## Development Growth System

The development of endogenous growing theoretical accounts enabled the finding of a state ‘s economic growing through the economic system itself.

In the original Solow Model, for case, the growing rate at the balanced growing way is given through the growing rates of both labour force and technological advancement, both exogenic to the economic system. ( Jones [ 1979 ] , Aghion and Howitt [ 1998 ] , and Romer [ 1996 ] ) For farther penetration on endogenous growing theoretical accounts see Romer [ 1990 ] , Grossman and Helpman [ 1991 ] , and Aghion and Howitt [ 1998 ] .

Among those theoretical accounts one can mention to the Schumpeterian attack which introduced perpendicular inventions by integrating a existent phenomenon: new inventions make old engineerings obsolete.

The Schumpeterian Model of Endogenous Growth takes proficient advancement ( inventions ) as a chief determiner to explicate economic growing. Such theoretical account assumes that the arrival rate of inventions is random harmonizing to a Poisson Distribution.

On the Poisson distribution see Stevenson [ 1978 ] .

This hypothesis allows that, even in the presence of low rates of technological advancement ( and hence low income growing ) , economic systems can rush up their growing and better their place in the universe ranking of income per capita. However this hypothesis does non explicate the inactiveness in long tally growing rates that sustain positioning stableness in that ranking for certain economic systems. Empirical facts suggest an account that include a deterministic constituent or tendency in add-on to the random constituent for the invention creative activity procedure.

One could believe that the stock of physical capital affects research productiveness. Henceforth, the larger the economic growing ( and its investing rates ) the larger the capital stock and hence the faster the arrival rate of inventions that will hike the rate of economic growing, and so on. Nevertheless, correlativity between capital stock and arrival rate of inventions ( and economic growing ) is non clearly supported by empirical facts: since the morning of capitalist economy one has seen several states, in several minutes in clip, roll uping big sums of capital without any direct reflect on growing rates in subsequently decennaries.

The former should be able to explicate an crystal clear fact: the larger ability for developed states to bring forth more technological inventions vis-a-vis developing states.

Harmonizing to the New-Schumpeterian literature inventions do present both a tacit and local character. Dosi et Al. [ 1994 ] argue that engineerings can non be merely copied and used in any topographic point. Windows of chance for catch-up require that states invest in infra-structures and institutional alterations for a long period of clip before such engineerings ‘ full potency can be exploited. Without old investings in instruction, R & A ; D and other scientific and proficient activities, really small can be achieved through import of engineering.

See Fagerberg [ 1994 ] , pp. 1156 for farther penetration.

If one assumes, as it seems to be supported by empirical facts, that the organisational and institutional environment are relevant in finding technological advancement, so such “ national systems of invention ”

The construct of “ national systems of invention ” was created by the New-Schumpeterian literature in the 1980s through a mix of historical informations, theoretical developments and empirical analysis. Basically it consists of an institutional building that enhances technological advancement in complex capitalist economic systems, through a system that joins together flows of scientific cognition and information which are necessary to the invention procedure. These institutional agreement involve houses, authorities bureaus, universities, R & A ; D sections, etc. and these articulate with the educational, fiscal, industrial and concern sectors such that creative activity and diffusion of technological invention is possible.

are more acute to make inventions and therefore in pulling capital flows.

This statement is equal to the post-Keynesian position – if the expetect returns in on domestic investing are high comparative to old old ages, but they are overcome by other economic systems ‘ investings, capital will flux outwards. This theory applies non merely geographically but besides in a clip position. An economic system that consistently presents higher returns on investing will pull foreign capital in all phases of its growing rhythm, even though penchant for liquidness will be low in this instance. In this sense, international capital flows are non merely a beginning of instability but besides divergency between economic systems ( Dow [ 1986/87 ] ) .

Therefore, investing rates increase in states where such an environment is more developed, taking to a geographical bunch of certain types of industries. These bunchs favour interaction between providers of technological parts and constituents, merchandises, users, related industries, etc. This interaction and networking stimulates technological invention every bit good as the creative activity of new goods and procedures because, inter alia, it takes advantage of scientific cognition and information flows which are necessary to the invention procedure ( Dosi et al. [ 1994 ] , Lundval [ 1998 ] ) . Therefore, this interaction is cause and effect of investing and geographical bunch of economic activities in some topographic points.

Investing in physical capital, though non a sufficient status, it is a necessary one to invention and technological advancement which favours the development of the “ national system of inventions ” and vice-versa ( Nelson [ 2005 ] , Dosi et al [ 1994 ] ) . Common interaction between the development of the “ national system of inventions ” and proficient advancement has been studied, e.g. by Bernardes and Albuquerque [ 2003 ] .

In this sense there is a virtuous rhythm that introduces a tendency constituent into the proficient advancement in some state ( or part ) : geographical bunch of economic activities favours technological invention and investing, which at the same clip stimulates geographical bunch and overall betterment of the “ national system of invention ” .

De facto, inventions occur chiefly in parts that are characterized by geographical bunch of companies or industries. For illustration, we have the Sillicon Valley in the USA, manner industry in Milan, fiscal services in London, inter alia. Porter [ 1990 ] , reports a figure of success instances in occupation creative activity, income boosting, international fight, related to economic activities located in some peculiar geographical infinites around the universe.

Castro and Gonalves [ 2002 ] besides try to explicate the being of a tendency constituent in proficient advancement. They show in a stochastic growing theoretical account that the interaction between two stationary procedures can ensue in a bi-modal distribution of income per capita. The Poisson procedure that describes the arrival rate of innovations has its parametric quantity determined by the figure of workers in the R & A ; D sector, while their handiness is given by graduated table additions from the productive sector. Credence by consumers of a new merchandise is a stationary procedure that determines possible graduated table additions. Consequently, the invention procedure is linked to the society ‘s ingestion form and therefore the 1s more prone to absorve newer goods are more acute to introduce faster. Through simulations they find a Markov Switching matrix with a dominant chief diagonal together with elements purely less than one, i.e. , there is the non-zero chance for an economic system to leap from one income group to another, nevertheless it is more likely to remain in its existent place.

Another account for the income per capita ranking inactiveness is given in Krugman [ 1991, 1993 ] . He shows that the relation between imperfect competitory constructions, increasing returns to scale and geographical bunch generate positive externatilities that stimulate investing and economic growing rates. Porter [ 1990 ] besides elaborates the statement that economic bunch increases the rate of return on investing.

In Romer [ 1990 ] the rate of creative activity of new thoughts may hold a positive or negation relation with the figure of inventions occurred up to that point in clip. In the first instance, the larger the stock of new thoughts the higher the chance of new inventions – mentioning Isaac Newton, Jones [ 2000 ] writes: “ If I got farther than other it was because I was standing on the shoulders of giants ” . The 2nd instance refers to the fact that easy and obvious thoughts are more prone to be discovered foremost. Subsequent thoughts are more and more hard to be generated. Therefore, the research worker productiveness would increase ( lessening ) through clip if, following the old logical thinking, the positive ( negative ) consequence dominates the negative ( positive ) one.

However, one should oppugn the followers: why are n’t income additions equally disseminated across the economic infinites, either states or parts? Using the same analogy as above, why was merely Newton who stood on the shoulders of giants and non all research workers at his clip? The theoretical account proposed in the paper takes geographical bunch of economic activities as a relevant factor to productiveness growing of R & A ; D activities.

In this instance, the random character of inventions would non be eliminated: the addition in geographical bunch favors research end product which can either hold a positive success in bring forthing new inventions or non. However, the higher the R & A ; D productiveness, the higher the chance of new inventions occur. That is, in add-on to the random arrival rate of inventions harmonizing to a Poisson distribution map, invention besides present a tendency constituent associated to the enlargement of the “ national system of invention ” .

Technical advancement has a local and silent character. Therefore, parts and states do non show the same ability to accomplish productiveness additions from their research workers, in the sense that these geographical bunchs are non uniformly spread.

Finally, there is non a limited or finite stock of obvious thoughts. On the contrary, the inventions ‘ history suggests that their skyline is infinite as it is human creativeness. In this sense, if there is an reverse relation between the figure of occured inventions till the present minute and the creative activity of new thoughts, this relation is more than compensated by the positive relation in developing geographical bunchs, accretion of thoughts and R & A ; D productiveness additions.

This paper is organized as follows. We begin, in Section 2, by showing general characteristics of Endogenous Growth theoretical accounts through a brief overview of its chief subscribers ‘ accomplishments. Then we summarize the basic Schumpeterian Growth theoretical account a la Aghion and Howitt [ 1998 ] which constitutes the workhorse for our chief freshness in this paper. In Section 2.1 is our chief intent to supply a theoretical extension to the basic Schumpeterian Growth theoretical account to account for the above mentioned phenomenon ( unite a random constituent together with a deterministic tendency to explicate the arrival rate of inventions over clip ) , which is complemented by exemplifying simulations for five ad-hoc economic systems over a clip period of 150 old ages. As it is every bit of import to supply some empirical grounds we aim at proving the part of invention and technological advancement ( through two different placeholders: patents per 100.000 dwellers and a Intellectual Property Rights Index a La Ginarte and Park [ 1997 ] ) in explicating different growing rates of income per capita around the universe. In this sense, we departure from the strictly theoretical attack to research some of the determiners of per capita income growing, by utilizing a panel dataset of 73 states between 1980 and 2005. Our last exercising in Section 4 consists in taking a clip series between 1963-2005 to measure the being of a long-term relationship ( through cointegration trials and Granger causality analysis ) between existent GDP growing and the entire figure of patents ( registered in the US market ) in Australia. One is able to happen a big literature on the right measuring of the invention procedure and technological diffusion, nevertheless no consensus has yet been found. We expect to lend with an extra attack to this on-going argument and hence aid to lend to policy devising towards income convergence.

Endogenous Growth Models – an overview

The documents of Aghion and Howitt [ 1992 ] , Grossman and Helpman [ 1991a ] and Romer [ 1990 ] are the best good known for their expressed treatment of the theoretical theoretical accounts of endogenous technological advancement. These theoretical accounts generate Steady-State equilibria in which the rate of technological advancement is affected by economic inducements to the R & A ; D sector. Such theoretical accounts relate policies that allow for a non-competitive supply behavior and economic growing encouraged by technological invention.

Harmonizing to Jones [ 1995 ] all R & A ; D theoretical accounts portion the alleged “ scale-effect ” , i.e. , if the degree of R & A ; D resources – measured, e.g. , by the figure of scientists – doubles, so the Steady-State growing rate of per capita income should besides duplicate. However, the “ scale-effect ” anticipation is non consistent with empirical grounds.

The kernel of the “ scale-effect ” on Romer, Grossman-Helpman and Aghion-Howitt theoretical accounts is summarized through the equations:

in which Y is income, A is productiveness or cognition and K is physical capital. Labour is used non merely to bring forth end product ( L_ { y } ) , but besides to detect new thoughts ( cognition ) ( L_ { A } ) . Equation ( 1 ) is the production map and ( 2 ) it is the beginning of the “ scale consequence ” . The latter implies the entire factor productiveness growing will be relative to the entire figure of workers in the R & A ; D sector. One consequence normally found in R & A ; D based theoretical accounts it is that with a changeless portion of labour engagement in that sector, such rate is relative to the size of the labour force.

The anticipation that the economic system ‘s growing rate is relative to the size of its labor force has been mostly refuted in recent literature. Historical experience is the most obvious grounds against such claim. Labour force has increased dramatically over the last 25, 50 or even 100 old ages, but the growing rate of the bulk of states has been kept changeless ( Jones, 1995 ) .

Jones seeks to build a plausible theoretical account that would extinguish this anticipation every bit good as maintain the other features of endogenous theoretical accounts. However, by extinguishing the “ scale consequence ” , Jones altered some of the long term deductions of the Solow ‘s growing theoretical account. In other words, per capita long term growing depends merely on parametric quantities that are, normally, exogenic and therefore alterations in policies – such as R & A ; D or capital accretion subsidies – do n’t hold any consequence on growing. More specifically, the Steady-State growing rate depends on the growing rate of inventions which is itself dependent from the exogenic population growing rate and hence it reflects the connexion between invention and research workers – an innovation demands an discoverer. However, as in the theoretical accounts of Romer, Grossman-Helpman, and Aghion-Howitt, and perversely to Solow ‘s, growing is endogeneized through R & A ; D.

These consequences suggest a polish to the term “ endogenous growing ” . In Jones theoretical account, growing is endogenous in the sense that technological advancement which generates long term growing consequences from R & A ; D activities done by net income maximizing agents. However, long term growing is non endogenous, as it was in the AK theoretical account or in the theoretical accounts by Romer, Grossman-Helpman and Aghion-Howitt, in the sense that traditional alterations in policies had existent consequence. Hence, harmonizing to Jones, one should be utilizing the nomenclature of “ semi-endogenous ” growing to mention to his theoretical account.

Harmonizing to Jones, an of import option modeling attack for the R & A ; D equation – in which cardinal consequences from Romer, Grossman-Helpman and Aghion-Howitt would be kept without enforcing “ scale effects ” – assumes that entire factor productiveness depends non on the measure of workers, but alternatively on the labour engagement in the R & A ; D sector:

One can easy see by looking at equantio ( 3 ) that R & A ; D determines the growing rate of entire factor productiveness and that R & A ; D subsidies ( by increasing I? ) , will increase the growing rate at the Steady-State degree. However, this specification is, for several grounds, unsatisfactory. First, ( 3 ) is inconsistent with the microfoundations that define the R & A ; D theoretical accounts as developed by Romer, Grossman-Helpman and Aghion-Howitt. These microfoundations imply that new thoughts are discovered by persons and that on these evidences the entire figure of inventions is inequivocally linked to the figure of people making R & A ; D. A specification that disregards “ scale ” such as ( 3 ) possesses the unreal deduction that one economic system with merely one unit of labor can bring forth at least the same measure of inventions as another with one million units.

Apart from this “ graduated table ” job, R & A ; D based theoretical accounts are intuitively really attractive. After all, in such theoretical accounts growing appears as a consequence of inventions that come out deliberately from rational agents that maximize net incomes and, even more significantly, they possess consistent microfoundations. Due to their attractivity it is desirable to happen a theoretical account that would maintain the basic construction of these theoretical accounts and at the same time would extinguish the “ scale consequence ” .

Romer, Grossman-Helpman and Aghion-Howitt define A as the stock of cognition or engineering in some economic system. Knowledge is merely the accretion of thoughts developed by people. In Jones ‘ simplest theoretical account, alteration in cognition, A, is equal to the figure of people engaged in detecting new thoughts times the rate at which the R & A ; D generates new thoughts.

The rate at which scientists or research workers find new thoughts is a map of the stock of cognition in the economic system. If one assumes a positive spillover consequence on the production of cognition so I? should be an increasing map of A.

Most likely, the most obvious thoughts are discovered foremost, hence the opportunity of a research worker to happen new thoughts is diminishing with the stock of cognition. On these evidences, Jones [ 1995 ] argues that the arrival rate of inventions is:

In this last equation the rate of invention lessenings with the degree of cognition when I† & lt ; 0. If I† & lt ; 0 so we face a positive outwardness. The arrival rate of new thoughts is independent from the stock of cognition when I†=0. The parametric quantity I† measures the grade of outwardnesss in the R & A ; D procedure.

Finally, if one considers the possibility of similar findings are taking topographic point at the same clip so the figure of inventions created by L_ { A } will be smaller than ab initio predicted. In this sense, L_ { A } will be replaced, in the R & A ; D equation by L_ { A } ^ { I» } , with 0 & lt ; I» & lt ; 1.

Taking this last alteration into ( 4 ) and ( 5 ) we get:

In equilibrium L_ { A } =l_ { A } . l_ { A } captures the duplicate consequence in the R & A ; D procedure.

Due to the usual cognition features – non-rival and non-excludable good – we are of course guided to theoretical accounts with technological spillovers, which are really of import in the growing literature. Informations that discoverers can non avoid being made available to the public sphere are extremely important for the creative activity of new thoughts. Harmonizing to Grossman and Helpman [ 1991b ] technological spillovers occur in two cardinal ways: 1 ) houses can get informations from others in collectible market minutess ; and 2 ) the Godhead of new information does non possess the legal resources to avoid its usage by others.

Several mechanisms exist for the extension of technological spillovers. Some informations might be acquired through the simple review of the merchandise that was generated from that invention ; others are made available to the populace by the discoverer that does non implement rational belongings rights.

Barro and Sala-i-Martin [ 1997 ] develop a theoretical account with similar features to Solow and Cass-Koopmans ‘ theoretical accounts. As mentioned before, new growing theoretical accounts try to work with endogenous technological advancement. However, Barro and Sala-i-Martin do do a nexus between the deductions of recent theories of long tally growing and the deductions of neoclassical growing theoretical accounts convergence hypothesis, by utilizing a theoretical account with technological diffusion. Harmonizing to the writers, long term growing depends on the find of new merchandises or engineerings in developed states. A relevant feature in the theoretical account is the fact that imitation is cheaper than invention, which makes more likely for states to prefer to copy alternatively of introducing. Henceforth, the low cost of copying suggests that followings should turn at greater rates and be given to catch-up the technological frontier leaders. As the figure of inventions likely to be copied lessenings, so it does with the economic system ‘s growing rate. The function of the increasing cost of imitation is similar to the diminishing returns to capital premise in the neoclassical theoretical account and this fact besides ensures besides a conditional convergence mechanism.

Harmonizing to the writers, due to market imperfectnesss and a low degree of right of first publication protection, the publication of new finds is non Pareto Optimal. The deficiency of ordinance on rational belongings rights does non promote taking states to farther innovate, but it stimulates followings to copy. Barro and Sala-i-Martin besides worked on issues like revenue enhancement, infra-structure and governmental policies on rational belongings rights which they argued as cardinal determiners of the research end product. We proceed we a speedy description of the basic Schumpeterian Growth theoretical account a la Aghion and Howitt [ 1998 ] which is traveling to be the footing substructure for our theoretical extension and some simulations.

The basic Schumpeterian Endogenous Growth Model

Harmonizing to the Schumpeterian growing model the invention procedure takes topographic point when the innovation of a new intermediate good replacements and overcomes an older one merchandise, thereby increasing the technological parametric quantity, A, by a changeless factor I? & gt ; 1 ( the alleged procedure of “ Creative Destruction ” ) . In this theoretical account we have a fixed gift of labor input, L, which has two viing utilizations: it can either be allocated to the production of new intermediate goods, or alternatively in R & A ; D on the production of new thoughts.

In Aghion and Howitt [ 1998 ] such theoretical account is clearly and extensively presented in Chapter 2. For the intents of this paper we will merely chalk out the chief equations and economic intuitions behind the theoretical account. This theoretical account, even without population growing, captures the indispensable feature presented throughout their book, i.e. , the fact that new goods and production procedures will of course get the better of their less productive ascendants. The theoretical account has some of import chief equations that will be discussed below. We start by the typical homogeneous single public-service corporation map, given by:

which translates a additive intertemporal penchant individuality.

Where, y=A.x^ { I± } is the production map of concluding ingestion goods, x is an intermediary good and A is a technological parametric quantity ( 0 & lt ; I± & lt ; 1 )

Furthermore, the economic system ‘s restraint, where the merely productive factor is labour, is L=x+n, with x and n the sum of workers used in the production of intermediate goods and research activities, severally. Furthermore, assume that one worker, x, generates a sinlge unit of the intermediate good, x.

The following important equation is the usual arbitrage status w_ { T } =I».V_ { t+1 } such that w_ { T } is the pay rate, T bases for the figure of inventions that indiscriminately arrive harmonizing to a Poisson chance procedure with arrival rate n.I» . The parametric quantity I» indicates the research productiveness in engineering and V_ { t+1 } is the expected price reduction pay-off of invention at clip t+1.

From the manufacturer side we besides have the plus equation

where Iˆ is the manufacturer net income of the intermediate good ten.

With this scene, Aghion and Howitt [ 1998 ] so derive the demand for the intermediate good every bit good as the dynamic equations of the theoretical account which will be used to calculate the Steady-State equilibrium. So, the demand for the intermediate good is given by:

The manufacturer net income outputs:

Finally, we end up with two stage line equations that describe the balanced way of this dynamic theoretical account where

is the arbitrage equation such that I‰_ { T } =w_ { T } /A_ { T } is the productiveness adjusted pay rate.

The 2nd 1 is the labour market equilibrium equation:

with x_ { T } =x ( I‰_ { t } ) a diminishing map of the adjusted pay rate I‰_ { T } .

The Steady-State equilibrium is defined as the stationary solution of the system ( 11 ) and ( 12 ) , with I‰_ { T } =I‰ and n_ { t } =n. Given that at the Steady-State equilibrium the productiveness adjusted pay rate is changeless through clip, so rewards, net incomes and concluding end product addition by a factor I? every clip an invention takes topographic point ( see Figure 1 ) . At the Steady-State, the arbitrage and labour market equations are given by:

Furthermore, as in the Steady-State the productiveness adjusted net income status is equal to Iˆ_ { T } = [ ( 1-I± ) /I± ] I‰.x= [ ( 1-I± ) /I± ] I‰ . ( L-n ) , so ( 13 ) can be rewritten as:

Harmonizing to ( 15 ) at the Steady-State the research degree N is a diminishing map of I± , which corresponds to the monetary value snap of demand of the monopolizer of intermediate goods in this economic system. Hence, competition has negative inauspicious effects on economic growing: the larger the competition, the smaller would be the monopolizer rents appropriated by successful discoverers, therefore the smaller are the inducements to introduce. The mean growing rate at the Steady-State is given by: g=I».n.lnI?

One can so compare the decentralised solution or laissez-faire, with the 1 that would be obtained by a cardinal contriver. The mean growing rate chosen by the societal contriver who wishes to maximise the expected present value of ingestion is:

On a societal public assistance position, the cardinal contriver chooses the degree of N that maximizes Uracils:

In order to compare the mean growing rate of the decentralised economic system, g, with the centralised version, g^ { a?- } , one merely necessitate to compare the values of N and n^ { a?- } . In that instance, one should look at ( 15 ) and ( 18 ) which determine N and n^ { a?- } , severally.

In ( 18 ) the societal price reduction rate is r-I».n^ { a?- } ( I?-1 ) while in ( 15 ) the private price reduction rate is r+I».n. The last is greater than the former. Perversely to the private house, the cardinal contriver does take into history that the benefits from the following invention will go on everlastingly. Therefore, R & A ; D under individualistic tend to be deficient. In ( 15 ) we have the factor ( 1-I± ) which does non look in ( 18 ) . It reflects the inability of the private monopolizer in allowing the entire end product flow. In this instance, research tend to be smaller under laissez-faire. Finally, the factor ( I?-1 ) that appears in the numerator of ( 18 ) replacements I? in ( 15 ) , as the private research house does non internalise the loss to the predating monopolizer caused by the invention. On the other manus, the societal contriver considers that an invention destroys the societal return of the predating invention. This consequence tend to bring forth more research under laissez-faire.

In amount, due to the conflicting consequences from the three effects discussed supra, the mean growing rate under laissez-faire can be either larger or smaller than the optimum growing rate, g^ { a?- } . In other words, the consequence about which of the analysed equilibria in the best 1, is in fact equivocal.

Comparative Staticss on the Steady-State

Aghion and Howitt [ 1998 ] besides analyse the Comparative Staticss on the Steady State rate of growing. At the Steady-State degree the flow of concluding end product that is produced between the T and t+1 invention is:

y_ { T } =A_ { T } x^ { I± } =A_ { T } ( L-n ) ^ { I± } which needfully implies that y_ { t+1 } =I?y_ { T }

One should present the undermentioned inquiry: What happens to the way of flow of ingestion good in existent clip, i.e, as map of I„ ?

From the last equation the log of concluding end product ln Y ( I„ ) additions by lnI? each clip an invention occurs. Nevertheless, the clip elapsed between two consecutive inventions is random, which will arise a random measure map with each measure matching precisely to lnI? & gt ; 0 and with the clip interval between each measure exponentially distributed with parametric quantity I»n ( see Figure 3 ) .

The writers so stop up with a really simple look for the mean growing rate in a Steady State:

Uniting this last equation with the old comparative-statics analysis on the Steady State degree of research N, we can now measure the impact of alterations in parametric quantities on the mean growing rate of concluding end product. A decrease of the involvement rate, R, and in the grade of market competition, I± , or an addition in the size of the labor market, L, will increase N and so g.Increases in the size of invention, I? , and/or in the productiveness of R & A ; D, I» will besides further growing, both straight ( I»lnI? ) and indirectly ( n ) .

Extensions to the Schumpeterian Model: presenting a deterministic constituent

As antecedently discussed empirical facts suggest an account for the creative activity of inventions that involve both a random and a deterministic constituent ( tendency ) . In this subdivision we assume that in add-on to the original Schumpeterian spillover consequence ( related to Newton ‘s “ standing on the shoulders of giants ” celebrated claim ) , each invention that takes topographic point in some part increases the efficiency degree of the bing figure of research workers in that part, in the sense that they become more productive. Therefore, the arrival mean rate of inventions in that part would increase without changing the random constituent of inventions overall.

The new hypothesis to explicate the determiner of research productiveness in new engineerings in composed by a random constituent with a Poisson distribution, I» ( I» is the mean figure of inventions per unit of clip ) and besides a deterministic constituent that is a positive ( but diminishing ) map of the entire figure of inventions, t. Therefore, after a new invention this event is reflected into a higher mean rate of reaching of new inventions in that part. In other words, inventions still have the random character despite the fact that its mean arrival rate grows with the entire figure of inventions. When T is zero the lone determiner of research productiveness of new engineerings is the random constituent harmonizing to a Poisson statistical distribution.

In order to present this tendency consequence into the old basic Schumpeterian theoretical account, we will change the chief equations of the theoretical account as described below:

That is, the mean arrival rate of inventions is now given by a^ { I?t } .I» . This rate grows at a diminishing rate with the entire figure of inventions, t. In this sense, the plus equation becomes:

The monopolizer net income is derived as in the basic theoretical account discussed before:

From equations ( 21 ) , ( 22 ) and ( 23 ) we get:

Dividing both sides by A_ { T } it yields:

Multiplying ( 24 ) by a^ { -I?t } /a^ { -I?t } , we have

in this instance, when ta†’a?z ,

When ta†’a?z , in the Steady-State we have

The corresponding curves to ( 26 ) and ( 27 ) in the ( n, I‰ ) -space are negatively and positively sloped, severally, as one besides gets in the basic theoretical account.

Graphic representation indistinguishable to Calculate 1.

However, equations ( 13 ) and ( 26 ) are different since that latter does non include involvement rate, R, in its denominator, nor the random constituent, I» , in its numerator. This happens because when ta†’a?z , so the term a^ { I?t } .I»a†’a?z independently of the size of I» . In other words, the deterministic constituent of the mean arrival rate of inventions, a^ { I?t } , tends to rule the random constituent, I» , as the figure of inventions grow.

Similarly, in a clip period with a little figure of inventions, the random constituent dominates the deterministic 1. Consequently, the chance of alterations in the universe ranking of income per capita is higher, the smaller the figure of inventions occurred in all states. Therefore, harmonizing to this theoretical account, as the entire figure of inventions, T, grows through clip, it is less likely for a developing economic system to accomplish the frontier of developed states as clip goes on.

Under these hypothesis, alterations in universe ranking of income per capita occur when the stochastic consequence dominates the tendency one. Intuitively, this happens when “ extremist inventions ” cancel the acquisition consequence antecedently accumulated. This means that the new technological paradigms unfastened range for larning chances and that economic systems linked to old/stagnant governments may happen themselves in a technological trap.

Equally far as the involvement rate is concerned, in the basic theoretical account, a autumn in this rate increases the fringy benefit or R & A ; D through increasing the present value of the monopolizer net income and thereore it leads to a higher equilibrium in the figure of research workers, n. However, in the augmented theoretical account, when ta†’a?z , the involvement rate does non impact the equilibrium degree of research workers since the mean rate of reaching of inventions is now infinite.

Historical Long Run Simulations

In order to better understand our part to the original theoretical account, Figures 5 to 12 illustrate, through simulation, the income per capita growing way in both the basic Schumpeterian Model and in its drawn-out version ( s ) . In this simple exercising, T=150 old ages and I ( T ) is the chance denseness map of the cumulative figure of inventions ( harmonizing to a Poisson procedure ) . The variable I? , exogenic and changeless, represents the size of each invention. Take I?=USD 500 at current monetary values. At the beginning of the period ( which goes from 1850 to 2000 ) , assume that income per capita of all economic systems was similar. At the terminal of the period assume that the largest income per capita economic system takes the value USD 50.000, the mean income per capita economic system is USD 12.000 and that the lowest 1 is USD 2.500.

Therefore, the I ( T ) scope goes from 0 to 100 inventions and the estimated I»s for the terminal of the period for the highest, mean and last income per capita are 0.67, 0.16 and 0.033 severally ( Figure 5 ) .

In Model 1, the arrival rate of inventions follows a random Poisson distribution procedure. Hence, this theoretical account incorporates entirely the random constituent and in the long tally one is able to acquire the mean rate of reaching of inventions given by I» . Given that this parametric quantity differs between economic systems and it is changeless over clip, so the long tally places of each economic system in the universe ‘s per capita income ranking will non alter ( Figure 2 and 3 ) . Furthermore, this theoretical account should be altered so that it is allowed to explicate some of the existent universe stylized facts.

In Model 2, the mean rate of reaching of inventions is given by the term a^ { I?t } .I» , where T is the figure of inventions and a & gt ; 1. For simpleness in the undermentioned simulations assume I?=1 ( this artifact does non change the chief decisions of the Model ) . The term a^ { T } introduces a memory constituent to the arrival rate of inventions: the higher the figure of inventions occurred at a certain point in clip, the larger the chance of future reachings of inventions at that clip period. In this theoretical account there is a tendency for rich economic systems to increase their distance to poorer 1s. As the clip goes by, the smaller the chance is for some economic system to accomplish the developed state ‘s higher income per capita degree – though it is non wholly impossible as there is the possibility of “ extremist inventions ” .

Model 2 nowadayss the same trap as Model 1: in the long tally one is able to acquire a changeless value for I» . If this parametric quantity is changeless and differs between economic systems we will acquire precisely the same consequence as in Model 1: in the long tally each state ‘s place in universe ‘s income ranking is inactive ( Figure 8 and 9 ) .

For simpleness in calculating Model ‘s 2 income per capita degrees, we assumed that the term a^ { T } grew from 1850 to 2000, i.e, each twelvemonth the variable T grew one unit in each state and we took a=1,005. This artifact does non change the theoretical account ‘s chief decisions.

The original theoretical account takes I» as a invariable in the long tally equilibrium. However, Castro [ 2000 ] develops an interesting thought by doing I» a random variable which presents a long tally stationary distribution. He develops a two sectors model with increasing returns to labor in the production of new goods, which are created indiscriminately in clip.

Aiming at work outing this job in Model 2 above, one used Castro ‘s theoretical account. In this context, we estimated five series of 150 random Numberss with a Poisson distribution and mean equal to one, matching to I» , for the period 1850-2000.

The coevals of the five series of random Numberss was done utilizing the computing machine plan MatLab, version 7.0

We reached Model 3 where I» is no longer a long tally invariable, alternatively it has a random but stationary distribution. Furthermore, the adoptive size of each invention was, in this instance, I?=USD 250 at current monetary values and a=1,005.Given this alteration in the theoretical account one is now able to detect alterations in place in universe ‘s income per capita ranking through clip. In this sense, in Model 3 income per capita distribution does non meet any longer to a lasting province.

States HH, H, M, L and LL presented after 150 old ages 156, 133, 121, 146 and 169 inventions severally, in Model 3. As one can detect in Figure 10 to 12, in 1875 state M had the largest income per capita, nevertheless in 1900 state LL occupied the top place and stayed in that place until 1950, recovering it once more in 1975 boulder clay 2000. Likewise, state L had the lowest income per capita in 1875 and until 1950, but between 1975 and 2000 its income per capita was above state M ‘s. The staying states present income per capita paths that cross each other several times over the 150 old ages.

Overall this is the consequence that better adjusts to empirical facts, comparatively to Models 1 and 2. The alteration in Model 3 allowed alterations in place in universe ‘s income ranking through clip. In this context, one might be able to explicate several historical facts occurred between the 19th and twentieth centuries, such as why the UK was an economic leader throughout the nineteenth century, while from twentieth century onwards the US presented itself as the most developed state in the universe up to day of the month. Besides several alterations in place allowed for states like a hapless USSR in the beginning of the twentieth century achieve the degree of developed states in the 2nd half ; or even the instance of Argentina which had the opposite development relation to USSR. In any instance, over the last 150 old ages few economic systems presented big fluctuations compared to universe ‘s mean income, as reported by states HH, H and L.

The consequences presented above besides show an of import facet of Model 3: the fact that at the beginning of the period of 150 old ages the random component, I» , dominated the deterministic one, a^ { T } , in finding the mean rate of reaching of inventions. However, by the terminal of the period the term I» is dominated by a^ { t } . In other words, the chance for alterations in place in universe ‘s income ranking is larger, the lower figure of inventions. This fact can be observed in Figure 12 ( rapid climb in of Figure 11 ) : the first 80 old ages ( up to 1930 ) there is intense alteration in the distribution of income per capita between the five states. After that period, which coincides with post-World War II, fewer alterations take topographic point in the distribution of universe ‘s income.

Finally, one should detect that before making the Steady-State degree, we have:

If we assume that I‰ is changeless, so:

After each invention, the term r/ a^ { I?t } decreases, doing it necessary to increase the entire figure of research workers allocated to the research procedure needed for the following invention, n_ { t+1 } , in order to maintain I‰ invariable. That is, when the factors that explain the invention procedure have a tendency constituent in increasing the mean rate of reachings, so the figure of research workers and, hence, the economic system ‘s growing rate would besides be increasing over clip, ceteris paribus ( i.e. , by insulating the random component in the invention procedure ) .

Incentive Mechanisms to Innovation

As seen before, it is an undoubtful fact that economic growing depends, ultima ratio, on the production of new thoughts ; however competitory markets do non allow inducements for that to go on, since if consumers were to pay merely the transmittal costs of new thoughts, so the gross obtained would be deficient to cover the production costs. Historically, societies have used a assortment of mechanisms to further the production of new thoughts. Some of them, like right of first publications and patents, guarantee the pioneer the monopoly in the production of the goods that use those new thoughts. Others, may include direct subsidies to R & A ; D.

The creative activity of monopolies or the ascription of R & A ; D subsidies bring about serious jobs. Copyrights and patents do make deficient inducements to original R & A ; D as discoverers are non able to to the full allow the consumer excess or cognition spillovers to other research workers. For illustration, harmonizing to Kremer [ 1998 ] Michael Milken ( an of import altruist ) would pay 1000000s of dollars for an effectual drug against prostate malignant neoplastic disease. while pharmaceutical industries would non take that into history when specifying their research budget, as they are non able to to the full pull out their excess. Copyrights and patents besides create a inactive deformation in the monopolist monetary value and they encourage non-productive disbursements in “ rearward technology ” . Inactive deformations appear when people that value a good above its fringy cost do non devour it at the monopolist monetary value. For case, AZT high monetary value makes its usage by HIV pregnant adult females in developing states hard and this fact increases the figure of septic kids ( californium. ibidem ) .

When there is complete and symmetric information between agents every bit good as entire committedness in some economic system, R & A ; D subsidies can really be more good than making a new set of deformed monopoly monetary values, coming from a system of patents ( Spence [ 1984 ] ) . However, before research is realized, authoritiess may non cognize its exact costs and expected benefits and so this creates large deformations when it comes the clip to pay the discoverer ( moral jeopardy jobs ) .

Harmonizing to Kremer [ 1998 ] in 1839 the Gallic authorities combined a patent system with direct governmental support to R & A ; D through the acquisition of the Daguerreotype picture taking patent and so by doing it available to the public sphere. After this, the technique was spread all over the universe and it was capable to significant betterments. The acquisition of patents has the potency of non merely extinguish deformations in the monopoly monetary value but besides decrease the inducement to non-productive disbursement in “ rearward technology ” . Furthermore, it encourages original research.

The greatest challenge to the patents acquisition system is monetary value finding. Kremer [ 1998 ] suggests a mechanism in which the patent private value would be determined in auction. The authorities would offer itself to purchase the patent for that private value multiplied by a fixed mark-up, the latter being merely plenty to cover the difference between the private and societal value of the innovation. Aiming at selling at the existent monetary value a little figure of patents indiscriminately selected would be sold at the highest command. Research workers would so make up one’s mind whether to maintain the patent or sell it at the given monetary value and furthermore any patents acquired by the authorities would be made available to the public sphere. Guell and Fischbaum [ 1995 ] suggested that the authorities should utilize its power to purchase pharmaceutical drugs and that Judgess should make up one’s mind the monetary value of the patent. Meanwhile, this method would deter any discoverer to introduce as the concluding monetary value could be much lower than the existent 1. So this scheme – of patents acquisition by authoritiess – should be used together with the traditional patent system. In this manner, discoverers would have an extra mark-up above the market value and this would maintain the inducements to maintain on innovating.

Shavell and Ypserle [ 2001 ] presented a system in which the discoverer could take between the patent and some premium. Their chief decision is that right of first publication and patent systems do n’t possess any cardinal advantage over the reward strategy. Comparing both systems, on the one manus the wages strategy is better than the patent one as it avoids deadweight loss coming from exerting monopoly power. On the other manus, inducements to put in research are imperfect in both systems. In the instance of patents the inducement of invest is ever inappropriate as the monopoly net income is lower than the societal excess. As far the reward strategy is concerned, inducements are governed from the wages itself which are non linked to any societal public assistance step. The optimum system, in which the discoverer chooses between the patent or the wages, is better than the patent system by itself. the ground is that the expected societal public assistance is improved as there is no deadweight loss when the discoverer chooses the wages. However, it is deserving indicating out that in the right of first publication and patent systems later inventions might be blocked by patent proprietors who do non let betterments in bing engineerings ( Fullerton and McAfee [ 1999 ] ) .

Scotchmer [ 1999 ] provinces that the usage of a wages system is limited as there is scarce information on the existent benefits generated by the innovation before it is created. Furthermore, patents besides distort the way of future research due to the inordinate inducement to develop patented goods alternatively of replacement or complementary 1s ( Mansfield et al. [ 1981 ] ) .

The fact that thoughts are non-rival goods suggest that research should be financed through public resources. However, public attempts to finance advanced research are impeded by asymmetric information between the research worker and the public contriver. When the authorities pays for the research production factors alternatively of paying for its concluding end product, so it is harder to supervise the research worker ‘s attempt. Nadiri [ 1993 ] suggests that the return rate of private financed inventions is much larger than the populace financed 1s. Another job with the direct subsidy from authoritiess to R & A ; D activities are the lobbying force per unit areas that distort both possible benefits and costs of inventions ( Romer [ 1993 ] ) .

The last manner to excite invention is through tourneies. These tourneies have had an of import impact in universe ‘s economic development. Fullerton and McAfee [ 1999 ] cite the illustration of an English railroad in 1829. Harmonizing to the writers, tourneies continue to pull fiscal resources and they are more effectual than traditional contracts as they avoid moral jeopardy issues.

In this subdivision we were able to show different, but every bit apt, innovation inducement methods, in add-on to the patent system and which may hold even better consequences every bit far as societal public assistance is concerned.

Methodology and Data

In this subdivision we aim to associate invention and technological advancement with income per capita growing by utilizing a panel information set, which presents, as one of its chief advantages, the complete disclosure of single heterogeneousness. Panel information sets suggest the being of distinguishing features between persons which may be or non changeless through clip ( in this sense pure clip series or transverse subdivision analysis that do non take this fact into history will ever bring forth colored estimations ) .

On the other manus, harmonizing to Hsiao [ 1986 ] panel informations sets provide a larger sum of information, larger informations variableness, less collinearity between variables, higher figure of grades of freedom and increased appraisal efficiency. Including a sectional constituent into a clip series survey allows for greater variableness in the information and this fact reduces eventual collinearity jobs between variables, specially in theoretical accounts with distributed slowdowns.

Additionally, Balgati [ 1996 ] argues that surveies that use panel datasets make it easier a more efficient analysis of accommodation kineticss, whereas cross-section sets transmit a false thought of stableness. Therefore, panel informations allows to fall in single behavior diverseness with the being of accommodation kineticss, even if they are potentially distinguishable. A larger handiness of information increases the appraisal efficiency belongingss, that is, panel datasets allow to place and step effects which are non pure and merely identifiable in surveies utilizing entirely cross-section or clip series datasets.

Harmonizing to Wooldridge [ 2002 ] one can separate between fixed and random effects when utilizing a panel information specification. The former is more appropriate for the instances where we have an thorough sample from a given population or when we want to calculate single behavior. When mentioning to fixed effects we should be covering with theoretical accounts whose coefficients can change between persons or through clip, even if they remain fixed, i.e. non-random, invariables. In the undermentioned subdivision we will utilize fixed effects due to two chief determiners: 1 ) panel information increases the entire figure of available observations ; and 2 ) the purpose to command our theoretical account in footings of concern rhythms through one-year silent persons included as explanatory variables. Random consequence theoretical accounts assume that some of the person and clip periods specific behavior is neither known – non discernible -nor mensurable. Hence, in big panel samples we can ever stand for these single particular effects as a random variable.

The Econometric Model

The proposed theoretical account to discourse the impact of invention and technological advancement on income per capita growing is the followers:

where,

I?= invariable

I±_ { ot } = clip changeless silent persons coefficient

X_ { 1it } = invention placeholder 1

X_ { 2it } = invention placeholder 2

X_ { jit } = control variables

Iµ_ { it } a??N ( 0, I? )

and y_ { it } is the income per capita growing rate for the selected states under examination ( beginning: Summers and Heston Penn World Table 6.2 ) . The sample is composed by 73 states

On the Appendix 2 we present the complete list of states considered in our analysis together with drumhead statistics and correlativity matrices.

( random choice based on informations handiness for all variables of involvement ) , where I represents the state and t the clip for the period 1980-2005 ( subdivided in five period of 5 old ages each ) . It is besides our intent to compare the public presentation of the full dataset counterpart the OECD group. From Figure 13, one can detect that a random sample of states ( from the full dataset ) present a positive norm GDP growing rate for the period under examination. One can detect that poorer states in this tabular array ( e.g. Cyprus, Latvia, Lithuania, or Slovenia ) had a much larger standard divergence than the remainder, hence connoting a more volatile behavior over the period 1980-2005.

The concluding aim is to capture the consequence of invention on per capita income growing, therefore we will utilize as a placeholder the ratio of entire figure of patents to 100.000 dwellers registered in the US market.

We use the US market as a manner of extinguishing differences in statute law across states in our sample.

, every bit good as the Intellectual Property Rights Index ( ab hinc IPR, as in Ginarte and Park [ 1997 ] ) . In Figure 14 we present the same descriptive statistics for the same choice of states as before for the period 1980-2005. The purpose is to prove if the coefficient on the two placeholders for invention is statistically greater than zero, i.e. , the higher the figure of patents / 100.000 dwellers in US market ( and/or the IPR index ) , the larger the growing rate of existent GDP per capita.

Data refering the figure of patents is obtained through the United States Patent and Trademark Office web site which divides patents in six classs: Utility Patent, Design Patent, Plant Patent, Reissue Patent, Defensive Publication and Statutory Invention Registration. Utility Patent, besides called “ patents for innovation ” is the class chosen for the intent ( full definition available in www.uspto.gov ) .

This paper differs from the bulk of the literature in this topic every bit far as the pick for the invention placeholders and the usage of recent informations ( up to 2005 ) . While other surveies take the R & A ; D disbursement ( production factor in the procedure ) , our takes the patent ( result of that procedure ) . Harmonizing to Crosby [ 2000 ] , proving through empirical observation new theoretical accounts of economic growing is highly hard as usual steps to quantify the impact of invention on growing are imperfect in its construct. Normally one uses the R & A ; D disbursement or some labour statistics variable in this sector. However, R & A ; D informations step factors in the production procedure while the patent quantifies the end product of the procedure. Additionally, one is able to happen a big database on patents leting the usage of clip series. On the other manus, when one takes patents as the “ ideal ” step, it is best to be cognizant of the trouble of quantifying its value. That is, some patents are really of import, while others are non. Harmonizing to Jones [ 2000 ] the job in utilizing patents is the fact that non all inventions are patented.

Devinney [ 1994 ] estimated the relation between patent and economic growing utilizing a panel information set and he found a positive and statistically important relation between them. On the same line of concluding Crosby [ 2000 ] argued that the figure of registered patents in Australia has a positive correlativity with this state ‘s labour productiveness and economic growing between 1901 and 1997. Park [ 1999 ] constructs a panel utilizing a sample of 60 states between 1960 and 1990 and provinces that R & A ; D did non explicate per capita growing rate.

In add-on to this technological variable it is necessary to account for a assortment of control variables, as other factors besides influence income per capita growing rate. These variables are:

a?™ degree of income per capita: within the treatment on economic growing it is of great importance to measure whether convergence is taking topographic point among the selected states in income per capita. Beginning: Summers and Heston Penn World Table 6.2

a?™ gross fixed capital formation ( % GDP ) : engineering is incorporated in capital goods. Therefore, one should anticipate physical capital to positively impact some economic system ‘s growing rate. The relation between engineering and investing may happen in two ways: on the one manus, investings do make inducements to better engineering ; on the other manus, technological advancement generates a higher productiveness addition and it accordingly increases the return on physical capital investing. Grossman and Helpman [ 1993 ] show that entire factor productiveness is higher in states with high investing rates. This variable consists of betterments in the efficient usage of land, acquisition of new machinery, building of new roads, schools, infirmaries, etc. together with the net alteration in the stock of goods houses hold to confront unexpected fluctuations in production or gross revenues. Beginning: World Bank Development Indicator 2008

a?™ Intelectual Property Index: inventions are non-rival goods with low fringy cost and they are easy reproduced or copied through “ rearward technology ” , so rational belongings rights are critical in guaranting profitableness and inducement to new inventions. This index goes from one to five, the latter uncovering a high patent protection. This index is constructed through a assortment of features of states ‘ governments and it covers five factors: coverage, engagement in international understandings, loss of protection, legal support mechanisms and continuance. Beginning: Ginarte and Park [ 1997 ]

a?™ Schooling: due to its scientific nature R & A ; D procedures are intensive in the usage of human capital. Secondary schooling normally requires a minimal threshold for admittance. This variable consists in the mean figure of old ages of secondary instruction of entire population and it will be used to capture the direct consequence of instruction in GDP per capita growing. Beginning: Barro and Lee 2000

a?™ Openness ( % GDP ) : measured as the amount of exports and imports it has two effects in R & A ; D disbursement: 1 ) states with competitory disadvantage in bring forthing new engineering will be induced to go forth behind R & A ; D activities ; and 2 ) R & A ; D transference has a positive impact on domestic R & A ; D. Source: World Bank Development Indicator 2008

a?™ Foreign Direct Investment ( FDI ) : FDI influxs have an equivocal consequence on economic growing. On the one manus, FDI can make positive outwardnesss through import of new techniques and capital goods. On the other manus, houses having FDI will, most likely, vie with domestic manufacturers which have less technological ability. Beginning: World Bank Development Indicator 2008

a?™ Fertility Rate: it represents the figure of kids born in some period of clip. Beginning: World Bank Development Indicator 2008

a?™ Life Expectancy at birth: it indicates the figure of old ages a new-born will populate. Beginning: World Bank Development Indicator 2008

a?™ Mortality Rate & lt ; 1y.o. : it represents the figure of kids that die before finishing one twelvemonth old ( per 1000 kids born each twelvemonth ) . Inclusion of this variable, every bit good as birthrate rate and life anticipation, will depend on possible correlativity job. Like anticipation and mortality rate will be used as a placeholder to human capital. Beginning: World Bank Development Indicator 2008

a?™ Urban Population ( % ) : portion of population life in urban countries in each state, harmonizing to UN definition. Urbanization ratio measures possible consequence of geographical bunch in economic growing. Beginning: World Bank Development Indicator 2008

a?™ Population Growth Rate: it includes at occupants independently of legal position, with exclusion of refugees. Beginning: World Bank Development Indicator 2008

a?™ Population Density: entire population divided by the state ‘s country. Beginning: World Bank Development Indicator 2008

Consequences

The chief intent of this paper is to verify whether the invention procedure has a direct consequence on GDP per capita growing rate. In order to carry through this undertaking we use panel informations with fixed effects. One should detect that it is non our intent to measure all channels through which technological advancement can impact GDP per capita growing rate. Before showing our chief consequences it is deserving detecting how Entire Factor Productivity – the Solow Residual or the “ step of our ignorance ” – is explained by variables such as Patents per 100.000 dwellers, Patents per capita and the IPR index In Figure 15 we present the appraisal end product for the three common options when covering with panel informations sets, i.e. , pooling, fixed and random effects

Through a Hausman Test to compare the theoretical accounts utilizing fixed and random effects, we concluded that the former provides better estimations in this instance at a 1 % significance degree.

. In all instances all variables are positive and important at the 1 % degree, connoting that so they are of import in finding the TFP ( in degrees ) for the sample at custodies.

TFP series were computed following the methodological analysis described in Appendix 1.

It is besides interesting to describe that the correlativity coefficients TFP-Patents and TFP-IPR were both positive and increased their value between 1980 and 2000 from 0.29 and 0.67 to 0.30 and 0.75, severally.

An alternate analysis is provided in Appendix 2 which graphs in Figures 23 and 24 two scatterplots of the growing rates of TFP and patents for two minutes in clip, 1990 and 2000. In this instance the correlativity is harder to see in 1990, but one can place a positive sloping relation in Figure 24.

The chief consequences are presented in Table 16 which is composed by five columns

We besides performed the same analysis taking “ Patents per capita ” as our technological placeholder. This alteration did non altered the non-significance associated with this variable ( and its discrepancies ) . Furthermore, even if one thinks of possible multicollinearity jobs associated in holding “ Patents per 100.000 dwellers ” and “ IPR ” both in the appraisal process, this possibility was disregarded through single appraisal with either one variable or the other.

. The first 1 shows all explanatory variables as described before ; on the 2nd and 3rd columns the theoretical account is estimated with all variables for the full set of states under examin