15 | India at the turn of the 20th Century - Findings from the 1901 Census of British India

The definition of a town given in the Imperial Census Code was as follows:

Town includes -

(1) Every municipality of whatever size.

(2) All civil lines not included within municipal limits.

(3) Every cantonment.

(4) Every other continuous collection of houses, permanently inhabited by not less than 5,000 persons, which the Provincial Superintendent may decide to treat as a town for census purposes.

The definition was practically the same as that adopted in 1891, but there have been some differences in its application owing to the latitude allowed to the Provincial Superintendents in respect of places with a population exceeding 5,000 persons but not under the operation of the law relating to municipalities.

In some provinces, e.g., in Burma, all such places have been treated as towns, while in others, e.g., in Madras, this has not been done unless they appeared to possess a truly urban character. In a few instances, in Bengal, places which are not municipalities and which have less than 5,000 inhabitants have been classed as towns for special reasons.

Comparison with Modern Census Definitions

The modern Census of India uses a more formalized dual definition:

Statutory Towns: Places with a municipality, corporation, cantonment board, or notified town area committee.

Census Towns: Places satisfying all three criteria:

1. Population ≥ 5,000

2. Density ≥ 400 persons per sq. km.

3. At least 75% of the male main workforce engaged in non-agricultural activities.

Cities

The 1901 Census adopted an international statistical convention: A city equals a settlement with at least 100,000 inhabitants. This threshold is much higher than the census definition of a town (5,000 population). There were only 29 cities in the whole of India, with an aggregate population of 6.6 million, constituting little more than 2 per cent of the total population of the British Empire.

In England nearly a third of the population is massed in cities, in Germany a sixth, and in France more than a seventh. Gait comments that the growth of cities is comparatively recent, and that, even in Europe, a hundred years ago the three countries (Britain, France and Germany)  had between them only 7 cities with a combined population less than a third as great as that of the Indian cities at the present time.

The 1901 Census reveals an India that was overwhelmingly rural. Out of a total population of 294.4 million, only 29.2 million people lived in towns, yielding an urbanization rate of just 9.9 percent. Urbanization was highly uneven across regions. British India as a whole was slightly less urbanized (9.6%) than the princely states and agencies (11.4%), challenging the common assumption that urban development was concentrated primarily in directly administered territories. 

Among the major provinces, Bombay Presidency stood out as the most urbanized (19.1%), followed by Berar (15.2%), Punjab (11.4%), Madras (11.2%), and the United Provinces (11.1%). In contrast, Bengal (5.2%) and Assam (3.0%) remained predominantly rural despite their large populations. Bengal alone contained over 74 million people but only 3.9 million urban residents. The United Provinces had the largest urban population in absolute terms (5.3 million), reflecting its enormous demographic size rather than a particularly high level of urbanization.

 The princely states exhibited even greater variation. Baroda State was the most urbanized major territory in India, with nearly a quarter of its population living in towns (24.0%). Bombay States (18.1%), Rajputana Agency (14.5%), and Mysore State (13.0%) also displayed relatively high urbanization, reflecting the importance of princely capitals, administrative centres, and commercial networks. At the other extreme, Bengal States (1.2%) and Central Provinces States (3.8%) had very limited urban development. The distribution of towns was equally uneven. The United Provinces contained 453 towns, the largest number in India, followed by Madras (234), Bombay Presidency (202), and Bengal (182). Together, these figures highlight that urban India in 1901 was not defined by widespread industrialization but by a limited network of ports, administrative capitals, cantonments, and market centres embedded within an overwhelmingly agrarian society.

The report highlighted contrasts of marriage customs between Europe and India, stating marriage as an almost universal religious obligation in India. Four categories were defined that set rules to govern marriage customs: Endogamy, Exogamy, Hypergamy and Prohibited Degrees (in which specific direct close relatives were to be avoided). The report discusses that customs such as infant marriage and prohibition of widow remarriage spread to lower castes through imitation.

The data revealed that one-third of the Indian female population was unmarried, seven-tenth of those were under the age of 10. However, very few remained single after the age of 20, unless they were suffering from severe afflictions. Widows represented 1 out of 6 in females, however widowers were almost double i.e. only 1 out of 18. This reflected the strict prohibition against widow remarriage present in many communities. By contrast, the proportion of widows in Europe was noted to be only 1 out of 13.

The report also noted that Hindus tended to have the highest rate of early marriage, followed by Muhammadans and Buddhists. Infant marriage was noted to be most frequent in the provinces of Berar and Bihar, while least in Burma, Mysore and Sind.

A growing sentiment against infant marriage was noted among educated classes and the reformist sects like Arya Samaj and Brahmo Samaj.

The 1901 Census revealed that migration within India was limited, with 90.7 percent of the population residing in the very district where they were born . This relative lack of mobility was largely attributed to the social restrictions, which made life uncomfortable away from one's social circle, and a deep-rooted economic dependence on local agriculture (Gait, 1913). 

Migration was officially categorized into five distinct types (Gait, 1913): 

(i) casual (minor village-to-village swaps),

(ii) temporary (travel for work or festivals),

(iii) periodic (seasonal harvest moves),

(iv) semi-permanent (city employment), and

(v) permanent (colonization of new territories)

One of the most significant streams involved over 500,000 people from the United Provinces and Bihar moving to Bengal to find work in jute mills, coal mines, and as field laborers . The tea industry in Assam also drew a massive labor force of roughly 750,000 immigrants, specifically seeking "jungly" coolies from the Chota Nagpur plateau who were better able to withstand the climate . 

Burma served as another major destination, attracting 475,328 immigrants from Madras and Bengal to support its rice harvest and oil industries. On the international stage, nearly 1.37 million Indians had emigrated to various colonies like Ceylon, Mauritius, and Natal, while only about 642,000 foreign-born individuals were counted in India . Within provincial boundaries, large-scale colonization occurred in the Punjab, where irrigation projects like the Chenab colony transformed barren wastes and drew over 600,000 settlers.

Table: Volume of Migration for Selected Provinces and States (1901)

This table summarizes the inter-provincial migration balance recorded in the 1901 Census. It highlights the major "exporting" regions like the United Provinces and Madras versus "receiving" areas like Assam and Burma (Risley & Gait, 1903; Gait, 1913). 

The plague first arrived in India in late 1896 through the port of Bombay. It was likely brought by rats on steamships arriving from China (Dyson, 2018). From the city's docks, the disease quickly spread across the country using the growing railway and road networks. By the time of the 1901 Census, the recorded death toll was already nearly half a million people. Many people fled the cities in panic, which ended up carrying the infection into rural areas (Dyson, 2018).

The impact of the plague varied greatly across different parts of the Empire (Dyson, 2018). The Punjab was hit the hardest, recording its highest death rate in 1907 with over 675,000 deaths in a single year. Bombay and the United Provinces also saw very high mortality, together making up three-quarters of all plague deaths in India. In contrast, the disease did not spread much in Madras or Bengal.

Maps showing the impact of Plague on population decline from 1891 to 1901. (Source: Census 1901, Part I)

Table: Reported Deaths from Plague in 1901.

Note: For the United Provinces and Punjab, figures represent combined male and female registered deaths (Gait, 1913).

Specifically, in the Bangalore Civil and Military Station (a key part of Mysore), the plague significantly damaged the local economy by hitting essential jobs very hard between 1891 and 1901. While a portion of the population drop was due to military changes, plague caused sharp declines in most civilian sectors. Industrial workers suffered the largest numerical loss, with their ranks shrinking by over 4,000 people (Row, 1973); this hit those involved in making food, providing fuel, building houses, and working with leather especially hard. Agricultural workers and merchants also saw their numbers fall by roughly 21% and 17% respectively. In contrast, a few sectors like personal and household services and those living on independent income managed to stay stable or even grow slightly during this crisis.

These frequent outbreaks acted as a serious detriment to industrial progress, as the constant threat of disease and high mortality made it difficult to open new factories or maintain a steady workforce. For many cities, this meant that the usual flow of new workers from rural areas was interrupted, as many people were deterred from moving to large centers like Calcutta where the disease was known to be present.

We delve deeper into the impacts of Plague on the Indian society at the time, in a separate article.

The 1901 Census recorded 147 distinct languages as vernacular within the Indian Empire. These forms of speech were classified into four primary families: Malayo-Polynesian, Indo-Chinese, Dravido-Munda, and Indo-European. The Indo-European family, particularly the Aryan sub-family, comprised the largest group with over 221 million speakers.

A massive administrative effort, the Linguistic Survey of India under Dr. Grierson, worked simultaneously with the census to explore and refine these classifications (which was otherwise complicated). This research demonstrated that unwritten languages often possess feeble vitality and are frequently displaced by more dominant neighbors (Gait, 1913). As a result, linguistic identity was often fluid, as tribes adopted the language of a superior civilization to improve their status.

Significant inaccuracies and discrepancies existed as millions of people returned the broad term "Hindi" without specifying distinct languages like Western or Eastern Hindi (Gait, 1913). In the Punjab, the 1901 returns for Panjabi were often inflated by the inclusion of Lahnda and other dialects that were not separately distinguished by uneducated enumerators (Gait, 1913).

To process such complex data, Commissioner Risley introduced the "slip system," which used specialized slips to simplify the tabulation of demographics like language and caste (Gait, 1913). This administrative innovation ensured greater uniformity across the fundamental tables used to classify the Empire’s diverse subjects (Srivastava, 1983). It replaced the older "abstraction sheets" where enumerators manually counted categories using ticks.

Census experts warned that language should not be used as an absolute proxy for race, as entire communities could change their linguistic identity within a few generations. In Burma, for instance, hill tribes frequently became Shan or Burmese speakers to facilitate travel and business (Gait, 1913).

Moreover, linguistic boundaries were acknowledged as conventional approximations, as dialects typically merge imperceptibly at their geographical borders. In Mysore, Kannada remained the dominant vernacular, spoken by 73 percent of the total population (Row, 1903). Ultimately, the 1901 report categorized these diverse Aryan tongues into "Inner" and "Outer" families based on their unique grammatical development and historical migration patterns.

This table presents the most widely spoken languages in India according to the 1901 Census, highlighting the numerical dominance of the Aryan and Dravidian families.

Literacy was defined as the ability to both read and write a language. In Mysore, only 4.8 percent of the population was literate, with 8.9 percent of males and 0.6 percent of females possessing this skill (Mysore Census Commissioner, 1901).

The 1901 Census simplified data by using only two groups: "literate" or "illiterate". Literacy was strictly defined as the ability to both read and write a letter in any language (Gait, 1913). People who could only sign their names or decipher familiar religious texts were considered illiterate. In the entire Indian Empire, only 53 people out of every 1,000 were literate. A massive gap existed between the sexes, with one in ten males being literate compared to just one in 144 females.

Education levels varied greatly based on a person’s religion. Zoroastrians and Jains were the most educated communities in India. High Hindu castes, such as Brahmans and Kayasths, maintained a historical lead in learning. However, many lower castes and Animistic tribes faced social barriers and poverty that prevented school attendance. In Mysore specifically, the total literacy rate was 4.8 percent, with 8.9 percent of males and 0.6 percent of females being able to read and write (Mysore Census Commissioner, 1901).

English literacy was extremely rare, claimed by only 68 males and 7 females per 10,000. This knowledge was found mostly in large cities like Bombay or among professional classes. High castes like the Baidyas of Bengal and the Prabhus of Bombay had the largest numbers of English speakers. Social rules often made it difficult to spread education to women or the lowest classes. Despite these challenges, census experts noted that female education was beginning to grow faster than male education during this decade (Gait, 1913).

This table summarizes literacy rates across major religious communities, showing that Zoroastrians and Jains were the most educated, while Animistic tribes and Muslims recorded the lowest rates. 

The census report tried to provide a detailed analysis of the religious landscape of the Indian empire in 1901. It categorized the population into several major and minor faiths and explored their underlying beliefs and social dynamics.

Of the total population, 70% were classified as Hindu, 21% as Muhammadans (Muslims), 3% Buddhists, 3% Animists and 1% Christians. Other remaining groups such as Jains, Sikhs, Parsis and others, made up nearly 1% of the remaining total population.

Table: The Population by Religion (1901 Census)

The report describes the religious landscape of early 19th Century India as defined by belief in magic and superstitions. It quotes Baines, the Census Commissioner of 1891 census, who described Hinduism as the “tangled jungle of disorderly superstitions”. The report highlighted the influence of caste over religious morality. It argued that the sanction attached to the breach of religious morality was the fear of caste penalties rather than fear of divine punishment.

The 1901 census attempted to count and analyze one of the most comprehensive and controversial racial and social compositions of India. It utilized physical measurements (anthropometry) and ethnographic observations to classify the Indian population and theorize the origins of the caste system.

The census identified a total of 2,378 main castes and tribes and 43 races or nationalities across the Indian Empire. The 1901 Census shifted from an occupational classification to one based on social precedence recognized by native public opinion.

Certain functional and social groups were widely distributed across the Empire. Brahmins were a widely diffused group, numbering nearly 15 million. They represented nearly 10% of the population in United Provinces and Rajputana, while dropping to 2-3% in Bengal and Madras. Chamars totaled over 11 million, forming 12% of the population in the United Provinces alone. Other widely distributed castes included Kumhars and Telis. Many caste groups remained localized (such as Marathas in Bombay province) or identified as Tribes, outside of the Hindu fold.

The report noted that caste distinctions had spread beyond Hinduism. Muhammadans were often divided into the Ashraf (noble) and Ajlaf (low/converts) classes, with the latter containing functional groups like the Jolaha (weavers), who numbered nearly 3 million across India. Even ancient Christian communities, such as the Syrian Christians, maintained distinct social markers.

A special article by the Censusnama team delves deeper into the caste and racial classification during 1901 done by H. Risely and its impact on later politics and scientific knowledge production.