Let me take a stab at contributing to this debate:
First, I feel a little background on the inequality-poverty-globalization debate is in order here.
The common, liberal argument (the one espoused by Becker and others) holds that over the last 25 years or so, historic trends in poverty and inequality have been reversed. This is the view held by the leading international financial organizations and the other influential illuminati (Martin Wolf, Wolfensohn, Xavier Sala-I-Martin, etc). Additionally, the main driver for this historic turn of events is seen to have been economic globalization—the phenomenon marked by increasing economic integration marked by the liberalization of trade, finance, and open markets. The success of neo-liberal globalization—as evidenced by the trends in poverty and inequality—therefore reinforces and legitimizes globalization along neo-liberal economic frames.
Much of the debate centers on arguments concerning the political economy of statistics—how is globalization measured, how is inequality measured, how is poverty measured? Because the metrics chosen generally determine the outcome of the trend, it is vitally important to know the rationale behind those metrics—there are potentially huge implications.
Let's look at poverty first. The World Bank (which is almost the monopoly supplier of figures) uses $1/day and $2/day lines for measuring the poverty headcount (both are PPP adjusted). Using these numbers the WB holds that both the absolute number of people and the proportion of people living at absolute poverty has declined over the last 20 years. Critics have taken issue with both the metrics they've chosen and with the methodologies of their studies.
1. The $1/day figure, while intuitive, is also rather arbitrary. Additionally, small changes in that base figure lead to huge changes in the headcount. Ex: if it is $1.20 there are hundreds of millions more people that need to be included even though 20¢ is not a huge amount of change for someone living on the edge of survival.
2. The figures are based on household surveys which tend to have fairly large margins of error and hence there is a real uncertainty of accuracy—this is especially true of China and India—the largest countries in the world did not faithfully dedicate themselves to participation in the relevant WB studies (China missed both, India missed one).
1. The WB says poverty went from 1.4BB in 1980 to 1.2BB in 1998. However the methodologies used to find those numbers changed midstream and the earlier figures were not calculated back with the new formulae. This makes it almost impossible to speak knowledgeably about poverty trend lines.
2. The figures are based on a basket of goods and services, not a "basic needs line" that might have been more accurate if one were really interested in measuring the plight of those at the absolute poverty line. This would incorporate things like clean drinking water and calorific intake—not stuff like the price of haircuts.
Almost without exception the World Bank has chosen numbers and methodologies that, while of some limited use, tend to skew the poverty trends in the best possible light. Thus the true poverty figures are likely less rosy, if still rather hazy.
Now let's turn to the inequality trends. Well, what is the trend on world income distribution? It depends. And it is a heated argument. There are a number of ways of measuring global inequality—most show inequality increasing. The World Bank calculates its figures in a way that shows inequality has decreased in the last 20 years. Common ways of calculating inequality include:
1. The WB Method: Uses Per Capita GDP by PPP, using a coefficient like the Gini for income distribution, with countries weighted by population. This measures person-to-person inequality worldwide. By this measure inequality is decreasing.
2. B/n Country Inequality w/States Weighted Equally: measures country-to-country inequality. Useful for comparing state-level economic policy effectiveness. By this measure inequality is increasing.
3. Income Polarization: Uses decile measurements rather than the Gini Coefficient (or similar). Ex: Compares income growth for richest 10% of population vs. income growth for poorest 10%. By this measure inequality is increasing.
4. International Purchasing Power: Uses market exchange rates, not PPP. Useful for assessing geopolitical influence, power, and prestige. By this measure inequality is increasing.
There is nothing wrong with measuring inequality the way that the World Bank and the "usual elite" measure it. However, the conclusions they draw from that measurement are hugely contentious at best, and purposefully misleading at worst. The WB insists that its measure reflects the true overall inequality—fine—but it takes that person-to-person result and equates it with an economic policy result. This is misleading on two fronts. First it is not a generalized trend—the WB's figures are wholly dependent on the incredible performance of a single country—China. Secondly, by weighting by population instead of equally by country, the respective country performance is made irrelevant and thus extending the argument to respective economic policies (which the WB does) is ridiculous.
There are also two other huge factors to keep in mind when assessing inequality figures:
1. With-in country inequality has been increasing in almost every country in the last 20 years. Both in DCs and in ACs—from China and Zambia, to the US and Britain. (look at the work of Milanovich, Krugman, Galbraith, UN Industrial Development Org, among others.)
2. Even if relative income gaps are closing, absolute income gaps are still increasing. If a country with $1000/Capita grows at 6% and a country with $30,000/Capita grows at 2%--the absolute income gap still widens for generations.
Now let us turn to globalization. Or, what is the relationship between globalization and inequality and poverty? The standard liberal argument holds that the perceived positive trends in poverty reduction and inequality are the result of the rising integration of poorer economies into the world system—i.e. liberalizing, globalizing countries have performed better. The WB (look at the Dollar and Kraay 2000/2001 study) measures "globalizing" countries by the change in the ratio of trade to GDP. This skews the results in a number of ways:
1. Relatively closed economies like China and India are counted as globalizers
2. Countries that were already quite open when the data set began, say in 1980, can be counted as non-globalizers. Think of extractive economies.
3. This study glosses over the chain of causation. It assumes open markets lead to growth, but couldn't growth have led to open-markets?
So what are the conclusions on this (non)convergence? Well, if the World Bank figures are suspect on global poverty and inequality trends, what are the implications? They're potentially rather grim. For it then seems that the current structure and nature of the global economic system promotes uneven development. It could be plausibly argued that the global financial order that was ushered in after the collapse of the Bretton Woods system has re-ordered the value of various sorts of economic activities.
And indeed, some of the more recent advances in economic theory—perhaps not so coincidentally occurring precisely at the time of the most relentless pursuit of the Washington Consensus policies—have started to perhaps show some of the root causes of this uneven development. Building on earlier development and industrial theories espoused by List, Schumpeter, and Raul Prebisch, modern heterodox economists like Romer, Krugman, Stiglitz, and Arrighi have taken the field against modern global market fundamentalism. And they have identified new trends in industrial rewards and interesting implications resulting from global supply and value chains. However, just as important questions remain for political observers—why would the World Bank and other elites champion questionable trends? And perhaps even more importantly, what can we do about it?