Today, like it or not, all of us live in the age of the algorithm. Some fairly big decisions in our life--be it a choice of university, getting a home mortgage, a reasonable car loan and the cost of our health insurance are now made by mathematical models, not human beings.
Banking titan J.P. Morgan testifying before congress over 110 years ago stated that credit was something that a person brought with them when they applied for a loan. It was as much about character as it was about financial creditworthiness. Not so any longer!
A model, often grouped under the term “Big Data,” determines whether a loan is approved and what terms you will pay. Early on, many of us thought that this would be a wonderful situation as bias, cronyism, and discrimination would be eliminated.
In a recent book entitled WEAPONS OF MATH DESTRUCTION (Crown, 2016), Cathy O’Neill says that the reverse is happening. The subtitle to the book is “How big data increases inequality and threatens democracy.”
Now, be aware that this book was NOT written by some bomb tossing emotional left winger. Ms. O’Neil has a Harvard PHD in Mathematics and has served as a quant at the prominent hedge fund D.E. Shaw. Disillusioned with the world of finance, she has shown involvement and sympathy for the Occupy Wall Street Movement. What she does, for sure, is expose the dark side of Big Data.
Her principal thesis and it is a very well reasoned argument is that many people are stuck where they are in America today because Big Data has, to a certain degree, locked them in a life of mediocrity. Smart kids from certain zip codes will not get approved for student loans, or may pay significantly higher rates for car insurance and auto loans.
Hourly workers are sometimes victims of modeling according to Ms. O’Neil. Take someone working at a fast food restaurant or even a casual dining establishment. Employees sometimes get their hours the day before the next day’s schedule is announced. This becomes a child care nightmare for many. Others close a store and then open it the next morning but do not know that until midday. The firm, often a franchise, has a sophisticated algorithm that determines the optimum use of employees. Great for them but it messes with the lives of many staffers. When Starbucks was alerted to this issue, they began publishing workers hours a week ahead of time which was a great help to many.
On the plus side, Big Data is a marvelous thing to those of us who are affluent. We can get great deals on airfares, sales on high ticket items, or reviews of hot new and reasonably priced restaurants that many Americans would never be able to afford. Amazon knows our every move in purchasing but we do not mind much as their logarithms place us in certain demographic and lifestyle “buckets” and we are offered prohibitively great deals on many items. The WALL STREET JOURNAL reported recently that even Neiman Marcus loyalists are now abandoning their favorite store and buying the same luxury goods online at a significant discount to the set price of the high end Dallas retailer.
Do I buy Ms. O’Neil’s thesis? Yes and no. As a marketer, Big Data gives you an edge that is incomparable to any tool that you could work with in the past. It is far easier to forecast who will buy certain items if you have a plethora of data points about a prospect and his or her lifestyle. If you are loaning money, it only stands to reason that you charge more to someone who has a lower probability of paying you back.
To me, Ms. O’Neil’s most powerful point is that there is no human element involved anymore in so many financial agreements. Some people have errors in their credit history and no machine double checks the veracity of them or tries to clear them up. Also, some people regardless of the health of their personal balance sheet are highly trustworthy.
Ms. O’Neil writes with great clarity. I highly recommend it and encourage you to reach your own conclusions. If you work in any aspect of marketing or finance, Big Data is here to stay and needs to be evaluated carefully.
If you would like to contact Don Cole, you may reach him at firstname.lastname@example.org
or leave a message on the blog.