Should We Use PPP-Adjusted Data to Discuss Global Income Inequalities?
Most economists, even my favorite ones, use Gross Domestic Product adjusted for purchasing power parity (PPP) when they compare the “wealth of nations.” In my last post, I made a short case for why comparing income levels in the world economy should use national income calculated at foreign exchange rates (FX). Given that my assertion falls in the minority position in the world of social science, I would like to expand on that claim.
After my justifications, I will also argue that Gross National Income (GNI – formerly known as Gross National Product/GNP), rather than Gross Domestic Product (GDP), is a better indicator of wealth levels in the world economy. I hope that readers will be able to sift through the econo-speak and get at the substantive issues at hand.
GDP is an overall measure of the flows and stocks of economic activities linked to market processes within a particular political territory. Yearly increases in GDP are the “value added” by the market to any existing and new economic activities over an annual period. When GDP per capita is measured at FX-based prices, this means that the market “value” of the overall set of market-linked economic activities within a certain territory is denominated in the national currency. This can then be converted at the international exchange rate to a common currency (usually US Dollars).
The GDP figure by itself (or when divided by population size to get GDP per capita) does not tell us what portion of the country’s wealth is distributed in labor incomes, property incomes, or entrepreneurial incomes – the “wages, rent, and profit” of classical political economy. Nor does it tell us anything about inequality of income within a country. GDP is simply the total product (hence, gross).
The concept of GDP has a long history of criticisms. Feminist and ecological critiques argue that many economic activities – constructive (e.g. child-rearing) and destructive (e.g. pollution) – are not always assigned “value” in the market. Without a market value, these activities are not recognized as “production” in the accounting of a country’s total wealth. Alternative measurements do exist that attempt to calculate what wealth levels of a country would be like if reproductive labor or ecological destruction were taken into account.
Another critique of FX-based GDP is that international currencies fluctuate in the short term. Therefore calculating a country’s wealth via its exchange rate with the world economy distorts the “real” value of its economy. The World Bank attempts to rectify this with the Atlas method, which adjusts for fluctuations in currencies and also in variations of inflation rates.
The critique of FX-based GDP that resulted in the PPP “project” by the 1960s, however, was that there were observable differences in prices for goods and services between countries. Tradable goods and services, ones that can be bought and sold on the world market, can differ due to the national effects of tariffs, subsidies, and taxes. Non-tradable goods and services, especially the latter, differ widely around the world, with prices for many services usually much lower in poorer countries. The PPP advocates argue that income levels calculated at FX rates do not take into account these differences in price levels. PPP adjustments, in theory, recalculate FX-based national income using estimates of prices for goods and services in that country. As a result, PPP-based GDP is supposed to give a better picture of relative consumption levels in the world economy.
There are lots of methodological problems that still exist in the construction of PPP data: many estimates are not made from direct observation of prices at all but from complex calculations, the quality of various goods and services are often not the same and attempts are made to factor this in, benchmark countries are used to calculate PPP incomes for other countries with not so great accuracy, and the backward historical projections in changes in PPP-based income use growth rates from FX-based data. All of these problems are compounded when calculating PPP-based income for poorer countries. The University of Pennsylvania, who has undertaken a large portion of the PPP project, estimated that in its 5th set of PPP benchmarks (released in 1991 with analysis of 64 countries), the range of accuracy of its income data for countries with less than a tenth of the US’ income was 60% up or down. As I pointed out in my last post, the most recent 2008 PPP benchmarks (the first to actually contain any systematic price data from China) have once again radically changed our understandings on poverty and consumption for large parts of the global South.
However, even if PPP-based data was entirely problem-free in its collection, and it accurately reflected average consumption levels for each country, a more serious problem exists when using them to understand global inequalities in wealth. With PPP-based data, we attach more importance to domestic economic processes in analyzing the reasons for the relative wealth of nations, and minimize the role of economic processes that take place across borders. If PPP-based income represents the command over economic resources a population has within its territory, FX-based income represents the command a population has over world economic resources. If the latter is more relevant for understanding why countries “catch up” or do not “catch up” with Western wealth levels, then PPP-based incomes underestimate the inequality between states of global income distribution.
It is understandable, then, that most economists like PPP-based data, since their models of growth are based on domestic factors of production (land, labor, capital). The best-known model of growth, and the one implicitly referred to by Hadi-Esfahani and Pesaran in their article, is associated with economist Robert Solow. The Solow theory of economic growth argues that, given population growth is equal to zero, the accumulation of capital is driven by technological progress. Or, as Hadi-Esfahani and Pesaran put it, “new ways of producing more output given [a set amount of] inputs.” Therefore, they write, “…the study of economic growth can be viewed as the analysis of the factors that enhance or hinder the acquisition and use of technology.”
In a wider understanding of “technological innovation,” like the one used by the economist Joseph Schumpeter, “new ways of producing more output given inputs” means something more than simply the invention of new technology or the upgrading of existing technology. It broadly includes the introduction of new methods of production (i.e. the assembly line), new goods and services for the market (i.e. the automobile), new sources of supply for production (i.e. rubber grown in Brazil instead of Malaysia for tires), new trade routes and markets to sell existing goods and services, and new forms of organization that combine the various factors of production (i.e. the vertically-integrated modern corporation). This is why Schumpeter argued that the “entrepreneur” – any person or organization who accomplishes one or more innovations out of these types – is the driving force of capitalism. Recurrent economic growth is based on “creative destruction” in the market: “creation” of new profit-oriented innovations that result in the “destruction” of existing methods of production.
Schumpeter, however, also pointed out that no single technological innovation could permanently guarantee a source of profits. Competition by other economic agents would eventually increase as the innovative methods were emulated, leading to a decline in profits for the original agent. The key to capitalist success, then, was to continually shift the pressures of competition elsewhere, either by generating a continuous stream of innovations within a particular organizational area (i.e. the auto industry in its early stages) or by shifting the area itself in response to other agents’ competition (i.e. moving out of cars circa 1982 and into computers). As Schumpeter portrayed, the most successful economic agents in capitalism are those that “are aggressor by nature and wield the really effective weapon of competition.”
As Giovanni Arrighi pointed out, this “creative destruction” in the market via continued rounds of technological innovation (or, better put, entrepreneurial activity in all its forms) historically clusters the accrual of profits not only over time, but also over space. In other words, those businesses that can continually innovate in world economy tend to be grouped in particular states. This is largely because states have a big effect on how the “weapon of competition” affects their domestic constituencies. States vary in their abilities to control access to the most lucrative niches in the global production of goods and services, provide higher levels of infrastructure to support continual innovations, and create a political climate that most favors capitalist entrepreneurs over other actors. Some states do this more effectively than others because greater economic command, at least over recent world history, can be translated into greater political command.
If there were no states, and the world market was all contained in a single political entity (the dreaded “world state” of the US right), the accumulation of capital would be quite volatile over time, and we would not see such a clustering of wealth in particular areas as exists today. In other words, economic command over world resources by states and their population matters. If we want to understand the sources of capitalist growth over periods of history longer than a few years, such as Iran’s relative position during the 20th century, it matters more than economic command over domestic resources.
With that lengthy but necessary exposition let’s return to the original question: should we use PPP or FX-based data to understand relative wealth levels? The ability to continually generate technological/entrepreneurial innovations depends in the long run on participation in the world economy. In business school speak, closing the wealth gap between a poor country and the West requires “moving up the value chain” within areas of global production of goods and services, as occurred in South Korea over the second half of the 20th century. This is easier said than done, as I attempted to show in my last post, but in any case one needs a relational lens to even begin to discuss the reasons. FX-based GDP per capita is a better measure of the concept of economic command over world resources, and therefore the relative ability to sustain the continued innovations required for capitalist growth. PPP-based income data inflates a poorer country’s relative command over world economic resources, and therefore should not be used when looking at the few successes and many failures in “catching up” with the West, or when the words “economic strength and size” are invoked.
After that, though I do not wish to add any more confusion, it is rather easy to explain why per capita Gross National Product/Income (GNI) is better to use than per capita Gross Domestic Product/Income (GDP). GDP calculates the total output of production within the borders of a given political territory, no matter who generates it. GNI calculates the total output produced by all the “nationals” of a given political territory, no matter where that output is produced. Methodologically, GNI includes net income (wages, rents, profits) from foreign production (the current account balance). GNI, then, focuses on the owners of economic production. Usually there is not too much difference between the two figures, but given that sometimes it does matter, when we pick a measure to understand the relative wealth of nations, FX-based GNI per capita is the better indicator.
Economists might respond that they care most about relative standards of living in the world and not relative economic command over world resources. As I said before, there are uses for PPP adjustment as well, even with all of its methodological problems, and consumption levels are a valid use. However, there are also substantive reasons why consumption levels may not be the best indicators for relative standards of living and well-being in the world. That discussion must wait for another post.
You might notice that in the last post I used FX-based GDP per capita to show Iran’s relative wealth levels over the last 50 years. I did this so that it would be easier to discuss the article in question. In a future post I will further discuss Iran’s trajectory with FX-based GNI per capita data, though the overall trends will not be starkly different than the tables I already posted.
Thanks to Roberto Patricio Korzeniewicz for consultation on this post.
