Indian Journal of PsychiatryIndian Journal of Psychiatry
Home | About us | Current Issue | Archives | Ahead of Print | Submission | Instructions | Subscribe | Advertise | Contact | Login 
    Users online: 2929 Small font sizeDefault font sizeIncrease font size Print this article Email this article Bookmark this page


    Advanced search

    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Email Alert *
    Add to My List *
* Registration required (free)  

   1931 And 1941
    Days of Future i...
    Article Figures

 Article Access Statistics
    PDF Downloaded431    
    Comments [Add]    
    Cited by others 1    

Recommend this journal


 Table of Contents    
Year : 2012  |  Volume : 54  |  Issue : 1  |  Page : 32-36
The census of India and the mentally ill

1 Consultant Psychiatrist, Sitaram Bhartia Institute, and Fellow, Nehru Memorial Museum and Library, Teen Murti House, New Delhi, India
2 Department of Psychiatry, Molecular Genetics Laboratory, National Institute of Mental Health and Neurosciences, Hosur Road, Bangalore, Karnataka, India

Click here for correspondence address and email

Date of Web Publication3-Apr-2012


Context/Background: Epidemiological data have long been considered essential for documenting incidence of disability and planning services. India has been conducting census operations for a long time, and this information may be relevant in the current context.
Aims: To document the prevalence of insanity, and discussions about treatment and disability arising out of mental illness in India (1850-1950).
Settings and Design: The material used was located at the British Library and the Wellcome Library, London; the Teen Murti Library, Delhi, and web-based archives.
Materials and Methods: We have retrieved and summarized the coverage of psychiatric illness in previous census reports from the 19 th and 20 th century.
Statistical Analysis: None, this relies upon historical archives and documents.
Results and Conclusions: Differences in incidence and prevalence of insanity, as well as biological and psycho-social factors in the causation, and outcomes, of mental illness are all discussed in these census reports. Comparisons are often drawn to other countries and cultures, and impressions drawn about these differences and similarities. Similar concerns persist to this day. Disabilities and mental illness were not enumerated since the census of 1941 and have been restored only recently, and this lacuna has hampered planning in the post-Independence era. As we debate policy and plan interventions using contemporary census data, it may be useful to remind ourselves of the issues, then and now.

Keywords: Epidemiology, census, history, psychiatric illness, disability, health planning

How to cite this article:
Sarin A, Jain S. The census of India and the mentally ill. Indian J Psychiatry 2012;54:32-6

How to cite this URL:
Sarin A, Jain S. The census of India and the mentally ill. Indian J Psychiatry [serial online] 2012 [cited 2021 Sep 16];54:32-6. Available from:

   Introduction Top

As we wait for the report of the census of 2011 and its partial inclusion of disability (that caused by mental illness being only partially covered) in the enumeration, it may be worthwhile to examine some aspects of the coverage of the mentally ill in previous census operations. This inclusion in the current and future census operations is meant to help plan services for the disabled and the disadvantaged. Census reports of the past, and the analysis and interpretations they offer, would help us understand how services were envisaged, and compare those efforts to what are being conceived now.

   1871 Top

The very first attempt to estimate the number of mentally ill in India and the differences in rates of insanity was made in the census of 1871. [1] The census of 1871 was also the first time that estimation of the number of "imbecile, idiot or lunatic" was attempted in the UK itself, and the India data thus offered a useful counterpoise. The number of insane and idiots in India was estimated at 67,000 (1 in 2700), a proportion which was one-eighth that of England and Wales (where the total number of lunatics was estimated as 39,567 of whom 35,790 were in Asylums, in a population of 22.7 million). [2] Absence of over work, over excitement, and low rates of intoxication in India were thought to be one reason for these low rates, while it was also suggested that a chronic starvation perhaps did not supply enough nutrition to the brain (to allow the furor of insanity to occur). Differences between Europe and India were also being observed clinically by then. Commenting on the trends in outcomes for admissions into the Bengal asylums (1848-1870), in the Indian Medical Gazette, [3] it was noted that the proportion of admissions who recovered (55-63%) was higher than in England (34%) and Victoria, Australia (49%). It was suggested that the better outcome was a reflection of the fact that most admissions were for those who were recently ill and the numbers of chronic residual ill were rather small. Death rates were also high, so the outcome for patients was either recovery or death (30-36%, as compared to 20-30% in England and Australia). The census thus detected lower incidence and a background of better recovery for mental illness in India. Within India, Mysore region was reported to have a higher rate of insanity than the adjoining Madras. Some reasons (psycho-social and biological) for these differences were suggested.

   1881 Top

By the time of the next census in 1881, [4] serious concerns were already in place about the errors in enumeration caused by inadequately trained field staff. This was particularly worrying as the rates detected (35/100,000) were a fraction of those in Europe (which varied from 130/100,000 in England to 38/100,000 in Sweden). Sweden and northern Europe were at that point in time considered underdeveloped and poor, and thus closer to the social situation in India. Even more surprisingly, there were inexplicable differences within India, with Burma recording 100/100,000 while Cochin and North West Frontier had 17/100,000. These differences were examined for the influence of age, gender, geography, religious factors, and interactions between these. No consistent pattern emerged, so it was assumed that the variations were caused by errors in enumeration. Specifically, it was thought that those with milder forms or periodical insanity (and were currently normal) had not been counted at all.

The census of Mysore (being an independent territory) was done separately. [5] The initial count of 1871 had found that Mysore had a higher rate of insanity that many other parts of India, which prompted a specific question from the Government regarding the possible reasons for this, and the census of 1881 tried to answer these. Dr. Houston and Dr. Henderson, from the Asylum in Bangalore (now NIMHANS), who had lived in Mysore for many years, said they could not think of any possible reason for the higher rate of insanity in terms of geography or temperament. However, the Census Officer (Mr. B. L. Rice) speculated that consanguinity, social breakdown, and the large and rapid demographic changes in Bangalore caused by migrations from Maratha, Hyderabad, Madras, and Malabar regions (over the previous two centuries on account of the various Anglo-Mysore Wars and the British occupation) had given rise to a very "vagrant and bohemian" lifestyle that could account for this increased prevalence of abnormal behavior, that could be linked to the higher prevalence of insanity in Mysore.

   1891 Top

This census [6] discovered a general fall in the number of mentally ill, caused, as the report discreetly puts it, by the "general condition of the people" having deteriorated. The famines of this decade over southern and western regions of India caused a great decrease in the number of mentally ill, [7] along with a reduction in the numbers of persons afflicted, in all the disabilities. As a group, the disabled suffered the worst, as the proportions declined dramatically. However, by now, a more detailed description was being attempted. Various causes (consanguinity, brain disease, epilepsy, intemperance, etc.) were identified. These included the high incidence of cretinism in certain areas, the location of shrines in the north-west that looked after the mentally ill (leading to an increase in prevalence in certain areas, e.g. Shah Shuja Mosque in Lahore that looked after people with microcephaly, who were called "Shah Shuja ke chuhe (mice)"), the litigious nature of particular societies (especially in the north-east), as also the effects of consanguinity, polygamy, and polyandry (on the male:female ratios of insanity in case of the last two). Intemperance for alcohol and cannabis also predisposed to insanity, especially in tribal populations in central India (whereas tribals as a group across India were otherwise thought to be "densely sane").

Several kinds of diagnosis were entered into the census returns. Variations within India and comparisons to Europe continued to be made, with prevalence rates in India being close to that of Italy, but much lower than that of the UK. The prevalence of insanity was thought to be related to an index of social development, and the relative positions of various European countries kept shifting in comparison to India. The famine itself was acknowledged only in passing, by the fact that the returns from Mysore and Madras were accurate, though they showed a marked decrease in the numbers of the ill. It was also noted, "The spread of hospitals, and of dispensaries in the rural parts of the country must have tended to check the growth of insanity, since the patient, who perhaps may be only temporarily deranged or epileptic, can be attended to by trained agency before the disease is confirmed." An early version of the district mental health plan (DMHP) was perhaps already in place, along with preventive aspects of psychiatry! The actual impact of the famine would not be acknowledged till the next census.

   1901 Top

In this census, the effects of the famine were formally acknowledged. The number of mentally ill had declined by 11% all over India, a huge decrease. It was suggested, "To some extent this result may be due to better diagnosis, but the main reason seems to be that the insane suffered far more than the general population in the years of scarcity of food confirmed by the fact that the decrease is even greater in the native states than it is in British territory." Bombay, Bengal, Berar, and Assam - all showed huge decreases in the rates of the mentally ill, attributed to the unhealthiness of the decade and an unusually high mortality amongst mentally afflicted.

The census then went on to address concerns about etiological factors proposed earlier and also draw inferences about specific issues. Since the Hemp Commission had already made its findings public, a more careful reading of the census confirmed that levels of ganja use in the community did not affect the incidence of psychosis. The effect of consanguinity received "little support" as local variations and habits were too diverse to draw any conclusions. For every five males afflicted, they could count on only three females, and their conclusion was that this was more to do with seclusion in the purdah/zenana, or enforced widowhood, rather than an actual gender imbalance. The gender imbalance, however, was worst in the Punjab and United Provinces, but not so skewed in Bengal and Madras (a trend that persists to this day). The effects of race and religion (compulsory data points) and caste (an optional data point) on prevalence of insanity were documented. Differences between the Mongoloid and Dravidian groups, and of particular clusters within these broad groupings, were detected. Prevalence by ethnic, religious and caste groupings suggested that the prevalence declined gradually, being the highest in Eurasian and Parsees, followed by Muhammadans, and within the Hindus, "the liability to insanity varies roughly with social position - the highest castes suffer most and the lowest castes least of all."

   1911 Top

These records [8] extended observations in more detail. Though the prevalence of insanity had increased by 9% since the previous census, it was still 14 times less than in England and Wales. As earlier, highest rates were reported from the fringes of the empire (Burma and the north-west), but local variations were also high (within Assam, Lushai Hills reported a rate 8 times more than the plains). The gender difference was noted, but now interpreted with a "feminist" tinge. It was noted that in communities where women were more empowered (e.g. in Cochin), the rates of affliction were equal to men. Thus, the "protection of custom" in more traditional societies prevented mental illness in women! The variations based on caste were not thought to be useful at all, as within the Brahmins, there were regional differences even within southern India, with prevalence declining in order from Malayali, Canarese, Telugu, and Tamil Brahmins. The effect of geography itself was not consistent enough to be meaningful anymore. Similarly, the effects of habits, temperaments, consanguinity, or social practices were noted, but were dismissed as having doubtful validity. The effect of religion was also discounted.

Interestingly, local ideas about the causes of insanity, and treatments were also collected. The list of causes included worms in the head (much later identified as delusional parasitosis), neglect of worship, curse by Yogis, Tantric practices going out of control, and demon possession. Treatments included certain herbs as medicines ("susumi saag" is mentioned by name), amulets of various kinds, giving patients obnoxious material (mud from particular tanks, soup of toads), and others. Hereditary influences were recognized, and observations were made that insanity often skips generations and also that while insanity was transmitted by the father, imbecility was transmitted through the mother.

   1921 Top

This report [9] was prefaced by a description of the difficulty in correct identification of insanity. The discrepancy from Western data was now ascribed primarily to errors of detection. One pertinent account quotes a medical man (an unnamed Indian) who described a person who had been ill for a long time as normal, saying that the person was "not exactly mad, but like this for ten years." It was suggested that in India, the lay mind (including the Indian medical professional) conceived insanity only as acute mania and madness, and other forms were simply not reported. The underreporting was also suggested by cross comparisons. Physical disabilities (deaf-mute) were much more frequent than the insane, and the Indian population was, according to the reporting officer, the only population in the world (usually, in contrast, the numbers of insane were far greater than the deaf-mute) to report this. The gender inequality was now thought to arise from gross inaccuracies in reporting. The theoretical assumptions suggested that the rates of illness should be much higher in women (female infants were neglected, early child rearing, brutal obstetric care, widowhood common) in view of the stresses they were exposed to. The census data, which showed lower rates in women, were thus probably erroneous. Most tellingly, data from those in private practice suggested that female insane patients were actually more common. The reporting office was thus forced to conclude, "It is evident that the returns of the incidence of insanity are incorrect and useless."

Various correlations imputed in previous reports (religion, caste, diet, consanguinity, ganja use) were firmly discarded as non-significant. The societal response to the mentally ill was commented upon for the first time. As the editor comments sadly, "The lunatics' life is not a happy one. The congenital idiot is often kindly treated, but those who develop it later receive little sympathy … cooling and mud baths if violent, bound hand and foot or have a heavy log fastened to his ankle."

Summing up the data [Figure 1], the census counted 72,907 people in British India, of whom 10,757 (14%) were in the asylums. No similar data were available from the Native States, as the mentally ill were confined to jails in these regions (a tendency that persists to this day in many areas). Implicit suggestions were made that future plans for the mentally ill include provision of care for a larger proportion in the asylums at least at some point in their lives.
Figure 1: Census of 1921 showing regional disparities in the numbers of mentally ill within India[9]

Click here to view

The census of Mysore [10] listed less than 900 as mentally ill in the whole state, a quarter of whom were in Bangalore (due to the presence of the asylum there). The proportion classified insane were 15/100,000 of those living, three times less when compared to Baroda (45/100,000) and 30 times less than the UK (449/100,000). It was observed that the highest numbers of the insane in Mysore were recorded in 1871. There was a sharp decrease in 1881, caused by the famine of 1878, when the total infirm fell from 18,480 to 7836, especially the mentally ill, who decreased from 2980 to 767 (it was finally acknowledged that almost two-thirds of the mentally ill died). The influenza epidemic of 1918 was also listed as a possible cause for the second decrease in the number of infirm between the census of 1911 and 1921 (total infirm decreased from 12,245 to 9936, while the insane decreased from 1334 to 869). The effects of the famine and viral infections on the epidemiology of mental illness thus had an early recognition. The asylum at Bangalore itself housed 181 subjects, only 1 more than a decade earlier. However, almost 11,000 men and 7500 women were listed as beggars, vagrants, witches, wizards, etc., which is still a common category error in defining mental illness in the community.

   1931 And 1941 Top

The census of 1931 [11] did not have anything new to offer. Insanity was still listed as one of the infirmities. Differences in accurate definitions still worried the compilers. One hilarious example was of a person who was listed as insane, deaf-mute, blind, and leprous in Lucknow. When the returning officer checked this remarkable entry, he found that a particularly disgruntled lady enumerator had listed her husband under all these categories. All the differences between regions and groups that so worried previous census officers were thought to arise mainly from mistakes in enumeration.

The census of 1941 is not easily available in the public domain in its entirety. In any case, infirmities and insanity were not included in the enumeration, ostensibly due to inherent faults in detection methods. It was abandoned in the UK at the same time (probably because the proposed universal health care plans made these data redundant). However, independent reviewers were already aware of the need for reform in social and health care in India and the disinclination of the departing Imperial Government to be pro-active in this regard. [12],[13] Whether the category was excluded for the methodological difficulties or as a cover-up for the glaring deficiencies in services for the disabled (existing and planned) is thus debatable. Infirmities were not included again, even after Independence, in the census returns (perhaps for the same reasons and same debates) till 1981, and the mentally ill were by then firmly excluded from this list, as by now insanity had been transformed from an infirmity to a treatable disease and a veritable industry of medical and psychiatric epidemiology had begun exploring the same questions.

   Days of Future in the Past Top

Counting the numbers of the mentally ill was integral to the census operations in India in the beginning and began at the same time as in the UK. However, as the definitions of insanity changed, and subjective symptoms and professional interviews supplanted "common knowledge," the data that were being generated were thought to be inadequate and erroneous.This was probably one of the major reasons for the difference between prevalence rates from India as compared to Europe. However, a similar transformation occurred within the UK, where the early 20 th century census counts were thought to underrepresent the mentally ill, as a medical diagnosis could no longer be appropriately derived by enumerators or by the public itself.

Moreover, the academic responses to the enumeration allow us to interpret the documents from various perspectives. Social development (or progress) was thought to be inexorably linked to increase in insanity in the 19 th century. India was thus placed close to southern Europe or Scandinavia (in terms of social progress) as compared to Western Europe (Britain, France, and Germany). It is interesting that Scandinavia is now at the top of the Human Development Index, while India has diverged in a rather extreme manner and is now near the bottom of the list. In addition, various other social factors peculiar to India (cannabis use, consanguinity, migrations, climate, and geography) were explored as possible reasons for differences in the rates of insanity, but these causes were eventually discounted.

The better outcome of acute psychoses was well recognized by the early reports of 1881. The effects of environmental events (the famine of 1875-1885, the influenza epidemic of 1918-1919) were also considered. Apart from exploring etiological factors and clinical differences, the census data also provided a partial backdrop for planning delivery of services (number of asylum beds). The total number of ill is contrasted repeatedly with the number of beds available in the asylums and highlights the big gap. The urge for reform in medical services [13] and social care that began with the Legislative Assembly debates [14] , resulting in the Moore-Taylor Report to the Bhore Committee [12] , relied upon the same data sets.

The omission of this primary data about insanity, in particular, from 1941 onward, perhaps handicapped adequate planning for care of the mentally ill. The advent of drug treatments reduced the contribution and significance of previous administrative responses (largely asylum and hospital based) to provision of care. However, the provision of services for the mentally ill, and hesitant inclusion into the general health services, has been further handicapped by the lack of epidemiological data, the stigma of asylum-based services, and inadequate inclusion of psychiatric developments in medical training. In any case, despite the initial enthusiasm about drug treatments, a large proportion of the severely ill, it is gradually realized, recover only partially, and providing services for long-term care for them has become a challenge. The society, in general, and the family, in particular, have been expected to provide this (in sharp contrast to the earlier observations about the life of a lunatic in the community not being a happy one). Changes in social and family structures in a transforming society, as India is, may make it necessary to re-develop and extend institutional responses of varying kinds again. Recent debates about the nature of mental health services and whether they are better classified under health and medical services or as a general category under services for the disabled (infirm) reflect a return to the categories of the 19th century Imperial Administration.

The historical reading of the impressions about mental illness in data from previous census reports and contemporary attempts at interpretation reflects a concern with psychological, sociological, and biological issues that resonates right till the present. Successive reports tried to analyze these concerns and also document the increasing reach of services and the impact of this on the mentally ill and the society. It is interesting to see how, as we supposedly learn more about mental illness, and its impact, many debates remain unresolved, while the way ahead not uncommonly excites a feeling of déjà vu. Perhaps, the more things change, the more they remain the same.

   Acknowledgments Top

We would like to thank the Wellcome Institute of History of Medicine Library, London (SJ), and the Nehru Memorial Museum and Library (AS) for support.

   References Top

1.Waterfield H. Memorandum on the Census of British India 1871-72. London: Eyre and Spottiswoode; 1875.  Back to cited text no. 1
2.Census of England and Wales. 1871. Available from: [accessed in 2011].  Back to cited text no. 2
3.Editorial. Indian Med Gazette 1871;6:268.   Back to cited text no. 3
4.Plowden WC. Report on the Census of British India taken on the 17th February 1881. London: Eyre and Spottiswoode; 1883.  Back to cited text no. 4
5.The Census of India. Mysore: W 7869/24, British Library; 1881.  Back to cited text no. 5
6.Baines JA. General report on the Census of India, 1891. London: Her Majesty's Stationery Office; 1893. p. 236.  Back to cited text no. 6
7.Jagadisha T, Jain S. Better outcome of schizophrenia in India: A natural selection against severe forms? Schizophr Bull 2009;35:655-7.  Back to cited text no. 7
8.Gait EP. The Census of India. Vol. 1. Calcutta: Government Press; 1913.  Back to cited text no. 8
9.Marten JT. The Census of India, Vol. 1, Part 1. Calcutta: Government Press; 1921.  Back to cited text no. 9
10.Thyagarajaiyar VR. Superintendent of census operations. Mysore, Bangalore: Government Press; 1921.  Back to cited text no. 10
11.Hutton JW. The Census of India. Vol. 1. Delhi: Government Press; 1933. p. 253.  Back to cited text no. 11
12.Mills JH, Jain S. A disgrace to a civilized community: Colonial psychiatry and the visit of Edward Mapother to south Asia, 1937-1938. (In Psychiatry and Empire ed. Sloan Mahone and Megan Vaughan; (Cambridge Imperial and Post-Colonial Studies) 2007.   Back to cited text no. 12
13.Hill AV. Health, food and population in India. Int Rev 1945;21:40-50.   Back to cited text no. 13
14.Records of the Legislative Assembly. New Delhi: 20 February 1936.p. 1290-1.  Back to cited text no. 14

Correspondence Address:
Sanjeev Jain
Department of Psychiatry, Molecular Genetics Laboratory, National Institute of Mental Health and Neurosciences, Hosur Road, Bangalore 560029, Karnataka
Login to access the Email id

Source of Support: None, Conflict of Interest: None

Rights and PermissionsRights and Permissions


  [Figure 1]

This article has been cited by
1 ICD11 and DSM5: The Indian dilemma
Jain, S.
Asian Journal of Psychiatry. 2013; 6(3): 269-270