? )*& E>:=L 0A> <HNGML BG MA>
W@NK> K>?>K MH MA> GNF;>K H? E>:=L HKB@BG:MBG@ ?KHF >:<A LHNK<> :G= MA>BK
>O>GMN:E HNM<HF> 0A> :GGHM:MBHG H? I>K<>GM:@>L BG MA> W@NK> K>?>K MH MA>
IKHIHKMBHG ?HK : @BO>G E>:= LHNK<> MA:M K>LNEM>= BG : LHE:K BGLM:EE:MBHG GHM
MA>G I>K<>GM H? E>:=L HO>K:EE
UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
=B= GHM A:O> : K><HK=>= LHNK<> GHG> :G= GH HMA>K LHNK<> <HG
MKB;NM>= MH FHK> MA:G  H? MA> MHM:E BGLM:EEL F:DBG@ K>?>KK:EL MA> FHLM
:MMK:<MBO> E>:= LHNK<> ?HK )*& LHE:K +HM:;ER K>?>KK:EL IKHOB=> : LBFBE:K
KHE> BG F:KD>MK:M> LHE:K E>:= @>G>K:MBHG PA>K> MA>R :K> >LMBF:M>= MH
<HGLMBMNM> ;HMA MA> E:K@>LM LHNK<>  H? BGLM:EEL :G= A:O> MA> EHP>LM
:LLH<B:M>= <NLMHF>K :<JNBLBMBHG <HLM 78
3.2. LMI acquisition pipeline
-KHLI><MBO> )*& <EB>GML ?HEEHP : <NLMHF>K :<JNBLBMBHG IBI>EBG> LBFB
E:K MH MA:M H? F:KD>MK:M> LHE:K =BO>K@BG@ IKBF:KBER MH =>M>KFBG> MA>BK
IKH@K:F >EB@B;BEBMR K:MA>K MA:G :;BEBMR MH I:R ?HK LHE:K )>:=L LM:KM PBMA
: IK>L<K>>GBG@ IKH<>LL MH =>M>KFBG> MA>BK IKH@K:F >EB@B;BEBMR ?HEEHP>=
;R LN;FBMMBG@ :G :IIEB<:MBHG +>QM BGLM:EE>KL <HFIE>M> :G BGI>KLHG LBM>
OBLBM MH >O:EN:M> M><AGB<:E LNBM:;BEBMR =>O>EHI : M>GM:MBO> LRLM>F =>LB@G
:G= NEMBF:M>ER BG=B<:M> :IIKHO:E ?HK BGLM:EE:MBHG HK GHM #B@ LAHPL :M
PAB<A IHBGML :G= ?HK PA:M K>:LHGL =NKBG@ MA> L<K>>GBG@ :G= :IIEB<:
MBHG IKH<>LL : E>:= >QBM>= MA> <EB>GM IBI>EBG> 0ABL W@NK> HFBML :GR E>:=L
MA:M P>K> GHM <HGM:<M>= ;R $.&!
?>P BGLB@AML >F>K@> ?KHF MA> :G:ERLBL H? MA> E>:= :<JNBLBMBHG
IBI>EBG> #BKLM : E:K@> ?K:<MBHG H? MA> E>:=L MA> IK:<MBMBHG>K <HGM:<M>=
=B= GHM K><>BO> LHE:K  0A> MPH E:K@>LM K>:LHGL P>K> BG>EB@B;BEBMR
MH I:KMB<BI:M> BG MA> IKH@K:F =N> MH @>H@K:IAB< BG>EB@B;BEBMR  H? :EE
E>:=L PBMA K>LHEO>= LM:MNL :G= =BLBGM>K>LM BG LHE:K HK EHLBG@ <HGM:<M
EHLM  M MA> MBF> H? :G:ERLBL :KHNG=  H? MA> E>:=L :K>
LMBEE BG MA> :<JNBLBMBHG IBI>EBG> ?HK ?NKMA>K IKH<>LLBG@ ->G=BG@3:BM
EBLM :G= P>K> GHM :G:ERS>= =N> MH MA>BK NGK>LHEO>= LM:MNL
IIKHQBF:M>ER BG EHLM E>:=L  P>K> =>>F>= >EB@B;E> ?HK LH
E:K =NKBG@ IK>L<K>>GBG@ PAB<A BFIEB>L O:EN:;E> E>:=L P>K> EHLM 1G=>K
LM:G=BG@ K>:LHGL ;>ABG= EHLBG@ <HGM:<M :?M>K BGBMB:E BGM>K>LM PHNE= A>EI
BFIKHO> <HGO>KLBHG K:M>L H? MA> IKH@K:F 3ABE> MA> HMA>K ;:KKB>KL MH
:=HIMBHG :K> KHHM>= BG ?>:MNK>L BGA>K>GM MH IKH@K:F =>LB@G HK <HG=BMBHG
H? : @BO>G ;NBE=BG@ MA> <:NL>L ?HK EHLM E>:=L :K> E>LL :II:K>GM 0A>R F:R
;> HPBG@ MH FBLMKNLM H? NMBEBMB>L <HG<>KGL :;HNM EHG@M>KF F:BGM>G:G<>
>QI>GL>L HK NG<>KM:BGMR HO>K MA> ;>G>WML H? MA> IKH@K:F 78
0A> .* IKHOB=>= =>M:BE>= BG?HKF:MBHG ?HK =BLJN:EB?RBG@ : E>:=
PAB<A P>K> @KHNI>= :L  BG>EB@B;E> @>H@K:IARE>:= K>LB=>= BG : @>
H@K:IAR BG>EB@B;E> ?HK I:KMB<BI:MBHG BG MA> )*& LHE:K IKH@K:F  LRL
M>F =>LB@G <HGLMK:BGML >@ NGLNBM:;E> KHH? :K>:  AHNL>AHE= A:= BG
<HF> AB@A>K MA:G IKH@K:F EBFBM  NG:O:BE:;E> BG<>G
Fig. 2. )*& E>:= :<JNBLBMBHG IBI>EBG> OBLN:EBS>= :L : /:GD>R =B:@K:F 0A>
IBI>EBG> <:M>@HKBS>L E>:=L PAH BGLM:EE>= LHE:K P>K> EHLM=NKBG@ MA> IKH<>LL
P:BMEBLM>= HK I>G=BG@ :G= MAHL> <:L>L MA:M P>K> NGJN:EBW>= ?HK L>O>K:E K>:
LHGL -K>L<K>>G JN:EBW>= E>:=L :K> E>:=L MA:M P>K> =>>F>= JN:EBW>= ;:L>= HG
:G BGBMB:E :IIEB<:MBHG MH MA> )*& IKH@K:F :G= IKH<>>= ?HK ?NKMA>K <HGLB=>K:MBHG
MBO>:IIEB<:MBHG H<<NKK>= :?M>K MA> IKH@K:F A:= KNG HNM H? ?NG=BG@
 MA> E>:= P:L GHM : AHF>HPG>K :G=  FBL<>EE:G>HNL HMA>K K>:LHGL
LN<A :L :EK>:=R HPBG@ : LHE:K LRLM>F 0AHN@A K>?>KK:EL P>K> : E:K@>
LHNK<> H? E>:=L :G= NEMBF:M>ER BGLM:EEL GHM :EE K>?>KK:EL P>K> LN<<>LL?NE
#HK >Q:FIE> : <EB>GM <HNE= K>?>K : E>:= MA:M K>LB=>L BG :G BG>EB@B;E> >G
LNL MK:<M HK PAHL> AHF> P:L =>>= K>LMKB<M>= ?HK LHE:K
"EB@B;BEBMR MH I:KMB<BI:M> BG :EB?HKGB:L )*& LHE:K IKH@K:F P:L MA>
E:K@>LM LBG@E> =KBO>K H? E>:= =BLJN:EBW<:MBHG 0ABL LI:GL @>H@K:IAB< BG>E
B@B;BEBMR  AB@A BG<HF>  :G= GHM ;>BG@ : AHF>HPG>K 
0A> :EB?HKGB: //% IKH@K:F K>JNBK>L MA:M I:KMB<BI:GML B HPG :G=
EBO> BG MA>BK AHF> BB A:O> :G BG<HF> ;>EHP  H? *& :G= BBB MA:M
MABL AHF> ;> =>>=K>LMKB<M>= EHPBG<HF> AHNLBG@ HK EH<:M>= BG : %1!
=>LB@G:M>= <>GLNL MK:<M BG<EN=BG@ "GM>KIKBL> 6HG>L :G= 0:K@>M>= "<H
GHFB< K>:L /RLM>F =>LB@G <HGLMK:BGML  <HGLMBMNM>= MA> L><HG=
E:K@>LM <:M>@HKR H? =BLJN:EBW>= E>:=L ,? MA>L> MA> F:CHKBMR  P>K>
=BLJN:EBW>= =N> :G HE= KHH? ?HEEHP>= ;R <H=> BLLN>L  LHE:K LA:=
BG@  :G= KHH? MRI>  ;HNM  H? :EE JN:EBW>= E>:=L P>K> NG
JN:EBW>= =N> MH NG:O:BE:;E> BG<>GMBO> B> MA> IKH@K:F ?NG=BG@ P:L
>QA:NLM>=
3.3. Prediction models
&FIKHOBG@ MA> IK>=B<MBHG H? E>:= <HGO>KLBHG :G= K>?>KK:EL A:L LN;
LM:GMB:E BFIEB<:MBHGL ?HK IK:<MBMBHG>KL ;><:NL> BM <:G EHP>K MA>BK <NL
MHF>K :<JNBLBMBHG <HLML PAB<A A:L ;HMA IKBO:M> :G= LH<B:E ;>G>WML
$BO>G MA> K:K> H<<NKK>G<> H? HNK M:K@>M <E:LL>L  H? JN:EBW>= E>:=L
:K> EHLM :G=  H? <EB>GML IKHOB=> K>?>KK:EL P> <HFI:K> FH=>E I>K
?HKF:G<> IKBF:KBER NLBG@ balanced accuracy, : F>MKB< MA:M <HGLB=>KL MA>
FH=>EL I>K?HKF:G<> BG IK>=B<MBG@ ;HMA MA> IHLBMBO> :G= G>@:MBO> <E:LL
:E:G<>= :<<NK:<R BL =>WG>= :L MA> :KBMAF>MB< F>:G H? MA> FH=>E L>GLB
MBOBMR :G= LI><BW<BMR K:G=HF @N>LL FH=>E A:L : ;:E:G<>= :<<NK:<R
H?  #HK >Q:FIE> IK>=B<MBG@ :EE HNM<HF>L MH ;> BG MA> IHLBMBO> <E:LL
A:L : ;:E:G<>= :<<NK:<R H?  ;><:NL>  H? MA> MKN> IHLBMBO>L :K>
B=>GMBW>= ;NM  H? MA> MKN> G>@:MBO>L R : LBFBE:K EH@B< : K:G=HF
@N>LL FH=>E P>B@AM>= HG MA> IHLBMBO> :G= G>@:MBO> <E:LL ?K>JN>G<B>L
:ELH A:L : ;:E:G<>= :<<NK:<R H?  0ANL :  ;:E:G<>= :<<NK:<R
K>IK>L>GML MA> XHHK H? NL>?NE IK>=B<MBO> FH=>EL
&FIKHO>F>GML BG :<<NK:<R H? IK>=B<MBHG EHP>K MA> <HLM H? IKH@K:F
:=FBGBLMK:MBHG OB: K>=N<BG@ MA> >??HKM G>>=>= MH :<JNBK> IKH@K:F I:K
MB<BI:GML E>:= EHLL HK MH :<JNBK> G>P E>:=L K>?>KK:EL HGLB=>K MA> ?HE
EHPBG@ BEENLMK:MBO> >Q:FIE> ?HK : FH=>E MA:M IK>=B<ML K>?>KK:EL PBMA :
L>GLBMBOBMR H?  :G= LI><BW<BMR H?  #B@ :G= PBMA K>?>KK:E
IK>O:E>G<> H?  &G MABL >Q:FIE> ?HK MA> GNEE <:L> MA> IK:<MBMBHG>K
PHNE= K>JNBK> <HGM:<MBG@  <EB>GML MH IKHOB=> K>?>KK:EL MH NEMB
F:M>ER K><>BO>  LN<<>LL?NE K>?>KK:EL &G <HGMK:LM MA> IK:<MBMBHG>K
<HNE= NL> MA> IK>=B<MBO> FH=>E MH BG?>K MA:M  <EB>GML FB@AM IKHOB=> :
K>?>KK:E :G=  FB@AM GHM ,? MA>  BG?>KK>= K>?>KK>KL  :K> :<MN
:EER K>?>KK>KL MKN> IHLBMBO>L :G=  :K> GHM ?:EL> IHLBMBO> ,? MA> 
BG?>KK>= GHGK>?>KK>KL  PBEE GHM IKHOB=> : K>?>KK:E MKN> G>@:MBO>
:G= PHNE= A:O> ?:EL> G>@:MBO> 0:D>G MH@>MA>K MA> IK:<MBMBHG>K
PHNE= K>JNBK>  H? LM:?? MBF>  MH :<JNBK>  H? K>?>K
K>KL  %HP>O>K NL> H? : IK>=B<MBO> FH=>E BG MABL <HGM>QM G><
>LL:KBER F>:GL =BLK>@:K=BG@ LHF> E>:=L MA:M :K> IK>=B<M>= MH A:O> G>@:
MBO> HNM<HF>L PAB<A <HNE= A:O> >JNBMR BFIEB<:MBHGL ;NM PBMABG MA>
)*& IHINE:MBHG
HMA FH=>EL L:P : LN;LM:GMB:E BFIKHO>F>GM BG MA> IK>=B<MBHG H?
<EB>GML MA:M IKHOB=>= K>?>KK:EL HK ;><:F> EHLM E>:=L #B@  ,O>K:EE MA>
EHLM E>:= FH=>E  A:= : AB@A>K ;:E:G<>= :<<NK:<R MA:G MA> K>?>KK:E
FH=>E  :G= ;HMA ;>G>WM>= ?KHF BG<ENLBHG H? IKHIKB>M:KR =:M:
*H=>E L>GLBMBOBMR F>:LNK>L MA> IKHIHKMBHG H? IHLBMBO>L MA:M :K> <HKK><MER
B=>GMBW>= >@ MA> I>K<>GM:@> H? K>?>KK>KL MA:M :K> <HKK><MER B=>GMBW>= :L IKH
OB=BG@ : K>?>KK:E *H=>E LI><BW<BMR BL MA> BGO>KL> MA> IKHIHKMBHG H? G>@:MBO>L
MA:M :K> <HKK><MER B=>GMBW>= >@ GHGK>?>KKBG@ <EB>GML B=>GMBW>= :L LN<A -K>O:
E>G<> BL MA> H;L>KO>= ?K>JN>G<R H? MA> M:K@>M O:KB:;E> BG MA> =:M:
UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
Fig. 3. #B<MBMBHNL >Q:FIE> =>FHGLMK:MBG@ :IIEB<:MBHG H? IK>=B<MBO> FH=>E ?K:F>
PHKD MH BFIKHO> HI>K:MBHG:E >?W<B>G<R BG IK>=B<MBG@ )*& K>?>KK:EL :FHG@
<EB>GML
Fig. 4. :E:G<>= :<<NK:<R ?HK E>:= EHLL :G= K>?>KK:E FH=>EL NLBG@ IN;EB<HGER
IKBO:M>HGER HK IKBO:M> :G= IN;EB< =:M: MH MK:BG FH=>EL 0A> .:G=HF #HK>LM
FH=>E NGLNKIKBLBG@ER :<AB>O>L ;>MM>K I>K?HKF:G<> MA:G MA> +:UO> :R>L
:<KHLL :EE <HGW@NK:MBHGL K:G=HF @N>LL FH=>E A:L : ;:E:G<>= :<<NK:<R H?

 EHLM E>:=  K>?>KK:E O>KLNL IN;EB<ER :O:BE:;E> =:M: HGER &G
:EE MA> FH=>EL MK:BG>= MA> LI><BW<BMR P:L AB@A>K MA:G MA> L>GLBMBOBMR
+>QM P> IKHOB=> :==BMBHG:E =>M:BEL H? MA> FH=>E WM >QIE:G:MBHG H?
?>:MNK> BFIHKM:G<> :G= =BK><MBHG H? BGXN>G<> :G= BGM>KIK>M:MBHG H? MA>
FH=>E NL>?NEG>LL B? :IIEB>= BG <HGM>QM
3.3.1. Lost lead model
L GHM>= >:KEB>K =>LIBM> IHM>GMB:EER K><>BOBG@ : ?K>> LHE:K LRLM>F
HO>K  H? :EE JN:EBW>= )*& E>:=L P>K> EHLM=N> MH E:<D H? BGM>K>LM HK
=KHI H?? 0ABL LN@@>LML MA:M BFIKHO>F>GML MH MA> IK>=B<MBHG H? E>:= HNM
<HF>L <HNE= K>=N<> :G EHPBG<HF> LHE:K IKH@K:FL HI>K:MBHG:E <HLML
#HK >Q:FIE> BFIKHO>= FH=>EL :EEHP IK:<MBMBHG>KL MH >BMA>K ?H<NL K>
LHNK<>L HG AB@A IKHI>GLBMR E>:=L HK <HGO>KL>ER INM :==BMBHG:E >??HKM MH
<HGO>KM IHM>GMB:E EHLM E>:=L 3BMA :G BFIKHO>= IK>=B<MBHG H? : E>:=L
IKHI>GLBMR : IK:<MBMBHG>K FB@AM :IIER =B??>K>GM F:KD>MBG@ :G= HNMK>:<A
LMK:M>@B>L MH =B??>K>GM L>@F>GML 0AHN@A MABL BL :EK>:=R : <HFFHG IK:<
MB<> BG MA> F:KD>MK:M> LHE:K BG=NLMKR LN<A : LMK:M>@R <HNE= BGMKH=N<>
>MAB<:E BLLN>L K>@:K=BG@ =BL<KBFBG:MBHG HK >JNBM:;E> :<<>LL MH MA> )*&
IKH@K:F
,NK IK>?>KK>= LI><BW<:MBHG MA> 4$ FH=>E NLBG@ IKBO:M> IN;EB<
=:M: IKHOB=>= : LN;LM:GMB:E BFIKHO>F>GM BG MA> :<<NK:<R  HO>K
: K:G=HF @N>LL 0:;E>  0A> FH=>E P:L :ELH FHK> :<<NK:M> MA:G NL
BG@ IN;EB<HGER =:M: :G= :L <HFI:K>= MH : +:UO> :R>L FH=>E 1LBG@ MA>
?K:F>PHKD =>O>EHI>= :;HO> /><MBHG  :G= PBMAHNM : IK>=B<MBO>
FH=>E MA> IK:<MBMBHG>K PHNE= IKH<>LL  E>:=L MH A:O>  EHLM :G=
 K>F:BGBG@ %HP>O>K MA> 4$EE FH=>E L>GLBMBOBMR :G= LI><BW<BMR
BFIER MA:M @BO>G  E>:=L  :K> BG?>KK>= MH ;> EHLM :G=  GHM
,? MA>  BG?>KK>= GHM MH ;> EHLM E>:=L  PHNE= IKH<>>= MH MA> G>QM
LM>I H? >O:EN:MBHG :G=  PBEE ;> EHLM "??><MBO>ER MABL F>:GL MA:M MA>
IK:<MBMBHG>K PHNE= K>JNBK>  H? LM:?? MBF>  MH :<JNBK>
Table 3
*>MKB<L ?HK EHLM E>:= FH=>E ;R FH=>E :G= =:M: MRI> K>IHKM>= ?HK MA> M>LM =:M:
L>M
*H=>E
0RI>
!:M:
0RI>
-K>O:E>G<> />GLBMBOBMR /I><B?B<BMR -K><BLBHG # :E:G<>=
<<NK:<R
4$ -N;EB<      
EE     
+:UO>
:R>L
-N;EB<     
EE     
 H? MA> E>:=L  HK :  BG<K>:L> BG E>:=L @>G>K:MBHG >?W
<B>G<R
-K>=B<MBO> FH=>EL :K> MK:=BMBHG:EER NG=>KLMHH= :L H??>KBG@ EBMME> >Q
IE:G:MHKR IHP>K "GL>F;E> F>MAH=L BG <HGLMKN<MBG@ :G= <HF;BGBG@
L>O>K:E ANG=K>= >LMBF:MHKL WM MH I:KMB<NE:K HNM<HF> @>G>K:EER MK:=> BG
M>KIK>M:;BEBMR ?HK :<<NK:<R 78 0AHN@A IK>=B<MBHG BL HNK NEMBF:M>
@H:E MH BFIKHO> FH=>E BGM>KIK>M:;BEBMR P> =>O>EHI /%- O:EN>L
/%:IE>R ==BMBO> >Q-E:G:MBHG 78 HG MA> MK:BG>= FH=>E /%- BL :
@:F>MA>HK>MB< ?K:F>PHKD ?HK >QIE:BGBG@ HNMINML H? F:<ABG> E>:KGBG@
FH=>EL :G= PHKDL ;R :LLB@GBG@ :G BFIHKM:G<> O:EN> ?HK : LI><BW< IK>
=B<MBHG 78 +>QM MA> /%- O:EN>L :K> <HGO>KM>= ?KHF EH@H==L MH BF
I:<M HG IKH;:;BEBMR 0A>L> O:EN>L :K> MA>G IEHMM>= BG : ?>:MNK> BFIHK
M:G<> IEHM #B@  PAB<A LAHPL MA> K:GD>= F:@GBMN=> H? BFI:<M HG MA>
HNM<HF> O:KB:;E> :L P>EE :L : <HG=BMBHG:E =>I>G=>G<> IEHM #B@ 
PAB<A LAHPL AHP NGBO:KB:M> <A:G@>L BG : ?>:MNK> O:EN> BFI:<ML MA> IK>
=B<MBHG /I><B?B<:EER PBMABG MA> <HG=BMBHG:E =>I>G=>G<> IEHM >:<A =HM
K>IK>L>GML :G H;L>KO:MBHG MA> Q:QBL BML O:EN> :G= MA> R:QBL MA> BFI:<M
HG IK>=B<M>= IKH;:;BEBMR B> MA> HO>K:EE IHLBMBO> HK G>@:MBO> BFI:<M HG
MA> IK>=B<MBHG ),"// <NKO> WM BL IKHOB=>= ?HK <HGMBGNHNL O:KB:;E>L
MH OBLN:EBS> MA> =BK><MBHG:E BGXN>G<> H? >:<A H? MA> MHI ?>:MNK>L
#B@ LAHPL MA> ?>:MNK> BFIHKM:G<> ?HK MA> )HLM )>:= 4$EE FH=>E
?HK MA> MHI  O:KB:;E>L :G= BGXN>G<> H? MA> K>F:BGBG@  O:KB:;E>L
0ABL MRI> H? W@NK> BL :DBG MH : MK:=BMBHG:E >??><M LBS> <:E<NE:MBHG BG MA:M
Fig. 5. #>:MNK> BFIHKM:G<> IEHM ?HK MA> )HLM )>:= 4$EE FH=>E LAHPBG@
K:GD>= BFIHKM:G<> H? MHI  O:KB:;E>L :G= <NFNE:MBO> <HGMKB;NMBHG H? K>F:BG
BG@  O:KB:;E>L 0A> ?>:MNK> BFIHKM:G<> BL <:E<NE:M>= NLBG@ MA> FH=>E /%-
O:EN>L <HGO>KM>= BGMH F>:G :;LHENM> BFI:<M HG IKH;:;BEBMR #HK >Q:FIE> MA>
?>:MNK> +NF;>K H? K>?>KK>KL PBMABG  FBE>LA:= :G :O>K:@> :;LHENM> BF
I:<M HG MA> HNMINM IKH;:;BEBMR ;R :IIKHQBF:M>ER  0A> W@NK> =H>L GHM
LAHP =BK><MBHG H? BGXN>G<> PAB<A BL H;L>KO>= BG MA> <HG=BMBHG:E =>I>G=>G<>
IEHM
UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
Fig. 6. )>:= EHLL <HG=BMBHG:E =>I>G=>G<> IEHM =>FHGLMK:MBG@ F:K@BG:E >??><M H? : LBG@E> ?>:MNK> HG MA> IK>=B<M>= E>:= EHLL IKH;:;BEBMR 0A> <HG=BMBHG:E =>I>G=>G<>
IEHM BL <HGLMKN<M>= NLBG@ /%- O:EN>L ?HK MA> MHI M>G ?>:MNK>L H? AB@A>LM BFIHKM:G<> MH FH=>EL ;:E:G<>= :<<NK:<R K:GD>= ;R MA> ?>:MNK>L F>:G :;LHENM> /%-
O:EN> 0A> /%- O:EN>L :K> MA>G <HGO>KM>= BGMH >JNBO:E>GM BFI:<M HG IKH;:;BEBMR 0A> Q:QBL LAHPL MA> O:EN> H? MA> ?>:MNK> >@ MA> GNF;>K H? K>?>KK:EL PBMABG
 FBE>L #HK : @BO>G QO:EN> : IHLBMBO> IKH;:;BEBMR RO:EN> BG=B<:M>L MA> E>:= P:L FHK> EBD>ER MH ;> EHLM 0A> F:@GBMN=> H? /%- O:EN>L BG=B<:M>L MA> =>@K>> H?
BGXN>G<> HG MA> IK>=B<MBHG .>F:BGBG@ ?>:MNK>L :K> NL>= BG IK>=B<MBHG ;NM GHM LAHPG ?HK <HG<BL>G>LL
BM =>FHGLMK:M>L MA> =>@K>> H? BFIHKM:G<> ?HK >:<A FH=>E ?>:MNK> >Q
IK>LL>= BG M>KFL H? MA> F>:G :;LHENM> BFI:<M HG IKH;:;BEBMR #KHF MABL
P> H;L>KO> MA:M <HG<>GMK:MBHG H? K>?>KK>KL P:L MA> FHLM BFIHKM:GM O:KB
:;E> ?HK IK>=B<MBHG A:OBG@ : F>:G :;LHENM> BFI:<M H?  HG MA> IKH;
:;BEBMR H? E>:= EHLL +>QM :O:BE:;BEBMR H? MA> E>:=L K:<> MA>BK BG<HF>
:G= AHNL>AHE= LBS> P>K> :EE LB@GB?B<:GM <HGMKB;NMHKL
#B@ LAHPL MA> /%- <HG=BMBHG:E =>I>G=>G<> IEHM ?HK MA> 
FHLMBFIHKM:GM ?>:MNK>L BG MA> )HLM )>:= 4$EE FH=>E -HLBMBO>
/%- O:EN>L R:QBL BG=B<:M> MA> E>:= P:L FHK> EBD>ER MH ;> EHLM #HK
MA> FHLM BFIHKM:GM ?>:MNK> GNF;>K H? K>?>KK:EL PBMABG FB=>
<K>:L>L BG MA> <HG<>GMK:MBHG H? K>?>KK>KL P:L <HKK>E:M>= PBMA IK>=B<MBG@
E>:= EHLL #HK BGLM:G<> E>:=L PBMA S>KH K>?>KK>KL PBMABG  FBE>L P>K>
:IIKHQBF:M>ER  FHK> EBD>ER MH ;><HF> EHLM E>:=L MA:G MA> F>:G
O:EN> PA>K>:L A:OBG@ WO> K>?>KK>KL G>:K;R =><K>:L>= MA> IKH;:;BEBMR
H? : EHLM E>:= ;R  K>E:MBO> MH E>:=L PBMA S>KH K>?>KK>KL #HK ;BG:KR
O:KB:;E>L LN<A :L .>@BHG :R K>: MA> BGM>KIK>M:MBHG BL ;>MP>>G
MA> MPH HNM<HF>L B> <EB>GML MA:M K>LB=>= BG MA> :R K>: P>K> 
E>LL EBD>ER MH ;><HF> EHLM E>:=L MA:G MAHL> MA:M =B= GHM .>:=>KL LAHNE=
;> <:K>?NE MH :IIK><B:M> MA:M MA> /%- <HG=BMBHG:E =>I>G=>G<> IEHML
=H GHM F:D> <:NL:E <E:BFL >@ P> <:G HGER BG?>K MA:M BG<K>:L>L BG MA>
GNF;>K H? K>?>KK:EL P:L correlated PBMA : =><K>:L> BG MA> IKH;:;BEBMR H?
: EHLM E>:=
/RGMA>LBSBG@ MA> ?>:MNK> BFIHKM:G<> :G= <HG=BMBHG:E =>I>G=>G<>
IEHM P> H;L>KO> MA:M I>>K >??><ML P>K> LB@GB?B<:GM <HGMKB;NMHKL MH MA>
IK>=B<MBHG %:E? H? MA> MHI  ?>:MNK>L :K> IKHQB>L ?HK LI:MB:E I>>K >??><ML
GNF;>K H? K>?>KK>KL GNF;>K H? )*& HK F:KD>MK:M> LHE:K BGLM:EE:MBHGL
G>:K;R :G= =>GLBMR H? )*& BGLM:EEL :G= MA>K> BL BGXN>G<> ?KHF ;HMA
F:KD>MK:M> :G= )*& BGLM:EE:MBHGL %HP>O>K MA>K> BL LHF> :F;B@NBMR BG
MA> =BK><MBHG H? BGXN>G<> :L BG<K>:LBG@ K>?>KK>KL :G= <HG<>GMK:MBHG H?
BGLM:EEL BG MA> MK:<M P>K> G>@:MBO>ER <HKK>E:M>= PBMA IK>=B<MBG@ E>:= EHLL
PA>K>:L <HG<>GMK:MBHG H? )*& BGLM:EE:MBHGL P:L IHLBMBO>ER <HKK>E:M>=
0AK>> H? MA> MHI  ?>:MNK>L :K> AHNL>AHE= <A:K:<M>KBLMB<L BG<HF>
AHNL>AHE= LBS> K:<> :G= MPH :K> G>B@A;HKAHH= E>O>E <A:K:<M>KBLMB<L
:R K>: 1MBEBMR-$" 0AHN@A MA>R :K> GHM BG<EN=>= BG MA> MHI 
MA>K> :K> L>O>K:E >GOBKHGF>GM:EK>E:M>= O:KB:;E>L BG MA> MHI  >@
A:S:K=HNL P:LM> =B>L>E >FBLLBHGL P:M>K <HGM:FBG:MBHG
3.3.2. Referral model
!>LIBM> ;>BG@ HG> H? MA> E:K@>LM :G= AB@A>LM IKHI>GLBMR E>:= LHNK<>L
HGER  H? )*& :=HIM>KL IKHOB=>= : K>?>KK:E E>:=BG@ MH :G BGLM:EE L
UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
=BL<NLL>= BG /><MBHG  PBMA BFIKHO>= IK>=B<MBHG H? K>?>KK:EL :
IK:<MBMBHG>K FB@AM IKBHKBMBS> F:KD>MBG@ <HLML MH MA> L>M H? <EB>GML IK>
=B<M>= MH A:O> : AB@A>K K>?>KK:E IKHI>GLBMR HK M:K@>M IKHLI><MBO> <EB>GML
PBMA : LBFBE:K IKHWE>
L BG MA> EHLM E>:= FH=>E P> <HFI:K> MA> :<<NK:<R BG IK>=B<MBG@ K>
?>KK:EL NLBG@ 4$ O>KLNL +:UO> :R>L FH=>EL :G= NLBG@ =:M: ?KHF IN;
EB<HGER MH IN;EB< :G= IKBO:M> LHNK<>L 0:;E>  :E:G<>= :<<NK:<R H?
MA> 4$EE FH=>E P:L  EHP>K MA:G MA> EHLM E>:= FH=>E 0A> :==B
MBHG H? IKBO:M> =:M: FH=>LMER BFIKHO>L MA> FH=>E :L =H>L NLBG@ MA>
4$ O>KLNL +:UO> :R>L LI><BW<:MBHG )BD> MA> EHLM E>:= FH=>E LI><B
W<BMR BL AB@A>K MA:G L>GLBMBOBMR F>:GBG@ MA:M : IK:<MBMBHG>K PHNE= ;>
FHK> :<<NK:M> :M IK>=B<MBG@ <EB>GML MA:M =H GHM IKHOB=> K>?>KK:EL MA:G
MAHL> MA:M =H
1LBG@ MA> ?K:F>PHKD =>O>EHI>= :;HO> /><MBHG  PBMAHNM : IK>
=B<MBO> FH=>E MA> IK:<MBMBHG>K PHNE= G>>= MH IKH<>LL  <EB>GML MH
@>G>K:M>  K>?>KK:EL %HP>O>K MA> 4$EE FH=>E L>GLBMBOBMR :G=
LI><BW<BMR BFIER MA:M H?  <EB>GML  :K> BG?>KK>= MH ;> K>?>KK>KL
:G=  GHM ,? MA>  BG?>KK>= K>?>KK>KL  PHNE= :<MN:EER IKHOB=>
: K>?>KK:E :G=  PHNE= GHM %HP>O>K ;><:NL> H? MA> EHP IK>O:E>G<>
MABL BL LMBEE : LN;LM:GMB:E BG<K>:L> BG >?W<B>G<RMA> IK:<MBMBHG>K PHNE=
K>JNBK>  H? LM:?? MBF>  MH :<JNBK>  H? MA> K>?>KK>KL
 HK :  BG<K>:L> BG K>?>KK:E >?W<B>G<R 0AHN@A MA> K>?>KK:E
IK>=B<MBO> FH=>E IKHOB=>L LB@GB?B<:GM IK>=B<MBO> BFIKHO>F>GM 
HO>K : K:G=HF @N>LL BM BL :ELH >OB=>GM MA:M :==BMBHG:E NGH;L>KO>= O:KB
:;E>L BGXN>G<> MA> K>?>KK:E IKH<>LL 0A>L> <HNE= ;> NGH;L>KO>= K>?>K
K>KK>?>K>> <A:K:<M>KBLMB<L :G= :MMBMN=>L :G= ;>A:OBHKL
#B@ LAHPL MA> ?>:MNK> BFIHKM:G<> ?HK MA> .>?>KK:E 4$EE FH=>E
?HK MA> MHI  O:KB:;E>L :G= BGXN>G<> H? MA> K>F:BGBG@  O:KB:;E>L
HGMK:LMBG@ PBMA EHLM E>:= ?>:MNK> BFIHKM:G<> IEHM ?>:MNK>L A:O> :
EHP>K BFI:<M HG MA> K>?>KK:E IKH;:;BEBMR EBD>ER K>E:M>= MH MA> HO>K:EE
EHP>K FH=>E :<<NK:<R %HP>O>K EBD> MA> EHLM E>:= FH=>E =BLM:G<> MH
MA> G>:K>LM K>?>KK>K BL MA> FHLMBFI:<M?NE O:KB:;E> HG IK>=B<MBHG :??><M
BG@ MA> F>:G :;LHENM> IKH;:;BEBMR ;R  ==BMBHG:E BFI:<M?NE O:KB
:;E>L P>K> PA>MA>K MA> <EB>GM P:L BG<>GMBOBS>= MH IKHOB=> : K>?>KK>K
K>LB=>= BG MA> :R K>: :G= A:= BGM>KG>M :<<>LL#B@
#B@NK> LAHPL MA> /%- <HG=BMBHG:E =>I>G=>G<> IEHM ?HK MA> 4$EE
.>?>KK:E FH=>E -HLBMBO> /%- O:EN>L R:QBL BG=B<:M> MA> <EB>GM P:L
FHK> EBD>ER MH IKHOB=> : K>?>KK:E L PBMA MA> EHLM E>:= FH=>E I>>K >??><M
O:KB:;E>L P>K> BFIHKM:GM MH IK>=B<MBHG :G= MA>R P>K> <HGLBLM>GM BG MA:M
IKHQBFBMR MH K>?>KK>KL :G= A:OBG@ ;>>G K>?>KK>= MA>FL>EO>L P>K> ;HMA
IHLBMBO>ER <HKK>E:M>= PBMA IK>=B<MBG@ : K>?>KK:E &GM>K>LMBG@ER DGHPBG@
MA> GNF;>K H? >QBLMBG@ LHE:K BGLM:EE:MBHGL G>:K;R P:L GHM H? LN;LM:GMB:E
O:EN> BG IK>=B<MBG@ K>?>KK:EL NGEBD> EHLM E>:=L "<HGHFB< ?:<MHKL :ELH
A:= : LMKHG@ BGXN>G<> HG MA> K>?>KK:E FH=>E<EB>GML MA:M K><>BO>= :
FHG>M:KR K>P:K= ?HK =HBG@ LH BG<K>:L>L BG :GGN:E -2 >E><MKB<BMR @>G>K
:MBHG :G= AB@A>K >E><MKB<BMR >QI>G=BMNK>L P>K> :EE IHLBMBO>ER <HKK>E:M>=
PBMA IK>=B<MBHG H? K>?>KK:EL
1GEBD> MA> EHLM E>:= FH=>E >GOBKHGF>GM:E O:KB:;E>L P>K> GHM LB@GB?
B<:GM BG IK>=B<MBG@ K>?>KK:EL I:KM ?KHF MK:?W< =>GLBMR PAB<A BL IHLB
MBO>ER <HKK>E:M>= MA>K> :K> GH HMA>K >GOBKHGF>GMK>E:M>= O:KB:;E>L BG
MA> MHI  &GLM>:= G>B@A;HKAHH= LH<BH=>FH@K:IAB< O:KB:;E>L LN<A :L
=BLMKB;NMBHG H? K:<>>MAGB<BMR :G= :@> H? K>LB=>GML :K> K>IK>L>GM>= K>
I>:M>=ER #BG:EER IKHOB=BG@ K>?>KK:EL P>K> IHLBMBO>ER <HKK>E:M>= PBMA BG
M>KG>M :G= ?HK LF:EE>K AHNL>AHE=L 0PH ?>:MNK>L P>K> <HFFHG MH ;HMA
MA> EHLM E>:= :G= K>?>KK:E FH=>ELMHI BFIHKM:GM ?>:MNK>L <HNGML=BL
M:G<> MH G>:K;R K>?>KK>KL :G= AHNL>AHE= LBS>
Table 4
*>MKB<L ?HK K>?>KK:E FH=>E ;R FH=>E :G= =:M: MRI>
*H=>E
0RI>
!:M:
0RI>
-K>O:E>G<> />GLBMBOBMR /I><B?B<BMR -K><BLBHG # :E:G<>=
<<NK:<R
4$ -N;EB<      
EE     
+:UO>
:R>L
-N;EB<     
EE     
Fig. 7. #>:MNK> BFIHKM:G<> IEHM ?HK MA> .>?>KK:E 4$EE FH=>E LAHPBG@
K:GD>= BFIHKM:G<> H? MHI  O:KB:;E>L :G= <NFNE:MBO> <HGMKB;NMBHG H? K>F:BG
BG@  O:KB:;E>L 0A> ?>:MNK> BFIHKM:G<> BL <:E<NE:M>= NLBG@ MA> FH=>E /%-
O:EN>L <HGO>KM>= BGMH F>:G :;LHENM> BFI:<M HG IKH;:;BEBMR #HK >Q:FIE> MA>
?>:MNK> !BLM:G<> MH G>:K>LM K>?>KK>KA:= :G :O>K:@> :;LHENM> BFI:<M HG MA>
HNMINM IKH;:;BEBMR ;R :IIKHQBF:M>ER  0A> W@NK> =H>L GHM LAHP =BK><MBHG
H? BGXN>G<> PAB<A BL H;L>KO>= BG MA> <HG=BMBHG:E =>I>G=>G<> IEHM
4. Conclusion
"G<HNK:@BG@ MA> :=HIMBHG H? LHE:K ;R )*& AHNL>AHE=L :==K>LL>L
F:GR IHEB<R @H:EL R>M BM BL NG<E>:K AHP MH >??><MBO>ER L<:E> MA>L> IKH
@K:FL "QI>KB>G<> PBMA )*& LHE:K BG :EB?HKGB: IKHOB=>L L>O>K:E E>LLHGL
MA>K>BG !>LIBM> MA> G>:KMHM:E <HLM LN;LB=BS:MBHG :<JNBKBG@ <EB>GML ?HK
)*& IKH@K:FL K>F:BGL <A:EE>G@BG@ PA>K> E>LL MA:G  H? LHE:K E>:=L
E>= MH BGLM:EE:MBHGL E:K@> ?K:<MBHG H? MA> E>:=L P>K> =BLJN:EBW>= =N> MH
MA> BGMKB<:M> <HF;BG:MBHG H? @>H@K:IAR HPG>KLABI :G= BG<HF> >EB@B
;BEBMR <KBM>KB: 0AHN@A MA>L> <KBM>KB: :K> BFIHKM:GM ?HK >GLNKBG@ )*& IKH
@K:F ?NG=L K>:<A MA> =>LBK>= IHINE:MBHG MA>R :ELH BFIHL> :==BMBHG:E
:=FBGBLMK:MBO> ;NK=>G -KH@K:FL FB@AM A:O> :==BMBHG:E BFI:<M ;R EHHL
>GBG@ HK LBFIEB?RBG@ >EB@B;BEBMR <KBM>KB: HK ;R HMA>K F>:GL MH F:D> BM
>:LB>K MH =>M>KFBG> MA> >EB@B;BEBMR
/>O>K:E F><A:GBLFL P>K> NL>= ?HK :<JNBKBG@ )*& IKH@K:F E>:=L ;NM
GHG> FHK> >??><MBO> MA:G I>>K K>?>KK:EL +HM HGER P>K> K>?>KK:EL MA> LBG
@E> E:K@>LM LHNK<> H? E>:=L ;NM MA>R P>K> :ELH HG> H? MA> FHLM >??><MBO>
BG M>KFL H? I>K<>GM:@> <HGO>KLBHG ==BMBHG:EER =BLM:G<> MH G>:K;R K>
?>KK>KL P:L MA> FHLM BFIHKM:GM O:KB:;E> BG BG<K>:LBG@ IK>=B<MBO> I>K?HK
F:G<> ?HK ;HMA MA> EHLM E>:= :G= K>?>KK:E FH=>EL -K:<MBMBHG>KL :K> >G
<HNK:@>= MH BG<HKIHK:M> K>?>KK:E F><A:GBLFL BG ?NMNK> )*& IKH@K:FL :L
: F>:GL H? L>>=BG@ :G= L<:EBG@ )*& LHE:K :=HIMBHG
-K>=B<MBHG ?HK : LHE:K E>:= MH ;><HF> EHLM P:L FHK> LN<<>LL?NE MA:G
IK>=B<MBG@ PA>MA>K : <EB>GM PHNE= IKHOB=> : K>?>KK:E 0ABL LN@@>LML MA:M
PABE> E>:= LM:MNL <:G K>:=BER ;> IK>=B<M>= NLBG@ :O:BE:;E> =:M: K>?>KK:EL
F:R K>JNBK> :==BMBHG:E >QIE:G:MHKR O:KB:;E>L 0A>L> <HNE= BG<EN=> MA>
G:MNK> H? MA> K>?>KK>KK>?>K>> K>E:MBHGLABI E>O>E H? L:MBL?:<MBHG PBMA MA>
LHE:K LRLM>F :G= HMA>K :MMBMN=>L :G= GHKFL =B?W<NEM MH H;M:BG ?KHF H;
L>KO:MBHG:E :G= L><HG=:KR =:M: LHNK<>L :EHG> #NMNK> PHKD <HNE= M>LM
:==BMBHG:E LI><BW<:MBHGL ?HK IK>=B<MBG@ K>?>KK:EL HMA IK>=B<MBO> FH=
>EL P>K> >O:EN:M>= PA>G MK:BG>= HGER NLBG@ IN;EB<ER :<<>LLB;E> =:M: MH
MAHL> MA:M PHNE= ;> :<<>LLB;E> HGER MH MA> )*& IKH@K:F :=FBGBLMK:MHK
IKBO:M> =:M: LB@GB?B<:GMER BFIKHO>= IK>=B<MBHG BG MA> EHLM E>:= FH=>E
;NM HGER FH=>LMER BG MA> K>?>KK:E FH=>E "O>G B? : ?>:MNK> <HNE= ;>
UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
Fig. 8. .>?>KK:E <HG=BMBHG:E =>I>G=>G<> IEHM =>FHGLMK:MBG@ F:K@BG:E >??><M H? : LBG@E> ?>:MNK> HG MA> IK>=B<M>= IKH;:;BEBMR H? IKHOB=BG@ : K>?>KK:E 0A> <HG=BMBHG:E
=>I>G=>G<> IEHM BL <HGLMKN<M>= NLBG@ /%- O:EN>L ?HK MA> MHI M>G ?>:MNK>L H? AB@A>LM BFIHKM:G<> MH FH=>EL ;:E:G<>= :<<NK:<R K:GD>= ;R MA> ?>:MNK>L F>:G :;
LHENM> /%- O:EN> 0A> /%- O:EN>L :K> MA>G <HGO>KM>= BGMH >JNBO:E>GM BFI:<M HG IKH;:;BEBMR 0A> Q:QBL LAHPL MA> O:EN> H? MA> ?>:MNK> >@ MA> =BLM:G<> MH G>:K
>LM K>?>KK>K #HK : @BO>G QO:EN> : IHLBMBO> IKH;:;BEBMR RO:EN> BG=B<:M>L MA> <EB>GM P:L FHK> EBD>ER MH IKHOB=> : K>?>KK:E 0A> F:@GBMN=> H? /%- O:EN>L BG=B<:M>L
MA> =>@K>> H? BGXN>G<> HG MA> IK>=B<MBHG .>F:BGBG@ ?>:MNK>L :K> NL>= BG IK>=B<MBHG ;NM GHM LAHPG ?HK <HG<BL>G>LL
DGHPG ?HK ?NMNK> BG?>K>G<> BM LMBEE F:R ;> BG?>:LB;E> MH K>EB:;ER @:MA>K
/H PABE> =:M: @:MA>K>= MAKHN@A : LNKO>R FB@AM BG<K>:L> IK>=B<MBO> I>K
?HKF:G<> BM <HNE= :ELH ;> BG?>:LB;E> MH <HEE><M :M L<:E> ?HK MAHNL:G=L H?
IHM>GMB:E <EB>GM AHNL>AHE=L L IHEB<RF:D>KL <HGLB=>K :IIKH:<A>L MH
K>F>=R BG>JNBMB>L BG :<<>LL MH K>G>P:;E> >G>K@R @>G>K:MBHG IKH@K:F
BFIE>F>GM:MBHG LMK:M>@B>L :G= M><AGBJN>L :L FN<A :L )*& BG<>GMBO>
IHEB<R =>LB@G PBEE LA:I> MA> K>:<A :G= I:<> H? IKH@K:F NIM:D>
0A> IK>=B<MBO> FH=>E K>LNEML A:O> IK:<MB<:E BFIEB<:MBHGL MH IK:<MB
MBHG>KL :G= MA>BK O:EN> LAHNE= ;> <HGLB=>K>= BG <HGM>QM #HK BGLM:G<>
:IIERBG@ MA> E>:= EHLL FH=>E L>> 3.3.1 BG IK:<MB<> PHNE= K>LNEM BG
 H? LM:?? MBF> MH :<JNBK>  H? MA> E>:=L  BFIKHO>F>GM
:G= MA> K>?>KK:E FH=>E L>> 3.3.2 PHNE= K>LNEM BG  H? LM:?? MBF> MH
:<JNBK>  H? MA> K>?>KK>KL  BFIKHO>F>GM 0AHN@A MA> FH=>E
BG<K>:L>L MA> >?W<B>G<R H? LM:?? MBF> I>K HNM<HF> BM K>LNEML BG ?>P>K HNM
<HF>L HO>K:EE ;R OBKMN> H? =BL<:K=BG@ <EB>GML IK>=B<M>= BG MA> G>@:MBO>
<E:LL 3A>MA>K MA> BG<K>:L> BG IK>=B<MBHG BL O:EN:;E> =>I>G=L HG <HG
M>QM #HK BGLM:G<> PA:M BL MA> HIIHKMNGBMR <HLM H? LM:?? MBF> &? MBF> HK
IKH@K:F K>L>KO>L :K> L:O>= ;R BFIKHO>= IK>=B<MBHG <:G MAHL> K>
LHNK<>L ;> :IIEB>= ?KNBM?NEER MH HMA>K F:MM>KL ELH MA> O:EN> <HNE= =>
I>G= HG MA> :FHNGM H? <HLM MH IKH<>LL : E>:= HK K>?>KK:E K>E:MBO> MH MA>BK
HK@:GBS:MBHG:E O:EN> #BG:EER MA> )*& IKH@K:F <HNE= ;> <HGLMK:BG>= ;R
:O:BE:;E> ?NG=L PAB<A PHNE= >FIA:LBS> HI>K:MBHG:E >?W<B>G<R HK ;R
MA> HO>K:EE L<:K<BMR H? E>:=L >FIA:LBSBG@ IK>L>KOBG@ E>:=L
0A> FH=>EL P>K> <:EB;K:M>= MH HIMBFBS> MA> balanced accuracy PBMA
:G >JN:E >FIA:LBL HG FH=>E L>GLBMBOBMR :G= LI><BW<BMR %HP>O>K :
IK:<MBMBHG>K FB@AM A:O> =B??>K>GM IKBHKBMB>L =>I>G=BG@ HG IKH@K:F
LMK:M>@R &G I:KMB<NE:K : AB@A>K >FIA:LBL HG FH=>E L>GLBMBOBMR <HNE= ;>
P:KK:GM>= B? IKH@K:F O:EN> BL =>KBO>= ?KHF <HKK><MER B=>GMB?RBG@ <EB>GML
EBD>ER MH IKHOB=> K>?>KK:EL :M MA> >QI>GL> H? AB@A>K ?:EL> IHLBMBO>L -K>
=B<MBHG H? K>?>KK:EL <HNE= ;> >LI><B:EER O:EN:;E> :L : F><A:GBLF ?HK
L>>=BG@G>P K>@BHGL HK B? MA> FH=>E <HNE= ;> :IIEB>= :M MA> E>:= @>G
>K:MBHG IA:L> /BFBE:KER ?HK E>:=L PAB<A HNM<HF> BL FHK> BFIHKM:GM MH
<HKK><MER IK>=B<M PAH PBEE ;><HF> EHLM HK GHM EHLM &G MA> ?HKF>K :
IK:<MBMBHG>K FB@AM :=CNLM MA>BK F:KD>MBG@ M:<MB<L MH IK>O>GM E>:= EHLL HK
O:KBHNLER BG<K>:L> ?H<NL HG AB@A IKHI>GLBMR E>:=L
LN;LM:GMB:E <HG<>KG H? F:<ABG> E>:KGBG@ BL MA> IHLLB;BEBMR H? FH=>E
;B:L MH MA> =>MKBF>GM H? <HFFNGBMB>L HK HNM<HF>L NG=>KK>IK>L>GM>= BG
MK:BGBG@ =:M: 78 0AHN@A P> LN@@>LM MA:M )*& IKH@K:F HI>K:MBHG:E
<HLML <HNE= ;> K>=N<>= MAKHN@A BFIKHO>= IK>=B<MBHG H? EHLM E>:=L :G=
K>?>KK:EL <E>:KER IK:<MBMBHG>KL LAHNE= IKH<>>= <:NMBHNLER 0H MA> >QM>GM

UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
MA:M MA>K> :K> =B??>K>G<>L BG IKH@K:F HNM<HF>L IK>=B<MBO> FH=>EL
<HNE= IK>?>K>G<> HG> LH<BH=>FH@K:IAB< @KHNI HO>K :GHMA>K E>:=BG@ MH
BG>JNBM:;E> IKH@K:F I:KMB<BI:MBHG 0H LHF> >QM>GM MABL BL :EK>:=R LN@
@>LMBO> BG MA> ?>:MNK>L MA> FH=>EL B=>GMBW>= :L A:OBG@ MA> FHLM IK>=B<
MBO> I>K?HKF:G<> >@ GNF;>K H? LNKKHNG=BG@ BGLM:EE:MBHGL K:<> HK
AHNL>AHE= LBS> -K>>QBLMBG@ ;B:L>L H? MA> IK:<MBMBHG>K <HNE= ;> BGMKH
=N<>= =NKBG@ FH=>E =>LB@G ?HK BGLM:G<> <AHB<> H? FH=>E ?>:MNK>L MA:M
BFIEB<BMER =BL<HNGM MA> >QI>KB>G<>L H? NG=>KK>IK>L>GM>= @KHNIL ,MA>K
<HG<>KGL BG<EN=> NL> H? FH=>EL BG :IIEB<:MBHGL MA:M IKBHKBMBS> IKH?BM HO>K
>JNBMR :G= ;B:L>= BGM>KIK>M:MBHG HK :IIEB<:MBHG H? : FH=>E >O>G B? MA>
FH=>E BML>E? BL NG;B:L>= %HP>O>K MA>K> :K> L>O>K:E IK:<MB<:E LM>IL MA:M
HG> <:G M:D> MH IK>O>GM =:M:=KBO>G FH=>EBG@ ?KHF ?NKMA>K >GMK>G<ABG@
>QBLMBG@ BG>JN:EBMB>L ;>MP>>G :=HIM>KL :G= GHG:=HIMBHG 0A>L> BG
<EN=> B >GLNKBG@ MA:M =:M: NL>= MH MK:BG FH=>EL :K> K>IK>L>GM:MBO> H?
MA> M:K@>M IHINE:MBHG :G= B? GHM BB HO>KL:FIEBG@ HG NG=>KK>IK>L>GM>=
IHINE:MBHGL MH <HGLMKN<M MA> MK:BGBG@ =:M: :G= BBB NL> H? >JNBMR;:L>=
JNHM:L ;R LH<BH=>FH@K:IAB< @KHNI ?HK IKH@K:F HNM<HF>L "GLNKBG@
>JNBM:;E> :<<>LL MH MA> ;>G>WML H? =><:K;HGBS:MBHG K>JNBK>L ;HMA : <:K>
?NE <HGLB=>K:MBHG H? MA> IKHFBL> H? =:M:=KBO>G F>MAH=L ;NM :ELH MA>BK
?:BK :IIEB<:MBHG
Declaration of Competing Interest
0A> :NMAHKL =><E:K> MA:M MA>R A:O> GH DGHPG <HFI>MBG@ WG:G<B:E
BGM>K>LML HK I>KLHG:E K>E:MBHGLABIL MA:M <HNE= A:O> :II>:K>= MH BGXN
>G<> MA> PHKD K>IHKM>= BG MABL I:I>K
Acknowledgements
0A> :NMAHKL MA:GD +:BF !:K@AHNMA ?HK :LLBLM:G<> BG <:E<NE:MBG@ BG
LM:EE>= ;:L> F>:LNK>L NLBG@ MA> GHGIN;EB< 0K:<DBG@ MA> /NG =:M: L>M
G:G= $HOBG=:K:C:G ?HK :LLBLM:G<> BG IKH<>LLBG@ =:M: :G= *:=>EBG>
$>H<:KBL ?HK >=BMBG@ 3> :ELH MA:GD 0HGR .>:F>L GGBD: 0H==EB<D
:G= (BF 3HELD> ?HK MA>BK ?>>=;:<D HG MA> LMN=R =>LB@G :G= MH $.&! E
M>KG:MBO>L LM:?? BG<EN=BG@ /:K:A N<<B '>??K>R HE>F:G 6:<A #K:GDEBG
:G= )BS: +H;>E ?HK MA>BK ?>>=;:<D BG BGM>KIK>MBG@ =:M:
0ABL PHKD P:L :NMAHK>= ;R MA> +:MBHG:E .>G>P:;E> "G>K@R ):;HK:
MHKR F:G:@>= :G= HI>K:M>= ;R EEB:G<> ?HK /NLM:BG:;E> "G>K@R )) ?HK
MA> 1/ !>I:KMF>GM H? "G>K@R !," NG=>K HGMK:<M +H !" 
$, #NG=BG@ P:L IKHOB=>= ;R MA> 1/ !>I:KMF>GM H? "G>K@R
,?W<> H? "G>K@R "?W<B>G<R :G= .>G>P:;E> "G>K@R /HE:K "G>K@R 0><A
GHEH@B>L ,?W<> 0A> OB>PL >QIK>LL>= BG MA> :KMB<E> =H GHM G><>LL:KBER
K>IK>L>GM MA> OB>PL H? MA> !," HK MA> 1/ $HO>KGF>GM 0A> 1/ $HO
>KGF>GM K>M:BGL :G= MA> IN;EBLA>K ;R :<<>IMBG@ MA> :KMB<E> ?HK IN;EB<:
MBHG :<DGHPE>=@>L MA:M MA> 1/ $HO>KGF>GM K>M:BGL : GHG>Q<ENLBO>
I:B=NI BKK>OH<:;E> PHKE=PB=> EB<>GL> MH IN;EBLA HK K>IKH=N<> MA>
IN;EBLA>= ?HKF H? MABL PHKD HK :EEHP HMA>KL MH =H LH ?HK 1/ $HO>KG
F>GM INKIHL>L 0ABL K>L>:K<A P:L I>K?HKF>= NLBG@ <HFINM:MBHG:E K>
LHNK<>L LIHGLHK>= ;R MA> !>I:KMF>GM H? "G>K@RL ,?W<> H? "G>K@R "?W
<B>G<R :G= .>G>P:;E> "G>K@R :G= EH<:M>= :M MA> +:MBHG:E .>G>P:;E>
"G>K@R ):;HK:MHKR
Appendix A. Supplementary data
/NIIE>F>GM:KR =:M: MH MABL :KMB<E> <:G ;> ?HNG= HGEBG> :M AMMIL
=HBHK@C>KLL
References
78 0$ .>:F>L !BLMKB;NMBHG:E =BLI:KBMB>L BG K>LB=>GMB:E KHH?MHI LHE:K IHM>GMB:E :G=
I>G>MK:MBHG BG ?HNK <BMB>L BG MA> 1GBM>= /M:M>L "G>K@R .>L /H< /<B  
 AMMIL=HBHK@C>KLL
78 ! /NGM>K / :LM>EE:GHL !* (:FF>G !BLI:KBMB>L BG KHH?MHI IAHMHOHEM:B<L
=>IEHRF>GM BG MA> 1GBM>= /M:M>L ;R K:<> :G= >MAGB<BMR +:M /NLM:BG  
 AMMIL=HBHK@LS
78 / HK>GLM>BG )3 !:OBL 0A> !BLMKB;NMBHG:E "??><ML H? 1/ E>:G "G>K@R 0:Q
K>=BML 0:Q -HEB<R "<HG     AMMIL=HBHK@

78 , /B@KBG *" *HHG>R .HH?MHI /HE:K 0><AGB<:E -HM>GMB:E ?HK )HPMH
*H=>K:M> &G<HF> %HNL>AHE=L BG MA> 1GBM>= /M:M>L +:MBHG:E .>G>P:;E> "G>K@R ):;
+.") $HE=>G , 1GBM>= /M:M>L 
78 * KHPG ' %N;;L 2 4BGRB $N *( A: .HH?MHI LHE:K ?HK :EE EHLBG@ MA>
@:I ;>MP>>G MA> M><AGB<:EER IHLLB;E> :G= MA> :<AB>O:;E> "G>K@R .>L /H< /<B 
  AMMIL=HBHK@C>KLL
78 $) :K;HL> +. !:K@AHNMA  %H>G .% 3BL>K &G<HF> 0K>G=L H? .>LB=>GMB:E
-2 =HIM>KL G :G:ERLBL H? AHNL>AHE=E>O>E BG<HF> >LMBF:M>L 
78  !K>AH;E ) .HLL )B?MBG@ MA> AB@A >G>K@R ;NK=>G BG F>KB<:L E:K@>LM <BMB>L
%HP >G>K@R >?W<B>G<R <:G BFIKHO> EHP BG<HF> :G= NG=>KL>KO>= <HFFNGBMB>L

78 / BK= ! %>KGTG=>S -HEB<R HIMBHGL ?HK MA> LIEBM BG<>GMBO> &G<K>:LBG@ >G>K@R
>?W<B>G<R ?HK EHPBG<HF> K>GM>KL "G>K@R -HEB<R   
78 ( $BEEBG@A:F * %:K=BG@ ! .:ILHG /IEBM &G<>GMBO>L BG .>LB=>GMB:E "G>K@R
HGLNFIMBHG "G>K@R '    AMMIL=HBHK@
78 '' HHD ) BK= 1GEH<DBG@ LHE:K ?HK EHP :G= FH=>K:M>BG<HF> K>LB=>GML
F:MKBQ H? WG:G<BG@ HIMBHGL ;R K>LB=>GM IKHOB=>K :G= AHNLBG@ MRI> +:MBHG:E
.>G>P:;E> "G>K@R ):; +.") $HE=>G , 1GBM>= /M:M>L  AMMIL
=HBHK@
78 * #HPEB> * $K>>GLMHG> 3HE?K:F K> MA> GHGFHG>M:KR <HLML H? >G>K@R
>?W<B>G<R BGO>LMF>GML E:K@> 1G=>KLM:G=BG@ EHP M:D>NI H? : ?K>> >G>K@R >?W<B>G<R
IKH@K:F F "<HG .>O     AMMIL=HBHK@
:>KI
78 0$ .>:F>L <HFFNGBMR;:L>= :IIKH:<A MH EHPBG<HF> K>LB=>GMB:E >G>K@R
>?W<B>G<R I:KMB<BI:MBHG ;:KKB>KL )H<:E "GOBKHG    
AMMIL=HBHK@
78 $ /HNMAP>EE ' *NKIAR 3>:MA>KBS:MBHG ;>A:OBHK :G= LH<B:E <HGM>QM 0A>
BGXN>G<>L H? ?:<MN:E DGHPE>=@> :G= LH<B:E BGM>K:<MBHG "G>K@R .>L /H< /<B
  AMMIL=HBHK@C>KLL
78 (/ 3HELD> *HK> :EBD> MA:G =BY>K>GM -KH?BE>L H? AB@ABG<HF> :G= EHPBG<HF>
KHH?MHI LHE:K :=HIM>KL BG MA> 1GBM>= /M:M>L "G>K@R .>L /H< /<B   
AMMIL=HBHK@C>KLL
78 ) %B@@BGL ) )NMS>GABL>K >K>FHGB:E "JNBMR )HPBG<HF> >G>K@R :LLBLM:G<> :G=
MA> ?:BENK> H? LH<BH>GOBKHGF>GM:E IHEB<R /H< -KH;E    
AMMIL=HBHK@
78 /& /BG@E>#:FBER ??HK=:;E> /HE:K %HF>L //% -KH@K:F G= AMMIL
PPP<IN<<:@HO$>G>K:E:LIQB= :<<>LL>= N@NLM  
78 )HP&G<HF> /HE:K -HEB<> $NB=> :EB?HKGB: )HP&G<HF> /HE -HEB<R $NB=>
G= AMMILPPPEHPBG<HF>LHE:KHK@;>LMIK:<MB<>LLBG@E>?:FBER<:EB?HKGB:
:<<>LL>= N@NLM  
78 ' HE>F:G NMAHKLBGM>KOB>P PBMA '>YK>R HE>F:G $.&! EM>KG:MBO>
,NMK>:<A 2- 
78 3HH=*:<D>GSB> 1/ .>LB=>GMB:E /HE:K -2 NLMHF>K <JNBLBMBHG  NKK>GM
:G= -KHC><M>= HLML :G= A:GG>E /MK:M>@B>L 
78 * *H>SSB  &G@E> ) )NMS>GABL>K , /B@KBG +HG*H=>EBG@ "QIEHK:MBHG H?
.>LB=>GMB:E /HE:K -AHMHOHEM:B< -2 =HIMBHG :G= +HG=HIMBHG +:MBHG:E
.>G>P:;E> "G>K@R ):; +.") $HE=>G , 1GBM>= /M:M>L  AMMIL
=HBHK@
78  HEEBG@>K ( $BEEBG@A:F ->>K >??><ML BG MA> =BYNLBHG H? LHE:K IAHMHOHEM:B<
I:G>EL *:KD /<B     AMMIL=HBHK@
FDL<
78 (/ 3HELD> -  /M>KG 0 !B>MS "QIE:BGBG@ BGM>K>LM BG :=HIMBG@ K>LB=>GMB:E LHE:K
IAHMHOHEM:B< LRLM>FL BG MA> 1GBM>= /M:M>L 0HP:K= :G BGM>@K:MBHG H? ;>A:OBHK:E
MA>HKB>L "G>K@R .>L /H< /<B    AMMIL=HBHK@
C>KLL
78 ' 5N 6 3:G@ . .:C:@HI:E ' 5N 6 3:G@  *:CNF=:K . .:C:@HI:E
!>>I/HE:K F:<ABG> E>:KGBG@ ?K:F>PHKD MH >?W<B>GMER <HGLMKN<M : LHE:K
=>IEHRF>GM =:M:;:L> BG MA> NGBM>= LM:M>L =>>ILHE:K F:<ABG> E>:KGBG@
?K:F>PHKD MH >?W<B>GMER <HGLMKN<M : LHE:K =>IEHRF>GM =:M:;:L> BG MA> 1GBM>=
/M:M>L 'HNE>   AMMIL=HBHK@CCHNE>
78 % 6A:G@ 5 2HKH;>R<ABD ' )>M<A?HK= ( ):DD:K:CN !:M:=KBO>G :@>GM;:L>=
FH=>EBG@ PBMA :IIEB<:MBHG MH KHH?MHI LHE:K :=HIMBHG NMHG @>GML *NEMB@>GM
/RLM     AMMIL=HBHK@L
78 * *BE=>G;>K@>K -! %HP> *BEC:GB<A %HNL>AHE=L PBMA LHE:K BGLM:EE:MBHGL
:K> B=>HEH@B<:EER =BO>KL> :G= FHK> IHEBMB<:EER :<MBO> MA:G MA>BK G>B@A;HNKL +:M
"G>K@R    AMMIL=HBHK@L
78 !:OB=LHG " !KNKR  )HI>S . "EFHK> . *:K@HEBL *H=>EBG@ IAHMHOHEM:B<
=BYNLBHG :G :G:ERLBL H? @>HLI:MB:E =:M:L>ML "GOBKHG .>L )>MM  
78 (/ 3HELD> (0 $BEEBG@A:F -3 /<ANEMS ->>K BGXN>G<> HG AHNL>AHE= >G>K@R
;>A:OBHNKL +:M "G>K@R    AMMIL=HBHK@L

78  />D:K 2 .:B -K>=B<MBG@ ):M>GM HGLMKN<ML ?KHF -:LLBO> !:M:L>ML
/B@GB?B<:G<> :G= ,IIHKMNGBMB>L /H<B:E /<B>G<> .>L>:K<A +>MPHKD .H<A>LM>K +5
 AMMILI:I>KLLLKG<HF:;LMK:<M :<<>LL>= N@NLM  
78  -:EF ->>K >??><ML BG K>LB=>GMB:E LHE:K IAHMHOHEM:B<L :=HIMBHG FBQ>=
F>MAH=L LMN=R H? /P>=BLA NL>KL "G>K@R .>L /H< /<B    AMMIL
=HBHK@C>KLL
78 * $K:SB:GH ( $BEEBG@A:F /I:MB:E I:MM>KGL H? LHE:K IAHMHOHEM:B< LRLM>F
:=HIMBHG 0A> BGXN>G<> H? G>B@A;HKL :G= MA> ;NBEM >GOBKHGF>GM ' "<HG $>H@K
    AMMIL=HBHK@C>@E;N
78 ! +HEE !:P>L 2 .:B /HE:K HFFNGBMR ,K@:GBS:MBHGL :G= :<MBO> I>>K >??><ML
BG MA> :=HIMBHG H? K>LB=>GMB:E -2 "G>K@R -HEB<R    AMMIL
=HBHK@C>GIHE
78  /B@KBG ' -E>LL " !KNKR !BYNLBHG BGMH G>P F:KD>ML >OHEOBG@ <NLMHF>K
L>@F>GML BG MA> LHE:K IAHMHOHEM:B<L F:KD>M "GOBKHG .>L )>MM   
 AMMIL=HBHK@

UNCORRECTED PROOF
B. Sigrin et al. Energy Research & Social Science xxx (xxxx) 102417
78 " 3BEEB:FL . :KO:EAH " %BMMBG@>K * .HGG>G;>K@ "FIBKB<:E =>O>EHIF>GM H?
I:KLBFHGBHNL FH=>E ?HK BGM>KG:MBHG:E =BYNLBHG H? K>LB=>GMB:E LHE:K .>G>P "G>K@R
   AMMIL=HBHK@CK>G>G>
78 0 A>G $N>LMKBG 4$HHLM L<:E:;E> MK>> ;HHLMBG@ LRLM>F BG -KH< G=
 * /&$(!! &GM HG? (GHPE !BL<HO !:M: *BG (!!   II 

78 1/ >GLNL NK>:N  F>KB<:G %HNLBG@ /NKO>R  AMMIL
PPP<>GLNL@HOIKH@K:FLLNKO>RL:AL
78 $ :K;HL> + !:K@AHNMA / "EF:EE:A / #HKK>LM>K ( (KBLMBG: / % ! *BEELM>BG
' .:G= 3 HMMHG / /A>KPHH= " ,/A:N@AG>LLR 0K:<DBG@ MA> /NG -KB<BG@ :G=
!>LB@G 0K>G=L ?HK !BLMKB;NM>= -AHMHOHEM:B< /RLM>FL BG MA> 1GBM>= /M:M>L 
"=BMBHG ):PK>G<> >KD>E>R +:MBHG:E ):; )+) >KD>E>R 1GBM>= /M:M>L
 AMMIL=HBHK@
78 ,"%% :E"GOBKH/<K>>G  ,"%%  AMMILH>AA:<:@HO
<:E>GOBKHL<K>>GK>IHKM<:E>GOBKHL<K>>G :<<>LL>= N@NLM  
78 / (:N?F:G / .HLL>M ->KEB<A )>:D:@> BG =:M: FBGBG@ ?HKFNE:MBHG =>M><MBHG
:G= :OHB=:G<> BG -KH< MA  * /&$(!! &GM HG? (GHPE !BL<HO !:M: *BG
LLH<B:MBHG ?HK HFINMBG@ *:<ABG>KR +>P 5HKD +5 1/  II 
AMMIL=HBHK@
78 ) K>BF:G .:G=HF #HK>LML *:<A )>:KG   
78 ' "EBMA '. )>:MAPB<D 0 %:LMB> PHKDBG@ @NB=> MH ;HHLM>= K>@K>LLBHG MK>>L '
GBF "<HE     AMMIL=HBHK@C
Q
78 /* )NG=;>K@ /& )>> 1GB?B>= IIKH:<A MH &GM>KIK>MBG@ *H=>E -K>=B<MBHGL
BG =O +>NK:E &G? -KH<>LL /RLM NKK:G LLH<B:M>L &G<  AMMIL
IKH<>>=BG@LG>NKBIL<<I:I>KA:LA::=<=?=;
;LMK:<MAMFE :<<>LL>= />IM>F;>K  
78 * (NAG NBE=BG@ -K>=B<MBO> *H=>EL BG . 1LBG@ MA> <:K>M -:<D:@> ' /M:M /H?MP
KMB<    AMMIL=HBHK@CLLOB
78  *HG= 1/ .>LB=>GMB:E /HE:K -2 NLMHF>K <JNBLBMBHG  NKK>GM :G=
-KHC><M>= HLML :G= A:GG>E /MK:M>@B>L $0* .>L>:K<A 
78 ' /<:OH / (HKHL>< " $N>KK>KH  ->GGBG@MHG - !HN@AF:G )HP&G<HF>
:KKB>KL /MN=R -:KM ,O>K<HFBG@ :KKB>KL MH "G>K@R "?W<B>G<R :G= .>G>P:;E>L
?HK )HP&G<HF> NLMHF>KL :G= /F:EE NLBG>LL HGMK:<MBG@ ,IIHKMNGBMB>L BG
!BL:=O:GM:@>= HFFNGBMB>L 
78 $ /AFN>EB 0H "QIE:BG HK MH -K>=B<M /M:M /<B    AMMIL
=HBHK@/0/
78 / *NEE:BG:MA:G ' /IB>LL *:<ABG> )>:KG IIE "<HG II   
78 /* )NG=;>K@ $$ "KBHG /& )>> HGLBLM>GM &G=BOB=N:EBS>= #>:MNK>
MMKB;NMBHG ?HK 0K>> "GL>F;E>L G= 
78 KB:G =E>LL:G=KH :MAR ,+>BE 0HF ):$:MM: HGL<B>GMBHNL <E:LLBW<:MBHG
=:M: L<B>GMBLML @NB=> MH =BL<KBFBG:MBHG:P:K> <E:LLBW<:MBHG B@ !:M:  
 AMMIL=HBHK@;B@
78 2 "N;:GDL NMHF:MBG@ &G>JN:EBMR %HP %B@A0><A 0HHEL -KH?BE> -HEB<> :G=
-NGBLA MA> -HHK /M *:KMBGL -N;EBLABG@ $KHNI 

... For instance, houses of worship directly engage with and can influence the behavior of their congregations (Van Cappellen et al., 2016) and can serve as conduits for public policy implementation (Evans and Hudson, 2014;Flórez et al., 2017). Third, installers or policymakers could leverage non-residential influence by installing "seed" systems, following the rationale that installing a seed system promotes subsequent adoptions (Zhang et al., 2016;Sigrin et al., 2022). Installers could use seeding to promote business in new markets, and policymakers could use seeding to promote deployment in underserved communities, an increasingly common policy objective (Carley et al., 2021). ...
Article
Full-text available
Household decisions to adopt rooftop solar photovoltaics are partly driven by social influence. Previous research on solar adoption influence has focused on influence among residential peers. Here, we expand the framework of solar adoption influence by exploring the influence of non-residential installations on residential adoption decisions. We use staggered differences-in-differences to estimate non-residential influence effects using a large data sample of residential adoptions. We also critically evaluate prevailing frameworks for solar adoption influence. We find that non-residential installations are associated with accelerated residential adoption rates, on the order of 0.4 additional residential adoptions per quarter per non-residential installation. We show that non-residential systems exert a continuous, long-term influence on residential adoption decisions. We explore separate results and influence mechanisms for solar installed on commercial buildings, government buildings, and houses of worship. The results suggest that non-residential solar adopters could serve as partners in policies to “seed” residential adoption in underserved communities.
... Recognizing that referred households were more likely to go solar than cold contacts 35 , GRID Alternatives created a referral rewards programme: existing LMI solar adopters are offered a conditional US$200 incentive for each referred household that installs solar. Such a programme is in line with the for-profit solar industry, though the latter typically offers higher rewards (US$500-2,000). ...
Article
Full-text available
Subsidized energy assistance programmes are a popular policy tool for promoting energy justice, but, like other social benefits programmes, are often undersubscribed. To improve uptake, some programmes have turned to social influence strategies, such as asking programme participants to refer their peers. Here, through a field experiment with California’s low-income solar programme (N = 7,676), we show that referral behaviour depends on how existing participants are approached. Adding behavioural science strategies to a referral reward increases peer referral rates, referral quality and ultimately solar adoption. Compared with only reminding existing adopters of a potential US200rewardforreferralsthatresultinadoption,addinganappealtoreciprocitythroughanoncontingentUS200 reward for referrals that result in adoption, adding an appeal to reciprocity through a non-contingent US1 gift—and further combining this gift with a simplified referral process—leads to 2.6–5.2 times as many solar contracts. These results highlight the potential of behaviourally informed peer referral programmes to accelerate equitable access to clean energy.
Article
Full-text available
Studies across multiple disciplines demonstrate the importance of peers in shaping energy-related behaviours. Research on this process is wide ranging, from documenting spatial peer effects in the adoption of rooftop solar—when an individual’s behaviour is influenced by the behaviours of neighbours—to showing how neighbour comparisons can be used to reduce household electricity consumption. However, gaps exist in our understanding of how and why these peer effects occur. In this Review, we examine recent findings on social influence in energy behaviour and discuss pathways through which social influence can result in peer effects. We propose a conceptual framework for predicting which social influence processes will most often result in peer effects, depending on the targeted energy behaviour. We also review the limitations of social influence as well as evidence for when it is expected to be the strongest.
Article
Full-text available
Climate risk mitigation requires rapid decarbonization of energy infrastructure, a task that will need political support from mass publics. Here, we use a combination of satellite imagery and voter file data to examine the political identities of US households with residential solar installations. We find that solar households are slightly more likely to be Democratic; however, this imbalance stems primarily from between-neighbourhood differences in partisan composition rather than within-neighbourhood differences in the rate of partisan solar uptake. Crucially, we still find that many solar households are Republican. We also find that solar households are substantially more likely to be politically active than their neighbours, and that these differences in political participation cannot be fully explained by demographic and socioeconomic factors. Our results demonstrate that individuals across the ideological spectrum are participating in the US energy transition, despite extreme ideological polarization around climate change.
Article
Full-text available
The rooftop solar industry in the United States has experienced dramatic growth—roughly 50% per year since 2012, along with steadily falling prices. Although the opportunities this affords for clean, reliable power are transformative, the benefits might not accrue to all individuals and communities. Combining the location of existing and potential sites for rooftop photovoltaics (PV) from Google’s Project Sunroof and demographic information from the American Community Survey, the relative adoption of rooftop PV is compared across census tracts grouped by racial and ethnic majority. Black- and Hispanic-majority census tracts show on average significantly less rooftop PV installed. This disparity is often attributed to racial and ethnic differences in household income and home ownership. In this study, significant racial disparity remains even after we account for these differences. For the same median household income, black- and Hispanic-majority census tracts have installed less rooftop PV compared with no majority tracts by 69 and 30%, respectively, while white-majority census tracts have installed 21% more. When correcting for home ownership, black- and Hispanic-majority census tracts have installed less rooftop PV compared with no majority tracts by 61 and 45%, respectively, while white-majority census tracts have installed 37% more. The social dispersion effect is also considered. This Analysis reveals the racial and ethnic injustice in rooftop solar participation. © 2019, The Author(s), under exclusive licence to Springer Nature Limited.
Article
Full-text available
Latent constructs are fundamental for theory-building and practice in various disciplines of the social sciences. Passive datasets and machine learning approaches provide an opportunity to predict latent constructs, thereby offering a strong potential to build scalable, granular, and theory-grounded models for various phenomena of interest in social sciences. However, there is little guidance for researchers to develop prediction models for latent constructs, especially given that latent constructs are different from observable variables in several keys ways. To address this gap, and drawing from both theoretical and applied literature across the social sciences, we first develop a taxonomy to classify latent construct models based on their ontology and measurement. Second, we synthesize the relevant literature on using passive data to predict latent constructs. Finally, by applying the developed taxonomy to various applications in the literature, we offer guidance for predicting latent constructs from passive data. By bringing together loosely tied research in the context of scaling up latent construct models, this paper makes two significant contributions: 1) identify key significant and specific gaps in this research area, 2) provide theoretically- and empirically-grounded guidance for social scientists interested in using passive data for developing scalable latent variable models.
Technical Report
Full-text available
This report serves to expand upon previous NREL research investigating the technical potential of rooftop solar in the United States, aiming to improve the understanding in the residential sector, particularly for low-to-moderate income households. Technical potential is a metric that quantifies the maximum generation available from a technology for a given region and does not consider the economic or market viability. A unique contribution of this work is to estimate rooftop solar technical potential of residential buildings per U.S. Census Tract by income, building type, and tenure. The underlying data—as documented in Gagnon et al. 2016—are a series of light detection and ranging (LiDAR)–based scans of the building stock in 128 metro regions. Using LiDAR scans, as opposed to aerial imagery, allows researchers to infer the building footprint and the unshaded roof area, azimuth, and tilt for each distinct roof plane, although roof age or other structural concerns are not considered. This data then is intersected with U.S. Census Bureau socio-demographic and building stock data at the tract-level to better understand how rooftop solar technical potential is allocated among different building types. This research also uses statistical techniques to estimate rooftop potential in areas not covered by the LiDAR scans. Collectively, these data and methods are used to address three research questions: 1) What is the quantity and spatial distribution of rooftop technical potential, stratified by income, building type, and tenure? 2) Among low-to-moderate-income households, what is the feasibility of achieving parity in solar access across income groups? 3) What is the quantity of technical potential for non-profit buildings, and to what extent might these buildings “oversize” systems on their roofs to export excess generation to LMI households?
Article
Estimating the technical potential of carbon-abatement options involves straightforward calculations, while estimating the achievable potential is more challenging. We illustrate this by examining solar photovoltaics (PV). We estimate that a 92% gap exists between the achievable and technical potential for rooftop solar in Georgia and it is shrinking slowly, while in contrast the state’s solar farms are rapidly expanding. Closing the gap between the technically possible and achievable levels of solar PV requires an understanding of why some electricity providers promote utility-scale solar and not solar rooftop systems. Our financial analysis suggests that rooftop PV on buildings in Georgia could deliver substantial economic benefits to host customers, if utilities offered favorable net metering rates and if host customers, in turn, curtailed their consumption. Comparing the slow diffusion of rooftop solar in Georgia with California and Massachusetts that are rapidly solarizing, underscores the need to develop methodologies that account for utility business models, policy interventions, and behavioral factors. A sensitivity analysis of alternative net metering policies, rebound and curtailment behaviors, a stakeholder analysis, and an assessment of barriers and accelerants documents that these factors are as important as techno-economic drivers in explaining solar technology transitions. Given the numerous barriers to adoption by low- and moderate-income households, an equitable solar transition requires business models and policies that foster participation by all.
Article
Single-family residential rooftop solar adoption in the United States has not occurred equitably across the country, nor across socioeconomic and demographic groups. In response, state and local governments have developed solar equity programs, primarily focused on increasing adoption by low- and moderate-income (LMI) households. This study merged national datasets that estimate rooftop solar potential, the distribution of rooftop solar systems, and census tract- level socioeconomic and demographic characteristics to answer three questions. First, how are spatial distributions of rooftop potential and penetration similar and different across cities? Second, how is rooftop penetration distributed across non-LMI and LMI communities in different cities? Third, how do the relationships between rooftop penetration and local socioeconomic and demographic characteristics, identified as barriers to solar adoption, differ? Using GIS, bivariate and multivariate analyses, these questions were examined in four US cities – Riverside and San Bernardino, California, Washington, DC, and Chicago, Illinois – to understand both universal and distinct local manifestations. Findings include: higher rooftop potential existed in some LMI communities; higher rooftop potential did not necessarily translate to higher rooftop penetration, especially if higher potential was in LMI communities; and beyond income, other socioeconomic and demographic characteristics such as race/ethnicity, limited English proficiency, age of housing stock, and internet access were associated with rooftop penetration. While there remains great potential for expanding rooftop solar to LMI households and communities, understandingexplor the local dynamics of solar potential and penetration may inform better policy development and implementation.
Article
The U.S. residential rooftop solar market has seen rapid growth in the last decade, much of it concentrated among higher-income households. To address inequities in solar access, a growing number of policies aim to put solar on the roofs of low-to-moderate income (LMI) households. As much of the extant literature on solar adoption is based on the behavior of wealthier and more highly-educated households, questions arise about how to best reach potential LMI adopters. Do the same factors guide their decision-making, or do LMI households have fundamentally different motivations and barriers? To answer these questions, this study surveyed over 500 low-income adopter households in California who received their PV systems for free through a nonprofit. Their responses were compared to those in an existing dataset of high-income PV adopters in the same state. Results suggest that high-income and LMI solar adopters are more alike than not. Both groups are drawn to novel goods and have strong pro-environmental norms, with LMI households scoring slightly higher on these measures. The relative importance of different motivations for and barriers to getting solar were also similar. Taken together the results suggest that even when PV is highly-subsidized, it may be most attractive to people who fit the profile of “early adopters.”
Article
We develop a new parsimonious model of residential solar diffusion that, with only two regression parameters and one independent variable, reasonably explains empirical observations. Additional solar customers resulting from an increase in Net Present Value (NPV) are modeled as a normal distribution. This leads to adoption as a function of NPV being the integral of the Gaussian, producing the error function, which demonstrates S-curve behavior commonly seen in technology diffusion. Empirical analysis for five regions (three U.S. states: Arizona, California, and Massachusetts; and two countries: Germany and Japan) from 2005 to 2016 shows a consistent relationship between annual adoption per million households and NPV. Non-linear regression indicates good agreement between data and the error function model, the adoption rate peaking at an NPV of 7100/kWwithstandarddeviationof7100/kW with standard deviation of 4110/kW. Consumer purchases of rooftop solar across multiple regions are explained with a single variable, making this model simpler than traditional diffusion approaches. A novel implication of the model is that the subsidy cost to stimulate additional solar adoption increases as the technology becomes cheaper. This is because the same subsidy is paid to all consumers, including those who would have purchased solar without subsidy.
Article
We developed DeepSolar, a deep learning framework analyzing satellite imagery to identify the GPS locations and sizes of solar photovoltaic panels. Leveraging its high accuracy and scalability, we constructed a comprehensive high-fidelity solar deployment database for the contiguous US. We demonstrated its value by discovering that residential solar deployment density peaks at a population density of 1,000 capita/mile², increases with annual household income asymptoting at ∼$150k, and has an inverse correlation with the Gini index representing income inequality. We uncovered a solar radiation threshold (4.5 kWh/m²/day) above which the solar deployment is “triggered.” Furthermore, we built an accurate machine learning-based predictive model to estimate the solar deployment density at the census tract level. We offer the DeepSolar database as a publicly available resource for researchers, utilities, solar developers, and policymakers to further uncover solar deployment patterns, build comprehensive economic and behavioral models, and ultimately support the adoption and management of solar electricity.