Charts have a starring role in humanity’s grand timeline.
The ancients tracked the dots in the night sky and plotted celestial patterns, creating their versions of cosmic PowerPoint slides, with clear calls to action for life on earth.
Through the millennia, our quest to understand the world has continued. The methods and metrics we use are now more specialised, more sophisticated. They provide new layers of insight and nuance.
During the Covid-19 lockdowns, for instance, seismic charts revealed a whole
In 1999, climate scientist Michael Mann and his research partners created a graph of average northern hemisphere temperatures over six centuries (it was later extended to 2,000 years). The graph resembled a hockey stick, with relatively stable temperatures followed by a sharp rise in the 20th century.
The graph became a powerful form of evidence, and a key visual representation, of global warming and modern climate change, helping shape public opinion globally and influence climate policy.
.
Keeling curve
More than 40 years before Mann’s hockey stick, climate-science pioneer Charles Keeling developed a new technique to accurately measure the concentration of carbon-dioxide in the air. In doing so, he also created the Keeling curve, in 1958, which is the longest uninterrupted record of atmospheric CO2.
The sharp upward swing of this curve provided the first significant evidence of the rising levels of CO2 in our air. In a single frame, it provided a snapshot of the impact of human activity on the planet, acting as an early indicator that we were entering the Anthropocene: the geological age we now find ourselves in, defined largely by human activity.
.
Lorenz curve
Developed by economist Max Lorenz in 1905, this curve depicts income or wealth inequality in a country, by telling us what percentage of total wealth is in the hands of a given percentage of the population.
In a world where everyone had the same wealth, the graph would be a straight diagonal line. In the real world, the curve is shaped like a bow, and the more curved it is, the more unequal the wealth distribution.
The Gini coefficient, which is a statistical measure of economic inequality in a society, is based on the Lorenz curve.
Comparing Lorenz curves across time is a good way to measure whether levels of inequality are rising or falling in a society. Incidentally, the richest 1% of the Indian population currently own 40.1% of the country’s wealth — which is more than at any point since 1961.
.
Laffer curve
Developed by economist Arthur Laffer in 1974, this is a theoretical representation of the relationship between tax rates and government tax revenues.
What’s interesting here is that a too-low tax rate is obviously not sustainable for a government. But a too-high one, it turns out, isn’t sustainable either.
Raising taxes beyond a certain point (the point varies, of course, based on the state of the economy and the phase it’s in) stifles entrepreneurship, commerce, spending — and compliance. And so, hypothetically, a government that imposed a tax rate of 0% would earn 0 in revenue; but so would a government that imposed a rate of 100%.
The trick, and the purpose of the chart, is to find that optimal rate that allows a government to maximise tax revenue without crossing that point at which the choking effect on spending and growth kicks in.
There are theoretical and empirical criticisms of the Laffer curve’s applicability, but it did influence Reaganomics (an era of tax cuts that did, at least temporarily, see interest rates, inflation and unemployment fall, boosting the US economy in the 1980s).
.
Bell curve
The bell curve (or the curve of normal distribution; indicated by the red line above) is a graph that shows how data is spread out around the average value. For example, if one measures the height of a large group of adults, most people will be close to the average height, with only a limited number being either very short or very tall.
The bell curve is the natural pattern of most of the physical world (most rocks are hard; most pillows are soft; most corn grows within a certain timeframe, to a certain height).
Most things, in other words, follow the 68-95-99.7 rule, which states that 68% of data points lie within 1 standard deviation (SD) of the mean, 95% lie within 2 SDs and 99.7% within 3 SDs of the average.
.
Fat-tailed distribution
In a normal distribution, there are very few data points at the extreme ends of the curve. These extreme values, or outliers, are critical in risk assessment, because one has to account for the likelihood of extreme events.
The fat-tailed distribution (a range of types are depicted above) has a larger proportion of extreme values at either end. One typically sees this graph when the natural order of things is disturbed.
The thing about the fat tail is that the unlikely event is still rare; but when it does occur, because systems were not built to accommodate it, the impact can be dramatic, unpredictable and long-lasting.
For two recent examples, think of the pandemic, and the global economic downturn of 2008.
Extreme weather events, which were once uncommon, are becoming more frequent, which is another fat tail we urgently need to engage with.
For more on this, there is the former options trader Nassim Nicholas Taleb’s engrossing book, The Black Swan: The Impact of the Highly Improbable (2007; no relation to the 2010 psychological thriller about a ballerina).
.
Exponential curves
Legend has it that the man who invented the game of chess was asked, by his king, to pick a reward. He said he wanted grains of rice, starting with one grain in the first square of his chessboard, with the number doubling in every square thereafter.
Why didn’t he ask for land or gold? Seems like a lost opportunity, people said.
Well, then as now, the human brain struggled to understand exponentials. We are not used to thinking of growth in geometric proportions.
Do the math and the chess reward would have amounted to about 18.5 quintillion grains, which is 52,000 times the rice reserves in all of the Food Corporation of India’s warehouses.
How can that be? Think about the million-billion comparison that is now virtually an adage: a million seconds is about 11.5 days; a billion seconds is about 31.7 years.
(Incidentally, after billion comes trillion, quadrillion, then quintillion.)
Exponential graphs are characterised by rapid growth (or decline). The rate of change is not constant, but it increases (or decreases) multiplicatively. The values grow slowly at first, then skyrocket.
Think of the jump in the number of Covid-19 cases in 2020. The investment mantra of “start early and benefit from compounding effect” is another illustration.
.
The Kaplan-Meier plot
This chart is used in clinical trials, to analyse chances of survival and other outcomes over time. It has a stepped shape, with each step representing an event (such as a relapse or a side-effect), making it easier to identify at what intervals certain events occur.
It allows researchers to compare how well different treatments work, through a visual inspection. With survival rates, for instance, if the drop is steep, events are occurring sooner and more frequently for that group.
The Kaplan-Meier plot is often used in conjunction with other metrics, such as the hazard ratio (relative risk measure of different treatments).
.
Heat maps
A heat map is a visualisation technique in which colour is used to represent data in a grid, with each cell’s colour or intensity of shade corresponding to a value.
A heat map provides a quick visual summary, and is intuitive. They are a staple of business reporting and are used, to great effect, to represent price changes of stock baskets, sales and distribution patterns, even employee performance.
One Subscription.
Get 360° coverage—from daily headlines
to 100 year archives.
Archives
HT App & Website