A Geographic Perspective on Attacks on the Hindu Community in Bangladesh

This post returns to the topic of religious cleansing in Bangladesh, the same subject as a post a few days ago. While I won’t be revisiting this topic continually, as this blog is supposed to be more than tracking current events, I do find myself keeping tabs on this particular country and issue. As I’ve said previously, I find issues of identity and violence particularly compelling within the discipline of Geography and considering that I wrote my thesis on Bangladesh… let’s just say I can’t keep away!

For this post, I wanted to map out recent attacks on the community as reported in a battery of news reports from BDnews24 (while I’d prefer the Daily Star, BDnews24 has a working RSS feed, well done). From my thesis I ended up acquiring district-wise population of religious communities in Bangladesh, but only for 1991. Apparently, these numbers were captured in 2001 or they were not reported. Incidentally, the center-right Bangladesh National Party was in power in 2001. Guess who was part of the ruling coalition? If you guessed the Jamaat-e Islami Bangladesh I owe you a drink. At any rate, not reporting the numbers of religious minorities during that particular regime is a little… odd, to say the least. Keep in mind that academics, and some NGOs, have been discussing the “missing millions” of Bangladeshi Hindus for at least a few decades at that point in 2001. The Census results for 2011 seem to be available, but in a useless format (PDF) so it will take me some time to get it converted to Excel and then into a more user-friendly GIS format (stay tuned!). More to the point, religious statistics are also in the 2011 results (you really have no idea how excited I am right now!!). And to make it even better, individual villages, unions, and sub-districts (upzilas) have that information – this is unprecedented! Well folks, I’m sorry but there will be some more Bangladesh-centric blogging!

Back on topic, obviously I don’t have the 2011 data available (since I literally just found it!) so we’ll have to make do with the 1991 until I crunch it so, at the very least, there will be an update.

As we can see on the map below, the Hindu community (at least the folks who decided that they would self-identify, which isn’t a foregone conclusion in a society that has routinely targeted minorities for prejudice and violence) is primarily concentrated in peripheral districts outside the capital at Dhaka (which is the central city figuratively and literally). The largest concentration is in the southwest, in Khulna and Barisal division (those borders are not shown). As I indicated in the previous post, Khulna “should have” gone to India in 1947. Joya Chatterji (1999) has a great article on the making of the Bengal frontier. The deciding factor for Khulna being provided to Pakistan, rather than India, lies not in human geography (since Khulna was majority non-Muslim at the time) but in economic geography. Calcutta, one of the principle trade centers for British India, was going to India. However, its link to the Ganga River was bisected by a Muslim-majority district, Murshidabad. In order to secure Calcutta’s link to the rest of India, the Indian National Congress was willing to trade Khulna for Murshidabad. The other aspect, which Chatterji points out deftly, is the internal politics at work. She argues that INC leaders were already thinking ahead to dominating electoral politics and so cut-out Khulna and the likely voter base there. You can identify, roughly, where Murshidabad is located. There’s a noticeable lack of Hindu communities north of the Khulna concentration, just across the border lies Murshidabad (in West Bengal state, India).

Attacks on the Hindu Minority Community and its Population, 2013/1991 (via ME!)

Attacks on the Hindu Minority Community and its Population, 2013/1991 (via ME!)

Overlaying recent attacks reported by BDnews24 (there are 1, 2, 3, 4, 5 reports mentioning 7 distinct sub-districts), there are only a few conclusions that we can make, none of them strongly. I’ve heard elsewhere that Human Rights Watch has a report out condemning violence and noting at least 100 attacks, that would be useful material for updating this work. In the meantime, the attacks betray JIB’s diffuse reach in Bangladesh. It would seem that JIB, and their supporters, are either avoiding targeting the community in the largest metropolitan areas (the divisional capitals) or I’ve missed the reports (which is likely). However, temples are being burned to the ground in the district towns, such as Gazipur Sadar and Faridpur Sadar (labelled Gazipur/Faridpur S.). From an organic state framework, these attacks are happening in relatively state-powerful areas – district towns. One hopes that these attacks aren’t occurring with the complicity of the local police…

Moreover, the attacks are occurring where there are relatively large number of Hindus (like Morrelganj in Bagerhat district, Khulna, the southern most) and small numbers of adherents (Shibganj in Chapain Nawabganj district, Rajshahi, the western most). In Nawabganj, Hindus made up 4.8% of the population in 1991, in 2011 they made up 4.0%. In Shibangaj itself, Hindus make up only 2.8% of the population (2011 data).

The decrease is starker in Bagerhat district. In 1991, 22% of the population claimed to be Hindu, in 2011 that proportion dropped to 18%. In Morrelganj, the minority Hindu community makes up 10.5% of the population.

When I had initially started this particular geographic exercise I hypothesized that JIB activists and supporters might not target the Hindu community in large concentrations (like Khulna) because of the likelihood for retaliatory attacks. Instead, I hypothesized that the JIB would target the community in places where it was less populous (like Rajshahi). Obviously I was wrong and right on each count, respectively, forcing me to amend my hypotheses as I continue to dig.

I could amend this model to include the fact that local JIB leadership will target, whoever and wherever they feel like, population-size be damned. Additionally, I should account for decreases in the Hindu population. Why? Because if I know its decreased, based on a Census report, than local militant Islamists certainly know it. What do they have to fear, there’s less Hindus now then there were in 1991. On the other hand the minority community of Hindus, while the largest minority religious group in the country (by far), is fast dwindling. After being routinely targeted over the decades and assuming that the Government, now as well as then, isn’t going to assist, and being acutely aware of being a minority group – perhaps the only recourse is simply to leave the country? In the end, the Islamists get one of the things they want: a fully “Muslim” Bangladesh.

This is one of the problems with the 24-hour news cycle and humanity’s inherent attention deficit disorder. We’re quick (or not) to jump on obvious genocides: Nazi Germany, Rwanda, Bosnia. But slow-burning problems, like the displacement of millions of Bengali Hindus, fails to spark a debate. Even within India, a country one would think would be more sensitive to this situation, this is scantly (if ever) reported. On the other hand, the plight of Tamils in Sri Lanka has been grabbing headlines in the Times of India for weeks, literally. But, then again, Sri Lanka’s treatment of Tamils (even before the emergence of the Liberation Tigers) leaves much to be desired, but that’s for an upcoming post.


Bangladesh and Hindus: A Geography of Cleansing

A special edition post reflecting the news from Bangladesh on the execution sentence issued to one of the primary leaders of the Jamaat-e-Islami Bangladesh (JIB). For those not in the know, the JIB is an Islamist political party (a party seeking to replace the “democratic” state of Bangladesh with one based on shari’a [Islamic] law). While I could write at length (and have elsewhere) on Bangladesh, the salient pieces of information here are the country’s lack of coherent state identity (are we Bengali or Islamic?) and it’s history of accommodating Islamist ideology (that the center-left Awami League is attempting to reverse). The execution order comes on the heels of a trials focused on accusations of war crimes (genocide) committed by senior JIB leaders during the country’s independence conflict in 1971 (independence from Pakistan). Estimates vary but between 1 million and 3 million people in what was East Pakistan lost their lives. Unsurprisingly, supporters of the JIB, including party activists and the party’s student wing (Islami Chhatra Shibir), have started rioting in the country.

The riots themselves are the subject of this special edition post for they betray the JIB’s view of Bangladesh and their answer to the “state identity” question. This current view conforms to the reason its senior leaders stand accused of war crimes, as we shall see. Generally, these riots highlight the nexus between identity, geography, and people, in other words, human geography.

As reported in Bangladeshi press, JIB activists “torch[ed]” a Hindu temple in the southeastern district of Noakhali of Chittagong division. Tensions between the Muslim and Hindu communities have been long standing (since at least the early 1900s) and due, in part, to British imperial policy. As most colonial and imperial powers learned, it was easier to govern areas by empowering a community (often a minority) over the others. The history of Islam of India (and Bengal) extends much before the arrival of the British, of course, generally Islam came via the sword (Mughal and Turkish conquerors) who were generally Persian and elitist. The other avenue were through Sufi monks, who often adapted Islam to local practices, it was this latter avenue that generally won converts among local population groups. However, the most “Indians” (keep in mind a unified political entity of India didn’t exist until the British left, and technically still doesn’t). Because of the existing bureaucratic and administrative structure of (Muslim) Mughal “India”, it was easy for the imperial power to co-opt as it replaced Mughal authority with British authority. And it just so happened that the Muslim community was, based on the entirety of Bengal province, a minority.

As the population of the province grew, attracted by potential economic gains thanks to Calcutta’s link to the wider British empire, the British decided to sub-divide Bengal to more easily govern it. The province underwent two or three “partitions” before being officially partitioned in 1947, during Indian/Pakistani independence. Generally, the partitions and combinations angered one community or another (as land, power, and people are inseparable). But the big wrench was the 1947 partition.

At this point Indians had to answer the question, based on Pakistan’s departure, are we a country of Indians (despite Pakistan’s removal, and adjusting for the north-south divide) or are we a country of Hindus (because of Pakistan’s removal, though allowing for a north-south regionalism). Pakistan, until 1971, had it easy – “we are a country of Muslims” – united. Of course, in practice the government was dominated by Punjabi Muslims in West Pakistan. In 1970-1971, when East Bengal managed to elect Sheikh Mujibur Rahman as a Prime Minister, he wasn’t allowed to take his position. Bangladesh, with India’s help, gained its independence in 1971 after enduring a civil war that killed millions and accusations of atrocities against the Pakistani Army and Bengali collaborators (bringing us to the trials against the JIB leadership). However, Bangladesh now had an existential crisis – what defines the state? Are they Bengalis? Perhaps not since West Bengal state (primarily Hindu) is part of India. Are we Muslims? Perhaps not since we just broke away from “Pakistan”.

Obviously, the JIB sees the answer as Bangladesh is a “nation” of Muslims, pan-Islamism in regards to Pakistan. This is why their senior leadership committed war crimes against Bengali civilians during the independence movement as part of the al-Badr and Razakar militias. This is why JIB activists targeted a Hindu temple during this unrest, because they are not a legitimate part of the body politic (to the Islamists) . The situation for Hindus is so bad in Bangladesh that a number of academics have discussed the “missing millions” in Bangladeshi Hindu population. While some are murdered, most are forced out of their homes, their lands and property confiscated, and wind up in India. While the JIB may certainly engage in this activity there’s bound to be additional culprits. A systematic study of the links between political parties and social organizations with this problem has yet to be undertaken by Bangladeshi academics (and if it has I don’t know about it).

The geographical link lies in the relative concentrations of the Hindu community within Bangladesh. Primarily, according to past Census data (1991 and 2001), the community is concentrated in the country’s southwest in Khulna division (see map below), which shares a border with West Bengal state. This relatively large concentration of Hindus was actually used as a reason to include that division with India before the 1947 partition. While the countries were largely organized along social lines (Pakistan being Muslim, India being non-Muslim), there were notable economic exceptions. In the case of Bengal, urban areas were allowed to control their hinterlands. Thus, Calcutta ended up receiving hinterlands in Murshidabad that were primarily Muslim, while combined Pakistan (through Chittagong) received non-Muslim hinterlands in the Chittagong Hill Tracts. Khulna was somewhat of a surprise decision, assigned to Pakistan despite having a large Hindu minority population.

Hindu Community in Bangladesh, 1991/2001 (via wikipedia)

This geographic disconnect between monolithic state identity and local-level realities continue to be a source of instability for both countries. Indeed, another Bangladeshi press article reported that hardline elements of the “sizeable population of Urdu-speaking Muslims” (interesting since Bengali is a separate language) sympathize with the JIB. Moreover, West Bengal “hardline Islamist radical” elements are attempting to enter Bangladesh.  The article reports that India is calling for the sealing of the border between West Bengal and Bangladesh.

Geographic Impact of Demography: Haryana’s Male Preference

The preference of much of South Asia’s families for male children has been well documented in both academic and popular papers. Often, male preference is attributed to “cultural” factors grounded in both social/community preference and perception (like status or standing in the community) and in economics (males viewed as more likely to provide money for the family). While the historical extent of male preference is outside the scope of this post, we can safely say that male preference at birth (manifested by female infanticide, the killing of female foetuses, and sex-selective abortions) has been ongoing for at least the past 30 years, probably longer. The population impact has been terrible, according to United Nations Population Fund paper published in 2010, India is “missing” over 600,000 girls, and that’s only from prenatal selection (or sex-selective abortion) between 2001 and 2007. In other words, that does not count infanticide making the number of missing higher.

Today’s post looks at the demographic impact of this “cultural” preference on geography through the case of Haryana, a state in India’s northwest. But first, how do we know there is a preference among families and society for males? Aside from the anecdotal “cultural” evidence, depictions in media and so on, there are demographic statistics we rely on. As part of the preliminary results of the 2011 Indian Census, the Indian government published the total sex ratio (the number of women per 1,000 men) of its states and territories based on the results from the 2011 and 2001 censuses. Available on wikipedia (as well as the Census website, and reproduced below), we find that Haryana is among the worst states for the total sex ratio, 877 women per 1,000 men. The total sex ratio is reflective of the population at-large, and while it does reveal male preference over time it doesn’t reflect current levels (or the persistence) of male preference at birth. The child sex ratio (defined as the ratio of girls to 1,000 boys aged 0 to 6) is another measure of sexual imbalance in a young population, but it is also influenced by such factors as non-registration of female children. Finally, the sex ratio at birth (defined as the number of female babies born per 1,000 male babies) captures male preference at birth, including sex-selective abortions, but misses infanticide. When looking at these numbers, its important to keep in mind what other factors (besides the one you’re looking for) is potentially influencing the statistic.

Haryan’s child sex ratio betrays a continuing preference for males (and potentially, under-reporting of females). According to the Census data, the 2011 child sex ratio was 830 girls per 1,000 boys (compared with 877 girls per 1,000 boys in the total population). The child sex ratio has only a slight disparity between rural and urban areas with rural areas reporting 831 girls and urban areas reporting 829 girls per 1,000 boys. The all India child sex ratio is 914 girls per 1,000 boys (as reported in the Economic Times), as that article notes, this is actually a decrease from 2001, when the ratio was 927. The Haryana data also betrays a significant decrease in the child sex ratio, from 964 girls to the current ratio of 830. This is very large decrease and I wonder if it is a typo (perhaps 864?) or an example in a collapse in the reporting of female children in the system?

Even without the revelation of the lower sex ratio between the decadal censuses, the lower sex ratio is significant. Depending on the demographer you ask, a “normal” sex ratio is just over 1,000 girls to 1,000 boys. The reason for this is that females are naturally stronger than males, not strong as in physical strength, but in terms of physiological resilience. Female babies are stronger and women tend to live longer. Examine population pyramids for any country and you will almost always find larger numbers of elderly women (especially in the 80s, 90s, and 100+ age ranges) than men. Sorry guys, but we are the weaker of the species and we try to make up for it with physical prowess (which kills or maims us and generally shortens our life spans anyways, see: war).

While this low child sex ratio in Haryana (and India in general, with the notable exception of Kerala state) implies a continuing problem for the next generation or two, the geographic behavior of Haryana residents has already been affected. According to the Times of India article that inspired this post, a study funded by the Norwegian embassy in India found that the “skewed” sex ratio in Haryana is forcing “poor, upper caste youths” to “hunt for brides” in other states of India; Assam, Odisha, Bihar, Maharashtra, and West Bengal. One of the key findings of course is that the unbalanced sex ratio in Haryana is disproportionately affecting poorer upper caste men. For a bit of background, India is infamous for its caste which, generally, organizes the population into four broad-based classes and “out-castes” (from which we derive the term “outcast”). Of course, each caste group is further sub-divided so that even among the top caste (or Brahmin) there are higher sub-castes than others. What makes the caste system infamous is its rigidity, its direct and indirect influence on your life chances, and of course, that it still exists (though steps are being taken to lessen its influence). Its similar to civil rights movement in the United States, and that’s not even completed yet, everything is a work in progress. At any rate, rich upper caste youth in Haryana have a greater chance of marrying other upper caste women in Haryana (since there’s less women they can be pickier).

Another very interesting finding is that “shortage of women is not common across all caste groups in the conjugal regions, but is endemic in dominant caste groups of Jats and Yadavs.” The implication is that this situation is a rich person’s problem. Like most societies, the rich in India have the most vested interest in maintaining the status quo: the caste system (of which they sit atop), dowry, and other such “traditional” notions. But the interesting turn of events is that some of the dominant caste group’s are, literally, unable to propagate their own communities. While more detailed census information would be needed to ascertain whether the population is decreasing, that some upper caste men have to leave the state to find a wife is indicative of the problem. And of course, the article discusses one of the ways that women can climb the caste ladder – their families lie about their caste to the new-comers.

Ultimately, these two developments: potential population decline of high caste groups and the progeny of upper-caste and not-upper caste marriages will have to be absorbed by Haryana’s, and probably India’s, society. This is unlikely to be a problem unique to Haryana, as the table of child sex ratios below reveals. It seems that demography is either going to force high caste population groups to either abandon their traditional preference for males or their preference for marrying only other high caste females. In effect, demography is challenging the future viability of the caste system. And there’s already inkling of social discontent, as the article reports, out-of-town brides as stigmatized as “bought” (or “sold” by their parents) by other, presumably locally-born, women. One wonders what will become of the children, who will have to grow up with similar taunts and questions of their actual caste? And since this is a widespread problem in Haryana, might future communities and cultural organizations appear there emphasizing far away places like Odisha or Assam? Might this, eventually, be the spark that eventually overcomes regionalism in India, creating an actual “Indian” identity?

Total Sex Ratio of India, 2011/2001 (via Census of India)

Total Sex Ratio of India, 2011/2001 (via Census of India)

From Earth Snapshot: Haze over Bangladesh

For those of you with readers, if you haven’t subscribed to Earth Snapshot yet, you should really consider it. Multiple daily satellite images of Earth, aside from their beauty they’re a valuable learning tool in their own right.

A few days ago, Earth Snapshot released this view (and reproduced below) of the eastern Bay of Bengal showing Bangladesh, Burma (Myanmar), and parts of northeastern India. The caption for the image discusses mountains natural ability to block smoke and haze from traveling over them. As we all know, heat rises, but when it is forced upwards by a mountain chain the air cools, forcing it back down if the mountain ridges are too high. This is an important aspect of the precipitation cycle, warm moist air low-lying areas and oceans are eventually forced upwards when they are pushed into mountain chains, as the air cools the moisture is released as precipitation, watering the mountainside and traveling downhill towards rivers and, eventually, larger bodies of water. In this image, we see haze from India and Bangladesh (attributed to agricultural fires and pollutants) locked out of Burma by the eastern Himalaya.

What is most striking to me is the clear delineation of the border between Bangladesh and the northeast Indian state of Meghalaya. Meghalaya’s landscape is dominated by the Khasi Hills, which are lush and heavily forested. Further north beyond the forested hills are the agricultural lands along the Assam Plain (on either side of the Brahmaputra river). The eastern boundary of Bangladesh with India and Burma is less discernible thanks to the forested Chittagong Hills, foothills of the eastern Himalaya. Finally while Bangladesh is overwhelmingly agricultural land (Dhaka is just obscured by a cloud), we can still make out the Sundarbans in the country’s south, at the western mouth of the combined Ganga-Brahmaputra (Padma-Jamuna) river system. This extensive mangrove forest is home to a wide variety of flora and fauna (including the endangered Royal Bengal tiger). Beyond the wildlife, the Sundarbans also absorb some of the devastation brought by cyclonic activity which routinely ravage the country.

In contrast with the haze over Bangladesh, the skies over Burma are clear allowing us to clearly see the mouth of the Irrawaddy river and its floodplain.


haze over India and Bangladesh (via eosnap)