Table 1: Number of Districts in Each Census Year

per their current bound-Arunachal Pradesh 1 4 4 5 4 2

Census 2001 State/UT Number of Districts 1971 1981 1991 2001aries. For each district it Assam 9 0 1 2 7 14

(1) (2) (3) (4) (5)

Bihar 7 240 24 180

also provides the popu-

Andaman and Nicobar Islands 1 2 2 2

Chandigarh 1 0 0 1 0 0

lation in the previous

Andhra Pradesh 21 23 23 23

Dadra and Nagar Haveli 1 0 0 1 0 0

census adjusted to the

Arunachal Pradesh 5 9 11 13

Daman and Diu 2 0 0

Assam 10 10 23 23 current boundaries. This

NCT Delhi 1 0 0 1 0 0 Bihar 17 31 42 37 would be the number

Goa 3 0 0 0 2 0 Chandigarh 1 1 1 1 obtained by netting out

Gujarat 19 0 0 19 0 0 Chhattisgarh 16 the population of the Haryana 2 1 15

1 9 0 Dadra and Nagar Haveli 1 1 1 1

areas transferred in Himachal Pradesh 5 3 4 10 0 2

Daman and Diu 2 2

and out of the district. Jammu and Kashmir 3 4 7 14 0 0

Delhi 1 1 1 9

In a large number of Karnataka 19 0 0 16 2 2

Goa 3 3 2 2

Kerala 4 0 8 6 2 6

cases, when a district

Gujarat 19 19 19 25 Lakshadweep 1 0 0 1 0 0

Haryana 7 12 16 19 is partitioned or only Madhya Pradesh 40 2 3 45 0 0

Himachal Pradesh 10 12 12 12 affected by one area Maharashtra 26 0 0 16 0 14

Jammu and Kashmir 10 14 14 14 transfer, this table is suf-Manipur 2 0 4 3 0 5

Jharkhand 18

ficient to deduce the

Meghalaya 0 50 5 00Karnataka 19 19 20 27

population of individual

Mizoram 0 3 0 3 0 0

Kerala 10 12 14 14

regions transferred.

Nagaland 0 70 1 06Lakshadweep 1 1 1 1

Our third source is

Orissa 13 0 0 13 0 0

Madhya Pradesh 43 45 45 45 the footnotes to the Pondicherry 4 0 0 4 0 0

Maharashtra 26 26 30 35 Manipur 5 6 8 9 Appendix to Table A-2. Punjab 4 0 8 7 0 5 Meghalaya 2 5 5 7 These footnotes detail Rajasthan 18 0 8 21 2 4

Sikkim 2 0 2 4 0 0

Mizoram 3 3 8 individual transfers bet-Nagaland 3 7 78 TamilNadu 9 2 5 12 9 0

ween affected districts,

Orissa 13 13 1330 Tripura 1 0 2 1 0 2

together with the

Pondicherry 4 4 4 4 Uttar Pradesh 49 2 5 41 8 14

number and year of the

Punjab 11 12 12 17 West Bengal 14 0 2 14 0 3

government notification

Rajasthan 26 26 27 32 India 271 62 79 313 54 99

that effected the trans-(1) In columns (2), (3) and (4), districts are listed as per 1981 jurisdictions, while in columns (5), (6) and (7), districts are listed as per 1991 jurisdictions.

Sikkima 4 4 4 4

fer. A typical entry from

Tamil Nadu 14 16 21 30 (2) The summary of intercensal boundary changes between 1991 and 2001 can be obtained from columns (8), (9) and (10) of Table 3.

Tripura 3 3 34 a footnote on page 142 Uttar Pradesh 54 56 63 70 of the Census 2001 national volume (Gov-figures reported by the census volumes Uttaranchal

13 ernment of India 2005) reads as follows: without attempting any adjustments.3 West Bengal 16 16 17 18

We refer to a district B as a “child” of

Fifty-eight villages with an area of 147.29 sq

Statesa 19 22 25 29

another district A, if any inhabited areas

km and population of 57,365 were transferred

Union Territories 10 9 7 6

to Khamanon tahsil of newly formed Fateh-have been transferred from A to B during

Districtsa 356 412 466 593

garh Sahib district vide Punjab Govt Notifica-the period under consideration. District A

a While Sikkim was not a part of the Union of India in 1971, later Indian Censuses allow us to trace boundary changes in tion No 2/3/92-RE.II(i)/4412 dated 9.4.92.

is referred to as a “parent” of B. Unchanged

Sikkim back till 1971. In this and all following tables, we use this information; however, the totals for states and districts in India Occasionally, the above sources from districts are single parent and single child do not include Sikkim in 1971.

(1) If a Union Territory is not further subdivided in any given the national volumes are insufficient to districts and partitioned districts share a census year, we count it as a single district in that year.

figure out individual boundary changes.1 single parent. For these two classes of

volumes.2 There are also some small areas corresponding to 1971 because populations

in 1987. In the same year, the UTs of Mizoram and Arunachal

close to state borders whose jurisdiction is for boundaries corresponding to the parent

Pradesh were also accorded statehood. Sikkim was added to the Union of India in 1975. Mizoram was created as a UT from disputed. There is no systematic approach district are simply the sum of child popula-Mizo district of Assam in 1972. In the same year, the UTs of Manipur and Tripura also acquired statehood. to classifying these areas. They are some-tions. To illustrate, the district of Muzaf

(4) Following Census documents, we count Delhi as a UT in

times included in the population figures farpur in Bihar in 1971 is partitioned into

1971, 1981 and 1991, but as a state in 2001. The Autonomous State of Meghalaya was created on 2 April 1970, but it acquired for both states and sometimes excluded the three districts Muzaffarpur, Sitamarhi

full statehood only in 1972. Census 1971 documents list it as a state, and we follow this convention. completely. We simply use the population and Vaishali in 1972 and Sitamarhi is

70 october 10, 2009 vol xliv No 41

further partitioned in 1994 to form Sheohar district. To obtain a time series on (say) the literate population of Muzaffarpur as per its 1971 boundaries, one would simply aggregate the literate population in the child districts for subsequent census years. It is also possible to create a panel based on district boundaries of later census years if we make certain assumptions about the distribution of the population. In the above example, an estimate of the number of literates in Sheohar in 1971 can be obtained by multiplying the data for Muzaffarpur in that year, by the product of two terms. The first is the fraction of the 1971 Muzaffarpur population that was in Sitamarhi by 1981, and the second is the fraction of the Sitamarhi population in 1991 which fell within

Table 3: Summary of Boundary Changes

the boundaries of Sheohar. The greater the uniformity in the distribution of the literate population across the original district, the more accurate would be a time series generated in this manner.

When a district has more than one parent, we cannot use the above procedure to obtain district panels spanning multiple census years. For example, a large part of present-day Champawat district was formed from Champawat tahsil of Pithoragarh district in 1997 (with the remaining part coming from Nainital district). Champawat tahsil was part of Almora district until Census 1971, and was transferred to Pithoragarh district only during the 1971-81 period. No part of the area within Pithoragarh’s boundaries in 1971 was part

This table lists the number of Census 2001 districts that had remained unchanged from, or were clean partitions of, districts covered in previous census years. The residual category is "Other".

of Champawat in 2001. Yet, if we mechanically and successively applied weights based on population shares over the census years, we would attribute 59% of the population of present-day Champawat as coming from Pithoragarh in 1971.4

In the above example, an entire tahsil was transferred across districts. However a large number of boundary changes involve a portion of a tahsil or taluk being transferred to a different district. In such cases, the tables tell us the number of villages that were transferred, but do not identify the villages themselves. This makes it impossible to track the transferred region across multiple census years. For these reasons, we only provide weights across adjacent census years for districts with multiple parents. To create longer panels, albeit with some error, our tables must be combined with a visual examina-

EPW

Table 4: Districts with Boundaries Unchanged between 1971 and 2001 period. Our contribution lies in using these

State/UT District

figures to construct a comprehensive

Andhra Pradesh East Godavari, Guntur, Karimnagar, Khammam, Krishna, Kurnool, Mahbubnagar,

classification of district-level boundary

Medak, Nalgonda, Nizamabad, Prakasam (Ongole), Warangal, West Godavari

changes for the 1971-2001 period.

Assam Karbi Anglong (Mikir Hills), North Cachar Hills

Chandigarh Chandigarh

3 Summary of Findings6

Dadra and Nagar Haveli Dadra and Nagar Haveli

Table 1 (p 70) lists the number of districts

Daman and Diu Daman, Diu

Gujarat Jamnagar, Kachchh, Rajkot, Sabar Kanta, Surat, Surendranagar, The Dangs in each state for each census year. The

Himachal Pradesh Bilaspur, Kinnaur, Kullu, Mandi total number of Indian districts went up

by 15.7% and 13%, respectively, in the first two decades and then by 27.3% during the 1991-2001 decade. New districts were created at very uneven rates across the Indian states. No new districts were created in Gujarat and Orissa until 1991 and only one or two new districts appeared in several other states (Himachal Pradesh, Punjab, Rajasthan, Madhya Pradesh, West Bengal, Andhra Pradesh and Karnataka). Most boundary changes during the 1971-91 period were concentrated in Tamil Nadu, Bihar, the

a One village with area 1 sq km and unknown population was transferred from Tehri Garhwal district to Dehradun district between 1971 and 1981.

the availability of public goods. Most recently, Bharadwaj et al (2008) map districts of British India in 1931 with corresponding areas in Pakistan and India in 1951 to study labour migration induced by the Partition of India.

Maps for successive census years are available in the Indian Administrative Atlas 1872-2001 (Government of India 2004) published as part of the documentation of Census 2001. Map No 15 in this publication provides a list of newly created districts in each decade between 1951 and 2001. Appendix I tracks changes in the spellings of district names in each decade, while Appendix III provides an “Equivalence Table of Administrative Divisions 1961-2001”.5

While these administrative maps are extremely useful in pointing out the principal antecedents of districts over long periods of time, visual comparisons of district boundaries based on these maps is necessarily coarse. Differences in population density are ignored and districts which either lose or receive small areas may appear unchanged. To illustrate, Murthi et al (2001) use administrative maps to classify 368 districts as being unchanged between 1981 and 1991, and another 42 districts as partitioned into two or three 1991 districts each (see their Table 2). Our own analysis reveals that there were only 313 districts that were unchanged between 1981 and 1991, and another 23 that were partitioned. Many of the differences in classification arise because small transfers are neglected when maps are used. If for example, we ignore transfers of up to 5% of the original district’s population, we get 358 unchanged districts and 35 districts being partitioned bet ween 1981 and 1991. These errors of misclassification are likely to get compounded in studies that cover longer time periods.

We are aware of one other study that relies on the population table we have used in this paper. Reddy (1988) in his historical study of changes in agricultural labour uses them to isolate 98 regions with unchanged boundaries over the 1881-1981

october 10, 2009

the number of districts. The only exceptions were the states of Assam, Himachal Pradesh, Andhra Pradesh and Kerala, where the number of districts remained unchanged between 1991 and 2001.

Table 2 (p 70) and Table 3 (p 71) show how districts can be placed in our threefold classification – those with unchanged boundaries, clean partitions of existing districts and a residual category of all remaining districts. Some states such as Bihar are good candidates for inclusion in multi-year panels because, even though the number of districts increased significantly, new districts were created by partitioning existing ones. In contrast, Haryana and Gujarat experienced a complex reorganisation of district boundaries. This makes it much harder to create long demographic panels for these states.

Table 4 lists the names of all districts whose boundaries remained unchanged throughout the 1971-2001 period. Table 5 lists all partitioned districts, and for each new district it provides the population share it received of the parent district. Table 6 contains districts that are neither unchanged nor partitioned. As argued in Section 2, for these districts it is only possible to define meaningful population weights for adjacent census years. Tables 6a-6c provide these weights.

vol xliv No 41

As discussed in Section 2, it is possible to create a 1971-2001 district panel for the entire country as long as we are willing to amalgamate the districts in Tables 6a-6c into composite regions with unchanged boundaries. Table 7 tells us the size of the panels that can be generated in this manner. For the whole country, we can define 267 such regions for the 1971-2001 period, 334 for 1981-2001 and 414 for 1991-2001.

Table 8 lists the composite regions individually. In some cases the complexity of underlying transfers means that such regions can be quite large. To cite the two largest composite regions by population for the period 1971-2001: the entire state of Punjab except Kapurthala district forms one composite region, while the entire state of Haryana forms part of one composite region that also includes two neighbouring districts (Bulandshahr and Meerut) of the state of Uttar Pradesh. In general, there are very few such regions that cross state boundaries.

4 Conclusions

The purpose of this article is to provide data on the composition of all Indian districts over the 1971-2001 period that can enable the construction of district panels. We conclude with a few caveats on using our data.

First, we have restricted ourselves to an examination of population shares and have neglected changes in district areas. The resulting classification is a natural one for the study of demographic questions but may not be the most appropriate for studies of land-use or climate, where weights based on areas transferred may seem more appropriate. Our main reason for not supplementing population changes with a study of areas transferred is the difficulty in obtaining a consistent time series of district areas. Different administrative authorities are responsible for the compilation of district and sub-district areas; the office of the surveyor general provides district areas while areas of jurisdictional boundaries below the district level (blocks, tahsils, villages) are based on numbers provided by land revenue authorities and municipal bodies (Government of India 2005: 25). These two sets of figures are often inconsistent.

Second, we have limited ourselves to changes in boundaries across census years. Much of the research on India uses

Economic & Political Weekly

EPW

household survey data from the National Sample Survey, the National Family Health Survey and other sources. The years for which these data are collected do not match census years. To create a panel based on these data sets, researchers can use the footnotes to the Appendix to Table A-2 which we have mentioned in Section 2 because these notes contain the year in which each transfer was made.7

A final caveat relates to the creation of multi-year district panels. The information provided here only allows for these to be created for unchanged districts, single parent districts, and the amalgamated regions described in Section 2 and listed in Table 8. For districts with multiple parents, it is not in general possible to use data on transfers in successive years to generate population weights over longer periods, as discussed in Section 2. If the data presented here are used for this purpose, they should certainly be combined with an examination of maps from the census publications to minimise errors. Ultimately, the only way to create error-free panels for all districts is to track the movement of villages across district boundaries. This is possible but time-consuming for the years before 1991 when census village data was not available in digital form. We leave this for future work.

Notes

1 This is the case for changes in Himachal Pradesh during 1971-81, Haryana during 1981-91, Assam during 1971-91, and Manipur during 1981-91.

2 In one case, Haryana, both the national and state volumes are inadequate in identifying boundary changes during 1981-91. Several transfers during this period in Haryana involved small numbers of villages and neither central nor state volumes list the population of transferred villages. As noted at the bottom of Table 6b, our figures for selected districts in Haryana over this period are, as a result, approximate.

3 For example, due to the changing course of the river Ganga, the status of 18 villages has been under dispute between the states of West Bengal and Jharkhand. The populations of these villages for 2001 have been included in both Maldah district of West Bengal and Sahibganj district of Jharkhand (Government of India 2005: 24). In contrast, an area of 13 sq km which has been long disputed by Pondicherry and Andhra Pradesh has been excluded from the population of both regions (ibid 169).

4 From Tables 6a-6c, the product of the population weights is computed as 75.57%×100%×78.44% = 59.28%

5 More detailed maps are also available from three other sources: (i) the state-wise Administrative Atlas series published for the 2001 Census; (ii) the state-wise Census Atlas series published with every Census since 1961; and (iii) the District Census Handbooks, of which there is one volume per district in every census year.

6 As mentioned, Tables 5 to 8 are posted on the EPW web site alongside this paper.

7 However, these notes are sometimes incomplete. Aside from the previously mentioned case of Haryana during 1981-91, information on the year of the transfer is missing in the tables for Uttar Pradesh and Arunachal Pradesh during 1971-81 and for West Bengal during 1981-91.

References

Banerjee, Abhijit and Lakshmi Iyer (2005): “History, Institutions, and Economic Performance: The Legacy of Colonial Land Tenure Systems in India”, The American Economic Review, 95(4): 1190-1213.

Banerjee, Abhijit and Rohini Somanathan (2007): “The Political Economy of Public Goods: Some Evidence from India”, Journal of Development Economics, 82(2): 287-314.

Bharadwaj, Prashant, Asim Khwaja and Atif Mian (2008): “The Big March: Migratory Flows after the Partition of India”, Economic & Political Weekly, 43(35): 39-49.

Government of India (1996): General Population Tables and Primary Census Abstract, Census of India 1991: Series 8: Haryana (Chandigarh: Director of Census Operations, Haryana).

Murthi, Mamta, P V Srinivasan and S V Subramanian (2001): “Linking Indian Census with National Sample Survey”, Economic & Political Weekly, 36(9): 783-92.

Reddy, Rammanohar (1988): “The Incidence of Agricultural Labourers in India: Variations over Time and Space, 1881-1981”, PhD thesis, Jawaharlal Nehru University.

Table 6c: Boundary Changes in ‘Other’ Districts (1991-2001) Table 6c: (Continued)

Census 1991 1991 District Population 2001 District Share of 2001 Share of 1991 Census 1991 1991 District Population 2001 District Share of 2001 Share of 1991 State/UT in 1991 District in District in State/UT in 1991 District in District in 1991 District 2001 District 1991 District 2001 District

(1) (2) (3) (4) (5) (6) (1) (2) (3) (4) (5) (6)

Andhra Pradesh Adilabad 2,082,479 Adilabad 100.00 100.00Bihar Begusarai 1,814,773 Begusarai 100.00 100.00 Anantapur 3,183,814 Anantapur 99.91 100.00 Dhanbad 2,674,651 Bokaro 27.11 49.86 Cuddapah 0.09 0.13Dhanbad 72.89 100.00 Chittoor 3,261,118 Chittoor 100.00 100.00 Giridih 2,225,480 Bokaro 32.77 50.14 Cuddapah 2,267,769 Cuddapah 100.00 99.87Giridih 67.23 100.00 Hyderabad 3,145,939 Hyderabad 100.00 100.00 Hazaribag 2,843,544 Chatra 21.55 100.00 Nellore 2,392,260 Nellore 100.00 100.00Hazaribagh 64.57 100.00 Rangareddi 2,551,966 Rangareddi 100.00 100.00Kodarma 13.88 100.00 Srikakulam 2,321,126 Srikakulam 100.00 100.00 Khagaria 987,227 Khagaria 100.00 100.00 Visakhapatanam 3,285,092 Visakhapatnam 100.00 100.00 Munger 3,060,027 Jamui 34.36 100.00 Vizianagaram 2,110,943 Vizianagaram 100.00 100.00 Lakhisarai 21.12 100.00 Arunachal Pradesh East Siang 99,643 East Siang 72.12 100.00 Munger 30.84 100.00 Upper Siang 27.88 100.00 Sheikhpura 13.68 99.59Lower Subansiri 155,978 Lower Subansiri 53.32 100.00 Nalanda 1,997,995 Nalanda 99.91 100.00 Papum Pare 46.68 100.00 Sheikhpura 0.09 0.41Upper Subansiri 50,086 Upper Subansiri 100.00 100.00 Patna 3,618,211 Patna 100.00 100.00 West Siang 89,936 West Siang 100.00 100.00 Gujarat Ahmadabad 4,801,812 Ahmadabad 95.54 100.00 Assam Barpeta 1,385,659 Barpeta 100.00 100.00Gandhinagar 4.46 19.89 Bongaigaon 807,523 Bongaigaon 100.00 100.00 Amreli 1,252,589 Amreli 86.76 83.03 Darrang 1,298,860 Darrang 100.00 100.00Junagadh 13.24 7.93 Dhubri 1,332,475 Dhubri 99.39 100.00 Banas Kantha 2,162,578 Banas Kantha 91.63 100.00 Kokrajhar 0.61 1.00Patan 8.37 17.48 Goalpara 668,138 Goalpara 100.00 100.00 Bharuch 1,546,145 Bharuch 74.27 100.00 Golaghat 828,096 Golaghat 100.00 100.00Narmada 25.73 88.54 Jorhat 871,206 Jorhat 100.00 100.00 Bhavnagar 2,292,026 Amreli 9.69 16.97 Kamrup 2,000,071 Kamrup 100.00 100.00Bhavnagar 90.31 100.00 Kokrajhar 800,659 Kokrajhar 100.00 99.00 Gandhinagar 408,992 Gandhinagar 100.00 37.96 Marigaon 639,682 Marigaon 100.00 100.00 Junagadh 2,394,859 Junagadh 80.40 92.07 Nagaon 1,893,171 Nagaon 100.00 100.00Porbandar 19.60 100.00 Nalbari 1,016,390 Nalbari 100.00 100.00 Kheda 3,440,897 Anand 47.74 100.00 Sibsagar 907,983 Sibsagar 100.00 100.00Gandhinagar 0.33 1.07 Sonitpur 1,424,287 Sonitpur 100.00 100.00 Kheda 51.93 100.00

Continued

october 10, 2009 vol xliv No 41

Economic & Political Weekly

october 10, 2009 vol xliv No 41

Table 6c: (Continued) Table 7: The Number of Consistent Regions between Each Census Year and 2001

Census 1991 1991 District Population 2001 District Share of 2001 Share of 1991 For each state, this table counts the number of regions with unchanged boundaries that can be used

State/UT in 1991 District in District in for panels between 2001 and each census year 1971, 1981 and 1991. This includes districts that are 1991 District 2001 District unchanged or partitioned, as well as the composite regions listed in Table 8

(1) (2) (3) (4) (5) (6)

Census 2001 State/UT Number of Regions 1971 1981 1991

Hardwar 1,124,488 Hardwar 100.00 100.00

(1) (2) (3) (4)

Kanpur Dehat 2,138,317 Kanpur Dehat 60.95 100.00

Andaman and Nicobar Islands 1 2 2

Kanpur Nagar 39.05 25.67

Andhra Pradesha 18 21 22

Kanpur Nagar 2,418,487 Kanpur Nagar 100.00 74.33

Arunachal Pradesh 4 8 11

Mainpuri 1,316,746 Etawah 0.43 0.51

Assam 6 6 22 Mainpuri 99.57 100.00

Bihar 10 23 28 Mathura 1,931,186 Hathras 14.53 24.89

Chandigarh 1 1 1 Mathura 85.47 100.00

Chhattisgarh 5 6 6 Mau 1,445,782 Mau 100.00 100.00

Dadra and Nagar Haveli 1 1 1 Meerut 3,447,912 Baghpat 29.88 100.00

Daman and Diu 2 2 2 Meerut 70.12 100.00

Delhi 1 1 1 Muzaffarnagar 2,842,543 Muzaffarnagar 100.00 100.00

Goa 1 1 2 Nainital 1,540,174 Champawat 2.67 21.56 Gujarat 12 12 12

Nainital 37.28 99.88 Haryanab 1 3 6

Udham Singh Nagar 60.05 100.00Himachal Pradesh 7 11 12

Pithoragarh 566,408 Champawat 26.44 78.44 Jammu and Kashmir 7 14 14

Pithoragarh 73.56 100.00Jharkhand 5 6 12

Rae Bareli 2,322,810 Fatehpur 0.01 0.01 Karnataka 16 16 18

Rae Bareli 99.45 100.00 Kerala 4 8 13

Sultanpur 0.55 0.50Lakshadweep 1 1 1

Madhya Pradesh 35 37 37 Maharashtraa 20 20 30

Saharanpur 2,309,029 Saharanpur 100.00 100.00

Siddharth Nagar 1,707,885 Sant Kabir Nagar 5.85 8.67 Manipur 2 4 8

Siddharthnagar 94.15 100.00Meghalaya 2 4 4

Sultanpur 2,558,970 Sultanpur 100.00 99.50Mizoram 1 3 3

Tehri Garhwal 580,153 Rudraprayag 10.32 29.87 Nagaland 2 4 7

Tehri Garhwal 89.68 100.00 Orissa 13 13 13

West Bengal Calcutta 4,399,819 Kolkata 100.00 100.00Pondicherry 4 4 4

Murshidabad 4,740,149 Murshidabad 100.00 100.00

Punjab 2 33

Nadia 3,852,097 Nadia 100.00 100.00

Rajasthan 20 23 26

North 24 Paraganas 7,281,881 North Twenty

Sikkim 3 4 4

Four Parganas 100.00 100.00 Tamil Nadu 12 16 20

South 24 Paraganas 5,715,030 South Twenty Tripura 1 2 2

Four Parganas 100.00 100.00 Uttar Pradeshb,c 35 37 45

* Census operations were not held in Jammu and Kashmir in 1991 due to political unrest.The population figures reported for this state in 1991 are interpolated numbers reported in Census 2001 documents. Uttaranchalc 4 5 5

(a) an uninhabited portion of one village, with an area of 0.07 sq km, was transferred from Alappuzha

West Bengal 14 15 17 district to Pathanamthitta district in 1996.

India 267 334 414

(b) uninhabited portions of two villages, with a total area of 0.33 sq km, were transferred from (a) In 1971 and 1981, one composite region cuts across the state boundaries of Andhra Pradesh Ernakulam district to Idukki district in 1997.

and Maharashtra. (b) In 1971 and 1981, one composite region cuts across the state boundaries (c) an uninhabited portion of one village, with an area of 0.01 sq km, was transferred from Kollam of Haryana and Uttar Pradesh. (c) In 1971 and 1981, one composite region cuts across the state district to Pathanamthitta district in 1992.

boundaries of Uttar Pradesh and Uttaranchal.

(d) an uninhabited portion of one village, with an area of 5.03 sq km, was transferred from (1) When a region cuts across state boundaries, it is included in the count of regions for each of Pathanamthitta district to Kottayam district in 1995. those states. However, the totals for India do not double-count such regions.

Economic & Political Weekly