IS the middle class around 70 million (or nearly 40 per cent of Pakistan’s population) as a recent article in this newspaper suggested?

After the publication of this article titled “Consumption conundrum” by Sakib Sherani (an ex-bank executive), a former World Bank official Shahid Javed Burki writing in another Karachi daily seemed to endorse this view although he had put this estimate around 40 million people or 24 per cent of the population in his article in Dawn in November 2010.

Both the estimates appear to be exaggerated and are apparently based on inadequate analysis and on statistics whose reliability is suspect. Further, the studies are based on certain debateable assumptions and one has to examine these carefully to avoid reaching conclusions that may be divorced from the reality.

It is important to note that little consensus exists on the specific parameters that define the middle class. Merriam Webster defines the middle class as follows: “a class occupying a position between the upper-class and the lower-class ... a fluid heterogeneous socioeconomic grouping composed principally of business and professional people, bureaucrats, and some farmers and skilled workers sharing common social characteristics and values.”

In layman’s terms, people who are neither rich nor working class or poor and have some education and skills are most likely to form the middle class. A growing middle class is typically associated with brighter growth and economic prospects for a country and has significant economic as well social and political implications. Asian middle class growth is a development with global significance and has been described as once in a generation investment opportunity.

Development institutions such as the World Bank use absolute measures of income to define poverty and use income distribution indicators to study income inequalities in countries and related trends.

According to the World Bank, 60.2 per cent of Pakistan’s population had a daily income of $2 a day in 2008. Some use this income measure to define poverty. However, the Bank’s own measurements do not seem to be very reliable. The Bank had put the percentage of poor in Pakistan (using $2/day measure) at 73.8 per cent just three years earlier in 2005. And it is not just me who has doubts about these statistics. The Asian Development Bank study (cited in the “Consumption conundrum” article) noted that Pakistan surveys had reliability issues.

Moreover, it is difficult to argue that in just a few years nearly 16 per cent of the population moved out of poverty particularly during a period when the GDP growth declined from 6-7 per cent to 3-4 per cent and the inflation hit double digits. Slower economic growth and higher inflation tend to increase poverty as the people on the lower end of the working classes are among the most vulnerable.

However, if we take the World Bank’s poverty estimate at its face value, 60 per cent of Pakistan clearly cannot form part of the middle class. It is tempting to put the rest in the middle class. However, this approach would be too simple. Why? Even according to the World Bank data that Burki uses to support his estimate, 80 per cent of the people live under $3 dollar a day or less. If we define the top two per cent as the rich, only 18 per cent of the population would fall into middle class category. This would mean 31 million people in 2010 using the World Bank figures for income and population and ignoring any skewed income distribution.

Now let’s review what the much cited Asian Development Bank report (Middle Class Size in the Past, Present,and Future: A Description of Trends in Asia, published in September 2010) has to say about the size of the middle class in Pakistan.

The ADB report uses estimates of income measures to divide the population in ranges. According to the report, 23.2 per cent of people in Pakistan live on less than $1.25 a day, 36.6 per cent of people survive on $1.25 to $2 a day, 35 per cent live on daily income of $2-$4 a day and only 7.33 per cent of the population has income of $4 a day or more. At least in one respect, these findings are consistent with the World Bank data that puts the percentage of poor in Pakistan at 60 per cent using the $2/day yardstick.

However, the question of middle class is a tricky one. All those who are not living above the poverty line cannot be put in the middle class. A study (’Estimating the size of the middle class in Pakistan’ by Dr Durr-e-Nayab) conducted by the Pakistan Institute of Development Economics (PIDE) in 2011, reported that the size of the middle class in Pakistan is around sixty million.

The study argued that the concept of the middle class is a multidimensional phenomenon that cannot be adequately captured by the existing definitions based merely on households’ income or expenditure categories. The measure suggested in the PIDE study is a composite of five weighted sub-indices of factors deemed to be important for being part of the middle class, namely, education, occupation, income, lifestyle and housing. This sounds like a fair approach but it cannot be ignored but income and education drive occupation, housing, and lifestyle, the last indicator being very subjective.A simple fact is that there are working class households who live on the border of poverty, have tight budgets and do not have the purchasing power that is typical of a middle class household. In Pakistan, a household with six people and an average income of $3/day per person had an average monthly income of about Rs 44,000. This would put that family in working class with little room for spending on a motor-bike or a car, eating out, health, education or other such items. I am sure head of such a household would not be amused if she was told she belonged to middle class.

Many upper or middle class families in Karachi, Islamabad, or Lahore who employ domestic servants would readily relate to this as it is common in families with 6-7 members to have the adults work in such jobs and supplement the main bread-winner’s salary to achieve a total household income of Rs40,000 or so.

In this context, it is useful to look at other indicators. In 2007, Pakistan had 1.44 million passenger cars in use and 2.7 million motorcycles. Even if we assume all those had a motorcycle belonged to a middle class household and assuming six members per family, the number of people belonging to the middle class households would not have exceed 25 million in 2007.

While it is true that the production of motorcycles has touched one million mark in Pakistan it is the actual total usage (because old motorcycles are discarded) that is relevant for the discussion of size of the middle class notwithstanding the fact that both the public sector departments (e.g. police, militias) and private sector businesses (banks, courier and private security companies, etc.) buy motorcycles and those sales are not part of family or consumer expenditure. The same observation applies to passenger cars with current annual sales figure of 160,000 units.

There is another indicator for the size of the middle class; female literacy in the youth. According to the United Nations Educational, Scientific and Cultural Organisation (UNESCO), 71 per cent of the eligible girls did not go to secondary school in 2009. The low rate does not support the case that Pakistan’s middle class accounts for nearly 35-40 per cent of its population.

Based on the analysis of both the World Bank data and the Asian Development reports as well as other indicators such as the demand for big ticket consumer spending items like cars and motorcycles, I think the size of the middle class in Pakistan is about 20-25 million (or 11-14 per cent of the population) with a total income of $30-40 billion or 17-22 per cent of national income.

Now that is a large enough size for many multinational companies to be attracted to Pakistan, for a lot of restaurants to be full in the cities and many shopping malls to be busy till late in the evening. But we should not get carried away with statistics that present a rather rosy view of the living standards of our people, an overwhelmingly majority of whom live under or just near the poverty line.