political Economy of Patent Protection and Economic Growth

political Economy of Patent Protection and Economic Growth

著者 池下 研一郎

journal or

publication title

金沢大学経済論集 = Kanazawa University economic review

volume 34

number 2

page range 275‑297

year 2014‑03‑31

URL http://hdl.handle.net/2297/36855

Ⅰ　Introduction

Intellectualproperty right(hereafterIPR)protection hasbeen an importantpolicy issue.Forthelasttwo decades,many developed countriessuch astheUnited States

－２７５－

Contents

Ⅰ　Introduction

Ⅱ　A SimpleModelofInnovation and IPR 　1.Household

　2.IPR Policy 　3.Producers 　4.R&D Sector 　5.NationalIncome 　6.TheEquilibrium Path

　7.GovernmentwithoutPoliticalDonation

Ⅲ　TheEffectsofPoliticalDonation 　1.Introducing PoliticalDonation 　2.Governmentwith PoliticalDonation

Ⅳ　Concluding Remarks

Ec onomi c Gr owt h*

Ke n- i c hi r o I ke s hi t a

艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶

＊Thisstudy ispartly supported by Grant-in-Aid forScientificResearch from Japan Society for Promotion ofScience(No.25380290).

†Faculty ofEconomicsand Management,InstituteofHuman and SocialSciences,Kanazawa University,Kakuma-machi,Kanazawa,Ishikawa,920-1192,Japan.

　E-mail:[email protected]

－２７６－

and Japan havestrengthened IPRs.Additionally,sincetheAgreementson Trade- Related AspectsofIntellectualProperty Rights(TRIPs)hasbeen approved aspart of the Final Act of the Uruguay Round, many developing countries also have strengthened their IPR protection. In developed countries, strong IPR policies enhancethereturnsto research and developmentand increaseproductivity and long- run growth rate. In developing countries, strong IPR policies attract foreign investmentand technology,promoting economicgrowth.

Theabovediscussion impliesthatfirmsthatengagein research activity desire strong IPR protection.According to theCentreforResponsivePolitics(2012),the amountofmoney spenton lobbying in theUnited Stateshasbeen increasing from 2000,and reached $ 3.54 billion in 2010.In particular,thehealth-related industry, which includespharmaceuticaland medicalcompanies,spendsmorethan $ 524 million per year on lobbying activities¹⁾. Since we know that these industries strongly depend on domesticIPR policies,itisnaturalthatIPR policiesareaffected by politicalactivitiessuch aslobbying orpoliticaldonations.

Many researchersdiscusshow IPR policiesaffectdevelopmentand economic growth.In termsoftheoreticalstudies,Judd (1985)examined how patentlength affectsthemarketequilibrium path by using an exogenousgrowth model.However, anumberofpapershaveinvestigated therelationship between IPR policiesand economicgrowth afterthedevelopmentofendogenousgrowth theory by Romer (1990), Grossman and Helpman (1991), and Aghion and Howitt (1992). In particular,Goh and Olivier(2002),Iwaisako and Futagami(2003),Kwan and Lai (2003),and O’Donoghueand Zweimüller(2004)examined thedomesticeffectsof IPR policies on economic growth using an endogenous growth framework. In general,thesestudiesfound thatstrengthening IPR policiesincreasesproductivity and thelong-run growth rate²⁾.In termsofempiricalstudies,Gould and Gruben (1996) examined the relationship between IPR protection and per capita GDP growth for95 countriesin theperiod 1960-1998 by using an index ofIPR protection constructed by Rapp and Rozek (1990).Gould and Gruben (1996)find thatIPR policiesaffecteconomicgrowth significantly and thiseffectisrelatively strongerin

－２７７－

open economies.

Further,many researchersalso havestudied thedeterminantsofIPR policies.In particular,Nordhaus(1969)and Scherer(1972)started astudy ofoptimalpatent duration. Klemperer (1990) and Gilbert and Shapiro (1990) examined optimal patentbreadth.In general,strengthening IPR policieshastwo effectson thewelfare. Firstofall,by strengthening IPR protection,thegovernmentcan providegreater incentivesforinnovation,and thiswilllead to consumersenjoying betterproducts. Thisisamarginalbenefitofstrengthening IPR protection.However,marginalcost ofstrengthening IPR protection ismaking consumerssufferfrom monopoly prices and dead-weightlosses.Theoptimalpatentpolicy balancesthemarginalbenefits with marginalcosts.

In atheoreticalstudy,Iwaisako and Futagami(2003)used avariety-expansion growth modelbased on Romer(1990)to show thattheoptimalpatentlength that maximizessocialwelfareisfinite.Laiand Qiu (2003)and Grossman and Lai(2004) also developed theNorth-South trademodelto analyzeinternationaleffectsofIPR protection, and showed that the governments of developed countries chooses strongerIPR protection than thoseofdeveloping countries.However,thesestudies assumed that the government is non-corrupt and maximizes household utility. Recently, Eicher and García-Penalõsa (2008) showed how private incentives to protect IPR affect economic growth. In particular, they have assumed that the privatefirmsthemselvesprotecttheirIPR and show thatmultipleequilibriacan emerge.

In thispaper,weconstructan endogenousgrowth modelthatonly incorporates 艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶 1)Thefinancialsectorand communications/electronicssectoralso spend alotofmoney on

lobbying activities.Fordetails,seetheCentreforResponsivePolitics(2012).

2)Somestudiespointoutthepossibility thatstrengthening IPR protection discourageseconomic growth.Forexample,Goh and Olivier(2002)developed agrowth modelwith an upstream sectorthatproducesdifferentiated inputsand adownstream sectorthatproducesdifferentiated finalgoods.They showed thattightening IPR in thedownstream sectorweakenstheincentive to innovateand decreasestherateofeconomicgrowth through themarketsizeeffect.

－２７８－

innovation asasourceofeconomicgrowth.Using thismodel,weconsiderhow politicaldonationsfrom firmsto thegovernmentaffectitsIPR policy,innovation, and welfare.Ourmodelissimilarto thatofEicherand García-Penalõsa(2008)in thatthefirmsmakeeffortsto protectIPR.However,weassumethatthegovernment decidesitsIPR protection taking into accounthousehold utility and theamountof political donation. This assumption is different from that in other literature. In general,politicaldonation isregarded asarent-seeking activity.Many researchers have investigated the relationship between rent-seeking activities and economic development.Oneview in theliteratureisthatrent-seeking activitieslowerthe growth rate,becausethey distorttheallocation ofresourcesand weaken capital accumulation and research activity³⁾.On theotherhand,politicaldonation aimed at strengthening IPR protection may enhancetheincentivesforresearch activity and promoteeconomicgrowth.In thispaper,weconsiderpoliticaleffortsby firmsand theimportanceofpoliticaldonationsin policy-making process.

Thispaperisorganized asfollows.In Section 2,weconstructthebasemodel withoutpoliticaldonations.In thissetting,weassumethatthegovernmentchooses itsIPR protection to maximizehousehold utility.Weshow thatwelfare-maximizing IPR protection is weaker than growth-maximizing protection. In Section 3, we introducepoliticaldonationsinto thebasemodel.In particular,weassumethatthe governmentpaysattention notonly to household’swelfarebutalso to theamount ofmoney offered by thefirms.In thissetting,weshow thattheamountofpolitical donations distorts optimal patent policy. In Section 4, we provide concluding remarks.

艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶 3)Forexample,based on an empiricalanalysis,Mauro (1995)findsthatrent-seeking activities

haveanegativeeffecton economicgrowth.Murphy,Shleifer,and Vishny (1991)arguethat rent-seeking activitiesrewardstalentmorethan entrepreneurship does.Thisimpliesthatrent- seeking activitiesdiscourageeconomicdevelopment.

－２７９－

Ⅱ　A Simple ModelofInnovation and IPR 1.Households

Weconsideraclosed economy with afixed numberofidenticalhouseholds.We normalizethetotalnumberofhouseholdsto unity.Weassumethateach household isendowed with Lunitsoflaborand suppliesitslaborinelastically.Households consumetwo kindsofgoods,homogenousgoodsand acontinuum ofdifferentiated goods. The representative household maximizes lifetime utility over an infinite horizon.Thelifetimeutility U^hisgiven by:

(1) whereρ isadiscountrate.D(t)isaconsumption index ofdifferentiated goodsat timetand Y(t)isconsumption ofhomogenousgoods.β ∈(0,1)isaparameter which determinestheproportion expenditureon differentgoods.Many kindsof differentiated goods exist in this economy and n(t) denotes a measure of differentiated productsinvented beforetimet.Each differentiated good isindexed by somerealnumberj∈[0,n(t)].D(t)isrepresented by aDixitand Stiglitz-type function:

(2)

whereand x(j,t)isconsumption ofthej^thvariety ofdifferentiated productattimet and α∈(0,1)isaparameterwhich determinesthepriceelasticity ofdemand.

Undertheseassumptions,thehouseholds’optimization problem can bebroken down into two stages.In thefirststage,thehousehold choosestheoptimaltimepath ofexpenditurein orderto maximize(1)subjectto thefollowing lifetimebudget constraint:

(3) whereE(t)isinstantaneousexpenditure,r(t)istheinterestrate,w(t)isthewage,

－２８０－

T(t)islump-sum tax,and W(0)istheinitialassetholding.In thesecond stage,the household determines how to allocate a given expenditure across differentiated goodsand homogenousgoods.By solving thesecond optimization problem,the demand functionsofdifferentiated goodsand homogenousgood aregiven by:

(4)

(5) Next,wereturn to thefirststageofhouseholds’optimization.To derivethetime path ofexpenditure,wedefinetheidealpriceindex ofD(t)as

(6)

Using thispriceindex,werewritetheflow ofutility aslogβ ^β(1－ β)^{(1－ β)}＋ logE(t)－ βlogPD(t). Therefore, the solution of the above dynamic optimization problem isgiven by thefollowing Eulerequation:

(7)

2.IPR Policy

In theliteratures,therehavebeen somewaysto formulateintellectualproperty rightin an economicmodel.Forexample,variousfeaturesofintellectualproperty and patentlegislation (namely,patentduration,patentbreadth,exogenousrateof imitation and cost of imitation) are considered and analyzed. In this paper, we assumethatthegovernmentchoosesan effortofitsenforcementpolicy,which is given by ω ∈[0, 1]. Precisely speaking,ω represents the probability that an invented good isprotected from competition by theenforcementofapatent.Oncea patentisenforced by thegovernment,thepatentholdercan enjoy theexclusiveright to produceand sellthegood.Forsimplicity,weassumethatthelifetimeofapatent

－２８１－

isinfiniteifthegovernmentdecidesto enforceapatent.Therefore,differentiated productsprotected by patentsexistin thefraction ofω and thosenotprotected in thefraction of1－ω.

To enforceω,thegovernmentmustemploy labor.A lump-sum tax collected from householdsisused to employ thelaborforces.In particular,weassumeγ ω ^θ units of labor are employed by the government in order to enforce ω. This representssocialcostofpatentenforcement.Moreoverweassumeθ ＞1,which impliesthatthemarginalsocialcostofpatentenforcementisincreasing.

3.Producers

Labor is the only factor of production. We assume that one unit of labor is required to produceboth differentiated and homogenousgoods.Thehomogenous good isalwaysproduced in competitivemarkets.Sinceahomogenousgood isthe numeraire,apriceofahomogenousgood isequalto theunitcostsofproducing it. Thisimpliesthatw(t)= 1,wherew(t)isthewagerate.

Firmsin thedifferentiated goodssectorproducetheirgoodsbased on designs created by R&D activity. Prices of products protected by patents are given as pm= 1/α,from thedemand (4),becausethepatentholdershavetheexclusiverights to produceand sellthosegoods.xm(t)denotesthedemand fordifferentiated good which areprotected by patentsand π(t)denotespatentprofits.Therefore,theprofit isgiven as

(8) On theotherhand,atechnology forproducing aproductisimmediately imitated if apatentisnotenforced fortheproduct.Thus,thepriceofunpatented differentiated goodsispc= 1,becausesuch goodsareproduced in acompetitivemarket.xc(t) denotesdemand foran unpatented products.Obviously,unpatented productsdo not provideprofitsto producers.

－２８２－

4.R&D Sector

In thispaper,new designsofdifferentiated goodsareinvented by research and development.Weletv(t)denotethevalueofanew design ofdifferentiated good.

Thisisequivalentto thesum ofthediscounted presentvalueoftheexpected profits from timet.Therefore,wehave

(9) By differentiating v(t)with respectto t,wederivetheno-arbitragecondition as

(10) Next,weconsiderthetechnology ofproductdevelopment.Ifafirm engaging in R&D activity employsLn(t)unitsoflabor,itcan producen ^・(t)unitsofnew designs ofdifferentiated goods,according to thefollowing knowledgecreation function:

(11) where a represents a parameter which determines the productivity of product development. We assume that firms can enter into R&D activity freely. Firms financethecostsofproductdevelopmentby issuing equities.Onceafirm succeeds in developing anew design ofadifferentiated product,thisdesign createsavalueof v(t)when itisprotected by apatent.Sinceω istheprobability thatan invented good isprotected by theenforcementofapatent,theexpected valueofanew design isgiven by ωv(t).However,a/n(t)unitsoflaborarerequired to inventanew blueprint.Sincethefree-entry condition requiresthatthevalueofpatentmustnot exceed thecostofproducing it,weobtain

(12)

5.NationalIncome

The final equilibrium condition equates savings with investment. The total incomeofthiseconomy consistsofwagefrom thelaborsupply and dividendsfrom

－２８３－

equities. As mentioned before, the measure of products protected by patents is ωn(t).Theaggregateincomeisgiven asw(t)L ＋ ωn(t)π(t).In thispaperwe assumethatthegovernmentcollectstaxesfrom householdsin orderto enforceits patent policy. Therefore, disposable income isw(t)L＋ ωn(t)π(t)－w(t)γ ω ^θ. Saving is given by the difference between disposable income and aggregate expenditure,E(t). Moreover, saving finances research investment through the financialmarket.Thus,wederivetheequilibrium condition asfollows:

(13)

6.The Equilibrium Path

In thissection wederivetheequilibrium path oftheeconomy.Dividing theboth sidesof(10)by v(t)gives

(14) Next,weconsiderthedemandsofdifferentiated goods.Using pc= 1 and pm= 1/α, thedemand function ofdifferentiated goodsprotected by patentsis

(15) Similarly,thedemand function ofunpatented differentiated goodsis

(16) By substituting (15)into (8),theprofitπ(t)isexpressed asfollows:

(17) Using thisexpression,wecan rewritetheno-arbitragecondition into (14)asfollows: (18) Herewedefinez(t)≡E(t)/(n(t)v(t)).By using (7)and (18),thechangein z(t)is expressed by thefollowing differentialequation:

－２８４－

(19) On theotherhand,dividing (13)by n(t)v(t)and using w(t)= 1 and (12)yields

(20) Substituting (20)into (19),wehave

(21) Thesteady stateoftheeconomy isdetermined so that^・z= 0.Weletz*denotethe valueofz(t)in thesteady state.Therefore,wecan derivez*asfollows:

(22) Thephasediagram ofthismodelisshown inFig.1.Wecan easily seethatthe uniquesteady state,z*,isunstablebecause1/ω ispositive.Sincewecan interpret z(t)astheinverseoftotalassetsmeasured by utility,z(t)isavariablethatcan jump.

Therefore,z(t)jumpsto thesteady statevaluez*in thiseconomy.Thisimpliesthat ourmodelhasno transitionaldynamicsand theequilibrium path jumpsto thesteady stateinstantaneously⁴⁾.

－２８５－

Next,wederivethegrowth rateofn(t)in thesteady state,g.Substituting (12) into theexpression ofz*given by (22),wederiveE(t)in thesteady state.Sincewe can show thattheexpenditureE(t)isconstantand dependson ω,thesteady-state valueofexpenditureisexpressed asfollows:

(23) Obviously,theexpenditureE(ω)isadecreasing function ofω.Ifthegovernment would like to strengthen patent protection, it has to collect more tax. This tax collection reduceshousehold’disposableincomeand expenditure.

Substituting (22)into (20)yieldsthegrowth rateofn(t)in thesteady stateas follows:

(24) Wecallgtherateofinnovation in thesteady state.Wearenow ready to analyze how thestrength ofIPRsprotection affectsinnovation and welfare.Sincetherateof innovation gisafunction oftheeffortofpatentenforcementω,wecan express g= g(ω).Differentiating gwith respectto ω yields

(25) Letω^gdenotethepatentprotection thatmaximizestherateofinnovation.From (25),ω^gmustsatisfy thefollowing relationship.

(26) Fig.2showstherelationship between therateofinnovation and patentprotection.

From Fig.2(a),wefind thatthereisan uniqueinteriorsolution ofω^gwhen thecost ofenforcing thepatentpolicy (γandθ)issufficiently large.Moreover,weeasily find that the growth-maximizing level of patent protection is larger when the 艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶 4)In otherwords,theequilibrium path which isnotin thesteady statecannotsatisfy rational

expectations.Fordetails,seeGrossman and Helpman (1991).

－２８６－

economy hasalargerpopulation (L),higherproductivity ofresearch activity and manufacturing (a)and lowercostofpatentprotection (γ)⁵⁾.

艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶艶 5)On theotherhand,therateofinnovation becomeszero when thelevelofpatentprotection is

too weak.Wecan derivean infimum in which therateofinnovation ispositivefrom (24). Here,wefocuson thecasewheretherateofinnovation ispositivebecauseweareinterested in an equilibrium path whereresearch and developmentisconducted.

－２８７－

7.GovernmentwithoutPoliticalDonation

Wenow analyzethegovernmentpolicy ofpatentprotection.Hereweconsider thecasewherethegovernmentisnon-corruptand maximizeshousehold utility U^h. In otherwords,thegovernment’sobjectivefunction Wisgiven by

(27) Forsimplicity,weassumethatthegovernmentsetsω attime0 and doesnotchange thispatentpolicy aftertime0.

Next, we derive the household utility as a function of the patent policy. Combining D(t)= βE(ω)/PD(t),(6),pc= 1,and pm= 1/α yields

(28) Substituting (28)and Y (t)= (1－β)E(ω)into βlogD(t)＋ (1－β)Y (t)yields

(29) By integrating (29)from 0 to∞with n(t)= n(0)e^{g (ω)t},wecan deriveU^hasfollows:

(30) whereΛ0isconstantand independentfrom ω.

Letω^b denote the level of patent protection that maximizes the welfare represented by (30). We find the condition thatω^b must satisfy by solving dW(ω)/dω = 0 asfollows:

(31)

Thesecond term of(31)showsthatdead weightlossincreaseswith thepatent protection.Thethird term impliesthattheexpenditureisadecreasing function ofω becausemoreresourcesareneeded to strengthen thepatentprotection.Obviously, the second and third terms of (31) are positive. Therefore, we can state the

－２８８－

following proposition.

Fig. 3 shows the relationship between IPR protection and welfare. When the economy is larger (largerL), the cost of enforcing patent protection is smaller (smallerγ),and theconsumerhasagreaterpreferencefordifferentiated goods (largerβ),thewelfare-maximizing levelofpatentprotection ishigher.

proposition 1Thewelfare-maximizing patentprotection isweakerthan thegrowth maximizing patentprotection.

－２８９－

Ⅲ　The EffectsofPoliticalDonation 1.Introducing PoliticalDonation

Next,weconsiderthepossibility oflobbying and politicaldonations.Weassume that firms which produce differentiated goods protected by patents engage in lobbying activitiesin orderto strengthen theirmonopolisticpower.In particular, firmsofferafraction oftheirprofitsto thegovernmentaslobbying activities.When thegovernmentiscorruptand willing to receivemoney,lobbying activitiesmay distortIPR policy.

In thispaper,each firm which producesdifferentiated goodsprotected by patents gainsaconstantfraction ofprofit,φ.Therefore,thefirm offersafraction (1－ φ) ofprofitto thegovernment.In thiscase,themarketvalueofanew design ofa differentiated good,(9),ismodified as

(32) From (32),itturnsoutthatthevalueofnew design issmallerascompared with the casewherethereisno politicaldonation.By differentiatingv(t)with respectto t,we derivetheno-arbitragecondition as

(33) Next,theequilibrium condition ofnationalincome,(13)mustbechanged.In this section, we have assumed that a constant fraction of profits is sent to the government.Therefore,householdsreceiveω φn(t)π(t)asadividend and (13)is modified asfollows:

(34) Next,weanalyzetheequilibrium path oftheeconomy.Sincewecan derivethe equilibrium path in thesameway asin Section 2,weomitthederivation ofthe equilibrium path.Asin Section 2,thiseconomy hasno transitionaldynamics,and jumpsto thesteady state.Wecan show thattheexpenditureE(ω)issameas(23).

－２９０－

On theotherhand,therateofinnovation in thesteady stateisgiven by

(35) Sincetherateofinnovation gisafunction oftheeffortofpatentenforcementω and φ,wecan expressg = g(ω,φ).Differentiating g with respectto ω yields

(36) Similarly,differentiating gwith respectto ω yields

(37) (36)impliesthatthegrowth-maximizing patentprotection ω^gisthesameasthe casewithoutlobbying.In otherwords,φ doesnotaffectthegrowth-maximizing patentprotection.On theotherhand,(37)ispositive.When thefraction ofpolitical donation ofprofitsissmaller,theincentiveto engagein research activity becomes higher.Thiseffectstimulatesinnovation in theeconomy.

2.Governmentwith PoliticalDonation

Next,weexaminehow thegovernmentchoosesitspatentpolicy when political donation exists. The government decides patent protection considering not only household utility butalso thegainsfrom politicaldonations.In thissection,the utility function ofthegovernmentisgiven astheaverageofhousehold utility and thesum ofthediscounted valueofpoliticaldonations.

(38) whereζisaparameterwhich determinestheextentofgovernmentcorruption.When ζissmall,thegovernmentvaluestheamountofpoliticaldonation.WedefineU^gas follows:

(39) By using (35),thetotalamountofpoliticaldonation isexpressed asfollows:

－２９１－

(40)

Substituting (40)into (39)and usingr(t)= ρon theequilibrium path revealsthatU^g issimplified as

(41) (30)and (40)imply that(38)isrepresented asafunction ofω.Letω^ddenotethe levelofpatentprotection thatmaximizesthewelfarerepresented by (38).Wefind thecondition thatω^dmustsatisfy by solving dW(ω)/dω = 0 asfollows:

(42)

whereΓ（ζ,φ）isdefined as

(43) Using (42)and (43),wecan examinehow politicaldonation affectspatentpolicy. First,wefocuson theeffectofζ .ζisaparameterthatdeterminesthepreferenceof government. A largerζimplies that the government gives much attention to household utility.Differentiating (43)with respecttoζyields

(44) (44)ispositive.Thisshowsthatwhenζislarge,thesecond and third termsof(42) becomes large and ω^d become smaller (SeeFig.4). This implies that if the governmentisnotcorruptand paysmuch attention to household utility,itchooses weakerpatentprotection.On theotherhand,acorruptgovernmentchooseshigher patentprotection.In particular,wecan show thatforany φ,Γ（0,φ）= 0.This impliesthatthesecond and third termsof(42)vanish and (42)correspondsto (26). Intuitively,when thegovernmentiscorruptand paysmuch attention to political

－２９２－

donations,ittendsto havetoo much incentiveto strengthen patentprotection in orderto receivealotofmoney from firms.Therefore,weobtain thefollowing proposition.

Next,weexaminetheeffectofφ.Theparameterφ representsthefraction of profitsthatispaid aspoliticaldonations.A largerφ impliesthatfirmsmakeefforts to offeralotofmoney to thegovernment.Differentiating (43)with respectto φ yields

(45) Therefore,thesign of(45)dependson therelationship betweenζandζ*,defined as (46) Thisshowsthat∂ Γ（ζ,φ）/∂ φ< 0 ifζ > ζ*.In thiscase,an increasein φ reduces proposition 2Ifthegovernmentiscorruptand paysmuch attention to political donations,thegovernmentstrengthenspatentprotection and increasesinnovation.

－２９３－

the second and third terms of (42), and the government chooses higher patent protection (SeeFig.5).On theotherhand,∂ Γ（ζ,φ）/∂ φ> 0 ifζ< ζ*.In this case, an increase in φ increases the second and third terms of (42), and the governmentchoosesweakerpatentprotection (SeeFig.6).Intuitively,wecan find two effectswhen φ increases.Firstly,when φ ishigher,an increasein ω raises therateofinnovation g(ω)moresharply.Thehigherrateofinnovation increases household utility and politicaldonation.In otherwords,an increasein φ raisesthe marginalbenefitofstrengthening patentprotection.Secondly,an increasein φ directly reducestheamountofpoliticaldonation.Thiseffectdecreasesthemarginal benefitofstrengthening patentprotection.When ζ > ζ*,thefirsteffectdominates and themarginalbenefitofstrengthening patentprotection increasesbecausethe non-corruptgovernmentconsidersthehousehold utility seriously.Thisimpliesthat whenζ ＞ ζ*, an decrease in φ, which corresponds to higher profit fraction allocated to political donation, only discourages innovation and the government choosesweakerpatentprotection.On theotherhand,when ζ ＜ ζ*,thesecond effect dominates, and the marginal benefit of strengthening patent protection decreasesbecausethegovernmentpaysmuch attention to politicaldonations.Asa consequence,thisimpliesthatwhen ζ ＜ ζ*,an decreasein φ raisestheamount ofpoliticaldonation and acorruptgovernmentchoosesstrongerpatentprotection.

Hence,wederivethefollowing proposition.

proposition 3Ifthegovernmentisnon-corruptand paysmuch attention to house hold utility (ζ ＞ ζ*), a higher fraction of political donations discou rages innovation and the government chooses weaker patent protection. On the other hand,Ifthegovernmentiscorruptand paysmuch attention to politicaldonation (ζ ＜ ζ*),a higherfraction ofpoliticaldonationsencouragesinnovation and the governmentchoosesstrongerpatentprotection.

－２９４－

－２９５－

Ⅳ　Concluding Remarks

Using an endogenousgrowth modelthatonly incorporatesinnovation asasource ofeconomicgrowth,thispaperhasexamined how politicaldonationsfrom firmsto the government affect its IPR policy, innovation, and welfare. Firstly, as a benchmark,wehaveassumed thatthegovernmentisnon-corruptand paysattention only to household utility and thatthereareno politicaldonations.In thismodel,we have shown that welfare-maximizing patent protection is weaker than growth- maximizing protection becausestrengthening patentprotection damageshouseholds by increasing dead weightlossand decreasing expenditurelevel.

Next,wehaveconsidered thecaseofpoliticaldonations.In thiscase,when the governmentiscorruptand paysmuch attention to politicaldonations,ittendsto havetoo much ofan incentiveto strengthen patentprotection in orderto receivea lot of money from the firms. Therefore, the government strengthens patent protection and increasesinnovation.Thisexplainsthepatentpoliciestend to be strengthened in countrieswherepoliticaldonation isimportantforpoliticians,such asin theUnited States.

On theotherhand,wehaveobserved two differenteffectswhen thefraction of profitsallocated to politicaldonationsishigher.When thegovernmentisnoncorrupt and paysmuch attention to household utility,ahigherfraction ofprofitsallocated to political donations discourages innovation and the government chooses weaker patentprotection.Thisisbecauseahigherfraction ofprofitsallocated to political donations weakens the incentive to engage in research and development. As a consequence, this effect decreases the marginal benefit of strengthening patent protection. On the other hand, when the government is corrupt and pays much attention to politicaldonations,ahigherfraction ofprofitsallocated to political donations encourages innovation and the government chooses stronger patent protection. This is because a higher fraction of profits allocated to political donationsraisestheamountofpoliticaldonation.Thiseffectenhancesthemarginal benefitofstrengthening patentprotection.In particular,thelatterresultisinteresting

－２９６－

becauseahigherfraction ofpoliticaldonation leadsto astrong IPR policy and a higherrateofeconomicgrowth in countrieswherepoliticaldonation isimportant forpoliticians.However,IPR policy in thiscasemay betoo high when weconsider the household utility. In other words, when political donation is important for policymaking,thegovernmenttendsto havean incentiveto choosean excessive levelofpatentprotection.

References

［　 ］1Aghion, P. and P. Howitt (1992)“A Mode, l of Growth Through Creative Destruction,” Econometrica,vol.60,323-351.

［　 ］2CentreforResponsivePolitics(2012)“Lobbyi, ng Database,”Availableat 　　http://www.opensecrets.org/Accessed May 15,2012.

［　 ］3Eicher,T.and C.García-Penalõsa(2008)“Endoge, nousStrength ofIntellectualProperty Rights:ImplicationsforEconomicDevelopmentand Growth,”European EconomicReview, vol.52,237-258.

［　 ］4Gilbert,R.J.and C.Shapiro (1990),“OptimalPatentLength and Breadth,”RAND Journal ofEconomics,vol.21,106-112.

［　 ］5Goh,A-T.and J.Olivier(2002)“Opt, imalPatentProtection in aTwo SectorEconomy,” InternationalEconomicReview,vol.21,1191-1214.

［　 ］6Gould,D.and W.Gruben (1996)“The, RoleofIntellectualProperty Rightsin Economic Growth,”JournalofDevelopmentEconomics,vol.48,323-350.

［　 ］7Grossman,G.M.and E.Helpman (1991),Innovation and Growth in theGlobalEconomy, MIT Press.

［　 ］8Grossman,G.M.and E.L.-C.Lai(2004)“I,nternationalProtection ofIntellectualProperty,” American EconomicReview,vol.94,1635-1653.

［　 ］9Iwaisako,T.and K.Futagami(2003)“Pa, tentPolicy in an EndogenousGrowth Model,” JournalofEconomics(ZeitschriftfürNationalfökonomie),vol.78,239-258.

［10］Judd,K.L.(1985)“On , thePerformanceofPatents,”Econometrica,vol.53,567-586.

［11］Klemperer,P.(1990),“How Broad Should theScopeofPatentProtection Be?”TheRAND JournalofEconomics,vol.21,113-130.

［12］Kwan,Y.K.and E.L.-C.Lai(2003)“I,ntellectualProperty RightsProtection and Endogenous EconomicGrowth”JournalofEconomicDynamicsand Control,vol.27,853-873.

［13］Lai,E.L.-C.and L.D.Qiu (2003),“TheNorth’sIntellectualProperty RightsStandard for theSouth?”JournalofInternationalEconomics,vol.59,183-209.

－２９７－

［14］Mauro,P.(1995),“Corruption and Growth,”QuarterlyJournalofEconomics,vol.110,681- 712.

［15］Murphy, K. M., A. Shleifer, and R. W. Vishny (1991)“The, Allocation of Talent: ImplicationsforGrowth,”QuarterlyJournalofEconomics,vol.106,503-530.

［16］Nordhaus, W. (1969),Invention, Growth and Welfare: A Theoretical Treatment of TechnologicalChange,Cambridge,MA:MIT Press.

［17］O’Donoghue,T.and J.Zweimüller(2004),“Patentsin aModelofEndogenousGrowth,” JournalofEconomicGrowth,vol.9,81-123.

［18］Rapp,R.and R.Rozek (1990)“Be, nefitsand CostsofIntellectualProperty Protection in Developing Countries,”NationalEconomicResearch Associates,Inc.Working Paper,No.3.

［19］Romer,P.M.(1990),“EndogenousTechnologicalChange”JournalofPoliticalEconomy, vol.98,S71-S102.

［20］Scherer,F.(1972)“Nor, dhaus’Theory ofOptimalPatentLife:A GeometricReinterpreta- tion,”American EconomicReview,vol.62,422-427.