Feed on

Many forecasters make predictions based on extrapolating past trends into the future. Malthusians famously did this with population growth rates, materials alarmists do it with resource forecasts, debt-alarmists do it with borrowing forecasts, and so on. Here is perhaps a silly analogy, but one that unfortunately is necessary to make the point that it is almost irresponsible to use prior growth rates to predict future growth rates, even for many processes that we have a solid understanding of, but certainly for processes that are complex.

Consider that a typical oak tree grows by 2′ per year. On the basis of watching your seedlings grow to 10 feet tall after 5 years, and 20 feet tall after 10 years, would you argue that in 500 years’ time, your oak tree would stand at 1000 feet tall?

Or consider that a large newborn baby is about 10 pounds. And typically that baby will grow to weigh about 20 pounds after his first year of life. Would you take anyone seriously who tapped you on the shoulder and said, “hey, Wintercow, your son Isaac grew to 20 pounds after his first year, where are you going to put all 5,120 pounds of him when he is 10 years old?”

2 Responses to “Weekend Thought: If Present Trends Continue …”

  1. Harry says:

    This is the curse of Microsoft Excel in the hands of people who never understood quadratic equations or calculus.

    Isaac is aptly named.

  2. Harry says:

    I always got a kick out of reading explanations for a drop in a good company’s stock when quarterly earnings disappointed street forecasts by a penny or two. Before the accountants were done with the numbers, the Chairman of the Board did not know what the earnings would be down to the penny.

    Now that we have transparency with Sarbanes-Oxley, Dodd-Frank, and a reinvigorated SEC, the forecasts of financial analysts are sure to improve.

Leave a Reply