Figures Don't Lie, But Liars...

A Little Statistical Alchemy Can Solve Many Problems

Mark Twain offered his decided opinion on maths, stats, and "figures" on more than one occasion.  Famously, he opined that figures don't lie, but liars figure.  There are myriad other quotes, of course (lies, damned lies, and statistics is perhaps the most famous), a fact of which I was reminded when I came across an article this weekend in the New York Times.  The author, Claire Cain Miller, weighs in on one of the current flavours of the month - "Big Data" - and more specifically, how the current Comstock Lode of data of all sorts (patient records, consumer purchasing habits, web site visits) is opening vast new opportunities when coupled with ever-expanding computing power.

Ms Miller's thesis is that the explosion of data "ore" has created a large demand for analytically-minded people, quoting a recent Harvard Business Review article that "data science is the sexiest job of the 21st century"

Data scientists are the magicians of the Big Data era. They crunch the data, use mathematical models to analyze it and create narratives or visualizations to explain it, then suggest how to use the information to make decisions.

The article goes on to cite the growth in universities (Columbia, North Carolina State, USF) who offer a catalog of degrees and graduate programmes in data analytics.  According to Ms Miller, schools cannot mint people fast enough to meet the rising demand.

As a person who makes my living (and a very good one) as a data science "magician," I am of course sceptical of much of the hoopla.

Since much of my work deals with creating economic models, I am quite familiar with what is called (in econometric theory) "revealed preferences."  Simply put, we can say what we value as often as we like, but what we truly value is revealed by how we allocate our resources.

To anyone who lives (or has lived) in Silicon Valley, or works in a scientific company (I work in "Big Pharma"), think about who in your circle of colleagues or within your company really gets the big money.  In economics, there is never really a "shortage" of anything - if the thing exists, supply will meet the demand at the given price.  If there is a temporary insufficiency, prices will rise (quickly, perhaps) until either the supply is increased, or the demand decreased.

THAT, and not effluvia about "Big Data" will allow you to judge the validity of this statement.

Have wages in data analytics risen dramatically?  How do they compare to other similarly demanding fields?

I wrote this piece - tongue somewhat in cheek - nearly three years ago to the day about the value of an MBA. It was a reflection on another piece in HBR.  I'm reminded of it again, as I consider the puff piece on "Big Data."

It all boils down to one simple data point - a nugget of silver, if you will - hidden in the article: the typical salaries offered to data "magicians."

From the article:

North Carolina State University introduced a master’s in analytics in 2007. All 84 of last year’s graduates in the field had job offers, according to Michael Rappa, who conceived and directs the university’s Institute for Advanced Analytics. The average salary was $89,100, and more than $100,000 for those with prior work experience.

$89,100 is nothing to sneeze at; $100k is even better.  Both are well above the median income in the US.

But I suggest a proper yardstick would be to compare the salaries (starting, and with experience) against those of people with a master's degree in business - an MBA.  How many top-level MBA holders with "prior work experience" would consider a salary in the $100,000 range to be a serious lure?

From the excellent blog "Poets and Quants," I offer the following analysis:

Median Starting Salaries, MBA Grads, 2002-2012

Interesting. These are median salaries, not means, so it's not an apples-apples comparison, but essentially, a new MBA will start at about $1000 per year more than the graduate in the "sexiest" field of the 21st century - one where schools putatively cannot turn out grads fast enough.

I don't have any data, but I suspect that that difference gets larger as time goes by.  Take a quick look around the executive suite and see who is in there.  The guys running the company - and this includes the famous "Lean In Gal" (Sheryl Sandberg) are not "data magicians," despite working for tech companies.  Of course, Mark Zuckerberg, Sergey Brin, and Larry Page are all technical people, but I suspect that the lesson here is, if you want to be the big cheese at a tech company, either actually create the product, or get a marketing degree.

As blogger Steve Sailer commented, analysing data is fine work, and you can do well with that if you've an analytical mind.  Just don't expect to get paid like the sales guy.