Data, like the scriptures, typically reveals the hand of its creator – which, in both cases, is human. This is a prosaic way of saying that data is not a god: it is fallible and, unlike faith-based dogma, the best studies are the ones that will tell you exactly what their limitations are and why.
This is mostly true of a 2015 series of reports, based on a multiyear study published in the peer-reviewed journal Demographic Research. It looked at the future of religion – how age, fertility, mortality, migration and what they call “religious switching” (opting out of one set of beliefs and possibly choosing another) might shape the world we live in, from population growth to economic progress.
The findings of these reports make for fascinating reading, but also present rich opportunities for hyperbole: whereas the version published in Demographic Research highlighted the study’s focus on the religiously unaffiliated, the Pew Research Centre chose to prominently subtitle its publication Why Muslims Are Rising Fastest. This is an example of how the rapid growth of Islam (true, but requires context; see below) easily becomes the rhetoric employed by the likes of Donald Trump.
According to the reports, between now and 2050 (the reasonable outer limit of the projections, based on current data) “Islam will grow faster than any other major world religion”, increasing from an estimated 1.6‑billion adherents in 2010 to about 2.76‑billion in 2050.
Over the same period, however, Christian believers will still make up most of the world’s population in total numbers (2.92‑billion in 2050) and claim the single largest share of the global population (31.4%).
What is significant, the report adds, is that by 2050 Christians and Muslims may be close to reaching parity for what could be the first time in history (Christianity holding its nominal lion’s share, with a 1.7 percentage point lead).
According to some Pew researchers, Muslims may have already outnumbered Christians between 1000 CE and 1600 CE, when Muslim populations were growing and Christians were being wiped out by the Black Death. But the relatively poor record- keeping – and general absence of things such as human rights, including that of religious freedom – of the Middle Ages mean such estimates are “fraught with uncertainty”.
There is an assumption that, over time, our better, bigger and smarter societies will increasingly reject the crutch of gods or organised religion. But the data shows the opposite. Globally, the numbers of people counted as unaffiliated (this includes atheists, agnostics and people who hold other religious beliefs) will grow over the next few decades.
Because these unaffiliated populations are concentrated in countries with older populations and lower fertility rates, the world population will increase at a much faster relative rate. This means that, although the number of non-affiliated, including nonreligious, people will increase, the proportion of these people will decline to less than it is today.
What we can see from data projections is that the religiously affiliated are increasing. Sub-Saharan Africa, in particular, will be the geographic centre of monotheistic growth in the next four decades, with high increases in population overall and for both Christian and Muslim populations specifically.
Developing countries with traditionally high birth rates and decreased infant mortality are likely to grow quickly in the near future, says the Pew report. But what these studies do not show is whether we are becoming more “religious” – not just in observance, but in that great immeasurable known as faith.
What does it mean to be a Christian, Muslim, Jew, Hindu? Or to be Catholic versus Protestant, or Apostolic Christian? Fundamental distinctions within seemingly homogenous metatheisms can be hidden in the big data, but are exposed in daily life and practice. For example, a new study, also published by the Pew Research Centre, showed that in Israel, although Jews make up 81% of the country’s population, there were strong differences between groups identifying as secular Jews, observant Jews and ultra-orthodox Haredi Jews – and these divisions can be as polarising (politically, religiously, economically) as opposing faiths.
And part of learning to work with data sets is accepting that the way in which we gather data, and which data we deem to be important or irrelevant, changes significantly over time. Population information gathered in, for example, apartheid South Africa typically excluded the homelands and undercounted demographics in township areas.
This does not mean we should discard older, potentially flawed data and censuses, but rather that we should see data for what it does and does not tell us.
The Tao of South African belief systems
Statistics South Africa has codes for a staggering 65 different religious affiliations or non-affiliations. The majority of these are various iterations of Christianity, from Methodist, Lutheran and Anglican to Pinkster, Presbyterian and Pentecostal (Judaism and Islam are given just a single code each) – which is perhaps to be expected in a country where, according to most recent figures, nearly 86% of the country self-identifies as Christian.
The smallest religious group specifically identified in the 2011 census was Taoists, numbering just 371.
The same survey also attempts to measure religious observance (in data, South African Muslims were more devout than Christians), which asked how often respondents attended religious services.
The Pew Research Centre reports give slightly different figures for South Africa – a Christian majority of just 81% in 2010, declining slowly. The difference in numbers can largely be attributed to the fact that the Pew survey is based on old census data, using the results from Census 2001, which were then adjusted.
Even the best data researchers and fact-checkers can only access the best information available at the time – but the difference in figures reveals the significance of small details: not just the proportion of Christians, but even the size of South Africa’s population. The Pew projections assume a South African population of 53.9-million in 2020. Statistics South Africa’s midyear population estimates indicated this figure was already 54.96-million by the middle of 2015, five years ahead of the Pew study’s projection.
These differences will become incrementally larger as the projections reach to 2050, and we must assume there will be similar divergences in other nations or regions.
Again, this does not make the Pew reports and others irrelevant – it’s not a case of lies, damn lies, religion and statistics – but it allows us to identify their limitations.
Islam is on the up, but …
The Pew Research Centre studies looked at the world’s population between 2010 and 2050. It found that, during that time, Muslims were the only population group projected to increase faster than the world’s population as a whole.
With the exception of Buddhists, all major religious groups (including nonreligious and unaffiliated) will grow in absolute numbers during this time. But proportionately their representation will stay the same or decrease.
The growing Muslim population is also going to remain concentrated in regions that already have large numbers of Muslims – India, the Middle East, Africa and Indonesia.
Even with the studies’ projected growth, only about 10% of Europe’s entire population and just 2% of the population of North America would be Muslim by 2050.
In Europe the parallel effect of declining fertility in the region and religious switching will mean that many countries with a current Christian majority will, over the next few decades, become majority “unaffiliated”.
The fertility curve
Women’s fertility rates are central to the studies’ projections, with mortality, migration and switching as added filters.
The world’s Muslim population is growing faster than other religious communities because, on a global scale, Muslim women have the highest fertility rates (3.2 children per woman, compared with a world average of 2.5).
But when this is broken down regionally, vast differences emerge. For example, Muslim women in Indonesia have fewer children (two) than Muslim women in Nigeria (6.5), and fewer than Christian women (2.6) in Indonesia.
Where you are matters just as much as what you claim to believe.
Nechama Brodie is head of TRi Facts, the research and training division of Africa Check.