حمایت از حقوق مالکیت معنوی و رشد اقتصادی درون زا

The main purpose of the paper is to examine the impact of intellectual property rights (IPR) protection on economic growth and welfare. To achieve this aim, we make use of an expanding-variety-type R&D-based endogenous growth model. We work out the transitional dynamics of a shock in IPR protection and account fully for the loss in current consumption and gain in consumption growth due to a tightening IPR protection. We find that there exists an optimal degree of IPR protection in our model. We then calibrate our model by US data, and found that under-protection of IPR is much more likely than over-protection. Moreover, in the case of over-protection, the welfare losses are trivial; whereas in the case of under-protection, the welfare losses can be substantial.

The main purpose of the paper is to examine the impact of intellectual property rights (IPR) protection on economic growth and welfare. To achieve this aim, we make use of an expanding-variety-type R&D-based endogenous growth model a la Romer (1990). The simplest way to model IPR protection is to assume that imitation is costless, and that stronger IPR protection lowers the rate of imitation. This is the approach we adopt here.1 The conventional wisdom in the literature about strengthening IPR protection is that it encourages investment in R&D, yet depresses current consumption. Since there is a tradeoff between current loss in consumption and future gain in growth rate, there is a possibility of the existence of an optimal degree of IPR protection. The existence and properties of such an optimum, however, have not been confirmed before in the endogenous growth literature, partly because it involves a full characterization of the transitional dynamics of the rate of innovation and fraction of goods imitated in the economy. If transitional dynamics are not considered, and one focuses only on the steady state, then the welfare analysis is misleading. In fact, steady state welfare is maximized when growth rate of consumption is maximized. This will be achieved by protecting IPR fully and forever. Obviously, such a corner solution is intuitively unappealing and also counter-factual, since the transitional welfare gains and losses are not taken into account. This paper shows that once transitional dynamics are taken into account, there exists a finite optimal degree of IPR protection. One contribution of our paper is that we compute the optimal IPR by working out the transitional dynamics of a shock in IPR protection and accounting fully for the loss in current consumption and gain in consumption growth due to a tightening of IPR protection. Specifically, we find that when the government announces an immediate increase in IPR protection, there is an immediate drop in current consumption and an immediate increase in the rate of growth of consumption, as well as overshooting of the rate of innovation. The instantaneous fall in consumption is caused by an expansion of the R&D sector, which bids up interest rate and induces more saving. On the other hand, higher investment in R&D leads to higher growth of consumption following the initial level drop. At the optimal level of IPR, the marginal cost due to current consumption loss is equal to the marginal gain due to consumption growth. We are able to compute this optimal level of IPR, the existence and significance of which has not been established in the literature. Another contribution of our paper is that it estimates the welfare loss due to deviation of the current IPR protection in the US from the optimal level. We conclude that under-protection of IPR is potentially a very important reason for under-investment of R&D. This is complementary to other work in the literature (e.g. Jones and Williams 1998 and Jones and Williams 2000). To assess quantitatively the welfare significance of optimal IPR protection, we calibrate our model by US data on long-term growth rate, mark-up factor in manufacturing industries, time rate of preference and intertemporal elasticity of substitution. The calibration results indicate that there is under-protection of IPR (relative to the optimal level) within plausible range of parameter values, and that under-protection of IPR is much more likely than over-protection. More complete computation indicates that in the case of over-protection, the welfare losses are trivial; whereas in the case of under-protection, the welfare losses can be substantial. One interpretation of this result is that the US should protect IPR much more than it currently does. Our specification of R&D draws from the “laboratory equipment” model used in Rivera-Batiz and Romer (1991). The dynamic analysis is similar to that of Helpman (1993). Helpman studied the effect of IPR protection in the South on the welfare of the North and the South in a two-region global economy. Contrary to our main objective, he neither attempts to find whether there is an optimal degree of IPR, nor is interested in whether there is under-protection of IPR in the global economy.2 There are by and large two types of R&D-based endogenous growth models: expanding-variety-type and quality-ladder type. O'Donoghue and Zweimuller (1998) construct a quality-ladder-type R&D-based endogenous growth model in the tradition of Grossman and Helpman (1991a) and Aghion and Howitt (1992). They merge the patent-design literature and endogenous-growth literature incorporating both length and breadth of patent in the quality ladder. They point out the shortcomings of the partial equilibrium patent-design analyses, which omit the general equilibrium effects. One of these effects is that when multiple industries use patent protection, the monopoly distortion effect can be greatly diminished. In our model, rather than the tradeoff between static and dynamic efficiencies, the central issue is the tradeoff between loss in current consumption and gain in consumption growth when IPR protection is strengthened across industries. Futagami et al. (1996) study optimal patent length in a Grossman–Helpman-type quality-ladder model. Although they identify an optimal patent length under certain conditions, there is no transitional dynamics as in our model. Nonetheless, their work is an interesting complement to our paper. Our result suggests that there is under-investment in R&D due to under-protection of IPR. This under-investment result echoes that of Jones and Williams (1998), who find that the US under-invested in R&D by a factor of 2–4. Jones and Williams (2000) is an attempt to account for the relative importance of the various externalities that are responsible for the deviation of the market equilibrium from the optimal level of R&D, such as knowledge spillovers, “stepping on toes” effect, and creative destruction. Our quantitative result points out that under-protection of IPR is another important reason for under-investment of R&D. Our contention that IPR protection is an important determinant of growth is echoed by the work of Jones (2001), who argues that property rights protection is responsible for the emergence of the “industrial revolution” in the 20th century. Our closed-economy result should be readily extended to an international setting. For example, Lai (1998a) finds that the rate of innovation increases with stronger IPR in both the North and the South as long as direct foreign investment (DFI) is the major channel of international technology diffusion. Accordingly, an optimal degree of IPR should also exist in a global economy. However, in a two-country setting, there are additional issues to be addressed, e.g. How are the effects of Southern IPR protection different from those of Northern protection?; and, Should there be harmonization of IPR standards? Moreover, the channel through which international technology diffusion occurs can affect the results.3 Section 2 lays out the model, and derive the dynamics when there is an immediate increase of IPR protection. The optimal degree of IPR protection is derived. In Section 3, we calibrate the model to the US economy. Since closed-form solution is not possible, we solve the dynamic general equilibrium numerically, and compute the optimal degrees of IPR protection that correspond to different assumed actual monopoly durations of the innovators. Section 4 concludes with some discussion on future extensions.

It must be borne in mind that our model is, in a sense, a metaphor that tries to capture the fact that development of new intermediate goods increases labor productivity, and that strengthening IPR protection lengthens the average duration of monopoly position of the innovators of these intermediate goods. A few caveats need to be mentioned. First, recall that we treat IPR protection as the part of “imitation rate” (rate at which monopoly profit is eroded) that can be influenced by government policy. This is a policy package that includes patent length and breadth, protection of trademarks, copyrights and trade secrets, and the degree of enforcement. Some may argue that even anti-trust policies can sometimes be treated as IPR polices. Accordingly, a strengthening of IPR in the US means a change in the whole package of policies in favor of the innovators. Consequently, any change in only one dimension of the package, e.g. an increase of patent length from 17 to 25 years, might have limited effect on the overall protection of IPR.8 Second, just like other one-sector, highly aggregated, macroeconomic models, ours cannot capture things such as sectoral differences. However, we gain in tractability, which allows us to calibrate the model using easily available macroeconomic data. Third, we have made use of a very simple model to illustrate the tradeoff between loss in current consumption and gain in consumption growth due to increase in IPR protection. While we gain from simplicity of the model, we might not have sufficiently captured certain important aspects of the economy. For example, the expanding variety model is subject to the criticism that it fails to capture obsolescence of goods—goods stay in the market forever. The absence of obsolescence might lead to an over-estimation of the degree of IPR under-protection that we have found to prevail in the real world. However, it is straightforward to show that the qualitative aspects of the results would be preserved if an exogenous rate of obsolescence is incorporated in the present model. It is endogenous obsolescence that is of substantive interest and has not been captured in this paper. We would extend the current model to incorporate endogenous obsolescence along the line of Lai (1998b) and the result would be reported elsewhere. Recently, there has been debate on how to reconcile the historical trend of increasing R&D-to-output ratio and a constant growth rate (see for example, Jones (1995) and Kortum (1997)). Our model features constant R&D-to-output ratio, which is at odds with the fact. However, we believe our results will not be altered qualitatively even if we adopted a modification such as in Jones (1995), which features increasing R&D to output ratio.9 Another extension we are carrying out is to assume a more general imitation technology to capture the fact that the rate of imitation is also dependent on the knowledge accumulated from past imitations. In this way, we allow for transitional dynamics of the hazard rate when there is a shock to IPR protection. Preliminary results show that this would give rise to multiple steady states and substantially richer transitional dynamics. It would be interesting to examine the qualitative and quantitative impacts of changes in IPR protection in such a more general and presumably more realistic model.