From the beginning of the 1990’s, global inequality, in particular personal distribution of income and wealth, became the subject of an intense economic debate, addressed by different points of view: theoretical, practical and of policy (Atkinson:1993). The topic, in the past years, has been addressed in regard of new frameworks, especially of the enhancement of the globalisation’s process. Amartya Sen (2002) has indeed recognized inequality both within and between countries as the main challenge. He identifies the crucial issue in the division of the potential gains from globalisation, between rich countries and poor countries and between different groups in the same country. If some countries have progressively reduced the gap among the value of the average income in comparison to the world average income, others have widened such discrepancy instead.
The factors, that have strengthened the increase of global inequality, are numerous and vary greatly. Due to different national contexts and to the level of development, some of them are specific and endogenous, they mostly depend on socio-economic and demographic variables. Nevertheless, these factors are not sufficient to explain the level of global inequality. It is, indeed, necessary to take in account also exogenous factors that have influenced, to different extent and according to specific institutional context, global inequality. Among all of them, globalisation is one of the most remarkable. Globalisation it is a broad and complex phenomenon, which can be interpreted by different points of view. It can be understood as a sociological and institutional process or as a merely economic process. It is defined ‘in terms of an outcome variable: increased share of trade and direct foreign investments in GDI’ (Milanovic:2006:p11). In this essay, in regard of the literature taken in consideration,the economic definition of globalisation will be addressed.
Several authors have questioned the existence of a relation between globalisation and global inequality, reaching a wide variety of findings. The aim of this essay is, indeed, to discuss whether the measurements are important in the understanding of this relation. In the first place, the controversial debate on the topic is briefly discussed. In order to clarify the different arguments, the methodologies and the data used are questioned and investigated, as these choices influence the outcomes.
The first researches on the identification of a global inequality trend were conducted in the 1980s. Bourguignon – Morrisson (1983) were two of the pioneers in the area and they ended their analysis in 1992. Their findings observe the absence of an inequality trend overall. Neither Schulz (1998) found any data that could suggest a tendency in global inequality; he argued, instead, that inequality has settled at a high level. Dollar and Kraay based their work on their studies and they argued, instead, that global inequality has increased considerably until 1980, following a period of stabilization and it finally started to decline in the recent era of globalisation (Dollar and Kraay:2002). Globalisation, indeed, seems to have an indirect impact on their measure of global inequality. Briefly, following the logic of their main argument: they proved that the acceleration of growth is the driven force in ending the global inequality increase trend and, according to the authors, the countries that have become more ‘globalized’, in the sense of being more open to trade and to direct foreign investment, have grown faster.
As empirical evidences of their argument they analyse India, China and Vietnam; although, as Galbraith (2002) noted, these countries might not be consider appropriate examples of ‘globalized’ countries. India, indeed, began its economic growth, due to the rigid capital controls and the long-term development assistance, which protected the economy from the debt crisis, occurred in Latin America and elsewhere. China’s progress was firstly induced by the agricultural reform, secondly by a strict program of industrialization and only thirdly by the openness to the world market. Moreover, China and Vietnam are governed by their communist parties, which can not be considered ‘Washington consensus’ countries (Galbraith:2002). Historically it seems that the rapid economic growth in these countries occurred before the market liberalization. However, especially the economic growth, which surely raises the incomes of the poor, does not necessarily reduce inequality (Galbraith:2002; Ravallion:2003); ‘growth, [indeed], raises the incomes [of the poor] by about as much as it raises the incomes of everybody else’ (Economist magazine, 2000, referring to Dollar and Kraay(2002).
Bhalla (2002), using very similar methodologies to Sala-I-Martin (2002), reached the same conclusion of Dollar and Kraay. Specifically, he argued that globalisation is driven by two elements: firstly, the ‘law of one price’ and the ‘law of one wage’, which will probably lead to catch-up growth in term of prices, including wages. Secondly, the investments of willing developing countries in education might create the conditions of what Arthur Lewis recognized as ‘unlimited supplies of labor at almost every education level’ (cited in Bhalla:2002). Given these circumstances, Bhalla (2002) claimed that world inequality has and will decrease constantly. Theory strongly criticised by Milanovic (2006:8), who, instead, believes that in global inequality pattern ‘there are zigzags’ (i.e. he calculated an increase of 3 Gini points between 1988 and 1993 and a decline of 1 Gini point in the followed five years). Discontinuities explained by his recent theorization of the Kuzents Wave, which, roughly, claims that global inequality is highly influenced by what he called the second technological revolution and the consequential technological innovation: the second Kuznets curve in the modern era (Milanovic:2016).
The lack of unanimity on the identification of a trend in global inequality and on the demonstration of globalisation’s influence in this trend, lead to question and investigate the methodologies, the data and even the definitions used by the authors. These choices influence the predictions on the direction of inequality, as the literature on globalisation and inequality showed. Three key distinctions could be drawn in the debate: the measurement of inequality, the availability and the quality of data, and whether the measure are population-weighted or not. (Mills:2009, Milanovic:2002, Dollar & Kraay:2002, Bhalla:2004
The majority of the studies adopt the Gini coefficient in the measurement on income inequality, which involves the use of data on the distribution of the income and its mean. It has been introduced more recently alternative measures, as the mean logarithmic deviation income (MLD) index of inequality, but the Gini coefficient represents the dominant preference (Mills:2009). Household surveys provide the data on income distribution and national accounts data (NA) and surveys are the two possible sources for the mean. The research community result be divided in two: who does prefer the exclusive survey means (mostly of the World Bank researchers) and who prefer the use of national accounts data (some World Bank authors and the rest of the research community) (Bhalla:2004). Nevertheless, survey data are not available annually, requiring the use of a considerable amount of NA information in the ‘only survey data’ and combining these two aggregates, which differ in the sources, is not ‘distributional-neutral’ (Milanovic:2006). Furthermore, NA include the institutional and NGO population, while the survey excludes it, meaning that the two sources have different coverage (Bhalla:2004). Due to this difference in definitions, it is very likely that, at any particular point, the survey mean will deviate from the NA mean (Bhalla:2004). In addition, surveys tend to not fully represent capital income and to underestimate especially high-income groups (Mills:2009, Milanovic:2006), therefore miscalculating inequality itself.
The lack of consistency in the availability of survey data is sometimes balanced by assumptions in the calculations. Millanovic (2006:9) commented the extrapolation of income distribution data from quintiles or from Gini coefficients by Bhalla and Sala-I-Martin as ‘very dubious assumptions e.g. that income distributions do not change over time or change in a certain (linear) fashion or everyone in a country has the same income’. Suppositions that may biases inequality result itself, in the sense that the possible slight difference occurred due this ‘measurement error’ could be interpreted as evidence of a change or of a trend (Millanovic:2006). Reading this change in relation to the global context could bias the interpretation of the trend: for example, the decreasing trend in global inequality identified by Bhalla it is seen through the lens of his definition explained earlier.
Another interesting case in readjusting the lack of data is the exclusion of Soviet Union in the measurement of global inequality between 1970 and 2000 by Sala-I-Martin. Milanovic (2002) is strongly critical of this preclusion, as he believes that, if it was included, Sala-I-Martin could not reach the same result on the decrease of global inequality. And the exclusion of Soviet Union in these particular years, specifically during its collapse and its inclusion in the liberal market, might deceive the understanding of the relation between globalisation and global inequality.
A third issue identified is whether inequality measurements should be weighted by population size or not. Roughly, distinguishing between what Milanovic refers to Concept 1 “inequality among countries’ mean incomes” and Concept 2 “inequality among countries’ mean incomes weighted by countries”. Un-weighted population studies are often used to determine cross-national differences, as each country is equal; population-weighted studies, instead, are useful to have a better global picture of inequality. Although, it is interesting to note the two contradictory trends identified by the two different measurements. Un-weighted population studies observe ‘a clear divergence in the world incomes during the past twenty years’ (Milanovic:2006). Differently, concept 2 calculation identified a decreasing trend in global inequality in the past twenty years, due precisely to the relatively high weight of notably China and India, but also Vietnam and Bangladesh (Dollar & Kraay:2002). “Large countries therefore figure prominently in measures of global inequality, rather than ‘distorting the picture’ as suggested by Pitts” (Dollar & Kraay:2003:182). Nevertheless, the exclusion of China and India from the measure shows a widening trend instead, as proved by several authors as Millanovic (2006) and Wade (2004). It is necessary to have a complete understanding of the choice made by the authors in calculating global inequality to interpret the findings in relation to phenomenon as globalisation; otherwise the assessments made could be biased. For example, the confirmation of the positive influence (direct of indirect) of globalisation on global inequality claimed by Bhalla and Dollar & Kraay using concept 2, should be also identified in comparing the cross-national impacts’ differences using concept 1.
In conclusion, the empirical evidences provided by several author in the last year can not show an agreement neither on the direction of the trend in global inequality and neither on how this trend is influenced by globalisation. The interpretations made can hardly be generalised, due to the multiple differences in the method of collection of the data, in the estimation techniques, in indexes used to measure global inequality and in the number of the countries considered in the test samples. Some of the measures show a substantial stability, other an increase and more besides a decrease. Therefore, it is important to relay of these measurements in the understanding of the relation between globalisation and global inequality only if the ambiguities, the one explained in this essay and more, are acknowledged previously. The brief review of globalisation and inequality literature demonstrated indeed the importance of being aware of the critical issues in the measure of inequality, in the choice of data and in whether population-weighted or un-weighted analyses. It demonstrated also that, despite the fact that globalisation appears as a wide and complex construct, it is possible, to some extent, to examine its impact on nations and individuals within them. Moreover, other complications are given also by the difficulty in decoupling globalisation itself and other exogenous factors. It is therefore necessary not only acknowledge the disputed listed previously, but also the definition of globalisation given, which can vary by study in study. Once an agreement is reached in these concerns, the relation between globalisation and global inequality could be more appropriately understood.