Tuesday, 18 June 2019

World population figures are often inaccurate

I did a search on Google to find out what the world’s population is and came up with a page called World Population Review that features what it calls the “US Census Bureau world population estimate”. On the page along with the figures themselves, you can find this:
Although the [world population figure] given above seems very precise, it is important to remember that it is just an estimate. It simply isn't possible to be sure exactly how many people there are on the earth at any one time, and there are conflicting estimates of the global population in 2016.
It looks like Google relies on the same source for its country population statistics. If you type “How does Google get its population figures” into Google you won’t get a straight answer but under the global population figure Google publishes it references the World Bank as the source. Elsewhere on the same page it references both the World Bank and the US Census Bureau. Google also provides population figures for individual cities around the world. For the figure for Amman, it references the World Bank and the United Nations as sources for its information.

On the World Bank website it doesn’t show individual city population figures, however, but it does show individual country figures. The sources used to arrive at the figure for Jordan, for example, is described as follows:
( 1 ) United Nations Population Division. World Population Prospects: 2017 Revision. ( 2 ) Census reports and other statistical publications from national statistical offices, ( 3 ) Eurostat: Demographic Statistics, ( 4 ) United Nations Statistical Division. Population and Vital Statistics Report (various years ), ( 5 ) U.S. Census Bureau: International Database, and ( 6 ) Secretariat of the Pacific Community: Statistics and Demography Programme.
It seems that the World Bank uses the same footnote for all countries, because it’s hard to see the relevance of item number six, given above, to a population calculation for Jordan. The same text appears, furthermore, on the page with the population figure for Indonesia and for the page with the figure for Japan. Further down in the notes for these country figures given by the World Bank is this:
Current population estimates for developing countries that lack (i) reliable recent census data, and (ii) pre- and post-census estimates for countries with census data, are provided by the United Nations Population Division and other agencies. The cohort component method - a standard method for estimating and projecting population - requires fertility, mortality, and net migration data, often collected from sample surveys, which can be small or limited in coverage. Population estimates are from demographic modeling and so are susceptible to biases and errors from shortcomings in both the model and the data. 
Which is not very encouraging, to be frank. For its purposes, the United Nations’ Population Division has a website and there is a page on it dedicated to explaining “methods for estimating population size, mortality and fertility with the emphasis on situations when data from censuses and civil registration are insufficient or unreliable”. The page contains access points to a number of different “manuals”. For example, Methods for Estimating Total Population for Current Dates, is dated 1952 and contains a number of different sections but it seems to me to be completely outdated. I think it is worthwhile wondering why a more recent set of guidelines has not been provided by this organisation. I would be very alarmed to think that in 2019 the UN is using the same calculation methods it used half-a-century earlier! One chapter however (let’s plug on regardless) provides this as a guide to the limitations of accuracy (in 1952, for goodness’ sake):
Most current estimates of population have two components: (a) a "base figure", that is, a count or estimate of the population at a previous date, and (b) a "time adjustment", that is, an allowance for population increase or decrease since the previous date. The accuracy of the estimate, of course, depends on the accuracy of both components.
So, when it comes down to it you are inevitably led to the conclusion that figures presented to people doing searches on Google are only as good as the relevant authorities in each target country.

Now, if Google is asking the Australian Bureau of Statistics about population figures for Australia or for Sydney then I can safely say that they will be pretty accurate even though, due to the kinds of visas some people have who live in, say, Sydney, it’s very difficult to know for sure how many people are actually resident in the country at any one time. Not all Sydney residents are citizens and the visiting population (especially due to the large number of overseas students living in Sydney) is considerable. But at least you can believe that there is a fair level of accuracy implicit in the given figure. If you drive around Sydney you will get a sense, therefore, of what a city with 5.5 million people feels like, whether you use public transport or whether you use a car to get from place to place.

It’s this sort of experience that makes me suspect the figure that Google provides for Amman, the capital city of Jordan. The figure they give is four million. But even at rush hour in the morning it only takes about 10 or 15 minutes in a taxi to get from the downtown area (where many of the shops are located that locals use to buy things like food and clothing) to the city’s periphery. In Sydney, even during quiet times, it takes this long just to drive from one point inside the metropolis to a suburb five kilometres away. At rush hour it can take much longer.

Even given the different traffic densities it is hard to see how you could fit four million people into a city of Amman’s size. There are very few (if any) high rise apartment blocks there and most people live in concrete buildings with two or three storeys.

If the authorities in Jordan are loose with their population figures there must be a reason for them to be so. It might have something to do with geopolitics. Like a peacock that fans out its tail when it’s approaching a peahen, Jordan might be using its population figures as a from of display. Other kinds of birds use this sort of display in order to make themselves look bigger when they are challenging another member of the same species for some reason: like a seagull standing up straight and walking very fast and making a cawing sound when it is trying to scare competitors away from a potato chip that is lying on the ground and that it wants to eat.

You can dream up any number of reasons why a country like Jordan, which regrettably has a tank museum (I’m not making this up to suit my purposes), might want to make itself look bigger than it is in fact. In addition to the tank museum, Jordan’s government does things like this (see photo below).

When I was in Amman this piece of artillery was visible placed in the Boulevard Abdeli Mall. The shopping centre is in the “nice” part of town and it contains restaurants and boutiques. The cannon on display had a sign next to it that pointed out (in English and Arabic) that the equipment on show had been used in the 1940s to announce the end of fasting on days during Ramadan. It was Ramadan when I was in Amman and nowadays there are still cannons used to mark the point in the day (around 7.30pm) when people, according to the dictates of Islam, are once again permitted to eat. But the installation had other uses as well, apart from showing how the end of fasting was brought to people’s attention in the recent past. The mannequins on display, with their distinctive Jordanian red-and-white headscarves, had a clear purpose.

The military plays a visible role in many such countries due to the realities of inter-country politics. Just across the river from Amman, of course, sits Jerusalem. Israel, for its part, publishes its own population figure (8.5 million) and it seems credible given the small size of the place and given the density of the population in the parts of Jerusalem I visited on my recent trip to the Middle East.

My misgivings about the population figures given for Amman or for Istanbul (15 million, again according to Google) unfortunately scale effortlessly to the whole world. So the current estimate published by Google (around 7.5 billion) is probably bunkum. Going by what I saw in the Middle East, it’s probably in fact something more like three billion but, like Google’s estimate, this figure is actually just a random figure that has been pulled out of the air. As far as food goes, from what I saw people living in the Middle East have nothing to worry about. There is plenty of good, cheap food available for a price, although in some places if you are a tourist it is likely that they will charge you more than they change a local for the same items.


marcellous said...

"If you drive around Sydney you will get a sense, therefore, of what a city with 5.5 million people feels like"

... where there is a very high rate of car ownership, low density of housing (with predominantly "nuclear" families occupying housing) and where the city is organized in a wheel-and-spoke fashion with highly centralized employment in the CBD.

"My misgivings about the population figures given for Amman or for Istanbul (15 million, again according to Google) unfortunately scale effortlessly to the whole world. Going by what I saw in the Middle East, it’s probably in fact something more like three billion"

Matthew, you can't be serious. {?]

India and China, both countries which have tried to limit their populations so presumably are not interested in overestimating them, reportabout 1.3 billion each. Allow 500 million for the EU (at this stage I guess this includes the UK!) 300 for the US and that's well over 3 billion before we take into account the rest of the Americas, Africa, the rest of Asia and the former USSR.

The figures may be rubbery (hat tip to the late Philip Lynch) but I don't think they can be that rubbery.

Also, I don't see what is so out of date about a methodogy, even if proposed in 1952. It would be figures, rather than methodology, which would date, surely?

Matthew da Silva said...

I'm perfectly serious but you are free to believe whatever the UN or the World Bank, both of which freely admit that their figures are complete fabrications, tell you. They say on their websites that their figures are not accurate and they explain how they reach their totals. None of this information can credibly give a Google user confidence in the figures that are so baldly thrown about all the time in the media.