Monday 19 October 2009

Why Let Facts Spoil the Narrative?

I've just finished reading a very shakey due diligence report, in which one of the key questions to review was the impact of the recession on the company under investigation.  The report author covered the beneficial effects of recession on services (comparable to those of the target company) that people mainly use in their own homes.  The report talked about increases in satellite and cable TV subscriptions, growth in Dominos deliveries, and the trend to "staycations", and implied that, as a result, everything would be OK.

A question kept coming to my mind as I ploughed through this nonsense.  This question was: "But the recession has been happening for about a year - why don't they just look at what's been happening to the company?".  The analyst could have looked at sales, customer churn, average customer value, new sign-ups.  And they could have looked at them before, during, and after recession (now we're coming out of it for a short while).  They could have compared the company's performance to changes in disposable income, or employment, or interest rates, or consumer confidence.  They could have just looked at whether they rose or fell.  The data was available and staring them in the face; but they didn't look at any of the vast array of facts at their disposal, and instead indulged in this staycation narrative.  I'll let you guess whether the facts confirmed, contradicted, or made irrelevant, the report's conclusions .

I kept asking myself why any sane analyst would display such disregard for information.  After some reflection on examples of similar behaviour, here's my conclusion: given the choice between some compelling facts and a compelling narrative, people will often prefer the narrative.  From everyday observation, there are legion examples of people ignoring or skimming over facts that might get in the way of a good story.

This preference can be, literally, fatal; and if you'll indulge a longer-than-usual post, I'll illustrate it with a historical example.

A nineteenth century physician called Ignaz Semmelweis analysed the high incidence of childbirth mortality of Puerperal fever at one of the wards of Vienna General Hospital.  He noticed that Puerperal fever was high in wards where the same doctors also conducted post-mortems, and showed that if doctors washed their hands with chlorine solution after working with cadavers, then Puerperal fever incidence declined dramatically.

His data is hard to challenge:



Unfortunately, Semmelweis's facts didn't fit the narrative of the day.  Prevailing theories of health related to the balance of the four humours of the body, and the role of "bad air" in the spread of disease.  In fact, his implication, that lack of cleanliness in the surgeon was a cause of the disease spreading, was considered insulting to the gentlemen who administered medicine and surgery.

Semmelwies was roundly criticised, and his observation and recommendations were dismissed by the mainstream, despite their obvious life-saving results.  It was only after Pasteur's work into germ theory became accepted 20 years later that the establishment embraced the findings of, the by then dead, Semmelweis.

So coming back to my point.  There will always be an accepted or acceptable narrative to explain anything, be that the four humours of nineteenth century medicine, or the various dubious adspeak marketing theories we hear today.  We can pretty much guarantee that by blindly following the narrative, we will be proven as gullible, closed-minded and wrong as those olden-day physicians.  Alternatively, we can ignore the narrative for a moment, and just have a quick  look at the facts...

Copyright Latitude 2009. All rights reserved.


Related links.

On Ignaz Semmelweis
http://en.wikipedia.org/wiki/Ignaz_Semmelweis#Ideas_ran_contrary_to_established_medical_opinion

On truth, bias and disagreement
http://www.econtalk.org/archives/2009/03/klein_on_truth.html



Latitude Partners Ltd
19 Bulstrode Street, London W1U 2JN
www.latitude.co.uk

No comments:

Post a Comment