Pondering Analytics Of Cholera On A London Walk

It’s peak travel season in London, and the crowds are even larger than usual as the city celebrates the 200^th anniversary of victory at the Battle of Waterloo with dozens of special events. On Friday morning, Registered Guide Hilary Ratcliffe of London Walks leads a group of international tourists on a tour of “Unexpected London.” Ms. Ratcliffe’s tour includes something really unexpected: a model example of analytics in action, including lessons about why the same math can lead either to meaningful change, or no results at all.

At the start of the walk, Ms. Ratcliffe tells us about Edwin Chadwick, who oversaw Britain’s Poor Law Board during a period when cholera outbreaks were a significant problem for London’s people.

Chadwick published a report (Report on the Sanitary Condition of the Labouring Population of Great Britain) asserting, among other things, that diseases such as cholera were the result of filthy living conditions. He misunderstood the true cause of cholera, believing it to be a result of breathing toxic air, but he was correct in making the connection between a dirty environment and disease.

Analytics alone does not explain patterns in data. Analysts who find potentially useful patterns test to verify that they occur consistently, but explanation is often left to other types of experts. Chadwick’s model was wrong, but still useful. Cholera was, indeed, related to filth. Still, the British government did not immediately invest in better sanitation facilities for London’s affected communities. Six years passed before Chadwick was able to get momentum behind a movement to eliminate cesspools.

Modern data analysts often face similar frustration. It’s one thing to produce and excellent analysis, and another to get management to use it as a basis for action.

Over a decade later, in 1854, physician John Snow added to the understanding of cholera. He analyzed the progress of one cholera epidemic in a manner that was unconventional at the time; he mapped it, as part of a larger statistical study that is now known as a landmark of epidemiology. His maps suggested a connection between cholera and water sources, a key to learning that the true cause of cholera is a water-borne microbe.

But it was not until 1859 that work began on an improved sewage system for London. Why, when the best analysis of the time indicated that better sanitation was needed to eradicate cholera, did it take 17 years to get serious action underway?

Back to the guide for an explanation. Early cholera epidemics affected London’s poor, and only the poor. When later epidemics took hold, the victims included wealthy Londoners, even members of Parliament. When the problem directly affected them, they finally took serious action.

Analysts today, whether they call themselves statisticians, physicians, data miners, data scientists, or by any other name, still struggle to make the case for putting analysis into action. Perhaps this historical example give a clue to motivating managers. When they can see how the problem affects them personally, they will take action.