Market failures within poor institutions: the effects of bureaucrats rent-seeking activity

Transcription

1 Market failures withi poor istitutios: the effects of bureaucrats ret-seekig activity Davide Ifate, Jaa Smirova Departmet of Ecoomics ad Statistics, Uiversity of Calabria Arcavacata di Rede (CS) Italy Abstract This paper develops a model which describes the iteractio betwee the stregth of istitutioal eviromet ad the optimal allocatio of resources i a cotext where the State itervees to correct for market failure. The chages i the stregth of the istitutioal eviromet is treated usig a parameter that accouts for the retseekig activity of bureaucrats ad the admiistrative capability of govermet. The model demostrates that i a weak istitutioal eviromet, bureaucrats ret-seekig activity may represet, through adequate State policies, a mechaism which ca be used to stimulate the adoptio of good productio techology, icreasig social surplus ad to eutralize market failure with less distortios. JEL classificatio: H3; H11; P3 Keywords: istitutioal eviromet; market failures; ret-seekig 1. Itroductio Recet research o the role of istitutioal eviromet o ecoomic performace has demostrated that istitutios play a fudametal role i ecoomic developmet. The establishmet of a appropriate istitutioal ifrastructure has bee show to be essetial for achievig ecoomic developmet. I fact, the presece of solid istitutios such as costitutioal order, political ad fiacial stability, market securities, competitio policies, solid legal systems etc. are ecessary compoets for successful ecoomic growth i developed coutries as well as i those udergoig developmet. Empirical research has show that istitutioal quality cotiues to improve amog developig coutries, maily i trasitio ecoomies (Havrylyshy ad Va Roode 1999, Campos 2000, Djakov 2002, ad others). Likewise, recet studies (Campos 2000, Recaatii ad Ryterma 2000, EBRD 2002, Di Tommaso, Raiser, Weeks 2007) have demostrated that istitutios should be cosidered as importat determiats of ecoomic activity sice high istitutioal quality sustais ad improves ecoomic performace. However, while there is much agreemet regardig the importace of establishig strog istitutios for successful ecoomic outcomes, the questio o how a better istitutioal eviromet might be created ad what should be its characteristics remais udefied. I fact, may developig coutries are still uable to establish strog istitutios which ca surpass the embedded poor istitutioal eviromet. As some authors argue (Djakov et al. 2003, Acemoglu 2006, Raja ad Zigales 2006), societies may cotiue to harbour weak istitutios for various reasos, such as usuccessful istitutioal trasplatatio, poor iitial distributio of factor edowmets or the persistece of the elites prefereces for a poor istitutioal eviromet. I such weak cotexts characterised by istitutios which are icomplete 1

2 ad workig poorly ad where the markets suffer distortios which hider the achievemet of the first best allocatios, mechaisms, which would otherwise ot fuctio for strog ecoomies, could actually improve ecoomic outcomes. Oe such cotroversial mechaism is the positive effect of ret-seekig activities durig istitutioal developmet. While, ret-seekig is ormally cosidered harmful for ecoomic performace, some theoretical ad empirical models (Shleifer ad Vishy 1994, Acemoglu ad Verdier 2000, Meo ad Sekkat 2005, Ifate ad Smirova 2007) have demostrated that activities such as bribig, red tape ad corruptio may actually play a positive role whe the ecoomic coditios are weak. I fact, ret-seekig activities are frequet whe a weak govermet attempts to implemet its objectives, as the asymmetric iformatio betwee the State ad its agets leaves lot of room for opportuistic behaviours. However, despite the high opportuistic costs, State itervetios are clearly ecessary, particularly, i the presece of market failures or i the process of public goods submissio. I these situatios govermet failure, e.g. corruptio, lobbyig, bribig, red-tape, etc., may be tolerated. The issue is therefore ot so much the elimiatio of such opportuistic activities, but about establishig equilibrium betwee State iefficiecies ad market distortios. This paper discusses the beeficial cotributio of ret-seekig activity o ecoomic performace withi a weak istitutioal eviromet. I particular, we aalyze how ret-seekig mechaisms may ifluece the allocatio of resources i State ad productio sectors i a istitutioal eviromet where the bureaucrats ejoy discretioal powers whe the State attempts to correct for market failures. The model preseted here is based o two strads of literature, oe dedicated to istitutios ad their role i ecoomic relatios ad the other to the effects of ret-seekig activities durig State itervetio i the ecoomy. The importace of istitutios for ecoomic performace, is quite a ew area of research with much empirical work carried out, but little from the theoretical poit of view. The mai difficulties that theoretical research dealig with istitutios faces lie i what costitutes a istitutio ad the complexity of the relatioship betwee various istitutios ad ecoomic cocepts, makig it complicated to itroduce istitutioal idicators ito models (Nelso ad Sampat 2001). I fact, the majority of the theoretical papers studyig istitutios aalyze a sigle istitutio or a set of similar istitutios withi a ecoomic framework. Oe approach is to aalyse the orgaizatioal aspect of istitutios. A good example is the study by Huag ad Xu (1999) who preseted a model which shows how ecoomic growth rates deped o whether the fiacial istitutios are orgaised as merged or cetralized bodies. Usig equilibrium ivestmet levels, the authors foud that higher ecoomic growth rates are attaied whe istitutios are cetralized. A similar approach is the aalysis of the quality of a sigle istitutio. Likewise, Skaperdas ad Syropoulos (2001) modelled ecoomic performace as a fuctio of a sigle istitutio, such as trade security, which ca be cosidered a iformal istitute ecessary for ecoomic developmet. The authors showed how various trade regimes differ from each other i terms of trade security comparig the levels of social welfare. The model also idetified various coditios uder which coutries may impose trade restrictios because of high levels of trade isecurity, represetig a uusual isight to the lik betwee iteratioal trade ad istitutios, a coectio which is rarely explored i the ecoomic literature. 2

3 Aother way to evaluate fuctioality of istitutios is to aalyze how efficietly they have bee itroduced ito ecoomic system. Usig this approach, Jack (2002) demostrated the effect the itroductio of a ew istitutio has o social welfare i trasitio ecoomies, aalysig the iteractio betwee operatig eterprises ad a ew istitute which has bee itroduced by govermet ad comparig the levels of welfare. It was demostrated that the presece of effective public istitutios i trasitio ecoomies leads to a closure of ieffective govermet eterprises. The actual developmet of ew istitutios was cosidered by Grossma (2001) from the poit of view of the creatio of property rights. The two models preseted used a geeral equilibrium approach where agets allocate time ad effort to the creatio of property rights. While i the first model the agets appropriate the commo resources, i the secod the agets already have the iitial claims ad create their effective property rights. The author idetified the first order coditios for agets cosumptio fuctio to achieve equilibrium allocatio betwee the time ad effort required for creatig property rights. Fially, a alterative way to iclude istitutios i theoretical models is the itroductio of idicators that reflect ot a sigle istitutio, but a set of them. This approach makes it possible to udertake a broad aalysis of the role of istitutioal eviromet i the ecoomy. Usig this approach, Brezis ad Verdier (2003) costructed a model that studied the diffusio of democracy as well as the process of privatizatio amog former socialist coutries, itroducig parameters which measure the effectiveess of a repressio apparatus. The model showed that where effectiveess is low, the optimal actio of authority is the reliquishig of political power. Similary, Esfahai (2002) itroduced a parameter represetig the stregth of the istitutioal eviromet ad showed a way by which istitutios ifluece the relatioship betwee State itervetio ad eterprises. The aalysis demostrated how istitutioal factors such as evaluatio of public fuds by private agets, reliability of govermet policies, corruptio etc. iflueced the owership of eterprises ad the degree of State itervetio i the idustrial sector. Further developed structural growth model (Esfahai ad Ramı rez, 2003) examied the mutual effects of ifrastructure ad ecoomy, usig a istitutioal parameter that ecompasses the variables that ifluece the adjustmet rate for capital ad ifrastructure. The model was tested empirically, itroducig variables that reflected the effectiveess of differet istitutios. The secod strad of literature, we base our research o, regards the theoretical cosideratios of how the presece of ret-seekig activities iflueces the relatioship betwee the State ad the market. May studies have examied the effect of retseekig i the cotext of State itervetios, studyig the appearace of ret-seekig, govermet cotrol over ret-seekig behaviours, the iteractio of ret-seekig o differet levels of hierarchies, ad the implicatio of ret-seekig i State-market relatioship. This latter is of great iterest to ecoomists who have geerated a abudat amout of empirical evidece ad theoretical demostratios regardig the egative effect of ret-seekig o govermet itervetio. However, there is some work showig that ret-seekig, i certai circumstaces, may have a positive ifluece o the ecoomic performace. These studies are maily based o situatios regardig correctio of market failures, submissio of public goods ad other processes where State itervetio i the market is required. A iterestig model demostratig the positive effects of ret-seekig was offered by Auriol ad Beaim s (2000) who aalyzed the cosequeces of public sector 3

4 corruptio i a growth model. The model udertook a stable equilibrium approach ad demostrated how corruptio iflueces the icome redistributio. The authors showed that equilibrium with corruptio may be preferred over oe where corruptio is abset as corruptio mechaisms bypass bureaucratic red tape. Guriev (2004) preseted a model that deals with market failures, itroducig corruptio ad red-tape ad evaluatig the mechaism of their itegratio. The author showed that, at geeral equilibrium, the level of red tape is above the social optimal level due to the presece of corruptio. It is demostrated that, eve though corruptio may have positive effects, its overall effect is destructive ad decreases the social welfare. A similar result is obtaied by Ifate (1999) who preseted a growth model i which the presece of ret-seekig is determied edogeously ad depeds o the differet reward structures of the techologies used i the productio ad ret-seekig sectors. Aalyzig the allocatio of agets betwee the two sectors, it was foud that ret-seekig may geerate positive effects but ca ot represet a ever-lastig positive mechaism. I fact, i the log-ru ret-seekig produces egative exteralities ad detracts resources from the productio sector. As ret-seekig usually appears i the relatioship betwee the pricipal ad the aget, with the former beig more ofte the victim of corruptio due to the fiacial gais the agets obtai from the pricipal. However, Olse ad Torsvik (1998) preseted a alterative model demostratig that the prospective corruptio ca actually beefit the pricipal. The authors showed that corruptio may represet a commitmet mechaism that leads to a immediate egative static effect, which, however, has a positive dyamic effect that is evidet i the log ru. Corruptio i both private ad public sectors was cosidered by Acemoglu ad Verdier s (1998) geeral equilibrium model where the State has a role of reiforcig the cotracts i the private sector. As the authors showed, prevetig corruptio ca actually be very costly ad optimal allocatio may ivolve some degree of corruptio. This result, as they argue, cofirms the experiece of uderdeveloped coutries that do ot have sufficiet sources to prevet ret-seekig activities. Successively Acemoglu ad Verdier (2000) developed a model that describes State itervetio i the market i a attempt to correct for market failure, ad cosidered the allocatios of agets i the presece ad absece of corruptio. I the model the govermet eutralizes market failure by allocatig agets betwee the State ad productios sectors, providig the etrepreeurs with the icetive to choose good productio techology. The framework of Acemoglu ad Verdier (2000) model has bee adopted i our work, itroducig a mechaism that also reflects the ifluece of istitutioal eviromet. The defiig characteristic of our approach is the ability to capture the iteractio betwee chages i the istitutioal eviromet ad resources allocatio, usig a sigle parameter that reflects how the level of bureaucrats ret-seekig activity is correlated to the admiistrative capability of the State (Esfahai 2000). Our model ca therefore determie optimal resources allocatio whe state officers are able to ejoy high ret whe istitutios are weak as opposed to whe little ret ca be derived whe istitutios are strog. The results of our model cofirm that the ret-seekig activities of state officers may ideed cotribute to the correctio of market failures by improvig the secod-best allocatio of agets withi State ad productio sectors. We show that, i a weak istitutioal eviromet, ret-seekig ot oly stimulates the agets to opt for good productio techology but also elarges the rage of positive exterality geerated by 4

5 good techology adoptio, icreasig social surplus. I additio, our research demostrates that ret-seekig activities i a poor istitutioal cotext become beeficial oly whe the State ca perceive ad utilize the opportuistic behaviour of the agets to improve ecoomic performace. We therefore delieate a set of key policies i the form of taxes ad subsidies that eable the State to create efficiet coditios withi poor istitutios, so that ret-seekig actually works to icrease social welfare. The remaider of the paper is orgaized as follows. I sectio two we discuss how istitutioal eviromet is itroduced ito the model of Acemoglu ad Verdier (2000) ad the trasformatios which were ecessary for such a modificatio. The basic model is preseted i sectio three ad sectio four discusses its implicatios ad derives some prepositios. Sectio five exteds the model, aalyzig the outcomes whe ret-seekig activities from differet sources is icorporated. Sectio six cocludes. 2. Settigs of the model We start by costructig a model based o that of Acemoglu ad Verdier (2000), where the govermet aims to eutralize market failure by itroducig bureaucrats ito the productio sector. Market failure occurs due to the fact that etrepreeurs opt for bad techology as good productio techology is more costly. As a cosequece, the positive exterality that would have bee produced through the use of good techology is ot geerated. To iduce the productio sector to adopt good techology, some of the agets become bureaucrats whe called o by the govermet to moitor ad report o the techology choices made by etrepreeurs. The govermet i tur imposes taxes o the etrepreeurs who use the bad techology ad, at the same time, trasfer subsidies to those who use the good techology, modifyig their pay-offs. This mechaism establishes a allocatio of agets such that positive exterality is geerated ad thus social surplus icreased. At this poit we itroduce ito the Acemoglu ad Verdier (2000) model a parameter that reflects the istitutioal eviromet to aalyze how chages i the stregth of the istitutios ifluece allocatios of agets ad productio of positive exterality. To provide the aalysis, however, some of the settigs i the origial model must be modified. It is ecessary to first suppose that whe the istitutioal eviromet weakes, bureaucrats are proe to practice ret-seekig activity. To itroduce ret-seekig ito the model it is assumed that the govermet gives subsidies ad collects taxes through the bureaucrats, givig them direct access to the fuds ad allowig bureaucrats to extract a premium or persoal ret durig the trasfer of the public fuds to/from a productio sector. As discussed i the first sectio, this mechaism was proposed by Esfahai (2000) who argues that, for self iterested bureaucrats, each dollar uder govermet cotrol is worth more tha a dollar, because it ca be utilized for private purposes ad thus icreases the bureaucrats rets. Accordig to Esfahai (2000), the size of the premium depeds o various istitutioal factors, such as fiacial ad ecoomic stability, efficiecy of the legal system, market security etc.; poor istitutioal quality permits the extractio of high rets by bureaucrats, while withi strog istitutios, ret is egligible or ull. Premium level also reflects the admiistrative 5

6 capability of the govermet which refers to a set of bureaucratic istitutios developed to cotrol ad promote the activities of the etrepreeurs. If the govermet is admiistratively capable, it collects taxes ad releases subsidies with less distortios ad bureaucratic costs, decreasig the volume of the premium. For the sake of simplicity, i our model, the premium is expressed as i terms of the delay i the trasfer of specific taxes from bad techology etrepreeurs to the govermet ad of the subsidies from the govermet to the good techology etrepreeurs. Hece, the subsidies ad the taxes become available after a delay 1, with a short delay cosidered welfare icreasig. Therefore, the bureaucrats represet a istitute that trasfers subsidies to the etrepreeurs who use good techology ad collects taxes from those utilizig bad techology. They are self-iterested ad extract ret from the subsidies ad taxes which iflueces their pay-offs. The amout of ret that bureaucrats ca extract depeds o the stregth of the istitutioal eviromet. I a weak istitutioal eviromet, where the govermet is ot admiistratively capable ad istitutioal quality is poor, the bureaucrats ca extract a high premium, while, i a strog istitutioal eviromet, the premium is egligible. I our model, aother assumptio must be added to overcome the problem of egative subsidy which was idetified by Acemoglu ad Verdier (2000). I fact, i their model, etries of the State budget, which were utilized to pay the wages of bureaucrats ad subsidies to good techology etrepreeurs, come from the taxes extracted from bad techology etrepreeurs. Cosequetly, the presece of bad techology etrepreeurs is ecessary to balace the State s budget. However, i the fial equilibrium, bad techology is completely elimiated. Such a allocatio of agets could ot therefore, be supported by a State that has o fuds to pay the bureaucrats ad good techology etrepreeurs. To overcome this problem, Acemoglu ad Verdier (2000) allowed for a egative subsidy, whereby good techology etrepreeurs were also taxed. I our model, both positive ad egative subsidies ca be applied to achieve a fial equilibrium by itroducig a additioal source of etry for the State i the form of a geeral tax paid by the etrepreeurs. We suppose that the geeral tax is paid whe the etrepreeurs are moitored by the bureaucrats 2, i.e. depeds o the probability of bureaucrats moitorig. The State budget is therefore composed of both the specific tax paid by bad techology etrepreeurs as well as the geeral tax paid by the etrepreeurs. Based o the above assumptios we ca ow costruct a model that aims to aalyze how chages i the stregth of the istitutioal eviromet ca ifluece the allocatio of agets i both State ad productio sectors, supportig the eutralizatio of market failure ad the geeratio of positive exterality. 3. Model The model cosiders two types of productio techologies chose by the etrepreeurs () of the productio sector. Good productio techology geerates a positive exterality (β) together with a output (y) that gives beefits to the other agets 1 A example of such ret-seekig activity ca be the iterest rate gaied from delay i paymets of subsidies ad i the trasfer of taxes, a commo practice i trasitio coutries (Saha, 2000). 2 The itroductio of such taxes is i respose to the reality surroudig weak istitutios where tax evasio is commo practice. 6

7 (with the total umber of agets beig 1), ad requires cost (c). Bad productio techology does ot produce ay exterality ad does ot require ay cost, producig the same amout of output as good techology. As the etrepreeurs are iterested i maximizig the pay-off that they ca extract from the productio ad sice the applicatio of bad productio techology implies a greater pay-off, they are ot motivated to choose the good techology ad positive exterality is therefore ot geerated. To icetivate the choice of good techology, the govermet iduces some of the etrepreeurs to become bureaucrats (1-) to moitor the techological choices of the etrepreeurs, trasferrig subsidies (s) to the etrepreeurs who use good techology (x) ad collectig taxes (τ) from the etrepreeurs utilizig bad techology (-x), while both types of etrepreeurs pay a geeral tax (t) to the govermet. Bureaucrats moitor etrepreeurs radomly with the probability give by: p( ) = 1! Whe bureaucrats moitor etrepreeurs, the pay-offs of the good ad bad etrepreeurs are give by:! = y + β x c p() t + p() s good! bad = y + β x p() t p() τ (1) Pay-offs thus deped o the output produced, positive exterality received ad o taxes or subsidies paid/received oce beig moitored. The govermet is iterested i maximizig the social surplus (SS) that is positively correlated to the etrepreeurs output ad the positive exterality: SS! y + (" # c)! x (2) With o govermet itervetio, whe etrepreeurs do ot apply good techology, the allocatio of agets is give by =1, x=0. Oce the govermet itervees, it aims to allocate the agets such that =x, so to obtai exclusively good techology adoptio ad thus maximizig social surplus. Give the above settigs, the followig sectios aalyze how chages i the premium that bureaucrats put o subsidies ad specific taxes ifluece both the allocatio of agets at the secod-best equilibrium poit ad the amout of positive exterality which ca be geerated by good techology etrepreeurs The allocatio of etrepreeurs betwee good ad bad techology We start by aalyzig how the premium which bureaucrats extract from public fuds iflueces the allocatio of etrepreeurs betwee the good ad bad techologies. To maximize the social surplus the govermet has the followig set of costraits to respect: 1. Liability costrait. The total amout of the geeral ad specific taxes paid to the govermet does ot exceed the value of etrepreeurs output:! # y " t (3) 2. Techology costrait. To iduce the etrepreeurs to use good productio techology, the pay-off of good techology etrepreeurs is greater tha those of bad techology etrepreeurs, implyig the followig iequality: 7

8 " c s +! $ 1 # (4) 3. Govermet budget costrait. The govermet does ot sped more tha it ears, hece the amout of taxes it collects is at least equal to its total expeses: 1! 1! 1! "(! x) "# + " t" $ ( 1! ) " w + " x" s (5) where the amout of specific ad geeric taxes the State collects is o the left ad the value of wages paid to the bureaucrats plus the value of subsidies released for good techology etrepreeurs is o the right. 4. Allocatio of talet costrait. To iduce some agets to become bureaucrats the pay-off of a bureaucrat is greater tha the pay-off of a good techology etrepreeur: 1 # w +!" x + $ ' % ( & ) "( s + *)" + y +!" x # c 1 # 1 #, # "t + "s (6) where o the left there is the wage ad the positive exterality a bureaucrat receives plus the premium a bureaucrat extracts from specific taxes ad subsidies. Substitutig the above iequalities ad solvig them for x, we obtai the etire set of costraits the govermet must respect: y"( 1! ) 2 x $ + #"( 1! ) " c (7) This costrait defies the allocatio of etrepreeurs betwee good ad bad techology ad clearly shows that the umber of good techology etrepreeurs is positively related to the premium which the bureaucrats ca extract from subsidies ad taxes The allocatio of agets betwee State ad productio sectors To aalyze how the premium extracted by bureaucrats affects the allocatio of agets betwee etrepreeurs ad bureaucrats, we first cosider a State which is attemptig to eutralize market failure give costrait (7). The costrait set is ocovex i x ad social surplus is liear i x, so that the maximizatio of social surplus has two solutios. The first solutio is give by =1 ad x=0, where oe of the etrepreeurs uses good techology. This solutio gives the level of social surplus that presumes market failure is ievitable: SS MF = y (8) The secod solutio correspods to equality =x, where all the etrepreeurs use good techology. I this case the social surplus is give by: SS = " ( y + #! c) (9) Clearly, the State itervees i the productio sector oly whe SS MF! SS. As optimal outcome from State itervetio requires =x, i substitutig x for i (7) we obtai the secod-best allocatio of agets ( E ): ( )! 2" y + c "# + 4" c" y + c 2 "# 2 E 2" ( c! y + c "#) (10) 8

9 d Sice E! 0 is valid, the premium o public fuds (γ) is positively related to the d" dimesio of the productio sector at the secod-best allocatio poit. Clearly, the level of social surplus that correspods to this allocatio is also positively related to the premium The level of positive exterality produced by good techology etrepreeurs Here we aalyze the relatioship betwee the level of positive exterality produced by the adoptio of good techology ad the premium. To do this we compare the level of social surplus i case the market failure has bee eutralized (SS) to that with the market failure ( SS MF ) ad defie the threshold level of positive exterality (! TH ) over which ay level of β gives a level of social surplus that is greater tha SS MF : % c" y" # + c 2 "# + ( c $ y) " 4" c" y + c 2 "# 2 ' & (! TH $ 2" y + c "# + 4" c" y + c 2 "# 2 ( ) (11) "# TH " Sice! 0 ad E d#! 0 the TH! 0. As a result, a higher premium the " E "# d" bureaucrats put o subsidies ad specific taxes actually decreases the threshold level of positive exterality over which State itervetio i the ecoomy is optimal. 4. Aalysis of the model s outcome The above results ow allow us to derive some propositios. We first suppose, that istitutioal eviromet becomes weaker (γ icreases) aroud the time the state itervees i the productio sector by itroducig some bureaucrats, so that the bureaucrats ca have more discretio power ad thus extract higher rets. Iequality (7) suggests that the icrease i γ leads to a upward shift of the costrait curve (Figure 1). While poit =1, x=0 (that defies market failure) remais uchaged i such a shift, the poit of secod-best solutio moves up alog the lie =x, passig from E 1 to E 2, demostratig that there is a icrease i the umber of good techology etrepreeurs ( E ) i accordace with (10). Therefore, the icrease of γ leads to a better agets allocatio ( E2 > E1 ), reachig a ew secod-best poit, E 2. We ca ow aalyze the mechaism of govermet itervetio i detail. I order to operate with parameters the govermet ca cotrol, (10) ca be re-writte as:!#(" + s) + ( t + s) E!#(" + s) + ( s + t) + 2 #" (12) Careful examiatio shows that sice =x, ay icrease i γ requires chages i some edogeous parameters if a icreased i E is eeded. To prove this, we rearrage the terms of the costraits the govermet must respect to satisfy (12). Techology costrait (4) thus takes the followig form: s +! " c + ( s +!) (13) 9

10 1 =1 x=0 Costrait Sets =x E 2 γ 1 < γ 2 1/2 E 1 SS SS SS ' MF E1 E2 x Figure 1. The secod-best allocatio of agets i a weakeig istitutioal eviromet. Budget costrait (5) ca be preseted as: x $ #! + t " w! + s (14) The talet costrait (6) becomes: w ) y! c + %& s# ( 1! ")! t! $ #"'( # 1! (15) As (15) shows, growth i γ decreases w. At the same time, from (14), it follows that the decrease i w implies higher x. Sice the State searches for =x, whe x icreases, icreases automatically. However, as (13) shows, a icrease i suggests chages i s, τ, ad/or c (τ i tur depeds o the levels of y ad t) 3. As a result, icreasig γ i equatio (12) does ot suggest a automatic icrease i E, but a chage i govermet policy that leads to a icrease i E. Obviously, there is a wide rage of policies the State may implemet, such as chagig geeral taxatio, bureaucrats wages, the value of subsidies etc. It is worthwhile to ote that the costrait set imposes precise liks betwee some of theses parameters ad chagig oe of them would automatically chage others. For example, with a icrease i γ, oe of the reasoable strategies the State may choose is to reduce bureaucrats wages (w). I fact, i coutries sufferig from istitutioal vacuum, the govermet has little cotrol over icreasig disorder ad bureaucrats ret-seekig activities become sigificatly profitable. Therefore, i such weak istitutioal cotexts, 3 Clearly, the cost of good techology (c) ad the level of etrepreeur s output (y) caot be cosidered uder the State s policy. 10

11 govermet policy may use the icrease i γ to lower bureaucrats wages, icreasig subsidies i the productio sector. Applyig our model, we ca demostrate the applicatio of such a policy. Usig the four costraits (3, 4, 5, 6 ad the requiremet x=), the depedece of w o γ ca be obtaied for the secod-best poit: ( ) 2 4 c c "! + c "! + " " y w y # 2 (16) The secod term of the above equatio is positive, thus cofirmig that icreasig γ decreases bureaucrat s wage. It is ow possible to evaluate whether a decrease i w ca lead to a icrease i s. Rearragig the set of costraits we derive the equatio for s: s +! c #" + c #" ( ) c # # y 2 (17) which simply becomes s +! + w = y. I fact, i lowerig w, the State chooses a higher s, ad the remaiig variables are adjusted respectively. Therefore, a decrease i bureaucrats wages is used to support a icrease i the subsidies for good techology etrepreeurs. We have demostrated that a policy which reduces bureaucrats wages so to icrease subsidies may represet a solutio for eutralizig market failure i a weakeig istitutioal eviromet. This meas that, i usig a appropriate respose to the icrease of ret-seekig activity, the govermet ca eutralize market failure with the better allocatio of agets. I fact, despite ret-seekig activities ad weak istitutios, it is possible to icrease the umber of good techology etrepreeurs employig less bureaucrats who are also paid lower wages. Fewer bureaucrats ecessary for achievig a secod-best allocatio of agets also reduces the bureaucratic costs resultig from the loss of output produced by etrepreeurs. It is therefore possible to give the followig propositio: Propositio 1: I a weak istitutioal eviromet, the icrease i ret-seekig activity of bureaucrats may eable the govermet, through adequate policies, to eutralize market failure with a larger productio sector, smaller bureaucratic costs, thus, icreasig social surplus. Now we ca aalyze how ret-seekig activity iflueces the threshold level of positive exterality produced by good techology. As show above, a higher premium put by bureaucrats o public fuds implies a lower threshold level of positive exterality produced by good techology (11). This mechaism is described by Figures 2 ad 3. Figure 2 demostrates that icreasig the discretio power of bureaucrats (γ 2 >γ 1 ) leads to a better fial allocatio of agets E1< E2, passig from poit E 1 to E 2. This chage leads to a higher level of social surplus, i.e. SS 2, istead of SS 1 (Figure 3), that correspods to a lower threshold level of positive exterality (β TH2 < β TH1 ). I fact, i Figure 3 it is easy to see that the threshold level of positive eterality, over which State itervetio becomes optimal, shifts to the right with a icrease i premium. Hece, β TH2 offers a wider rage of feasible good techology applicatios ad higher social surplus with respect to β TH1. The level of β TH, expressed i the followig way, clarifies this process:! TH y E " y + c (18) 11

12 x x 1 = x (, γ 1) x 2 = x (, γ 2) 1 γ 2 > γ 1 E 1 E 2 x = E1 E2 1 Figure 2. Allocatio of agets i a weakeig istitutioal eviromet. SS SS 1 = (y + β TH1 c) SS 2 = (y + β TH2 c) y β TH2 < β TH1 SS MF = y E1 E2 1 Figure 3. Productio of positive exterality i a weakeig istitutioal eviromet. Therefore, a icrease i E, clearly leads to a decrease i β TH. This mechaism may be explaied i the followig maer: eve with a lower level of β TH, a icrease i the umber of good techology etrepreeurs, resultig from a icrease i premium establishes the level of social surplus beyod SS MF. Hece, i a weak istitutioal eviromet it is possible to actually elarge the rage of feasible good techology which ca be adopted, such that eve techologies offerig a low level of positive 12

13 exterality become worthwhile for State itervetio. The followig propositio ca thus be formulated: Propositio 2. I a weak istitutioal eviromet, adequate govermet policies ca utilize the ret-seekig activity of bureaucrats to lower the threshold level of positive exterality at which State itervetio i the productio sector is optimal, thus elargig the rage of feasible good techology which ca be adopted. 5. Extesios of the model: differet premiums o public fuds I the previous sectio, γ was cosidered the sole parameter describig bureaucrats behaviour. To obtai a more detailed aalysis of the impact of the stregth of the istitutioal eviromet o the allocatio of agets, γ ca be exteded ito two ew parameters λ ad δ that respectively represet the premium bureaucrats put o subsidies ad that placed o specific taxes. Clearly, distict premiums may represet differet sets of istitutios ivolved i the mechaism of extractig taxes or gratig subsidies ad may thus ifluece the secod-best allocatio of agets differetly. Modifyig the settigs of the model with respect to the talet costrait, bureaucrats pay-off is chaged as follows: 1 # w +!" x + $ ' % ( & ) "( s "* + + ",) y +!" x # c 1 # 1 # - # "t + "s (19) Combiig costraits (3) (4) (5) with this, the ew costrait set is ow give by: 2 2 ( y "( 1! ) % (( 1! ) "( y! t) "( +! *)% x * ) & # + & # + * "( 1! ) (20) ' " c $ ' " c $ The above costrait set leads to the followig secod-best allocatio of agets betwee State ad productio sectors: 2 2 ( y! t) "( $! #) + 4" c " y + c "# + 4" c "( y! t) "( $! #) 2" [ c! y + c "#!( $! #)" ( y t) ]! 2" y + c "#! 2" * = E! (21) This set of expressios thus presets a more complete model which allows for a more detailed aalysis of bureaucrats behaviour. The followig sectio applies these settigs to the developmet of State policies that aim to fid the respose to the retseekig activity of bureaucrats Ret-seekig as a fuctio of premiums o public fuds Here we describe the dyamics of govermet itervetio i the productio sector to eutralize for market failure, as a fuctio of the differet premiums bureaucrats ca extract from taxes ad subsidies. We first suppose, that the govermet cotrols ad x by establishig w, τ, s, t, so that (21) is satisfied ad the allocatio of agets ca reach poit E 1 (Figure 4). I the iitial stage where the allocatio of agets moves from poit A to poit B ad the umber of bureaucrats starts growig, these bureaucrats first face bad techology etrepreeurs who must pay the specific tax τ to the govermet. The icreasig umber of bureaucrats, through the moitorig process, augmet the pay-off of good techology etrepreeurs, givig the icetive to their appearace. As see from Figure 13

14 4, passig from poit A to poit B, the umber of good techology etrepreeurs would thus icrease from 0 to x B. 1 A =x B C D E 3 1/2 E 1 E 2 SS 3 SS 1 SS 2 x B x C x D x Figure 4. The secod-best allocatio of agets i a weakeig istitutioal eviromet: a positive effect of ret-seekig, give differet premiums o specific taxes ad subsidies. Subsequetly, the bureaucrats icrease their discretio power sice the istitutioal eviromet at poit B becomes weaker (i.e. δ icreases). Rearragig iequality (20) to mirror this icrease i ret-seekig behavior, we obtai: 2 2 ( ) ( ) ( ) $( ) ( ) 2 & y ' 1( # & 1( ' y ( t # & 1( ' ( y ( t ) # x* ) $! + + ' $! + * ' 1( (! (22) % ' c " % ' c " % ' c " dx* Therefore,! 0 which meas that the costrait curve moves forward with d" a icrease i δ. Hece, the allocatio moves to poit C where the umber of good techology etrepreeurs icreases (x C ). The shift of the costrait curve also implies that poit E 1 moves upward alog lie =x to the poit E 2, as show Figure 4, represetig a higher level of social surplus. Oce good techology etrepreeurs appear, the bureaucrats also begi to extract ret from subsidies. Referrig to equatio (22) ad aalyzig the derivative dx *, it becomes evidet that the premium o subsidies may have both a positive or d! dx* egative ifluece o the allocatio of etrepreeurs. I fact,! 0 whe d" ( 1! ) c # "( y! t) (23) Coditio (23) thus defies the iterval of good techology cost that guaratees the positive impact of λ o the umber of good techology etrepreeurs. Applyig 14

15 appropriate substitutios, it is easy to see that (23) actually represets the coditio of positive subsidy, s! 0. The case of egative subsidy will be discussed below. I the coditio where (23) is valid, a icrease i λ moves up the costrait set curve ad correspods to a more itesive use of good techology which would shift the allocatio of agets to poit D (Figure 4). At poit D, the bureaucrats cotiue extractig premiums λ ad δ, promptig etrepreeurs to switch to the good techology that offers greater pay-off. Moreover, the State utilizes bureaucrats ret-seekig activity to modify the exogeous parameters, allocatig agets to x E = E ad reducig the umber of bureaucrats (e.g. the situatio moves to poit E 2 ). Therefore, the more itesive the ret-seekig activity of bureaucrats 4, the greater the x E, ad market failure actually becomes eutralized at a higher level of social surplus Negative subsidy for good techology etrepreeurs I the situatio where subsidy is egative, i.e. the case where (23) does ot hold, modifyig the costraits the govermet must respect (ote, the liability costrait τ y - t is maitaied, sice s τ ), the costrait set becomes: & y (( 1' ) # &( 1' ) (( y ' t) # &( 1' ) (( y ' t) # x* ) $! + + ( $! + * ( $ ' ( 1' )!" (24) % ( c " % ( c " % ( c The secod term i iequality (24) shows that the icrease i premium o bad techology for which etrepreeurs are taxed (τ) icreases the umber of good techology etrepreeurs (x*). At the same time, a raise i premium o subsidies of good techology etrepreeurs (λ) icreases x* if the third term of iequality (24) is positive. Elaboratig this expressio ad applyig appropriate substitutios, we fid dx* that, as i the above case of positive subsidy, the coditio of! 0 is oly possible d" whe s 0. Hece, regardless of whether the subsidy o good techology etrepreeurs is egative or positive, the mechaism of ret-seekig may actually cotribute to achievig a better allocatio of agets i the productio sector whe istitutios are weak. O oe had, this result cofirms that of Acemoglu ad Verdier (2000), showig that it is possible to apply a egative subsidy to eutralize market failure. However, our model also shows that whe good techology etrepreeurs are taxed, the umber of good techology etrepreeurs is reduced compared to the case where good techology etrepreeurs are give subsidy. The followig equality cofirms this statemet: "! s * = E c + "! (25) s Therefore, usig egative subsidies istead of positive subsidies with good techology etrepreeurs, the fial allocatio of agets is give by a lower umber of 4 The premium ca ot icrease idefiitely to reach the first-best allocatio of agets (=x=1) sice it would require bureaucrats to receive a egative wage. The maximum possible level of premium at the secod-best allocatio poit is give by substitutig w=0 the costrait set ad settig =x, hece: " 1. = c y max! 15

16 good techology etrepreeurs ad greater umber of bureaucrats. The followig propositio ca thus be formulated: Propositio 3. The State may utilize ret-seekig to improve the secod-best allocatio of agets regardless of whether the subsidies o good techology etrepreeurs are positive or egative. I additio, a better secod-best allocatio of agets is obtaied whe the State releases positive subsidies Sesitivity of good techology adoptio to the premiums o subsidies ad taxes Iterestigly, the secod-best allocatio of agets behaves differetly to the chage of the premium o subsidies or specific taxes. To demostrate it, from (24) we ca fid the ratio of the first derivative of both premiums: d x * d# s = (26) d x *! d" Equatio (26) demostrates that chages i λ or δ elicit a differet respose i x*, idicatig that the State may actually choose which premium to utilize ad which policies to apply whe eutralizig market failure. Such a choice could be valuable i a situatio whe differet istitutios are ivolved i the processes of subsidizatio ad taxatio. For example, subsidizig may be give a primary role if the State does ot ited to improve the correspodig istitutio i the ear future but iteds to impose as much good techology as possible. The followig propositio ca thus be formulated: Propositio 4: As a respose to the bureaucrats ret-seekig activity, the State may have at its disposal various policies for correctig market failure sice the secodbest allocatio of agets behaves differetly to the chage of the premium placed by bureaucrats o specific taxes or subsidies. 6. Coclusios We have preseted a theoretical model for aalyzig how chages i istitutioal eviromet stregth affects the allocatio of resources i both State ad productio sectors i a situatio where the govermet itervees to corrects for market failure. We treat the chages i istitutioal eviromet with a sigle parameter that accouts for both the ret-seekig activity of bureaucrats as well as the admiistrative capability of the govermet. I costructig the model we suppose that bureaucrats, itroduced to eutralize market failure, are ivolved i trasferrig subsidies ad taxes to/from etrepreeurs who apply good or bad productio techology, respectively, becomig, therefore, a source of ret for bureaucrats. I a weakeig istitutioal eviromet, a icrease i bureaucrats discretio power iflueces the choice of which agets become the etrepreeurs or the bureaucrats as well as the bad or good techology type which etrepreeurs choose. The model demostrates that, i a weak istitutioal eviromet, the State may actually use the bureaucrats ret-seekig activity to achieve a secod-best agets allocatio wherei more etrepreeurs opt for good techology. I fact, by implemetig adequate State policies as a feedback mechaism to guide ret-seekig behaviours, market failure ca be eutralized with lower bureaucratic costs ad with 16

17 higher level of social surplus. We also show that, i a weak istitutioal eviromet, the State may utilize bureaucrats ret-seekig activity to exted the rage of feasible good techology adoptio i the productio sector. The model also provides a detailed aalysis that cosiders how differeces i the ret which bureaucrats ca extract from taxes ad subsidies iflueces the secod-best aget allocatio. I fact, we fid that the secod-best allocatio of agets chages as a fuctio of the ret which bureaucrats maage from differet sources. Therefore, to eutralize market failure, the State may chose amog various policies, depedig o the state of the istitutios ivolved i taxatio ad subsidizatio. Fially, our model also shows that while both positive ad egative subsidies ca be applied to good techology etrepreeurs to eutralize the market failure, a secod-best allocatio of agets with the adoptio of more good techology ad greater social surplus is achieved whe the subsidy is positive. As show, i a weak istitutioal eviromet, bureaucrats ret-seekig activity ca actually work as a costructive mechaism to achieve a better allocatio of resources. However, the positive effect of ret-seekig is ot everlastig. I fact, oce the istitutioal eviromet is stroger, the search for ret becomes a egative factor that exhausts the ecoomic resources ad is detrimetal to ecoomic performace. That this caot be demostrated is a limit of this model. Future improvemets to the model would eed to cosider how more sophisticated istitutioal parameters ca be icorporated so to overcome this limit ad allow for a theoretical descriptio of how iitially positive ret-seekig behaviour the becomes detrimetal whe the origially weak istitutio becomes stroger. As is, the model oetheless offers a suitable framework for guidig a empirical cross-coutry research to test the propositios regardig the effect of ret-seekig o ecoomic performace i a chagig istitutioal eviromet. Refereces Acemoglu D., Verdier T., Property Rights, Corruptio ad Allocatio of Talet: a Geeral Equilibrium Approach The Ecoomic Joural, 108 (450), Acemoglu D., Verdier T., The Choice Betwee Market Failures ad Corruptio The America Ecoomics Review, 90(1), Acemoglu D., A simple model of iefficiet istitutios Scadiavia Joural of Ecoomics, 108(4), Auriol E., Beaim M., "Stadardizatio i Decetralized Ecoomies" America Ecoomic Review, 90(3), Brezis E.S., Verdier T., "Political Istitutios ad Ecoomic Reforms i Cetral ad Easter Europe: A Sowball Effect" Ecoomic Systems, 27(3), Campos N., Cotext is Everythig: Measurig Istitutioal Chage i Trasitio Ecoomies World Bak, policy research workig paper Di Tommaso M., Raiser M, Weeks M., Home Grow or Imported? Iitial Coditios, Exteral Achors ad the Determiats of Istitutioal Reform i the Trasitio Ecoomies Ecoomic Joural, 117 (520), Djakov S., Glaeser E.L., La Porta R., Lopez-de-Silaes F., Shleifer A., The New Comparative Ecoomics NBER Workig Papers Series, Workig Paper 9608, April. EBRD Trasitio Report, EBRD, Lodo Esfahai H.S., Istitutios ad Govermet Cotrols Joural of Developmet Ecoomics, 63(2), Esfahai H.S., & Ramirez M.T., "Istitutios, ifrastructure, ad ecoomic growth" Joural of Developmet Ecoomics, 70(2),

18 Grossma H. I., Papers ad Proceedigs of the Hudred Thirteeth Aual Meetig of the America Ecoomic Associatio, Guriev S., Red Tape ad Corruptio Joural of Developmet Ecoomics, 73(2), Havrylyshy O., va Roode R., Istitutios Matter i Trasitio, But So Do Policies Comparative Ecoomic Studies, 45(1), Huag H., Xu C., Istitutios, Iovatios, ad Growth America Ecoomic Review, 89 (2), Ifate D., Etrepreeurship, Ret-seekig ad Ecoomic Developmet Federico Caffè Cetre, Roskilde Uiversity, Research Report, 9/1999. Ifate, D., Smirova J., Istitutioal Eviromet ad Trasitio Ecoomies i proceedigs of the Sixth Iteratioal Scietific ad Practical Coferece "State Regulatio of the Ecoomy. A Regioal Aspect", Uiversity of Nizhi Novgorod Russia. Jack W., Istitutioal Desig ad the Closure of Public Facilities i Trasitio Ecoomies Ecoomics of Trasitio, 10(3), Méo P.G., Sekkat K., Does corruptio grease or sad the wheels of growth? Public Choice, 122(1-2), Nelso R.,N., Sampat B.N., "Makig Sese of Istitutios as a Factor Shapig Ecoomic Performace" Joural of Ecoomic Behavior ad Orgaizatio, 44(1), Olse T., Torsvik G., Collusio ad Regulatio i Hierarchies: A Case of Beeficial Corruptio Iteratioal Ecoomic Review, 39(2), Raja, R.G., Zigales L., "The Persistece of Uderdevelopmet: Istitutios, Huma Capital or Costituecies" CEPR Discussio Papers, Recaatii F., Ryterma R., Disorgaizatio or self-orgaizatio? World Bak Saha B., Red Tape, Icetive Bribe ad the Provisio of Subsidy Joural of Developmet Ecoomics, 65(1), Shleifer A., Vishy R.W., Politicias ad Firms The Quarterly Joural of Ecoomics, 109(4), Skaperdas S., Syropoulos C., "Gus, Butter, ad Opeess: O the Relatioship betwee Security ad Trade" America Ecoomic Review, 91(2),