Report No. 17161-BIH Bosnia and Herzegovina Public Expenditure Review (in Two Volumes) Volume II: Technical Annexes November 26,1997 . Poverty Reduction and Economic Management Unit Europe and Central Asia Region "i1.si f" ;:..E .E- 3, E I:i , E " E, : I ; :,, , E , 5" ti,l F: , i 'i ,,i,, ;;,;, 3, E !c ; j j  j , ,i,3, E !  E, E",3 rT,!,iriili : E 1:  ! ;,  :  i :!4 1,. . ,,,5 :3,:; , ;, !w,'? ?  -E] ,tt ::.. E: ;E.iE.: , :Eii .E .E .: Ei .i l': j .E,J !Ii"','",,,!:iii,:,.. 11,`.I,!i E! -t, i,3! ti,r% ,:!. ii i.ii: i! ;,i :1? - ;i E ; 1 : ,-  E1 ; i i i't 1 ,i!k!It !Ji!YTt:::.iF ::E.EqEE.E3;:IEE! rYiiv;f ri,E: :i ., :: .El : 'i : : - , :;,,i .. .,!t.: :: .i" 1; !  i E,i  E .i 'El - EE-i1, .:.,: 'E. :E:i E. ::: ;i::: :: Ei E,EE .: E. :t,ti E: :i E 'i :tv': .: . -: ..lE:: - :; .2iiiEi-E,EE.:. . .I : . . ....- -E, ii:E!1:1 .;Et,.5: .j . : t.EE:. .:E,;::::;;E:z, ;i:!:E.:.I :1. .::: ::,::: .i;,....-. ,- ...! .Ej:;E .i::.: i: :E:iEE ;!:E:EEi it :,E;: .i E: :jt,EEit :: . !.iiE!  ! !i i t ! I i:. :: E, i.: E.EF ': tEE' E -; it .E I it ! .E.:::- it i E, :E;iit :t,l t t!!ii; .;i. Et ;::it ,E i :! :,.i'! .i it : i i: .1 '! i,1tEl .t ;EE,"t::;.E .:.E:.:EE:i J:.,,.2 .:i 1 : 1 E: i  ?", 1 ! ti ! Eii 1 :. i , i  1....I-:... .,,.,- ! ', E ,i.:  :5::;. :E ECE:;I: :; 'i,,E EJjE:. :. !: it ,E:. -,  . ! . E ; 1!-,E ::: :. E-, - :. :. .:. i. E, t ;!: ,-. :E E. EE  tj iEz,::-.'i E:1.1 E.,, ....I.EEE: :,.::; ,-i,tti;,E - .E:,l. 1! li .!i;i', :: :; .::i,': i;-Eit J1j! J! it i .1 .;E . ij:E i .E tE! E, E.: .:0 ,:q E;:,EI.ltl :E....,, -:., .: .E ....E:.....]; E , EEE: 1: : .-. ::,::.tt:.E::t .E' :EE'::;,E,  i: .: ,:::,,i .: I:t: I I::..EEE: 'i;ii: -i..::::::ii i !J.i,,,i .EE: :i. Ei ; ! izi ;E :-JiJ.EE, ;:-E.... ,t ::--,.V:E- :E:; .;  z: :1E :::- ..,.1 iij1'O :,,;[Lj!i!,Ie - .-ii :1V! i Ad :l'!:. ,,.,,IE E;:5.7,E.::tzi:s :: .: ;i :: . ': :: t:,?i E!. Pti:i:!:,,,! Eii-ijji Jj!":!"i:;j-i, ITi :1, ; :" E, !i Ei"11 I, i i E : !  I,,",. :E . E:,.3-11,I-, ,;  ,,,  iji, : 3 , : E, E : ,L,:, :11 I:1- , i: Ei t1 'L 1. 1..'E  - E .1 it &t:; ,; ji,, ,j ii E F ,, t, i ji,,, E E; 2: ,: .E.- :l .E. :,:..:,.,: C ::E' :E ,EEi 5i" 5; .:i: E: ,.iz,: .: :: :,. ,,Y ,.-. !-  , :EE;; .  t ,.,.;i . .. E ', .: EEE,E;: .. :lzE:E, :E ;E , '! .i .t :t,Et1,;,,! J'. :O,., :,: j, :: ':,:::z?-,.:z.::E.:..:7:jIEEE.t, ...iE :.E E E: EEE,Eiit .:'i :--E.:EEr : E, :j. .: ,.E i,..:E". ., ..it .El,E,i.i '!ji! it, 't!. "i i ;T E-.. ,EE:E ., iF..iiE': Z.:::...:.:; .. ..: ii E: E, iljji: ,i ii .IE. .Z :EEE;i4iE: E: ,,j :E 1:-i:::E;:": E. _... .. . . .... .1j;:E.1: :EE:i.:E-,.,i 'i .E..E 'E. '. .!iii:EiiiijjEitiit iEji ;: j,- ..;..:.;,,iE-:i:E.::::EltE_....! A:tiE.tz .im: E ': ,.O."t ? , El, ': : E.EE:t ,E .::it ::E.,::E-::iE.; ,; :i:.;EE: E: El E?EE.E::A, iE: i :i it7i  h i; ; .! :: E: EEE:E ': liEEt;? -s Z, .-,.., .E ,;E E:... ... ,;;yq, ; ,   " E ", It jit dJit:u", "':!, E ,I,,,,,E1j,.: : t, :;E: : E! : t;,-E E: 1E. t, E E'iEE !,-,-,,": ]: j  . E:: E ii :%"s...,.,J:LE :OJ, P,LE :.E'i ; ,E E !  ,?  :: : ,.-,:ts] : ,cii :! , -,,  '? -t E- -., -;.;,. :; i-it t ,:..:I t: :. :, ;:', " E. zE,E. :! E EEs .-i  E, t, IEEEit .:: :....)  ii ti,, EE. :Z: :;:. . ..it : :..:.. ,E-E,o 'i  .. E;E...,.:O.:ti:EE. ':  EE: Er, :E:E.? .i .:.;,E;. !.mi., ,::-. F E,: 'i . EE: ., E, i0: ,  ... . .:;; .2:'iEE.,:I,.: .:Ei IE i :: .;EE :E :::E.E:.:,::::.:-:-,E:.:::- i%tE,:i E, :: .t 7EE. ..E .: . E;;:I.1 t:: : .; : :::::E JE::E:: ,:,::?EE;i:.!:EEi!,EiE:E : iti. ..EEE.:,E, :E:E,' : : E.: .:E--:.::%iE:.:,::.: iE.t.:Ez.; .." -,,.:O.E:: :; .;F:: iz E, .EEE;J;E: 'iE,.,o :i :: .j " E: .i E .,;:, : .:E:!,:E it ,Ei  " i, li il,i i ,i ,.!,,,,, i]L.E ;iL i, :, E: E, -, j ;j ,, Q i p;,i:, i%:: i !;: U- ii,!!,!! .!J,,!,! it,! it !, L,:! 1 ii,,!j!,,!Ll!,i :EE1.  E .. o i: : ,E . :"i- E `2 "El:', : ; :  , : - :: :i E- .: '. E.: E  .t: - E---- ,  v",;,irt,:k,EE t, EEY,i!,,U i EiEU!,! ;1i:,&LL',::d,i: .,..-': .il .: .E .t i:,! i, E: :i .E:E:.,-:.,,.i"':.: : '.";... .:E.-i:i EEW': .iiE; EEE:..EE:; .:.7t:,,E;,:t': ,:' .:.o,itE; : ,. i !: !i  jI i :i;iii,i7iEE ; iiI1: ! iiE]!!.,] :i .::.: :;..:,.: , ': ,:1 7E ;EEI.EF .:!.i!E:JE: .:. 't 1t .E.: .:,: .iE1:it1-.,E:::; .;E5t i:::::-::... ;::EE:, ;': .:;:I-.- E:  ... .: :,,EE t4  : .E E jl..jjiiE, : ...:.,-..:I:-.: 'IEi !i J'j 1!i t E: tt .iEE;.:E:;E.E4 :, -:,E ,.: .,.; : ::,EE,:.:.EEi;:; :t.: .: ;:;i'lE:i 'i E ,; ,:' . ,: ;' .E ,:  ,;;E, .1 i : :::': ;::7:;,:,..;EE: ::;.;iEitE1E:EE:EE:EiE'i EE.: .! E'. ...E, .i:- ,E:: ..E -E .E:E. :E E.: '::,:. : .; ,:: ,.EE,:;1, - :-:EE:E::i:EE;,,. ':E::: ..i E. z,;:it ,E ,E :E.E,..I :.,:,:;E, , E.pE:E.:Exi;E..E: EI ::;.:. :.: i il:: ,EE.:.:tj, i i7]Ei . ,j j ;::i:  .: - :; .-,: j E,::E:iEI.:.:.EiEt,iF:E: :E .t -:,.; ....E:::,Et ':t sEEE I, EO-E.:E;::: 7E,;iit.:,,E.; .i .; :E ,: ti ; J, :i 'i E . :-.: ,; ,;  :.:EEii,; .; .: : 7 ; E,. .; :. .:,;.EE - ;;: .E :. : :::.1.::-i5,; E .:iE;.tEEl::.:.;-'!:- :E.i;-': ,E :: :::.E.E ,:.EE 'i zEi E E:  E .': - iE:EE- .EEE : .:::EZ. ,:. EIt:::'i ji -i E! :i%-!,,:_A:j! E,l;z ,;,ii ., .j !t E: .i Jjt .! : ;E!,;I E! -],! !:,i,. i 'i1j:,: ,E;:']. !,: .::E ,,:1,j: :jj,;E J', !, ....EE,;,OEE, : . 3, j, .:EE,3, ,E itE,:,,z,E, I :E5'! : .? ,i :i : Ett':1:itEttEEE:i ':!E: ::EEli E,: ti .i:::.E .E.E, .-,E.. Ei .,;E- :. :-E ,: ze , E:E.:EE:;-E E,;:i ; it::E, 7 , , :..E:: E :-,p. ,iE!i%t .ii ;:J!EiE.i .1 ] .1i 1:: .i .'E .: ': 't. E,.: .!:-ii EEE, ,E .;it .? ;i': Et E::..E::Ei,E-iE:: It .j 1: !iE:EEtiE;: ,E.:7E:,:EiIEA EE,it '::  ;E, I:EE.-j,E:: E: , : -t-,. j. '? E, : :j '2. : :, , E.PE, :: , ,E , .::',.E ,E ;TIE - 'i.E.i E, : .iE:,i;E ;EEEE, E-.ic Ei;;E ::iEii,i Eiii. , F:i .iE;.j E ,E .: FE,.. 1... .. ::tE jl, E: .:, -E:,,,--.i.: .EZ:;E, .:;.. :Ej:EE,.t. E: : EitEE::i, ,EE;E1 Z- iEt,E:: -i,ii.E. E. E ,:EEEiE:  :j -! i: 1, [,!:"! ;) .!E ii i'l t , 1 jE !L t "! :, i i : L"!i ! j t  L! [], : i It LE  ; : E !, ': ! E tl!::EE E .. tE ,Et . .:ji :.,, .1:,: : E.J1,:; ,-EE ;, .E? , ;  ;,:, E:E  i,!  ,::E k .,,E! I , u i  i E; 1, "?j ;:!,"J-; j: ;E ;1 i 1i' 1".1 :! i,i:, i, ,v! .Qgf:;" E: :EEi.,F E E-;,: .: :E ;E. . ..,... -..,, ,:E, E:  : it E] .i 1, j :j EE.::Eii:Ei .: ;iE.;EE.:'i itiEEt'i:EE:,jE-':,.,E::. E: E. E.  EE.:E: t E::.I:!-1 , . ;Ej:; , z:., 1: ::,:,: .I: ;. :E EE::. . .:, . ,:EE E- ...; ; E;,:EE :..--i  :. ::.i ,i4:7..: E :1  -,E .:;::,E.., ,i ,;,;E EEitE:E:EE:;it :1 E ': 5:E:15Flt iEEE.EE3E:: .:, , ::5 :. . , . .,;..iE, E"E, : ,. Ei,. Eii;.:,:E : . : , :E:EE ,,.. . . .11 I .: . .. ... .,, E, E ,: :E, : EEE ,:tit:;: t .i:E:.!:;E, :i :E .i-.E.E,, i ':E:E, .::E,:!; E: ,i:tEE. i? t: t EE. EtEiEE; ] E: .E;'.] E,iE:,;: .:E .;EE:: :,, E E:.E : .E itE:j1i it .! Ell!,ti E ,, E, itEEii ,-E: I E , Eiz .: .ii. ..:-,.. 4i ,iEE.;.-tEi: eiii,,t: E,l,it  i,,,i .. iF :: : I::,: , E tI . -i .i .;-i It E : , ,.E.--:, -::, E,E:ii7 , -: it, Ez:,-,E,A E J:iii,,  j1:1 jj,_. :.,. .E ,E E': .:; .; .EE: .il, : '.Ii ,::,: .i.::::,..: ;.,, . - .,,E: : ,k:.:: 1.E'.. E: .i .:E:,EEE:.': E. : .:.-.,-..-,,..-, :,:. E, '. E, ,f ,  ,. , t , E'. ,Q ]t il  ': j ij U'Ll EL&!, E; ; <  i :: ; q, .E,,,It'! ,n] t.EE!,E : i, ,,,,,, ,;tj, ,,-t, ., -l-ttt : !?;IE.,- . tl. ,, ; E  t tl,,  : : E- - ,  :: E,:: : ,--': %:::,:;I.E.,:-:EE:EE:E:E't , ,.-`.,izE.. .. . - - . 5.. .. -.." ....,:E:iz :E,i:.; ..5I' -.- I . .. . . .11i: I.: ..::.;;:: :E,;;:;':..:::j7:; i .7- ;E, :15F"E:it, :: .EEE::,.1E.: ,:- , :,:jz,EE::E: .: t- E.::.:jit .: E t;.: .i::,:,:: .; , ..:..- -I..!.,.I.:,..-. -.. . . .I.. .. :.I.::.-: E, Eit::,:i ,Jijtlj;,E,.,;Ei-..E .E ;,E., ::.:E,:.jpE] : ;: !E.:. . .Eo , jiEE E, ::::EE:-,:O::i,-  11..  I... .E I, E.E: -,.,: ;',Et,; .: E, :E,:j ir E: :E .:;,.iE,.i ,:.tE!jIt El ,iE,...iii: . E ti itjiitt ..,,j.2;':EE:ij '.tEiE::: :zi,: Z' E z it , .5 ..E:EtE I: ...?: :. :iE ,E- :. : "... E. t:tE.5 E;! . F E'.iii .1::t,t .: ,E .j: ::Etl'E Eit,'.;Y::: Eit .: ji ;,i,.::5.1,:.E.!IE.::E:"E.. i:t, :E:;: :1 .: E; ;,.;.:i E, E; g E .E :, ': ,E:EE:::E ,.::: ". s E,E,I.:Ei l.".E. it:,E. ;, ii.E :- :,,,:E ,i .:.EEi:itEi'z:E .i E: z jij: .:,El it : .,: E: :E::  imi.: E.:.j, E i :, ,E.o E l: :E"E, E.1 ,E..  E , : ,  E: E Ej E; ti :, ..jE S. tEEE: .: ':E:.i .i ,j,Ei.;,Et'ij.;i . E j:1:: :EE7;: 'i EEjE: .i .ii: j; i, : .ji E.:E.: : i:::E:E,;,:E:::; i , :,..i :E,; ii-..:,.':EitE::E E '::EE:E 37. :i:;-..: ':.V. . .iE,.--,.,.. -.--i : ':EE.E,E.ooE. it.-E El 'i .F:.i E: : :2t:i .ij.., :-::;.'iEi--.:.. :. . .: . E  ,. tE, i 1 E . E: :!-: ..EE.E. :li .t , i ;'% ,i :: it..FF.,.: d? ..:ji: , it , EtE :; ,. :E,.; .;E_ : ,!E.E.J:;5.E 'i E:ijt:i .E E .- 1:- :EE .::z, .::- : 4.-;:, ;1: ,v ; , t,i : i;tE :1 , ,:i:! ii A, , 'E"j, : ,, " ,: .,; s " :,:; : i, ,' , , : t: E i ;i !, 1!. i-,IE E, ; i ,,z ,  t i, i,. L, E  EE, t " ;, ::: -,,"EE: :" , ;; ,Et,%F1,,,,, . lz,,::':.II-.,;-i,UL:EEO 1iE : '14 E, ',ii,"tt :1, E. EEE. ".Ei . E ;: :EEi EE ;;I.i ti;EE: .t E , ;. -:,4i.;. :E:.-........lt. E:5::.E ; EZ it:::EE.-,; ijURti ,. .E.E.:-::::E.1E:?.iEI-EEE , E.. E. E, 1;:E ; :,,IE,!I. -:- -Et:: ::,: E] ? EE:-:- -E E 'i E,:7'i ii t: 'iE._i, , .,.z..I- .IttE  it Ji:tj Et ij, ::.E .Fl.i E!;jEi:-,-:.E.5:-.:-: 5.f.1 .i.it it it !: EE Et,. .t EE.; :, .EitEE,E iE': :iE.: .:: ::tE .: :EtE : .E:: .E Et .:E:..,:SE i :_!EE, E:::,,:,,.': .:..1i :;it,t E , .: .i:itEE' ]; E: :i,E;:E,Eti ':E:.:,t :ij ':ii :, E; it .i ::E; 5E .t,t.:I? ,"::::E:, : .t-.i, . ti.t E:,ir E'i i  ..: I  ". . E, - E,, ;..- ,tE EE:, E:E ;-,E:- . , :E :,..E :; 'i t?'tj, ::4:z. ': E- ..iJE:! :t:,i! t ,  ...E ; ., o:, :E ] E; ! ,":EEfit E.E :.:F:11, .1 , i.. ,,, ! t. Et?E,E : E:;iEE:, :. itEEi .: t, E;E t: :, ,:,:j, .,t' -,j'E"E-,,EZ.,,., ,! :i, E j:E'li; i! t t- i ! ,...E li i-: i 1 . iE, i 1 , !:,: E! qi i:,! 4;4i it :_', :t,ji,;],.NJ, .t ;r,i:. ,, ,T,]",:;.j:D,;,-:" . Jt-- z, ,:j .,; Q7,,, E , 5: !,L, ,i: .,,,;Itl-,. -'i .,n.i 3- t" :L. 1%, : !Ei :1i: !..,, ,,,: 2':t. EE: EE ;; :it ,:E : 7-l::. . ::- E: EEi - :: .E .1 E" .E,:. :; E; iz :i E :E,zl.,EE::E :E- :E. E; i:,,.;..,'..;:. . .. .-.. .; - ,E: ,,EiE .i;E; :E :::: - iE,,E:E:: ? ,z.iE: ElE ,EEE:,it ':;E; ': ., :.,:,-,.t, t, :U iF AF 4i . t,iti i, : ii :-1:5 -::.l5 :: :: .:ElitEE:it. itE . . .:I.. . . ..E. E:i E,...E,i'l :iE': .i ;i: :EtE ,.:,... ,.E1E.-E:Ei E t:.E:1I,1i;E :E.-i .A- .El.. EIEEii it .E .titzEt.k.;;-,..Ei :, i -i':::E:.:.E: ,..:-:: .: .. .E"% : . : !: . E 1 :,:,. E .,..,.i: Etititj:: t,i .! .: .; ,t : itE: liE.E:: : ';,::, EEE:; :, .;i '::zii, i ::EE! .Eo.: ':O:,EE,; 'i,-.EE.;:'i .: .E , .,::Ei ;::;.:... ,E: ii i,,:E:E.: -i :,:.. ..:5 , .-.i.Z.1,,.-. . .. .,... -. t:E., -F -i V,:t E!ji' ij :..;:,i..E, .t,:-i ,i, : iE..:E.- :iit EE., ,E.E.;Eli: .1 P:E:E.ji";E  E iE ., ,, lj I: Ti El J 1,  .i li,R , ,,,', .iEE,"E '.E.EE-.....-.i .. ..  ,.:i- ;, ,- -" . ! jq,i : , :g, E :iE, j ;,   i i -!E',sz : ;i ;,:- ". " E; ,..;,l ,i El p; , E ': ,E ,; : ,E : ,; !isit,,i .: .E ;:E'], :I't, ,1 z;,-',;" -El" -,_3 ":,qEE J,3, " .:,, ,;E,;, 3,t"'B! t !  ,.E. 1,  .:.:t!It:i:'f1":E:iz,-.:::,-E.- .:.-:i 3 E "E-i t3 EE, :, i : 55 -I ` . .,,.,,.,,E., , E ;, E:i E,.. E, E v E-ti% ?4 ti: .,:-:.:;...E:.-,.-E.: .:-: E,' .., ,..it ..,.. i]i: t,it .,...-;-'...::-.,::;:: 't  :".. : Ei ; :E. ..- -.- .,.:-li:.lE . l:. . .EE,,. : ,:,.' titE-.:::E0: EE  t,;.;EEEi; it .1... -.--.. .. . . ,q... tiF,.-"..E': lE .E: E::EJE.EI-..;.,.EEE,,;.: E; :E;E.,. , ;,t .-E : ", : ..j 'i:;j E:.E .iE11 ,E .:Eit: E:.i .i ;i5jE::E;E :EEZ ;: l: 1: ::iE,:: - E.:- .;j-E:,E.t:,:i:"EE:i t::."-:.iE.tit :, .E :- i.;E ':, i,l .: Et..: .i,j i ::.EE.",EEE:;.- i i iiiE.,jj,;i ': E: :E .: E li E,E:: ;E .,vEi:,!t::E,.:E ,P: E.:, EE-.t ..i.E E: it J:ji:Ei i i ,- t. ..:.: ., :I:iEE: , :.t :::t;,5 E-EE E:. ,,;.,E1..: .iE,:E.: E :::i:,5,E.;Eit ;: ,i,i:'i i i E1Et :E:ttEit::; E:,E,,I EEil,EE,:; .;E:. .EE..;.:,,:1::;'::li': -, .:, :.x,-, I I'Zi '!!i .:,i- : ., :: ;.i:; E: i. i,: .E ,.: E.E,: :. E:, :E, :: ., . ,.,_.-..-,:.:..I:::zEE:; .E .. ;: .: 7 .iE_ii:z:::..: E;.E!i-E,-E: E! .i t it E, ': i:,j -E 'I i ".....,..-I.':E:,: t: .:::.i:,E:::,;:j ,E :: .,.,_..:.: ,.:iR.!i FEE5:;; .z:i5.i"::Ei:....:E ,..........I::,.- t,.,. iz ,,;zE-E. t EE. ;, E: 7. ic..'.... I ,1.EE':i': 5: ; :: .. : ,5.-,-E. '.E.-,;,!;E:E :i :E:;:i: "t, i : .. :t E E; i  .t,E!::. E, E .. :- iI:; : ,51Ei :.EiE !EE , E,. i...E . EEF. ...... ,...EE-!tj :E E; ! !E E ;! i ..,'. -."IE.E, E.i,i ,...,.EE'4: '. i.-,'"..--.,  E .:,:E:tiit-,EJ :..j EE!Ei ;: :t , 1:i't;.; iij-.:iE: :E !1 : ,1 .: ..E :, i.E: . ,:,i;E.::: : . 2.::.E :.. .. : , ,: .;,.:.:,. .;;:--,-:....-E:7;.:E,7.:E..I,,.:,:ii :i ,:. .E,:,;E F . !FEii, !i i - !:,.,ii iI , ,,Ek E. 7:E,E : :': ': ::,;...,.:1 ',:E,...;:.-.. ,..4 ::t ., .i :2E::Eit ,EE:i :E7:., litiit iti;:i'k ,:xE., tI .1.-.E E-t :Ett E .:E .,i .:I :E it -E i -i: tt! EE :,-:.:E- E t , E.'! 1 E, i  Er j ,:E  f l; ,,,,:,,U i:, iii,"E,:: :-k  !,:-tt;i , -;,"',EJt;,,li ,f Er -,,.-E!---.---: i'E%siii : E: i, ", Mi,Dzi,:],, !,i E, i';, i , i, ;, -:,,;...:E::.:j , 3i, i, :3E,1, :"j!f'T,Ei E i ,! " 1, E'i ,7.;,-E i.i,. i 11E,. Ei. , .., : .E E5 ;:L E:E:P EE: .,E.i,..E il: Ei 5z,,::.,,,E;;: E : ,E..:El. . Et .i .: it t! :: E--:2:::E .;.:::: i- ::!i;i L : :E:!%i E: ;::;:,jt::E-E! !i.: -i .Oi-:;;7:... :E:EEi.:E:i:0, .:E,.E i::?i,E .E .; E, :: .t E:E,E.:E. i. :,, t.EI'i ,:Fi :: ,7 Et -:::E :!i:.E 'i:iE:tiIEEEl,EE-i:EE; ; El :iit :: . . .. ... ..:-:iit ;: .t!Jj J:: E  .] !], E;lI:, E. V: :. .EE EZEliF:EE,t E--E..,...E";:E:E .E:.E"  -.-; EE; .E.,:;,iE, :. : .. :, iE.E .E: .i ;IE. .E:;% :E ,E-.EE:,--..E.," , .. . :..:...,..i.:EE ,Eit 'iE,E:j E, , EE; ::EE:2Eit:E.EE;I,:EE:; ,,,..EE:E:.:Ei-:Z :::Eit::it,E :t!EE:,ti .:EE:E.E ;: .]:EE, E:.1lP :; iE- ,;:.E.: : j,-: E: itE;:,i:i E' i iEE: it . 1. .: E ,i ? ': .: .E,Eit .i:::.-.-.;,E..Eiil: .EE.:tEitEE. : .: :;,: :i ;?:7E: :iE,.,Z:EE, .i. EE: :i,; , E EE.jl: F. E:  i': ::E &,,:I t,':,3;".,-E, EF%`Q: :2ii  ,, : .; E?E:.,-: . E ,- Eit  :.E; . E.EE, E, -. ,L , 2" E ,:, E.it, E,EE':EEl. ; . 7 :: :-it,E! .1-- .. ., -...-t.j, E,:;, -E ,z-, .i,,j t'! itit;qii i ,":,Ell ......"it, "" "'i, , , 1 ;,,§,;:, E, i;,,:, E, E ,,;,,,:,,Ei, ., ii,  p,"i IEE 'ki. ";, ,,-:, ,"i'tt Ei%E,,4,i :. it .oi:: I ::  , :: E,j:  !;!i ,1E .," ,. .E.-ii.: .;tF.. tE:: .. ijj1iiE!i .EEE,-:EEEE;: t: -EE': .E :: .iE': ':EEE::E7:.....  : .:.i.:: .EI 1 .: : .:.: .,..'i.I..-.:-Ei:-i .; 7 EE:: 'i E t::::.i;F :E 'i 'i. :E::.i .:,: l  ::7 ...,:.:;...,,.,.: -,,...:, iEE iizE:t::,,,,:::;.:: E::E,:;:.,.': ... ..;: .:..:. ,, E5;; ;' : :E;, E z. .1. ..1i .: .:.,:.EE:;: E. .: P :i tE;::Et:i E.-E .; i :::.E,::i': ..::i .l: :i .: E: 1E. i: t.-it t,,:: E. E;Eii::;-EE'i E] E. :;Z:.E, :E:::..EE ;:IE:E .iE.EE:EEtiltE:,-..iFEi .7,i::.7  z.: s ,i;:Et .:  ,,E: ,:. E: .E - E:1 : Et ,E ; E; :j .t .:o, ', :;:E, it :Eii:%i::E .i,E;,. 1:i,.,,-:,E.E:jE: :: :E -j-',:,E -..: t :i:EE-i F:E'. :: 'i .;:. -... .,.Eit5:E .iEFP :i ,ii-.; ,E E'tE,:it .; E: :,i .: lil.,-.-.: E .E -EE.: ,E .I. :,  E ,E E .i ;,-l :E: ::j-.E:;, .0, . .; , 'tE, .: ;-:....E: E.i._ ...':p:. ,i.iE,iiFEZ. .t .F E. .E FI.;EEE ,. t.:;.:E.,,--..E,P Ei1ii E,,ij , ,, , , ". . :. .-E E,,,: EIE-,    :" :t :,, E.,ii .E.-  E: J-,E:: E...i .; ii 'i .:, E: :, j,,! ,,,J..,it ii EE I i: :: :21 i: -v3Ill; 11I-,,,%v i11 t:  L 3E ,tl;, i't ' E j i" . ,,I,it E ,, .1,:! t    .,;i"", : , ,t,"i  E. t, s,,, :  " E,E-i,"i , ! !, i , t- ! ,J. ii_ ,.",: ::-. :1at. " '. : ;-.:14 -, P iil,i! :!  : .Z'i- E t t  :.  S -3E; E: : :,E:.  :.:EE,-:lrEE,:: ;:Ei:EE-:.:-,.,,-::E:j it, E. ;:,;t4E. E:... .:E,E]b.5il:!ili, E, ,'..'.-., ....,....... -E itEE: ? ,;: E:EE.EE:;;,El:Ej: ': , i, ::Ej.EiE: : : :, ,: :E E. R'i;E& .t,t .: .i tj E:,it'.E -,E .E.. . ., . _ _t. E .:, ji, . : .E . ,.i,.IE.:t:i'. :Ejjl-.-..-....:;: , , ,Z; ,. Et .t,f : ,:-! E! : E: ;,.:: :E E.!,IE 1, E. EEE E:: : EEE '::E::t:::E: .?:itE":; 'iiE.E:.E . iEE EEE.i t.E ,E :E .2::.i.EEii .. ., E: itE?,,: .t E E;J it .E  .. . 't'- , Ei E: ,::.:E;: : .Sti::: .: ,.,E, t: ..ii.EE: . .:.P;., : .;:,::: -., ..I.:.1-.. :E-..:;E,:EE.  1, EEE it .:.: -i .? .];E:EEl ;Ei:;. ,:1 :EEE;:. .EE E .'.-,.. I,II ,. , ; .E? : .,-. ::, E.:. E I:E. E. E,. EE.,, - .. :EE z - ::, EF t:;E : .:E.: ,:.:;, :., ,::E ", .E . . .E:.:E._.: .: :Eii:Ez.i; .., :-:.iII-..5E ::,.E'; .E! . .: _:j ,E . . .; :;.; :; 11 It'.. :.EE: .. E ,  ..E:IqiEEt E-i .:, it::NE .Eitji .: .i :Et E2:,. - .., .:.,:-.: :. , .: . . .; .-E.,-:t. .i ,E :: tEEE .,: :: :- :E:,I.: E. E; E .:::,E. .i .: .: .E. E. ..IE Ei: '::ji :.E:: ij.tEEE:l.: .: ':,iF-,.:].EtitEj:t ,;j-iEi'iEt:E ,.: Z:tE, : .E ; .1.: eE: VE , : E.E ,," :t;:  :: ;t,E; EitE.E: ,7E::;..:.;E: iE:ii.E  ,: E&;. i:,:, .:EZ. ti :!!, tj :! i. 1. F:-ii:  Ik  :;: : :E:::;E. E., ::-i,;. :,E "E:: :EE:E:.:EE. :; : it .i i : 'iE.: E: E: .: :E:E.EE.: .E;:EE::E:;,.. . . . .1, .:.-t : .E. ,.,.i :::z : :: .: .: i": ': jE:E: ::.;j-;:; "': i :;: l:.E,E.Ei:.:,_E-.;:::::iiII.: E, :E ..F:: .E:.i.:it E:EE,E:E:?E ,:,VE:Zji E: EE,.j >t!,E;: :.- k i", ,t::E, !  .   - ! -E !, it!. :, ;,t,:,, ;,'ij t ,,!a'; :  iit .iE. i:: .E,I .1 E,-,i, :. E.: ,EE ,; :;'l,  E ",,."t"Eit:, .L! J ]: E,;U", ,i ; ,,, 'it 4,E,,,, :, 1,  , 1 ;!,;i,WE..i it", 'it :,,, E!'JO !LLI: !:,,! .,i,l t,E- it i,,,:!: !t ,"t:6,1,ERA,:I1,,,l t',E j i,:,;l i 2:" E;E , EEEE E:E, E,E : E  j EE E.E: E E : EE : I .E ;.E.:i- l_,. q. :: ,,:::;i,:::Ej i',,iEE lj j E: .EEE: li7.:. , . , . .1.. :. E E:: ;:,:E-.:,,- ':EEEEt:?EitEE. E; ;:E:z, :..:1 q iF ii :EEE; !:t. :E, E::.E:EE,ii :E:?. i,': .j ,::." E.... - ,E , t :: ;. E Ei!! '! E E:. i EE :E 1 ii :i i ., i : E, E: :: E;EE : ;..i;. ::E :E E, EEj: 7 E .E::;, ,,i ::j ;j i E ,t. E.:EE.t.E:.: !t : : ji! :E'! !Jj;4.E E..,:iA.-A....I- ..W. .: : :: ,Ek :::;.: : EE..:::W.::.: :,-:4- . --II, .:1 .]t E.:.. .i ;it .E E. . sit:i E;;; I'  E! ilIIl :] .!;! i, :j. : .ji.,:. ::E..:7.: tiiiit: :;E5E: .EE?i5,i .:.I--EE,_,-.i I .1 .:,.,E.:: j .:':EtEt:E.i :: :E ; :-!: 'i t,.tE.,,.EE.!!Ii.: ., I.E ,.;E5:.:E:,E: i:.E;:EE; 2 i :t j;:'i , t ,: g E ., :, ,,E, i.- .i.: .: ,Ek E- ";-.;,,:, ;,.., 't, t;,g , E:,,E, t, l- ,iz :i,l-i:, E,1. :  ,'Ec,], :::,,.,l :5 : :,:, L 3 ,,, : ,. i:; t.! E'. i"t,7J, I ii,k.E it .-,EE1 1"i 'i  --::::. ,-3,1' ;,,,;'. 1:  :I,,:l, ,,.,..., !": Eiiit, L 1  ;  3 E,;_3j,,.,:,---"i.,-', ,::..E :: .:,- , :: .:.:-,.'i,. . .1-Z.-,Z5 'El.':,"',-: ., ,, :,;E: :. : _ ::, 'i.E.i'E: : E EE: i.E.;:, ;E  Zi..- E , E-:E;: E :-.E:,:,llE:.iE,, ,, ..,,::E EE':t-E i:..i..iE;_i.: E' E: i, ;itE::j: it j i ]EtE  ;: E. i,, zi .:..-i-:.: .::E Y. E zE.E. .E ;..,j, EF,:iEEE-I..:i..E-: -_ E .,E .t! .E,:.; .: .:; .. .i,t-EE: ,;.E:,.C.:,: ,: ,..E: i:EECi,E::. E:EE:: ,E:E  ! it -t E: ;E E E: :::-:,Et, E::EE:Eji .E-i zrE:2 ;iitiE:. .. .t E t, ::EE: ,E: :E:E,.-....::i:E ;7 ; E-:.-::z. .:::El E : , .: _ ...,__... .E;EEEF,.., -. . . . . E .I I ` E E J5 7 -;:,E j : . : !   :: : :: -: :. .E .i E. ': :,: .i E ii ,E E: .: - i E . E. P E: is ; ; - -.i . ".,.; .. . .. . . .. . t Fi E:t  .   ,FE .; E.- .E " . :- .: .. .; E; .- ; : . : , ,.,itjj E z  ..E E : i.t:E. : .: .-:E -.i7.ic '-. .. ,.- :," :': Z:: -,,-,.l::. ..ffi&W . .A, ,. ., L! it.:...:.: : E.:;:iIE:: ; :;,: E; :; - :5:15:,; .:  1: .:I.: .. : . :E- 1:. ...,E-: , 'i.E E ,., , . ': .: ....,.-E, :E-t:j,, .; ..--.-.. '....-.-E .; **.:t!,;,,-1: ,:, :::E -.. :::;1, :,- ! -:!..:.:Z: ': -?, t: t,,. 6 L ! ,,? !: E: it.i..i,i, i.,,. i'. .tiI:E: :,- E,,I'E" E,I, .:: , i , EE: , ,J,  ", .: ,.: 1. I ;  & V 7;` ,-. ;: : :.. ,;, : ,E , - : ., -1.  . , , E - 5... "El '%l, E -i: .S ." :-;-iEit 4;1:E.:;0404iloi* il.- ., :: E ,E'-"I , I I ::11 ; ". I :" ;,;j , ': i, VJij",,,, E,,- :E ; , ,", ! 't , E i " ;:5j. ... E: ...,-ii--_..:,3 - -!:.:: E,i E. .:tE;,_EE:::,:::,::E,:j 2 . ':,:-E:E.-:.:E. . 'i ., E. Et ;  .: 7:,E! it .i - 5: :- :-. .:-,it;:i,:,'::: :..- :- E.,-:'.t:Epj: E,:,:,EE..1. -E::I. :. I':.,I:Ei,  :.E:,:,iE:EE, ;E :iiE;:E,iE:::,. E :; .? .EF7ii :;JcE."I.:..: ttFi:t?:E ,; 5y::,,.ti WEE.i lit ...tE,:.-li.-..-.:.: ,:,.. .1.S. ;, ......E, .:EE1E.:I.::E .--... i . ;I: ,; : EE ...::  ., ?.1,.::.IE; :EE;,E.: i .F"EEi: -iIiF5, :.: .:,..E :: :1 .tE5 , : ,.; i5: E tttj .E%oi 'i E EiE':iiEE:IE-:-::: EE'i.iEiF Et.E::it;:, ': 'i, E..E::..E. 'ii.5i ,.E.;t., ji. E.': tE:E., :,:" ...,. ii:. E::. ;.EE: -::E, ,..E:,;EE::1 .t' ; E:.i.. .:,E :E t...0. E: :TEEEitEE..,.,. EE:t.. -EEEt': .:;,,,,::;Eil:5. .:;": ...::E::,:I,::;E:EI, ,E.EE-E]t:.E;.i 5i,E E:;..EE:.E, E,;EEEE.];E:::-i .t .E;:- -:EE: . E:E.EEE: .:;E::.;, ,.E .E E?El.E: ::E?iztii :ii!] t!:E; j i! 1: EE] it : itti .tiii , :E E;:,:. !i!];E,t. : 1: t, E, '..,.:...:ii. E: ..E::,. :;-:..j,.:,E .; .; j j: Ej : .:....E: ": : ::.E-jEEjEEE,.i E' :E, :E",.,,,.-:EE .::': .: .Eiz:: i:ii .i-iEE:: 'i . :,.i E,:Ei l::11- ;iEI,,. ..:.. ,- - ..,... .--.l:.,:: .i;E.E .E.ti:i,.:E::.:: .. :2EE:E, E.2 i E ; iEEE.5: :E j,:.,:,E .::iEiiEfit ii ' ii i :jIj. i  .:,:.;:,:. .: .:.., ::.. . .. .I..::. . ;EEE-E E--, ti i ;j .! i: i :i i i  ,. -.. .: .- : -,, -:1..:.i--,2 il .: .:ZE ::,,,'i,E:E .: J;..:::E,:, ,j:?:i, i: ;E i ,: I.. EE Z, E,, E. i.: '.I,:: .: l:;,. .-: -,., .1- . .;,-:E,7-, - ,E-Li, E.- E,E J ! ;i., `  "? il,:E1,;E,i, :,E, EI E', I ...,- , ,;%E t,,: :, l'; E , E, ,,, 'O 5, "t ,i. .. .E% . :: :., :;; .. :; ;E ::: E..- ,-,f ,q, ".-"", ,.,;,A, 3"r ;,& Ei'; &,,Ej,. i,;, ,,l,:,-,.c, ,, -ii t 1%;'Q ..: ::. :: ,.-.:: ".: :1 _ . .: - -.., :. ..:.E....E:; ,-.,;,: E,_...E. .: ': , .: ::: .; -:1,,.."; .,: .:-;':., d:.; ,EIi IE ;E:.,:,. E ,p ti 5. ..,:.,,E i'v : ;- -i i, , E E. ....- .. i:..;.... .. . .: E. :i   i :1i::: :;,,.ZE:EE, -E.EE,E, ..  ;, :::: cm .:- : :E:: il. : :EE :. :.. - .ii': , E: : .EE: .-.EEE:: .: : E] :: .i. ":- :. :: .. i,:E:-i .. - ,-.,Eti,E::. ': ::E:..,E5iE.i.:EE::,.:;. ,..,:  .:,..... E,...i;E.:EEE:,:::.:E.E.:EE., it .:t :: '::Z--.. ,,..E .E :..: .: :;E .i,.E ':i - , . . : ':,, E.. .E:-." :.,.:Ei:.. I.,:,,.: : ;"E:: .:;2:E1:; : ,, - :. . .:---i.-.. , :E,I-EEi E:..i .;E: : EE,itE-.,.E, .: , , E. ., it   Eli ':E:. E:-,E::,:E:! ::::,:, .,: .i .E,t :i:;.,-tr;.::.. ,I l, ;";Z::Z,E," E, E: E, S%i,s:ti E E , ji .ji0,,R, : . :5ilt,.! 2: ", ,E :Ot:titc.k',:i,,l,,,,i E4 i: I " ,E ,;,,'j, EJ-,4": I W,Eii it...,,!,,,'i,,,, zr%!; ::A -.:..E I;EE:;i:,EE,EE., i. E; E, J;,ti lk;, 1 .., E:.il - ,, ;E::Ei E: .i : ; E;E:E .: E Eji .EE .i jE .:::t.;:;: : - jF t. . , ,' E ::: :I.; E :. l:, ...-::, .:E: - ,EE i ; .::Eii:-,jiEiE ti .EE:1 .::E. ,;  I'll . E:i : y : ,  : , ....  EiE EE, E: .E ...-. ... .: :, .: :: E, a  .: it E";E:.E ,E .i it,, "2 .E .: z. E .1,.': :. i: O;".E;: : ::;.I ,I.ti .; ., t:j E E ,EE...,-:II ...:..::..,,i: E'i:t E t! .,:. :EE:,.E.: : i,- :I..E? i:,: E. ;:::: ;:,,.::in ,; l,i .!!: .E i: E;--,--., ::."Ic,- 1: .,,i ,,t5Eit,izii t :E:-,,Iilbi , ,41,,.1 il,ki, ,,I,,.-,, ,. ... . .?,i%ilm E , 't, E, I;,,, 2:,rtz's _ ., _ ., ., ,..., .1.,.., l" . ,: E:: E, ': ;, : .!,,',,.., -i..-...:;,Ei-iEi:- . ,:.:, E .. ... ,.. EEE;%, .E-.: i1it ,E_..i: .::i:- : .:;- ., .. .,:L...EE,..EH,iE  d."o,Ei::E': it: itEE, ;IEj E." S:. 3o: % ;:,r "...,! . ..EEi it;,..- ,.. .,I"EI ; EE !E E .;'Ot .I.II..El .. .. :., E.-:E: E' : '. E J ..,E,. .. ":: :..:, E ::.t .:.-.S. E ,,:E .;. E?"Y: ,b ::Et % ., ii . ..;I Currency Equivalent 1.5 Deutsche Mark = 1.0 US Dollar (1996) Abbreviations and Acronyms DM Deutsche Mark FRY Federal Republic of Yugoslavia GDP Gross Domestic Product IBRD International Bank for Reconstruction and Development IDA International Development Association IFI International Financial Institutions MFTER Ministry for Foreign Trade and Economic Relations NGO Non-governmental Organization OHR Office of the High Representative RS Republika Srpska SFOR Stabilization Forces SFRY Socialist Federal Republic of Yugoslavia UN United Nations UNHCR United Nations High Commission for Refugees Vice President - Johannes F. Linn Country Director - Christiaan J. Poortman Sector Director - Pradeep Mitra Sector Leader - Hafez Ghanem Task Manager - Khaled Sherif Acknowledgments This report was prepared by a team of Bank staff and consultants led by Mr. Khaled Sherif, Senior Country Economist, (ECSPE). The report benefited from extensive comments and suggestions from Mr. Wei Ding, Principal Economist, (ECSPE) and Ms. Mary Sheehan, Sr. Country Officer, (ECC04). The PER team consisted of: Sebnem Akkaya, Pedro Rodriguez, Robert Palacios, Hana Polackova (ECSPE); Jacques Bure (ECSIN); Richard Hamilton (ECSEG); Michael Mertaugh (ECSHD); Guy Ellena (ECCHU); Alexei Remizov, Marie Therese Schurrer (ECCBA); Oliver Campbell-White (AFTPI); Randi Ryterman, George Clarke, Luke Haggarty, Philip Keefer (DECRG); April Harding (PSD); John Crihfield, Kelly Edminston, William Fox, Malcolm Green, Monica Burns, David Turnbull, Stephen Harmston, Kerry Johnstone, Robert Forster (Consultants); and Jean Tesche (US Treasury), The PER team met with representatives from the State, Entity, and selected canton and municipal governments of Bosnia and Herzegovina, as well as senior managers of selected nonprivate enterprises. The mission would like to extend its thanks to the excellent cooperation it received from government, authorities, enterprises and agencies. The mission received exceptional cooperation and would like to extend its gratitude for the excellent cooperation of representatives from bilateral donors, multilateral agencies, and private firms, including Michael Markels (US Treasury); William Mako, Rocky Ho (Price Waterhouse); Craig Buck, Mike Sarhan (USAID); and Major Bob Bishop (SFOR). The mission would also like to acknowledge the invaluable assistance of the Resident Mission. Special thanks and gratitude are extended to Mr. Saumya Mitra and Ms. Lejla Zaimovic who both facilitated the mission's work and ensured its success. This document was formatted and developed by Ms. Erlinda Inglis (ECSPE), Ms. Rose Nguelie Djieya, Ms. Mary Morgan and Ms. Stephanie York (ECC04). The document was edited by Ms. Meta de Coqueraumont. BOSNIA AND HERZEGOVINA PUBLIC EXPENDITURE REVIEW Table of Contents Volume 2: Technical Annexes 1. Assumptions Underlying the Medium-Term Economic Framework ...........................1 General Framework .1 High Case ....................................................2 Low Case .5 2. Military Expenditures .9 3. Public Administration in Entity and Cantonal Governments .12 The Federation .12 Republika Srpska .15 4. Pensions .19 5. Sector by Sector Review of Infrastructure Policy .23 Transport .23 Roads .23 Railroads . 26 Telecommunications .30 Water .35 Energy .36 6. Subsidies to Industry .44 Government Spending .44 Lines of Credit .45 Liabilities of Nonprivate Enterprises .47 Implicit Subsidies in the Tax System .48 Subsidized Utility Costs .50 "Waitlisted Workers" .50 Boxes: 1.1 Medium-Term Framework: Underlying Assumptions of High Case Scenario 1.2 Medium-Term Framework: Underlying Assumptions of Low Case Scenario 6.1 Whither the Conglomerates? The Case of UNIS Tables: 1.1 Medium Term Outlook, High Case Scenario, 1996-2003 1.2 Medium Term Outlook, Low Case Scenario, 1996-2003 2.1 Troop Strength 2.2 Estimated Expenditures on Military Salaries for the Federation and RS 2.3 Estimated Annual Military Expenditures for the Federation and RS 3.1 Public Service Staff Numbers and Costs 3.2 Federation Employment and Salaries 3.3 Sarajevo Canton: Budgeted Salaries 3.4 Republika Srpska: Employment and Salaries 3.5 Republika Srpska: Number of Employees by Employing Institution 4.1 Projected Budget Results for Sarajevo-Based Pension Fund, 1997 5.1 Road Sector Funding 5.2 Road Sector Revenues 5.3 ZBH Financial Activities for 1996 5.4 Investment Required to Reach Pre-War Railroad Infrastructure 5.5 Estimated Maximum Billed Revenues 5.6 Telecom Revenues in 1996 5.7 PTTRS Expenditure Plans for 1995 and 1996 5.8 PTTBH International and Domestic Revenues 5.9 Income Statements of Three Energy Utilities 5.10 Liabilities of Three Energy Utilities 5.11 Payment Sources for Household Collections, January 1997 5.12 International Energy Rate Comparisons 5.13 Natural Gas Rates in BiH and Other Transition Economies, 1996 5.14 Externally-Financed Investment Programs - the Energy Sector 5.15 Company Projections of Cash Positions 6.1 Maximum Interest Rate Subsidies from Credit Lines 6.2 Number of Workers in Waiting in Industrial Enterprises in Bosniac-Majority Areas in the Federation, March 1997 Charts: 1 . Composition of Public Expenditures Medium Term Scenario: High Case 2. Composition of Public Expenditures Medium Term Scenario: Low Case Annex One Assumptions Underlying the Medium-Term Economic Framework 1. Bosnia and Herzegovina needs to further develop a new governance structure, undertake structural reforms, and implement a reconstruction program, while simultaneously maintaining macroeconomic stability. The pressure to keep a tight fiscal stance will continue, in part due to a very constrained resource base. The currency board approach to monetary policy, mandated by the Dayton Agreement, will render money financing impossible and impose a very tight constraint on the government's ability to borrow domestically. Therefore, foreign financing is the only viable source for meeting most financing needs, including to normalize Bosnia and Herzegovina' s relations with external creditors. Whether the country is able to increase access to foreign financing over the medium term depends primarily on its ability to maintain momentum in developing market based economic structures. 2. A tight fiscal program alone will not guarantee fiscal sustainability in Bosnia and Herzegovina. Fiscal efforts need to be complemented by progress on the structural measures -- privatization, trade liberalization, institution building, financial market development -- that are key to the development of a dynamic private sector. Equally important is that Bosnia and Herzegovina immediately address the severe mismatch between revenues and expenditures through comprehensive fiscal reform. This will be vital if the country is to smoothly adjust to diminishing donor assistance over the next few years. 3. This annex presents an illustrative medium-term fiscal framework that analyzes two scenarios. The high case scenario assumes that reforms are both strong and quick enough to generate a robust recovery, ensure substantial reconstruction financing and debt relief, and develop a sustainable fiscal situation. In contrast, the low case assumes that reform is implemented less vigorously, justifying only gradual disbursements of reconstruction financing and delayed and less comprehensive debt rescheduling. General Framework 4. The basic structure of the macroeconomic and fiscal framework underlying the high and low cases is described in Boxes 1.1 and 1.2, respectively. Included are assumptions about: (i) growth; (ii) the shares of revenues and the various functional categories of expenditures in GDP; (iii) the implied time path of external government borrowing; (iv) balance of payments, external debt and creditworthiness; (v) the disbursement path of the US$5.1 billion in external support for reconstruction; and (vi) the terms and timing of debt rescheduling. The basic framework is illustrative, particularly regarding debt rescheduling assumptions, and is in no way intended to prejudge the outcome of future negotiations with Bosnia and Herzegovina's external creditors. It is deliberately simple and presents projections at the aggregated country level to account for a weak data base and as-yet unsettled issues, in particular, regarding international obligations. 2 Bosnia andc Herzegovina.: Public Expenditure Review Box 1.1. Medium-Term Framework: Underlying Assumptions of High Case Scenario The momentum of economic recovery that started in 1996 is assumed to continue based on implementation of a well-coordinated reconstructioii program and rapid progress in institution-building and market reform. Maintenance of a tight fiscal stance stimulates stability and hence rapid recovery. The fiscal deficits are assumed to be financed predominantly by external funds, reflecting the monetary rules of the currency board arrangement and limited potential to borrow from domestic capital markets. Key tax and expenditure policy measures, focused in particular on expanding the tax base and rationalizing the composition of expenditures, are assumed to be successfully implemented to ensure continued growth as donor assistance scales down. Sources of growth and employment generation principally, as in 1996, are the reconstruction program that is financed by the international donor community, and to a lesser extent, recovery in the domestic market and the moderate recovery of exports. The increase in exports is assumed to be largely stimulated by continued rapid recovery in the exports of natural resources, such as wood and timber, especially during the next few years. About 90 percent of the US$5.1 billion reconstruction program will finance imports for reconstruction and, to a lesser extent, humanitarian needs. Grants are assumed to comprise 70 percent of this financing and concessional loans account for the remaining 30 percent. Imports of other goods and services, mainly by the private sector, are assumed to be income elastic, and are estimated to increase proportional to GDP growth, using GDP as the measure of income. Private transfers will finance large portions of these imports. Foreign investment is expected to resume beginning in 1998, and commercial lending to start in 2001. Real GDP is assumed to recover to about 60-70 percent of its pre-war level by 2000. This increase in income would permit strong recovery in investments and an improved level of private consumption. It is also assumed that additional financing needs over and above donor reconstruction financing will mainly be provided by lending on concessional terms until 2000, in particular, through balance of payments and fiscal support by donors to facilitate the transformation of the economy. Financing at non-concessional terms resumes thereafter. Simulation results illustrate that debt sustainability indicators improve to manageable levels as early as 2000, which should make Bosnia and Herzegovina eligible for borrowing from non-concessional sources thereafter. For Paris Club pre-cutoff debt this scenario assumes a stock reduction of 67 percent of the total stock of arrears in 1998. The remaining 33 percent is rescheduled with a six year grace period. Post-cutoff arrears are assumed to be repaid within six years with a three year grace period. London Club creditors are assumed to reschedule 50 percent of the debt under the same terms as the Paris Club pre-cut off debt, while the rest is assumed to be bought back at 10 percent on the dollar by the end of 1997. Other commercial creditors are assumed to provide treatment comparable to that given by the Paris Club for pre-cutoff debt by the end of 1997. High Case' 5. The high case assumes continued success with stabilization, rapid progress in institution- building, significant increases in the pace of structural reforms -- led by significant liberalization of trade, price, investment, and exchange policies, reduction in tax burdens, particularly those on labor, and enterprise privatization; and rapid development of the financial sector -- led by restructuring and privatizing of public-owned banks. The donor community is assumed to fully support these moves through fast disbursing reconstruction assistance- and provision of comprehensive debt relief, including cash flow and stock relief. The resulting strong economic growth brings about a rapid recovery, with GDP reaching some 60-70 percent its prewar level by 2000. Note that the high and low case scenarios exclude pension expenditures, and the revenues from the earmarked taxes designed to fund these expenditures. Annex One: Assumptions Underlying the Medium-Term E&onomic Framework 3 6. By vigorously implementing its reform agenda, Bosnia and Herzegovina maximizes the beneficial impact of donor assistance by developing a sustainable fiscal outlook by the time donor assistance begins to phase down, toward the end of the decade. Tax reforms, accompanied by institutional reforms to broaden the tax base, ensure revenues strong enough to support public expenditures at about 40 percent of GDP (Table 1.1 ). Table 1.1: Medium Term Outlook, High Case Scenario, 1996-2003 High Case Scenario (in percentage of GDP) 1996 1997 1998 1999 2000 2001 2002 2003 L. Total Revenues 56 69 57 47 38 36 36 37 Taxes 33 28 29 30 31 31 31 32 Nontax revenues (excluding grants) 3 5 5 5 5 5 5 5 Grants 20 36 23 12 3 0 0 0 II. Current Expenditures (excluding interest) 49 51 44 38 32 30 29 31 III. Current Savings (I - 11) 7 17 13 9 6 6 6 6 IV. Investments 14 29 21 13 7 7 7 7 V Primary Balance (I-II-IV) -7 -11 -8 -4 -1 -1 -1 -1 VI. Interest 1 2 2 1 1 1 1 1 VII, Total Expenditures (II+IV+VI) 64 82 67 52 40 38 38 39 VIII. Budget Balance (- VII) -7 -13 -10 -5 -2 -2 -3 -3 IX. Financing 7 13 10 5 2 2 3 3 Foreign 7 13 10 5 2 2 3 3 Borrowing 7 13 10 5 3 4 4 3 (Repayment) 0 0 0 0 -1 -1 -1 -1 Memo items Real GDP growth (%) 50 35 27 20 12 10 8 5 Exports of goods and services (in % of GDP) 20 23 24 25 26 27 28 28 Imports of goods and services (in % of GDP) 20 23 24 25 26 27 28 28 Current Account Balance (in % of GDP) -26 -33 -24 -14 -6 -3 -3 -4 Total External Debt (in % of GDP) 115 99 49 42 38 35 34 34 Debt Service (in % of total exports) 87 30 7 7 8 9 9 7 Interest Payments (in % of total Revenues) 1 3 3 3 3 3 3 3 Source: 1996-97 are estimates based on official data; 1998-2003 are staff projections. Note: Assumes that accumulation of contingent liabilities ceases in 1997. Pension expenditures, analyzed separately in Annex 4 of this report, are excluded from the scope of the projections in the table. This has no major impact on the budget balance and financing requirement as pensions are more or less self financed 7. This performance will require a considerable reduction in the public sector's share of the economy. In addition, the composition of current expenditures, reflecting these reform polices, will have to change. In particular, spending in productive sectors, notably education, should 4 Bosnia and Herzegovina: Public Expenditure Review increase.2 Government administration will also increase for a number of reasons: (i) the wage and salary structure should change to minimize inter-Entity inequality and public/private sector disparity; and (ii) off-budget administrative expenditures should be moved on-budget (Chart 1). Chart 1. Bosnia and Herzegovina: Composition of Public Expenditures Medium Term Scenario: High Case (in percentage of Total Expenditures) 100 , Government Adm. O3Defense O3 Police and Justice _13Social Welfare 50 - 2 8 M Education o H-ealth 0EEconomic Develop. EInterest 3 13~~~ Investment 13Other 0 1997 2003 8. Reduced current expenditures have the effect of stabilizing current savings (revenues less non-investment expenditures) at a level sufficient to finance the investments needed to sustain growth. Even if the private sector eventually accounts for the bulk of capital investments, as is the case in market economies, in the medium-term government funded investment will play an important role in supporting and eventually carrying forward the donor-funded reconstruction program. General government investment spending of around 7 percent of GDP could be a sustainable medium-target, roughly equivalent to the average level in small and medium-sized OECD economies. 9. The tight fiscal stance, ruling out spending in excess of domestic resource capacity, leads to a significant decline in the primary deficit (balance of the total revenues and expenditures before interest payments), as donor concessional aid phases down toward the year 2000. Government interest payments stay at about 3 percent of total revenues, reflecting the favorable impact of lending and debt rescheduling by donors at highly generous terms. The budget deficit stabilizes at a moderate level, which prevents cumulative government borrowing. Similarly, the external current account deficit is projected to decline as reconstruction-led import growth moderates, and exports continue to recover, which reduces the need for debt financing. 2 Pensions, which are discussed in detail in Annex Six and Eight, are also excluded from the sectoral breakdowns. Annex One: Assumptions Underlying the Medium-Term Economic Framework 5 Creditworthiness indicators improve considerably over time, as reflected by improving debt service and debt-to-GDP ratios. Thereafter, financing at commercial terms is assumed to be available and affordable to finance modest budget deficits in line with improvements in the debt servicing capacity. Low Case 10. The low case assumes slower progress in institution-building, failure to aggressively implement structural reforms, and, accompanying, gradual disbursement of reconstruction financing and delayed and less comprehensive debt rescheduling, that only results in cash flow relief. Economic recovery is projected to be slow under this scenario, with GDP recovering to only about 40 percent of its pre-war level by 2000. Fiscal pressures arising from social demands, needs of enterprises and workers in industry, and high debt service obligations are assumed to be considerable under this scenario. Box 1.2: Medium-Term Framework: Underlying Assumptions of Low Case Scenario The low case scenario illustrates the possible consequences of slow progress in building cooperative economic institutions and implementing a program of economic adjustment and transformation. In the absence of rapid and comprehensive fiscal reforms, the existing ineffective tax system and the distortions in the allocation of public expenditures are assumed to remain largely unchanged in the medium-term. Failure to make timely reforms will continue to limit the efficiency with which scarce resources are allocated, and will feed back to lower GDP growth. As in the high case, the fiscal deficits are assumed to be financed predominantly by external funds, reflecting the monetary rules of the currency board arrangement and the limited potential to borrow from domestic capital markets. The low case scenario assumes that US$5.1 billion in external reconstruction assistance will be disbursed over a longer period (seven-eight years compared to three-four years in the high case). On this basis, real GDP and real exports are projected to recover to about 40-50 percent of the pre-war level by 2000 rather than the 60-70 percent recovery assumed in the high case scenario. Import growth is also projected to be slower, reflecting limited external financing and slower income growth in the economy. The inflow into the economy of non-debt creating financing, such as private transfers and foreign investment, is more limited compared to the high case scenario. Simulation results of the external account illustrate that additional financing needs over and above donor's reconstruction financing significantly increase over time. Financing is assumed to be provided entirely on concessional terms through 2000, and at mixed terms (i.e., including non-concessional financing) thereafter. Debt service obligations are projected to be higher under this scenario than in the high case due to: (i) debt service on rescheduled debt being higher because of further accumulation of the debt as a result of delays in Paris and London Club deals, and because of the difference in the terms of the deals (interest pt-'nents are incurred on a higher base since there is no debt stock reduction, and amortization starts early); and (ii) sizable financing gaps that are filled at harder terms after 2000. Debt sustainability indicators improve only marginally. The debt service ratio begins to increase over time to levels which imply that servicing of the external debt would be very difficult. This scenario assumes a flow rescheduling by the Paris Club on Naples terms with a 67 percent reduction in the present value of debi ccrvice payments for pre-cutoff debt by mid-1998 (a year later than in the high case). Post-cutoff arrears are assumed to be repaid within six years, with a three years grace period. Commercial creditors, including London Club, are assumed to provide treatment comparable to that given by the Paris Club for pre-cutoff debt, but one year after the Paris Club rescheduling arrangement. 6 Bosnia and Herzegovina: Public Expenditure Review 11. The analysis under the low case scenario demonstrates that the inability to undertake the structural and fiscal reforms that will promote growth, despite the availability of exceptional donor assistance, would lead to development of a fragile macroeconomic framework in the medium-term. Achieving the needed reforms at a later point in time, in the face of diminishing donor assistance, will certainly be much more difficult than if the reforms were made today (Table 1.2). Table 1.2: Medium Term Outlook, Low Case Scenario 1996-2003 Low Case Scenario (as percentage of GDP) 1996 1997 1998 1999 2000 2001 2002 2003 I. Total Revenues 56 72 56 49 43 38 35 34 Taxes 33 33 32 31 30 29 28 28 Nontax revenues (excluding grants) 3 5 5 5 5 4 4 4 Grants 20 33 19 13 8 5 3 2 I1. Current Expenditures (excluiding interest) 49 55 46 41 38 33 32 31 III. Current savings (I-1I) 7 16 10 8 5 5 3 3 11' Investments 14 27 15 12 8 6 4 4 V. Prinarv Balance ('1-11-1 ) -7 -11 -5 -4 -3 - I - I - I 11. Interest I I 1 2 2 2 2 2 VII. Total Expenditures (IH+ll+ VI) 64 84 63 55 48 41 38 37 1VI11 Buedget Balance (1-1 7) -7 -12 -6 -6 -5 -3 -3 -3 IX. Financing 7 12 7 6 5 3 3 3 Foreign 7 12 7 6 5 3 3 3 Borrowt,ing 7 12 8 9 7 7 7 7 (Repayment) 0 0 -I -2 -2 -4 -4 -4 A'Ietno itenms Real GDP growth (%) 50 20 15 10 8 6 6 5 Exports of goods and services (in %ofGDP) 20 22 23 24 25 26 26 26 Imports of goods and services (in % of GDP) 73 67 50 40 36 32 30 29 Current Account Balance (in % of GDP) -26 -33 -21 -13 -9 -5 -4 -3 Total External Debt (in % of GDP) 114 1()8 99 99 93 89 85 82 Debt Ser\ice (in % of total exports) 87 36 24 19 17 22 22 20 Interest Paxyments (in % of total Revenues) I I I I I 1 4 4 Source: 1996 is an estimate based on official data, 1997-2003 are staff projections. Note: Assumes that accumulation of contingent liabilities ceases in 1997. Pension expenditures. analyzed separately in Annex 4 of this report. are excluded from the scope of the projections in the table. This lhas no major impact oi1 the budget balance and financgin requirement as pensions are mor-c or less self financed. Annex One: Assumptions Underlying the Medium-Term Economic Framework 7 12. Delays in tax reform, lower incomes and profitability of businesses under the low case results in lower tax revenues. The government's bias towards current expenditures would be exacerbated by: (i) delays in restructuring and privatization of enterprises that would likely lead to increased pressures for subsidies and transfers to enterprises that would be used for operational rather than investment purposes; (ii) high and increasing social support transfers caused by, among other reasons, failure to achieve strong recovery and employment generation; and (iii) delays in reducing the large size and high cost of the government administration. The bias can be seen in rapid declines over time in the current budgetary savings. The result is that public investment levels will be squeezed as donor assistance diminishes (Chart 2). The composition of current expenditures also remains biased towards unproductive sectors, which absorb more than two thirds of allocations by 2003, compared to about 40 percent in 1997. Such an outcome, in turn, would work against the objective of developing private activities in Bosnia and Herzegovina, and would feed back into lower overall growth in income. It is worthwhile noting that what derails the recovery process under this scenario is not the size of government, but the speed and quality of government reforms. Chart 2. Bosnia and Herzegovina: Composition of Public Expenditures Medium Term Scenario: Low Case (in percentage of Total Expenditures) 100 _ MGovemenAd. NNIM M~~~~~~~~Defense mom = E3 Police and Justice S ~~ t; 0; 0014 13 ESocial Welfare E3 Education ElHealtls 1 133 6 ; = IEconomic Develop. 33 E3 I~~~~~~~~~~~~6mnterest fllnvestrnent 1 2 E MOther 0 1997 2003 13. The maintenance of tight fiscal rules under the low case results in a modest budget deficit, stabilizing at around 3 percent of GDP. Its manageability is, however, bounded by the availability of foreign financing. Bosnia and Herzegovina's ability to mobilize the total amount of needed financing suggested by the low case is questionable, given its weak commitment to reform on the one hand, and a negligible improvement in its creditworthiness on the other. The latter is partly due to lower growth in GDP and export revenues and partly to the nature of the debt rescheduling assumptions (Box 1.2). Under this scenario, in order to meet obligations on the rescheduled debt alone, the country would need about DM 400 million a year in financing 8 Bosnia and Herzegovina: Puiblic Expenditure Review between 2000-2003, equivalent to four percent of GDP. It does not seem plausible for Bosnia and Herzegovina to obtain significant additional concessional financing in excess of what is needed to finance priority obligations (i.e. on the rescheduled debt and credits obtained from donors in the context of the reconstruction program). Commercial lending would likely be limited and also less affordable. This would impose additional constraints on public expenditures and, given the bias towards unproductive sectors, risk further lowering overall GDP growth. Annex Two Military Expenditures 1. The heavy burden of military expenditures and transfers clearly suggests that reduced spending must first be considered in these areas when attempting to put Bosnia and Herzegovina on the path to sustainable fiscal policy. Obviously, military spending is substantially driven by the level of political tensions. Recommendations to reduce military expenditures are therefore meant primarily to underscore the economic and fiscal consequences of continued tensions. That is, they are meant to emphasize the importance of continuing intensive and good-faith efforts to reduce tensions on all sides. 2. Military expenditures were off-budget in 1996 so there are few available data on actual expenditures prior to 1997. Defense expenditures (excluding benefits to veterans) make up 41 percent of the 1997 Federation budget of DM 615 million, and 20 percent of the consolidated Federation budget (which includes funds from cantonal and municipal budgets). In Republika Srpska, budgeted military expenditures are 26 percent of total expenditures for 1997. These ratios do not include substantial off-budget items. In order to evaluate the total fiscal burden that military spending creates, it is therefore necessary to estimate all actual military expenditures. There is no accepted estimate of military expenditures, although all parties agree that the Federation budget for military expenditures will not even cover wages and salaries in 1997. A rough attempt at estimating total expenditures is reported below. 3. The most reliable infonnation available is troop strength. At the height of the war, SFOR estimated that the Croat-majority army had 46,700 soldiers, the Bosniac-majority Republic of Bosnia and Herzegovina army had 152,000 soldiers and the army of Republika Srpska had 151,000 soldiers.' After the Peace Agreement, demobilization was quite rapid; current troop levels are estimated at 56,000 in the Federation and 18,000 in Republika Srpska. According to SFOR, the Federation intends to reduce troop strength further to 35,000. In contrast, Republika Srpska is not expected to reduce troop strength from its current size (Table 2.1).2 Table 2.1: Troop Strength (in thousands) Federation RS Croat-majority Bosniac-majority Estimated Maximum Size in War 46.7 152.0 151.0 Current Force Size (March 1997) 18.0 38.0 18.0 Estimated Future Size 10.0 25.0 18.0 Source: SFOR estimates. However, there is considerable disagreement over these figures; e.g., the Deputy Ministry of Defense reports that the maximum troop size for the army of the Republic of Bosnia and Herzegovina was 210,000. 2 While the Federation Army is currently over twice the size of the Republika Srpska Army, this is likely to be in part due to tensions between the Croat and Bosniac Federation partners: neither is confident that the other will be available in possible conflicts with third parties. 9 10 Bosnia and Herzegovina: Public Expenditure Review 4. Based on average salaries of enlisted men (provided by the Federation Ministry of Defense) and estimates of salaries for officers from anecdotal evidence collected by SFOR, it is possible to estimate roughly how much salaries would cost for the present Federation army and for a Federation army of 35,000. The calculations are based on the assumption of both a "7 percent officer" and a "10 percent officer" structure (Table 2.2).3 Table 2.2: Estimated Expenditures on Military Salaries for the Federation and RS (in DM millions) 10% Officers 7% Officers Federation RS Federation RS Current Force Size 376.1 56.2 363.0 52.3 Estimated Future Size 233.1 56.2 225.0 52.3 Source: SFOR estimates; Federation Ministry of Defense data. Note: As of March 1997, average salaries for enlisted men in the Bosniac- and Croat-majority portions of the Federation army were, respectively, 300 DM and 900 DM per month. Average officers' salaries were estimated to be 1,000 DM and 1,500 DM per month. Exact figures for salaries in the Republika Srpska army are not known, although they are thought to be less than salaries in the Federation. SFOR estimates 200 DM for enlisted men and 800 DM for officers. 5. Based on the estimates of expenditures on military salaries, Table 2.3 presents estimates of total military expenditures for the current and future estimated force size. These numbers were calculated by assuming that salaries make up approximately 55 percent of total military expenditures in the Federation and Republika Srpska.4 For all middle income countries for which data were available, military personnel expenditures were approximately 40 percent of total military expenditures. The actual share of expenditures on personnel in Bosnia and Herzegovina is not likely to be this low because diplomatic protocols and financial difficulties will constrain operational expenditures (e.g., training exercises and maneuvers), the internationally-funded "Train and Equip" program lowers non-personnel costs (equipment and training), and because the army uses a relatively low level technology. Using a lower share of personnel expenditures would of course increase estimated total military expenditures.5 3SFOR contacts indicate that low technology levels may make a "7-8 percent officer' assumption most reasonable. 4SFOR contacts suggest that a reasonable estimate of salary expenditures would be 55 percent of total (excluding the costs of the "Train and Equip" program), based on comparisons with the United States. In the United States Army in 1996, approximately 41 percent of expenditures were on military pay (and other military personnel costs). An additional 13 percent went towards pay (and other personnel costs) for civilians in the military. The estimates for personnel costs for Bosnia and Herzegovina do not include civilian pay. It was suggested that the figure for Republika Srpska should be lower than the 55 percent estimate for the Federation because Republika Srpska is likely to have higher equipment costs due to its outmoded equipment (and because it does not participate in the Train and Equip program). Therefore, the estimates of total expenditures for Republika Srpska may be too low. 5A higher share might not be appropriate because salary costs are very low in Bosnia and Herzegovina relative to the U.S., while many goods that the military requires (such as diesel, spare parts, and ammunition) are priced intemationally. Annex Two. M'ilitary Expenditures 11 Table 2.3: Estimated Annual Military Expenditures for the Federation and the RS (in DM million) 10% Officers 7% Officers Federation RS Federation RS Current Force Size 683.8 102.1 659.3 95.0 Estimated Future Size 423.5 102.1 409.0 95.0 Source: SFOR estimates; Federation Ministry of Defense data. 6. The total, conservatively estimated expenditure required for the Federation to maintain the arny at its current level is between DM 660-680 million DM per year. This is far greater than the amounts allocated to the military in either the Federation or consolidated budgets. In the Federation, salaries alone would cost over DM 350 million, nearly half as much again as the current Federation military budget (and greater than the consolidated Federation military budget). In the short term, some of the shortfall is likely to be covered by grants for the military from donor countries. In the medium term, a reduced Federation army of 35,000 is estimated to cost some DM 420 million to maintain, which would increase over time with increasing professionalization. As the flow of donor funds slows, this level of expenditures will become increasingly difficult to maintain.6 7. For Republika Srpska, a similar calculation yields an estimate of military expenditures in Republika Srpska of some DM 95-102 million for 1997. This estimate is slightly greater than the DM 89 million budgeted for the Republika Srpska Ministry of Defense. Spending is considerably less than for the Federation army, both because the army is much smaller and because salaries are lower. However, defense spending is not more fiscally sustainable in Republika Srpska, despite the similarity of budgeted and likely expenditures, because budgets in that Entity significantly overstate revenues (Republika Srpska has reportedly operated with several months of arrears in soldiers' wages). Unlike the Federation army, the Republika Srpska army is not expected to shrink, and so there will not be any savings in the medium term. 6 It is possible that money budgeted for public order and safety is also being used to fund military or paramilitary activities (in both the Federation and Republika Srpska). However, SFOR note that they do not include individuals in paramilitary groups in their estimates of force size. Hence, these groups are also omitted from cost estimates. The Federation has budgeted DM.35 million for public order in 1997 (with an additional DM 176 million at the sub- Entity level), while Republika Srpska has budgeted DM 52 million for public order and safety. Annex Three Public Administration in Entity and Cantonal Governments 1. A summary of all government service employment by employing institution as of early 1997 is shown in Table 3.1. The remainder of the annex describes in greater detail issues of public administration in sub-State governments. Table 3.1: Public Service Staff Numbers and Costs Total government sector staff Staff nos. (%) Civil staff Avg. gross Total civil exp. Staff exp./ Staff exp./ exp. (DM) wage (DM) total exp. GDP (DM/month) State of BiH 900 0.72% 15,213,600 1409 320,342,420 5% 0.26% RS 35,000 27.82% 58,683,402 140 289,331,922 20% 1.00% RS municipalities/ health: 7,600 6.04% 12,831,659 141 18,330,942 70% 0.22% estimated Federation-federal government 5,100 4.05% 59,847,500 978 360,581,000 17% 1.02% Federation-cantons 69,500 55.25% 576,019,304 691 1,001,250,189 58% 9.78% Federation-municipalities 7,700 6.12% 52,329,330 566 74,756,185 70% 0.89% Grand total 125,800 100% 774,924,795 513 2,064,592,658 38% 13.16% Notes: GDP World Bank estimate; exchange rate as of April 10, 1997. All other data taken from tables below. Total staff numbers rounded to nearest 100. The Federation 2. The Federation of Bosnia and Herzegovina has largely implemented the new institutional structure introduced by the Washington Agreements, broadly complying with the decentralization required under the constitution. The Federation now employs fewer than 5 percent of the total number of government employees in the country, excluding military personnel (Table 3.1). Of the total 5,114 civilian employees in the Federation goverment, about 80 percent are employed in three ministries: Finance, Interior, and Justice. The remaining 24 ministries and other institutions each account for no more than one to two percent of total employees (Table 3.2). There may be significant overhead costs associated with maintaining so many institutions, however, and the present structure should be reviewed with the aim of rationalizing the number of ministries. 3. While the Federation has largely implemented the new institutional structure, important issues remain. Work is still needed to rationalize the large number of ministries which now perform few functions; re-allocating the 600 staff who have been engaged on 'State' functions; defining the Federation's role in relation to the cantons; and developing the capacity to promote efficient delivery of public services at the cantonal level. Employment levels in government are below the average for comparable countries, although service levels may be lower, particularly given the paucity of complementary inputs (such as police cars for police personnel). No doubt there will be pressure on the various administrations to recruit more staff if, and when, the armies de-mobilize and as refugees return. This should be resisted, particularly if the productivity of existing staff is low. 12 Bosnia andl Herzegovina: Public Expenditure Review 13 Table 3.2: Federation Employment and Salaries Federation of BiH Staff lbos. % Civil staff Avg. gross Total civil Staff exp./ Federal government exp. DM wage exp. DM total exp. DM/month I Federal Parliament 82 2% 1,020.000 1,037 1,381.900 74% 2 Presidency 14 (% 182,900 1,089 264,200 69% 3 Federal cabinet 100 2% 1,455.200 1,213 2,923,000 50% 4 Ministry of Interior 1.319 26% 15,828,000 1,000 20,576,400 77% 5 Ministry of Justice 552 11% 6,044,000 912 12,709,100 48% 6 Ministry of Finance 2,074 41% 23,631,100 949 29,676,200 80% 7 Ministry of Energy, Mining ald Industry 94 2% 1,218,200 1,080 2,083,100 58% 8 Ministry of Transport and Commiiuniications 37 1% 437,300 985 20,633,400 2% 9 Min. of Social Affairs, Refugees, 162 3% 1,683,500 866 2,104,300 80% and Displaced Persons 10 Ministry of Health 144 3% 1,672,500 968 6,674,100 25% 11 Min of Education, Science, Culture and Sport 59 1% 751,800 1,062 5,477,300 14% 12 Ministry ofTrade 73 1% 958,300 1,094 1,197,800 80% 13 Min. of Physical Planning and Environment 33 1% 400,000 1,010 570,100 70% 14 Min. of Agriculture and Water Managemelit 65 1% 768,300 985 11.060,300 7% 15 Constitutional Court II 0% 179,700 1,361 284,100 63% 16 Supreme Court 44 1% 596,600 1,130 885,700 67% 17 Federal Prosecutor's Office 13 0% 203.200 1,303 253,200 80% 18 Federal Legal Office 5 0% 54,800 913 68,500 80% 19 Penalty Council ofViolations 12 0% i75,100 1,216 218,800 80% 20 War Veterans Agency 19 0% 219,100 961 180,273,800 0% 21 Geodesic and Property Rights Agency 35 1% 410,300 977 532,000 77% 22 N/A 5 0% 58,600 977 73,200 80% 23 Bureau for Statistics 80 2% 937,900 977 1,188,300 79% 24 Bureau for Meteorology 50 1% 586,200 977 732,200 80% 25 Federal Archive 8 0% 93,700 976 117,100 80% 26 Agency for the reserve of goods 12 0% 140,600 976 175,700 80% 27 Bureau for programming development 12 0% 140,600 976 175,700 80% Sub-total operations 5,114 100% 59,847.500 302,309,500 20% 28 Other expenditures 41,100,000 29 Reserves 17,171,500 Total 5,114 100% 59,847,500 360,581,000 17% Source: Draft Federation 1997 budget. 4. The 1994 Washington Agreements created ten highly autonomous cantons in the Federation. Apart from the limited functions specifically reserved for the federal government, all powers belong to the cantons. As of March 1997, the cantons were at varying stages of development. At one extreme, Sarajevo -- which has had a city administration for many years -- appears to have formed its organization, including new functions such as police, and put in place a draft budget. It is now probably the single largest administrative unit in the country, with an annual budget of some DM 430 million (Table 3.3). Anecdotal evidence suggests that it is functioning well. At the other extreme, Mostar continues to be divided between the Bosniac and Croat communities, with no common cantonal functions yet established except a common police force of 100 officers (deployed on April 4, 1997) to operate in the town's central zone. One particular barrier to the formation of the new cantons is the continued existence of two payments bureaus within the Federation, which unnecessarily complicates canton financing. It is essential that these are effectively unified as soon as possible. 14 Annex Three: Public Administration in Entity and Cantonal Governments Table 3.3: Sarajevo Canton: Budgeted Salaries Budget heading Total Salaries Total Expenditure (DM) (DM) Parliament of Sarajevo Canton 384,000 732,000 Cantonal Government 1,038,000 16,046,000 Ministry of Justice and Administration 6,462,000 9,450,000 Ministry for Problems of Soldiers 95,000 36,069,000 Ministry of Traffic and Communication 330,000 34,534,000 Ministry of Town Planning, Housing and Public Utilities 1,700,000 29,272,000 Ministry of Internal Affairs (Police) 21,380,000 35,840,000 Ministry of Economy 696,000 40,775,000 Ministry of Finance 310,000 25,250,000 Ministry of Reconstruction, Development and Environmental Protection 210,000 3,379,000 Ministry of Health 65,258,000 72,922,000 Ministry of Education, Science and Information 48,120,000 74,790,000 Ministry of Culture and Sport 4,160,000 9,924,000 Ministry for Labor, Social Affairs, Displaced Persons and Refugees 4,084,000 22,430,000 Inspectorate 1,980,000 2,570,000 Professional Fire Brigade 1,980,000 3,060,000 Institution for Construction of Sarajevo Canton 780,000 1,032,000 Institution for Town Planning Sarajevo 768,000 2,688,000 Department for Common Services 570,000 3,002,000 Headquarters of Civil Protection 195,000 333,000 Political Parties, Associations and other organizations N/A 1,310,000 Reserve 4,000,000 Total 160,500,000 429,408,000 Source: Draft Sarajevo Municipality 1997 budget. 5. Given the federal structure of public administration, decisions on government employment are now made by each of the 11 main institutions: the federal government and the ten cantons. Each institution sets its own labor laws and employment practices. There seems to be some consensus, however, that it would be wise for the Federation to set some defining framework and cantons seem ready to follow a lead set by the Federation government. In the meantime it appears that each institution is following established eastern European practice, by allowing each ministry (both at federal and cantonal level) to recruit its own staff, with numbers subject to some measure of central control. No institution interviewed had a central database of personnel records, which is a severe control weakness and exposes all institutions to the risk of unauthorized payments. Moreover, recruitment procedures have not been regularized to ensure merit-based, rather than politically-based, civil service development. 6. Practically no information is available about municipality staffing and costs, and so estimates have been made, based on the little information available in Sarajevo and East Mostar. The data on the municipalities in these areas has been used to estimate the numbers of staff in all municipalities, based on the number of staff per capita. On this basis there would be 7,000-8,000 municipal employees, but this is subject to verification. Bosnia and Herzegovina: Public Expenditure Review 15 Republika Srpska 7. Republika Srpska has not had to implement institutional change on the same scale as the Federation and it appears that its institutions are largely formed. A full list of ministries as of early 1997, with employee numbers and costs is in Table 3.4. Table 3.4: Republika Srpska: Employment and Salaries Republika Srpska Staff nos. % Civil staff Avg. gross Total civil Staff exp./ exp. DM wage exp. DM total exp. DM/month Presidency 30 0.08% 113,618 316 1,457,903 8% National Assembly 95 0.27% 295,530 259 1,830,958 16% Government 148 0.42% 348,475 196 3,430,102 10% Constitutional Court 13 0.04% 46,821 300 70,232 67% Supreme Court I1 0.03% 34,878 264 69,757 50% Republic Public Prosecutor 3 0.01% 9,800 272 19,600 50% Republic Public Defense (Attorney) 29 0.08% 66,216 190 99,325 67% Ministry of Defense 638 1.79% 1,340,037 175 112,733,194 1% Ministry of Foreign Affairs 12 0.03% 33,805 235 878,238 4% Ministry for Interior Affairs 12,453 34.99% 24,292,000 163 59,295,142 41% Ministry of Education 16,914 47.53% 22,195,072 109 44,698,164 50% Ministry of Finance 681 1.91% 1,558,942 191 39,709,842 4% Ministry of Justice 1,815 5.10% 3,291,494 151 4,867,648 68% Ministry for Management and Local Self 0 0.00% 37,071 - 51,900 71% Government Ministry for Industry and Technology 0 0.00% 37,071 - 51,900 71% Ministry for Health and Social Protection 157 0.44% 76,285 40 5,250,571 1% Ministry of Energy and Mining 17 0.05% 49,248 241 2,931,285 2% Ministry for Agriculture, Forestry and Water 80 0.22% 225,350 235 2,640,691 9% Ministry for Transport and Communication 81 0.23% 188,476 194 9,521,010 2% Ministry of Science and Culture 58 0.16% 133,255 191 758,085 18% Ministry ofTrade and Tourism 150 0.42% 418,857 233 29,611,423 1% Ministry for Urban Planning etc. 27 0.08% 71,428 220 421,428 17% Ministry for Information 1,179 3.31% 1,627,712 115 5,827,715 28% Ministryforthe lssuesofSoldiersandLabor 46 0.13% 116,571 211 68,740,942 0% Ministry for Economic Relations with Abroad 0 0.00% 31,428 - 44,000 71% & Development Ministry for Refugees and Displaced Persons 388 1.09% 928,856 199 3,179,714 29% Ministry of Confessions 8 0.02% 23,142 241 32,400 71% Ministry for Sport and Youth 5 0.01% 12657 211 17,714 71% Republic Bureau for Geodesic and Property 23 0.06% 1205142 4,366 1,748,000 69% Rights' Affairs Republic Customs Administration 342 0.96% 642714 157 1,028.342 63% Financial Police 113 0.32% 368571 272 737,142 50% Republic Foreign Currency Inspectioni 0 0.00% 45.142 - 81,142 56% Republic Bureau for Statistics 40 0.11% 87277 182 130915 67% Republic Bureau for Meteorology 23 0.06% 47.314 171 66,240 71% Republic Bureau for Planning 6 0.02% 14.514 202 20,314 71% Republic Secretariat for Legislature 3 0.01% 8,670 241 12,138 71% Total 35,588 100.00% 60,023,439 141 402,065,116 15% less Ministry of Defense 638 1,340.037 112,733,194 1% Total civil affairs 34,950 58,683,402 289.331,922 20% Total includinig Ministry of Defense, plus Financing the Organs of BiH 2,545,714 23,564,000 11% Source: Budget data obtained from the Republika Srpska budget. Staff data were obtained from Republika Srpska government, but in a form which does not match the employing institution to a budget heading, which has, therefore, been deduced (see Table 3.5 below). 16 Annex Three: Public Administration in Entity and Cantonal Governments Table 3.5: Republika Srpska: Number of Employees by Employing Institution Institution Assumed budget heading (not confirmed) Total number of staff Constitutional Court Constitutional Court 13 Supreme Court' Supreme Court II Republic Public Prosecutor Republic Public Prosecutor 3 Rep. Public Defense (Attorney) Republic Public Defense (Attorney) 29 Courts Ministry of Justice 1396 Presidency Presidency 30 National Assembly National Assembly 16 Rep. reconstruction & development Ministry for Refugees and Displaced Persons 12 Assembly representatives National Assembly 79 Government Government 144 Ministry of Defense Ministry of Defense 638 Ministry of Finance Ministry of Finance 22 Min. of Education & Culture Ministry of Education 32 Rep. Bureau for Pedagogy Ministry of Education 56 Culture at Republic level Ministry of Education 388 Primary schools Ministry of Education 10,846 Secondary schools Ministry of Education 4,468 Tertiary education Ministry of Education 1,124 Min. of Health Ministry for Health and Social Protection 31 Min. of War Veterans Ministry for the Issues of Soldiers and Labor 25 Min. of Interior Ministry for Interior Affairs 12,453 Min. of Justice Ministry of Justice 27 Min. of Foreig.n Affairs Ministry of Foreign Affairs 8 Min. of Transport & Communications Ministry for Transport and Communication 21 Min. of Flight Ministry for Transport and Communication 26 Min. of Trade Ministry of Trade and Tourism 150 Min. of Agriculture Ministry for Agriculture, Forestry and Water 68 Min. of Religion Ministry of Confessions 8 Min. of Information Ministry for Information 14 Min. for City Planning Ministry for Urban Planning etc. 27 Min. for Industry & Energy Ministry of Energy and Mining 17 Rep. Secretariat for Legislation Republic Secretariat for Legislature 3 Rep. Administration for Public Revenue Ministry of Finance 659 Rep. Geodesy Admin. Republic Bureau for Geodesic and Property Rights' Affairs 23 Customs Admin. Republic Customs Administration 342 Rep. Statistics Bureau Republic Bureau for Statistics 40 Rep. Meteorology Bureau Republic Bureau for Meteorology 23 Rep. Bureau for Health Protection Ministry for Health and Social Protection 126 Rep. Planning Bureau Republic Bureau for Planning 6 Rep. Water Management Directorate Ministry for Agriculture, Forestry and Water 12 Rep. Directorate for Succession of ex-Yugoslavia Ministry of Foreign Atfairs 4 Rep. Department for Refugees Ministry for Refugees and Displaced Persons 101 Min. of Information - Document Center Ministry for Information 2 International Press Center Ministry for Information i I Srpska Radio and TV Ministry for Information 541 Srpska News Agency Ministry for Information 154 Srpska Publishing & Printing News Company Ministry for Information 445 Publishing House Ministry for Information 12 Museum of RS Ministry of Science and Culture 19 Archives Ministry of Science and Culture 30 Prisons Ministry of Justice 392 insurance Company 33 Insurance Company 9 Rep. Directorate for Roads Ministry for Transport and Communlicationi 31 Bureau for State Monuments Ministry of Science and Culture 9 Miii. for Refugees and Displaced Persons Ministry for Refugees and Displaced Persons 275 Air Ministry for Transport and Communication 3 Rep. Protocol Government 4 Min. of Sports Ministry for Sport and Youth 5 Min. of Work Ministry for the Issues of Soldiers and Labor 21 Financial Police Financial Police 113 35,574 Note: Budget headings do not match institutions, and rough estimates have had to be made. Bosnic and Herzegovina: Public Expendilure Review 17 8. The government comprises 134 institutions, but only 36 budget heads, and it is not immediately obvious which staff are funded from which budgets. A tentative allocation of institutions to budgets is illustrated in Table 3.5. More than 80 percent of civil servants are employed in just two ministries: Education and Interior Affairs. The great majority of ministries each employ fewer than one percent of the public service and, as with the Federation, there would appear to be a case for rationalizing the number of ministries. 9. There is no cantonal government in Republika Srpska, but there are about 55 municipalities. Data on these, and on the Entity's health service, are not held centrally. Municipal and health sector employment has been estimated at about 7,600 based on the same ratio of municipal and health workers to population as in the Federation. 10. Some 16,438 government employees in Republika Srpska or 46 percent were in the education sector, leaving around 19,150 in general government administration. The latter is equivalent to about 2.5 percent of the Entity population and is much higher than the one percent average that general administration reaches in most countries of Eastern Europe and the Former Soviet Union. Much of this difference is due to the very large Ministry of Interior Affairs. Excluding this Ministry, total public sector employment may be somewhat higher than the regional average but not unduly so. 11. In contrast, employment in education, at 2.2 percent of population is lower than the regional average of 3.1 percent. Data on health workers in the Entity could not be collected on this mission, as they are funded separately from both central and municipal expenditures. Until further resources are devoted to a more accurate estimate, it has been estimated that there are about 5,600 employees in the health sector. If correct, fewer than one percent of the population are employed in the health sector, lower than the regional average of 1.5 percent. 12. Similarly, no central data exists for employment in the 55 municipalities, but there are probably no more than about 2,000 staff, well in line with, or less than, the regional average. Possibilities for efficient decentralization, if and when this becomes realistic, would not be constrained by excessive local employment today. Overall, government service employment in Republika Srpska, at some 43,000 staff (including health and municipalities), is equivalent to about 5.7 percent of population. 13. In 1997, the legal minimum wage in Republika Srpska was set at about DM 40 per month. However, salaries range from about DM 80 to DM 280. The compression ratio is about 7:1; in practice it is increased to about 8:1 by bonuses and allowances for senior personnel, which may amount to 15 percent of salary. However, very few, if any, employees appear to be paid the minimum wage, and so the apparent compression ratio is probably over-stated. Chronic wage arrears, in any event, remove much meaning from the wage calculation. 14. The criteria for recruitment appear loose and highly informal in Republika Srpska. Although formal education is clearly given an important role, other non-meritocratic criteria appear to be relevant. Among these criteria is a preference for ex-soldiers, especially where they are returning to a position that they held before the war. The implications for the skill profile of 18 Annex Three: Public Administration in Entity and (C'antonal (overnmenLts government employees, and consequently for the effliciency of public administration, may become significant. 15. Of greater concern from a fiscal point of view is that Ministry requests for personnel are reviewed and approved by the Cabinet on a case-by-case basis, and it seems that there is no systematic attempt to reconcile total requests with fiscal objectives. Instead, the "current reserve" item in the preliminary budget is used to cover excesses in the wage bill. As in the Federation, the absence of manpower control is a significant handicap to an efficient budget process. 16. The Ministry of Finance does have a central payroll and database. Although limited to the information needed to determine the validity of the salary payment, this system offers some protection against the emergence of 'phantom' workers and unwarranted payments. Annex Four Pensions 1. Pensions in BiH are managed by three separate institutions. Each pension plan suffers from a severe mismatch between promised benefits and actual revenues, the resulting accumulation of arrears, and a failure to adequately target benefits. The Sarajevo-based fund is the administrative successor to the pre-war Bosnian pension fund and has maintained the record-keeping link between the past and the present. In early 1997, the fund was receiving contributions of 24.5 percent of gross wages (21.6 percent of total labor costs) from approximately 200,000 workers' and was paying pensions to around 185,000 beneficiaries in the Federation. The scheme continues to operate under the former Yugoslav legislation with various exceptions. For example, its extra- budgetary status was reinstated only in February 1997. 2. The second Federation pension fund is headquartered in Mostar. It receives contributions of 17.5 percent of gross wages2 from about 45,000 workers and pays pensions to more than 43,000 pensioners in the Federation. There are about 185 staff working in 28 branch offices and 76 staff in the central office in Mostar. Although a series of decrees regulates the activities of the fund, both the Sarajevo- and Mostar-based funds will continue to exist in a legal vacuum until legislation is passed at the Federation level. 3. Finally, the pension scheme of Republika Srpska has branches in each municipality in addition to a central office in Pale. There are about 160,000 workers required to contribute to the fund and 158,000 pensioners. The contribution rate is 22 percent of gross wages. In fact, the number of contributors is far less because many firms do not pay taxes and have accumulated significant arrears. On the other side of the equation, pensioners had not received any payments during the first quarter of 1997. The formula for determining benefits, along with the other rules, are determined by separate legislation. 4. Retirement age in all three systems is 55 for women and 60 for men, five to ten years younger than in most OECD countries. The use of early retirement was growing before the war, and was extensive in Republika Srpska. This has now been restricted by measures such as the elimination of the possibility of purchasing serx ite years. Benefits in the Sarajevo-based and Republika Srpska funds are related to past earnings. Since its establishment in 1993, the Mostar-based fund has paid a flat cash benefit determined by the amount of revenue collection and does not record arrears. I While 260,000 workers were officially employed in the Bosniac-majority area in December 1996, only about 200,000 were actually being paid during that month. 2 In the Croat-majority area, all payroll taxes are paid by the employee so that gross wage equals total labor cost. 19 20 Bosnia and Herzegovina: Public Expenditure Review 5. The existence of three pension funds in Bosnia is an anomaly which can be directly traced to the political conflict. It poses several problems in terms of economic efficiency. First, it has led to excessive administrative costs as the funds are forced to replicate record-keeping and other functions. The cost of administering pensions in neighboring countries such as Hungary or Albania is generally less than two percent of total expenditures. The Mostar-based fund, with fewer than 100,000 participants, is not able to take advantage of economies of scale and charges more than fifteen percent in overhead. This translates directly into lower net pensions. The existence of separate funds is also likely to inhibit labor mobility. 6. Pension spending in Bosnia grew dramatically in 1996 in the Federation, the first year of post-war economic recovery, because of a significant recovery-driven increase in the total wage bill and in worker contributions to the pension systems. Average benefits for the 185,000 pensioners in the Sarajevo-based pension fund rose from around DM 11 in 1995 to more than DM 55 in 1996, taking into account several months of nonpayment. Pension levels remained steady for pensioners in the Mostar-based fund at around DM 75. The lowest pension levels were found in Republika Srpska where the average net pension during 1996 remained less than DM 25. 7. Rising pension expenditures in 1996 were possible because of a sharp increase in revenues, due to a rapid increase in wages and employment in the Federation. Employment increased from 221,171 to around 260,000 while the number of paid employees increased from around 150,000 to around 200,000 between January and December 1996. Average net wages almost tripled from about DM 74 to DM 225 during the same period. The tax base for pensions and other social insurance programs rose by nearly 400 percent. Despite a full seven percentage point reduction in the earmarked pension tax for the Sarajevo-based fund in March, revenues increased by 280 percent. In addition, external donors provided DM 15 million to the Federation pension funds. In the Federation, recovery meant convergence of pension levels in the two funds while for Republika Srpska, the average pension levels fell even further behind those paid to Federation pensioners. 8. As shown in the main text, payroll tax rates in BiH are already excessive. As a result, except for external donor support, spending in the two funds is likely to be driven by available revenues, and benefit levels will, therefore, rise only as the total wage bill (in the formal sector) increases. 9. Table 4.1 below presents a baseline scenario for monthly pension revenues and spending in the Sarajevo-based fund during 1997, and demonstrates that wage growth will be insufficient to fully fund unchanged benefit levels. Revenue growth is consistent with the "high case" macroeconomic assumptions and is based on a constant payroll tax rate for pensions. A constant labor share of GDP is assumed so that the wage bill will Annex Four: Pensions 21 grow at the same rate as GDP, in this base case at 20 percent.3 A detailed breakdown is presented to show how increasing revenues translate into higher net pensions, administrative costs and transfers to cover pensioners' health insurance.4 Table 4.1: Projected Budget Results for Sarajevo-Based Pension Fund, 1997, under Baseline Assumption, DM Million Except where Stated Payroll tax Number of Average net Transfer to Admin. Pension fund revenues pensioners pension (DM) Health Fund Costs expenditures Jan-97 19.4 185,000 88 2.7 0.4 19.4 Feb-97 19.4 185,200 88 2.7 0.4 19.4 Mar-97 19.5 185,400 88 2.7 0.4 19.5 Apr-97 19.5 185,600 88 2.7 0.4 19.5 May-97 19.8 185,800 90 2.8 0.4 19.8 Jun-97 19.8 186,000 89 2.8 0.4 19.8 Jul-97 20.5 186,200 92 2.9 0.4 20.5 Aug-97 21.4 186,400 96 3.0 0.4 21.4 Sep-97 22.2 186,600 100 3.1 0.4 22.2 Oct-97 23.1 186,800 104 3.2 0.5 23.1 Nov-97 24.0 187,000 108 3.4 0.5 24.0 Dec-97 25.0 187,200 112 3.5 0.5 25.0 Total 253.6 N/A. 1,144.3 35.5 5.1 253.6 Average 21.1 186,100 95.4 3.0 0.4 21.1 Note: Revenues are forecast by the Sarajevo Pension Fund. This projection is consistent with a 20 percent growth in the tax base over the course of the year. The fund is assumed only to have payroll tax revenues. The payroll tax rate is assumed to remain constant throughout the period at 24.5 percent of gross wages. 10. Projected revenue growth leads to a significant increase in average net pensions. However, even if only a small net increase in the number of pensioners is assumed, pensions at the end of 1997 are not likely to be more than DM 110 per month. With the current distribution of benefits, this will leave many pensioners with less than DM 50 per month and more than 100,000 pensioners will receive less than DM 100 per month. Meanwhile, some pensioners will continue to receive more than DM 300 per month. In the Mostar-based fund, net pensions are likely to continue to be slightly lower than in the Sarajevo-based fund. Given similar growth of the taxable wage base net pensions should increase, but will remain below DM 100 per month. 3 The assumed wage and employment growth must generate this 20 percent increase in the wage bill. In these projections, employment growth in the Bosniac area was assumed to continue at its 1996 pace of about 3,000 per month while wages were assumed to grow at an annualized rate of seven percent. 4 Pension expenditures differ from Pension Fund expenditures in that transfers for health insurance are not included in the former in order to avoid double counting in the consolidated government budget. 22 Bosnia and Herzegovina: Public Expenditure Review 11. The situation is not as clear in Republika Srpska, where the pension scheme received direct transfers from the central budget in 1996 and is counting on such transfers again in 1997. Planned expenditures for 1997 would increase pension spending to more than seven percent of GDP in Republika Srpska. It is more likely, however, that the anticipated revenues from both the central budget and payroll taxes are vastly overestimated. If the wage bill grows by 20 percent, and central budget transfers remain at 1996 levels, total spending will be constrained to about 54 million DM or about 3.9 percent of projected GDP. Of this, as much as one third would be needed for administration and health insurance payments. Assuming no increase in the number of pensioners, this would allow for a slight increase in average net pensions to about 29 DM. Annex Five Sector by Sector Review of Infrastructure Policy Transport 1. The war caused substantial damage to the transport system and interrupted all maintenance activity for more than four years. Since the end of the hostilities, transport infrastructure conditions have improved significantly, but services provided and management are still far from the pre-war level. In the road sector, more than half of the main network still needs to be rehabilitated, maintenance is urgently required, and maintenance institutions need to be organized, equipped, and trained. In the railway sector, repair of infrastructure is being completed on most of the network, but the lack of cooperation between the three companies has prevented trains from running on more than small segments of the network. Roads 2. Road traffic increased significantly throughout Bosnia and Herzegovina in early 1996 due to increased security and demand for goods transport (road transport is still the main means of delivering goods). The resurgence of trade is not likely to be strong enough in the medium term to allow traffic to reach its prewar levels, however. The two major consequences are that: (i) it will be difficult to deal with the existing road maintenance backlog (estimated at US$450 million in 1996), because revenues based on road user taxation will remain low; and (ii) there will likely be significant retrenchment among road contractors since future investment and maintenance programs will not be sufficient to support the existing private and public capacity for road maintenance and construction. 3. Both the Federation and Republika Srpska began to dedicate resources to road maintenance and rehabilitation in 1996. The programs matched the sector priorities: the urgent repair of damaged roads and the removal of bottlenecks due to landslides, destroyed bridges and blocked tunnels. Funds have come from local sources (levies on fuel, car registration and custom taxes) and the donor community (principally in the Federation), as well as investments undertaken by international troops. 4. In Republika Srpska, the lack of external resources has restricted all but the most urgent repairs and maintenance, with estimated expenditures of US$10 million in 1996. Works have been performed by the 10 remaining pre-war Putevis (state-owned road agencies). The 1996 program was financed with funds from road user taxes, which totaled US$13 million. In addition the Republika Srpska budget provided an estimated US$100,000 for wages. The Republika Srpska Ministry of Transport and Telecommunications (RSMTC) plans to allocate US$15 million for maintenance in 1997 plus several million dollars for repairs and rehabilitation, financed by expected road tax revenues of US$21.4 million. It is unlikely that Republika Srpska will be able to increase its commitment to road maintenance by the targeted 50 percent needed (para. 9-11) given likely government revenues. The Entity counts on strong commitments from donors, which might provide up to US$46 million, at most half of which would be disbursed in 1997. 23 24 Bosnia and Herzegovina: Public Expenditure Review 5. On the Federation side, a donor-funded Emergency Transport Reconstruction Programn in 1996 focused on the reconstruction of fourteen bridges and the rehabilitation of 50 kilometers (km) of roads plus equipment and technical assistance (World Bank funding of US$20 million disbursed, in addition to other donors' funding). Investments were also undertaken in rehabilitation operations by the international military forces (local contracting estimated at US$7 million); and routine maintenance and winter operations (US$10 million financed from the Federation budget, automobile registration fees and direct contracting by international military). With respect to maintenance expenditures, US$6.6 million remained at the Federation level to be allocated to main roads, and US$3.3 million were distributed among the ten Cantons to operate secondary roads. Operating costs were covered through: (i) the Emergency Transport Reconstruction Program (ETRP), which provided US$0.9 million to assist the PIU; and (ii) the Federation budget which provided the FMTC with an additional US$0.6 million for wages, facilities, cars and telephone expenses. 6. The Federation 1997 program is oriented to the maintenance of the main road network. The 1997 budget will transfer US$13.3 million to the FMTC to cover road maintenance expenditures, US$0.6 million for wages, and US$0.2 million for other operating costs. Beyond those transfers, the FMTC will rely on donor community support for 1997. There may be an additional donor commitment of US$40 million in 1997, of which half would be disbursed in 1997, and the remainder subsequently (Table 5.1). Table 5.1: Road Sector Funding (in US$ millions) Funding Sources 1996 1997 Fed RS Total Fed RS Total Maintenance financed by taxes (+SFOR) 10.0 10.0 20.0 13.3 15.0 28.3 Operating costs (wages, facilities, others) 1.5 0.5 2.0 1.5 0.5 2.0 1996 ETRP (First) 20.0 0.0 20.0 - - - Transfer from 1996 ETRP - - - 20.0* 3.0 23.0 1997 ETRP (Second) - - - 20.0 23.0 43.0 Total 31.5 10.5 42.0 54.8 41.5 96.3 Source: Ministries of Transport. *US$20 million represents disbursement from the US$40 million transferred in 1997 to the Federation from the 1996 ETRP. 7. Road sector revenues are described in Table 5.2. In the Federation, fuel and registration taxes are low. Fuel taxes (which were, on average, 35 percent of the price of fuel before the war), do not exceed 10 percent of the final price of any fuel. In addition, the reintroduction of fuel taxes in 1996 was slow.' As a consequence, the FMTC received only US$6.6 million from fuel tax revenues (August 1996). In addition to the levy on fuel the FMTC received an estimated US$2 million from car registration. If the fuel tax were restored to its pre-war level (i.e., 35 percent of the fuel cost), given the 1996 traffic estimates, yearly collections could rise to as much as US$70 million in the Federation. i From June 1996 to late 1996, Federation ZPPs were not authorized to collect fuel taxes from petrol stations, depressing fuel tax revenues to below levels in Republika Srpska. Annex Five. Sector by Sector Review of Infrastructure Policy 25 8. In Republika Srpska, tax collection for the roads sector seems to be more effective. The RSMTC received US$21.3 million in 1996, directly from the payments bureau (US$9.3 million from the collection of the fuel tax, representing 50 percent of the total collection; US$9 million from border fees; US$2.2 million from vehicle registration). Those inputs constituted the only budget for roads in Republika Srpska in 1996. Table 5.2: Road Sector Revenues Federation, US$ million Republika Srpska, US$ million 1996 1997 1996 1997 Revenues from road users 15.0* 25.0* 31.5 35.0* Transfer to the Ministries 8.6 15.3 21.3 23.0* * Estimate 9. Locally-generated resources are currently responsible for less than half of total spending on the road network, with the remainder being financed by donors. Engineering estimates indicate, however, that minimum local contributions will need to double to avoid deterioration of the existing network. That is, in addition to the remainder of the Transport Reconstruction Program financed by donors, the governments of Bosnia and Herzegovina will need to allocate about US$50 million annually to road maintenance activities over the next four years (US$30 million a year in the Federation and US$20 million a year in Republika Srpska). Even this level of expenditure is low, and scarcely enough to prevent further deterioration of the road system. Because crucial maintenance and construction activities will be postponed under this plan, beginning in the year 2000, over US$110 million should be allocated yearly. These projections do not include investments to expand the road system. 10. Even this level of financing would not be sufficient to completely achieve two objectives: (i) remove the existing backlog estimated at about US$450 million (including lack of maintenance and hostilities-related degradation, but excluding the costs of improving low quality road surfaces); and (ii) undertake the maintenance of the network to avoid further degradation, estimated to cost about US$92 million yearly.2 11. Eliminating the backlog has already started under the Emergency Transport Reconstruction Project, which contributed US$33 million to road sector rehabilitation. Further contributions from the international community are expected under the three- to four-year transport reconstruction program for a total amount of about US$220 million, principally to 2These estimates of backlog and required expenditure are based on work done in 1996, within the Federation of Bosnia and Herzegovina by IMG and Start Up Services in collaboration with the FPID, using documents on the prewar road network as described in the 1990 Third Highway Sector Project for Yugoslavia, and assuming road maintenance standards applicable in other European countries. In addition, the estimates are based on three figures: (i) the value of the backlog in 1986: a study undertaken for the First Highway Sector Project had identified the existence of a US$40 million backlog in 1986; (ii) the value of investments in 1986: according to the Third Highway Sector Project documents, the total inputs (including all road maintenance activities and investments) allocated to the road sector in 1986 amounted to US$90 million; and (iii) the depreciation of road assets: following road maintenance and activities standards developed in other European countries and given the structure of the road network in Bosnia and Herzegovina, the depreciation has been estimated, in the absence of new investments since 1986, at about US$92 million per year. 26 Bosnia and Herzegovina: Public Expenditure Review address the backlog on the main road network. This will still leave some US$240 million in road maintenance "backlog" after the year 2000. In addition, US$92 million would be needed annually to maintain the existing network at international standards. Moreover, if less than US$42 million per year is spent over the next three years on road maintenance, the backlog would continue to grow, leading to collapse of the road network. The expenditure recommendation of US$50 million per year until the year 2000 is, therefore, only slightly above the amount needed to protect the existing network, and significantly less than what is needed to bring the road network to average European standards. 12. The pressing revenue needs of the road sector, combined with the inevitable gradual reduction in donor financing across the entire scope of public sector activities, underlines the importance of increasing the efficiency with which local financing is collected. Cost recovery principles need wider application in the future in order to impose most of the burden for payment on road users through user charges, at least for maintenance expenditures. The governments should investigate the possibility of gradually shifting the road sector to a fee-for-service basis. The fees may consist, inter alia, of: (i) border fees/short-term permits charged on all foreign trucks entering BiH; (ii) license fees on motor vehicles and a supplementary license fee collected from heavy vehicles to ensure they pay in full for the damage they do to the road pavement; (iii) fines on overloaded vehicles; and (iv) a road maintenance levy added to the price of all transport fuels. The aim is to ensure that sufficient funds are available for maintenance of all roads. General budget funds would be reserved for upgrading, improvement, and new construction. Railroads 13. The railroad sector is running large deficits because of extremely low operating levels. This is not due to damage from the war or other capacity problems, but rather to continuing political disputes between the three operating organizations over inter-Entity transit. 14. In 1985, railroads in Bosnia and Herzegovina transported 19 million passengers and 32 million tons of goods, employing 23,000 people. Currently, trains run only occasionally and only on an intra-Entity basis, due to lack of agreement on interconnection arrangements. The three existing operators (ZBH, the Sarajevo-based company of the former Republic of Bosnia and Herzegovina; ZHB, the Mostar-based company; and ZRS, the Republika Srpska company) manage some passenger traffic and organize freight transport, but most demand remains unsatisfied, including that of potential private sector and military clients. Each operator tries to pay its numerous staff with its slight operating revenues and government subsidies. This puts a significant burden on government fiscal accounts. 15. In the Federation, the international military and donor community were very active from early 1996, as with the roads sector. The NATO-led Stabilization Force (SFOR) tried to align its military programs with those of the government and the donor community, including rehabilitation of the Volinja Doboj (US$3.5 million) and Doboj-Zvornik lines (US$0.7 million). The donor community focused on urgent repairs to existing lines that meet with minimal traffic requirements (simplified safety and signaling) and disbursed US$15 million in 1996 though grant money on soft loans, and an estimated US$21 million through equipment donation (mainly locomotives to the Federation). Annex Five: Sector by Sector Review of Infrastructure Policy 27 16. There has been no significant investment in Republika Srpska. However, the 1997 railroad investment program in Republika Srpska is ambitious: US$14.5 million for tracks (US$5.6 million for maintenance and US$8.9 million for rehabilitation), and US$14.5 million for electricity (US$4 million for maintenance and US$10.5 million for rehabilitation). The Republika Srpska Ministry of Transport and Communications (RSMTC) will rely on donor support to carry out such a program. 17. All three companies receive direct subsidies. In 1996, ZBH received DM 7.12 million from the Federation budget (but expected DM 53.7 million).3 ZHB received DM 740,000 from Croat-majority area cantons, and ZRS obtained DM 2.54 million from the Republika Srpska budget. 18. Even taking into account direct government subsidies, only ZHB, the Mostar-based company and the smallest of the three, seems to operate with a balanced budget. Subsidies from the cantons represent 70 percent of total revenues, although revenues from operations increased significantly in 1996 (DM 180,000 from transport of domestic goods, and DM 150,000 from the transport of goods through Croatia). In 1996, ZHB had 100 employees, with 400 additional unpaid staff on a waiting list. 19. In contrast, the financial performance of ZBH, the Sarajevo-based company, is poor. Losses for 1996 were evaluated at DM 56.6 million, including DM 49.3 million for depreciation (DM 34.5 million for the depreciation of infrastructure and DM 14.8 million for rolling stock and equipment). This represents 79 percent of the total budget (DM 71.5 million).4 Even on a cash basis, ZBH ran a deficit of DM 7.31 million. The deficit was financed through arrears, especially to workers. The 1996 revenues and operating incomes before depreciation are given in Table 5.3. 20. In 1996, subsidies from the Federation covered only 50 percent of operating losses (net of subsidies). Wages and charges (DM 12.3 million) represent 54 percent of the expenditures (without depreciation), much more than operating levels justify. Gross wages of DM 12 million were recorded. Net wages after taxes were DM 6.39 million for 3,000 staff and 2,000 on the waiting list. Estimated salaries for regular staff were DM 140/month in 1996; 1000 individuals on the waiting list were paid DM 80 per month; and an additional 1000 were paid DM 5 per month. 3These figures, which are part of the report on ZBH 1996 financial activities, were not found in the 1996 Federation budget. 4It is not clear whether the estimated DM 49.5 million depreciation represents war damages, and if it is related to all Bosnia and Herzegovina territory or only to the area ZBH managed in 1996. 28 Bosnia and Herzegovina. Public Expenditure Review Table 5.3: ZBH Financial Activities for 1996 (in DM million) Income Statement Infrastructure Transport Total Operating revenues Revenues on sold items 0.06 7.70 7376 Subsidies 2.72 4.40 7.12 Total operating revenues 2.78 12.10 14.89 Operating expenses Wages and charges 2.62 9.72 12.35 Material, energy* 0.36 2.41 2.77 Financing expenditures 0.01 0.81 0.82 Other working expenses** 0.96 5.31 6.27 Total operating expenses 3.95. 18.25 22.20 Expenditures less Revenues 1.17 6.15 7.31 e e.g. locomotive rehabilitation, rolling stock maintenance * e.g. recurrent costs, training. 21. The situation has not improved in 1997. In the first quarter of 1997, ZBH spent approximately DM 6 million marks for wages, liquid fuel, and debt repayment. Operations generated approximately DM 1.1 million in the first quarter of which 80 percent is noncash (i.e., inter-company receivables). The last salary payment to employees was in December, and was apparently financed with loans from state banks such as Central Banks, Commercial Bank Tuzia, Privredna Banka Sarajevo. 22. Assuming the same expenses for the next three quarters, a total of DM 24 million would be needed in 1997 to keep the company operating on a cash basis. Altogether, ZBH is expecting, optimistically, DM 6 million of revenue from current operations. The company will therefore accrue an additional DM 18 million in arrears by the end of 1997. In addition, the company has carried over approximately DM 10 million in accounts payable for loans, salaries and other supplies. The total arrears of the company at the end of 1997, including the DM 10 million, are, therefore, expected to be DM 28 million. There is no subsidy for ZBH planned in the Federation 1997 budget. However, the subsidy made in 1996 did not appear in the budget, either. In general, it is likely that pressures for subsidization will grow, as will arrears to suppliers and workers. 23. ZBH expects support from the donor community to rehabilitate the infrastructure over Bosnia and Herzegovina, although such support is unlikely to be forthcoming without better integration of the three networks. Some DM 188.5 million in investments over five years is estimated by ZBH to be needed reach prewar levels of railroad service (Table 5.4). 24. The situation of ZRS, the Republika Srpska company, is also financially unsound. Wage arrears, for example, as in other parts of the public sector in Republika Srpska, are likely to be significant. ZRS operated few trains for goods and passengers during 1996, but those which operated, did so on a regular basis, on the east-west Bosanki Novi - Banja Luka-Doboj line. The operating revenues have been slight (an estimated DM 2 million). These funds, coupled with a Annex Five: Sector by Sector Review of Infrastructure Policy 29 DM 2.54 million subsidy from Republika Srpska, were set aside to (partially) cover wages for 3,000 regular staff and 2,000 on the waiting list, and to purchase liquid fuel to operate the trains. Table 5.4: Investment Required to Reach Pre-War Railroad Infrastructure (in DM millions) Railroad line Track Electrification Total Sarajevo - Capljina 34.2 20.3 54.5 Sarajevo - Zenica 32.8 8.0 40.8 Zenica - Doboj 41.2 8.0 49.2 Doboj - Samac 1.1 8.0 9.1 Breza- Vares 0.6 - 0.6 Doboj - Tuzla 0.9 2.9 3.8 Tuzla- Brcko 4.4 1.5 5.9 Tuzla - Zvornik 1.9 0.6 2.5 Doboj - Bosanski Novi 4.9 4.5 9.4 Bosanski Novi Knin 2.8 9.9 12.7 Total 124.8 63.7 188.5 25. Although the number of employees is about one third of prewar levels, overstaffing remains significant for ZBH and ZRS. At most, an estimated 1,000 staff could be employed in labor-intensive maintenance of the network, although this work would be better performed by contracting with local firms. Traffic would have to rise to on the order of 600 million passenger- km and 2,000 ton-km of goods (half of the prewar average level) to justify the current size of the compames. 26. Management must therefore be allowed, and obliged, to follow sound business practices. A substantial amount of work will be required to change current management practices, to address the overstaffing issue, and to design specific programs to make up for the departure of many qualified staff during the war. 27. However, significant steps must also be taken to boost revenues and operating levels. None of the three companies can operate soundly as long as there is no agreement between the Entities, and no inter-Entity traffic. The current situation virtually guarantees financial disaster. Beyond the necessary agreement to operate trains, fees and tariffs have to be evaluated and levied. The Dayton-mandated Commission on Public Corporations (set up under Annex 9 of the Dayton Peace Agreement) tried to determine the fees the three companies would have to pay one another for using the infrastructure, but there was no agreement on rates. Agreement in the near- term is essential. As a first step, the authorities and managers of the companies should establish detailed target deadlines for improving billing and collection rates, eliminating subsidies and increasing tariff levels. A reasonable agenda would be to: (i) cover all cash operating costs by 1998; (ii) make cash flows positive by 1999; and (iii) fully recover costs within the next five years. 28. More efficient management must also address the maintenance backlog, so that the current efforts to rehabilitate the network and rolling stock will not be jeopardized. This means that the companies will have to allocate sufficient expenditures to maintain the existing network, 30 Bosnia and Herzegovina: Public Expenditure Review even if the resumption of passenger and freight traffic is very low. The required amount varies with the kind of service that will be provided. The first step the companies agreed upon consists of a transitional period during which trains would run at low speed (50 km per hour). During that period, only an estimated DM 20 million would be needed for emergency works per year. The second step would be to establish a tariff policy and improve maintenance so that trains can run faster and more safely. Customers, when charged, would then participate in the maintenance of the infrastructure and rolling stock. Annual maintenance under that second stage would need additional funding (a first estimate suggests DM 35 million per year). 29. If company costs were adjusted to operating levels, the fiscal burden imposed by the railroads would be trivial. However, the current system is accruing liabilities that will rise -to a minimum of DM 28 million in the Federation alone by the end of 1997. All governments have limited their subsidies to the railroads, but none have decreed that railroad debts are not the responsibility of, nor in any way guaranteed by, the governments. Government subsidies to the railroads should be no more than the DM 20 million needed to rehabilitate the networks to permit speeds of 50 km per hour. There should be no discussion of subsidies, however, until the workforce is rationalized, waiting lists are abolished, and inter-Entity and inter-company agreement on operations and tariffs is finalized. Telecommunications 30. The telecommunications sector combines post and telephone services. As in the case of railways, the institutional structure in the telecommunications sector is ad hoc. While there is one internationally-recognized telecommunications operator -- the Sarajevo-based PTT of the former Republic of Bosnia and Herzegovina -- there are in practice three service providers which operate the network in the territory of Bosnia and Herzegovina: the Sarajevo-based (PTTBH); the Mostar-based (PTTHB), and the Republika Srpska (PTTRS) operators. Lack of cooperation among the three service providers and the unresolved structural and policy framework of the sector is the single most important stumbling block to continued speedy reconstruction. 31. In all organizations phone services yield 95 percent of revenues, of which some 75 percent are from international calls. Donor resources had an important impact during 1996, especially on the domestic phone network. Some DM 55 million were committed by donors during the year, which brought resumption of quasi-normal use of telephones for a large number of households, as well as for all public and private firms (e.g., there were 236,000 lines in Republika Srpska in 1996, 181,000 of which functioning -- matching the prewar level.). The European Bank for Reconstruction and Development (EBRD) was the lead agency in mobilizing and coordinating these resources. Grants and credits were distributed among the three operators per donor preference, with most funding going (DM 26.8 million) to PTTBH, and significantly less to PTTHB (DM 0.6 million) and PTTRS (DM 0.95 million). Some DM 28 million in remaining funds will be implemented through 1997-1998 programs. 32. Despite the unclear legislative, regulatory, and political situation all three companies embarked on urgent reconstruction works, with PTTHB and PTTRS employing largely own resources. Work aimed at resuming domestic calls capability. Very little effort was aimed at Annex Five: Sector by Sector Review of Infrastructure Policy 31 completing direct communication links between the Entities, making telecommunications impractical between the different networks. PTTRS dedicated a significant part of revenues to rehabilitation in 1995 and 1996.. The final report of PTTRS for 1996, still to be submitted to the government, mentions investments in construction and rehabilitation of the network capacity amounting to DM 21 million (for offices, local cables, transmission system, and computer and power equipment). There is no review of the completion of works and the disbursement status in Republika Srpska, or of works undertaken in the Federation. 33. The three companies expect to finance 1997 rehabilitation programs from their own resources and the donor community, from which they expect half of total rehabilitation funds. As an example, the priorities in the 1997 program of PTTRS (total amount DM 36.4 million) are the main distribution system (DM 9.7 million), an international switchboard in Bijelina (DM 6.6 million; the only international switchboard in use is located in Belgrade, although PTTRS does not pay for the service provided), local optical (DM 2.4 million), the local network (DM 12.1 million), and other capacities (DM 5.6 million for transit exchange in Pale and Banja Luka and others). However, there is no provisions for funding to restoring inter-Entity connections. 34. Each company makes a profit from the services they provide to customers and from the deployment of GSM (Groupe Special Mobile, i.e., mobile phone) networks through contracts with private operators. PTTBH and PTTHB have already developed regional systems (in the context of an illegal framework and differing views on a future national consensus). PTTBH is deploying its network in Sarajevo, Tuzla and East Mostar, whereas PTTHB developed a separate system within most of the areas it serves. PTTRS is in the process of selecting a partner for the provision of similar services (Republika Srpska would own 10 percent of the operator firm to be created), and plans to charge its partner DM 30 million in exchange for the right to operate GSM technology in Republika Srpska. 35. Unlike other infrastructure sectors, there is no transfer from government budgets to the telephone companies, even for the postal services (which are paid for through subsidies from telecommunications revenues). Telephone revenues are significant because of the number of international calls, willingness to pay for services, and the willingness of the operators to disconnect the lines of bad customers. 36. Since there are no comprehensive annual reports, financial information for the three service providers is difficult to accurately assess. The maximum potential revenues billed can be estimated, however, on the basis of company information on the number of impulses ("telephone units") registered and the cost of an impulse (e.g., DM 0.04 in the Federation). On this basis, the estimate of total maximum billed telecommunications revenues in 1996 for the three service providers totals some DM 170.5 million (Table 5.5). 32 Bosnia and Herzegovina: Public Expenditure Review Table 5.5: Estimated Maximum Billed Revenues (in DM millions) PTTBH PTTHB PTTRS Total BH. DM million DM million DM million DM million Postal services 2.4* 0.8* 1.3 4.5 Domestic calls 27.0 8.0* 14.0 49.0 International calls 55.0 18.0* 44.0 117.0 Total 84.4 26.8* 59.3 170.5 * Estimated figure 37. However, these figures do not take into consideration lost revenues from non-paying customers. Households generally pay their bills because of the threat that their lines will be disconnected if they fail to pay. This is not the case for the public institutions and private firms whose activity is considered indispensable for reconstruction of the economy within each Entity and whose lines are not disconnected for non-payment. 38. PTTRS estimates that 70 percent of its major customers pay their bills. This estimate affects domestic calls and one third of the international calls (those from Republika Srpska abroad).5 As PTTRS provides 80 percent of its services to major customers, the estimate of Republika Srpska's 1996 telephone revenues must be reduced as follows: PTTRS domestic calls reduction: (DM14 million) x 80% x 30%=DM3.4 million PTTRS international calls reduction: (DM 44 million) x 80% x 30% x 1/3 = DM3.5 million 39. The same approach is applied to PTTBH and PTTHB, although the estimate of delinquent customers among major clientele is some 40 percent: PTTBH domestic calls reduction: (DM 27 million) x 80% x 40% = DM 8.6 million PTTBH international calls reduction: (DM 55 million) x 80% x 40% x 1/3 = DM 5.9 million PTTHB domestic calls reduction: (DM 8 million) x 80% x 40% =DM 2.6 million PTTHB international calls reduction: (DM 18 million) x 80% x 40% x 1/3 = DM 1.9 million 40. Thus an estimate of actual collected revenues for 1996 on the basis of the above calculations is closer to some DM 144.6 million (Table 5.6). The table also demonstrates that the telecom sector provides indirect subsidies totaling some DM 26 million to major public and private firms, in the form of telephone services that are billed, but not collected. Calls from foreign countries to Republika Srpska are paid through annual readjustments with foreign telephone companies, as agreed between PTTRS and those companies Annex Five: Sector by Sector Review of Infrastructure Policy 33 Table 5.6: Telecom Revenues in 1996 (in DM millions) Service PTTBH PTTHB PTTRS Total BH Service Revenue Service Revenue Service Revenue Service Revenue provided collected provided collected provided collected provided collected Postal services 2.4 * 2.4 0.8 * 0.8 1.3 1.3 4.5 4.5 Domestic calls 27.0 18.4 8.0 * 5.4 14.0 10.6 49 34.4 International calls 55.0 49.1 18.0 * 16.1 44.0 40.5 117 105.7 Total 84.4 69.9 26.8 * 22.3 59.3 52.4 170.5 144.6 Note: Se;-vice provided Revenues collected are lower that the Service provided, because it take into consideration the collection ratio. * Estimate. 41. Telecommunications expenditures are also difficult to accurately assess. There is little evidence for investments and maintenance matching material costs, and depreciation. Special expenditures, which are not defined or detailed, represent 25 percent of the total expenditures. As an example, Table 5.7 summarizes PTTRS expenditure plans in 1995 and 1996. Table 5.7: PTTRS Expenditure Plans for 1995 and 1996 (in DM millions) (DM million) Executed 1995 Planned 1996 Executed 1996 Announced revenues 30.0 59.3 Realistic revenues 25.5 52.4 Expenditures Material costs 5.1 12% 10.6 20% 14.8 20% Non-material costs 5.0 12% 6.8 13% 7.9 11% Salaries 2.3 6% 6.6 12% 5.0 7% Financial expenditures 6.4 15% 4.7 9% 0.7 1% Special expenditures 8.8 22% 2.8 5% 18.2 25% Expenditures without depreciation 27.6 - 31.5 - 46.6 - Depreciation 13.8 33% 21.3 41% 25.2 36% Total expenditures 41.4 100% 52.8 100% 71.8 100% 42. There is reason to believe that telecommunications revenues are substantially greater than the costs, including investments and depreciation, of the telecom companies. These high revenues are due, in large part, to the companies pricing policies, which charge high rates to international customers (whose demand is very inelastic). As argued in the text, a significant part of tariff policy reform should be to have telecom surpluses transferred to the government and placed in the budget. These surpluses are likely to be as much as DM 100 million per year. This could be accomplished by lowering telecom tariffs and taxing telephone usage. 43. The generally healthy financial situation of the sector allows the telecom companies to pay a large number of employees and waiting list staff. For example, PTTRS employs 1,600 people for telephone services, 1,400 people for postal services (although postal services are in deficit), and 700 people on a waiting list, (who are paid DM 50 per month). A similar situation 34 Bosnia and Herzegovina: Public Expenditure Review exists in the two other companies. Because of the stable salaries, work in the sector is considered one of the most desirable in Bosnia and Herzegovina. 44. PTTRS charges international rates that are, on average, 10 percent higher than those in Belgrade (and thus generally consistent with international standards). PTTHB also places a small "mark-up" on Zagreb-based tariffs. On the other hand, the international rates charged by PTTBH are highly overpriced and need to be modified. The tariffs were accepted in 1995 and 1996, due to the difficulty of network operation, and as an easy way of restarting PTTBH activities. Now, however, the corporation of international operators has requested that the government of the Federation order new tariffs. The changes would consist of a broad reduction in international rates, and an increase in domestic rates (at present households are charged DM 1 for the first 100 impulses on units, that they use in a month). The Federation Ministry has drafted a decree to reform tariffs along the lines described below. 45. The rates were initially intended to use the international community to revive telephone services in the Sarajevo area, and allowed PTTBH to complete necessary rehabilitation works without asking households to contribute. This situation, although it has been tolerated since late 1995, has to be addressed. This is why the Federation Ministry embarked on: (i) lowering international rates; and (ii) asking households to pay more realistic prices for domestic services they receive. The overall strategy is to maintain revenues at their current level. Proposed changes in PTTBH pricing policy are compared with Croatia tariffs in (Table 5.8). Samples given for international calls are based on the ten major foreign countries the Federation trades with. The estimate of annual revenues, is based on 1997 recorded impulses. Table 5.8: PTTBH International and Domestic Revenues international Croatia Current PTTBH Proposed Revenues Calls From: DM per min DM per min DM per min DM million To: Current Future Croatia 1.00 0.50 13.30 6.65 Germany 1.48 2.80 1.50 10.50 5.62 USA 2.66 5.00 1.50 6.30 1.89 England 2.24 2.80 1.50 3.20 1.71 Slovenia 1.48 2.80 1.50 2.90 1.55 Czech Republic 2.66 2.80 2.40 2.40 2.05 France 2.66 2.80 1.90 2.30 1.56 Italy 1.48 2.80 1.50 2.20 1.17 Turkey 1.78 2.80 2.40 1.30 1.11 Switzerland 2.24 2.80 1.50 1.10 0.58 Total International calls 54.63 32.35 Domestic calls Subscriptions and connections 5.61 15.72 Calls (current rate 0.04 / future rate 0.053 DM/impulse) 22.25 30.04 Total Domestic calls 27.86 45.76 GRAND TOTAL 82.49 78.11 Annex Five: Sector by Sector Review of Infrastructure Policy 35 46. Telephone pricing policy in Bosnia and Herzegovina, although higher than in neighboring countries, is approaching international standards. The telecom sector will need two to three years to reach a prewar level of equipment and management. Once network rehabilitation is complete (likely in late 1997), the companies should be able to run the sector with one third of their revenues (i.e., DM 50 million out of the DM 144 million for total revenues). This suggests that privatization, and replacing some portion of telecom tariffs with taxes, are feasible in the near term. Cooperation between the two Entities to regularize structural and policy frameworks in the sector is the single most important step to ensure this occurs. Water 47. Before the war water and waste management services in Bosnia and Herzegovina were good, both in coverage and quality. In urban areas piped water served 94 percent of the inhabitants and sewerage 72 percent, water production capacity kept pace with demand, and twenty four hour service was the norm. In rural areas 35 percent of the population was connected to piped water supply systems and 10 percent to sewerage systems, a rate comparable to most industrial countries. Most of the networks of the Federation were severely damaged during the war, and suffer from a extreme lack of maintenance. In Republika Srpska war damage seems relatively minor, but the maintenance burden rules out normal delivery. The Entity Ministries managed their own water resources and distribution capacities in place of the previous Ministry of Agriculture, Water Management and Forestry (which had 350 staff compared to the current handful in Republika Srpska, 15 in Mostar and 35 in Sarajevo). Most qualified personnel left the country during the hostilities, and currently only four students are specializing in water management at the University of Sarajevo. 48. There was no budget line for the sector in 1996 in either Entity budget. Sector rehabilitation relied entirely on donor commitments, which peaked at US$96 million, concentrating mainly on emergency works in Federation municipalities (currently no disbursement ratio is available). Rehabilitation was carried out according to priority lists of municipal infrastructure projects. Funds were allocated as follows: (i) 80 percent to water supply; (ii) 8 percent to institutional strengthening; (iii) 7 percent to waste management; and (iv) 5 percent to flood control. 49. Thus, for example, the Sarajevo water utility has no revenue and its budget relies exclusively on small subsidies from the municipality. Salary payments run months behind, and there is no maintenance fund (maintenance needs are estimated at DM 8 million per year for Sarajevo). The Municipality is supposed to pay for the electricity to opc-ate the pumps. This situation is common in Bosnia and Herzegovina, and with few exceptions (e.g., Mostar, Zenica), bill collection barely functions. 50. Although there is no apparent reason for delaying development of bill collection, even the company's managers think it would be difficult to collect money for a service like water, because of the difficulty in assessing consumption and the low income of the majority of households. In addition, water is important to the entire population, and therefore managers are convinced it would be almost impossible to disconnect the pipes of bad customers. As long as this situation 36 Bosnia and Herzegovina: Public Expenditure Review persists, not only will privatization remain impossible, but efforts to prevent a dramatic decline in the quality and reliability of water services will fail. 51. There are clear opportunities in the water sector. Donor comnmitments should be tied to maintenance, pricing, and collection of bills. Each municipality approached by donors for receiving funds for the rehabilitation of their networks should work out cost recovery principles. These principles would impose part of the cost on users, rather than on the general public through the tax system, and encourage more rational use of water and sewage services. Energy 52. Under normal circumstances, energy providers would cover all their costs from sales revenues and comnmercial borrowings, and pay taxes to government as well as dividends to the government bodies which hold ownership shares. Their effects on the fiscal situation of government would be strictly positive. In Bosnia and Herzegovina, however, the energy sector is a substantial fiscal drain, and is creating large unfunded liabilities for the different governments. In part this is due to the limits imposed on cost recovery by the temporarily low incomes of households, which account for more than 50 percent of sales by these enterprises, and to unusually large investments that are required to rehabilitate the facilities that suffered damages and lack of maintenance during the war. In large part, however, the fiscal drain is due to inadequate attempts at cost recovery and inefficient operations. 53. Detailed information has been obtained from the Federation power company, Electropriveda of Bosnia and Herzegovina (EPBiH), from SarajevoGas, and Toplane (the district heating company for Sarajevo). The financial and economic situation of Electropriveda Mostar (EPM) and Electropriveda Republika Srpska (EPRS) could not be analyzed to a similar extent, due to a lack of data availability and unresolved liability allocations between the three power enterprises. All financial data shown in this review are based on unaudited financial statements of the enterprises and therefore have to be treated with caution. 54. The war reduced the sector's activity drastically. More than 50 percent of the electric power system was destroyed; gas supply to BiH permitted only individual household consumption in Sarajevo; and the operation of the Sarajevo district heating system was reduced to less than 30 percent of its original capacity. Billing and collection was terminated for both gas and district heat in 1991. Consumption of electric energy continued to be billed throughout the war, but consumers' ability to pay deteriorated substantially, and collections plummeted. While part of the significant ensuing losses was covered by subsidies throughout the war, most losses accumulated as arrears to other utilities, which may partially offset each other; and as arrears to fuel suppliers, namely coal mines and Gazexport. In addition to arrears accumulated during the war, the enterprises had liabilities in the form of prewar long term credits, mainly with the World Bank, that have not been serviced since 1991. 55. As a result, the financial situation of the energy utilities has developed through 1995- 1996 as described in Tables 5.9 and 5.10. In 1996, SarajevoGas required DM 54 million in subsidies to avoid arrears. Toplane was also unable to collect sufficient revenues to offset costs, Annex Five. Sector by >Sector Review o1 JInfraiaructure Policy 37 but accumulated DM 7 million in arrears since subsidies covered only one-third of its operating deficit. While EPBiH maintained an operating surplus, it has more than DM 650 million in arrears with external creditors and internal suppliers; neither debt service nor depreciation, both of which are significant, were provided for in 1995 and 1996. They are, therefore, not included in the figures in Table 5.9. Table 5.9: Income Statements of Three Energy Utilities Income in 1995 Cash Expenses Collected Cash Direct Increase Increase in and 1996, (no depreciation Revenues Surplus/ Subsidies in Receivables from (DM million) or debt service) Deficit Payables Consumers EPBiH 263.0 301.0 38.0 0.0 0.0 15.8 SarajevoGas 71.0 17.0 -54.0 54.0 0.0 0.0 Toplane 11.0 0.5 -10.5 3.5 7.0 0.0 Table 5.10: Liabilities of Three Energy Utilities Liabilities End-1996, Arrears Inter - Arrears Fuel Debt by Creditor (DM million) Utility Suppliers EPBiH 0.0 300.0 356.0 SarajevoGas 0.0 0.0 65.0 Toplane- Sarajevo 15.8 0.0 0.0 56. The fact that only marginal investments in equipment and maintenance were made in the period 1991-1996 created additional, off-balance sheet liabilities. The backlog in investments and uncompensated deterioration of equipment contribute substantially to the overall reconstruction requirements of the war-damaged energy supply systems. Therefore, the Bank's reconstruction projects for electric power, district heat and the gas system compensate to an unquantifiable extent for the shortfall in maintenance and lack of investments during the war. 57. The financial situation of power utilities in the Croat-majority area of the Federation and in Republika Srpska is less clear, for two reasons. The first is that EPM and EPRS rely more heavily than EPBiH on hydropower (100 percent and 66 percent, respectively) making comparisons somewhat more difficult. The low operating costs of hydropower generation allowed EPM and EPRS to avoid the large arrears that EPBiH, for example, built up with coal suppliers.6 The second factor complicating finances at EPM and EPRS is the allocation of prewar debt between the three power enterprises. Since neither EPM nor EPRS have submitted 6 Prior to the war, 92 percent of the total energy consumed in Bosnia and Herzegovina was provided by the country's 13 hydropower and 12 thermal power plants p#oducing 17,600 GWh per year. Thirty coal mines supplied 18 million tons of coal per year to the thermal power plants. The natural gas system supplied 8 percent of the total energy consumed in BiH. Six hundred ten million cubic meters per year of natural gas were imported from Gazexport, Russia to supply the aluminum factory in Zvomik, the steel plant in Zenica, the cement plant in Kakanj and residential and commercial consumers in Sarajevo. Finally, 10 percent of the total population in BiH received heat from the country's mainly gas-fired district heating systems. In Sarajevo, the country's largest district heating system supplied 45 percent of the inhabitants with heat, using a 500 MW gas-fired boiler. 38 Bosnia and Herzegovina. Public Expenditure Review detailed information, no conclusion can be drawn at this stage, but will be subject to discussion in the context of the 1997 appraisal of the Bank's second electric power reconstruction project. 58. With the ongoing reconstruction of the sector, the energy utilities are resuming regular service for households and public consumers, although the recovery of industrial and commercial consumers still lags. Households are contributing twice as much to energy company revenues in the Federation as commercial and industrial customers. Billing and collection were resumed for gas and district heat in December 1996 after an interruption of five years, with collection rates in January 1997 reaching 60 percent for SarajevoGas and 35 percent for Toplane-Sarajevo. EPBiH had continued to bill its consumers when possible, during the war; and had achieved a collection rates of 75 percent by January 1997. The collection rates illustrate that revenues collected from consumers can at present cover only a fraction of the operating costs of the sector. With the resumption of billing and collections, energy utilities have three sources of revenues: collections from unsubsidized consumers; collections from subsidized individual and public consumers; and * direct subsidies to the enterprises. 59. The indirect subsidies contained in the collections for gas and district heat can be broken down into three categories, at least for Sarajevo canton: * The monthly canton subsidy to every household for gas or district heat consumption is DM 44.5 per household. * The maximum canton subsidy to demobilized soldiers through the passbook mechanism (explained below) is DM 50 per month per demobilized soldier for all household energy bills. . There is also a small canton subsidy to a very limited number of severely disadvantaged households, covering all utility costs. Table 5.11: Payment Sources for Household Collections, January 1997 Energy company Unsubsidized portion, 44.5 DM/month, % Passbooks, % EPBiH 90 0 10 SarajevoGas 20 70 10 Toplane-Sarajevo 15 35 50 60. Whereas EPBiH receives only passbook subsidies used for electricity consumption, gas and district heat are subsidized in various ways. In addition to benefiting from government subsidies to consumers, Sarajevo canton provides both SarajevoGas and Toplane direct subsidies, primarily to cover fuel costs. However, the direct subsidies are closely linked to the enterprises' liquidity and are being phased out as collected revenues increase. Annex Five: Sector by Sector Review of Infrastructure Policy 39 61. The difference in collection rates and subsidies between electric power, gas and district heat is mainly due to the fact that electricity can be switched off for individual non-payers. The collection of revenues for gas and heat is technically more difficult to enforce. In addition, gas and heat supply were considered vital during the war and Sarajevo canton decreed that their consumption would not be billed. There are also historical reasons for differences in collection practices. Even in 1991, subsidies for the consumption of district heat used to cover up to one third of consumers' annual heat bills. 62. Energy bills paid through the use of passbooks have accounted for DM 4.6 million of EPBiH revenues during the period December 16, 1996 through March 31, 1997. In January 1997, SarajevoGas received US$188,000 in passbook payments and Toplane received DM 150,000. The total allocation in the budget of the canton for passbook payments is, according to the utilities and canton officials, DM12 million for the period December 1996-April 15, 1997. The canton has expressed its intention to terminate the use of passbooks for energy consumption with the end of the current heating season, April 15, but to resume the program with the start of the 1997/98 heating season on October 15. 63. Electricity rates charged to industry in BiH are similar to prices in Western European countries, although collection rates are very low. However, household electricity rates are significantly below rates charged in Western Europe (Table 5.12). It is true that household and industrial customers are charged approximately the same rates in BiH. However, the costs of serving households are significantly higher. To reflect this fact, the tariff structure will need to be revised as households' ability to pay increases. Table 5.12: International Energy Rate Comparisons Energy type Bosnia and Herzegovina OECD Europe (1995) Natural Gas (US$/107 kcal) SarajevoGas (1997) Households 268.0 431.0 Industry 360.0 125.0 Electricity ($/kWh) EPBiH (1996) Households 0.030 0.127 Industry 0.065 0.064 EPM (1997 Households 0.046 Industry 0.094 EPRS (1997) Households 0.041 Industry 0.048 Source: For the OECD: Energy Agency, IEA Statistics, Energy Prices and Taxes, Third Quarter 1996. 64. Natural gas prices in Sarajevo, although they are comparable to most other transition economies (Table 5.13) are badly distorted compared to those in Western Europe. Household prices are below the average in OECD Europe and industrial prices are three times higher. 40 Bosnia and Herzegovina: Public Expenditure Review 65. Comparative price data for district heat are not available. In the absence of meters, Toplane has based monthly charges on the area of household premises. From data supplied by Toplane, the household price in early 1997 was equivalent to US$670 per 10 kcal, and the price to all other consumers was equivalent to US$950. These are very high when compared to prices charged by SarajevoGas for individual gas supply. At these rates, district heat will not be competitive with other sources of energy in the long run. The high gas sales price from SarajevoGas to Toplane and the tariff structure of both enterprises will be analyzed and subject to change in the context of the Bank's emergency reconstruction projects. Table 5.13: Natural Gas Rates in BiH and Other Transition Economies (1996) Country Households Industry SarajevoGas (1997) 224 300 Albania 70 70 Algeria 19 12 Belarus 22 75 Croatia 203 141 Poland 233 167 Romania 14 47 Russia 17-200 60 Ukraine 64 83 Source: World Bank survey of energy prices in transition economies (unpublished staff estimates). 66. Energy rates in BiH are intended to recover costs, to the extent permitted by social policy constraints. However, the cost structure of the sector exhibits significant distortions. Fuel costs count for the biggest portion of high current cash operating expenses. In the gas sector, especially, fuel costs are problematic. The delivered price of natural gas in Sarajevo was US$120 per thousand cubic meters during early 1997, although the price of the same gas at the Ukraine-Hungary border was only US$78. The charges for transit through Hungary, the Federal Republic of Yugoslavia and Bosnia itself appear to be greatly in excess of the true costs, and suggest a profit margin of at least 25 percent. In Western Europe, the difference between the border price and the end-consumer price is only 20 percent. 67. The Federation has assumed responsibility for the import liabilities of Energoinvest. Despite the high prices charged to end-users, and US$21 million in foreign assistance in paying for gas sold by the Russian company Gazexport, US$2.2 million in arrears were added to existing Federation liabilities for gas imported by Energoinvest during the winter of 1996-97. Total Federation liabilities now stand at US$10 million. This will be a significant fiscal issue in 1997- 1998, when expected donor assistance is reduced. 68. Republika Srpska entered into an agreement with Gazexport in March 1997 to cover current consumption and US$150,000 per month to pay off its US$4.7 million share of the total debt, accumulated since October 1995, of both Entities. The alumina factory in Zvornik is the Annex Five. Sector by Secior Review of Infrastructure Policy 41 principal customer for this gas in Republika Srpska, and 99 percent of the contract costs are to be paid for in-kind, with the factory's output. 69. Significant and chronic arrears have placed all imports from Gazexport in jeopardy. The company suspended deliveries on April 7, 1997, which threatens to severely damage the Zvornik alumina facility if the smelter cools down and the alumina hardens. This sectoral issue demonstrates the severe public expenditure issues that confront the two Entities, and the essential role that political tensions have played in exacerbating them. 70. While domestic coal suppliers are highly inefficient, their high costs are not passed on to the electric companies. On the contrary, Zenica Coal reports that it is required by the government to supply coal at an average rate of DM 50 per ton to Elecktropriveda BiH, when the local market price is as much as DM 70 per ton. 71. The energy companies suffer from overstaffing. Toplane-Sarajevo is a characteristic case, employing at least two times more staff than comparable enterprises in Western Europe. Limited information is available on other energy providers, but anecdotal information suggests that the pattern repeats itself across the energy sector. 72. The sectoral requirements for subsidies through the year 1999 will be a function of three main parameters: (i) operating cost recovery of the utilities; (ii) new investments; and (iii) payments due on arrears and debt. 73. The full operating costs calculated for the utilities, including depreciation, taxes, duties and debt service, currently exceed their respective cash operating expenses by 30 percent for Toplane-Sarajevo and 50 percent for EPBiH and SarajevoGas. Without accelerated reforms in the sector, the potential fiscal burden, amounting to as much as DM 150 million per year, is very large. In current negotiations with the World Bank, authorities have agreed that by 1999, enterprises will cover their cash operating expenses. At this point the sector will resume financing of depreciation and debt service out of its own resources, leading to coverage of full operating costs. 74. Through 1999, most maintenance and new investment expenditures will be financed under the World Bank's reconstruction projects. By rehabilitating the war-damaged energy facilities, the investments cover both reconstruction needs due to direct war impact and to shortfalls in maintenance. This implies that no significant maintenance gap will exist in 1999. The energy sector emergency activities underway in BiH with the financing of the World Bank and other international donors are described in Table 5.14. 42 Bosnia and Herzegovina: Public Expenditure Review Table 5.14: Externally-Financed Investment Programs-the Energy Sector Investment Program, (US$ million) Utility Total Effective Closing Investments Date Date Emergency Electric Power Reconstruction EPBiH 196.4 1996 1998 Project Emergency Gas System Reconstruction Project SarajevoGas 55.4 1997 1999 Emergency District Heating Reconstruction Toplane- 44.5 1996 1998 Project Sarajevo 75. Of the long-term liabilities and arrears of the energy companies in the Federation, there has only been significant progress in dealing with the existing prewar debt with the World Bank, which has been consolidated under the Bank's Consolidation Loan. Payment of interest for the Loan commence in 1997 on a semi-annual basis, and principal will be due starting in 2001. The fact that both the Consolidation Loan and the IDA Credits for the current reconstruction projects are given on concessionary terms is a subsidy for the investment program of the sector. However, the onset of principal payments in the year 2001 are another signal that the current financial situation of the sector's firms must be rapidly improved. 76. Sector-internal arrears, such as inter-utility arrears and arrears on fuel supply, remain an unsettled issue. Inter-utility arrears may be offset between the governrnent-owned enterprises. However, in the absence of a government decision in this matter, inter-utility arrears will, at present, be treated as de-facto liabilities of the respective enterprises. The DM 300 million arrears of EPBiH to the coal mines are still in question and are a major threat to the viability of the enterprise. However, the government is in no position to assume responsibility itself. 77. The arrears of SarajevoGas are another significant liability for either the company or the government. Since SarajevoGas was deprived of revenues during the war, it is unlikely to be held liable for outstanding payments for gas supply, but is expected to pay, and is paying, an increasing share of the cost of current natural gas supplies. The allocation of liabilities for the existing arrears for gas is unresolved. 78. The projected financial situation of the energy utilities (Table 5.15) through 1999 is based on the assumption that the following significant reforms will be accomplished: (i) the recovery targets for cash operating costs will be met; (ii) the gas utility will not be held liable for the existing arrears for gas; (iii) current fuel costs will be covered by the enterprises; (iv) the enterprises will service their long-term debt; (v) inter-utility arrears will not be offset or forgiven; and (vi) the existing tax system will apply. Annex Five: Sector by Sector Review of Infrastructure Policy 43 Table 5.15: Company Projections of Cash Positions Cash Deficit/Surplus 1997 1998 1999 Development, (DM million) EPBiH -282 166* 162* SarajevoGas -27 -12 -3 Toplane-Sarajevo -26 -3 2 TOTAL -335 151* 161* Note: Includes debt service but not depreciation. * The positive development of EPBiH's cash balance is contingent on the settlement of liabilities with coal mines and other suppliers in 1997. 79. To quantify the consequences of insufficient growth in collection rates, that could be caused by drastic cuts in consumer subsidies, a worst case scenario of 0 percent increase in collections has been calculated for 1997. In this case, the cash deficit in 1997 would increase by 30 percent for SarajevoGas and Toplane and by 10 percent for EPBiH. 80. A key factor for the economy in general, and the energy sector in particular, is the recovery of industrial and commercial energy consumers. However, industrial recovery is currently threatened because of the suspension of gas deliveries by Gazexport, pending the resolution of the unsettled liability for gas. The gas-dependent industries in Zenica, Zvornik and Kakanj and industrial and commercial enterprises in Sarajevo are severely affected by the current interruption. Annex Six Subsidies to Industry 1. As outlined below and in the main text, due in large part to the scarcity of revenues, subsidies from all levels of government to industrial enterprises are low. The 1996 and 1997 budgets of the State, the Federation, and Republika Srpska do not contain explicit references to industrial subsidies. The 1996 and 1997 budgets of selected cantons also do not identify explicit subsidies.' Interviews with government officials confirm the near absence of direct budgetary support for industry. Indirect subsidies are also likely to be low. This annex provides some estimates of the likely maximum values of direct and indirect subsidies to industry in Bosnia and Herzegovina. Government spending 2. Despite the non-existence of explicit budget lines for support to industry, some firms are likely to be receiving direct budgetary support from the government.2 Support to these firms is likely to be included in budgetary allocations to the relevant line ministries. An upper bound on the total potential amount of direct subsidies to industry could be estimated by aggregating the total expenditures of the four large ministries most likely to engage in assisting industry: the Ministries of Industry, Energy, and Mining; Agriculture, Forestry, and Water Management; Urban Planning, Utilities, Housing, and Environment; and Transportation, Communications, and Information. 3. The amounts spent by these ministries at the Entity level in 1996, excluding donor contributions, is on the order of 12 percent of total government spending. In Sarajevo canton, the figure was double, at about 24 percent.3 In most other cantons it is significantly less (Tuzla- Podrinja zero percent; Zenica-Doboj one percent; and West Herzegovina nine percent). If all of these resources were assumed to have been channeled to industry subsidies -- an unrealistic assumption given the responsibilities of these ministries -- the total amount would be DM 210 million in the Federation, or six percent of GDP. In Republika Srpska, a similar calculation would yield DM 27 million, or two percent of Republika Srpska GDP. 4. Indirect subsidies are likely to be a more important source of government support for industry in Bosnia and Herzegovina. Among the vehicles for indirect government subsidized interest rates through lines of credit to enterprises; nonprivate enterprise arrears that could become the fiscal responsibility of the government; implicit subsidies in the tax system; and I Budgets for all ten cantons were unavailable at the time of the mission. Budgets were obtained for Sarajevo, Tuzla-Podrinja, Zenica-Doboj, and West Herzegovina cantons. On the basis of interviews with the remaining six cantons, the priorities of the four cantons for which budgets were available appear representative of all ten. Sarajevo is the only canton reviewed which had a line item (DM 12 million, or less than 3 percent of the cantonal budget) to "restart production." 2 Interviews with government officials indicated that five firms in the Bosniac-majority areas of the Federation, two firms in the Croat-majority areas, and five firms in Republika Srpska receive some direct government support. Most are firms that produce strategic goods or are considered strong candidates for employment growth. 3 Includes the referenced DM 12 million to "restart production." 44 Annex Six: Subsidies to Industry 45 subsidized utility costs. In addition, payments to "waitlisted" workers constitute an important "negative subsidy" or indirect tax on firms. Each of these subsidies is discussed below. Lines of credit 5. Lines of credit contain small subsidies by design. A number of programs offer lines of credit to firms in the Federation, detailed in Table 6.1. (To date there is little, if any, access to credit for firms in Republika Srpska.) The Bosnia Reconstruction Finance Facility (BRFF) is a DM 38.9 million line of credit supported by US Agency for International Development (USAID). World Bank loans have financed three lines of credit totaling some DM 80.8 million: the Emergency Reconstruction Program (ERP) and two on-lending operations through the Transition Assistance Credit (TAC), one benefiting the Federation and the second benefiting the State.' 6. The interest rates associated with these loans, ranging from 2 to 7.5 percent, are lower than the estimated "market" rate of on the order of 22% and thus include an element of subsidy.5 The loans have terms of approximately five years. The total subsidy over the five years is calculated to be some DM 75.6 million, or an annual subsidy of about DM 23 million for the first two years, and less thereafter (Table 6.1). Table 6.1: Maximum Interest Rate Subsidies from Credit Lines Credit Line Total Amount Interest Term Market Subsidy for Maximum Disbursed Rate (Yrs.) Rate Period Subsidy per Year TAC-State 15,300,000 3.0% 5 22.0% 14,535,000 2,907,000 TAC-Federation 33,755,000 7.5% 3 22.0% 14,683,425 4,894,475 BRFF 38,921,119 7.5% 3 22.0% 16,930,687 5,643,562 ERP 31,700,000 7.5% 2 22.0% 9,193,000 4,596,500 55,342,112 18,041,537 BOR of which: 27,493,570 Founder Cash 5,900,000 7.0% 2 22.0% 1,770,000 885,000 Malaysia* 9,924,462 3.5% 5 22.0% 9,180,127 1,836,025 Japan 11,669,108 2.0% 4 22.0% 9,335,286 2,333,822 20,285,414 5,054,847 TOTAL 147,169,689 75,627,526 23,096,384 * Does not include funds lent to non-industrial organizations. 7. Although the subsidy implicit in each of the lines of credit is small, the loans are an important source of financial capital for firms. The total amount of credit disbursed based on 4 Further information on the use of lines of credit under the Transition Assistance Credit is provided in the Implementation Completion Report, dated June 10, 1997. 5While it is difficult to determine comparable bank interest rates (with comparable risk) due to the lack of medium- to long-term capital available in the markets, one bank in Mostar said that they could offer some (limited) three year money, but the interest rate would be 18-20 percent. The estimate of 22% used in the calculations is somewhat higher. Given the fact that interest rates have been falling steadily in Bosnia and Herzegovina, and are expected to fall even further in the short to medium term, the calculation of implied interest rate subsidy is likely to be an upper- bound estimate. 46 Bosnia and Herzegovina: Public Expenditure Review external assistance as of October 1996 is DM 147.2 million, nearly one-half of domestic credit extended by domestic commercial banks in the Federation. Hence, access to these funds is important to the financial viability of recipient enterprises. Misallocation of these resources could contribute to the development of an inefficient and unsustainable industrial structure. Generally, however, lines of credit based on loans from USAID and the World Bank were underwritten based on commercial criteria. They required collateral, and several factors indicate that credit appears to have been allocated on the basis of reasonable commercial criteria: * Default rates are low: The default rate for the BRFF is zero; the ERP has had one default in four months of repayment on 173 loans. This low default rate is not unexpected, however, given that both programs are less than one year old and grace periods on principal repayments have only recently expired on the new portfolio. * The pool of applicants was large and applications were rejected: A large pool of applications provides the opportunity for lenders to discriminate among firms in the distribution of credit. Lenders of ERP funds rejected 20 percent of applications received, lenders of funds for the BRFF rejected 39 percent, and lenders of TAC funds rejected 59 percent.6 It should be noted, however, that a large pool does not necessarily guarantee selection based on economic criteria. * At least one bank supplemented credit lines: Hrvatska Banka, Mostar, added its own funds to those disbursed under the TAC to meet the requirements of the firms. The bank supplemented funds with a mixture of short- and medium-term lending to meet the firms' requirements. Hrvatska Banka is one of the few banks to be able to provide slightly longer term financing, due to its ability to access longer-term capital in Croatia. * Political pressure on loan approvals was generally low: The Directors of both BRFF and ERP stated that credit decisions were not influenced by outside parties. 8. One concern that arises with credit line programs is whether loans are being used for productive activities. Participant banks and other organizations were unanimous in stating that the loans from these lines of credit had been used to invest in productive assets and for agreed working capital requirements, as stipulated in the loan agreements.! 9. One bank presents an exception to this generally positive picture of commercially-based credit allocation. The government-owned Bank for Reconstruction and Development (BOR) began with loans worth about DM 5 million to fifteen enterprises to purchase equipment and raw 6 Industrial Task Force Meeting of March 25, 1997. 7In interviews, representatives of participant banks and other organization stated that the Ministry of Finance in the Federation did not allow the TAC funds to be used for nonproductive expenditures, such as payments to waitlisted workers (paras. 24-28). Loans were to be used to restart production, primarily by providing funds to purchase assets and raw materials. A relatively small working capital component may include salaries, but only for actively employed workers. The BRFF disbursements are closely monitored and controlled. The director of BRFF estimates that at least 75 percent of loan principal is used to purchase or improve fixed and capital assets and 25 percent is used for working capital. He estimates that the purchase of raw materials accounts for over 90 percent of proceeds allocated to working capital. The balance of working capital can be used for salaries of currently employed workers or workers mobilized from the waiting list into active employment, but not wait listed workers. Annex Six: Subsidies to Industry 47 materials in an effort to restart production. The bank was subsequently funded by state banks, enterprises, and the Malaysian government, among others. Initial financing amounted to DM 36.14 million, of which DM 19.8 million was liquid. The enterprise borrowers in the BOR portfolio are predominantly majority government-owned enterprises in the Bosniac area of the Federation. Beneficiaries appear to have been selected by the bank's board.8 While there is anecdotal evidence that loans are made and monitored on a commercial basis,9 these efforts are undermined by lax collection and low repayment rates (BOR collects only 55 to 60 percent of loan payments due, and is not foreclosing on delinquent borrowers). Liabilities of nonprivate enterprises 10. Wage arrears, as well as selected other liabilities of non-private industrial enterprises, are a potential liability for the Entities and for local governments. The Law on Basic Employment Rights and Law on Labor Relations, which predate the creation of Bosnia and Herzegovina, grant workers a high level of income security. Although there are some ambiguities concerning the intention and force of these laws, wage arrears accumulated by nonprivate enterprises could potentially become liabilities of the government, if left unpaid by the firm. 11. Wage arrears do not, however, appear to be a significant issue. Interviews with finns in Bosnia and Herzegovina suggest that, in general, finns are not accumulating significant wage arrears. Neither managers, workers nor government officials act as if they view wage arrears as a potential liability of the state. Most firms appear to adjust wage rates to reflect available working capital, and generally do not hire workers they are unable to pay. Although some firms pay workers on "waiting lists" (paras. 24-28), these payments are generally much lower than wages paid to active workers. In no case was there evidence that firns are accumulating arrears because of non-payment to workers on waiting lists. In the case of arrears to active workers, they often accumulate at the minimum wage stipulated by law. 12. Data provided by the Payments Bureaux in Sarajevo and Mostar suggest that nonprivate enterprises in the Federation accumulated less than DM 10 million in new wage arrears during 1996.) The actual stock of wage arrears is unknown, but is unlikely to exceed DM 70-80 8 The Coordination Board for Reconstruction and Development within the State Ministry of Foreign Affairs. 9Collateral accepted by BOR includes the order of acceptance at the Payments Bureau, which allows a lender to freeze the overdue amount in the giro account of a defaulted borrower, and a guarantee from a Bosnian commercial bank. (However, the bank guarantee is of questionable value; and, in practice, collecting collateral for repayment in the case of borrower default is problematic at best). During interviews with one firm which received Malaysian funds through the BOR, it was stated that not only had the firm used the funds for productive investment (machinery and raw materials), but they had repeatedly been forced to demonstrate this, and had been visited by the Malaysian Ambassador who was checking on the use of funds. 10 This calculation is based on the assumption that "other short-term liabilities" consist entirely of tax arrears and wage arrears. Wage arrears are calculated as 'other short-term liabilities', less estimated tax arrears. A description of the method used to estimate tax arrears is provided below. 48 Bosnia and Herzegovina: Public Expenditure Review million.'" No aggregate data are available to estimate wage arrears for nonprivate firms in Republika Srpska. Two of the five firms interviewed in Republika Srpska were accumulating wage arrears, but only equal to the wage bill for one month. 13. A second liability of nonprivate enterprises that could potentially become a liability of the governments of Bosnia and Herzegovina is inter-enterprise debt arrears. These liabilities arise when nonprivate enterprises receive shipments of goods from their suppliers, but are not able to pay for them. In the Federation, nonprivate enterprises in industry may be net creditors of private enterprises. The stock of credits to customers and debts to suppliers of nonprivate enterprises equals DM 677.6 million and DM 433.9 million, respectively; thus, the stock of net credit provided by nonprivate to private enterprises is DM 243.7 million, or 7 percent of Federation GDP. During 1996, nonprivate enterprises added DM 110.4 million to their stock of credit and DM 68.4 to their stock of debt, for a net flow of credit equal to DM 42.0 million, or one percent of Federation GDP. Because nonprivate enterprises are net creditors of other enterprises, such arrears do not present a fiscal problem for the governments. Implicit subsidies in the tax system 14. Firms obtain implicit subsidies through the tax system in several ways: by running arrears; by evading taxes; and by avoiding taxes through the use of exemptions and "loopholes." Firms with tax arrears receive a subsidy to the extent that the government eventually forgives or finances these arrears. The stock of tax arrears can be viewed as an upper bound on total subsidies from tax arrears. The precise value of implicit subsidies depends on three factors: the share of tax arrears that survive possible administrative and judicial challenges by firms; the share of tax arrears that are the responsibility of firms and not households; and whether such arrears are ultimately repaid, inclusive of penalties at least equal to the opportunity cost of money. Although it is known that most tax obligations accrue to firms and not households, the other caveats cannot be assessed: neither repayment and penalty rates, nor the share of arrears that survive legal challenge, are known. 15. The current levels of tax arrears do not seem to be a significant source of financing for industry. Moreover, because the Federation tax burden on formal enterprises is high; any increase in tax collection should leave the government room for badly-needed reduction in rates, leaving the fiscal situation of the government unchanged but reducing the burden on taxpaying firms. Incomplete data from Federation tax authorities suggest that the average monthly addition to tax arrears in cantons that account for 40 to 45 percent of personal income might be more than DM 900,000 per month.'2 Crude calculations suggest that the total accumulation of arrears in 1996 in the Federation could be approximately DM 25 million.'` The exact distribution of arrears between private and nonprivate firms is not known. However, of the 25 firms with the I IThis calculation is based on the assumption that the share of wage arrears in the stock of "other short-term liabilities" is equal to its share in the flow of "other short-tern liabilities." 12 Una Sana, Posavina, Tuzla-Podrinja, Zenica-Doboj, Gorazde, and parts of Central Bosnia and Neretva I 'This calculation is based on the assumnption that arrears are a constant proportion of national incomlie over the year. Annex Six: Subsidies to Industry 49 largest volume of arrears in the Federation, only two are nonprivate. A larger share of the most serious debtors in Republika Srpska are nonprivate. 16. Tax exemption arrangements differ between the Federation and Republika Srpska. In the Federation, the principal issue is the disparate tax treatment of firms in the administration of exemptions, and the economic distortions this creates, rather than in their direct fiscal impact. The most common form of exemption in the Federation is from the Special Customs Tax (10 percent). The law provides that "equipment, ready-made products, parts, intermediate goods and raw materials that are not produced within the country and are imported exclusively for the purpose of the production process..." and "...equipment imported for the 'project of reconstruction of Bosnia and Herzegovina" are exempt from customs taxation.'4 The inclusiveness of these provisions suggests that no firm's imports of goods would actually be subject to tax, however the law makes all decisions regarding these exemptions subject to approval by the Minister of Finance, based upon recommendations by the competent ministry. Reports from private industry analysts suggest that this process of allocation is not done on purely technical grounds. 17. The direct revenue losses associated with customs exemptions in the Federation is likely to be small, however. Assuming that 10 percent of total imports in 1996 is excluded from the Special Customs Tax and that one-half of exempted products would not be imported if taxed, then the total loss in revenues would be equal to DM 5.8 million. 18. A more serious tax exemption issue exists in Republika Srpska, where the tax burden is low (only 17 percent of GDP). Recent information suggests that liquor and tobacco monopolies have paid no excise taxes to Republika Srpska government. Sales taxes are also a small fraction of general revenues. One indicator of the size of this loss is that 1996 excise and sales tax revenues in Republika Srpska were less than 22 percent of general revenues. In the Federation, excise and sales tax revenues were 57 percent of general revenues. Assuming that Republika Srpska could have taken in as much from these taxes as the Federation, as a proportion of total revenues, the losses were on the order of DM 232 million, or 16.6 percent of Republika Srpska GDP and 51.8 percent of its government expenditures in 1996.15 19. Tax authorities in the Federation and Republika Srpska do not estimate the size of the shadow economy, and making a precise estimate of subsidies to firms through tax evasion difficult. However, based on a conservative estimate that one third of income is not taxed, one tenth of this untaxed activity is industrial, and one half of this activity would not take place if 14 See "Information on Customs Incentives-Exemptions, Pursuant to the Customs Law", Federation Ministry of Finance, Federation of Bosnia and Herzegovina. Is Including expenditures based on donor credit. 50 Bosnia and Herzegovina: Public Expenditure Review taxed, the amount of lost revenue to governments in the Federation"6 is DM 27.7 million and in Republika Srpska'7 is DM 5.8 million. Subsidized utility costs 20. One important indirect subsidy to industrial enterprises is the subsidy implicit in the provision of some utilities. Rates for energy and water are significantly below levels needed to cover operating, maintenance and replacement costs; collection rates for industry are significantly below those for households, and service providers are discouraged from cutting off service to delinquent industrial customers (Annex 5). Officials at the aluminum company, Aluminij-Mostar, indicated that they expected to pay substantially less than the prewar price for electricity, once they restart production. Only the significant availability of hydroelectric power and its low marginal costs have allowed energy supply to remain as high as it has, given current collection rates. Especially in Republika Srpska, however, the absence of payment for services will hasten equipment breakdown and increase the number of power outages. "Waitlisted" workers 21. One institution that appears unique to enterprises in the former Yugoslavia is the enterprise's list of workers on "waiting lists." These are workers who are not actively engaged in production at present, but who are retained by the enterprise in anticipation of future work. According to enterprises, workers can be placed on a waitlist for six months. During that period, the enterprise is responsible for paying a basic wage, and meeting all other statutory entitlements (including contributions to the pension fund and other indirect taxes). After six months, waitlisted employees become the responsibility of the government. 22. Table 6.2 presents statistics on the number of workers in waiting in industry and mining in the Bosniac-majority areas of the Federation. As shown, the ratio of active workers to waitlisted workers is less than two to one, with demobilized soldiers comprising more than one- third of the workers in waiting. If enterprises fully complied with the compensation requirements of the law, maintaining these workers could be costly to the firm. The total estimated cost to firms in the Federation, based on full compliance, would have been DM 33.6 million;"8 the estimated cost to firms in Republika Srpska would have been DM 11.3 million.'9 16Assuming that the tax rate is 32 percent, which is equal to tax collections to GDP in the Federation in 1996. This methodology is based on the crude assumption that all taxes are the burden of firms. ]7Assuming that the tax rate is 17 percent, which is equal to tax collections to GDP in Republika Srpska in 1996. This methodology is based on the crude assumption that all taxes are the burden of firms. 1 This estimate is based on the following calculation: one-half the monthly wage times six months times the number of industrial workers in waiting. The number of industrial workers in waiting is estimated as follows: the number of industrial workers in waiting in Bosnia-majority areas times the number of all workers in waiting in the Federation divided by number of all workers in Bosniac-majority areas. The data source for the Federation is: Statistical Data on Economic and Social Trends, Institution of Statistics of Bosnia and Herzegovina, Sarajevo, April 1997. Annex Six: Subsidies to Industry 51 Table 6.2: Number of Workers in Waiting in Industrial Enterprises in Bosniac-Majority Areas in the Federation, March 1997 Canton Number of Number of Industrial Number of Demobilized Industrial Workers Workers in Waiting As A Soldiers in Waiting As A in Waiting Percent of Total Percent of Total Industrial Industrial Workers* Workers in Waiting Una Sana 5,112 52% 28% Tuzla-Podrinja 10,076 31% 30% Zenica-Doboj 10,048 31% 49% Gornjedrinski 377 60% 73% Cen'tral Bosnia 1,184 18% 21% Neretva 253 4% 33% Sarajevo 10,834 52% 37% Total 37,884 35% 37% Source: Statisticki Podaci 0 Privrednim I Drustvenim Kretanjima Na Podrucju, various cantons, March 1997. Note: These data include workers in both industry and mining. * Total industrial workers include industrial workers in waiting. 23. Not surprisingly, actual practice regarding payment to waitlisted workers appears to deviate substantially from the rules stipulated in the law. Some firms provide workers with only partial salary, others provide only assorted benefits such as food, and many others provide no compensation at all. Of the five firms interviewed in Bosniac-majority areas of the Federation, all maintained waitlists; of these, two paid no benefits to waitlisted workers, one paid a salary equal to the minimum wage to demobilized soldiers on the waitlist, one provided occasional income and benefits to all its workers in waiting, and only one compensated workers according to the law. The two firms interviewed in Croat-majority areas of the Federation maintained waitlists, but did not compensate workers on these lists. Only three of the five firms interviewed in Republika Srpska maintained waiting lists; only one of these provided any compensation to workers, which it did in compliance with the law. If one assumed that not more than 10 percent of firms in the Federation and five percent of firms in Republika Srpska comply with the law, then the total expenditures by firms in each Entity would have been DM 3.4 million and DM 0.6 million, respectively, or significantly lower than the fiscal implications of the law. 24. The continued existence of this policy exposes firms to possibly arbitrary enforcement actions by government. In general, implementation of a law that forces firms to operate in the shadow of the law is bad practice; it sets a precedent that can have adverse consequences for tax compliance in the long term. If the policy is ever strictly enforced, its cost- would be substantial. Payments by firms to workers in waiting would be a significant drain on the working capital of firms, at a time when working capital is in very short supply. This is an inefficient way for the government to avoid the fiscal burden of unemployment compensation. 19 This estimate is based on the following calculation: one-half of the annual wage times six months times the ratio of workers in waiting to active workers in industry. The annual wage is approximated based on social fund contributions, which are assumed to be 100 percent of wages. The ratio of workers in waiting to active workers in industry is based on the ratio for Bosniac-majority areas of the Federation. 52 - Bosnia and Herzegovina: Public Expenditure Review 25. Moreover, the waiting list "tax" on firms is a regressive way to deliver assistance to the most needy. Beneficiaries are determined by prior employment history and the current condition of the enterprise to which they happen to have had an employment relationship, rather than need. The scarcity of resources suggests that the waiting lists, and the resources diverted to them, should be changed to an equitable and fiscally sustainable policy for government delivery of benefits to the most needy.