Wednesday, June 25, 2008

The Size of the Indian Middle Class

(Everybody seems obsessed with size these days, including 90% of your SPAM folder) In a recent PBS documentary on the rise of the Indian middle class (brought to my attention by my friend Omar "He who is not AMOK" Karim), one of the numbers bandied about was that the middle class in India was 30% and defined as those earning between $5000 and $25000 per year. (The documentary is worth watching, if not for anything else, for observing the paranoia that seems to be afflicting Americans about how the rise of consumerism in India and China will impact the American way of life). In a subsequent discussion with my colleague Kapil Paranjape here, we tried to figure out how such a number was at all possible. Let me start with a very simple analysis. According to Nationmaster the Gross National Income (GNI) per capita in India is $128 (all numbers are for yearly incomes). Let us assume for extreme simplicity that the middle class is 30% and there is only one more class whose income is x. We will ignore for now say, the super rich and the super poor. Then it is clear that if N is the population of India .3 * 5000 * N + .7 *x * N = 128 * N (The N is fake anyway and can be factored out and we can work with GNI per capita but I am keeping it just for the sake of clarity). Notice that the above equation cannot be satisfied for any positive x. Even if we assume that there are many other classes earning y, z, etc. with suitable fractions, these will be added contributions to the left hand side (LHS) of the equation, and while we would not be able to solve the equation since there are multiple unknowns now and one equation, we could still conclude that positive solutions of x, y, z ... are not possible since the first term on the LHS is already substantially bigger than the RHS. One of the reports that Kapil dredged out from the Economic Times claims the GNI per capita to be $750. Note that this number on the RHS of the above equation still does not allow positive solutions for x etc. Now that its a free for all, since nobody seems to have any reliable numbers, let us keep $750 on the RHS, and 30% as a size of the middle class. Then the maximum earning (i.e the upper threshold) that the middle class can have is $2500 obtained by putting x=0 i.e. assuming no other strata of society (which, I think, is also the middle class dream!). Clearly for any positive x and keeping the fraction as 30%, the yearly average earnings of the middle class would be quite a bit below $2500. While this analysis has been fairly simplistic, it exposes some gross features of this problem which clearly need correcting. How do we correct this?
  • The size of the middle class is substantially lower
  • The yearly earnings of the middle class are substantially lower
  • Combination of both
  • The GNI per capita is substantially higher
  • Since black money is supposed to be 50% of the Indian economy, double the figures for GNI (which GNI?) per capita. The point being that black money is usually not invisible to GDP calculations which are based on turn-overs but black-money is often hidden from GNI which is usually based on IT revenue calculations. (this idea is due to Kapil).
Anyone have the true numbers or knows where to find them? Update: See the comments by Debajyoti. Assuming these are family incomes we are talking about (say 4 members per family) the first term on the LHS of the equation gets divided by 4 allowing a reasonable solution to the equation for a GNI per capita of $750 or $800. (The figure of $128 by all accounts seems far out!). Thanks Debajyoti for a Occam's razor solution. Of course the GNI per capita is indeed per capita since these are statistical quantities based on total tax returns divided by population and so on, whereas the $5000 for the middle class is based on a general perception of salaries of earning members of the middle class.


AMOK said...

Having been a student of India since early days, this blogger wishes to share some facts with the eminent fellow-bloggers. Speaking of facts, as you read on you will realize that Don Qui-Basu de La Mancha is running Amok, fighting windmills -- assigning blogger attributes, names and friendships as freely as he bandies a country's GNI, solves equations and manipulates statistics. No offence, but this blogger is only interested in baggage-free intellectual pursuits.

The Indian middle class of say 300-odd million seems to be everyone who subsists on more than $2/day. This is of course about $700/year. The GNI from various sources, including UNICEF is of the order of $800 or $900. Re-doing the eminent blogger's arithmetic, estimating the non-middle average at $1/day, 30% middle class and a GNI of $900, we find the annual income of the middle class to be, on the average, about $2000. Round numbers. [Where Dr. Basu obtains a GNI of $128 remains a mystery! The lucky 2^7? ]

Now to address the issue of paranoia. Nobody wants high prices, most certainly not the many politicians who profit from many of the international activities. Oil/Bush is an example, the eminent bloggers herein will have their own lists. They are eager to shift blame onto others so they can continue to get re-elected etc. Just as in India migrants from one part "take" the jobs of the locals, so too in America it is not the local economy or politics. It is those "others" who eat more and are to blame. Sad but true, this scenario is a constant in public life and must be recognized as such.

Rahul Siddharthan said...

Dear AMOK or OAK or Omar or whatever your name is,

You're trying too hard to be witty and it's really not coming off. Can you try to be articulate instead? That would be rather useful. I'd like to know what you're saying.

It would also be nice if you used your real name on your posts.


Rahul Basu said...

The number $128 as I explained, came from which is actually quite a standard site for statistical and economic data for different countries. (The fact that it is also 2^7 as Sir AMOK points out perhaps has deep numerological significance which escapes me right now). It is indeed somewhat low. As I also explained, other sources like the Economic Times give different numbers. In fact, a google search of this topic produces all kinds of numbers.

Your number of $2000 is close to my $2500 which I had suggested, to square with the GNI per capita of $750 that the ET quotes.

About the paranoia point which was made just in passing, I was merely pointing out what some people in the video claim, and which both George Bush and Condoleezza Rice have also mentioned in their speeches which created a furor here. You will perhaps recall, as an avid reader of my blog, that an early post of mine actually defended Bush as having merely stated a fact to which there was no need to react violently as people in India had done.

Your over-reaction to my post continues to be a mystery....which excess baggage are you exactly talking about?

Anonymous said...

One respect in which the documentary differed from Mr. Bush and
Ms. Rice was that they did have one person, the Asia correspondent of
Forbes (whose name I forget), espouse the view that this crisis of
over-consumption should make Americans think about how much
they are using of the world's resources.

The question remains. Is there a way to arrive at a level of
consumption such that if everyone tries to emulate it, it will still
be sustainable? Moreover, is it possible to convince those from the
first world to lower their levels of consumption to this
sustainable level? One can then try to convince the burgeoning
middle classes of the rising economies to limit themselves to these

This was also Dr. Pachauri's (I thought rather clever) response to a
question posed by the BBC during his joint interview with Mr. Gore
after the peace prize. The question was about the threat posed to the
environment by the rapid growth seen in India and China.


Rahul Basu said...

It's unlikely that either the first world or the third world (I really dislike these labels) peoples will voluntarily lower their consumption levels based on some arguments presented to them. Only when it pinches in the pocket is such a thing likely -- for example $4 gallon gasoline and so on. Americans are once again thinking of smaller cars and the Hummer has finally, thankfully been taken off the drawing board.

Unfortunately in India, it pinches the classes who can least afford higher prices. Thus, most of us may not notice higher prices for sugar, rice, or vegetables but the poor are being squeezed very badly. Thus as of now, there is no incentive for most people in the middle sections of the middle class to reduce consumption. There is still enough spare cash.

I am not sure how this argument works in the US presently with the rise in prices.

AMOK said...

Sire Basu, thank you for your additional notes and intellectual comments. You are absolutely correct, unless it pinches in the pocket book people don't care. The disposable income forms a cushion. In the US, based on news reports, there is a knock-on effect on the cost of living. Of course there is the whole housing market collapse that is playing a major role as well.

Your baggage will not be in excess once you cease the unwarranted juxtaposition of this blogger's ID with the names of your dear friends. Apologies to one of your other commentators who was unable to follow my witticisms. When a Greater Blogger like yourself popped the baggage question, it is understandable that a far, far Lesser Blogger was left entirely in the dust.

Debajyoti said...

Could it be that the definition of $5000 p.a. as a lower cutoff for being termed a part of the middle class referred to a typical family unit, and not to the income per person ?

If you take this as the starting point,
assume an average family size of 4 or 5, AND consider a fig of, say, $750 as the per capita GNI, the maths holds up.

And while some might think that $5000
p.a. for a family of 4 would not
qualify for a middle-class tag, my
hunch is that it would (think of a
single income family of, say, a school
teacher; a bastion of the Indian middle class)

Debajyoti said...

You could get more info from the
GoI pages. Specifically, those of
the Ministry of Statistics and
Programme Implementation

See, for example,

As you'd recognize, the official
Net per capita income is closer to
$800 (Rs 33,300) p.a. at 2007-08

Incidentally, the Finance Minister
had projected a similar number in
a lecture at Yale in 2005.

Debajyoti said...

On Occam's razor soln:

It just shows that one needs to do
some phenomenology rather only field
theory :-) Take heed, my dear fellow !

AMOK said...

Here we find that not only is the Indian (and Chinese) middle class creating these high food prices as discussed in this post-- but now also the mercury levels in the Great Salt Lake in Utah!! Is there no limit to the "ill" effects of progress of India and China?