# How to Lie with Statistics

0

Alex Otti OUTSIDE THE BOX alex.otti@thisdaylive.com

“There are three kinds of lies:
Lies,
Damned lies and
Statistics”
Above were the words of Benjamin Disraeli (1804-1881). Disraeli became British Chancellor of the Exchequer in 1852 and after running and failing election four times, he was elected British Prime Minister twice in 1868, and 1874. He broke records amongst other achievements of being the first and only Jewish Prime Minister of England up to this day.

How to lie with statistics, on the other hand, is the title of a famous book authored in 1954 by Darrel Huff. Huff wrote extensively on what to look out for before we believe numbers churned out by statisticians. He argued that in choosing samples as against the entire population, the researcher could build in his own bias. If for instance, a researcher has an outcome in mind, he can select a group of respondents that will give him the answer he is looking for.

He went further to posit that the theory of averages is fraught with fatal danger. How, for instance, do you come by average age and average height? Which average would you be talking about because, the trio of mean, median and mode are all measures that describe averages? There is also the issue of little things that matter or don’t matter as the case may be, described more technically as the margin of error.

The statistician may choose to ignore little things that matter or play up little things that don’t matter depending on what results he intends to achieve. On visual presentation, Huff argues that graphs and charts are subject to manipulation, as there is a tendency to force a straight line ignoring some points that go off the line. He also talks about one dimensional picture, ignoring other dimensions and semi attached figure, which he illustrated thus, “if you can’t prove what you set out to prove, prove something else and relate it to what you wanted to prove in the first place”.

Huff continues by arguing that sometimes, the difference between real and nominal numbers are glossed over once it suits the statistician. Also differentiating cause and effect is a matter at the discretion of the researcher. He concludes by warning against what he called, statisticulating, a concept he used to describe the manipulation and bamboozling of people with figures and numbers.

It is a known fact that many people are uncomfortable with numbers. Some people will tune off once numbers come into the discussion. Others who are otherwise very intelligent naturally cave in as soon as the other party begins to quote figures. They therefore give up their right to subject the matter under discussion to rigorous interrogation. People who understand this, use it very well to their advantage in presentations and arguments. It is like Queens English to the man who is barely literate.

A story is told of a very rich chief from my part of the country who is barely literate. He was at a big launching in the village hall when a professor was called up to speak by the MC. The professor spoke in very eloquent English and ended his speech by promising to give his moral support to the community. His speech received thunderous applause from the audience. Our chief who by nature is very competitive asked the man sitting close to him in Igbo “moral shopot, ofusa okpogh ole” meaning “How much money is represented by moral support”? Our chief intended to donate more than the professor so he could get a better applause from the audience. So we must put the statistics to strict question to find out what they mean to our daily lives or better still, what they represent in monetary terms.

We had been inundated in the past by growth numbers which tended to exaggerate the state and performance of the Nigerian economy. We were told we had GDP growth rates of about 6.5 per cent per annum in the last four to five years, though we are now aware that those numbers have drastically dropped to about 2 per cent. Our foreign reserves grew from \$3.7 billion in 1999 on return to democratic governance to \$45 billion in 2007 and then to an all-time high of \$63 billion by the end of September 2008.

The war chest, as it is often called, was to decline to \$47.7 billion by the end of December 2009 and had since continued to drop from \$37 billion by the end of May 2015 to \$27.8 billion as at end of last month. While analysts and economists celebrated the great numbers, we failed to ask questions like what these mean to the ordinary man on the street. How did these figures impact jobs for our youths? How many roads did we build within the period? How many megawatts of electricity did we add to the national grid? What statistics seldom highlight is that “hot money” also known as foreign portfolio investment is normally part of foreign reserve and anytime exchange rates become unstable, hot money takes a flight and foreign reserve is usually the casualty.

Closely related to this is the rebasing of our GDP in 2014 from \$269.5 billion to \$510 billion. Commendable though this action may be, given that the figures we were using were very obsolete (1990 base year figures), there were some outlandish statements which in my opinion had no relationship with the realities on ground. With rebasing, we became the 26th largest economy in the world and the largest economy in Africa. Our per capita income rose from \$1555 to \$2688, ranking us 121st from 135th position amongst the 195 countries of the world. Without going into further details, someone should ask, how come the 26th largest economy in the world ranked number 121 in terms of average income per person? And still talking of per capita income, the easiest way to illustrate it is that you add Alhaji Aliko Dangote’s \$16.7 billion net worth to Mike Adenuga’s \$5.3 billion and Mrs. Folurunso Alakija’s \$1.61 billion with my own “toro” and divide by four, each of us would end up with an average or per capita income of about \$5.9 billion.

If life worked like that, yours truly would be on permanent vacation in Honolulu or Hawaii. So the point one is making is that this measure says nothing about income inequality. I know someone would be quick to point my attention to another measure referred to as “the Gini coefficient” which measures inequality in income distribution on a 0 to 1 basis, 0 representing perfect equality (where everyone has the same income and 1 representing perfect inequality where only one person has all the wealth and the rest have nothing). My response is that this measure only works in theory as it has nothing to do with reality and is subject to all the biases highlighted above. Further interrogation of the per capita Income would lead one to asking one more question.

Granted that we are now the largest country in Africa by GDP, how come we still rank No. 15 in per capita income in the continent coming below Equatorial Guinea (\$20,581), Libya (\$6,623), South Africa (\$6,483) and Egypt (\$3,304). The answer simply lies in the fact that there is a difference between absolute numbers and averages. While in absolute terms we have more in the basket than other countries, we also have a whole lot more people to share it. That is not to say that our large population is our problem after all, China and India have populations several times more than ours. The major problem remains that we are individually not adding enough value to GDP. It is instructive that to make the story continue to sound very sweet, our analysts have paid little attention to the issue of taxation. With the rebased GDP, our tax to GDP ratio dropped to about 8 per cent. That means that for every N100 of our GDP, only N8 comes from taxes. When compared to countries like Sweden at 46 per cent, France at 45 per cent, Burkina Faso at 11.5 per cent and Tanzania at 12 per cent, it would be clear that something is definitely wrong with our tax administration system. That would bring us to the issue of productivity. Nigerians are known to be very hardworking people.

But we also have a lot of people who are not working either by reason of not being able to find something to do or finding less than what they have capacity to do. This is what is described as unemployment and underemployment. Statistics differ on these. Nigerian Bureau of Statistics (NBS) says the unemployment rate in Nigeria as at January 2016 stood at 10.4 per cent, while underemployment was about 18.7 per cent. When you add the two, you would wind down with a number of about 29 per cent which is not far from some international analysts’ estimates of about 30 per cent unemployment rate for Nigeria. Whichever number is correct, we must agree that most of our youths cannot find jobs to do and that can be seen in the number of idle hands roaming the streets and many eventually becoming available for less than noble roles including robbery, kidnapping and other forms of violence.

Just imagine that we are able to engage them productively, our GDP numbers will shoot up and the per capita income will begin to make more sense. The other issue lies in interrogating the component parts of the GDP numbers themselves. There is no gain saying that if the GDP is geared towards commodities rather than final products and services, then we expose our economy to the vagaries of commodity price instability. This challenge becomes more pronounced where the commodities are traded in their primary form. So if we export crude oil rather than refined products, we end up exporting jobs in addition to the identified problem of not being in control of the price of the product. The same scenario applies to our other resources including tin, iron ore, coal, limestone, lead, zinc and agricultural products.

So, we should organise ourselves in such a way that we export finished or at least secondary or tertiary products after having added some value to the primary products. This should start with crude oil that accounts for 35per cent of our GDP and over 90 per cent of our export earnings. By so doing, we would not only be creating jobs for our people, but also saving on the foreign exchange for import of refined petroleum products which accounts for close to 40 per cent of allocation in the foreign exchange market. We will also eliminate some avoidable logistics costs in terms of insurance, freight, clearing, loading and offloading. These costs are incurred both when the primary products are being exported and when the finished products are being imported.

Like I had pointed out earlier, there seems to be some areas that are not so attractive and our statisticians seem to avoid measuring them. Nobody seems to measure how many people and tonnes of goods are moved by air, road, sea and rail in-country. This measure will help understand logistic capacity improvements and the level of interaction of our people. Education is another area that we pay little or no attention to. Not a whole lot of numbers come out to tell us about enrolments, exam scores, failure rates at different levels, dropouts and placements. We also don’t seem to put the issue of housing in the front burner. Of course we read about estimates of the number of additional houses required but beyond that we don’t seem to have reliable statistical information for planning purposes.

The concern about food security doesn’t also seem to have attracted the required attention. For instance, it is estimated that over 50 per cent of food is wasted to logistics and storage deficiencies. If this figure is correct, just imagine what will happen if we were able to provide solution to the challenge.

While we must continue the campaign for more reliable data that would help the country plan for our people, there are other numbers like inflation rate or Consumer Price Index, which may be useful for both time series and cross sectional analysis but must also be interrogated with rigour to ensure that they have a bearing to the existential conditions of the ordinary man on the street. Otherwise, we will end up looking very good statistically while our problems remain with us.