Johan Fourie's blog

I'd rather be a comma than a fullstop

An ode to demography

with 5 comments

Global population growth according to the CIA World Factbook

Global population growth according to the CIA World Factbook

In his latest book ‘The Last Afrikaner Leaders’, Hermann Giliomee notes three ‘badly flawed assumptions’ of the Apartheid government: the carrying capacity of the homelands, the historic claim to the land, and the growth of the black population. While the first two have solicited much attention, the last one has, surprisingly, been neglected. In those years, the study of demography was still at an early stage. Jan Sadie, professor in Economics at Stellenbosch and South Africa’s pre-eminent demographer, greatly underestimated the growth of the black population. This was an important error. Given his calculations, Sadie believed that complete separation between the races was still a possibility. But what if he had known that the black population would grow more than five times faster than his projections? Would integration – however skewed and unfair – not have been the only option available to Verwoerd and his followers?

Instead of discussing this counterfactual world, the point here is that we tend to underestimate the power of demographic changes. To explain changes to GDP per capita, economists focus mostly on the GDP rather than the per capita. But population growth rates change rapidly and the impact of the demographic dividend - a rise in the rate of economic growth due to a rising share of working age people in a population – is often neglected. Perhaps one cause of the high South African unemployment rate is the large increase in the number of working age individuals that was born at the end of the 1980s? Or perhaps the slower growth in Europe is not only a consequence of rising debt, but because there are fewer people to pay it? Or perhaps the lower rates of crime in New York – as Freakonomics authors Steven Levitt and Stephen Dubner argued – is due to legalised abortion two decades earlier, removing a subsection of the population that would have had a higher probability of committing crime? The Census 2011, for example, suggests that South Africa’s fertility rate during the 2000s declined dramatically. What are the implications for the number of kids going to school? And what if the unemployment rate begins to fall in 2020? Do we congratulate ourselves for a thriving economy, or realise that it’s simply the result of a smaller denominator rather than a growing numerator?

Accurate demography statistics is also critical if we to understand development over time. In compiling national income statistics for various twentieth century African countries, Morten Jerven at Simon Frasier University noticed that for most African countries, official population estimates are scattered and often don’t add up. Colonial administrators most likely had to minimise costs and censuses were thus often little more than guesswork. And after colonisation, the new independent governments often had an incentive to overestimate the population numbers of their own constituencies, boosting population numbers far beyond their actual size. Morten and Ewout Frankema is busy with a large project to find archival sources to recalibrate African population sizes for the twentieth century. In the latest edition of Economic History of Developing Regions, Leandro Prados de la Escosura use these and other data to calculate new GDP growth estimates for twentieth century African countries.

Often, demographic indicators are all we have left of the past. In a new paper, Jeanne Cilliers and I use a new genealogical dataset to calculate demographic estimates of the settler population in the Cape Colony (Working Paper here). We find that the settlers often lived longer than many of their eighteenth century counterparts (i.e. the poor) in Europe – and that the average life span increased from around 40 years to 54 years during the nineteenth century, probably due to higher incomes and better medicine (which, incidentally, is higher than the life expectancy of South Africans in 2010 – 52.1 years). It also appears as though this rise occurs a decade or two earlier than the discovery of diamonds in the South African interior – which is when incomes probably began to increase. Demography is also about much more than incomes: household size was also high for most of the eighteenth and nineteenth centuries (about nine children per male), falling to four by the 1930s. In more recent work, Jeanne also investigates the marriage patterns of Cape settlers, showing that women’s age of marriage increased from about 20 at the beginning of the eighteenth century to around 26 in the 1840s. In contrast, men’s age of marriage declined from a high of 28 at the start of the period to be on par with women at the end. Such evidence has important implications for understanding household size, fertility and female education, for example.

It’s surprising that the field of demography doesn’t receive the attention it deserves. A better understanding of our future – and our past – depends on it.

Written by Johan Fourie

January 20, 2013 at 23:07

5 Responses

Subscribe to comments with RSS.

  1. [...] provinces, for example. As Tom Moultrie, UCT demographer, notes in a remark on one my previous posts:  ”If you were an MEC in the Northern Cape, how would you plan rationally, if StatsSA’s [...]

  2. Tom Moultrie

    February 11, 2013 at 09:50

  3. Your assertion about the Census 2011 showing rapidly falling fertility since 2000 is factually incorrect. Look at the population structure aged 0-14. These represent the estimated survivors of births in the year before the census, the year before that etc etc. The data show that fertility fell, by around 20%, and then rose again post 2007. This makes no sense at all given what we know about fertility decline in the country since the 1950s.

    Further, if you were an MEC in the Northern Cape, how would *you* plan rationally, if StatsSA’s own estimates suggested that the population of your province lay *somewhere, with 95% certainty* between 984 000 and 1.307m? (That’s a 14% range either side of the point estimate of 1.146m!).

    Tom Moultrie

    February 11, 2013 at 09:49

    • Thanks for pointing out these errors, Tom. I was just citing official Census documents, but will clearly have to double-check next time. I presume these errors are related to the dismissal of StatsSA personnel? (http://mg.co.za/article/2013-01-27-stats-sa-officials-censured-for-totally-wrong-results)

      Johan Fourie

      February 11, 2013 at 16:57

      • Mostly unrelated. The stories that are coming out are – largely procedural. Although we suspect that the quality of the finally released data may have been compromised by the delays that resulted. The big thing is the uncertainty, and the age distribution of the population under the age of 20. If true it would certainly challenge most of what we have understood for the last twenty years. But those concerns are all in the public domain.

        Tom Moultrie

        February 11, 2013 at 18:21


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 348 other followers

%d bloggers like this: