How many people died in 1971? ("The missing number")

Rumi_'s picture
Submitted by nazmul09 [Guest] on Mon, 25/02/2013 - 3:58am

So, how many people died in 1971?

There have been many studies on this (you can search online). However, the researchers have found it difficult to get the most accurate number since the Pakistani army killed a lot of people indiscriminately and arbitrarily, making one layer of mass grave on top of another. (Compare that to the Holocaust, the Nazi's literally recorded almost every Jewish victim that were put in prison and killed thereafter).

I wanted to do some calculation based on the population of Bangladesh in 1970 & 1975. I know that population based estimation cannot guarantee the most accurate number. However, I hope that it will still give you an estimate of the missing number of people in 1971.

Here I go.

1. According to UN estimates [1], Bangladesh population in 1965, 1970, 1975 are as follows:

1965: 57.792 million

1970: 66.881 million

1975: 70.582 million

Online Graphing

Do you see that the rate of increase suddenly dropped from 1965-1970 (approx 9 million) to 1970-75 (approx 3.7 million) ? This will be my focus point.

2. According to UN estimates, the rate of population increase in Bangladesh during the same years are as follows:

1955-60 2.85

1960-65 2.856

1965-70 2.92

1970-75 1.077

1975-80 2.66

Online Graphing

Again, see the drop in the rate of population icrease

3. I believe that this drop in rate of population comes mostly from three tragic scenarios.

i) 1970 Cyclone in East Pakistan , ii) 1971 war and iii) 1974 famine of Bangladesh.

Now, what would have happened if the rate of population increase followed the trend of 1960-65 and 1965-70? How many people would be in Bangladesh with that trend?

Let's take a middle figure of 65-70 and 75-80's population increase rate. It would be 2.79.

Starting from 1970's population, if rate of increase were 2.79, 1975's population should have been:

66.88 * (1 + (2.79/100) )^5 million = 76.74 million

The actual population at 1975 = 70.582 million

The difference = (76.74 - 70.582) = 6.22 million


Let's look at cyclone, famine and another important figure.

a) 1970 Cyclone: Up to 0.5 million people died in that cyclone . [3]

b) 1974 famine: Up to 1 million people died in the famine. Approximately, up to 0.5 million people died in the post-famine diseases. [4]

c) Migration of Hindu refugees in India: A huge portion of Hindu refugees stayed in India during 1971. The estimated number is 1.5 million. [5]

d) There is another factor, though (relatively smaller in number).

Migration of the biharis to Pakistan: In 1974, 170 thousand (0.17 million) Biharis went to Pakistan [6]. Add the 150 thousand Bihari's (a neutral estimate) who died in the revenge killing during the aftermath of 1971 war, that number comes to 0.32 million [7]

Overall, add the four above numbers:

0.5 + 1.5 + 1.5 + 0.32 = 3.82

So, approximately 3.82 MILLION PEOPLE DIED/LEFT BANGLADESH during 1970-1975. (I considered the maximum no. for cyclones and famines) Subtract that from the original:

6.22 - 3.82 = 2.4 million

Therefore, approximately, 2.4 million people went missing from our/Bengali's side during the 1971 war (I have already calculated the Bihari's in separate figure).

2.4 MILLION MISSING IN A SPAN OF 9 MONTHS? (Compare that to the killing of 6 million Jewish people by Hitler during 6 years)

I do not claim that my number is exact/perfect. But I believe that further work on this arena, by proper statisticians, will give us a much better estimate. Apart from the 1970-75 period, the curve of rate of increase in our population is pretty smooth (law of large numbers normally work, you know). Therefore, a good estimate of the four above factors (cyclone, famine, Hindus staying in India and Biharis migrating to Pakistan) should give you a good estimate of the number of Bengalis/Bangladeshis (accounting for the aboriginals) that died in 1971.

I hope that some statisticians will do further work in this arena. After all, I am a wireless communication researcher, not a statistician হাসি.

Additional Comments (added later):

As some people have pointed out, I have not used the effoct of death toll during '70 cyclone and '71 death on the expected number of people in 1975. I agree that this should be taken into account. If we do that, the overall number would come closer to 2 million (from 2.4 million).

However, note that, I used the maximum death toll, among all available sources, for 1970 cyclone (0.5 million) and 1974 famine (1.5 million). If you use some other lower estimates for these two tragic scenarios, the estimated death toll of '71 will increase.

Overall, my objective was not to provide 'a perfect number' for 1971 death toll. Many researchers say that the death toll will range from 1 million to 3 million ( Researchers like R. J. Rummel use 1.5 million [7]. My analysis tells me that the range of 1.5 to 3 million should be correct with a high probability.


1. (Table: total population both sexes)

2. (Table: rate of natural population increase)


4. (page - eight)





মহা_পুরুষ's picture

look at the above book, pages: 15, 17, 18, 19.

according to this, there has always been steady migration from Bangladesh to India.

So, I was wondering if it has been so, then should the migration during 1971 have any significant effect on the dip in growth rate in Bangladesh? (as we are assuming)

nazmul.islam's picture

I agree that there has been a steady growth. But, during the 1971 war, there was an influx (not as much as 1947, though). It should be counted, I believe. The 1.5 million number went to Wiki from a report on the (the link is not working, somehow ইয়ে, মানে...). This number might not be correct. A better estimate will help.

সাফি's picture

Great analysis

মরুদ্যান's picture


যদি তোর ডাক শুনে কেউ না আসে তবে একলা চল রে

অতিথি লেখক's picture

Wonderful analysis

কৌস্তুভ's picture

(1) correct reference for the broken link for no. 5:

but the article doesn't cite any sources. it reads:
"But the crucial question is - how many preferred to stay back? According to an unofficial estimate, while 9.27 million refugees returned by the end of March 1972, another 1.5 million refugees continued to stay in India. Many of them were Hindus who merged with post-Partition refugees to join the swelling ranks in settlements like the Promodnagore Colony in Dum Dum."

so this reference doesn't actually stand either.

(2) any references for saying "20-30 thousand Bihari's who died in the revenge killing during the aftermath of 1971 war"? Pakistani estimates claim 0.5 million died just during the war.
Ref 6 here is about the number of Biharis who moved to Pakistan, not about the number killed.

(3) if 0.5 million died in 1970, one should subtract that from 1970's population of 66.881 million before applying the growth rate. if you subtract that number after applying the growth rate to project to 1975, the gap would obviously look larger.

nazmul.islam's picture

I agree with most of your points.

(2) You are dead right. I was sloppy. A neutral estimate of Bihari killing is 150,000 (Rummel, This will reduce my total number by 0.1 million.

Also, I will spend more time on the no. of Bihari's that went to Pakistan during that time. It might be more/less.

(1) Yes, Hindu times does not show any source. Shall I look into the population statistics of West Bengal, Tripura during 1965-70 and 1970-75? Let me know if you have some other ideas.

(3) The gap would actually be smaller, no? Since, 0.5 million ppl died in 1970, they could not have affected "the expected population" at 1975. Taking that into account, the number of missing people would be 0.05 million less.

But, keep in mind, I took the upper bounds for cyclone and famine death (the data says up to 0.5 and 1.5 million). If these numbers are somewhat lower, the no. of ppl died in 1971 would be higher.

Please send me more links if you have. We need more/solid data before we go out with this.

Rumi_'s picture

Updated the no. of Bihari death and its effect on the overall death. Used Rummel's data for Bihari death estimation, that should be as neutral as it gets.

কিংকর্তব্যবিমূঢ়'s picture

I did some research on the issue, Nazmul covered most of it here. Here are some other facts worth considering:

1. The first academic debate against this calculation will probably come by highlighting the fact that Ayub Khan started the first Family Planning Scheme for Pakistan as a part of the country's Third Five Year Plan (1965–1970). The scheme's goal was to have a vast impact in the shortest time possible, with a reduction of the birth rate from 50 to 40 per 1000 by 1970. Now, arguments can be made that the drastic population drop was because of this program. Although the program officials boasted about its success, in her paper "Policy Making in Pakistan's Population Programme", Ayesha khan revealed the following:

"The National Impact Survey(1968-69),conducted by the federal Council's Training, Research and Evaluation Centre(TREC) and assisted by Johns Hopkins University, revealed that three years after the Scheme began, only 6% of married women were using any contraceptive method.It also showed that the level of IUD use and retention was much lower than the government had estimated. It suggested a wider range of contraceptives to be offered to the public, and a greater emphasis on outreach and motivation in order to reduce the gap between knowledge and practice."

link to her paper: []

So, the myth is busted.

I was also skeptical whether the program was active equally in East and West Pakistan, as the central government always paid more emphasis on the western part. However, I didn't find any reliable info on this, so not bringing it in the argument.

2. After independence, Bangladesh (predictably) did not launch any Family Planning program immediately. "In 1976 a national policy for population control and family planning (FP) was announced which employed thousands of full-time field workers, and developed information, education, and motivation activities. The implementation strategy entailed the integration of health and FP, maternal child health (FP/MCH) service delivery systems at subdistrict (upazila) levels with a wide choice of contraceptive methods and expanded good quality services."
Source: []

So it couldn't have been that the people of BD suddenly decided not to take any kids after independence.

3. During the nine month or war, commonsense says there will not be a lot of "child-producing activities" [I admit, I did not find a better term here.] So there may be a dip in birth-rate just after the war. But again, we are doing our calculation on a five year basis, so the effect of nine months should be minimized in the overall calculation.

nazmul.islam's picture

Thanks Fahim.

We will use more data and then improve our estimation.

ইয়াসির আরাফাত's picture

I have to say it's a very interesting way of finding the missing number of people during the liberation war in 1971. Dedicated research is necessary to develop the model and to come to a concrete conclusion.

The only drawback I saw in your calculation is to consider steady growth rate during 1971-1975 time-span compared to previous batches. '71 was a year of crisis, which should heavily influence the birth rate negatively. Also the famine in 1974 would do almost the same. Nevertheless, the end number looks quite convincing. I appreciate the effort.

Rumi_'s picture

Thanks. You have a valid point. I'll have to see what birth rates we can use for '71 and '74. I think that we can find a lower bound of ppl that died in '71 based on the four other factors that I mentioned in my analysis.

Guest's picture

For the 1974 famine death counts, Amartya Sen's book (chapter-9) says:

A case study of the 1974 famine in Bangladesh, which was associated with the floods of that year, and had an official mortality of 26,000


Rumi_'s picture

Thanks for providing the link. The official estimates are always very low and far from reality. Based on the other reports that I have read, 26000 does not seem to be an acceptable figure.

Although, 1.5 million '74 famine + post-famine death is the maximum number that I have read among all reports. I used the upper bound because I wanted to get a lower bound on the death toll of '71 war. I will have to go through other links to see if this number goes down. In that case, the death toll of '71 will go up. Inform me if you know of other analysis.

All these inputs enrich the future analysis.

Post new comment

The content of this field is kept private and will not be shown publicly.
Enter the characters shown in the image.