ترکیب مالکیت بین المللی موقوفات در یک مدل تعادل عمومی اعمال شده جهانی

The ability of comparative static models to capture the long-run effects of trade liberalisation is often limited by their inability to take account of capital accumulation and track foreign ownership. In this paper a method for endogenising capital and tracking foreign ownership is outlined. The mechanism adopted uses endogenous risk premium to explain how investors allocate their saving across regions. This mechanism is incorporated into the GTAP model and used to simulate the long-run effects of Asia–Pacific trade liberalisation. The results show that foreign capital ownership can significantly affect the projected long-run results of trade liberalisation.

The sectoral detail and general equilibrium nature of computable applied general equilibrium (AGE) models make them ideal for analysing the short-run effects of trade liberalisation. It is in the long run, however, where the short-run gains of trade liberalisation have led to further capital accumulation, that many additional benefits from trade liberalisation are expected to occur. When long-run mechanisms are introduced, however, applied general equilibrium models quickly become large and complicated; assumptions regarding the mobility of capital need to be made and the ownership of that capital and rental income needs to be taken into account. Some research into incorporating long-run behaviour into global applied general equilibrium models has been undertaken. In the case of the GTAP model (Hertel and Tsigas, 1997), previous attempts at simple comparative static long-run analysis have included Francois et al., 1996 and Walmsley, 1998. Both of these methods incorporate steady state characteristics, however, their assumptions regarding capital mobility differ significantly, with Walmsley (1998) opting for perfect capital mobility, while Francois et al. (1996) fixes the trade balance and hence, net capital flows. Both Francois et al., 1996 and Walmsley, 1998 have had to make certain unrealistic assumptions about the mobility of capital and the allocation of rental income to the owners of that capital. Willenbrockel (1999) showed using a two-country model that the practice of ignoring cross-ownership of capital could give rise to misleading results. Moreover, Willenbrockel (1999) suggested that in some cases even the sign of the welfare effect might be reversed. Tracking changes in the ownership of capital resulting from trade liberalisation is thus of paramount importance for long-run analysis. In the Dynamic GTAP model, developed by Ianchovichina and McDougall (2001), some foreign ownership of capital is incorporated by allowing households to invest both domestically and in foreign countries via a global trust. Although ownership is not tracked on a bilateral basis this allows them to track the flow of foreign income to and from abroad and hence make some conclusions regarding welfare. In this paper capital accumulation is dealt with in a comparative static framework and capital ownership is tracked on a bilateral basis. In addition to foreign capital ownership the existence of foreign guest workers and the foreign ownership of land are also examined, although to a lesser extent. The modifications are also made to the standard GTAP model. The model developed in this paper differs from the Dynamic GTAP model in several respects. Firstly, the comparative static nature of the GTAP model is maintained in this model. Secondly, foreign capital ownership is tracked on a bilateral basis, while the Dynamic GTAP model allows households to invest either in the domestic economy or in all foreign economies via a global trust. Finally, the mechanism for allocating investment across regions differs between the two models. In this model a household allocates saving across regions so as to equalise risk adjusted rates of return, where risk is determined endogenously. In the Dynamic GTAP model the household allocates saving between domestic firms and the trust in such a way as to keep the shares invested in each of these as close as possible to those established in the initial database, subject to certain adding up constraints. In both cases the models ensure that households tend to invest more at home than they do abroad. Once these modifications have been made, the long-run effects of Asia–Pacific trade liberalisation are simulated in order to examine the effects of incorporating foreign ownership into the GTAP model. Two simulations are undertaken. The first uses the revised model and database developed in this paper, while the second, uses a long-run closure developed by Walmsley (1998) for the standard GTAP model. Following the introduction this paper is divided into three broad sections. Section 2 examines the new equations incorporated into the GTAP model to track the ownership of endowments. Section 3 provides an analysis and discussion of how the initial data set, required for this extension, was created. The latter discussion includes details of the assumptions made and the necessary calibration. Section 4 is divided into two subsections. The first develops the short- and long-run closures for the revised model. The second provides a brief account of the results of an Asia–Pacific trade liberalisation shock using the revised version of the GTAP model and compares them with the long-run results obtained using the standard GTAP model (Walmsley, 1998).

The primary purpose of this paper has been to outline a method for incorporating long-run behaviour into a comparative static global applied general equilibrium model, which would also track changes in the ownership of capital. In addition, these changes also included incorporating foreign ownership of land and the labour incomes of foreign workers. The resulting model measures welfare in terms of income earned by permanent residents on endowments owned by permanent residents, rather than in terms of endowments located within the region. This has required a number of fundamental changes, which were implemented into the structure of the GTAP model and database. The most significant of these changes involved incorporating a mechanism for allocating the saving of permanent residents across regions to investment. It was assumed that the saving of permanent residents of t is allocated across regions such that their own risk-adjusted expected rates of return equate across regions of asset location, where risk is determined endogenously. In this model the level of risk associated with holding capital located in a given region is said to rise with relative exposure to that region. This is known as the perceived risk schedule. This schedule states that as the proportion of capital stocks, located in r and owned by permanent residents of t, increases relative to the corresponding global proportion, perceived risk also increases. Hence, expected risk-adjusted rates of return can be equated by moving along the perceived risk schedules of the different representative owners. The model was then used to simulate the long-run effects of an Asia–Pacific trade liberalisation shock. These were then compared to those obtained using a long-run closure with the standard GTAP model (Walmsley, 1998). The results showed that the incorporation of foreign ownership did have a significant effect on the projected results of a simulation. This concurs with Willenbockel (1999) findings in the two-country case where he concluded that the results could be affected by the inclusion of cross-country ownership. This paper has examined how the accumulation of capital over the long-run and the ownership of that capital can affect the predicted outcome of a policy shock. There are, however, many other long-run issues, which could also affect the outcome of a policy, such as endogenous growth in the form of endogenous technological changes or higher savings rates. Unfortunately, not all of these could be covered in this paper. In addition, the paper does not take into account any other long-term policies that the Asia–Pacific economies might be expected to undertake over the period.