Maybe size doesn't matter - thinking about censuses in India and BiH

08 September 2015 by Valery Perry

Maybe Size Doesn’t Matter- Thinking About the Censuses in India and BiH

DPC Blog Post
September 3, 2015
Valery Perry

( BHS verzija - BCS version )

I often laugh when well-intended scholars will include India as a possible model for Bosnia and Herzegovina (BiH) in terms of its multi-national federal political structure. The notion that India – a country of well over a billion people, with a unique ancient, colonial and post-colonial past and 22 official languages (22 legitimately different languages, not minor dialect issues) – could be relevant to anyone other than the occasional Balkan-focused political scientist seemed crazy. BiH’s 3.6 million people would constitute an inconsequential city in India; BiH’s municipal populations are nearly the size of the average elite wedding in nuptial-obsessed India.

So imagine my surprise when I arrived to Delhi for an end of summer trip to see that a big news story in Indian politics was related to the census results, an issue that has intrigued me in BiH for years.

The census in India is notable for a number of reasons. First of all, it is a massive exercise, as the results showed a population of over 1.2 billion people. Second, the 2011 census was the first since the British Raj-administered count in 1931 to include a question on one’s caste or tribe, a move seen as potentially controversial and inflammatory by some, but seen as necessary by others interested in using the data in support of leveling social welfare policies. Third, while initial data was released already in 2011, data on religious affiliation was only recently released on August 25, 2015.

Why the delay? Many analysts note that a number of elections were held in the intervening period, and there were concerns that data on religion would be manipulated in critical ways by various parties. There was particular apprehension that any change in the number of Muslims – in general and vis a vis the majority Hindu population – might be used to influence election campaigns and voting decisions. (When India’s 2001 census data was released in 2004, there was controversy on this matter.) In addition, experts pointed out that data on religious affiliation would in any case be difficult to compare to previous counts, as some contested areas in the north (primarily Jammu and Kashmir, but also Assam) were not counted in previous decennial censuses. The inability to simply compare apples to apples would likely leave analysis open to different interpretations and comparisons, as well as manipulation.

Sure enough, my perusal of the Times of India, The Hindustan Times and other news sources pulled me once again into the nerdy yet interesting issue of census enumeration and analysis. After the release of the long-awaited religious data, articles and infographics sought to break down the numbers. Some was basic, looking at the number of the smaller religious groups (Jains, Parsis, etc.), or looking at male/female ratios and trends among the various groups (generally reflecting sex selective abortion and female infanticide). And sure enough, many headlines noted that according to the 2011 count, 79.8% of the population is Hindu, and 14.2% Muslim, with the Hindu proportion of the population decreasing by 0.7% and the Muslim population increasing by 0.8% since 2001. In particular, some analysts focused on the word decrease in terms of the Hindu population, while others pointed out that the rate of growth of the Muslim population has continued to show a decline. Over a ten-year period, the change was nominal; however, in politics numbers and vectors matter, no matter how small. Depending on their political outlook, various analysts emphasized different interpretive angles. Overall, the admittedly limited English-language press I saw was even-handed.

While on a much smaller scale, these dynamics resemble those in BiH census politics. In BiH as well, preliminary results were released promptly; however, there is still no certainty on when the full results of the 2013 count will be released. Concerns that the release of results might coincide with the 2014 election campaign proved unfounded, yet in the absence of results in mid-2015 as hoped one might wonder about the impact of results announced in the run-up to the 2016 municipal elections as well. And just as in India, the delays are related to controversy surrounding how the religious/ethno-national affiliations will be counted, announced, and ultimately used, either in formal policy decisions or in public dialogue, debate and party rhetoric.

It would be interesting if BiH authorities would decide to release data piecemeal, to get non-controversial information out in the public space while continuing to determine how to aggregate the identity-related questions; it is worth reminding that the ethnicity/nationality, religion and mother tongue questions were never requiredby Eurostat or the broader EU/Brussels machinery, and were included solely due to BiH political party agendas. (The EU still went ahead and funded a substantial chunk of the census exercise even though it was well known that these unnecessary questions would cause problems down the line.) However, the piecemeal outcome seems less likely in light of core disagreements over which forms to even count as valid resident respondents.

These cases show that in any country, big or small, censuses are never simple technical exercises in counting bodies, but complex and often highly politicized exercises closely tied to politics, policies and budgets.

So perhaps it was unfair of me to think that BiH and India don’t have the occasional overlap. Let’s just be glad BiH doesn’t have to worry about the issue of caste……