You are here: Home > Learning > Articles

The Census of India 1881 - 1931

Related Links
Articles Main Page
Development Dark Facts

Dharampal, Written Bangalore, August 1993: A note for recollection and reference

It seems that even before they arrived in India and began to build exclusive settlements or factories1 the British were given to surveys of the places they settled in, to the carrying of censuses within them, and to the surveying and estimating of the wealth and military potential of the regions in which these settlements or factories were established. Thus the idea of the survey, census, etc. initiated by the British in India seem to go back to around 1600 AD., and by around 1750 when the British began to exercise military and political power over substantial regions of India such surveys had become quite frequent and common.

Two of the major detailed early post - 1750 surveys were of the district of Chengalpattu near Madras and of the Bazee and Chakran Zamin2 of the major part of Bengal and Bihar. The survey in Chengalpattu, held during 1767-1774, covered all the over 2,000 localities of the district, and took detailed account of all the land, water sources, houses, professions or community of the people who inhabited each and every house, the agricultural production, woods, cattle, sheep, goats, etc., of each and every locality, The Bengal survey of Bazee and Chakran Zamin, carried out in the 1770s, was equally, perhaps even more detailed. By around 1800 such surveys seem to have been carried out in most districts of British conquered India and seemingly continued from one year to another for large part of the 19th century.

In a way the survey was integral to conquest, and in time became a more effective tool of domination and control, and its results and conclusions were of major assistance in the making of state policy. However the idea of survey did not originate in India but was integral to British functioning in Britain itself. In time surveys led to the enumeration of each and every human being, besides the counting and surveying of land, houses, horses, cattle, sheep, machinery, tools, etc. The first Census was taken in Great Britain in 1801 and then at 10 yearly intervals thereafter till today.

The British began to think of an India-wide Census around 1850, but, the first such census could only be held on 17-2-1881 under the direction of a Census Commissioner at Calcutta and under the supervision of Census Superintendents in each of the provinces, presidencies, and Indian princely states. This first Census put the land area of India at 13,82,624 square miles and the population at 25,38,91,823. The textual part of the report of the all India Census was divided into chapters and it may be useful to mention them at this point.


(paras 1-29)

Area and Density



(paras 30-76)

Religion of the People



(paras 77-115)

Proportion of the Sexes



(paras 116-153)

Civil condition of the People [unmarried, married, widowed]



(paras 154-171)

Ages of the population



(paras 172-240)

Rates of Mortality and Duration of life



(paras 241-358)




(paras 359-385)

Statistics of Birth Place



(paras 386-404

Statistics of Instruction



(paras 405-433)

The Insane, Blind,Deaf, Mutes and Lepers



(paras 434-448)

Urban and Rural Population



(paras 449-691)

Caste Statistics



(paras 692-773)

Occupations of the People



(paras 774-789)

Movement of the Population



(paras 790-805)

Concluding Remarks


The first volume of the textual report of the 1881 census was accompanied by 3 other volumes of tabulated data relating to the above text and design. The four volumes together covered around 1,200 folio pages. This pattern was followed by the provinces, presidencies, etc. The pattern set up in 1881 was followed, with minor additions, supplementing, varying etc., till 60 years later, i.e. in the 7th British India-wide census in 1941. The census of 1941 was however a partial census as the British and their government in India were then engaged in a major European war. The published material therefore was much less voluminous. In all, the report of these seven censuses, 1881-1941, add up to around 1,20,000 folio pages, and have for sometime been available on 3,693 microfiche, produced in Switzerland, from the printed copies kept in the India Office Library, London. India, the subject of the census does not seem to possess a complete printed set, either for the whole of India, or a province of it.

What is available in the London India Office Library, and that which has been put on microfiche, is perhaps also only around half of what was printed for public or official use. Though it seems to include all the reports pertaining to the provincial and all-India levels, it does not include detailed material pertaining to the villages and towns, and perhaps some pertaining to taluks and districts which is stated to have been printed at least in the earlier census. Individual reports prepared by some of the smaller Indian princely states also do not seem to be included in the set on microfiche.

The more voluminous of the available reports are for the census of 1891 (on 722 microfiche), 1901 (on 751 microfiche) and that of 1931 (on 588 microfiche). Amongst them all the report on the Mysore census of 1891 (102 microfiche) seem to be the longest.

Though each census begins with a description of land area and topography of province and region to which it pertains, and in the later decades, especially in 1921 and 1931 to the decay of indigenous Indian industries, etc., or to the extent of the spread of the modern power driven industry, or even census of cattle, the main concern of each census was to gather information on the number of people inhabiting India and the racial, cultural and social characteristics of the people of India, as well as their economic divisions and activities. Till about 1911 there was major speculation and exploration about Indian religion and religious sects, and even more so about Indian social divisions described, listed, enumerated, and variously tabulated under the term “caste”. The interest in caste seems to be highest around 1891 when the census, especially for Punjab, NWP and Oudh (the present Uttar Pradesh), the Madras Presidency and the state of Hyderabad came out with what were termed as Index of Castes. The Index for Punjab listed over one lakh names of what was termed as sub-castes, that of Uttar Pradesh 54,000, for Madras Presidency around 30,000 and for Hyderabad around 5,000. The number of sub-divisions amongst the over 40 lakh Muslims, Hindus and Sikh Jats of Punjab in 1891 were listed as above 11,000. For numerous other groups such subdivisions ranged from 1,000 to around 5,000. In the Madras Presidency the Paraiyars had around 350 subdivisions and the Palli had around 365 subdivisions Some modification of the Punjab list was made in the census of 1911. But even this modification would have left Punjab with over 50,000 names of “castes”. The Punjab Jats for instance, still retained 4,473 subdivisions in the modified list of 1911.

The following table gives the number of subdivisions of the 15 selected Punjab castes as given in 1911 Punjab census.

Subdivisions of 15 selected castes in the Punjab



1911 (Revised)

1. Agarwal



2. Ahir



3. Awan



4. Biloch



5. Brahman



6. Chuhra



7. Fakir



8. Jat



9. Khatri



10. Lohar



11. Macchi



12. Musalli

(with Chuhra above)


13. Rajput



14. Sheikh



15. Sonar



It is probable that if all the provinces and presidencies and princely states had printed such indexes the total number of such names may have ranged around 3 to 5 lakhs. Though there possibly were major errors in these indexes and listings, it seems that each household at the time of enumeration was defining itself by indicating its lineage and the name which it gave was that of the kula, of which the household was a part, and not the name of its gotra, or sub-caste or caste, as seemingly assumed by these indexes.

According to these reports the number of the major castes in any province, presidency, etc., was in the range of 30-50 and these accounted for around 75% of the population of the province; or in provinces like Bengal, where the overwhelming majority of the Mussalmans, by stages began to be clubbed together as "Sheikh", they by themselves formed around 75% of the Mussalman population. Besides these major or numerous castes, there were around 100 to 300 smaller groupings perhaps totaling around 25% of the population, which quite possibly included amongst them large number of groups engaged in special crafts and professions. Additionally most provinces had 100-300 other minor groupings, each numbering less than a few thousand and many only in hundreds.

Though caste seems to have fascinated and perplexed those who planned and directed the census, the ratio between males and females, in the Indian population seems to have interested the British more. The knowledge of prevalence of infanticide and exposure of very young children, in European antiquity and later in the European middle ages, seems to have directed the attention of the British officers in this direction. Moreover their knowledge since about the mid 17th century of the practice of Sati in certain parts of India, and reports of the prevalence of female infanticide amongst certain groups of Rajputs in Gujarat and in certain districts of Uttar Pradesh may have also aroused their interest in this area. From this the belief arose that Indian society was given to female infanticide, or at least to neglect and maltreatment of its females.

The 1881 India wide census enumerated the total number of males as 12,99,41,851 and of females as 12,39,49,970 in a total 1881 population of 25,38,91,821. The eastern (especially Bengal and Orissa) and southern India (most of Madras Presidency, Mysore, Travancore) had slight majority of females, while the central, northern and especially the north western areas (Punjab, Kashmir, NWFP) had a varying excess of males. While the South and the East had an excess of females to males similar to Germany, Netherlands, Spain, etc., the Central and Northern regions were more like Greece which had 51.7 males to 48.3 females. Amongst large areas only the Punjab showed a much higher male ratio of 54.25 males to 45.75 females.

The stress on this low female presence and investigation around it had been set on by the British much before the 1881 census. A census taken in 1872 in NWP had made a special study of it and produced a 600 page report on the sex ratio in the NWP. It may be indicative of British high concern that Mr. W.C. Plowden who had directed the 1872 NWP census, was made the first Census Commissioner for the 1881 Census. It may also be mentioned here that while in large parts of India there was a slight excess of females over males, and in most they were about equal, in the latter part of the 19th century, after such manifest British concern there was a marked decline in the proportion of females in most parts of India. By the time the British were made to quit India, a further decline in the proportion of the female population (in comparison to its proportion before 1900) was of the order of 5%.

From one decade to another, the census came out with further elaborations of the older data as well as information on hitherto unrecorded aspects of Indian life. Much of such information perhaps was largely exotic in its nature as the details of the 18 and 9 phanas into which most of the communities in the Mysore state were said to group themselves, or the detailed rituals of a large number of Brahman communities, or for that matter of the jats and other communities of northern India, or that out of the total of 19,630 listed villages in Mysore 7,935 of which had the ending "Halli", 1,289 had the ending "Uru" and 1,770 ended with "Pura".

The early census especially have much discussion and comment on Indian religions, religious sects, castes, etc. A few comments from the Punjab census of 1881 may be reproduced here.

1. The effect of Hinduism upon the character of the followers:
“(Hinduism) can hardly be said to have an effect upon the character of its followers, for it is itself the outcome and expression of that character ... In fact the effect of Hinduism upon the character of its followers is perhaps best described as being wholly negative. It troubles their souls with no problems of conduct or belief, it stirs them to no enthusiasm either political or religious, it seeks no proselytes, it preaches no persecution, it is content to live and let live. The characteristic of the Hindu is quiet, contented thrift. He tills his land, he feeds his Brahman, he lets his womenfolk worship their gods, and accompanies them to the yearly festival at the local shrines, and his chief ambition, is to build a brick house, and to waste more money than his neighbour at his daughter's wedding”.

2. On Mussalmans (of Eastern Punjab)
“In the eastern portion of the Punjab the faith of Islam, in anything like its original purity, was till quite lately to be found only among the Saiyads, Pathans, Arabs and other Musalmans of foreign origin, who are for the most part settled in towns. The so called Musalmans of the villages were Musalmans in little but name. They practiced circumcision, repeated the Kalimah, or mahomadan profession of faith, and worshipped the village deities. But after the mutiny a great revival took place. Mahomadan priests traveled far and wide through the country preaching the true faith, and calling upon believers to abandon their idolatrous practices... But the villager of the East is still a very bad Musalman... As Mr. Channing puts it, the Musalman of the villages ‘observes the feasts of both religions and the fasts of neither’.”

3. The Impure and outcaste tribes
“I have said in the beginning of this chapter that the impure and outcaste races are not generally recognised by the higher castes as belonging to their religion, even though they may profess its tenets and observe its injunctions. These people include some 2,012,000 Hindus, 173,000 Sikhs, 492,000 Mussalmans and some hundreds of Buddhists... I am sorry to say that we are singularly ignorant of the practices and beliefs of these outcast classes. Generally it may be said that such of them as have not become Musalmans usually burn their dead and marry by Phera, while most of them have Brahmans to attend them in their ceremonies, though these Brahmans have become impure by association with their unclean clients, and have been excluded from communion by their unpolluted brethren.”

4. Effect of conversion upon caste
“The Musalman, Rajput, Gujar, or Jat is for all social, tribal, political and administrative purposes exactly as much a Rajput, Gujar, or Jat as his Hindu brother. His social customs are unaltered, his tribal restrictions are unrelaxed, his rules of marriage and inheritance unchanged; and almost the only difference is that he shaves his scalplock, and the upper edge of his moustache, repeats the Mahomedan creed in a mosque and adds the Musalman to the Hindu wedding ceremony. As I have already shown in the chapter on religion, he even worships the same idols as before, or has only lately ceased to do so. (This is much less true of the middle classes of towns and cities. They have no reason to be particularly proud of their caste; while the superior education and more varied constitution of the urban population weakens the power of the tribal custom. In such cases the convert not infrequently takes the title of Sheikh though even here a change of caste name or conversion is probably the exception.)”

5. Impact of Islamic Conquest on Caste
“Indeed it seems to me exceedingly probable that where the Musalman invasion has not, as in the Western Punjab, been so wholesale or the country of the invaders so near as to change bodily by force of example the whole tribal custom of the inhabitants, the Mahomedan conquest of northern India has tightened and strengthened rather than relaxed the bonds of caste; and it has done this by depriving the Hindu population of their natural leaders the Rajputs, and throwing them wholly into the hands of the Brahmans. The full discussion of this question would require a far wider knowledge of Indian comparative sociology than I possess. But I will briefly indicate some considerations which appear to me to point to the probable truth of my suggestion... We know that, at least, in the earlier and middle stages of Hinduism, the contest between the Brahman and the Rajput for social leadership, of the people was prolonged and... (see Muir's Sanskrit Texts, Vol. I). The Mahomadan invaders found in the Rajput princes political enemies whom it was their business to subdue and to divest of authority; but the power of the Brahmans threatened no danger to their rule, and that they left unimpaired.”

“The information which comes through the Census relates to a large variety of Indian social, cultural and economic aspects. But the more weighty of the information generally conveys a sense of unbelievable terror, horror and agony. To the people about whom such information is given it must have been, at least in a localised, immediate experiential sense very familiar. They had to live through its reality perhaps for a century or more and yet survive it as a people. But its value as well as its horror consists in its being a consolidated account or indication of what they must have been through.”

The Madras Census of 1891 mentions, "In Madras about 28 percent of the children born, die before the completion of the first year of life, and one half of them are dead before they reach the age of nine in the case of males, and before they are fourteen in the case of females”. And it proceeds to add, "In England the population is not reduced to one half until the 45th year for males and the 47th year for females." It can be taken that what was happening in Madras was more or less true for other parts of India too.

Another fact, equally horrifying, relates to the proportion of widows in the Indian population. The average India-wide figure for 1881 or 1891 which is given informs us that out of every 100 females (from the age 1 to the age of 60 and above) 20 were widows. The meaning is that in every average group of five females there were two unmarried girls, 2 married women and on the average one widow. In certain parts of India, and among certain groups, the proportion of widows amongst the total females went up to one-third (33%) and more, while in many others, perhaps in the remaining about 60% of the population it was below 15%.

There are many other details of this kind. But even considering half the children dying by the age of 9 and 14 and one fifth of the women living as widows implies a society not only of great sorrow but one which had been emaciated to the extreme. When and how such a situation was reached is a matter for deep extensive investigation. But it can perhaps safely be assumed that this phenomenon on this scale had started soon after the British began conquering India. The conquest was accompanied by prolonged plunder, enforced emaciation of the people and by the neglect and exhaustion of India's land and its other assets. The traditions and ways and beliefs of India may have introduced further complications in this enforced situation. The continual dying of the young would have led to not only wishing and awaiting the birth of yet another child but also to the tendency to have a larger number of children so that at least one or two of them would survive in situation where at least half of them were not expected to survive at all. Further, it would have also led to a considerable lowering of the age of formal marriage especially amongst those who wished to class themselves as twice-born, as being married was equal to being auspicious. If one were to die soon after such a marriage it was perhaps thought the right thing that one had at least already nominally entered the married state.

Given such a death-toll it is far more likely that the 100-150 years of active British intervention and conquest had halved the population of India by around 1880, or perhaps even a few decades earlier, as the black death had halved the population of major parts of Europe in the 14th and 15th century. If that is what, on investigation, is found to have happened, India around 1750 could have had a population of some 30-40 crores and not the 15 to 20 crores as assumed hitherto. It may also be worth considering that the notion of 5 persons constituting a household does not seem applicable to the 18th century and earlier Indian society. It is more likely, and the agricultural productivity data seems to support it, that the household in India was much larger and often included a few other persons besides the currently much celebrated nuclear family of husband and wife and 2-3 children.

Along with the holding of the census, various major policies of directing Indian society were being experimented and implemented at the ground level especially in the latter part of the 19th and early 20th century, amongst different strata of Indian society. In fact the census was a major tool for gauging the impact of such policies and measuring their success. Before the first India-wide census in 1881 one or more census had been undertaken in every province, princely state, etc., during 1850 and 1875. Their results were conveyed in 1875 in a 64 page memorandum prepared in London for presentation to the British Parliament. The total population of India under British control, including the princely "Indian" India added up to 23,88,30,958. The share of Indian India in this total, for which not many details are given in the memorandum, was 4,82,67,910.

The memorandum had then termed 87,12,998 persons as outcastes in a total Hindu population of 14,91,30, 185 and a total India-wide population of 19,03,63,048 for British India. This total included 4,08,82,537 Musalmans (1,95,53,831 being in Bengal which included Bihar and Orissa). The Musalmans were further grouped as Saiyads (7,90,984), Pathans (18,41,693), Moghuls (2,19,755), Sheikhs (47,00,320) and "unspecified" (3,26,74,800). It is to be noted that 30-35 years later the census of 1911 gave the number of Sheikhs in Bengal alone as 22, 952 ,944 and the number of Sheikhs for India as 3,21,31,342. The memorandum of 1775 gave the number of Sheikhs in Bengal as 10,69,497.

Besides including the "outcastes" (87,12,998 more than half of whom were in Madras) the 14,91,30,185 Hindus included 1,77,16,825 persons under the category "Aboriginal tribes or semi-Hinduised Aborigines", 5,95,815 native Christians, 11,74,436 Sikhs, and 28,32,851 Buddhists and Jains. 2 4,47,831 of the latter number were Buddhists living in the then British Burma.

The number of those known as outcastes (or later as depressed, untouchable, in 1931 the exterior castes, and finally on being grouped together on various provincial schedules termed as the Scheduled castes) more or less got multiplied in the same way, as the category "Sheikh". The 1931 census commissioner, J.H. Hutton, termed them as Exterior Castes and made their total come to 5,02,50,347 in a total 1931 Hindu population of 23,86,22,602. Similarly those called "Primitive tribes" in 1931 were enumerated as 2,46,13,848.

Policy seems to have been quite active about the achievement of literacy in English, and perhaps in all possible elimination of the mention of "Sanskrit" from the census schedules. While around a hundred world languages even including Estonian, Japanese, Chinese, etc., were put on the census schedule and on the language returns, Sanskrit seems to have been clubbed in the category "other languages". From around 1891 data was being collected on literacy for a fairly large number of selected castes. The Punjab literacy table for 1891, running over 238 folio pages, was titled "showing by caste, tribe and race, the literate population distinguishing those who know English from those who do not." In a total Punjab population of 2,51,30,127 the total number of literate was given as 7,99,177 males and 20,205 females. Those literate in English numbered 40,556 males and 4,887. Out of these 21,849 males and 4,116 females were of English origin and a small number were from other European countries or were Eurasians. But by then 4,193 Khatris, 2,684 Brahmans, 1,307 Aroras, 1,149 Banias, 1,319 Sheikhs, 763 Kaith, 736 Rajputs, 653 Jats, 565 Billoch, 517 Saiyads, and 1,375 Native Christians had become literate in English. English literacy had also reached 139 Sunars, 80 Barbers, 80 Lohars, 157 Takkhans, 35 Julahas, 25 Telis and 99 Jhinwars. Amongst the 771 Indian females who had become literate in English, the largest number 589 was of Native Christian females. There were also 28 Khatri, 19 Sheikh, 8 Billoch, 13 Brahman, 10 Rajput, 11 Saiyad, 2 Kalal, and 1 Chamar female who had acquired English literacy in the Punjab by 1891.

There is continuous caste-wise data on literacy and literacy in English for all provinces, states, etc., till 1941. The data seem to suggest as if there was a planned linkage between general literacy and literacy in English and that the former was not to overly out pace the spread of English literacy. A detailed analysis of this data may indicate that by and large the ratio between total general literacy and literacy in English continued to be in the range of 10 to 1 and in some less westernised regions as 15 to 1. Persons in many castes undoubtedly lagged far behind in the acquiring of any literacy and especially English literacy. There were however groups like the Brahmans, Khatris, Kayasthas, and others amongst whom in certain cities and towns around half the males had become literate in English by 1931.
1 sort of fortifications with buildings for storage, etc. And not really places which went for any production of commodities.
2 lands the tax of which had traditionally and historically been allocated to religious, cultural and charitable functions, and also towards the remuneration of local establishments including those of the police and militia systems.