For reasons not always well established, the middle class in modern societies is viewed with favourable eyes. Defined in terms that are often subjective, that group is seen as directly or indirectly underlying social stability and driving growth. Based on results from its Global Attitudes database, the Pew Research Centre, for example, found that members of the middle class defined as belonging to a specific income range were more satisfied with their social condition and positive about their future than the poor. That makes them a force for stability, when compared to the poor or near-poor who are likely to be dissatisfied by their condition and not as hopeful given past experience. And the very presence of the rich, who are always a small minority, is likely to generate unrealisable aspirations and a sense of injustice that cannot contribute to stability. What is more, the Pew study found that the middle classes are more likely to favour forms of parliamentary democracy, the rule of law and freedom of speech and expression, making it a force for political fairness and stability.
That’s a sociological view of the middle class. From an economic perspective, ever since it was argued that high growth in poorer economies (as South Korea once was, for example) could result in catch-up or the transition of a country from low income to high income status, the middle class has also been seen as a force for growth. A burgeoning middle class is seen as expanding the domestic market because of a rising share of income available for discretionary spending after satisfying its demand for basic necessities. It is also seen as supporting productivity increases by serving as the seedbed for innovation. Hence, whatever the initial growth trigger (such as net exports, for example), the presence and rapid expansion of a middles class is seen as contributing to sustaining the process and ensuring the transition to developed country status.
Empirical verification of this hypothesis requires defining what qualifies an individual to be included in the middle class, which would help define the size and growth of that class over time, and the association of the phenomenon with democracy and/or growth. The most recent such estimate is that available in a study released early July by the Pew Research Center, an independent think tank. The Pew study defines the threshold income for entry into the global middle class as $10 per day as measured in 2011 Purchasing Power Parity (PPP) dollars (about which more later). The $10 threshold is seen as one which considerably reduces the probability of an individual falling back into poverty with an income of less than $2 a day. The ceiling income for inclusion of an individual in the middle class is $20 in constant 2011 PPP dollars.
The news about the size of the global middle class in 111 countries accounting for 88 per cent of the world’s population based on that definition is not all too good. Between 2001 and 2011, which included the years (2003-2008) when many countries experienced an economic boom, “the share of the world’s population that subsisted on $2 or less per day halved from 29% to 15%. At the same time, the share of people classified as low income ($2-10 per day) increased from 50% of the world’s population to 56%. That 6 percentage point increase, in turn, was matched by a similar increase in the share of the world’s population that could be considered middle income (5 percentage points). Meanwhile, the share of the world’s population in higher-income categories barely changed between 2001 and 2011.”
The study concludes that: “While the doubling of the global middle class, from 7% in 2001 to 13% in 2011, is certainly a major shift with potentially major economic and political ramifications, it is important to keep in mind that, at the end of the first decade of the 21st century, the vast majority of the world’s population (71%) remained either poor or low income.” The upper bound used to cap the middle income category of $20 in 2011 PPP dollars may be questioned. But raising that ceiling to $50 does not make too much of a difference with the increase in the size of the middle class between 2001 and 2011 rising from 6 to 8 percentage points.
The income range used to define the middle class suggests that identifying the middle class as consisting of some proportion of the population earning incomes within a band on either side of the median (or central) income in a country is insufficient for the purpose. This would be particularly true if the issue of concern is the role of middle class presence in driving growth: a threshold income, that may differ from those around the median, is required to drive demand and growth at a pace that permits catch up; incomes must be at levels that ensure the educational access needed to acquire the capabilities to support innovation.
This does create one difficulty. If income has to be at a level that creates demand for commodities that drive high growth in a globalised and integrated world, the threshold income that includes a person in the middle class has to be globally defined. This has spurred interest in recent years in identifying a “global middle class” and assessing its distribution across countries, whether poor, “emerging”or developed. That exercise requires one important adjustment, however. Since the issue here is the purchasing power of the individuals to be included in the middle class, using a nominal income range defined in dollars to identify that class, for example, would be inadequate. Citizens in different countries make their purchases in domestic markets and the prices of the same commodities vary across countries even within the integrated European Common Market. Moreover, many commodities and especially services are not traded and varying costs of production and market structures across countries result in significant differences in the prices of these ‘non-tradables’. So taking a common set of dollar income ranges and converting them into local currencies at official exchange rates, to divide the population into low, middle and upper income groups would not do. The bundle of goods and services those incomes could purchase would vary significantly. So the income taken must not just be in dollars that are converted at official exchange rates, but dollars measured in purchasing power parity (PPP) terms after applying a conversion factor that takes into account, besides the official exchange rate, differences in the dollar prices across countries of specifically identified commodity bundles. An income denominated in PPP dollars for each country should then be capable of buying the same bundle of goods and services in the local market.
Once the PPP conversion factors have been identified, all that remains to be done is to identify the income ranges in the benchmark country that separate the population into low, middle and high income groups. Once that is done, the corresponding PPP dollar ranges for each country can be identified and a similar separation achieved. If such an exercise is resorted to, to measure the size of the middle class in a country in a particular year, the result would depend on (i) the income range used to separate out the middle class in benchmark dollars (in some ‘base year’); and (ii) the PPP conversion factors used to identify the corresponding PPP dollar income range in the country concerned.
The income range used is a decision taken by the individual or group conducting the exercise. PPP conversion factors, on the other hand, normally come from a common, internationally accepted database. However, these conversion factors vary over time, as new and improved price surveys are conducted to identify the price of an appropriately revised bundle of goods and services. Thus, till recently the conversion factor yielded figures in constant 2005 PPP dollars. But new, and presumably improved, conversion factors deliver estimates in constant 2011 PPP dollars. In April 2014, the International Comparison Program of the World Bank updated PPP to 2011 constant international dollars. The revised purchasing power exchange rates result in significantly increased estimated GDP per capita values for many countries.
Differences of these kinds have yielded varying estimates of the level of changes in the size of the middle class in different countries or across the globe. One definition of the ’global’ middle class (used by economists Branko Milanovic and Shlomo Yitzhaki) identified it as consisting of those with incomes that fell between the average income of Brazil (the floor) and Italy (the ceiling). That worked out to a range of around $12 to $25 per person per day in 2000 PPP dollars. If that definition is used, the middle class in the so-called emerging markets works out to around 400 million in 2005 or about 6 per cent of the world’s population. Unhappy with what that meant for growth and democracy, others tried arriving at a more “inclusive” definition of the middle class. Thus, Martin Ravallion, while still at the World Bank, attempted an estimate of a middle class consisting of those earning more than the World Bank’s poverty line income in developing countries of $2 a day in 2005 PPP dollars and less than $13 a day which was the US poverty line. That took the size of the middle class in emerging markets to 1.4 billion as early as in 1990 and 2.6 billion in 2005, or from one-third to one-half of the population of developing countries (The Economist, February 14, 2009). If that were indeed a high-spending middle class, the world economy should have been booming.
It is in this context that we should consider the estimates emanating from the Pew study. They seem to suggest that in the most successful years of the neoliberal project, the expected expansion of the global middle class, which in turn is required to sustain high growth and the transition of more than one or two emerging markets to developed country status, has not been realised. There have been exceptions of course. Three decades after the launch of the reform programme in China in 1978, the expected expansion of its middle class seems to be occurring. The middle class as defined by the Pew measure has grown from 3 per cent of its population in 2001 to 18 per cent in 2011. That implies the entry of 203 million people into the middle income category, which does seem enough to sustain high growth, if this expansion is not reversed.
India on the other hand has not performed well. To quote the study: “Although the poverty rate in India fell from 35% in 2001 to 20% in 2011, the share of the Indian population that could be considered middle income increased from 1% to just 3%. Instead of a burgeoning middle class, India’s ranks of low-income earners swelled. Many of these were people hovering closer to $2 than $10 in daily income, and thus still a ways from the transition to middle-income status.” This tells a very different story from what has been sold in India.
Using an income range of Rs.70,001 to Rs.105,000 per year in 1998-99 prices, the National Council of Applied Economic Research has reportedly found through its surveys that the share in India’s population of the middle class rose from 10.1 per cent in 1989-90 to 13.9 per cent in 1998-99 and 22.1 per cent in 2009-10. According to the report from The Economist referred to earlier, the NCAER found that the share of Indians earning $6-$12 a day in PPP dollars at 2000 prices, rose from 18 to 41 per cent of the population between 1995 and 2005, while the share of those earning $12-$60 rose only from 2 to 5 per cent over the same period. So if you give yourself some flexibility in defining the middle class income range and base yourself on the NCAER’s surveys, India too can be seen as having followed a Chinese-type trajectory, despite the fact that its reform effort came much later.
What the Pew study calls for is caution when building these stories of a burgeoning Indian middle class. That would also call for caution when launching celebrations based on assessments such as those from the IMF’s recent World Economic Outlook Update that India is likely to overtake China and become the fastest growing country globally in terms of GDP growth rates.
(This article was originally published in the Frontline, Print edition: August 7, 2015.)