A Comparative Examination of Sinhala, Tamil and Muslim Demographic Trends in Sri Lanka

 Chandre Dhamawardana … with highlighting emphasis added by the Editor, Thuppahi

It is a common belief that the Moor population, nearly 99% Muslim in faith, have high demographic rates and also have to capacity to have high birth rates because of Islamic laws that provide for the possibility of having several wives. Hence it is of interest to examine these contentions in the context of demographic data available from the Dept. of Census and statistics, Sri Lanka.

We examine the data for the whole of Sri Lanka, for the Ampara district, Madakalapuwa (Batticaloa) and for the  Colombo District. The respective plots of the  percentage population in each district is plotted, while the actual population figures are shown in the plot for the whole country.

National data

First we review the National ethnic demographic data, that were included in our study on claims of Genocide by pro-LTTE activists[1]. The  plot of the  national data and a selection from the data are given below (note that the Y-axis is logarithmic), and “Indian Tamil” data are not included in the Tamil total.

Let us look at a selection from the nation-wide data.

Year     Moor,    Sinhala     Tamil

1981    1047K    10980K    1887K

2018    2062K    16784K    2510K

average 1555K    13882K   2198K

The growth is over 37 years.

If you take the Moors, 1047K have grown to 2062K.

Hence  the total change is   1015K.

So the average change per year is 1015K/37=27.4,

This change is due to an average population of 1555K

So the growth rate of the Moor population per 1000K, i.e., per 1000,000 is  17.6K. Finally  we have

Moor growth rate per million is  17600

For the Sinhalese                 11300

For the  Tamils                       7700

So the National Muslim growth rate/million per year  is 1.56 times that of the Sinhalese during the last 4 decades.  The low Tamil growth rate becomes slightly larger than that of the Sinhalese if emigration of Tamils is included[1].

More Sinhalese have also emigrated than Muslims. But we don’t have enough data to correct for that. Our  guess is that when emigration is corrected for, the Sinhalese growth rate becomes 14000 per million/year compared to about 18000 per million/year for Moors.

Although the national growth rate of the Moors per million is some 1.6 times larger than, say the Sinhalese, there is absolutely no basis for the belief that the  majority ethnic group will be “over-run” by these minorities.  The majority growth rate for the whole population is an order of magnitude (approximately  10 time) more than that of the Muslims because the Sinhalese are more numerous.

Data for the Ampara district

In the following we look at Sinhala, Tamil and Moor populations, where “Tamil” does not include those designated as “Indian Tamil”. As an example of the analysis we use, the data for the Ampara district is given below

A selection of the data is given below for discussion. Here we have data only up to 2012.

Year     Moor,    Sinhala     Tamil

1981    161.7K    146.9K    77.8K

2012    282.5K    251.0K    112.81K

average 222.1K           168.8K 95.3K

Growth rate per million/year,  i.e.,  1000K/year for the three Ethnic groups are:

17.5K          16.9K    11.8K

That is,

Moor growth rate per million is  17545, nearly 17600

For the Sinhalese                 16879, nearly 17000

For the  Tamils                      11847. nearly 12000

So, the Muslim growth rate is  very similar to that of the Sinhalese during the last 4 decades. Since it is is a growth rate for a district, this is a rate which is moderated by people moving out of Ampara as they get wealthier. This is more true for the Moors than for the Sinhalese who are less mobile.

The linear projections (dashed lines) show that the Sinhalese will have a slight majority in the Ampara district by 2015. Projecting beyond that, say to 2035  are  not meaningful.

The Tamil growth rate of Tamils becomes  similar  to others if the emigration of Tamils is included.  But we don’t have enough data at the district level to correct for that. While the legal possibility of having multiple wives exists, most Muslims cannot afford to maintain three or four families, an in practice the effect of such a law has a lower impact than what one may believe, a priori. Furthermore, birth rates drop drastically when the literacy and prosperity of a society increases. While many Muslims are very rich, the average Muslim is impoverished and uneducated. These sociological factors must also be considered in understanding any high birth rates shown by any community.

Madakalpuwa (Batticaloa) District

 The Batticaloa district is very different from the Ampara district as it was affected by by the Eelam war. The Sinhalese population, which was already very small, dwindled during the war years, and has very slightly recovered since. Because of the smallness of the Sinhalese population, the y-axis has been plotted on a log-scale to bring out the low-lying data.

Year     Moor,    Sinhala     Tamil

1981    139.7K    1319K    170.6K

2012    242.7K     1771K    231.3K

Moor growth rate per million is       16690

For the Sinhalese                    -19033 (negative)

For the  Tamils                          15036

Hence, in the war-ravaged Madakalapuwa (Batticaloa) district, Moor population has done better than the Tamil population. Here again, this is partly because many Tamils who become some what prosperous move to other provinces and then to Western countries like Canada.

Colombo District

The Colombo district is the economic dynamo of the country, and there is significant emigration to Colombo of those in the “out-station” areas who become wealthy.  Similarly, there is an outflow of the even more wealthy emigrating to the West.

Year     Moor,             Sinhala  Tamil

1981     139.7K    1318.8K     107.6K

2012     242.7K           1771.3K  231.3K


rate of growth/million per year, Sinhala      9447

rate of growth/million per year, Tamil         9748

rate of growth/million per year, Muslim    17372

Hence the rates of growth of the Sinhalese per million per year in the Colombo district are comparable. However, there is a strong growth of the Moor population (nearly twice the rate). A significant part of this is due to the immigration of Moors from the Eastern province and the North during the war years.


The view that districts like Ampara are being rapidly populated by Moors has been used by various groups to raise the temperature of political militancy, and create competitive feelings among ethnic groups. There does not seem to be much justification for this view. The Moor population of specific districts like the Colombo District has certainly risen very fast, during the Eelam war years.

The national growth rate  of the Moors per million per year is some 1.6 times larger than, say the Sinhalese. However,  there is  no basis for the belief that the  majority ethnic group will be “over-run” by these minorities.  The majority growth rate for the whole population is an order of magnitude (approximately  10 time more) than that of the Muslims. However, no useful attempt can be made to project these ethnic populations beyond 2025 with any confidence.


[1] Chandre Dharma-wardana, OIDA International Journal of Sustainable Development, Ontario International Development Agency, Canada. ISSN 1923-6654 (print) ISSN 1923-6662 (online) http://www.oidaijsd.com ………….. Also available at http://www.ssrn.com/link/OIDA-Intl-Journal-Sustainable-Dev..html

****  ****


Chandre-Michael, 17 May 2020: “Gerry knows far more about Purana Villages in the Eastern Province interior and its evolution. He should be the one to write about it.. I have not done any historical research. Merely looked at census data which do not support the xenophobic hysteria against the Muslims. I have also looked at the Tamil claims about Canada Tamil diaspora. They have vastly exaggerated their numbers to show that they have a lot of voter power. So this is a highly controversial subject.”

Gerald Peiris to Michael & Chandre, 18 May 2020:

Michael, Chandre’s statistical analysis is flawless, and his comment about the  anti-Muslim hysteria is also irrefutable. What more is there to write about without repeating information that has already been documented by several other researchers including Gerry? ………..Best regards
Response From Chandre, 18 May 2020:
Thanks Gerry for a strong endorsement! However, I haven’t looked at Kandy, Puttalam etc.

Some one remarked that Muslim rates are very high in several other provinces that I have not looked at.

I doubt it, but who knows. Maybe some one should look at all 24 districts and do a comprehensive study.


Gerald Peiris: “An Appraisal of the Concept of a Traditional Homeland in Sri Lanka,” 26 April 2013, being a reprint, https://thuppahis.com/2013/04/26/an-appraisal-of-the-concept-of-a-traditional-tamil-homeland-n-sri-lanka/

Gerald Peiris: “A study of Contemporary Buddhist-Muslim Relations in Sri Lanka,” August 2017, https://www.researchgate.net/publication/319999061_A_study_of_Contemporary_Buddhist-Muslim_Relations_in_Sri_Lanka

Indrapala, K. (1970), “The Origin of the Tamil Vanni Chieftaincies of Ceylon,” The Ceylon Journel of Humanities 1(2), pp.111-140.

Keane, John (1905), “Report on Irrigation in Ceylon,” Sessional Paper XL V-1905, Government Printer, Colombo.

Turner, L. (1922), Population Census of Ceylon 1927, Government Printer, Colombo.

Roberts, Michael 2010 “Ethnic Identity in Sri Lanka’s Pre-capitalist Past,” https://thuppahis.com/2010/08/15/ethnic-identity-in-sri-lanka%E2%80%99s-pre-capitalist-past-shanie-darshanie-and-roberts/

Leave a comment

Filed under centre-periphery relations, communal relations, economic processes, governance, growth pole, historical interpretation, Islamic fundamentalism, island economy, life stories, modernity & modernization, Muslims in Lanka, politIcal discourse, self-reflexivity, Sinhala-Tamil Relations, sri lankan society, the imaginary and the real, transport and communications, truth as casualty of war, Uncategorized, unusual people, working class conditions, world events & processes

Leave a Reply