The Value Of Patents On Innovation Economics Essay

Posted on


Patent right is a impermanent belongings right given to an person for an innovation. It is a right and non a warrant for non leting others to double or sell the merchandise. In exchange for these rights the person must uncover all the information about the patented merchandise. A patent can be granted for a new merchandise, a new procedure, new machinery etc. Patentable capable affair is much more varied in US as compared to other states. In Europe it is concentrated more towards industrial and proficient pertinence and does non be given towards countries such as concern methods, surgical methods and familial stuff as in the US. A patent is merely granted in exchange of some valuable information. In US it is granted by the US Patent and Trademark Office ( USPTO ) and in Europe by the European Patent office ( EPO ) . Once granted it can be used for trading like any other belongings. The lone difference being that this right is impermanent and lasts for a maximal period of 20 old ages.

So far few surveies have been conducted that allow for a non-linear consequence of patent rights on invention. In this paper I present such a survey. I have used domestic patent filings ( as my step for invention ) to prove the void hypothesis that patent rights have no consequence on invention. I examine the patent rights index in 39 states out of which 19 are developed and 20 are developing. Data for the strength of patent rights is drawn from the Patent Rights Index as set Forth by Park ( 2008 ) .

As an overview, my consequences illustrate that patent rights have non additive relationships with invention. The consequences differ for developed and developing states. They show that a low degree of patent rights affects invention negatively up to a certain point after which invention is affected positively.

This paper is organized in the undermentioned manner. To get down with, in subdivision 2, I present a reappraisal of the literature concentrating on the theoretical model followed by the results of old empirical surveies. In subdivision 3, I discuss the empirical methodological analysis used, outline the information beginnings and depict the different variables used. In the following subdivision I report the empirical consequences of my arrested development analysis and construe the coefficients of the selected variables. The last subdivision which is the decision summarizes the overall findings of my work and provides ideas for farther research.


As pointed out by Katharine Rockett ( 2009 ) Patent rights can assist in the advancement of scientific discipline in two ways –

The first manner is the wages the patent holder gets for the invention which is besides called the wages theory. The patent rights given to the pioneer give them monopoly power to bring forth that merchandise. Through this power they earn net incomes. If they did non hold this power they will hold no motive to bring forth the merchandise as other parties will merely utilize the information and double the merchandise. Hence the net incomes generated by the granting of patent rights promote invention. However we know that monopoly power consequences in dead weight loss. This can be a heavy cost for the society despite the fact that something is being invented.

The 2nd manner patent rights aid in the advancement of scientific discipline is through the information that the invention generates. This is socially utile as the information can be used by other pioneers to bring forth farther societal additions. Hence it is of import that equal wagess are given for the invention so that the information is available publically.

In the same paper she besides talks about inducements to introduce without patent rights.

Let us see a state of affairs where there are no patent rights. We allow different persons to double the invention after the merchandise is sold in the market and they become familiar with it. If the innovation generates net income there will be more providers of the advanced merchandise. This will go on because every person will desire to double the merchandise to do some net income, peculiarly if it is fast and inexpensive. If there are more providers in the market for the same merchandise, each provider will non be able to bear down a high monetary value as all consumers will travel to the provider bear downing the lowest monetary value. Therefore, the merchandising monetary value of the merchandise will fall down and therefore the net income of the original pioneer. If this net income is less than even the cost it took for the pioneer to develop the merchandise, there will be no motive for him to bring forth the merchandise as he knows that he will incur loss.

There have been many expostulations to the statement chiefly because of the premises it makes. It assumes that pecuniary addition is the lone incentive for innovation and it is really easy, fast and cheap to double the merchandise.

The statement here is that persons do non care about the pecuniary benefits they get, but they invent to portion it with different people. Apache web waiter which is an unfastened beginning package ( OSS ) runs about 50 % of the web sites of the universe broad web. It was developed by extremely skilled coders. These coders put in their cognition and clip for the successful operation of this package without being paid. Lerner and Tirole ( 2002 ) and Lakhani and Wolf ( 2005 ) suggest that Open Source codification developers frequently invent to develop their ain accomplishments. By affecting in OSS programmes workers signal their capablenesss to employees and hence likely secure a better occupation.

However, Boldrin and Levine ( 2005 ) province that there were inducements to introduce in the agricultural sector even without patent protection. Till the early 1970s invention in agribusiness flourished without any patent rights. Breeders developed different assortments of seeds and sold them to husbandmans. The husbandmans could sell and reproduce these seeds without being taken to the tribunal.

To analyze the effects of invention in agribusiness Boldrin and Levine use maize outputs in US averaged by decennary.

U.S. maize output is measured on the y-axis, and old ages is measured on the x-axis. We can see from the graph above that till the 1940`s there has been no alteration in the maize output. As noted by them this has really small to make with deficiency of invention and is chiefly because of the fact that during this clip agribusiness went west towards poorer clime and dirt conditions and hence in order to keep a stable output uninterrupted invention was required. The 1940`s and 1950`s show a enormous addition in output. The ground behind this is the invention of improved intercrossed assortments which can defy heavy fertilisation. The chief statement here is that there was an addition in yield through invention when no patent Torahs existed as the first patent jurisprudence for maize was recorded in 1974 by Pioneer-Hi-Bred.

“ If we did non hold a patent system, it would be irresponsible, on the footing of our present cognition of its economic effects, to urge establishing one. But since we have had a patent system for a long clip, it would be irresponsible, on the footing of our present cognition, to urge get rid ofing it. ” – Fritz Machlup ( 1958 )

Some people argue that the patent system is non effectual even though it may be necessary. Survey grounds of Cohen, Nelson, and Walsh ( 2000 ) has stated that pioneers do non acquire the full benefit for their invention. Some sectors such as biotechnology and pharmaceuticals may acquire the full benefit but many first mover advantages ( E.g. – acquisition by making ) do non acquire much benefit. This is chiefly because workers try to better their productiveness by on a regular basis reiterating the same action and through minor inventions. They may non contrive something considered to be valuable by the patent system. Carnegie study conducted in 2000 by R & A ; D research managers reports that the patent system is non that effectual. The study reports that it received responses from 1087 houses for procedure invention and 1118 houses for merchandise invention. The houses were asked whether the undermentioned methods were effectual in accomplishing maximal benefit from invention. The following tabular array shows the different methods and their effectivity in per centum. The Numberss in parenthesis represent effectivity for pharmaceutical and medical equipment industries severally. These are the industries in which respondents stated that patents were the most effectual.




51.00 % ( 53.57 % , 50.97 % )

50.59 % ( 68.13 % , 49.24 % )

Lead clip

52.76 % ( 50.10 % , 58.06 % )

38.43 % ( 35.52 % , 45.15 % )



45.61 % ( 49.39 % , 49.25 % )

43.00 % ( 44.17 % , 49.55 % )



42.74 % ( 33.37 % , 52.51 % )

30.73 % ( 25.21 % , 32.12 % )


34.83 % ( 50.20 % , 54.70 % )

23.30 % ( 36.15 % , 34.02 % )

Other legal

20.71 % ( 20.82 % , 29.03 % )

15.39 % ( 16.04 % , 22.27 % )

It can be seen from the above tabular array that secrecy, lead clip and complementary fabrication are the most effectual manner to acquire appropriate additions from invention. Patents and other legal agencies are least effectual. Merely one tierce of the respondents feel that patents are effectual. Even pharmaceutical and medical equipment industries which are used as illustrations to why patent Torahs are of import have merely half the respondents holding that patents are effectual. Besides we can see that other methods such as secretiveness and lead clip are every bit effectual as patents. Hence patents entirely are non a important factor to impact determinations on invention. In the following subdivision I will be speaking about how weak patent rights affect invention.

James J. Anton, Hillary J. Greene, and Dennis A. Yao ( 2005 ) see the policy deductions of weak patent rights. They province that patents may miss strength because they have a important chance of being overturned or are comparatively easy to besiege. Patents may be overturned because a careful scrutiny of patents is rather dearly-won, and if the patent is non economically valuable it will ensue in a waste of resources. Weak patent rights may ensue in houses trusting more on secretiveness in order to acquire more benefit from their invention. However this does non ensue in transportation of economically valuable information between viing houses about the way of invention activity and besides about strategic motions and information about their rivals.

Equally far as developed states are concerned Landes and Posner ( 2003 ) and Allred and Park ( 2007 ) argue that stronger patent rights will impact invention positively as it will ensue in increasing the grade of appropriability of the returns to invention. Invention may besides be affected positively because it allows for cognition revelation instead that maintaining it as secret. This cognition made available publically will surely assist other persons to introduce other merchandises and procedures. Since, a stronger patent system may let others to form and commercialize the market better for follow on invention, it may besides ensue in increased invention through licensing. ( Kitch, 1977 ) .

However, Ziedonis, ( 2004 ) argues that patent rights may impact invention negatively as some houses or persons do non portion the engineerings with other challengers for strategic defence intents. They might non desire to unwrap engineering which is needed for others to introduce so that they can acquire more net income from their invention. Another ground for patent rights to impact invention negatively as pointed out by Bessen and Maskin ( 2000 ) and Shapiro ( 2000 ) is that it leads to increased dealing costs in the market for technological exchange. It can be really hard for persons to acquire permission for patented engineerings particularly if different right holders ain different constituents of a engineering. Licensing and cross licensing can be really difficult to obtain and therefore increase the dealing costs.

As for developing states, Sieback ( 1990 ) states that invention is affected positively because of cognition revelation from patents and appropriability consequence if the strength of the patent rights index is low. This is because of the followers. The optimum degree of patent strength depends on the features of the market and engineering as stated by Grossman and Lai ( 2004 ) . They point out that in developing states there are normally less extremist inventions. Hence continuance of the patent need non be really long as there is non much invention investing to retrieve. Therefore there is a little market size, there is a low capacity to introduce and hence the patent protection should non be that strong otherwise these states will non be able to take the full benefit of the appropriability consequence due to decreased competition. Hence a lower patent rights index for developing states will ensue in increased invention. As noted by the The Commission on Intellectual Property Rights ( 2002 ) , the developing states need some kind of a technological base or a set of accomplishments in order to vie with the inventions of the developed states and if there is a strong patent rights index it will be hard for them to larn through imitation and therefore will impact invention negatively.

Previous Empirical work –

Petra Moser ( Assistant Professor Department of Economics – Stanford University Stanford ) states that patent length has no consequence on invention. She conducted a survey on patent Torahs act uponing invention in the nineteenth century in her NBER working paper figure 9909. She took informations from two exhibitions. One is the Crystal Palace exhibition in London in 1851 and the 2nd is Centennial Exhibition in Philadelphia, 1876.


We can see here that exhibit per capita does non depend on patent length. Switzerland with zero patent length has the 2nd highest figure of inventions ( exhibits ) . Switzerland adopted their first patent Torahs in 1907. We can besides see that Russia with 10 old ages of patent length has fewer exhibits per capita than comparing with Denmark who has zero patent length.

Similarly, Lerner ( 2002 ) conducted a survey to analyze 177 events of patent reforms in 51 states over a 150-year period. The reforms covered alterations in continuance of rights and fees, passage of patent Torahs and restrictions of patent rights such as compulsory licensing. His survey found out that patent reform did non impact residential patent filings. Branstetter ( 2004 ) conducted a survey similar to that of Lerner. He examined patent reforms in 12 developing states from 1982 to 1999 and found out that the consequence of patent reforms on residential patent filings was undistinguished.

As for the deductions of weak patent rights a survey by Mansfield, Schwartz and Wagner ( 1981 ) clearly suggests that patents were non strong plenty to forestall violation or avoid close circumvention. They examined 48 inventions and within the following 4 old ages 60 % of them were litigated. If the houses know that its invention will hold weak rational belongings protection so there will be no motive to put in invention ( R & A ; D ) .

In contrast Sherwood ( 1997 ) conducted instance surveies for 18 developing states and concluded that hapless proviso of rational belongings rights prevents local invention and risk-taking. Similar to this survey Varsakelis ( 2001 ) provides a cross-country analysis of patent strength and R & A ; D and concludes that R & A ; D/GDP ratios are positively related to the strength of patent rights

As for individual state surveies Scherer and Weisburst ( 1995 ) examined the Italian pharmaceutical industry and found out that that the Italian pharmaceutical invention activities were unaffected by patent reforms. However, in contrast the survey conducted by Korenko ( 1999 ) utilizing more recent information found out that stronger patent rights affect the Italian pharmaceutical industry R & A ; D positively.

Allred and Park ( 2007 ) behavior a survey to see the non additive consequence of patent rights on invention. They use two informations sets. First is the panel for national patent applications and second is the panel for steadfast degree research and development ( R & A ; D ) . Their consequences show that the non additive consequence of patent rights on invention does be and that the consequences differ for developed and developing states. Invention is affected positively for developed states, but for developing states at that place does non look to be a significantly positive consequence on invention.


I have used informations on international patent applications as my step for invention. I am cognizant of the fact that patent applications are non a direct step for invention. Not all inventions might be patented as in some instances the pioneer may take non to patent or the invention might non be patentable due to capable affair. Variations in patent filings may non precisely reflect fluctuations in invention even though each patent application must declare and stand for a new invention, nevertheless, they do reflect leanings to register. Hence, utilizing international patent filings to mensurate invention is surely justified.

International patent applications include domestic patent filings and foreign patent filings. Foreign patent filings in most instances are a good step for international engineering diffusion instead than invention. This is because most first filings occur in the place state, with attendant filings made in the foreign states ( Hingley and Park, 2003 ) . Hence, I have merely used domestic patent filings as my step for invention.

Now traveling on to the theoretical account estimated. Count data econometric methods need to be used to carry on the analysis as domestic patent filings are non negative whole numbers. I have used negative binomial arrested development to acquire the consequences. This arrested development method is better suited to response measurings that are counts and because of the satisfactory representation for the variableness observed in count informations it provides an analysis similar to that of the normal distribution in multiple additive arrested development.

Poisson arrested development, the other count informations method is non used because of the limitation it imposes, viz. equality of the conditional mean and the discrepancy of the random variable. As noted in Greene ( 2007 ) negative binomial arrested development is used to gauge count theoretical accounts when Poisson appraisal is inappropriate due to over scattering. If the discrepancy is greater than the mean the distribution is said to expose over scattering. If informations whose sample discrepancy is greater than its sample mean is estimated by the Poisson theoretical account with the same mean and discrepancy, it will ensue in over scattering. As pointed out by Cameron and Trivedi ( 1998 ) negative binomial arrested development can be considered as a more flexible signifier for over spread count informations. Quadratic specification characteristic of this theoretical account is frequently a really good estimate to a more general discrepancy map.

The Negative Binomial distribution is given by –

E ( Y ) = i?­

Var ( Y ) = i?­iˆ«iˆ?i?­2/k )

Let Pit denote the figure of patent filings. The expected counts are modeled as an exponential map of state specific effects ( i??i ) , which can either be fixed or random, and a set of independent variables Xit:

E ( Pit i??iiˆ¬iˆ Xit ) = exp ( i??i + i??iˆ±PRIit + i??iˆ?PRI2it + i??iˆ?i?sit ) ( 1 )


– Xit includes the patent rights index and the other control variables Z, to be discussed below.

– i??iˆ±iˆ¬iˆ i??iˆ?iˆ¬iˆ i??iˆ?iˆ are the coefficients to be estimated

– The error term is given by i??it = Pit – Tocopherol ( Pit i??iiˆ¬iˆ Xit )

Data –

I have collected informations for 2 old ages ( 2000 and 2005 ) for a sum of 39 states to see the consequence of patents on invention. The variables used and their beginnings are as follows –

Dependent variables –

Data for domestic patents fillings is taken from the universe rational belongings organisation statistics database. I have used direct residential filings to stand for domestic patenting. Direct occupant filings refers to an application filed with an office of the province or stand foring the province in which the first named applier of the application concerned has abode.

Independent variables –

Following are the independent variables used –

Patent Rights Index ( PRI ) – Data for the strength of patent rights is drawn from the Patent Rights Index as set Forth by Park ( 2008 ) . The patent right index ranges from zero to five, where nothing is the weakest and five is the strongest. There are five constituents which make up the PRI. They are –

( a ) Coverage constituent – Which measures the topic affair that is patentable.

( B ) Membership component – indicates the extent to which states are participants to major international pacts on patent rights.

( degree Celsius ) Loss of rights constituent – indicates the limitations from the use of patent rights such as compulsory licensing and working demand

( vitamin D ) Duration – Indicates the figure of old ages the invention can be patented.

( vitamin E ) Enforcement Mechanism – refers to the handiness to enforcement processs such as preliminary injunction entered by a tribunal which is fundamentally keeping a party to go on with a class of behavior until the instance has been decided.

All of these 5 constituents are measured out of one and are so aggregated to acquire the concluding belongings right index.

Control Variables – The other control variables that my theoretical account analyses other than patent right index are – existent GDP, authorities disbursement to GDP, mean one-year loaning rate and homo development index.

As pointed out by Barro and Sala-i-Martin ( 2004 ) in the growing theory literature that invention is besides a map of authorities size, human capital, wellness capital, and fiscal factors. We can utilize these variables to see their consequence on invention.

Real GDP -The information for this variable is collected from the World Bank ‘s World Development Indicators ( ) . I have used the natural log of existent GDP as a replacement for overall market size, which should impact inducements to patent.

Human Development index – The information for this variable is taken from the human development studies by the United Nations Development programme ( ) . As pointed out in this study, this index is a combination of three dimensions of human development – ( 1 ) life anticipation at birth which measures a long and healthy life, ( 2 ) Education which measures big literacy rate and combined gross registration ratio for primary, secondary and third schools and ( 3 ) Standard of life which is measured by GDP per capita ( PPP US $ )

Average one-year loaning rate – The information is from the World Banks World Development Indicators. Since mean one-year loaning rate reflects clip penchant rate it is an of import intertemporal variable for invention.

Government disbursement to GDP – The information is from the World Banks World Development Indicators. It is the authorities outgo for goods and services ( including compensation for employees ) . It besides includes outgo on security and national defence but excludes military outgo.

The following subdivision is devoted to describe and construe the obtained econometric consequences.


The chief inquiry that I will be discoursing here is the relationship between patent strength and domestic patent applications.

Turning to the estimations of equation ( 1 ) , I will be describing the estimations utilizing the fixed effects negative binomial arrested development. Table 1 gives us the relationship between patent right index and domestic ( occupant ) filings. I have included merely the log of existent GDP as my control variable in this analysis. Regression including all the control variables is done subsequently. Column 1 studies the additive consequence of patent rights on invention for all states and the column 2 studies the non additive effects for all states.

Table 2: Domestic Patenting and Patent Right Index

Variable ( 1 ) All ( 2 ) All ( 3 ) Developed ( 4 ) Developed ( 5 ) Developing ( 6 ) Developing

states states states states states states

Changeless -9.362 -9.071 -4.921 -4.867 -2.752 -3.748

( 1.052 ) ( 1.041 ) ( 3.901 ) ( 3.853 ) ( 2.051 ) ( 2.969 )

Patent Right Index 0.462 -0.891* 0.318 -0.796 0.425* 0.632*

( 0.217 ) ( 0.19 ) ( 0.167 ) ( 0.223 ) ( 0.082 ) ( 0.158 )

Patent Right Index2 – 0.108** – 0.092* -0.082*

( 0.014 ) ( 0.023 ) – ( 0.021 )

Ln ( GDP ) 0.241* 0.278* 0.162 0.227 0.172* 0.154*

( 0.061 ) ( 0.067 ) ( 0.064 ) ( 0.151 ) ( 0.044 ) ( 0.039 )

Log Liklihood -859.65 -846.13 -519.29 -505.38 -483.62 -480.91

Liklihood Ratio – 7.31** – 6.73* – 4.17*

Test Statistic

Number Of 39 39 19 19 20 20


Notes: Standard mistakes are in parentheses. The likeliness ratio trial is used to prove the statistical significance of the quadratic patent rights term. *P & lt ; 0.05, **P & lt ; 0.01

Looking at column 1 we can see that patent rights have a statistically undistinguished consequence on domestic patent filings. However, column 2 shows a statistically important consequence. Furthermore a likeliness ratio ( LR ) 1 trial rejects the void hypothesis that the coefficient of the quadratic patent rights term is zero. Hence we can state that patent rights do impact the degree of invention. Since i??iˆ?iˆ?iˆ°iˆ the conditional relationship between domestic patenting and patent strength is u-shaped2. This means that at low degrees of PRI a strengthening of patent rights will negatively impact domestic patenting. Once the critical point is reached a strengthening of patent rights will impact the degree of domestic patenting positively.

Typically developing states have a low degree of PRI. Strengthening of PRI for these states will impede invention. This supports the literature which has argued that a high degree of PRI does non let for agents to copy and copy. Harmonizing to my survey this relates to states with a low degree of PRI. For states holding a PRI higher than a threshold degree of strength, a strengthening of PRI will impact domestic patenting positively. Typically developed states have a high degree of PRI. This shows that for these states stronger patent inducements will increase invention. This is chiefly because of the fact that it will ensue in promoting cognition revelation and bettering appropriability. The critical turning point as we can see from column 2 is equal to 4.10. In my sample 24 % of the states are below this point and includes states such as India, Brazil and Mexico.[ 1 ]

The following four columns study the consequences for dividing the sample between developed and developing states. For each sample we estimate the theoretical account with and without the quadratic term for the patent right index. In column 3 we see that patent rights index is statistically undistinguished. However, in column 4 we can see that the LR trial rejects the void hypothesis of no non-linear consequence at the 5 % degree. The developed state sample shows a U-shaped relationship between domestic patenting and patent rights index, but the U-shaped relationship is non every bit important as the U-shape for the sample for all states. The chief ground for this is the absence of weak PRI states from this sample. The critical value for PRI for this sample is 4.32 and merely 4 states are below this sample. For Developing Countries we see an reverse U-shaped relationship between domestic patenting and PRI. This shows that the states with low degree or no patent systems, that is which are below the critical point of PRI have a positive consequence on domestic patenting and states above the critical point have a negative consequence on domestic patenting. A ground for this can be that these states mostly conduct adaptative and imitative research. LR trial rejects the void hypothesis that the coefficient of the quadratic term is zero at the 5 % degree.

Hence we see that by uniting both samples we get a U-shaped relationship between domestic patenting and patent rights index. The consequences show that the consequence of PRI on invention depends on the strength of the index in each state. Patent rights affect invention positively after a certain degree of the index strength has been reached.

For the log of GDP variable we see that it has a positive influence on domestic patenting for all the three samples. High GDP means high end product for the economic system. This implies higher net incomes and higher retained net incomes for houses. Since houses look for cost film editing techniques to increase net incomes farther, they will utilize these maintained net incomes to put in R & A ; D disbursement and therefore, will ensue in a positive consequence on invention and domestic patenting. As we can see from the tabular array the t-statistic is important at the 5 % degree for full sample and besides for the development states sample.

In Table 3, I repeat the analysis including the other control variables to see if this changes any of the consequences found earlier. I have merely shown the arrested development consequences of the theoretical account with the quadratic patent rights term in order to conserve infinite. The consequences are similar to the consequences found before. A likeliness ratio ( LR ) trial which supports a nonlinear consequence of patent rights rejects the void hypothesis that the coefficient of square of patent rights variable is zero for all the three samples. For the full sample and developed states sample the significance degree is higher than in the instance of the old arrested development. The likeliness ratio trial statistic is important even at the 1 % degree for both these samples. This besides means a stronger U-shaped relationship between patent rights index and domestic patenting. For developing states sample we see the consequences are similar to the old consequence. The likeliness ratio trial is still important at the 5 % degree.

Table 3: Domestic patenting and the Patent Rights Index ( Including all the control variables )

Variable All Developed Developing

States ‘ states states

Changeless -7.819 -3.217 1.624

( 6.235 ) ( 3.098 ) ( 4.366 )

Patent Rights Index -1.561** -2.052* 1.793

( 0.230 ) ( 0.436 ) ( 2.241 )

Patent Rights index2 0.194** 0.239** -0.256*

( 0.027 ) ( 0.035 ) ( 0.064 )

Ln ( GDP ) 0.644** 0.323 0.226*

( 0.161 ) ( 0.161 ) ( 0.053 )

Ln ( Government Spending/GDP ) 0.469 0.298* -0.274**

( 0.426 ) ( 0.075 ) ( 0.038 )

Ln ( HDI ) 1.787* 0.227* 1.138

( 0.435 ) ( 0.055 ) ( 1.442 )

Ln ( Lending Rate ) 0.211 0.146* -0.539

( 0.134 ) ( 0.031 ) ( 0.270 )

Log Likelihood -832.12 -584.96 -466.51

Likelihood Ratio Test 6.94** 6.11** 4.01*


Number of Observations 39 19 20

Notes: Standard mistakes are in parentheses. The likeliness ratio trial is used to prove the statistical significance of the quadratic patent rights term. *P & lt ; 0.05, **P & lt ; 0.01

Looking at the control variables we see that the log of GDP variable generates the same consequences as earlier. However the T statistic is important at the 1 % degree. Log of HDI has a positive influence on domestic patenting. The ground for this is that a high HDI for a state consequences in more old ages of schooling, a high criterion of life and high life anticipation at birth for persons in that state. This implies that these persons will hold more instruction and experience every bit compared to states with a low HDI. This will surely ensue in a higher invention attempt. The t statistic is important at the 5 % degree for the full and the developed states sample.

The mean loaning rate affects innovation positively for the full sample. But if we look at the developed and developing states individually we see that it affects domestic patenting negatively for the development states sample. A ground for this can be the high loaning rates for these states as compared to the developed states. A high loaning rate would stand for a tightening of funding for invention as people will happen it expensive to borrow, therefore ensuing in a lower invention attempt.

Ln ( Government spending/GDP ) affects domestic patenting negatively. This is consistent with the fact that authorities size reflects revenue enhancement load or herding out effects. However, if we look at the developed states sample we can see that authorities spending/GDP affects domestic patenting positively. A possible ground for this can be that this variable incorporates multiplier effects of authorities disbursement on demand which motivates houses to introduce. Government is a big pioneer in developed states ( via support of defence, universities, and private industry ) .A Hence, a big public sector would be associated with a batch of publically financed innovation.A For a big public sector that is mainly pensions and wellness attention related, this could besides bespeak a strong demand pull for pharmaceuticals and medical equipment industries – major invention classs which are considered as illustrations to why patent Torahs are effectual.

I will be stoping this subdivision by comparing my consequences to Allred and Parks ( 2007 ) paper. It is surely hard to compare accurately to their consequences because it differs in footings of sample size, sample period, control variables and empirical methodological analysis. However after seting for a few of the methodological differences we can see that my consequences are consistent with their findings.

Table 3: Appraisal of patent theoretical account utilizing Allred and Park ( 2007 ) sample –

Dependant variable Patent Right index Patent Right Index2 Number Of Likelihood ratio

Observations Test statistic

( Full Sample )

Domestic Patent Filings 0.109 ( 0.082 ) 491

Domestic Patent Filings -0.744 ( 0.212 ) *** 0.158 ( 0.036 ) *** 491 16.80***

( Developed Countries )

Domestic Patent Filings 0.238 ( 0.126 ) * 184

Domestic Patent Filings -0.746 ( 0.401 ) 0.147 ( 0.057 ) ** 184 6.00**

( Developing States )

Domestic Patent Filings -0.256 ( 0.111 ) 307

Domestic Patent Filings -0.158 ( 0.3911 ) -0.022 ( 0.086 ) 307 0.20

Notes: Standard mistakes are in parentheses. The likeliness ratio trial is used to prove the statistical significance of the quadratic patent rights term. *P & lt ; 0, 05 ; **P & lt ; 0.01 ; ***P & lt ; 0.001.

I have merely reported the coefficient estimations and standard mistakes of the patent rights index variable in order to conserve infinite and to aggressively concentrate on the non additive consequence of patent protection.

Looking at the consequences for the full sample we see that their there is an undistinguished additive consequence of patent rights on domestic patent filings but a important non-linear consequence. This is similar to my consequences but merely at different significance degrees. Hence we get a similar U-shaped for the full sample as described earlier ( Table 1 ) .

For the developed states we see that there is a important additive and non-linear consequence of patent rights. The likeliness ratio trial statistic rejects the void hypothesis of no non additive consequence at the 1 % degree in both instances. Hence one time once more we get a similar U-shaped form for the developed states sample as described earlier.

However, mentioning to the developing states sample we see an undistinguished linear and non-linear consequence of patent rights on domestic patent filings. This is opposed to my findings of a important additive and non-linear consequence of patent rights. A ground for this is the long sample period and the big sample size which Allred and Parks paper has chosen. My consequences for this sample explain that if developing states start with zero or low patent system, invention will increase up to a certain critical degree after which patent rights will hold a negative consequence on invention. However, subsequently in their paper when Allred and Park include all the control variables to see the consequence of patent protection on domestic patent filings they get consequences for the development states sample similar to that merely described.

Equally far as control variables are concerned the consequence for the log of GDP variable is similar to that of Allred and Park. For the mean loaning rate variable, their paper reports a negative consequence on invention sing the full sample, which is opposed to my findings of a positive consequence. The ground for this can be that Allred & A ; Park have more underdeveloped states in their sample, which have relatively higher loaning rates than developed states. Due to high loaning rates, people find it expensive to borrow which consequences in tightening of the funding for invention and therefore ensuing in a lower invention attempt.

I have included the ‘human development index ‘ variable alternatively of the ‘years of schooling ‘ and ‘life anticipation at birth ‘ variables used by Allred and Park ‘s paper. We see that this variable has a positive consequence on invention. A high HDI consequences in more old ages of schooling, a higher criterion of life and a higher life anticipation at birth for persons. This means that persons populating in these states will hold more experience and instruction, which will surely ensue in a higher invention attempt. Hence we see that HDI as a control variable affects invention positively.


This paper has aimed to happen an empirical relation between patent strength and invention. Rather than analysing a individual state to see this consequence, I use 39 states, out of which 19 are developed and 20 are developing.

We see that that the consequence of Patent Rights Index on invention depends on the strength of the index in each state and that the non additive consequence of patent rights does be. For the full sample I get a U shaped relationship between domestic patenting and the patent rights index. This shows that patent protection affects invention positively at high degrees and affects invention negatively at low degrees.

Since, typically developed states have high degrees of patent strength and in the sample most of the developed states are above the critical degree, it shows that it will impact invention positively for these states. The ground for this is that a stronger patent system allows for cognition revelation instead than merchandise secretiveness. Another ground is that it allows pioneers to form and commercialize the market better for invention.

Developing states normally have a low degree of patent protection. This shows that for these states invention will be affected negatively. The ground for this is their inability to copy and accommodate. However, if these states start with nothing or a low patent strength invention can be affected positively but merely up to a certain degree.

The restriction that my paper has which suggests the demand for farther research is the simplicity of my step for advanced end product. Due to miss of clip I was required to utilize domestic patent filings as my step for invention. Ideally, I would hold liked to analyze a broad assortment of steps such as entire factor productiveness growing and R & A ; D disbursement. I would hold besides been able to place other effects had I collected informations for more old ages. However despite these restrictions my overall findings summarize that patent protection affect invention positively for developed states and affects negatively for developing states.