حسابداری رشد در اقتصاد باز : مقایسه بین المللی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|24929||2003||19 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : International Review of Economics & Finance, Volume 12, Issue 4, 2003, Pages 417–435
This paper identifies, measures, and compares the main factors explaining nominal GDP growth in two dozen open economies. The analysis goes beyond the standard Solow approach by disaggregating outputs, by accounting for terms-of-trade changes, and by being based on a flexible representation of the aggregate technology, with special allowance for foreign trade. The results demonstrate the overwhelming importance of capital accumulation and technological change in explaining real growth. Movements in the terms of trade play a significant role in many countries as well. The contribution of labor is found to be negligible, or even negative, in most European countries.
From 1967 to 1996, GDP growth averaged 48.8% in Turkey, but it barely reached 5.7% in Switzerland. How can such large differences be explained? Many reasons can be invoked. Besides the obvious fact that inflation has widely varied across nations, there are real factors, too, such as uneven rates of technological change, differences in the rate of capital accumulation, differences in employment growth, and uneven terms-of-trade movements. The purpose of this paper is to identify, to measure, and to compare the main factors explaining nominal GDP growth in a number of open economies. We will use an index-number decomposition that has a strong theoretical foundation, being based on the GNP/GDP function approach to modeling the production sector of an open economy.1 This technique builds on the pioneering work of Diewert and Morrison (1986) who examined the welfare effects of technological change and terms-of-trade shifts.2 Much of the recent work on growth accounting has focused on the effects of technological change and increases in domestic factor endowments, especially the supply of capital (see Mankiw et al., 1992, Pack & Page, 1994, Young, 1994a and Young, 1994b, for instance). A controversy has developed regarding the relative importance of these factors, with some authors stressing the role of technological change and productivity shocks, and others adopting a more conventional view by mostly crediting the process of investment and capital accumulation. It is likely, however, that in most cases both factors are simultaneously at work, and that any discussion as to the relative importance of these two forces can only progress on the basis of additional empirical evidence. In fact, it probably would be most useful to first refine the measurement of the various factors at work, and to ensure that all major growth determinants have indeed been taken into account. In this regard, it strikes us that previous work has generally neglected the impact of terms-of-trade changes. This is all the more surprising that foreign trade has often been cited as a major factor explaining economic growth.3 As noted by Diewert and Morrison (1986), an improvement in the terms of trade is similar to a technological progress since it makes it possible for a country to increase its net output for any given amount of domestic inputs. A deterioration, on the other hand, is equivalent to technological regress, and it reduces the net amount of goods that a country obtains for a given effort. With a few exceptions, growth accounting is based on the method developed by Solow (1957).4 That is, the technology is described by an aggregate production function. All outputs are aggregated, and two inputs—labor and capital—are considered. The functional form is Cobb–Douglas.5 The model then decomposes growth into three parts: the contribution of labor, the contribution of capital, and an unexplained residual that is interpreted as the contribution of technological change. This calculation can easily be done, based on the knowledge of the data alone, and of the average share of labor and capital in total costs. Solow's approach is clever, and it is based on a tight theoretical framework. It is both extremely simple and transparent. However, it suffers from a number of serious drawbacks, which, fortunately, can easily be remedied. First, by considering a single output, Solow's approach is excessively restrictive. Disaggregating outputs would be fairly straightforward, but it forces one to abandon the production function setting, or to assume that the technology is globally separable between inputs and outputs. Second, Solow's approach is not well suited to analyze open economies since it does not allow for imports and exports, and thus, it is incapable of assessing the contribution of terms-of-trade movements. Third, the use of a Cobb–Douglas functional form is unduly restrictive; it would be much preferable to use a flexible functional form instead.6 All three points are being addressed in what follows. Specifically, we will innovate by using a multiple-output GDP function instead of a production function, by using the Translog functional form rather than the Cobb–Douglas, and by explicitly modeling foreign trade, which will enable us to identify the terms-of-trade effect. The shortcomings of the Solow approach are likely to affect the so-called Solow residuals that are traditionally interpreted as the contribution of technological change and the result of exogenous shocks. The assumption that the shares of labor and capital are fixed through time, for instance, or the neglect of the effects of changes in the terms of trade and changes in relative output prices are likely to lead to a number of biases. The residuals presented in this paper could advantageously be used in lieu of the standard Solow (1957) residuals commonly used in empirical studies of real business cycles, and international comparisons thereof. The paper proceeds as follows. The aggregate technology is described in the next section, and the GDP growth accounting framework that we use is presented in Section 3. Section 4 gives a description of the data. Our estimates are presented in Section 5, and Section 6 concludes.
نتیجه گیری انگلیسی
One impression that emerges from Table 1 and Fig. 1 is that the causes of GDP growth vary greatly from period to period and across nations, even if one abstracts from domestic price inflation that is notoriously volatile. Nevertheless, the contribution of capital seems to be dominating the real growth picture. Technological change and productivity shocks vary greatly through time and across countries, and they are the prime cause of long-run growth in just over one third of the countries in our sample. Terms-of-trade changes do add up over time and should not be neglected. One point that distinguishes European nations is the low—and sometimes even negative—contribution of labor. Not surprisingly, the countries that experience the fastest population growth are also the ones where employment contributes the most to economic activity. European nations are not only characterized by low employment growth, which reflects both a slow growing population and a high level of unemployment, but also a rapidly shortening workweek. The residual Rt,t−1 comes close to the concept of unexpected, and transitory, productivity shock that is a driving force in real business cycle models. 23 In our opinion, the estimates of Rt,t−1 reported in Table 1 could advantageously replace the Solow residuals commonly used in empirical work on real business cycles and the international propagation of productivity shifts. Indeed, it is surprising that so many of these studies, which use very sophisticated times series analysis and other econometric techniques, rely on such crude measures of technological change. Unlike Solow residuals, Rt,t−1 is based on a flexible, multiple-input, multiple-output representation of the aggregate technology. Furthermore, it may be useful in empirical work on real business cycles to also take account of the GDP effects due to changes in the terms of trade, as measured by At,t−1; such changes are often unforeseen, and thus they can be associated with unexpected productivity shifts. The framework presented in this paper could easily be expanded to allow for additional inputs and/or outputs. It would be of particular interest to disaggregate labor according to skills and education in order to assess the GDP growth contribution of education and human capital. Another interesting research avenue would be to take public goods and infrastructure into account. These and other extensions may be undertaken in future work.