India is currently divided into 36 administrative states and union territories, which are further divided into districts and more granular units. Districts in India are the third geographic layer for data dissemination after national and state-level setups. Many agencies — such as the Census of India, Reserve Bank of India, Ministry of Health and Family Welfare, Directorate of Economics and Statistics, etc. — collect and compile data at the district level.
Though collecting data at the district level is common, comparing different data sets or data from different years at the district level can be difficult. There are three reasons for this problem:
- The number of districts reported by different agencies differ due to data reference date.
- Sometimes new districts are formed from multiple districts.
- In many cases, even after a district is split, the old name continues to be used for the parent district though it is no longer the same old district.
In the absence of more granular data, comparing district data across time and space poses practical difficulties.
Evolution of New Districts in India
The Census is our benchmark for detailed district level data. However, just 4 years after the last Census, its district boundaries are out of date.
As of December 2015, India has 684 districts and another 31 districts are expected in 2016. Around one-third (218 districts) of the total 684 districts were created between 1991 and 2015.
Normally, the Census of India is used as the reference source for number of districts and district composition, since the Census collects detailed demography data at sub-district, village, and ward level. However, the Census is only conducted once in every 10 years, and district boundaries change much more frequently. The last census was conducted in 2011 and it collected data based on 640 districts (as of December 2010). Since then, 44 new districts have been created.
Some agencies report an updated list of districts as and when new districts are created. To map data from other sources against Census districts, the data for new districts has to be adjusted based on this updated district information. Hence, it is essential for every data analyst working on time series district-level data to have a repository of information on district changes so that data comparison can be credible.
At SocialCops, data sanctity is at the core of our processes, so we monitor the 593 districts from Census 2001 for updates and the formation of new districts. Out of 593 districts of Census 2001, 88 districts have changed their geographical boundaries as of December 2015, and some districts have been split into multiple districts.
Similarly, out of 640 districts from Census 2011, 50 districts have changed their boundaries. Sometimes districts are split repeatedly. For example, Kurung Kumey district in Arunachal Pradesh was created from Lower Subansiri district in April 2001. But, in February 2015, another district (Kra Daadi) was created from Kurung Kumey. Similarly, Udalgiri district in Assam was formed in October 2003 from 808 villages of Darrang district and 19 villages of Sonitpur district. Then, on 15th August 2015, Sonitpur district was again divided to form Biswanath district.
4 Possible Reasons for Splitting Districts in India
There are many reasons for splitting a district. Generally, new districts are created for ease of administration so that the distance between the district headquarters and remote areas are shortened. This, in turn, helps with better monitoring of government schemes and maintaining law and order in remote areas. In addition, districts can be split based on their population size or various socio-cultural factors. We looked at the following four aspects:
Average Population per District
In India, neither average population nor population density are even across districts. In 2011, while the average population per district was 1.9 million, the most populated district had a population of 11.1 million (Thane in Maharashtra) and the least populated district had a population of only 8,004 (Dibang Valley in Arunachal Pradesh). Consequently, Thane district has since been split into two districts (Thane and Palghar).
In 2011, a little over 50 percent of India’s population was concentrated in 148 of the total 640 districts. Interestingly, 31 of the new districts expected in 2016 are in three states with high average population per district — West Bengal, Andhra Pradesh, and Telangana.
Population density is highly variable across districts in India. On average, 369 persons lived per square kilometer in 2011 in India. However, this figure varied from as low as one person per square kilometer (Dibang Valley in Arunachal Pradesh) to as large as 37,346 people per square kilometer (North-East district in Delhi).
It seems that population density has a limited role in splitting a district. Average population density for the 50 split districts of Census 2011 was 393, compared to 367 in non-split districts. However, if we take out the districts of Delhi that were split, the average for the remaining split districts is 346 persons per square kilometer versus 371 in non-split districts.
It is to be noted that two new districts were created in Delhi in 2012, but the boundaries of 8 of the 9 existing Delhi districts were also changed. As of the 2011 Census, 2 districts from Delhi (Central Delhi and New Delhi) were purely urban. After the boundary changes in 2012, these two districts now include a rural component as well!
Urbanization appears to be one factor in splitting districts. The proportion of urban population was a bit higher in the 42 split districts (38.1%) than the non-split districts (30.5%). If we include the Delhi districts, urban population rises. For the 50 split districts (including districts of Delhi), the pre-split urban population was 45.4% versus 29.6% for non-split districts.
Distance to District Headquarters
We looked at distance of villages from district headquarters for 631 districts that did not have an urban component in 2011. Distance indeed seems to have played a role in carving out new districts between 2011 and 2015. Overall, 39.1% of villages and 33.2% of population were 50 kilometers or more from the district headquarters across all districts. However, in the 48 non-urban districts that were later split, the corresponding figures were 50.7% and 42.0%. This means that, if an area is farther away from the district headquarters, it is more likely to be split into a new district.
The 44 new districts created between 2011 and 2015 are in 12 states only. Some of these states have high average population per district, while some have low population per district.
|Table 01: New Districts in India since 2011 (as of December 2015)|
District Name, Census 2011
New Districts created after Census 2011 (after December 2010)
|Arunachal Pradesh||East Siang||Siang|
|Arunachal Pradesh||Kurung Kumey||Kra Daadi|
|Assam||Dhubri||South Salmara – Mankachar|
|Assam||Karbi Anglong||West Karbi Anglong|
|Chhattisgarh||Surguja||Surajpur & Balrampur|
|Chhattisgarh||Durg||Bemetara & Balod|
|Chhattisgarh||Raipur||Balodabazar & Gariyaband|
|Delhi||District New Delhi|
|Gujarat||Ahmadabad||Botad district was created from Ahmedabad and Bhavanagar districts|
|Gujarat||Surendranagar||Morbi was created from Rajkot, Surendra Nagar & Jamnagar|
|Gujarat||Rajkot||Morbi was created from Rajkot, Surendra Nagar & Jamnagar|
|Gujarat||Jamnagar||Devbhoomi Dwarka & Morbi were created|
|Gujarat||Bhavnagar||Botad district was created from Ahmedabad and Bhavanagar districts|
|Gujarat||Kheda||Mahisagar was created from Kheda & Panch Mahal|
|Gujarat||Panch Mahals||Mahisagar was created from Kheda & Panch Mahal|
|Madya Pradesh||Sheopur||Agar Malwa|
|Meghalaya||West Garo Hills||South West Garo hills|
|Meghalaya||East Garo Hills||North Garo hills|
|Meghalaya||West Khasi Hills||South West Khasi hills|
|Meghalaya||Jaintia Hills||East Jaintia hills & West Jaintia hills|
|Tripura||West Tripura||Khowai & Sepahijala|
|Uttar Pradesh||Muzaffarnagar||Prabudh Nagar|
|Uttar Pradesh||Moradabad||Bhim Nagar|
|Uttar Pradesh||Ghaziabad||Panchsheel Nagar|
|Uttar Pradesh||Rae Bareli||Amethi|
Duplicate Names and Name Changes
While comparing district-level data between two sources across states, looking at district names alone can lead to errors. This happens when there are duplicate district names. There are 6 duplicate district names in different states in India that one should consider while mapping district data:
- Aurangabad district in Bihar and Maharashtra
- Bilaspur district in Chhattisgarh and Himachal Pradesh
- Bijapur district in Chhattisgarh and Karnataka
- Hamirpur district in Himachal Pradesh and Uttar Pradesh
- Pratapgarh district in Rajasthan and Uttar Pradesh
- Balrampur district in Chhatisgarh and Uttar Pradesh.
Similarly, there are many examples of district names changing for one reason or another. Sometimes the name changes for political reasons, while other times it happens because of a popular demand or a district’s history. For example, Gauriganj district in Uttar Pradesh had two earlier names: Chhatrapati Sahuji Maharaja Nagar and Amethi.
|Table 02: Variation in District Names|
|Andhra Pradesh||YSR||Kadapa, Cuddaph, Y S Reddy|
|Andhra Pradesh||Nellore||Shri Potti Sriramulu Nellore|
|Assam||North Cachar Hills||Dima Hasao|
|Bihar||East Champaran||Purbi Champaran|
|Bihar||West Champaran||Paschimi Champaran|
|Madhya Pradesh||East Nimar||Khandwa|
|Madhya Pradesh||West Nimar||Khargoan|
|Punjab||Sahibzada Ajit Singh Nagar||Mohali, SAS Nagar|
|Punjab||Shahid Bhagat Singh Nagar||Nawanshahr|
|Uttar Pradesh||Amroha||Jyotiba Phule Nagar|
|Uttar Pradesh||Hathras||Mahamaya Nagar|
|Uttar Pradesh||Rambai Nagar||Kanpur Dehat|
|Uttar Pradesh||Kasganj||Kanshiram Nagar|
|Uttar Pradesh||Bahjoi||Bhim Nagar|
|Uttar Pradesh||Gauriganj||Chhatrapati Shahuji Maharaja Nagar, Amethi|
We hope this blog will help reduce duplication of efforts in compiling district names. If you come across any useful information related to this blog, kindly let us know by leaving a comment or emailing us at email@example.com. For a more historical list of change in boundaries across countries, check here.
Collecting data in the field is always easier with a mobile app. We’ve helped over 150 partners collect over 20 million data points through our mobile data collection app Collect!