Shaping a Pension System: Distributive and Incentive Effects of the Russian Pension Reforms

Transcription

1 Unversà d Torno Unversà del Pemone Orenale Shapng a Penson Sysem: Dsrbuve and Incenve Effecs of he Russan Penson Reforms Irna Kovrova Doorao d Rcerca n Scenze Economche

2

3 Unversà d Torno Unversà del Pemone Orenale Shapng a Penson Sysem: Dsrbuve and Incenve Effecs of he Russan Penson Reforms Irna Kovrova Doorao d Rcerca n Scenze Economche Cclo XVIII Supervsore: Prof. Elsa Fornero Coordnaore: Prof. Alessandro Sembenell

4 Dparmeno d Economa S. Cogne de Mars Dparmeno d Scenze Economche e Fnanzare G. Prao Dparmeno d Sasca e Maemaca Applcaa alle Scenze Umane Dego de Casro Dparmeno d Polche Pubblche e Scele Colleve Dparmeno d Scenze Economche e Meod Quanav Dparmeno d Ssem d Produzone ed Economa dell Azenda Unversà degl Sud d Torno

5 Conens Acknowledgemens Inroducon......v 1. The Russan Penson Sysem: A Descrpve Analyss Inroducon Overvew of he pre-reform socal secury sysem The sysem durng he socals regme The ranson of he 1990s The 2002 penson reform The 2005 nnovaons Gender redsrbuon Socal secury money s worh measures The framework NPVR (Ne Presen Value Rao) RR (Replacemen Rae) Conclusons A Esmaon of he earnngs profles B The evoluon of he socal secury conrbuons and he penson s deermnans formulas Why Do Russan Pensoners Connue Workng and Wha Happens f Penson Benefs are Taxed for Those Workng? Inroducon Reremen models: whch lessons for Russa? A model of he Russan pensoners decson o connue workng Emprcal specfcaon Daa descrpon Esmaon resuls The Russan pensoners probables of workng and some polcy smulaons Would he susanably of he socal secury sysem be subsanally mproved by nroducng a penson ax for hose workng? Conclusons A Esmaon of he wage equaon B Calculaon of he number of pensoners who receved a penson as rerees n

6 3. Effecs of he Inroducon of a Funded Pllar on he Russan Household Savngs: Evdence from he 2002 Penson Reform Inroducon Savng behavor of Russan households: evdence from he economc leraure Daa examnaon Usng dfferen defnons of savngs Dsrbuon of savng raes across household ncome Age profle of he savng raes Mehod and daa Monhly savng raes : summary sascs for he sample used n esmaon and esmaon resuls Annual savng raes : summary sascs for he sample used n esmaon and esmaon resuls Effecs of he nroducon of a funded pllar on household savngs: evdence from he 2002 penson reform Household savng behavor and socal secury Household savng behavor and socal secury reforms nroducng funded componen Savng mplcaons of he socal secury reform n Russa Mehod and daa Summary sascs for sample used n esmaon and esmaon resuls Conclusons A Correlaons B The Tob regresson model Appendx 1: The Russan Longudnal Monorng Survey 105 Bblography 109

7 Acknowledgemens I would lke o acknowledge he help of many people durng my sudy. Elsa Fornero s he frs person o whom I am hghly ndebed for her enhusasc supervson, paence and encouragemen. Helpful and nsghful commens came from he dscussons wh Ugo Colombno, Alessandro Sembenell, Marghera Borella, Mchele Bellon, Rccardo Calcagno and Olga Marshkna. I wsh o hank Slva Maero for her careful revson of my Englsh. I am graeful o colleagues from CeRP (Cener for Research on Pensons and Welfare Polces) for he smulang and very frendly research envronmen. I would lke o hank he parcpans o CeRP and DSE (Doorao n Scenze Economche) semnars for helpful commens and feedback. I have also benefed from dscussons wh eam members of he projec on Assessmen of he labor marke mpac of he penson reform from he Maasrch Graduae School of Governance. I graefully acknowledge he fnancal suppor of he Unversy of Turn, CeRP and Fondazone CRT (Progeo Alfer). Needless o say, I am solely responsble for my errors and shorcomngs of hs work.

8 To Dusa v

9 Inroducon Inroducon 1. Socal Secury s a very mporan elemen for he well-beng of workers, her famles and he enre socey. I proecs people from a sharp declne of lvng ha may occur when hey lose her regular earnngs by lfng, among oher hngs, he lowes ncome elderly ou of povery. A well desgned socal secury sysem should have several feaures. Frs of all, should be adequae. The socal secury adequacy means o nsure a ceran lvng sandard n he old age, ha requres ncome ransfers from hgh-earner workers o low-earner ones. Second, a socal secury sysem should provde penson benefs whch are commensurae wh worker s hsory of conrbuons. The opmal balance beween socal adequacy and ndvdual equy s exacly wha one expecs from a developed sysem. I s no possble o mprove or even o manan he adequacy whou he fnancal susanably of he sysem. Tha s why a provson of benefs whch are susanable and robus o major shocks s among he prmary goals of each penson sysem. In s urn, he fnancal susanably s ndrecly nfluenced by he equy of he Socal Secury snce he laer plays a crucal role n ndvduals ncenves o parcpae and conrbue. Penson provsons mpac he behavor of boh he young and aged populaon and can creae deadwegh losses accordng o he sandard economc models ha are used nowadays o sudy households consumpon and labor supply neremporal decsons. Whle balancng proecon and dsorons, a well desgned socal secury program should mnmze he negave effecs. The nfluence of he socal secury sysem on ndvduals reremen, savng and labor supply decsons s a very mporan ssue o analyze boh heorecally and emprcally. v

10 Inroducon Ths research concerns he Russan socal secury sysem and ams a nvesgang emprcally some of he maers ha have been menoned above. The work s composed of hree chapers, whose conens are brefly summarzed n ha follows. 2. In 2002, Russa reformed s penson sysem movng from a defned benef o a defned conrbuon penson scheme wh he nroducon of a mul-pllar penson sysem. Improvemens of boh he adequacy and he fnancal susanably were among he major objecves whch he reform nended o acheve. Some oher mporan measures were mplemened n 2005, when he compulsory penson nsurance conrbuon rae was reduced by 8%, n order o ease he ax pressure on he employers and o rase he offcally repored salares. I was soon clear ha hs polcy nervenon could have heavly undermned he fnancal susanably of he sysem and ha, accordng o subjecve analyses, he new penson sysem could no have nsured o he rerees adequae, equable and susanable benefs. In he frs chaper of he hess, besde a bref descrpon of he evoluon of he Russan penson sysem snce he Sove mes, we analyze he above changes. Ths descrpve analyss s an mporan prelmnary sep before of he emprcal sudes n he second and hrd chaper, ha wll be done wh more sophscaed economerc ools. To provde a more objecve bass for a dscusson on he Russan Socal Secury feaures, we use he socal secury money s worh measures. More precsely, we compue he ne presen value rao, ha s he rao beween he presen values of benefs and conrbuons, and he replacemen rao, ha s he rao beween he average penson of each cohor and he average wage of ha cohor. Snce we do no have any nformaon abou he conrbuons o he socal secury sysem durng socalsm, we consder hese ndcaors only for he cohors born n 1970, 1975 and 1980, whch are supposed o sar her workng lfe afer he collapse of he Sove Unon n We show ha he 2005 reducon of conrbuon raes o he socal secury wll enal a reducon n he replacemen raes. We fnd ha, n he nex decades, he replacemen raes wll have a declnng srucure, ha can reduce ndvduals ncenves o conrbue o he sysem. v

11 Inroducon Moreover, we show ha he pos-reform sysem provdes a beer correlaon beween he worker s conrbuons and hs/her penson benefs, han he pre-reform sysem. Neverheless, he redsrbuve naure of he sysem s sll very mporan. In general, he redsrbuon s hgh no only because of he so called basc penson, desgned exacly for hs purpose and fnanced by a half of he payroll ax rae, bu also because he acual lfe expecancy s no adequaely refleced n he annuzed benefs of he noonal defned and he funded componens. In parcular, he new penson sysem keeps srong redsrbuve elemens owards women. Under he pressure of populaon agng, s ofen argued ha an ncrease of he normal sauory reremen age s he bes way o mprove boh he fnancal sably and he adequacy of a socal secury sysem. Unforunaely, he effecveness and he farness of hs measure s very doubful for Russa, where we observe a decrease n boh ferly rae and lfe expecancy. Takng no accoun ha women s lfe expecancy sgnfcanly exceeds men s one and, neverheless, women are elgble o rere fve years younger han men, we argue ha would be naural o ncrease he normal sauory reremen age only for females. For example, he reremen age for women could be rased o ha for men, as n many European counres. Anoher possble way o fnd he addonal resources o pay oday s pensons s lnked o he fac ha, n Russa, pensoners may receve full pensons and keep workng smulaneously. Noe ha, snce one can work and ge full penson benefs a he same me, he erm reremen becomes ambguous. In hs conex, under reremen we undersand he acual age when a person whdraws from he workforce and sars lvng on he penson benefs. A he same me, we defne he sauory reremen age as he age of elgbly of he penson benefs, regardless of he employmen saus. A pensoner s hen defned as an ndvdual who has reached he sauory reremen age. Gven he descrbed feaure of he Russan penson sysem and he above consderaons, one may hnk o ax penson benefs for he workng pensoners. Ths las polcy measure wll be furher analyzed n he followng chaper where an analyss of he labour supply decsons by workers/pensoners s underaken. Tha wll be our own conrbuon, based on economercs, o he polcy debae ha res o address he problem of fnancal sably of he Russan penson sysem. v

12 Inroducon 3. Pensoners parcpaon o he labor force n Russa s a very neresng phenomenon, because many Russans connue workng afer he sauory reremen age. The mos mporan reason why Russan pensoners connue workng s lnked o fnancal needs. Very ofen, due o he modes penson benefs and he low level of savngs, remanng a work s he only possbly o manan he pre-penson lvng sandard. Today, a grea number of low-pad and unaracve jobs, especally n he publc secor, are occuped by pensoners. Involvng more pensoners n work can also reduce he forhcomng labor shorage caused by a declne n he workng-age populaon. The concep of he early reremen s meanngless n Russa, snce penson benefs are gven o he ndvdual as soon as he or she reaches he sauory reremen age, regardless of whdrawal from he workforce. Thus, he sandard measures capurng he margnal penson ncenves such as he opon value, he peak value or he mplc ax rae are no nvolved n he ndvdual s decson o sop workng afer he sauory reremen age. In he second chaper, assumng ha a pensoner makes hs/her choce wheher o keep workng or no on he base of hs/her consumpon and lesure, under mperfec cred marke, we develop and esmae a smple model of he Russan pensoners decson o connue workng. The emprcal evdence shows ha age, healh and educaon are he key facors whch deermne he pensoners choce o reman a work. Moreover, he pensoners propensy o keep workng s hgher for urban dwellers, n parcular for he resdens of S.Peersburg and Moscow. We also fnd ha Russan pensoners have a srong propensy o reman a work, ha s an ncreasng funcon of wage and s decreasng n he penson benefs and ncome of he pensoner s famly ha he/she akes as exogenous. By employng he esmaed model, we show ha he nroducon of a penson ax for workng pensoners would no enal a dramac whdrawal of pensoners from he workforce f he ax depended progressvely on he wage. However, we argue ha hs nervenon canno subsanally mprove he fnancal susanably of he socal secury sysem. Neverheless, hose addonal resources could be effecvely used o mprove s adequacy. v

13 Inroducon 4. Socal secury sysem plays an mporan role n nfluencng households savng behavor. I s hen obvous ha a penson reform can be a crucal deermnan n reshapng households decsons on savngs. In he hrd chaper, we provde an analyss of he savng behavor of he Russan households. We fnd ha he overwhelmng majory of he Russan famles do no pu asde. Moreover, we show ha a very large poron of he Russan household savngs s due o he hgher-ncome households, ha s conssen wh he prevous fndngs for Russa and for he oher counres as well. An esmaon of he savng raes age profle shows ha he household savng raes do no vary subsanally wh he age of he household head. The savng raes are hgher f he household head has a hgher educaon, for households wh an employed and male head, for urban households, for marred couples and for famles wh no chldren; hs s n lne wh nernaonal experences. We hen urn o our man objecve, ha s he esmaon of he effecs of he nroducon of a mandaory funded pllar on households savngs. To hs purpose, we explo he 2002 Russan penson reform, when a funded componen was nroduced along he payas-you-go componen. We fnd emprcal evdence ha he nroducon of he funded pllar has a negave mpac on he Russan households savngs. The resul s no surprsng. The nroducon of he funded pllar n 2002 has gven he possbly o he parcpans no only o manage resources of hs pllar by choosng an nvesmen manager, bu also o leave he funded par of he penson o he famly members. Therefore, households can decrease her savngs due o expecaons of he hgher rae of reurn on her conrbuons and hey can reduce her famly savngs for rany days. 5. Fuure research may concern he nnovaons underaken n As we menoned above, n 2005 he compulsory penson conrbuon rae was reduced. In order o fnance oday s pensons, smulaneously wh ha measure, he conrbuons of he mddle-aged workers hose born beween 1952/1956 (men/women respecvely) and 1967 have been swched Usually, savng decsons are aken whn he household and no by each sngle member of famly. Tha s why we always alk on household savngs raher han on ndvdual ones. x

14 Inroducon from he funded pllar o he noonal conrbuon one. The furher neresng queson o sudy s hen how hose measures have mpaced he Russan household savngs. 6. In all hree chapers, we employ daases derved from he Russan Longudnal Monorng Survey (RLMS), he frs represenave sample of he Russan Federaon, whch has been organzed and coordnaed by he Unversy of Norh Carolna. The RLMS s a householdbased survey desgned o measure he effecs of Russan reforms on he economc well-beng of households and ndvduals. Noe ha workers born n 1952/1956 (men/women respecvely) and earler were no elgble o parcpae o he funded pllar snce he old workers have a relavely shor workng lfe o collec enough money n he funded pllar. x

15 Chaper 1 Chaper 1 The Russan Penson Sysem: A Descrpve Analyss 1.1 Inroducon Durng he las decade, here have been nense and endless dscussons abou how o srenghen he weak fnancal susanably and how o mprove he adequacy 1 of he Russan socal secury sysem. The resoluon of hese problems was among he man objecves of he 2002 penson reform and of he changes underaken n Alhough several mporan measures have been mplemened, wo fundamenal mechansms have no been appled ye. Frs, he normal sauory reremen age has no been ncreased, and s sll very low 55/60 years for women/men respecvely. Second, acual lfe expecancy a reremen has no been nroduced n he penson formula. Today, he need o readly pu hese mechansms no acon s wdely acknowledged and, as a consequence of hs, here s bg debae abou hese poenal polcy nervenons. The man arge of hs paper s o revew he Russan socal secury sysem, whch wll allow us a beer undersandng of he curren problems, as well as o provde a more objecve bass for furher dscusson. To hs am, we compue wo socal secury money s worh measures for he cohors born n 1970, 1975 and 1980, 2 he ne presen value rao and he 1 An adequae socal secury sysem seeks o nsure a ceran lvng sandard n he old age. 2 Unforunaely, we do no have any nformaon on conrbuons o he socal secury sysem durng socalsm, so we consder only cohors ha sared conrbung afer he collapse of he Sove Unon n

16 Chaper 1 replacemen rae. The ndvdual wage profles needed o compue hese measures are esmaed usng a daase drawn from he Russan Longudnal Monorng Survey (RLMS). Our esmaons show ha, whle earnngs wll end o ncrease across cohors n he nex decades, he replacemen raes wll end o decrease. Moreover, we fnd ha he 2005 reducon of conrbuons o he socal secury sysem enals a reducon n replacemen raes. We also look a he nra-cohor redsrbuon of he socal secury sysem. A penson sysem ha ncorporaes a guaraneed basc penson s necessarly redsrbuve. However, n a sysem provdng annuzed benefs, a subsanal redsrbuve componen comes from he sysemac dfferences n he lfe expecancy of he populaon. The mos srkng example n Russa refers o he redsrbuon from men o women. On he bass of our esmaes, we fnd ha he pos-reform sysem es benefs o conrbuons more han he pre-reform one dd. Neverheless, he new penson sysem remans sgnfcanly redsrbuve, n parcular n favor of women. Today, n Russa, lfe expecancy a brh s abou 59 years for men and 72 years for women and lfe expecancy a he age of 60 s abou 14 years for men and 19 years for women. On hs bass, we argue ha an ncrease n he normal sauory reremen age only for women would be a more proper decson han ncreasng he sauory reremen age for boh genders. For example, he reremen age for women could be equalzed o ha for men, as has been done n many developed counres where he equal rghs of genders are an mporan ssue n he legslaon process. The paper s organzed as follows. In secons , we brefly descrbe he man characerscs of he Russan socal secury sysem snce he Sove revoluon, gong hrough he 2002 reform o he las changes n Secon 1.5 analyzes he redsrbuon naure of he socal secury sysem owards women. In secon 1.6, we assess some socal secury money s worh measures. Secon 1.7 concludes. 2

17 Chaper Overvew of he pre-reform socal secury sysem The sysem durng he socals regme The frs pensons of he Socals perod were se snce he early 1920s for he Red Army veerans and her famles. A shor me laer, penson coverage was esablshed also for some caegores of workers and pensons were pad hrough a specal knd of labor unons. In 1936, was decreed ha all czens would be enled o he socal secury provson n her old age. A comprehensve penson sysem was creaed only n 1956 bu remaned far from an unversal one (Buckley, 1998). For example, dd no cover collecve farm workers, whch sared recevng penson benefs only n In wha follows, he mos mporan dsncve feaures of ha sysem are descrbed. The normal sauory reremen age was 55 years for women wh a mnmum senory of 20 years and 60 years for men wh a mnmum senory of 25 years. Agan, he collecve farm workers were dscrmnaed. To aan penson benef, hey were requred o be a leas 60/65 years old (women/men respecvely), wh a mnmum senory of 25 years for boh genders. Some parcular caegores of workers were allowed o rere before of he normal sauory reremen age and/or o receve hgher penson benefs. For example, ndvduals employed on dangerous or physcally heavy jobs had he rgh o rere 5 years younger, women a he age of 50 wh 15 years of job experence and men a he age of 55 wh 20 years of job experence. By manpulang hese dfferen reamens, he sae boh recrued ndvduals o unaracve jobs and creaed he ncenves o move o underpopulaed areas, as he norhern errores. The conrbuons o he socal secury sysem were pad by he sae-owned enerprses as a percen of her wages funds. The workers dd no drecly conrbue o he penson sysem and raher regarded pensons as a gf from he sae. Despe he wage level of he sove workers was very low n comparson wh he oher counres and he man par of workers earnngs were whdrawn for he sae s needs, sove people were convnced ha pensons were a socals achevemen. Needless o say, ndvduals could no op ou of he sysem. Penson benefs were compued accordng o he ndvdual s senory and he prereremen wage and pensons were se once and never ndexed. Thus, he oldes ndvduals 3

18 Chaper 1 receved lower pensons han hose who rered laer, regardless of he same qualfcaon and senory. Gven he above feaures of he sysem, he e beween benefs and conrbuons was very weak. The mporan people, for example hose holdng hgh posons n he communs pary and he Red Army, formed a specal group, whch go he so called personal pensons. People from ha group no only aaned hgher penson benefs bu also had a number of oher prvleges, such as specal medcal care and access o specal sores. In general, he level of penson benefs was low. Tha led o a subsanal gap beween he ndvdual s lvng sandard before and afer he reremen. Snce pensoners could work and receve full penson benefs smulaneously, many pensoners connued workng. In fac, n he perod , he number of workng pensoners ncreased by fve mes (Snyavskaya, 2001). To conclude, he sove penson sysem was an negral componen of he cenral planned economy wh s sae-owned ndusres, collecve farms, full employmen and srcly saeconrolled prces The ranson of he 1990s Afer he collapse of he Sove Unon, Russa sared mplemenng radcal marke-orened reforms. The counry faced unemploymen, hgh nflaon, resrucurng and prvazaon of enerprses. The sove penson sysem, desgned for a command economy under a oalaran governmen, could no be susanable anymore. A need o esablsh a new penson sysem ha would have been compable wh a marke-based economy arose. In November 1990 a new law called On he sae pensons n he Russan Federaon passed. Tha law had he followng arges: 1) o esablsh a unfed penson sysem for all workers, coverng he whole populaon n he Russan Federaon; 2) o abolsh he mos prvleged benefs (personal pensons); 3) o ncorporae he leas prvleged scheme (collecve farmers pensons) no he unfed sysem; 4) o esablsh a mnmum penson level; 5) o provde a proecve mechansm agans nflaon; 6) o connec ndvdual s penson wh boh wage and senory. 4

19 Chaper 1 In hs conex, a Penson Fund was creaed o ac as an ndependen exra-budge fnancal nermedary, based on he conrbuons by all employees and employers. Bu, already n 1993, he Penson Fund los s ndependence and became srongly opened o he nfluence by he governmen. The Penson Fund s resources were also used for purposes ha dffered from he nal am of provdng he nsurance-based penson benefs. For example, he Penson Fund was forced o pay he war nvalds pensons and o fnance he celebraons of he World War II Vcory Day. The socal secury conrbuon rae was se o 29%; 28% of he gross wage was pad by he employers and 1% of he afer ax employmen ncome was pad by employees. Overall, he new penson scheme resed on he prevous one. The normal sauory reremen age was no ncreased and remaned as had been esablshed jus afer he Ocober Revoluon, 55/60 years for women/men respecvely wh a mnmum requred senory equal o 25/20 years. The penson benef was se o be equal o 55% of he prevous wages averaged ou over eher he bes fve or he las wo years (a percenage ha was ncreased by 1% for each addonal year of servce over he mnmum requred senory, up o a maxmum of 75%). Moreover, he penson benef could no exceed he maxmum or deprecae below he mnmum level fxed by law. In he meanme, he number of pensoners was connuously growng, whle he coverage rao, ha s equal o he rao of workers o pensoners, was decreasng (see able 1.1 and fgure 1.1). The declnng srucure of he coverage rao was (and s) he resul of a connung reducon n he ferly rae and ncrease n adul moraly, especally among mddle aged males (Fornero e al., 2006). As well as he old socals penson sysem, he new one was hghly redsrbuve. Mos of he prvleged pensons were no abolshed. Moreover, new prvleged groups came ou. In he lae 1990s, one pensoner over seven sared o receve he penson benef before he/she had reached he normal sauory reremen age (Afanasev, 2003). To proec pensons agans hgh nflaon durng he ranson perod, a very complcaed mechansm of penson ndexaon and compensaon was esablshed. However, hs was no well desgned, so penson benefs n real erms fell back very rapdly. Whle penson benefs 5

20 Chaper 1 had been beween 60% and 100% of he average wage durng he lae 1980s (Buckley, 1998), durng he 1990s, hey ranged beween 20% and 40% of he average wage (see Table 1.2). Table 1.1: Penson Aged Populaon* n he Sove Unon and he Russan Federaon, Years Sove Unon Toal populaon 170,557, ,800, ,436, ,400, ,700,000 - Penson Aged Populaon 14,285,000 25,510,000 36,377,000 40,672,000 49,026,000 - Percen Penson Aged Russan Federaon Toal populaon 108,377, ,534, ,941, ,402, ,022, ,938,000 Penson Aged Populaon 9,361,000 13,827,000 19,987, ,000 27,196,000 29,931,000 Percen Penson Aged * Penson Ages are 55 years for women and 60 for men Source: Burckley, 1998 Noes: he daa are aken from Rashd e al.. (2002) and Fornero e al.. (2006) 2.1 Workers/Pensoners Fgure 1.1: Coverage Rao, Years

21 Chaper 1 The possbly o receve a full penson benef regardless of he whdrawal from he labor force was mananed on. Penson benefs of hose who decded o pospone he whdrawal of penson, connung o work afer he sauory reremen age, were ncreased by 10 percen for every year, bu only for he frs 3 years. Obvously, consderng he low lfe expecancy and he fac ha n absolue erms he ncenve was very low, he ndvduals clamed penson benefs as soon as hey were elgble. Table 1.2: The raos of he average penson o he mnmum lvng sandard and o he average wage o he mnmum lvng sandard o he average wage Noe: all numbers are expressed n percenages Source: Snyavskaya (2001). The dffcules of he ranson perod and he nrnsc drawbacks of he penson sysem weakened very much he e beween conrbuons and benefs. Gven hs sae of affars, workers were easly nduced by employers o offcally repor wages lower han he acual ones n order o avod a heavy ax burden. Because of boh ax evason and absence of a well desgned admnsrave machne provdng a severe monorng of he suaon, he procedure of collecng axes became very dffcul. As a resul, he Penson Fund expendures sared o grow whle he conrbuon base was exhaused; hs suaon led o he phenomenon of he arrears n he paymens of he penson benefs. In fac, durng he 1990s, penson arrears were subsanal and reached up o 3-6 monhs (Rashd e al. (2002)). The new pay-as-you-go scheme nhered from he socals pas was clearly no susanable. An mprovemen of he sysem by eher ncreasng payroll axes or decreasng pensons was no accepable, snce he socal secury ax was already hgh and pensons were already low. An aemp o reform he Russan penson sysem was done n 1995 under he World Bank s nsrucons. The governmen dscussed he esablshmen of a new mul-pllar penson srucure. Three pllars were planned: 1) a basc socal penson; 2) a mandaory labor penson; 3) a supplemenary prvae penson. The frs wo pllars were supposed o be 7

22 Chaper 1 fnanced hrough payroll axes and he hrd one by ndvduals savngs. Bu, he proposed reform dd no found a polcal suppor and was pu asde. The nex proposal for a new penson sysem wh a hree-pllar srucure daes The reform, whch was supposed o come no force n 1999, planned he creaon of he followng pllars: 1) a socal penson (he Sae penson provson for ndvduals who were no elgble for he Sae reremen nsurance); 2) a Sae reremen nsurance ha was composed of a pay as you go (henceforh, PAYG) componen and a funded componen; 3) a supplemenary penson nsurance and provson. Unforunaely, because of he 1998 fnancal crss, he realzaon of he projeced reform was no possble. Thus, he penson reform was agan posponed o he fuure. The economy of he counry was heavly h by he 1998 fnancal crss. Pensoners suffered mporan losses from he consequen devaluaon of he ruble and he burs of nflaon ha gravely reduced her real pensons. In 1998, a new penson formula was nroduced wh he purpose of ncreasng a dfferenaon n he penson benefs. The new formula was based on he so called ndvdual coeffcen. The ndvdual coeffcen was equal o he rao of he ndvdual s average wage (of eher he bes fve years or he las wo years) o he average wage n he Russan Federaon repored by he Russan Sascal Commee n he same perod, and could no exceed 1.2. Accordng o he new formula, he penson was equal o 55% of he average wage n he Russan Federaon (whch was approved by he governmen) mulpled by he ndvdual coeffcen and by he rao of ndvdual s senory o he mnmum requred senory. The prevous penson formula was no abolshed, and ndvduals could choose beween he old and he new one. In spe of he offcal clam ha he maxmum penson was no fxed n he case he penson was calculaed accordng o he new formula, lke was wh he old formula, ha was jus an lluson. In fac, he penson lmaon was se hrough boh he ndvdual coeffcen and he average wage. In addon, he new formula could be chosen only f an ndvdual sopped workng afer he sauory reremen age, oherwse he/she had o use he old one. Acually, he nroducon of he new penson formula made penson calculaon more complcaed and led o a decrease n he number of he offcally employed pensoners. 8

23 Chaper 1 The Russan Federaon enered he 21 s cenury wh an unsusanable and complcaed penson sysem ha needed o be radcally ransformed. I s well known ha because of he nsably, weakly developed fnancal markes and he shadow economy, a polcy-makng process n he ranson and developng counres s more dffcul han n counres wh developed marke economes. Whle conemplang a penson reform, polcy makers have o ake no accoun he nal suaon of he counry, n order o decde on he more approprae penson scheme. Barr (2000, 2001) argues ha s useful o dsngush beween prerequses whch apply o all reforms and hose over whch a polcy maker has a choce. Moreover, always accordng o Barr, some prerequses of he publc secor, such as he polcal susanably of a penson reform package, he admnsrave capacy o enforce axes/conrbuons and he capacy o manan macroeconomc sably are essenal for boh he sae and he prvae scheme. A he same me, he fscal susanably of sae scheme s of course mporan only for he sae schemes. As for prerequses of he prvae secor, he auhor menons, among ohers, requremens such as a suffcenly well-nformed populaon and well developed fnancal markes. Lookng a he suaon of he Russan Federaon, s evden ha some prerequses of he publc secor, such as he admnsrave capacy o enforce axes/conrbuons, are lackng for obvous reasons (n general, hs s a problem of all he ranson economes) and he same s rue for he above prerequses for he prvae secor consderng, for example, as for he sably and he well-beng of he fnancal markes, he no so far away 1998 fnancal crss. Furher on he drecons of he reform n ranson economes, Rukovsk (1999) clams ha he desred reformed penson sysem should be a parally prvazed sysem ha ensures an adequae mnmum benef and lnks benefs o lfeme conrbuons. In hs sysem, a share of he ndvdual s fuure penson should depend on he wage growh rsks, and he oher share on he capal marke rsks. To summarze, he world experence accumulaed by ha me showed ha a penson reform could go n dfferen drecons. A counry could choose o mprove an exsen pay-asyou-go scheme or o nroduce a mul-pllar model or o urn o a fully funded sysem. 9

24 Chaper The 2002 penson reform In 2002, Russa reformed s penson sysem. The reform was nended o solve he problems of he old sysem and had he followng arges: 1) o nroduce ransparen elgbly condons and smplfy he benef formula; 2) o mprove he fscal solvency of he penson sysem by ncreasng ndvduals ncenves boh o conrbue o he sysem and o work longer; 3) o mprove he adequacy of he sysem. Wh he reform, Russa nroduced a mul-pllar sysem and shfed he penson provson from a defned benef penson scheme o a defned conrbuon one. I s commonly clamed ha he advanage of he mul pllar sysem s gven by he possbly o beer dversfy rsks. However, Barr (2000) argues ha hs proposon s rue only f rsks are negavely correlaed. For example, economc and demographc rsks are common o boh fundng and PAYG schemes, whle polcal rsks (e.g., unsusanable PAYG sysems) and nvesmen/managemen rsks may be ndependen. Accordng o Wllamson e al. (2005), Russa has adoped he World Bank mul-pllar model no only because was economcally superor or more fnancally feasble han he alernaves, bu n large par because hs sep gave Russa legmacy n he world socey. The new penson sysem s composed of: 1) a basc benef componen desgned for redsrbuve purposes; 2) a noonal defned conrbuon (henceforh, NDC) publc scheme; 3) a mandaory funded penson pllar. Sarng from 2002, all men born before 1952 and women born before 1956 are covered by he new mul-pllar sysem. The old workers have a relavely shor resduary workng lfe and do no have me o collec enough money for he funded penson. Tha s why ndvduals born before 1952/1956 (men/women respecvely) are no elgble o parcpae n he funded pllar. 3 Many prvleged caegores remaned afer he reform. For example, he war veerans have very hgh pensons n comparson wh he oher pensoners. The compulsory penson nsurance conrbuons were se o be equal o 28%. Half of he penson conrbuons goes o fnance he basc penson and he oher half of he payroll ax s 3 Snce 2005 also ndvduals born before 1967 canno parcpae n he funded pllar (see secon 1.4). 10

25 Chaper 1 recorded n he personal accoun of each wage-earner. Snce 2002, workers born n 1967 and laer have o conrbue 2% of he before-ax wage o he funded pllar, percenage ha gradually wll ncrease o 6% n 2006; workers born beween 1966 and 1957/1963 (men/women respecvely), nsead, have o conrbue 2%. Fgure 1.2 presens he dsrbuon of he conrbuons o he penson sysem afer he 2002 reform. Source: Afanasev (2003) Funded par Funded par 2% 6% Insured par (NDC) Insured par (NDC) Insured par (NDC) 8% 12% 14% Base par Base par Base par 14% 14% 14% Born n 1967and laer Born n /1957 Born n 1952/1956 and earler Fgure 1.2: The 2002 reform: dsrbuon of he conrbuons o he penson sysem Afer he reform, n order o be elgble for he benefs, an ndvdual s requred o conrbue o he penson sysem a leas for fve years. The penson benef s calculaed accordng o he formula shown n Fgure 1.3. The penson rghs of ndvduals who sared workng before 2002 and ha conrbued o he pre-reform penson sysem were convered hrough he so called nal capal. The nal capal s based on he senory and he average wage over he years or over he bes fve years. Acually, he nal capal was almos dsconneced wh he ndvdual s conrbuons. The ndvduals wh nsuffcen years of servce receved only a mnmum nal capal, n spe of he fac ha hey conrbued for dfferen amouns and/or durng dfferen perods. A he end, many ndvduals from ranson cohors, whch parcpaed n boh he pre- and pos-reform sysems, were dsadvanaged by he converson. 11

26 Chaper 1 P=BP+IP/T+FP/T where BP basc par of reremen penson; IP nsured par of reremen penson ( s regsered n ndvdual accoun n he form of noonal defned capal); FP funded par of reremen penson ( s regsered n he specal par of ndvdual accoun); T lfe expecancy a reremen (T was nally se a 144 monhs (12 years) years and s ncreasng o 228 monhs (19 years) over me) Fgure 1.3: The new penson formula A new mechansm of penson ndexaon was also nroduced smplfyng and makng more ransparen he prevous one. A prce ndexaon s adoped for he basc penson. The nsured penson s nsead ndexed o he change n he rao of he oal wages o he oal benefcares bu, n any case, canno grow less han he nflaon rae. Snce, as we have seen, he coverage rao declnes, he NDC componen wll be, a he end, ndexed o prces as well as he basc penson. Noe ha, however, he prce ndexaon s far less han he growh rae of conrbuons and wll reduce he ncenves o conrbue o he sysem. Noe also ha all ndexes and growh raes are deermned by he governmen. I s mporan o sress ha he pensoners can connue o work and o aan full penson benefs smulaneously. The lfe expecancy a reremen (T n fgure 1.3) can be reduced f ndvdual sars recevng penson benefs laer han he sauory reremen age, bu, n any case, T canno be lower han 168 monh (14 years). Therefore, he pos-reform penson sysem, as well as he pre-reform one, does no provde srong ncenves o pospone he clam o he penson benefs. The normal sauory reremen age was no changed by he reform, remanng equal o 55 and 60 years for women and men respecvely. I s well known ha he reremen age s one of he crucal componens of he fnancal susanably of he socal secury sysem. The proposals of an ncrease of he reremen age n Russa have gven rse o anmaed dscussons bu hey had dffcules o be pu no pracce because of several polcal moves, and a low lfe expecancy n Russa s he man jusfcave reason. However, n he Russan Federaon, he lfe expecancy a he age of 60 s no sgnfcanly 12

27 Chaper 1 lower han ha n he oher counres wh ranson economes (see Table 1.3). Neverheless, many Easern European counres have ncreased he reremen age o renforce he fnancal susanably of he sysem. To make hs procedure less panful, some of hem have underaken a gradual ncrease (see Table 1.4). Table 1.3: Lfe expecancy (years) n seleced Easern European counres Men a he age of 60 Women a he age of 60 Russa Hungary Poland Lava Bulgara Source: Rashd e al. (2002) Table 1.4: reremen ages for women and men n seleced Easern European counres Curren Low Men Women Bulgara 2000 Increasng o 63 n 2005 by sx Increasng o 60 n 2009 by sx monhs/year monhs/year Czech Republc 1995 Increasng o (dependng on Increasng o 62 n 2006 by wo number of chldren rased) by four monhs/year monhs/year Esona 1998, n force Increasng o 63 n 2016 by sx monhs/year Hungary 1996 Increasng o 62 n 2001 by one year Increasng o 62 n 2009 by one every second year year every second year Lava 1998 Increasng o 62 n 2003 by sx Increasng o 62 n 2006 by sx monhs/year monhs/year Lhuana 1994, 2000 Increasng o 62.5 n 2003 by sx Increasng o 60 n 2006 by sx monhs/year monhs/year Poland 1998, n force , wh early reremen elmnaed 60, wh early reremen begnnng n 2007* elmnaed begnnng n 2007* Romana 2000 Increasng o 65 n 2015 by one Increasng o 65 n 2015 by one monh/quarer year monh/quarer year Slovaka 1988**` 60 Increasng o (dependng on number of chldren rased) Slovena *** 61 * Elmnaon of early reremen apples o hose covered by he new sysem,.e., hose born afer There wll be excepons for a narrow ls of occupaons, o be specfed n fuure regulaons. These pensons wll be separaely fnanced, no hrough he socal nsurance sysem. ** A new ac s under preparaon by he Mnsry of Labor, Socal Affars, and Famly ha wll ncrease he reremen age gradually o 62 for boh men and women. *** Reremen pror o he age of 63 for men and 61 for women enals penales (hs s a general rule, bu here are excepons for ceran groups of nsured persons). I s also possble o receve bonuses (.e. hgher accrual raes) f he workng perod s greaer han 40 years for men, or 38 years for women. Source: Fulz (2003) In Russa, lfe expecancy a reremen n he new penson formula s equal for all ndvduals n spe of he fac ha, among oher hngs, women lve much longer han men. Rashd e al. (2002) argue ha he nroducon of he acual lfe expecancy n he Russan penson formula and he ncrease of reremen age wll help o rase penson replacemen 13

28 Chaper 1 raes, o mprove he fscal solvency and polcal susanably of he penson sysem, and o reduce he labor marke dsorons caused by he hgh socal secury conrbuon rae. Overall, he man dfference beween he 2002 penson reform and he 1998 reform proposal s he nroducon of a noonal defned conrbuon (NDC) formula. The NDC componen has several posve aspecs: es closely benefs o conrbuons, gves an opporuny o reflec growng longevy; s less vulnerable o corrupon han he funded alernave and avods rsks pecular o he funded scheme (as an nvesmen, managemen rsk), whch are hgh n counres wh a young marke economy. Anoher aracve feaure of he NDC componen s he possbly o use s resources for he curren penson paymens, ha s no possble wh he resources of he funded componen. The major reason why mos of he Easern European and Asa counres (henceforh, ECA) have no urned o he fully funded sysem are he very hgh ranson coss. Insead, hey have nroduced a funded pllar (see Table 1.5). The nroducon of he funded pllar makes workers more responsble of her old age provson, snce he conrbuors can choose an nvesmen manager who wll manage her funded par of penson. As for he funded componen, Lnderman e al. (2000) argue ha should no be lower han 4%-6% of he payroll o be cos effecve and he nal bound of he conrbuons o he funded componen should be beween 7% and 10% of he payroll. Because of he heavy ranson coss, some counres (ncludng Russa) sared wh a level lower han 4%, plannng o ncrease gradually. 14

29 Chaper 1 Table 1.5: Characerscs of Penson Reform n Movng o a Mul-pllar Sysem n he ECA counres Counry and saus of sysem Sarng Dae 1 s pllar Sze of he Funded Pllar (share of payroll) Projeced penson fund asse n 2020 (%GDP) Workforce n funded pllar (2003) Hungary, legslaed and operang January 1998 PAYG DB 6% 31% 45% Swchng Sraegy Mandaory new enrans, Volunary ohers Kazakhsan, legslaed and operang January 1998 Guaraneed mnmum 10% 30% 100% Mandaory for all workers Poland, legslaed and operang January 1999 NDC 7.2% 33% 70% Mandaory < 30, Volunary Lava, legslaed and operang July 2001 (NDC January 1996) NDC 2% growng o 9% 20% 72% Mandaory < 30, Volunary Croaa, legslaed January 2003 PAYG DB 5% 25% - 30% 60%-70% Mandaory <40,Volunary Bulgara, legslaed and operang January 2002 PAYG DB 2% growng o 5% Mandaory < 42 Slovaka, legslaed January 2005 PAYG DB 9% Mandaory new enrans Esona, legslaed July 2002 PAYG DB 6% 20% 60% Volunary (op-ou +2%) Lhuana, operang January 2004 PAYG DB 2.5% Volunary Romana, parally legslaed and hen quesoned January 2003 PAYG DB 8% 30% 75% Mandaory >20 years from reremen Macedona, legslaed January 2005 PAYG DB 7% 26% Mandaory new enrans Ukrane, parally legslaed January 2003 PAYG 2% growng o 7% Mandaory new enrans Russa, parally legslaed and operang Kosovo, parally legslaed and operang Source: Holzmann e al. (2005) January 2002 NDC 2% (<35) o 6% (36-50) Mandaory < 50 January 2002 Mnmum 10% Mandaory 1.4 The 2005 nnovaons In 2005, he Russan Federaon reduced s unfed socal ax (henceforh, UST) from 36.5% o 26%. The purpose of he polcy measure was o ease he ax pressure on he employers and o smulae a rasng of he offcally repored salares and, consequenly, o ncrease he ax proceeds. Gven he ax srucure, he compulsory penson nsurance conrbuons were also reduced from 28% o 20%. Ths sep has led o a subsanal declne n he sae penson budge. In 2005, a par of he Penson Fund s budge defc has been covered by he Sablzaon Fund 4 and he oher by he specal reserves of he Penson Fund. 4 The Sablzaon Fund ncludes a varable surplus from he well-beng ol ndusry. The Sablzaon Fund s accumulaons were desgned, frs of all, for lqudaon of overseas debs. The use of hese resources o plug Penson Fund s defc has been long dscussed and crczed by a number of economss and polcans. 15

30 Chaper 1 I s known ha a shf o a mul-pllar sysem dvers conrbuons away from he pay-asyou-go componen o he funded componen. As a resul, he resources requred o fnance curren pensons are reduced. In order o fnd he addonal recourses o fnance oday s pensons, snce 2005 he conrbuons of workers born before 1967 have been swched from he funded par o he NDC one. Moreover, snce 2005 ndvduals born n 1967 and laer conrbue 4% of he before-ax wages o he funded pllar and 6% o he NDC componen and, sarng from 2008, 6% o he funded pllar and 4% o he NDC scheme. Fgure 1.4 presens he dsrbuon of conrbuons o he penson sysem afer he 2005 changes. Funded par 4% (6%) Insured par (NDC) 10% Insured par (NDC) 6% (4%) Base par Base par 10% 10% Born n 1967 and laer Born n 1966 and earler Fgure 1.4: Dsrbuon of conrbuons o he penson sysem snce 2005 Despe he measures amed a ncreasng he ncenves o conrbue o he penson sysem, apprecable resuls have no been reached ye. The reasons are dfferen. Frs, n many frms employees have sll a weak barganng poson. They have a choce wheher o agree o underrepor her wages (consequenly o reduce conrbuons o he socal secury sysem) or o leave he job. Second, workers ncenves o conrbue are undermned by he dsrus n socal secury, ha s based on bad experence. Moreover, he excluson of he mos acve par of populaon from he parcpaon n he funded pllar n 2005 enforced sll more he 16

31 Chaper 1 unbelef n he penson sysem and, n addon, oday s workers can observe how a perssen hgh nflaon and nadequae ndexaon erode her penson savngs Gender redsrbuon In hs secon, we provde some furher commens abou he redsrbuve feaures, especally concernng men and women, of he Russan socal secury sysem, sarng from some general consderaons abou hs pon. I s well known ha In any sysem provdng annuzed benefs here s a redsrbuve componen comng from he sysemac dfferences n lfe expecancy n he populaon (Damond, 2003). For nsance, hgh ncome ndvduals end o lve longer han low ncome ones and women end o lve longer han men. As for he women s/men s possble redsrbuve conflc, he opnons n he leraure dffer a lo. On he one hand, Schokkaer e al. (2003) rase a queson of wheher men should conrbue less or receve hgher pensons or rere laer han women because of her shorer lfe expecancy. On he oher hand, Gnn (2004) clams ha Lfe expecancy sascs are only gude o he rsk of moraly and many women de younger han men. Gnn also rases doubs on wheher a longer lfe should be consdered as a benef, snce among older people here s a prevalence of dsably and chronc llness. Moreover, Myles (2003) argues ha he redsrbuon of reremen wealh owards women can be consdered as a sharng of he husbands reremen wealh wh wves, and no jus as a subsdy from men o women. Myles also pons ou ha, hsorcally, women s longevy has been a form of care nsurance for men raher han free rdng. Le us now urn o he analyss of he Russan penson sysem, as far as hs redsrbuve conflc s concerned. In Russa, women lve n average 10 years longer han men and are elgble o rere 5 years younger. Noe ha hs sae of affars s very dfferen wh respec o oher European 5 The nflaon rae n he Russan Federaon was equal o 15% n 2002, 12% n 2003, 11.7% n 2004 and 10.9% n 2005 and always has exceeded he rae of he ndexaon of he pensons (daa are aken from he webse of he Hgher School of Economcs, hp:// 17

32 Chaper 1 counres where women s lfe expecancy a brh s 5-7 years hgher han ha of men and, n he majory of hese counres, he reremen age s equal for boh genders (see Table 1.6). Table 1.6. Lfe expecancy a brh and reremen age n seleced counres n 2002 Counry Lfe expecancy a brh Curren reremen age Men Women Men Women Ausra [ncreasng o 65 ( )] Belgum [ncreasng o 65 by 2009] Germany Ialy Norway Russa Span Sweden UK [ncreasng o 65 ( )] Source: Dobromyslov (2002) and U.S. Census Bureau, Global Populaon Profle (2002 A-55) Ths huge dfference n lfe expecances for men and women makes he Russan socal secury sysem clearly redsrbuve owards he laer. In he Russan pre-reform penson sysem, he redsrbuve componen from men o women was very hgh due o boh he penson formula self, whch dd no ncorporae acual lfe expecancy, and he fxed maxmum penson. For example, men could have conrbued more and for longer perods han women, havng also n average shorer lfe spans, bu, a he end, could have aaned he same penson benefs as women. As has been sressed before, he 2002 penson reform has ed more closely benefs o conrbuons. Today, workers ha conrbue less and/or ha conrbue durng shorer perods wll necessarly end up wh lower penson benefs han more acve conrbuors. Under he new penson scheme, he penson benefs of low-wage workers wll replace a lower poron of her pre-reremen wage han would be under he pror scheme (Wllamson e al. (2005)). In parcular, he pos-reform penson sysem s less redsrbuve owards women. More han ha, he new sysem s unfavorable o females no only because hey n average earn less han men, bu also because hey no longer receve penson benefs durng her perods ou of he labor force for chld bearng and chld rearng. Moreover, women wll be lvng longer wh lower pensons han wll men. 6 Accordng o he Russan Sascal Commee (Goskomsa), n 2002, he lfe expecancy a brh n Russa was equal o years for men and years for women (Goskomsa, Russan Sascal Yearbook, 2003). 18

33 Chaper 1 Despe he new penson sysem redsrbues less han he pror one, here s however sll a broad room for redsrbuon. Frs of all, he new penson formula, lke he old one, does no ake no accoun he dfference n he lfe expecances n he populaon. Second, he basc penson desgned for he redsrbuve purpose s fnanced hrough half of he oal socal secury conrbuons and s equal for all ndvduals who conrbued a leas for fve years. Le us assume ha men and women sar her workng lfe a he same age, for example a he age of 22, and ha hey have an unnerruped workng career and sop workng a he sauory reremen age. On he bass of hese assumpons, women wll conrbue for 33 years whle men for 38 years, bu hey wll receve an equal basc penson. In addon, due o he hgher average earnngs, men wll conrbue proporonally for a hgher amoun han women o fnance he basc penson. However, he man redsrbuve componen comes from he gender-neural annues n he suaon of a subsanal dfference beween men and women lfe expecancy. In wha follows, n order o provde a more objecve bass for dscusson and for furher analyses, we employ some socal secures money s worh measures o compare how men and women of he cohors born n 1970, 1975 and 1980 fare under he new penson sysem and how would have been whou he reforms. 1.6 Socal secury money s worh measures The money s worh measures are very useful ools o compare he benefs and he conrbuons of he dfferen ncome groups or cohors under he same socal secury sysem. Ther value s nsead more dubous n makng comparsons across dfferen penson plans snce he compuaon of he money s worh measures requres a number of problemac assumpons and, because of hs, s no possble o accoun properly for all he rsks and/or he ranson coss of he dfferen reform plans (Geanakoplos e al. (1998)). The boom lne s ha s possble o use such ools beng aware of her lmaons. For example, n our case, a comparson of he ndcaors of he un-reformed sysem wh hose of he pos-reform one would no be correc whle we can correcly compare ndcaors of he reformed sysem wh and whou he dscussed 2005 changes. 19