Most of the reduction in the variation in the inequality coefficient occurred between 2000 and 2017. During this period, Latin America entered its longest point of convergence within the selected time series for this paper. The reduction in the inequality coefficient (G) reached %, which resulted in a -convergence of . Within this variation, the mobility effect played a less important role; the reduction in inequality was mainly induced by -convergence, that is, by progressive growth. As a percentage of the variation in the inequality coefficient (G), the mobility effect represented only %. Thus, most of the growth process in Latin America, in this period, was marked by its progressive effect, which represented % of the -convergence. This means that the growth process has proportionally benefited the poorest economies in the region.
Again, regarding the data above, it is interesting to note the level of significance in each period. In all periods, -convergence is significant at a level of at least 5%, which is valid for the period from 2000 to 2017, when most of the convergence between Latin American countries occurs. Between 1995 and 2000, when there is a greater dispersion of the product per capita, the -value is not significant. In the period of 2013 and 2017, marked by the stagnation in important economies in Latin America, there is still a trend towards convergence, although -value is not statistically significant.
As a general conclusion, it is noted that both methods are consistent. In addition, there is a reduction in the dispersion of the product per capita in the analyzed period. The objective now is to better understand what role foreign trade has played in reducing this dispersion, or, in other words, as has now been established, in the -convergence. The main justification for using the alternative approach is exactly its advantage in allowing the variation in -convergence to be decomposed according to sources. In this specific case, these sources will be the components of aggregate demand, selected from special categories of foreign trade. This is because the sources linked to foreign trade weigh considerably in variation of regional inequality, in addition to being a source of an ever more volatile nature and conditioning important elements when plotting regional economic policy.
Decomposing Convergence Arising from Foreign Trade
This section describes the main results obtained from the proposed method, which show the effects foreign trade has had on the
-convergence of Latin American countries and territories. Thus,
-convergence is decomposed from the components of aggregate demand, with trade balances expanded according to selected categories. The results are shown in the
Table A1 and
Table A2. The first column lists the types of components of aggregate demand, while the second and third columns show the composition effect and concentration, respectively. Finally, in the last column, we find the total effect of the contribution of each component in the variation of inequality.
As shown above, the total variation of the coefficient was G = %, which can also be interpreted as -convergence, that is, the variation in the dispersion between the set of countries. Again, it should be noted that the components of internal absorption (household and government consumption, in addition to gross capital formation) contribute considerably to the regional convergence of product per capita. An initial synthesis of the data shows that exports and imports have important effects on the evolution of convergence, with exports corresponding to and imports to percent of the total -convergence. Thus, the various trade policies and integration strategies, by affecting the volume and patterns export and import specialization, also significantly affect convergence.
Nevertheless, although exports and imports significantly affect convergence, their effects do not occur in the same manner. While exports tend to reduce dispersion between countries, that is, to increase convergence, imports are one way in which dispersion tends to increase. In this paper’s time frame, of the reduction of 16.7% found in the dispersion of the product per capita, about 40% occurred due to variations in the export accounts. On the other hand, imports, with a participation of around 50%, significantly contributed to increased dispersion.
The importance of this paper is based, precisely, on opening these accounts to better explain how these effects are associated with the values mentioned above. Thus, a cross-section of foreign trade is taken according to patterns of export and import specialization in terms of technological content and categories of intraregional and extra-regional trade. The data are analyzed below.
Both the exports of goods and services contributed towards increased regional convergence, despite the fact that the exports of goods contributed towards increased regional dispersion, that is, to decreased convergence, by the composition effect. However, this is more than offset by the participation effect. By contrast, imports, whether of goods or services, contributed to the dispersion of the product per capita, due to both effects, given the regressive character of this source on convergence.
In the case of the variations associated with the technological content of Latin American international trade, one can see the progressive measures associated with exports, regardless of their technological structure, are mostly progressive, that is, they contribute to increasing convergence. An exception is the exports of services. This is due to the fact that this export segment is highly represented in the product, at around 35%, although, in a very heterogeneous way between countries, the source concentration index corresponds to 0.7586, indicating that few Latin American countries are specialized in service exports. Moreover, this category contributed to the convergence, precisely because both its participation and its concentration coefficient decreased over the period, which favors general convergence.
Again, regarding the results found for service exports, one can see that their regressive effect is dominant in the total accounts. Thus, total exports are regressive. As stated before, service exports weigh heavily in the product per capita and, consequently, in the total of goods and services exported. Thus, their regressive effect becomes dominant. Hence, this demonstrates the importance of opening trade accounts, as it is of fundamental interest to know what is in fact regressive or progressive in each subcomponent and which are dominant.
Service exports require a more complex and competitive production structure, which is still poorly developed in Latin American countries. This explains the high degree of concentration of this source, with only a few countries in the region meeting the conditions to specialize in exports of this type. However, exports of goods, whatever the subcategory, are progressive, but do not have a dominant effect on this balance. Based on these results, it would be interesting, regarding policy recommendations, if there were an agenda that allowed the poorest countries to become more competitive in the production and export of services, while, at the same time, in which all of Latin America could seek to expand its markets for goods exports.
On the other hand, regarding imports, one can see that the main categories, those with the largest share in the product per capita, are service imports with 18%, medium-technology goods at 10%, and those based on agricultural products with around 8% of the product per capita. This category, except for agricultural products, has a well-defined regressive pattern and contributes to increasing the dispersion of regional products among countries. Importantly, the import categories with greater weight in the Latin American product are shown to be regressive in terms of convergence. The other import categories are progressive, but have a smaller share in the total product and, therefore, have less effect on reducing their dispersion.
From the point of view of an ideal import policy to promote convergence and integration, the agenda would need to seek to reduce the deficit in transactions of medium-tech goods, since Latin America imports many more goods in this category than it manages to export, while it is also the basis of regional imports. This should come with a plan to reduce the volume of service imports. Another way to achieve a similar result would be to promote greater access by the poorest countries to international markets for services and medium-tech goods, increasing total imports and putting pressure on the trade balance.
The
Table A2 describes the
-convergence decomposition for components of aggregate demand, but the foreign trade accounts are displayed in terms of intraregional trade. Analysis of the progressiveness measures supports the perception that intraregional trade favors income convergence in Latin America. Both exports and intraregional imports are progressive. In the case of extra-regional trade, as expected, exports are progressive, since, as shown in the previous table, exports are strongly progressive towards convergence. On the other hand, extra-regional imports are regressive.
However, despite the progressive nature of intraregional exports, there has been a reduction in their participation in the Latin American product per capita. This goes against efforts designed to promote trade integration. Thus, the integration schemes seen in the region are not as effective at consolidating intraregional trade, at least in terms of its growth as a proportion of the product. For this reason, Latin American intraregional exports showed a composition effect of −4.3%, pressuring for greater dispersion (since this source is progressive), but reducing its share in the product.
The composition effect of intraregional exports is offset by their concentration effect, that is, although this type of progressive trade has required participation in the product, it has become even more uniform among the countries. The concentration coefficient of this source decreased by 12.2%, which led to a concentration effect of 13.5% in the decomposed convergence. As a result, the net effect was 9.2%, still favoring convergence, but much less than its potential. Thus, to some extent, the potential for the convergence of the regional economies is wasted by the absence of a policy that effectively promotes the dynamism of intraregional exports, which might lead to the growth of this source as a proportion of the product.
These data complement and support those observed by
Lima and Lo Turco (
2010), who also identified underutilization of intraregional trade in Latin America in terms of potential gains actually obtained. Thus, there is the added fact of the underutilization of this type of trade, not only in its scales, but also in its effect on convergence and, consequently, on Latin American integration.
In the case of extra-regional exports, a similar process occurs with the loss of their participation as a proportion of the product, while they are also more uniform among countries. This reinforces the perception that Latin America experienced a period of loss of dynamism in its exports in 1995 and 2017, contributing to a negative composition effect, only offset by the reduced concentration coefficient in both sources.
Finally, it should be noted that the concentration effect dominates the variation of total convergence, both in the case of convergence composed of technological intensity and of trade. However, the composition effect shows a negative variation, that is, it contributes towards increased dispersion by −9.47%; the concentration effect more than dominates the total variation of convergence. This is, obviously, as expected; one does not see a profound change in the composition of the components of aggregate demand in the total product per capita, whatever the parameter for decomposition, as mentioned above.
A final notable fact is the intensity of the measure of progressiveness of intraregional and interregional imports. In the case of the former, the intensity is much lower, which indicates a relationship of strong progressivity in this category. Thus, this pattern of trade based on the meeting of domestic consumption demands in Latin America by other countries in the region tends to greatly increase regional convergence. In the case of the latter, the concentration ratio of interregional imports is quite high, which indicates that the source has a strong tendency towards regressivity. Thus, when the domestic needs of Latin America are met by countries outside the region, this tends to increase product dispersion.