R&D and Economic Growth in Slovenia: A Dynamic General Equilibrium Approach with Endogenous Growth

Transcription

1 Received: 10 March 2010; Acceped: 10 December UDC (497.4) DOI: /PAN V Original scienific paper Miroslav Verbič Faculy of Economics, Universiy of Ljubljana, Slovenia and Insiue for Economic Research, Ljubljana, Slovenia Boris Majcen Insiue for Economic Research, Ljubljana, Slovenia Olga Ivanova TNO Bouw en Ondergrond, Delf, Neherlands Mija Čok Faculy of Economics, Universiy of Ljubljana, Slovenia This research was funded by he Slovenian Research Agency (ARRS) under he conrac No R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh Summary: In he aricle, we model R&D as a major endogenous growh elemen in a small open economy general equilibrium framework and consider several R&D policy scenarios for Slovenia. Increase of he share of secoral invesmen in R&D ha is deducible from he corporae income ax and increase of governmen spending on R&D urned ou o be he mos effecive suggesed policy measures. While he former policy measure is sill followed in par by an undesired ransfer of he ax relief o dividends, a moderae increase of governmen spending on R&D booss long-run produciviy in he economy, hus increasing he fuure value of firms, which is refleced in a desired dividend increase. The households ha would gain more uiliy from such policy scenarios are hose wih more skilled and highly skilled labour, bu no he very op earners in he economy. Key words: Endogenous growh, General equilibrium modelling, R&D, Slovenia. JEL: C68, D58, O38, O40. The opic of economic growh is among he mos essenial issues in macroeconomics, as i direcly affecs he living sandard of he populaion and he welfare level. As a resul, he search for fundamenal deerminans behind he growh process is an ongoing research heme. There are broadly speaking wo dominan heories; he neoclassical growh approach and he endogenous growh approach (cf. Rober J. Barro and Xavier Sala-i-Marin 2003). Neoclassical growh models assume ha produciviy growh is exogenous. This view has changed in he early 1980s. According o he new growh heory, he long-run economic growh is affeced by deliberae economic behaviour and human acions. Economiss agree ha he long-run growh poenial in per capia growh is deermined by advances in produciviy. Producion can o cerain exen be expanded exensively hrough invesmen in facor inpus and hrough employmen growh, bu in he long run inensive improvemens in produciviy are required. By working in a more efficien way, more can be produced wih given facor inpus. The major deerminans of he produciviy growh are invesmens in educaion and hus human capial, and in research and developmen (R&D). Along hese lines, he he-

2 68 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok ory of economic growh can be undersood o have urned ino he heory of produciviy growh. On he oher hand, he endogenous growh heory can be comprehended as an aemp o explain wha can no be explained in he Rober M. Solow (1956) model, i.e. he rae of growh of oal facor produciviy or he so called Solow residual 1. While hese ideas have been esed in a number of empirical sudies, hey are sruggling o find heir way ino general equilibrium modelling, which has led o a good deal of criicism. As Chrisian Ghiglino (2002) poined ou, endogenous growh heory has had some success in explaining he observed daa relaed o he process of economic growh, bu he resuls of he models are ypically very sensiive o heir microeconomic srucure. Therefore, valuable insighs can be gained by inegraing endogenous growh heory ino he framework of general equilibrium heory. The moivaion behind our work is o consruc and develop a dynamic general equilibrium model wih endogenous growh, driven by invesmen in educaion and R&D, which will enable us o analyze he impac of hese deerminans on economic growh in he conex of complex muual aciviy of economic agens ha is aking place in heir socio-economic environmen. Our conribuion o he exising model lieraure is a focus on a small open economy case of Slovenia, where a large par of he echnological change comes from abroad. In his aricle, we focus on inegraing R&D as a major endogenous growh elemen ino an iner-emporal general equilibrium framework, which has no been done before in he predominanly exogenousbased lieraure on economic growh in Slovenia 2. R&D can be inroduced eiher as a separae producion facor, or hrough is impac on oal produciviy. R&D aciviy resuls in new goods, new ideas and new knowledge, which are non-rival (cf. Paul M. Romer 1990; Gene M. Grossman and Elhanan Helpman 1991; Philippe Aghion and Peer Howi 1992). As such, i is a major source of susained economic growh, boh in erms of innovaion and hrough adopion of exising echnologies. Namely, he empirical sudies sressing a srong and enduring link beween R&D capial and oupu growh roughly sugges ha a one per cen increase in he R&D capial sock is found o lead o a rise in oupu of beween per cen (Zvi Griliches 1992; David T. Coe and Elhanan Helpman 1995; Jeffrey I. Bernsein 1998; Seven Globerman 2000). The poenial gains from improved efficiency can also be imporan, as innovaion is increased. However, hese relaionships should also be dependen on he qualiy of insiuions and is regulaion (cf. Maja Klun and Renaa Slabe-Erker 2009). Noneheless, while increases in human capial, R&D, and produc marke compeiion can improve macroeconomic performance, i akes ime before hese benefis are realised. On he oher hand, invesmens in R&D are subopimal as he privae and social raes of reurns diverge (Griliches 1988). This is because, in is R&D invesmen 1 There is vas lieraure available on measuring oal facor produciviy. Thus, one could only poin ou here some comprehensive lieraure reviews on his opic, e.g. M. Ishaq Nadiri (1970), T. K. Rymes (1983), Renuka Mahadevan (2003), Richard G. Lipsey and Kenneh I. Carlaw (2004), and Tao Kong (2007). 2 Readers are invied o consul Boris Majcen e al. (2005), Miroslav Verbič e al. (2006), Verbič (2007), Mija Čok e al. (2009), Verbič (2009), and Majcen e al. (2009) in order o obain an insigh ino he developmen of general equilibrium modelling of he Slovenian economy.

3 R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh 69 decision, a privae invesor does no accoun for he losses i imposes on previous innovaors, as well as he iner-emporal knowledge spillovers and consumer surplus i creaes. Exising lieraure suggess ha in he absence of axes and subsidies, he decenralized economy underinvess in R&D wih he primary impeus coming from he surplus appropriabiliy problem, arising from monopoly pricing of R&D oupu 3 (cf. Madanmohan Ghosh 2007). Thus, governmens in many counries provide various incenives o simulae R&D aciviies. Since mos benefis of R&D are inangible, hey are difficul o esimae economerically and here is lile empirical work ha evaluaes he relaive meris of hese incenives in a sysemaic way, such as using a dynamic general equilibrium framework. The ouline of he aricle is as follows. In Chaper 1 a curren lieraure overview on R&D-driven endogenous growh models is presened. In Chaper 2 a broad descripion of he general equilibrium model of he Slovenian economy is provided, while in Chaper 3 we show in deail how he R&D secor is modelled in a dynamic general equilibrium framework. In Chaper 4 he scenarios are described and he resuls of simulaions are presened, where we focus in paricular on macroeconomic and welfare aspecs. In he final chaper we summarize he main findings of he aricle. 1. Lieraure Overview on R&D-driven Endogenous Growh Models The endogenous growh lieraure capures he insigh ha he crucial force behind posiive growh raes is he eliminaion of he endency of diminishing reurns o invesmen in a broad class of capial goods, including R&D. Anecedens of his lieraure uilize heories of echnological progress, innovaion and imiaion (Romer 1987; Grossman and Helpman 1991), learning by doing (Nancy L. Sokey 1991), and populaion change, feriliy and human capial invesmen (Gary S. Becker and Barro 1988) in order o inroduce increasing or consan reurns o scale o he cumulaive facor of producion. Recen advances in he new growh heory idenify, among many ohers, he invesmen in R&D as a crucial deerminan of he long-run rae of economic growh (cf. Maria J. Alvarez-Pelaez and Chrisian Groh 2005; Ghosh 2007). A he beginning of 1990s, a firs generaion of endogenous R&D growh models appeared. The produc-variey model by Romer (1990) and he qualiy-ladder model by Aghion and Howi (1992) are pioneer models and he saring poin for many laer conribuions. The idea was o assume ha in order o se up producion, a firm mus incur a fixed cos in he form of an oulay on he final good; an oulay, which may be hough of as he R&D cos of developing is paricular variey. The major policy quesions of ha ime have been how an economy can susain a posiive growh rae and how innovaion policy can enhance addiional growh and welfare. In Romer s (1990) model, growh was driven by echnological change ha arises from inenional invesmen decisions, made by profi-maximizing agens. The 3 However, in some circumsances privaely financed R&D may be oo high as well (cf. Aghion and Howi 1992; Charles I. Jones and John C. Williams 2000).

4 70 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok disinguishing feaure of he echnology as an inpu was ha i is no a convenional or a public good; i is a non-rival, parially excludable good. Because of he nonconvexiy inroduced by a non-rival good, price-aking compeiion canno be suppored. Insead, he equilibrium is one wih monopolisic compeiion. The main conclusions were ha he sock of human capial deermines he rae of growh, ha oo lile human capial is devoed o research in equilibrium, ha inegraion ino he world markes will increase growh raes, and ha a large populaion is no sufficien o generae growh. Aghion and Howi (1992) developed a model of endogenous growh, in which growh is driven by verical innovaions ha involve creaive desrucion. Equilibrium was deermined by a forward-looking difference equaion, according o which he amoun of research in any period depends negaively upon he amoun expeced nex period. They analyzed posiive and normaive properies of saionary equilibria, and showed condiions for he exisence of cyclical equilibria and nogrowh raps. They demonsraed ha he growh rae may be more or less han opimal because a business-sealing effec couneracs he usual spillover and appropriabiliy effecs. In addiion, innovaions end o be oo small. Lakshmi K. Rau and Thirukodikaval N. Srinivasan (1994) employed a model wih endogenous populaion growh. According o his approach, he growh of populaion may induce echnological change when cerain resources, such as land, are fixed. As Eser Boserup (1981) and Julian L. Simon (1981; 1996) argued, populaion pressure gives agens incenives o develop new echniques of producion. In ha case, one would expec he resuling innovaions o bring abou changes in he echnology, defined as he se of all efficien echniques, such ha eiher he shape or he numbering of he conours ha make up he isoquan map will depend on he size of he populaion. Jones (1995) criicised R&D growh models wih expanding variey or growing qualiy of inermediae inpus ha have a scale effec of R&D employmen on he produciviy growh. His criicism was based on he ground ha he predicion of such models is widely a variance wih he facs of R&D employmen and produciviy growh in he advanced counries over he las fify years. He suggesed a model, which shares imporan feaures wih Kenneh J. Arrow s (1962) seminal paper on learning by doing. He believed ha growh per se is no endogenous, alhough if populaion is growing, per capia oupu may persisenly increase because of purposeful research effor due o increasing reurns o scale in he oupu secor. While he aggregae models of Romer (1990) and descendans (Jones and Williams 2000; Alvarez-Pelaez and Groh 2005; Thomas M. Seger 2005) reaed economies as closed, in a small open economy framework a large par of he echnological change comes from abroad. These models ake ino accoun ha echnological change resuls from profi-maximising R&D firms oupu of paens ha are purchased by capial producers in order o supply new varieies of capial equipmen. However, since he access o global knowledge is limied by is level of echnical sophisicaion and capaciy o absorb, domesic R&D is imporan boh for is direc impac on long-run economic growh and enhancing absorpive capaciy. Gerhard Glomm and B. Ravikumar (1994) modelled an infinie-horizon economy where he

5 R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh 71 sock of echnological knowledge enhances he produciviy of all households and echnological knowledge depends upon public secor invesmen in R&D. They esablished ha public policy affecs he growh rae of per capia income, bu has no impac on income inequaliy. More recenly, a second generaion of endogenous R&D growh models has appeared, in which he scale effec is eliminaed and he simulaneous expansion of inermediae goods variey and qualiy occurs under condiions ha make seadysae produciviy growh dependan on he raio beween inensive R&D employmen and oal employmen (cf. Alwyn Young 1998; Piero F. Pereo 1998; Howi 1999). Maria J. Freire-Serén (2001) inroduced ino an R&D-model a echnology of innovaion, based on expendiure ha generaes endogenous susainable growh in absence of any scale effec. Daisuke Ikazaki (2006) consruced an endogenous growh model ha incorporaes he R&D secor, he educaion secor, and environmenal issues. He suggesed, similar o Jones (1995), ha he populaion level does no affec he economic growh rae. Bria Bye e al. (2007) explored how innovaion incenives in a small, open economy should be designed in order o achieve he highes welfare and growh by means of R&D-driven endogenous echnological change, embodied in varieies of capial. Ghosh (2007) uilized a general equilibrium R&D model of endogenous growh via increasing capial variey, similar o he one of Xinshen Diao e al. (1999), in order o examine he impac of alernaive policies on produciviy and economic growh. His findings reveal ha direc incenives, such as subsidies o R&D aciviies, would have he highes produciviy impac, ha an increase in subsidies o he users of R&D capial would have a posiive bu smaller impac, and ha rade liberalizaion would have minimal effecs on produciviy growh via is impac on inernaional R&D spillovers. 2. Descripion of he General Equilibrium Model of he Slovenian Economy The model SIDYN 2.0 is a dynamic endogenous-growh general equilibrium model of he Slovenian economy, based on social accouning marix (SAM) for he base year, and parameer daa on consumer preferences, producion echnologies, accumulaion of human capial and composiion of oal facor produciviy 4. The model incorporaes he following economic agens: (1) five households grouped ino quiniles according o income level; (2) weny producion secors of boh goods and services; (3) invesmen secor; (4) naional governmen; and (5) exernal secor. Each agen in he economy supplies and demands a range of goods, services and facors of producion a prices defined by equilibrium on he corresponding markes. There are six ypes of producion facors in he model; counrylevel human capial differeniaed by hree skill (educaion) levels, secor-specific physical capial, secor-specific R&D sock, and secor-specific human capial sock. 4 This chaper provides a non-echnical descripion of our general equilibrium model (excep for modelling of R&D, which is provided in Chaper 4). For a echnical descripion of he model he reader is invied o consul Verbič e al. (2009) or Majcen e al. (2011).

6 72 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok The oupu level of he aggregaed commodiy of each of he weny secors is deermined by an opimal combinaion of hese producion facors. Boh households and firms make heir decisions under he assumpion of an infinie horizon wih perfec foresigh (raional forward-looking expecaions). All prices in he model are relaive prices, which is he usual assumpion of general equilibrium modelling. The iner-emporal problem is formulaed in discree ime for he purpose of numerical implemenaion. To keep he derivaion and calibraion simple, all ransacions are assumed o ake place a he end of each period, while decisions are made or planned a he beginning of each period. Households maximize heir iner-emporal uiliy given he budge consrain. They decide how much ime and money o inves ino a paricular ype of human capial a each period of ime. The consumers decisions associaed wih spending of heir money and ime are independen of each oher. The use of money posiively influences consumer uiliy via an increase in consumpion, whereas he use of ime for educaion and work exers a negaive influence upon he consumers uiliy level. Households do no inves in he secor-specific socks. Differen income caegories in he economy correspond o differen consumpion paerns and governmenal ransfers. Each ype of labour is supplied by he households and he res of he world. Afer domesic households decide upon he share of is labour endowmen spen on work and educaion, he res of he labour endowmen is labelled as leisure aciviy and represens he level of volunary unemploymen in he economy. I is assumed ha he labour flow from abroad does no choose o say volunarily unemployed. The human capial sock owned by he households is freely raded on he labour marke and is mobile beween he secors. Secor-specific human capial is no mobile beween he secors and hus no raded. I represens secor-specific knowledge accumulaed wihin he company, such as experience, repuaion and conacs. The firms choose invesmens ino heir physical capial, human capial and R&D socks, as well as labour inpus such as o maximize is presen discouned value. Invesmens made by he firms are financed using he oal savings available in he economy. The socks of secor-specific capial are accumulaed over ime via he new invesmens made by firms and he governmen. Gross prices for final goods are calculaed as he sum of he producer price, ranspor and rade margins and various axes and subsidies, where he ranspor and rade margins are he spending on ranspor and rade services, consumed in a cerain proporion o he commodiy iself. Invesmens ino physical capial are financed by he naional invesmen agen wih oal savings and are used o buy differen capial goods. The spli of he oal physical capial invesmen beween he paricular ypes of capial goods, such as machinery and buildings, is done so as o maximise he uiliy of he invesmen agen, which decides on how much of various capial goods are o be bough. The public secor is represened by a naional-level governmen, which collecs a range of axes, receives is share of dividends, and pays subsidies and ransfers o households and firms, as well as ransfers abroad. The revenues of he governmen consis of receips from personal income ax, corporae income ax, VAT, payroll ax, social conribuions, and impor ariffs. The governmen subsidies suppor inves-

7 R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh 73 men, producion, inermediae consumpion, household consumpion, and exporing. The governmen also consumes a range of goods and services, and invess in naional-level human capial and secor-specific R&D socks. The exernal secor incorporaes he represenaion of expors and impors, as well as annual labour inflows from EU15, new member saes and he res of he world. Modelling of he exernal secor is based on he assumpion of a small open economy, meaning ha he prices of expors and impors are exogenously fixed in he model. Expors and impors are defined by Slovenian oupu and income levels, as well as by he raio beween he prices of domesic and expored goods and services, and elasiciies of subsiuion beween domesic and foreign goods. Inflows of labour o Slovenia are defined by he changes in domesic real afer-ax wages and he elasiciy of he labour supply, which is assumed o be higher for he res of he world han for he EU15 and he new member saes. The model is build wihin he general algebraic modelling sysem (GAMS), which has become boh mos widely used programming language and mos widespread compuer sofware for consrucion and solving large and complex general equilibrium models. Wihin he GAMS framework, he dynamic general equilibrium model is wrien in Lars Mahiesen s (1985) formulaion of he Arrow and Gerard Debreu (1954) equilibrium model, i.e. as a mixed complemenariy problem (MCP). The key advanage of his formulaion is he compac presenaion of he general equilibrium problem, which is achieved by reaing variables implicily and hus significanly reducing he compuaion ime for higher-dimensional problems. To solve he model, i.e. o achieve convergence, a recen version of he PATH solver (Michael C. Ferris and Todd S. Munson 2000) is used, which is renowned for is compuaional efficiency. In conras o simpler models, such a large-scale model enables one o consider simulaneous changes in a variey of policy insrumens and provides ways o undersand shor-o-medium run responses by making i possible o observe he ransiion pahs of he modelled economy from one seady sae o possible-oher. Wih assumpions of longer ime-spans on he par of each agen, such a model provides a more realisic seup ha poins o he income disribuion effecs of permanen policy changes. 3. Modelling R&D in a Dynamic CGE Framework Economic growh in he model is endogenously deermined by he developmen of households human capial sock, sock of secor-specific human capial and R&D, used as facors of producion by he firms, as well as he developmen over ime of he overall oal facor produciviy (TFP). Wih he higher levels of R&D sock one needs o inves more; he higher is echnology developmen he more has o be invesed in R&D in order o obain new innovaions and o furher increase he R&D sock. The developmen of overall TFP is described by he following regression equaion and depends posiively upon he share of naionally produced R&D in he GDP and he share of foreign rade in he GDP (cf. Erik Canon e al. 2005):

8 74 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok XDrd, MX logtfp fpr erd log em log GDP sec GDP (1) where rd is a subse of commodiy ypes, consising of R&D services; fpr is he residual of he oal facor produciviy regression; e rd is he parial elasiciy of he oal facor produciviy wih respec o he share of R&D in he GDP; e M is he parial elasiciy of he oal facor produciviy wih respec o he share of foreign rade in he GDP; TFP is he oal facor produciviy level in he economy; GDP is he gross domesic produc; XD is he oal oupu of he domesic secor sec; and MX is he mean of oal impors and expors of commodiies. The laer is defined as: 1 MX sec, IMEU15sec, IMEU9sec, IMROWsec, 2 EXEU15 EXEU9 EXROW, (2) IMEU15 is he impor of commodiies from EU15; EXEU15, where is he expor of commodiies o EU15; IMEU9 is he impor of commodiies from he new EU member saes; EXEU9 is he expor of commodiies o he new EU member saes; IMROW is he impor of goods from he res of he world (ROW); and EXROW is he expor of goods o he res of he world. Expression (1) has been adoped in he curren version of our general equilibrium model in order o accoun for he domesic and inernaional spillover effecs of he R&D aciviy. The oal facor produciviy, represened by he above regression equaion is hen applied o all producion secors of he economy hrough he secoral producion funcion XD. The firms in he model are faced wih he iner-emporal profi maximizaion problem and wih he formulaion of heir invesmen decisions, relaed o R&D (cf. Sephen P. Cassou and Kevin J. Lansing 2004). In each period of ime, he firms produce one commodiy by secor, using physical capial, labour, secor-specific human capial and secor-specific R&D sock as inpus. I is assumed ha he firms operae under he following consan reurns o scale Cobb-Douglas echnology funcion: sec XD TFPaF K L HCS RDS Fsec FLsec FHCsec FRDsec sec, sec sec, sec, sec, sec, (3) where af sec is he scale parameer of he Cobb-Douglas producion funcion; K is he inpu of physical capial; L is he inpu of labour; HCS is he

9 R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh 75 inpu of secor-specific human capial sock; RDS is he inpu of secor-specific R&D sock; Fsec is he share parameer of he producion funcion, associaed wih FLsec physical capial; is he share parameer of he producion funcion, associaed wih labour; FHCsec is he share parameer of he producion funcion, associaed wih human capial sock; and FRD is he share parameer of he producion funcion, associaed wih R&D sock. The sock of R&D knowledge, used as inpu by firms each period of ime, is deermined according o he following law of moion: 1 RDs RDssec sec sec sec RDS ARDS RDS RD RDG (4) where ARDS sec is he scaling parameer of he R&D sock accumulaion funcion; RDssec is he share parameer of he R&D sock accumulaion funcion, associaed wih he new R&D invesmen 5 ; RD are he R&D services bough by he secor sec for invesmen in is R&D sock; and RDG are he R&D services bough by he governmen for invesmen in he R&D sock of secor sec. The firms choose invesmens in heir physical capial, human capial and R&D knowledge socks a each ime period, as well as labour inpus such as o maximize he presen discouned value of he firm, VF. Given ha he model has a finie simulaion ime horizon, {1,, T}, where T is he las simulaed ime period 6, we obain he following expression for he value of he firm: T T r VFsec DIV DIV T 1 1r 1r rg (5) where r is he seady sae ineres rae; g is he seady sae growh rae; and DIV are he dividends paid by secor sec. I is assumed ha he dividend paymens are equal o he value of he commodiy, reduced by he labour coss and he invesmens made by firms in physical capial, human capial and R&D sock. This leads o he following implici expression for he firms dividends, DIV : 5 Noe ha he share parameers of he laws of moion, associaed wih new invesmens, are inerpreed as depreciaion raes for differen ypes of capial used as inpu by he firms. 6 Afer ime period T i is assumed ha he economy will be on he seady-sae pah where all real economic variables grow wih he same annual rae unil infiniy.

10 76 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok spv spv xdv PD XD kv DIV I PK sec, sec, sec, sec, sec, sec, sec, sec, lev lev lv payrv PL L seccsechr sec, sec, sec, sec, sec, sec, io XD P micv PTM sicv secc, sec 1 secc, 1 secc, 1 1 secc, 1 1 sicv icv vaicv exsicv secc, 1 secc, 1 secc, 1 secc, 1,,, 1,, 1sicvedus, icv vaicv exsicv edus, edus, edus, rd, rd, 1,, 1 sicv icv vaicv exsicv, HC P micv PTM sicv kv sharehcv sec edus edus edus sec RD P micv PTM sicv kv sharerdv rd sec rd, rd, rd, rd, (6) where secc is an alias of secor sec; edus is a subse of commodiy ypes, consising of educaion services; sechr is a subse of commodiy ypes, consising of educaion and R&D services; spv is he oupu subsidy rae; xdv is he oupu ax rae; kv is he corporae income ax rae; lev is he employees social conribuion rae; lv is he employers social conribuion rae; payrv is he payroll ax rae; I is he privae demand for invesmen goods; PK is he reurn o capial; PL is he wage; PTM is he composie price of rade and ranspor margin; PD is he domesic producer price; P is he domesic sales price; HC is he human capial spending; io secc, sec is an inpu-oupu coefficien of commodiy secc used for producion in he secor sec; micv secc, is he inermediae consumpion rade and ranspor margin; sicv secc, is he inermediae consumpion subsidy rae; icv secc, is he inermediae consumpion ax rae; vaicv secc, is he inermediae consumpion VAT rae; exsicv secc, is he inermediae consumpion excise ax rae; sharehcv is he share of secoral invesmen in human capial sock, deducible from he corporae income ax; and sharerdv is he share of secoral invesmen in R&D, deducible from he corporae income ax. Invesmens made by he firms are financed using he oal savings available in he economy, i.e. he savings of households, he governmen, reained profis of he firms and he savings from abroad. Invesmen level in he secor-specific R&D is hus chosen such ha he firms discouned profis resuling from hese invesmens are equal o he coss of he invesmens. By maximizing he expression for he value of he firm (5) subjec o he firms oupu echnology (3) and he law of moion (4), one obains, afer some simplificaion, he respecive firs-order condiion:

11 R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh sicv icv vaicv exsicv RD P micv PTM sicv kv sharerdv rd, rd, rd, sec, rd, rd, rd, rd, 1 1 spv (1 spv ) xdv sec, 1 1 sec, 1 PD XD r seccsechr io XD P micv PTM secc, sec sec, 1 secc, 1 secc, 1 1 secc, 1 secc, 1 secc, 1 secc, 1 secc, 1 sec sec 1sicv 1 sicv icv vaicv exsicv FRD RDs. (7) Lef hand side of expression (7) represens he oal cos of he invesmen in R&D, and is equal o he righ hand side, which represens he addiional discouned dividends of he firms, resuling from he invesmen in R&D. The value of addiional dividends depends posiively upon he Cobb-Douglas share of he R&D sock in he producion funcion, afrd sec, and he Cobb-Douglas share ha represens he conribuion of new R&D invesmen o he oal sock of secor-specific R&D, RDs sec. In he governmen secor we model explicily he ax revenues, he governmen subsidies and he governmen consumpion of goods and services. The ax revenues of he governmen, associaed wih R&D, consis of he following expression: 1 sicv icv vaicv exsicv P micv PTM RD rd, rd, rd, rd, rd, rd, secc, secc (8) form: while he governmen subsidies, associaed wih R&D, have he following sicv kv sharerdv P micv PTM RD rd, rd, rd, sec (9) Finally, he equilibrium in he marke for R&D services is deermined by he following condiion: RD RDG I sec, rd, io XD secc, sec SVX TMX X rd, rd, rd, (10) sec sec seccsechr where SVX are he changes in socks of secor sec; TMX is he consumpion of secor sec for ranspor and rade margins; and, sec X are domesic sales of he good of secor sec of domesic and foreign origin. Expression (10) hus reveals ha he sum of firms consumpion of R&D services, governmen consumpion of R&D services, consumpion of he invesmen agen on R&D services, inermediae inpus in he producion of R&D services, changes in he socks of R&D services and

12 78 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok ranspor and rade margins on R&D services is equal o he oal domesic sales of R&D services. 4. Resuls of he Simulaions The groundwork for our analysis is he dynamic calibraion of he model and consequenly preparaion of he reference soluion. In he framework of performing he dynamic calibraion of he model SIDYN 2.0, we follow he sraegy of using he model o generae he enire dynamic pah of endogenous variables in order o accuraely reproduce he values of every endogenous variable in he base year. In his way we obain he reference scenario, which represens he auhenic sae of he economy. Analysis of he Slovenian economy, where we ake ino accoun possible developmens wih respec o he R&D and oher developmen policies, is hen performed by forming counerfacual scenarios and comparing heir oucomes o he resuls of he reference scenario. The counerfacual scenarios are based on varying he parameers of SIDYN 2.0, which are relaed o he modelling of R&D services. For he purpose of our analysis we disinguish beween model parameers and policy parameers; he former are of echnical naure and subjec o sensiiviy analysis, while he laer are of economic naure and subjec o policy analysis. In his aricle, we focus on he laer. Policy parameers, involved in modelling R&D decisions in a dynamic CGE framework, which are used in SIDYN 2.0, are he following: (1) he corporae income ax (CIT) rae in secor kv ; (2) he share of secoral invesmen in R&D ha is deducible from he CIT, sharerdv ; and (3) governmen spending on R&D in secor RDG. There are also several policy parameers, involved direcly in he producion of R&D, which will no be examined here 7. Table 1 Scenarios for he Analysis of Impacs of R&D on Economic Growh Scenario Descripion of he scenario SC1 Decrease of he corporae income ax rae, kv = kv, by 25% from 2009 SC2 SC3 SC4 Increase of he share of secoral invesmen in R&D ha is deducible from he corporae income ax, sharerdv, by 25% from 2009 Increase of governmen spending on R&D, RDG = RDG, by 10% per annum from 2009 Increase of governmen spending on R&D, RDG = RDG, by 20% per annum from 2009 Source: Auhors simulaions using SIDYN These secor-specific policy parameers are: (1) he inermediae consumpion ax rae, icv rd, ; (2) he inermediae consumpion subsidy rae, sicv rd, ; (3) he inermediae consumpion VAT rae, vaicv rd, ; (4) he inermediae consumpion excise ax rae, exsicv rd, ; and (5) he inermediae consumpion rade and ranspor margin, micv rd,.

13 R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh 79 Descripion of he scenarios examined in his aricle is given in Table 1. We have chosen he scenarios in such a way ha hey reflec a combinaion of policy measures discussed in he Slovenian public debae and hose proposed by Majcen e al. (2011). Le us firs examine he effecs of hese policy scenarios on R&D expendiure in Slovenia. As can be seen from Figure 1, scenario SC2 has a major effec on R&D expendiure of firms, while he effec of he remaining hree scenarios is less profound, as i does no provide enough incenive o change he behaviour of firms. Namely, decreasing he CIT rae by 25% (SC1) increases on average he R&D expendiure of firms by modes 0.45% wih respec o he reference scenario, while he impac of increasing he share of secoral invesmen in R&D ha is deducible from he CIT by 25% (SC2) amouns o %. I is hus raional for he firms o wihheld realised profis for invesmen in R&D in order o increase fuure profis. The effec of increasing governmen spending on R&D wih respec o he reference scenario is moderae; an increase of 10% per annum (SC3) increases he R&D expendiure of firms up o 1.3%, while an increase of 20% per annum (SC4) gives rise o an R&D expendiure increase of up o 2.6%. 8.0% Change wih respec o he reference scenario 7.0% 6.0% 5.0% 4.0% 3.0% 2.0% 1.0% SC1 SC3 SC2 SC4 0.0% Figure 1 R&D Expendiure of Firms in Slovenia, Source: Auhors simulaions using SIDYN 2.0. Wha is he raionale behind he dynamics presened in Figure 1 and how does i affec he res of he economy? Firs, one needs o noe ha we only observe in Figure 1 he R&D expendiure change of firms wih respec o he reference scenario, i.e. he response of he firms o economic policy, and no he oal R&D expendiure change. Producion of R&D services (no shown), which represens he oal supply of R&D services in he economy, i.e. funded by he firms and he governmen, increases on average by 0.33% in scenario SC1, by up o 4.1% in scenarios SC2 and SC3, and by up o 7.9% in scenario SC4 wih respec o he reference scenario. The Year

14 80 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok laer scenario herefore demonsraes he dynamics of increasing he R&D expendiure on he Slovenian economy mos disincly. In addiion, invesmen in R&D services be i by firms or he governmen induces an accompanying invesmen in educaion services. R&D services and educaion services are o a cerain exen complemenary facors of producion, which is necessary in order o consume addiional R&D services efficienly in he producion process. This phenomenon is obvious by looking a he effecs of he policy scenarios on educaion expendiure of households (Figure 2) and on human capial expendiure of households and firms (no shown) in Slovenia. 3.0% Change wih respec o he reference scenario 2.0% 1.0% SC1 SC3 SC2 SC4 0.0% Figure 2 Educaion Expendiure of Households in Slovenia, Source: Auhors simulaions using SIDYN 2.0. Namely, scenario SC1 induces on average an increase of educaion expendiure of households wih respec o he reference scenario of 0.3%, scenario SC2 of up o 1.4%, scenario SC3 of up o 1.3%, and scenario SC4 of up o 2.5% (see Figure 2). Human capial expendiure change deviaes from he educaion expendiure change (no shown) only in scenario SC1, where he human capial expendiure increases on average wih respec o he reference scenario by 0.45%. This indicaes ha only scenario SC1 induces significan firm-funded human capial invesmen, while addiional human capial change, induced by he oher hree policy scenarios, originaes mainly from household-funded educaion expendiure invesmen. Because of addiional invesmen in R&D and he induced invesmen in human capial, he labour demand of firms in he economy increases and he labour supply of households adjuss (Figure 3). The adjusmen is again leas disinc in case of scenario SC1 (up o 0.2% increase) and mos eviden in case of scenario SC4 (up o 1.7% increase wih respec o he reference scenario). Year

15 R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh % Change wih respec o he reference scenario 1.5% 1.0% 0.5% SC1 SC3 SC2 SC4 0.0% Figure 3 Labour Supply of Households in Slovenia, Source: Auhors simulaions using SIDYN 2.0. Le us now examine in more deail he effec of increasing governmen spending on R&D by 20% per annum (SC4) on household invesmen in educaion. Figure 4 presens he dynamic of his effec by labour ype. I urns ou ha households wih unskilled labour increase is spending on educaion he leas, while households wih highly skilled labour increase is spending on educaion he mos. This follows naurally by: (1) keeping in mind he naure of R&D invesmen; and (2) observing he dynamic of household income (no shown, bu follows closely he dynamic of GDP in Figure 7). Namely, R&D invesmen a key driver of economic growh yields new produc innovaions and adds o he knowledge base of indusry and he markeplace as a whole. As such, addiional R&D expendiure employs more highly skilled labour, bu also needs he suppor of skilled and unskilled labour in order o ransform R&D invesmen ino higher long-erm growh. As can be inferred from Figure 4, implemening scenario SC4 employs on average 0.44 percenage poins more highly skilled labour in comparison o skilled labour, and 0.64 percenage poins more highly skilled labour in comparison o unskilled labour. Addiionally, by wihdrawing labour from he producion process he households lose income. Households wih unskilled labour, which are on average also lower-income households, use higher share of heir income for exisenial needs and are less able o inves is own funds in educaion. On he oher hand, more invesmen in educaion reduces curren labour supply. Figure 5, which presens he dynamic of labour supply by labour ype, reveals his phenomenon. Labour ype depends primarily on years of schooling. Invesmen in educaion of unskilled and skilled labour causes no iniial decrease in labour supply wih respec o he reference scenario, bu leads o lower long-erm growh change. Conversely, invesmen in educaion of highly skilled labour pulls he (poenial) labour force ou of he producion process for a longer period of ime, bu leads o highes long-erm growh change. Year

16 82 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok 4.0% Change wih respec o he reference scenario 3.0% 2.0% 1.0% Unskilled labour Skilled labour Highly skilled labour 0.0% Source: Auhors simulaions using SIDYN 2.0. Figure 4 Educaion Expendiure of Households in Slovenia in Case of Increasing Governmen Spending on R&D by 20%, by Labour Type, Year 2.0% Change wih respec o he reference scenario 1.5% 1.0% 0.5% Unskilled labour Skilled labour Highly skilled labour 0.0% Year -0.5% Source: Auhors simulaions using SIDYN 2.0. Figure 5 Labour Supply of Households in Slovenia in Case of Increasing Governmen Spending on R&D by 20%, by Labour Type,

17 R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh 83 Invesmen in R&D (Figure 1) and he induced invesmen in educaion and hus human capial (see Figures 2 and 4) are followed also by addiional capial inpu (no shown). This is mos eviden in scenario SC4, where capial inpu increases wih respec o he reference scenario by 3.75% in he long run, bu also in scenarios SC2 and SC3, where we observe an increase of up o 1.5% and 1.4%, respecively. Capial inpu change is even more manifes in he R&D services secor, where i amouns up o 4.1% in scenario SC3, up o 4.2% in scenario SC2, and up o 8.1% in scenario SC4 wih respec o he reference scenario. Diverse policy measures obviously affec differen incomes o a differen exen, which is mos disinc in he case of real dividends (Figure 6). Namely, real dividends increase wih respec o he reference scenario on average by 4.6% in scenario SC1, by up o 3.4% in scenario SC2, by up o 1.4% in scenario SC3, and by up o 2.7% in scenario SC4. Obviously, decreasing he CIT rae (scenario SC1) no only provides more funds in he profi opimizaion process for invesmen, bu also leaves he firm wih more profi for sharing, which appears o be an aracive opion. On he oher hand, by increasing he share of secoral invesmen in R&D ha is deducible from he CIT (scenario SC2), i becomes raional for he firm o redisribue some profi from sharing o invesing in order o increase fuure profis. In hese wo scenarios, he dividend increase is sill primarily a direc effec (an undesired effec) of he policy measures. Increasing governmen spending on R&D (scenarios SC3 and SC4) booss long-run produciviy in he economy, hus increasing fuure profis and he fuure value of firms, which is refleced in he dividend increase. The laer hus becomes a side effec (a desired effec) of he policy measures. 5.0% Change wih respec o he reference scenario (in percen) 4.0% 3.0% 2.0% 1.0% SC1 SC2 SC3 SC4 0.0% Year Source: Auhors simulaions using SIDYN 2.0. Figure 6 Real Dividends in Slovenia,

18 84 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok As already indicaed, he dynamics of real household income change (no shown) and real GDP change (Figure 7) are similar, bu wih differen levels. Real household income increases noiceably wih respec o he reference scenario in scenarios SC2 and SC3 (on average by 1.0% in he long run) and scenario SC4 (by 1.9% in he long run). Real consumpion (no shown) and real saving (also no shown) increase accordingly in he same hree scenarios. Real GDP increases markedly in scenarios SC2 and SC3 (on average by 1.4% in he long run) and scenario SC4 (by 2.6% in he long run). This would indicae ha increasing he governmen spending on R&D by 20% per annum is he mos efficien policy measures wih respec o he long-erm economic growh, while decreasing he CIT rae by 25% and increasing he governmen spending on R&D by 10% per annum are roughly equivalen (hough inferior) policy measures. However, even if one neglecs he problem of comparabiliy of he analyzed policy measures, i is necessary o compare oher measures of well-being as well. 3.0% Change wih respec o he reference scenario 2.0% 1.0% SC1 SC3 SC2 SC4 0.0% Figure 7 Real Gross Domesic Produc in Slovenia, Source: Auhors simulaions using SIDYN 2.0. One such measure is he household welfare, which comprises he consumpion of maerial goods and services and consumpion of leisure 8. I urns ou ha he dynamic of aggregae welfare change (no shown) follows closely he dynamic of real GDP change. Namely, welfare increases noiceably wih respec o he reference scenario in scenario SC4 (by 1.6% in he long run), while he increase is less profound in he remaining hree scenarios (beween 0.5% and 0.9%). Figure 8 illusraes he dynamic of welfare changes by household ype in case of scenario SC4. One can observe ha his dynamic is very similar irrespecive of he income quinile, wih some Year 8 Household welfare in he model is defined in he form of is equivalen variaion as a share of income. The equivalen variaion represens he amoun of income needed o achieve he same uiliy level as in he reference scenario a presen prices.

19 R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh 85 divergence in levels in he course of ime. Deailed analysis of oher scenarios leads o he same conclusion; households ha would gain more uiliy in case of implemening he analyzed policy scenarios are he ones from income quiniles 2-4, i.e. he households wih more skilled and highly skilled labour, bu neiher he very op nor he boom earners in he economy. 2.0% Change wih respec o he reference scenario 1.5% 1.0% Quinile 1 Quinile 2 0.5% Quinile 3 Quinile 4 Quinile 5 0.0% Year Source: Auhors simulaions using SIDYN 2.0. Figure 8 Welfare Change in Slovenia in Case of Increasing Governmen Spending on R&D by 20%, by Household Type, Concluding Remarks The debae wheher long-run economic growh paerns can be bes explained by radiional or endogenous growh is far from seled, bu he noion ha educaion and innovaions can conribue o economic growh is nowadays widely acceped. This provided us wih he moivaion o develop a dynamic general equilibrium model wih endogenous growh, driven by invesmen in educaion and R&D. In he presen aricle, we demonsrae how R&D can be modelled as a major endogenous growh elemen in a small open economy general equilibrium framework, and consider several R&D policy scenarios for Slovenia, wih primary focus on macroeconomic and welfare aspecs. Economic growh is endogenously deermined by he developmen of human capial sock of households, secor-specific human capial and R&D socks of firms, and he oal facor produciviy. R&D aciviy in he economy is hus modelled as a secor-specific R&D aciviy and represens he key driver of economic growh, which yields new produc innovaions and adds o he knowledge base of indusry and he markeplace as a whole. The secor-specific R&D sock is accumulaed over ime hrough new invesmen made by firms, as well as by he governmen. The R&D invesmen of firms has counry-level spillover effecs via an increase of he

20 86 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok oal facor produciviy. The counry-level oal facor produciviy depends posiively boh upon he oal oupu of he R&D secor and he openness of he economy, represened by he share of foreign rade in he GDP. In he simulaions, we analyzed several policy scenarios ha direcly or indirecly increase he R&D expendiure. Increase of he share of secoral invesmen in R&D ha is deducible from he CIT and a moderae increase of governmen spending on R&D urned ou o be he mos effecive policy measures. By increasing he share of secoral invesmen in R&D ha is deducible from he CIT, i becomes raional for he firms o redisribue some profi from sharing o invesing in order o increase fuure profis. However, his policy measure is sill in par followed by a dividend increase, which is an undesired effec. On he oher hand, increasing governmen spending on R&D booss long-run produciviy in he economy, hus increasing fuure profis and he fuure value of firms. This is refleced in he dividend increase, which is a side effec (a desired effec) of he policy measure. The households ha would gain more uiliy in case of implemening he analyzed policy scenarios are he ones from middle-income quiniles, i.e. he households wih more skilled and highly skilled labour, bu neiher he very op nor he boom earners in he economy.

21 R&D and Economic Growh in Slovenia: A Dynamic General Equilibrium Approach wih Endogenous Growh 87 References Aghion, Philippe, and Peer Howi A Model of Growh Through Creaive Desrucion. Economerica, 60(2): Alvarez-Pelaez, Maria J., and Chrisian Groh Too Lile or Too Much R&D? European Economic Review, 49(2): Arrow, Kenneh J., and Gerard Debreu Exisence of an Equilibrium for a Compeiive Economy. Economerica, 22(3): Arrow, Kenneh J The Economic Implicaions of Learning by Doing. Review of Economic Sudies, 29(6): Barro, Rober J., and Xavier Sala-i-Marin Economic Growh: Second Ediion. Cambridge, MA: The MIT Press. Becker, Gary S., and Rober J. Barro A Reformulaion of Economic Theory of Feriliy. Quarerly Journal of Economics, 103(1): Bernsein, Jeffrey I Iner-Indusry and U.S. R&D Spillovers, Canadian Indusrial Producion and Produciviy Growh. Micro-Economic Policy Analysis Branch, Sraegic Policy Secor, Indusry Canada, Oawa, Working Paper 19. Boserup, Eser Populaion and Technological Change: A Sudy of Long-erm Trends. Chicago: Universiy of Chicago Press. Bye, Bria, Taran Fæhn, and Tom-Reiel Heggedal Welfare and Growh Impacs of Innovaion Policies in a Small, Open Economy: An Applied General Equilibrium Analysis. Research Deparmen, Saisics Norway, Kongsvinger, Discussion Paper 510. Canon, Erik, Ber Minne, Ae Nieuwenhuis, Ber Smid, and Marc van der Seeg Human Capial, R&D, and Compeiion in Macroeconomic Analysis. Cenral Planbureau, Den Haag CPB Documen 91. Cassou, Sephen P., and Kevin J. Lansing Growh Effecs of Shifing from a Graduaed-rae Tax Sysem o a Fla Tax. Economic Inquiry, 42(2): Coe, David T., and Elhanan Helpman Inernaional R&D Spillovers. European Economic Review, 39(5): Čok, Mija, Polona Domadenik, Tjaša Redek, and Miroslav Verbič Labour Marke Reforms in he Conex of Poliical Power Theory: The Case of Slovenia. Zbornik radova Ekonomskog fakulea u Rijeci Proceedings of Rijeka Faculy of Economics, 27(1): Diao, Xinshen, Terry Roe, and Erinc Yeldan Sraegic Policies and Growh: An Applied Model of R&D-driven Endogenous Growh. Journal of Developmen Economics, 60(2): Ferris, Michael C., and Todd S. Munson GAMS/PATH User Guide: Version 4.3. Washingon, DC: GAMS Developmen Corporaion. Freire-Serén, Maria J R&D-Expendiure in an Endogenous Growh Model. Journal of Economics, 74(1): Ghiglino, Chrisian Inroducion o a General Equilibrium Approach o Economic Growh. Journal of Economic Theory, 105(1): Ghosh, Madanmohan R&D Policies and Endogenous Growh: A Dynamic General Equilibrium Analysis of he Case for Canada. Review of Developmen Economics, 11(1):

22 88 Miroslav Verbič, Boris Majcen, Olga Ivanova and Mija Čok Globerman, Seven Linkages Beween Technological Changes and Produciviy Growh. Micro-Economic Policy Analysis Branch, Sraegic Policy Secor, Indusry Canada, Oawa, Occasional Paper 23. Glomm, Gerhard, and B. Ravikumar Growh-Inequaliy Trade-Offs in a Model wih Public Secor R&D. Canadian Journal of Economics, 27(2): Griliches, Zvi Produciviy Puzzles and R&D: Anoher Non-explanaion. Journal of Economic Perspecives, 2(4): Griliches, Zvi The Search for R&D Spillovers. Scandinavian Journal of Economics, 94(0): S29-S47. Grossman, Gene M., and Elhanan Helpman Innovaion and Growh in he Global Economy. Cambridge, MA: The MIT Press. Howi, Peer Seady Endogenous Growh wih Populaion and R&D Inpus Growing. Journal of Poliical Economy, 107(4): Ikazaki, Daisuke R&D, Human Capial and Environmenal Exernaliy in an Endogenous Growh Model. Inernaional Journal of Global Environmenal Issues, 6(1): Jones, Charles I R&D Based Models of Economic Growh. Journal of Poliical Economy, 103(4): Jones, Charles I., and John C. Williams Too Much of a Good Thing? Economics of Invesmen in R&D. Journal of Economic Growh, 5(1): Klun, Maja, and Renaa Slabe-Erker Business Views of he Qualiy of Tax, Environmen and Employmen Regulaion and Insiuions: The Slovenian Case. Inernaional Review of Adminisraive Sciences, 75(3): Kong, Tao A Selecive Review of Recen Developmens in he Economic Growh Lieraure. Asian-Pacific Economic Lieraure, 21(1): Lipsey, Richard G., and Kenneh I. Carlaw Toal Facor Produciviy and he Measuremen of Technological Change. Canadian Journal of Economics, 37(4): Mahadevan, Renuka To Measure or no o Measure Toal Facor Produciviy Growh? Oxford Developmen Sudies, 31(3): Majcen, Boris, Miroslav Verbič, and Saša Knežević The Effecs of Foreign Trade Liberalisaion and Financial Flows beween Slovenia and he EU afer Accession. Pos-Communis Economies, 17(2): Majcen, Boris, Miroslav Verbič, Ali Bayar, and Mija Čok The Income Tax Reform in Slovenia: Should he Fla Tax Have Prevailed? Easern European Economics, 47(5): Majcen, Boris, Miroslav Verbič, Olga Ivanova, and Sonja Uršič Research and Developmen, Educaion and Economic Growh: A Dynamic General Equilibrium Approach wih Endogenous Growh. Ljubljana: Insiue for Economic Research. Mahiesen, Lars Compuaion of Economic Equilibria by a Sequence of Linear Complemenariy Problems. In Economic Equilibrium: Model Formulaion and Soluion, ed. A. S. Manne, Amserdam: Norh-Holland. Nadiri, Ishaq M Some Approaches o he Theory and Measuremen of Toal Facor Produciviy: A Survey. Journal of Economic Lieraure, 8(4):