Re: [Marxism] Debunking Third World myths

["Joaquin Bustelo" <jbustelo@xxxxxxxxx>]

Mehmet writes: 

"This debate started when Paula posted a clip in which Hans Rosling
demonstrates a "detailed" statistical data to prove that, contrary to the
popular belief free market economy serves to the benefit of humankind
whether in "so-called" third word or in the developed countries. Although
some of the data that he presented are fairly idiotic like child mortality
rate and life expectancy, since a country certainly benefits from some of
the scientific developments unless she is located in another planet (with
the exception of Russia which inexplicably underwent a decline of life
expectancy in 90's), let's assume that these statistics are correct. I have
no doubt about the integrity of Mr. Rosling and his data."

Mehmet is referring to this talk:
<http://www.ted.com/talks/hans_rosling_shows_the_best_stats_you_ve_ever_seen
.html>. 

It's only 20 minutes, and actually you only need to get to the main economic
section less than halfway through the talk to get the gist of it. 

The talk pretends to be basically a promo for database graphing software
(which is here: www.gapminder.org), and a similar talk was given at Google
by Rosling's son (Google ended up buying Rosling's software, and it's pretty
good software). But the use Hans Rosling made of it in his TED presentation
is another matter altogether.

I looked at Rosling's TED talk when Paula first posted the link and had a
chuckle because it is a purely ideological construct that relies for its
effect on innumeracy and constant patter, like a three card monte street
corner hustler. But I hadn't planned on posting on it, figuring our kind of
people would see through it. Mehmet's post shows that in his gut, he knows
the conclusions of this Rosling guy can't be right, but can't really
articulate why, and even gives Rosling the benefit of the doubt.

But Rosling deserves no such consideration. He manipulates the presentation
of his data to serve his ideological ends. He does this primarily by
constructing graphs in such a way as to minimize the difference between
richer and poorer countries, and magnify differences among Third World
countries. His primary tool is using logarithmic scales for various
measures, especially of income and wealth. 

Logarithmic scales make the distance between $5 and $500 the same as between
$500 and $50,000. In other words, the distance between two data points
depends on how many times the second one is in relation to the first. There
is the same distance between 1 and 2 as there is between 2 and 4, 4 and 8, 8
and 16, and so on. Even the distance on a graph between 1,000,000 and
2,000,000 is the same as between 1 and 2. It is the ratio of the two numbers
that determines the distance, not at all absolute amounts.

In the Third World over the past four decades there's been a tremendous
decline in subsistence farming, with huge percentages of the population
moving into the money economy. In theory even the economic value of
subsistence production is taken into account in GDP figures; in practice
this is very difficult to do. For example, the house of a peasant family is
worth next to nothing in such an economy, but that is often not the case of
a comparable urban dwelling. The net result is that while some of the
increase in GDP in Third World countries represents a real increase in
living standards, a significant portion of it is illusory in terms of the
real standard of living. 

He takes the increases in the money value of GDP in a couple of different
forms --income per day and GDP per capita-- and does his time based
evolutions showing a "narrowing" of gaps between different countries and
groups of countries. For example (this isn't from his data, it is just to
illustrate) country A goes from $1/person/day in income to $10/person/day,
country B from $100 to $200. Rosling's logarithmic scales would show the gap
between the two countries NARROWING, they would move closer together on the
money axis of the chart, when in FACT the gap in income between the two
countries has almost doubled! (From a $99 difference at the outset to a $190
difference.)

He supplements this manipulation with non sequiturs presented in the most
emphatic way as if any child could see it. For example, at one point he puts
up a pie chart of world income. "This is 100% of the world's annual income
and the richest 20% take out of that about 74% and the poorest 20%, they
take about 2%. And this shows that the concept developing countries is
extremely doubtful. We sort of think about aid of these people here
[pointing to the rich] giving aid to these people here [pointing to the
bottom 20%]. But in the middle we have most of the world's population, and
they have now 24% of the world's income."

But what does that "middle" consist of? People with per capita daily incomes
ranging from $1 to perhaps $15 or $20 (this is a guess, the logarithmic
scale makes it hard to interpolate). Moreover, there is a very pronounced
peak towards the lower end of the $1-$10 segment, perhaps at $2 or $3 a day.
If a linear scale had been used, Rosling's fairly compact "middle" would
disappear, and the gigantic gulf between rich and poor become evident.

But in addition to that, how does this make the concept of developing
countries "extremely doubtful"? In reality, it tells us nothing about the
concept. From what that chart says, the income distribution in every country
could be the exactly the same, or all the rich people could live in one or a
handful of countries and everyone else in the rest. 

But for purposes of that presentation, Rosling needed to pretend he's
already undermined the idea of big difference between countries and regions
because there simply is no way to hide the big differences between countries
and regions that are evident in the next section of the talk. 

Thus Rosling's subsequent presentation of income by groups of countries
shows people in the OECD countries typically having incomes between $10 and
$100 per person per day. There is one peak in the chart there, and another
way to the left, at the low income end, perhaps $2 a day. His animations
over time show the distance between the "rich country" hump (i.e., incomes
enjoyed by the largest number of people in those countries) and the poor
country hump decreasing between 1970 and 2000 and the gap is projected to
decrease even more by 2015. 

But what that means in concrete terms is that the poor "hump" (the income of
the greatest number of poor people) went from perhaps 70 cents to two
dollars, whereas the rich people "hump" went from around $10 to perhaps $15.
The poor TRIPLED and the rich only increased by half, and that is the change
the logarithmic scale shows, the poor moved up much more than the rich. 

But the dollars-and-cents TRUTH is the poor went up by $1.30, the RICH by
TRIPLE that amount. Before the gap was nine dollars and change, and now it
is $13.

How does Rosling justify the swindle? "Of course it's a logarithmic scale
here, but our concept of economy is growth with percent, we look upon it as
a possibility of percentile increase." 

But does that make any sense when you're comparing per capita GDP and daily
incomes? Of course not. Rosling himself finds he can't address the subject
without using actual levels, money figures, but what his graphics and
presentation do is change the subject from the AMOUNTS to the relation
between the RATES OF CHANGE in per person GDP or income of different
countries and groups of countries.

Something very similar occurs with his health data. He plots child survival
to age five on a weirdo percent scale. The bottom division is from 50% to
70% and is just as big as the top division of 99.5% to 99.7%. To compound
the confusion, his other axis are things like logarithmic GDP per capita.

If he'd use the more conventional way of presenting morbidity and mortality
statistics, incidence per 1,000 or 100,000, in this case, births, and GDP,
we'd see that the poorer countries often have double, triple, quadruple, or
even ten times the number of deaths of children under age five than rich
ones do.

Now, you don't have to take my word for it, that Rosling's way of presenting
it in essence is a distortion of the data. Go to gapminder.org, click on the
"gapminder world" link, and there you will have one of the main graphs used
by Rosling, life expectancy versus income. On the bottom right, change the
scale for income from logarithmic to linear. 

You will see plainly that most countries in the world are bunched on the
extreme, low GDP, left of the graph. Switch back and forth between
logarithmic and linear scales, and pay close attention to the numbers on the
bottom, and you will se how using the logarithmic scale spreads out those at
the low end and scrunches together those at the high end, making it seem
like there is just a continuum of countries at different income and health
levels, whereas the linear scale shows the truth: there's a relative handful
of rich countries and tons of poor ones.

And then with linear scales select a couple of country's checkboxes, say,
Brazil and the United States, and play the time evolution from, say, 1940,
1950 or 1960 to today, and see how the income gap has become larger and
larger. 

Now, I'm not surprised Paula was taken in by this talk, as it fits perfectly
her preconceptions, and she's proved to be especially resistant to facts. 

But no one else should give any credence to Rosling's graphical quackery.

A final note: contrary to what Mehmet says, I don't believe there is nothing
"inexplicable" about the decline in Russian life expectancy in the 90's.
That was the result of the restoration of capitalism in Russia, the
elimination of the "safety net" of the social gains by working people that
had resulted from the 1917 Revolution. 

Joaquin


________________________________________________
YOU MUST clip all extraneous text when replying to a message.
Send list submissions to: Marxism@xxxxxxxxxxxxxxxxxxx
Set your options at: 
http://lists.econ.utah.edu/mailman/options/marxism/archive%40archives.econ.utah.edu