From Watts Up With That?

Comment by Kip Hansen — 14 July 2023

Many of you probably follow Statistica – I know I get emails almost daily with some visual presentation of some data set from somewhere. Over the last year, I have begun to suspect that some of the illustrations and presentations have political motivations (meaning: seemingly produced for propaganda purposes for some group – or maybe just representing the group bias of the employees, owners, or managers at Statistica).  This commentary is not about that issue, however.  [If readers have specific examples of that aspect of the “Statistica issue”,  I would appreciate a note in comments or an email to my first name at i4.net].

Here I just want to look at what has become a run-away bad habit in the field of science and data.  Here’s the illustration whose purpose is, according to Statistica, to “Empower… people with data”:

This illustration comes with a caption by Felix Richter: 

“According to United Nations’ latest projections of global population, India surpassed China as the world’s most populous country in April 2023. Having gradually closed the gap to China from more than 200 million people in 2000 to little more than 10 million in 2022, the UN Population Division predicts India’s population to reach 1,429 million in July 2023, surpassing long-time leader China by 3 million people.

In recent years, China’s population growth has slowed down notably due to its one-child policy before reaching an inflection point in 2022, when China’s population declined for the first time since 1961, when three years of famine had decimated the country’s population.

For India, which is currently expected to continue growing until the 2060s, its new position as the world’s most populous country will come with a new set of challenges, both domestically and internationally. These challenges include providing access to healthcare, education, and employment opportunities to an ever-growing number of people, all while finding its role in the global political and economic landscape.

Looking beyond India and China, the UN predicts a continental shift in population growth over the next few decades. With Europe’s population already declining and Asia’s and Latin America’s growth expected to turn negative in the 2050s, Africa is set to be the largest driver of global population growth for decades to come. By 2100, five African nations are expected to join India, China, and the United States among the world’s 10 most populous countries, with Nigeria projected to reach a population of half a billion before 2080. This demographic shift at the global level will require new approaches to managing resources, promoting sustainable development, and addressing issues such as poverty, inequality and access to healthcare and education.”

In my view, there are obvious outright errors in the data presented – if anyone wishes to chase those up, give us the scoop in comments.  One quick example – readers are encouraged to ferret out others – is the projection (allegedly by the UN) of China’s population in 2100.  The chart shows that “The UN Projects” that China’s population will be reduced to ½ its current level in the next 75 years.  Yes, dropping from 1,426 million to 767 million.  Even Chairman Mao couldn’t get rid of that many people that fast:  “In 1958, he launched the Great Leap Forward that aimed to rapidly transform China’s economy from agrarian to industrial, which led to the deadliest famine in history and the deaths of 15–55 million people between 1958 and 1962.” [ wiki ]  How is China going to get rid of over half a billion people in just 75 years?

Nigeria is shown with a projected population in 2100 of over 500 million (half a billion) – a doubling of current population.  That would be 1400 persons per square mile – or 541 persons per square kilometer.  The current population density in the United States is 94 persons per square mile.  I guess it is possible but I believe that national resources would be stretched exceedingly thin at that population level.

Those who wish can dig into the U.N.’s report (here’s the link to the Summary again).  Interesting but not my issue today.

My complaint is this:

We do not and cannot have data about the future.

Yes, I know Statistica labels the future data as projections….and blames those projections on the United Nations….but it does not do anything to visually distinguish between the actually counted (calculated, estimated, etc) data – real data from the past and of the present — and the projections from the U.N.   

Now, I don’t want to disparage the general public but it is my understanding that when experts speak, particularly using illustrations, images, graphs, that the general public simply accepts these statements (especially pictures – visuals) as true.

And, though I am loathe to bring up the issue once again, there is no uncertainty shown in the “projections” despite that fact that some of the projections (China, for instance) are very alarming.

What is the uncertainty claimed for the 2100 projections by the cited UN source?

“III. Long-range population projections to 2100  —  Long-range population projections are highly uncertain, especially for high-fertility countries still in the early stages of the demographic transition.”

In this case, Statistica has failed its customers (and the general public). Instead of empowering the pubic with knowledge, they have misrepresented HIGHLY UNCERTAIN projections about the future as if they had the same veracity, trustworthiness, as historical data.

The original UN report (duplicate link – .pdf) shows how this error can be avoided with illustrations similar to one in the U.N. report here on life expectancy.  The data is shown in solid colors and the projections of the future are 1) set off by a vertical dashed line, demarking a new graph area labelled clearly “Projections”  and 2) the projections are showed in dashed lines giving a visual effect of uncertainty (less solid). 

That’s a lot of words to say a thing that is so simple that we often overlook it:

There is no data about the future.

There is no evidence about the future.

Not yet, at least.

# # # # #

Author’s Comment:

Please spare me the nonsense that “We know the Sun will rise tomorrow” as an example of having data about the future. (Yes, this has been used by readers here in the past).

And, admittedly, this is somewhat of a nit-pick – but it is an important nit-pick – as the scientific principles of what is data and what is evidence  is so often violated. 

In the real world, the one in which we all live, there are children wracked with grief because they have been told repeatedly that the world is ending, that they will have no future.  They are told that “the scientific data shows this.”   A bigger lie could not be told.

A reasonable person can make reasonable decisions about his immediate future – that’s how we manage to live from day to day.

I will decide whether to go sailing this afternoon based on local weather conditions and my favorite weather program’s  guess about the still-hours-away future.  So far, those guesses encourage me to do so. Have a good day.

Thanks for reading.

PS: The weather prediction for wind on the local water turned out to be  wrong – the wind increased to a snotty 15-20 kts, variable and gusty, with a bit of rain mixed in instead of the predicted decrease to a pleasant 8-10 kts.