How Has Kashmir’s Religious Map Changed Since 1947?

The Kashmir conflict and Partition are often portrayed as being different stories, but in fact, the Kashmir conflict is partly a result of Partition. If it weren’t for the sloppy and illogical manner in which the British executed Partition, the Kashmir conflict would never have occurred. Unlike Punjab and Bengal, Kashmir was not technically part of the Raj. It was a princely state, actually called the Princely State of Jammu and Kashmir. Princely states, 565 of them, were a British creation that existed in a weird limbo between independence and colonial occupation. They were theoretically independent and allied with the British, but the British held all the power in them, and the princes were basically figureheads. While the British remained in India, the princely state vs British India split was largely a distinction without a difference.

When independence came, the princely states presented a problem, because technically they retained their independence. The British policy, after some confusion, was to give the princes the option of acceding to either Pakistan or India, regardless of the religion of their subjects or the geographical location of their territory. It would have been more logical for the British to have given each princely state to the religiously appropriate country, and partitioned Kashmir, the one large state with both Hindu and Muslim majority areas, along religious lines. While this policy would have infringed upon the rather dubious sovereignty of the princely states, the princes didn’t have the clout to stop it. Why the British didn’t take this option is unclear, but it likely wasn’t due to a healthy respect for the sovereignty of Indian rulers (it isn’t as if India invited the British in, after all). There is some evidence that the British were playing a game with the princes, in which they were flirting with the possibility of granting independence to some of the larger states, such as Kashmir, Hyderabad in central India, and Travancore in the far south. The princes were seen as the most pro-British element of the Indian political leadership, and some in London, including Winston Churchill, thought of them as a way to maintain a British foothold in the sub-continent. The Indian leadership considered independent princely states completely unacceptable and succeeded in obtaining the accession of all of the princely states which were contiguous with India through a mix of patriotism, religion, bribery, shady assassination attempts, and a few strategic invasions. It is a pretty fascinating story, which to my knowledge has never been fully told.

Anyway, Kashmir was one of the tough cases. It was a Muslim majority state, but had a Hindu ruler, Maharaja Hari Singh. At first he made some noise about independence, which, for what it’s worth (sadly, not much), is what the majority of Kashmiris actually wanted. Then, he led the Pakistanis to believe that he was moving towards accession, although India was of course also wooing him. The Pakistanis lost patience and infiltrated irregular fighters over the border to take Kashmir by force. The Maharaja desperately called New Delhi for military assistance and India agreed to send in troops, but on one condition…

The Maharaja had no choice but to agree to India’s demand to accede, and the Indians were able to take the vast majority of Kashmir, but not all of it. This set up the dispute of the territory, which continues to this day (both Pakistan and India claim Kashmir in its entirety). The Kashmir dispute has shaped both Indian and Pakistani politics, especially Pakistani. For now, lets just take a quick look at the pre-Partition demographics of the state.

Jammu and Kashmir had a total of about 4 million people, of whom 76.4% were Muslim, 20.1% were Hindu, and 3.49% Other, mostly Sikh and Buddhist. The two main cities were Srinagar, with 208,000 people (78.4% Muslim, 20.7% Hindu, 0.9% Other), and Jammu, with 50,000 people (60.7% Hindu, 31.6% Muslim, 7.8% Other, mostly Sikh).

The language map in Kashmir is a real mess. I won’t post it here, but if you’re interested take a look here. Kashmiri is the most common language, but is only spoken in the Valley of Kashmir, centered on Srinagar. In the far east of the state, speakers of Balti and Ladakhi, two closely related Tibetan languages dominate. In the southern Jammu area and the west around Muzaffarabad, most people speak Western Pahari languages that link other Paharhi languages such as Nepali with Punjabi and Hindi. Pothohari, the dialect in the Muzaffarabad area, is often classified as a Punjabi dialect. Gojri, a Rajasthani dialect which somehow ended up in Kashmir is spoken by a minority as well. In the far north, most people speak Shina, a language related to Kashmiri, or Burushaki, a language isolate with no known relatives.

Below is the much simpler religious map of the state. I used the 1931 Census data instead of 1941 because I was able to find more detailed data for the former. The template map I used also came from the 1931 Census.

Jammu and Kashmir Religion 1931

I don’t think I need to draw in zones like I did for the Bengal and Punjab posts, because it’s pretty obvious who lived where in Kashmir. The Hindus were a majority in the southern area around Jammu, and significant minorities in the west, and in Srinagar in the center of the Valley of Kashmir. Buddhists were the majority community in the very sparsely populated eastern part of the state. Muslims dominated everywhere else, and made up close to 100% of the population in the north of the state. Unlike in Bengal and Punjab, a partition would have been fairly easy, especially because the Hindus in the south didn’t share linguistic ties with the Muslims of the Valley and to the north. Since the de facto partition of 1947, India has administered most of Kashmir, including the overwhelmingly Muslim Valley of Kashmir. Pakistan got the north and the very western edge of the former princely state. Below is the same map with the 1948 ceasefire lines (or Line of Control) drawn in as best I could.

Jammu and Kashmir Religion with Borders

Obviously, the haphazard way in which Kashmir was split made religious considerations impossible, so the Line of Control bears no relation to the religious demographics of the state. Below is what the former princely state looks like today religiously. I used this map as a template. One quick administrative note: the border between China and India in the maps is different, but that might reflect genuine uncertainty on the part of the British as to where exactly the border fell (see the 1962 War for more detail).

Kashmir Religion Today

The Hindu minorities in what is now Pakistani Kashmir are gone, and the south is more Hindu. Overall, unlike in Punjab and Bengal, Partition did not have a big influence on the religious makeup of Kashmir. Perhaps the fact that the border was never open for immigration stopped people from moving, or the overwhelmingly Muslim nature of the Valley prevented the population from leaving. The Valley did have an economically influential Hindu minority known as the Kashmiri Pandits, most of whom have been forced to flee since the late 1980s, but Muslims have always made up made up at least 90% of the population in this critical central region. Finally, perhaps the fact that the Pakistan movement never caught on in Kashmir contributed to a reluctance to leave everything behind and move to Pakistan.

This last point is important in understanding Kashmir. The Kashmiris do not want to join Pakistan. They never have and probably never will. That doesn’t mean that they like the Indian presence in the Valley either (India has as many soldiers in Kashmir as the United States had in Vietnam). If given a choice, which won’t happen, the Kashmiris would choose independence. In a 2010 poll, 66% in Indian Kashmir said they would chose independence, while a whopping 6% favored a merger with Pakistan. India tacitly acknowledges that it is holding Kashmir against its will, but Pakistani politicians frequently make tear-jerking statements in support of their “brothers” in Kashmir and call for a referendum and self-determination for Kashmir. Don’t be fooled. Pakistan holds one third of Kashmir. Why not lead the way and hold a referendum? Well, a different 2010 poll put support for independence on the Pakistan side of Kashmir at 44%. Obviously, they think that a referendum would be too close to risk. Both sides have run roughshod over the Kashmiris at every turn.


How Did Partition Change the Religious Map in Bengal?

The Punjab gets most of the attention when it comes to Partition, probably because of how disastrously everything went there, but on the other side of India, the British divided another major province along religious lines. Partition in Bengal was more orderly, although not without some violence, and simpler, because of the binary Hindu/Muslim split, as opposed to the Hindu/Muslim/Sikh mess in Punjab. As in my earlier look at Partition in Punjab, I used the 1941 British Census data and this excellent map as my basic template. And just like last time, I’ll do a quick rundown on Bengal on the eve of Partition, as well as neighboring Assam, which was also subject to Partition (although India got all but one district).

In 1941, a total of 70.5 million people lived in Bengal and Assam combined, 62 million in Bengal, the rest in Assam (which encompassed all of northeast India at the time, not only the modern state of Assam). There were a few small princely states, but 96% of the population lived in British India. The religious breakdown in Bengal was 53.4% Muslim, 41.7% Hindu, and 4.8% other, mainly people practicing tribal religions. In Assam, 41.5% were Hindus, 31.8% were Muslim, and 26.7% were tribal. That may sound like a lot of people who were not Hindu or Muslim, perhaps a Sikh equivalent in Bengal, but the followers of tribal religions were not united in the same way the Sikhs were in Punjab, and most of them lived far away from the heavily contested areas.

The 1941 Census summary that I found doesn’t have language data, but it is safe to say that Bengali was the most widely spoken language in the area. Assamese would have been second, followed by all of the Tibeto-Burman (related to Burmese and Tibetan obviously) and Austro-Asiatic languages (related to Vietnamese and Khmer) spoken by the tribal peoples.

Despite a population twice that of Punjab, Bengal and Assam didn’t that many large cities. Calcutta was, of course, British India’s largest city, home to 2.1 million inhabitants. The only other cities with more than 100,000 people though, were Howrah with 379,000, Dacca with 213,000, and Bhatpara with 117,000. Hindus were the largest community in all four cities: 72.7% in Calcutta, 81.8% in Howrah, 60.7% in Dacca, 69.2% in Bhatpara, with Muslims making up almost all of the remaining share of the population. Three of these cities went with India at Partition, the other, Dacca (now Dhaka) eventually became the capital of Bangladesh. Below is the map of religion in Bengal and Assam as it was in 1941, six years before Partition.

Bengal and Assam Religion, 1941

In my post on Partition in Punjab, I divided the province into five clear zones based on religious composition. Unfortunately, Bengal cannot be so neatly divided. Nevertheless, I identified a southeastern, heavily Hindu zone centered on Calcutta, a central Muslim majority zone, and a northeast zone that was a mess of different religions, including Hinduism, Islam, Buddhism, Christianity, and tribal religions (the later three are all classified as “other” on this map). Below is the map with my zones drawn in:

Bengal and Assam with zones

I would have given Zones 1 and 3 to India, and 2 to Pakistan (remember Bangladesh was originally part of Pakistan). Interestingly, according to this map, that is exactly the partition proposed by the British Parliament in the Indian Independence Act of 1947, which was passed in June 1947 before the boundary award, and probably represented Parliament’s best guess at what the final division would look like. That isn’t what actually happened though. Below is the final boundary award:

Bengal and Assam Religion final boundary

The final boundary isn’t really pro-India or pro-Pakistan. It’s just very…eccentric. India ended up with large chunks of Muslim-majority territory on the west, and one random Muslim-majority slice of Syhlet in the east (Syhlet was the one district in Assam that went to Pakistan). Pakistan however got the overwhelmingly Buddhist Chittagong Hill Tracts in the far east, and the Hindu-majority district of Khulna, which was contiguous with Indian West Bengal. There appears to be no rhyme or reason for any of this. I used to assume it was because India wanted West Bengal to be contiguous, but it wasn’t immediately after Partition. It was made so later by transferring part of Bihar to West Bengal.

Maybe there was concern about breaking up areas that were economically interdependent. That hypothesis doesn’t pass muster though because Bengal’s economy was famously crippled by Partition. The jute-producing East was cut off from the jute-processing West, and the tea plantations of Assam lost the quickest route to the sea, and a major port in Chittagong. And remember that Calcutta was the largest city in British India? In part due to the economic stagnation brought on by the severing of Calcutta from its supply of jute, Calcutta has fared badly in the last 65 years. It is now the tenth largest city of the former Raj behind Mumbai, Delhi, Bangalore, Hyderabad, Ahmedabad, and Chennai in India, Lahore and Karachi in Pakistan, and Dhaka in Bangladesh. The British obviously didn’t give a damn about wreaking economic devastation on Bengal, so their strange partition can’t be explained by some kind of commitment to reducing the fallout.

Interestingly, the British also partitioned Bengal in 1905, publicly for administrative reasons, but it was actually a ploy to reduce Bengali influence. That partition, which was quickly reversed, electrified the independence movement and prompted the British to move the capital of the Raj from Calcutta to Delhi. So clearly there was no love lost between the British and Bengalis. Finally, the arbitrary way in which the British partitioned Bengal also makes me doubt that my theory that the boundaries in Punjab were part of a plan to spare the Sikhs undue suffering. The British just didn’t care enough about India to think up something like that. Or rather, perhaps they cared about India, but they certainly did not care about Indians.

Now, we reach the comparison section of the post. Below is the 1941 Bengal and Assam map again with no lines drawn:

Bengal and Assam Religion, 1941

Below is a map of Bangladesh, West Bengal, and northeast India as it looks today, using 2001 Census results for India and Banglapedia for Bangladesh:


Note that I added Arunachal Pradesh and Sikkim, which weren’t on the 1941 map. Also, there are some areas on the west of the map that have been added to West Bengal since Partition. Here, we have a tale of two Bengals. The western half (or third really) is basically unchanged since Partition. The Muslims of West Bengal stayed in India. Bangladesh however, is a different story. I really don’t even need to trace it out on the map, because the bright green of Bangladesh stands out from the muddled browns of West Bengal. The Hindus of East Pakistan/Bangladesh didn’t leave during Partition in large numbers, but instead steadily trickled out of the country over time, with a large spike during the 1971 War. There are still many more Hindus in Bangladesh (around 10% of the population) than in West Punjab (about 0.5%), and the province was spared the apocalyptic cleansing faced by the Punjab.

A few other features to note include how much more Hindu the eastern end of Assam has become. I don’t know what caused that. Another major shift in this region, which I didn’t show because I only have three colors to work with is the transition from tribal religions to Christianity in the far eastern states of India. That story isn’t relevant to Partition, but maybe I’ll revisit it at a future date.

Poverty in South Asia by the Numbers

Poverty-fighting measures tend to focus largely on sub-Saharan Africa, and rightly so given the desperate poverty in the region. Unfortunately, there is also a tendency to ignore South Asia. While it is true that the poverty rate in India and the other South Asian countries is lower than in the worst African nations, South Asia has the most poor people in a concentrated area. The Multidimensional Poverty Index, created by the University of Oxford and the United Nations to measure poverty in a more complete way than by income, is a good tool to compare poverty in different regions.

As expected, no country in South Asia ranks especially high by percentage of the population living in poverty. The poorest countries by percentage of the population are the usual suspects, including Niger, Ethiopia, Liberia, and the Central African Republic. By total numbers of people living in poverty though, three of the top five countries are South Asian, with small but impoverished Nepal also in the top fifteen. In total, South Asia has about 800 million people who are considered poor by the MPI standard, only a little less than the entire population of sub-Saharan Africa. Approximately three-quarters of these poor people live in India.

The heart of South Asian poverty is in a belt of high population states that runs from Rajasthan, on the Indian border with Pakistan, through the Hindi-speaking heartland in northern India, all the way east to Assam, plus Nepal and Bangladesh. This swath of territory represents about 1% of the land on Earth, but is home to 12% of the world’s population and 32% of its poor people as defined by MPI. To provide perspective, that is an area slightly larger than Alaska packed with about 850 million people, of whom nearly two thirds, or a little less than twice the United States’ entire population, live in poverty. Two Indian states, Uttar Pradesh and Bihar, that encompass the Gangetic Plain are the epicenter of this region. They make up 0.2% of the world’s area, 4% of its population, and 12% of its poor. Combined, they are about 75% the size of California, but with a population of 300 million (roughly the same as the U.S.), of whom three quarters live in poverty. The poverty belt is highlighted below in light blue, with Uttar Pradesh and Bihar in darker blue:

South Asia Poverty Belt

While Africa continues to be the most impoverished region of the world, with Niger scoring as the worst country per capita by MPI, South Asia has the highest concentration of poor people. For example, the poverty belt in South Asia has a combined area just 40% larger than that of Niger, but a population that is a whopping 5300% higher than Niger’s. Clearly, we should try to eradicate poverty in all areas in the world, but fighting poverty has the greatest chance to succeed in the Indian subcontinent, due both to its extraordinary concentration of poor people in a relatively small area, and to its political stability compared to Africa.

How Did Partition Change the Religious Map in Punjab?

July 2015 Update: See my more detailed look at Punjab here (but read this post too of course!).

Ever since I became interested in the Partition of India, I have been puzzled by the dearth of good maps showing the distribution of different religious communities in India on the eve of Partition in 1947. The main religions in India are few enough to make mapping possible but numerous enough to make it interesting; the British had carried out a detailed census of India as recently as 1941. All the information exists, and the story of Partition is one of the most consequential of the last century. So where are the maps?

I took matters into my own hands and made some new maps. For the map posted below, I used the 1941 Census numbers and this map as a base. The base map is one of the few decent maps available showing the pre-Partition religious situation in Punjab, and, more importantly for my purposes, it shows the districts and main princely states of the region.

A quick primer on Punjab in 1947: Most of the undivided Punjab region was part of the British Indian province of Punjab. Some medium-sized princely states were sprinkled in as well. Most Punjabi speakers lived in Punjab, though some lived (and still live) in what was then called the North West Frontier Province. The southeast and northeast of Punjab province was inhabited by non-Punjabi speakers. The Punjab region was home to about 35 million people, roughly 4/5ths of whom lived in Punjab province, the remaining 1/5th in the princely states.

The Punjab had seven cities with populations over 100,000. The capital, Lahore was the largest with 630,000, followed by the Sikh holy city, Amritsar, which housed 390,000. The other five were Rawalpindi, Multan, Sialkot, Ludhiana, and Jalandhar, all with populations between 100,000 and 200,000. All but Jalandhar and Rawalpindi had Muslim majorities. Those two had Muslim pluralities (or, if you prefer, Hindu+Sikh majorities). The overall religious distribution in Punjab, including the princely states, was 53% Muslim, 30% Hindu, 14.6% Sikh, 1.4% Christian, and 1% Other. Muslims were concentrated in the west, Sikhs in the center, and Hindus in the east. Hindus were also relatively prevalent in cities and Sikhs in rural areas.

Below is my new map, which takes the base map with districts colored simply by whether it was majority Muslim or non-Muslim, and adds two things. One is that it distinguishes between Hindus and Sikhs, so you can see where the “non-Muslims” in question were predominantly Sikh or Hindu. The other is the color gradient, which allows me to show districts where Muslims were 51% as different from those where they were 95%. In the map below, bright green signifies Muslims, blue is for Sikhs, and red for Hindus:

Punjab Religions 1941

The Punjab can be divided into five areas. One is the west, which was generally 80% or even 90% Muslim. The second is the center-west, which was majority Muslim, but typically around 60% and with large Sikh minorities. The third area is in the center-east, with no obvious majority religion. This is where much of the worst carnage during Partition took place. In some places, the Sikhs were a plurality, in some the Muslims, and in some the Hindus, but rarely was any one community a majority. The fourth area is to the southeast, in what is now Haryana. This part of the Punjab had a Hindu majority, but it was relatively narrow, and the communal split was Hindu/Muslim, with few Sikhs in the mix. In this map, Delhi is included as zone four, because communally and culturally, it was similar to the nearby parts of the Punjab. The fifth zone, which corresponds to the modern state of Himachal Pradesh was almost exclusively Hindu. Below is the same map, but with my zones drawn in:

Punjab Regions

Looking at this map, reasonable Partition lines are fairly obvious. Pakistan should get areas one and two, and India four, and five, with three being divided between the two, probably with most of it going to India. Below is the map again, with the claims made by Congress (in black) and by the Muslim League (in white), as per these maps, drawn in:

Punjab 1947 Claims

The difference between the two claims is stark. The Congress claim is maximalist: in addition to the heavily Hindu areas (4 and 5), they claimed all of zone 3, 2, and even a few parts of 1. I don’t know what the argument for giving those heavily Muslim regions to India would have been. Perhaps it was a negotiating tactic, or an attempt to keep the Sikh heartland undivided. The Muslim League asked for much less, only claiming zones 1 and 2 and most of the Muslim plurality parts of zone three. Below is the final boundary (in pink) drawn by the British:

Punjab Claims+Boundary

To my eyes, this looks like an extremely favorable result for India. No Hindu/Sikh majority district went to Pakistan, while several swaths of Muslim majority territory ended up in Indian hands. The explanation that comes to mind is that the British wanted to try to ease the damage Partition would do to Sikhs, who clearly got a raw deal with Partition. Their homeland was split in half, leaving many of their holiest sites, including the birthplace of the founder of Sikhism, abandoned in Pakistan. Lahore, which had been the capital of their early 19th century empire, also went to Pakistan. Unlike the Muslims, they didn’t even get a state out of the carnage, and in Punjab as it was then formulated, they would remain a minority. The British respected the Sikhs perhaps more than any community in India, because of their long service in the British India Army, and their loyalty during the 1857 revolt. Perhaps the generous lines on the map were intended to keep as many Sikhs in India as possible, and therefore reduce the number of uprooted Sikhs . My theory would also explain the very favorable lines in Sindh (or no lines: Sindh wasn’t partitioned despite a Hindu majority in the southeast) and Kashmir. The British expected the Muslim-majority Princely State of Jammu and Kashmir to accede to Pakistan, leaving the Hindus in the southern Jammu area in Pakistan. The British plan in Sindh and Kashmir balanced the pro-India lines in Punjab (obviously, India foiled the plan in Kashmir).

In any case, whether the British had complicated motives, or just didn’t know what they were doing, the lines were drawn, and all Hell broke loose. Virtually all of the Punjabis who found themselves on the wrong side of the new border left or died trying. It was one to the largest population exchanges in history (around 11 million people in Punjab crossed the new border). Here is the 1941 map again:

Punjab Religions 1941

Below is the religious picture of the greater Punjab region today (or ten to fifteen years ago when the data I used were collected). I added Buddhists in yellow, and since I couldn’t find any district-level data for Pakistan, I colored all of the Pakistani side the same color (97.2% Muslim, 2.3% Christian, 0.5% Other, which is the overall religious breakdown for West Punjab). I assumed that, with half a percent of the population, Hindus and Sikhs wouldn’t show up anyway. There is one religious map of Pakistan, which shows a Hindu majority in the desert south of Bahawalpur. I do not know what numbers this is based on, but I haven’t seen it anywhere else, so I’m ignoring it, at least until I find their data.

Greater Punjab Religions Today

Obviously, the Pakistani side is almost completely Muslim, while the Muslims have left the Indian side except in the area south of Delhi. A pocket of Buddhists has emerged in the sparsely populated far north, apparently mostly consisting of Buddhist refugees from Tibet. The Sikh population is completely concentrated in what is now the Indian state of Punjab, where they are a majority. In 1941, they were not a majority there, but the Muslims left and Sikhs from Pakistan arrived. Over all, Partition drastically changed Punjabi culture and demography in ways that would profoundly influence the courses of both India and Pakistan, and the maps tell the story in the simplest and most direct way.