We have become accustomed to presuming that what has happened in the past will give us a good idea of what may happen in the future, whether in determining insurance premiums, stock market movements, or the outcome of political referendums. That is changing, however. As has become evident in so many ways over the past two decades, we can no longer count on historical performance to serve as a guidepost to the future.
One need look no further than the outcome of the Brexit vote or the Colombian peace referendum to see how spectacularly the pollsters and bookies got it wrong. While UK voters were almost universally expected to reject Brexit, and Colombian voters — longing for peace after a half century of war — were naturally expected to embrace the recently signed peace agreement with the FARC, they did exactly the opposite. Is that the result of outdated polling methods? Is it a failure to truly understand the pulse and intention of voters? Or have voters around the world become so fickle and unpredictable that accurately predicting election outcomes is no longer possible?
Something similar may be said about historical U.S. economic performance, which would suggest that the U.S. has been due for a recession for some time. According to the National Bureau of Economic Research, which has tracked recessions on a monthly since 1854, between 1854 to 1919, there were 16 economic cycles, the average recession lasted 22 months, and the average economic expansion was 27 months. From 1919 to 1945, there were 6 cycles, recessions lasted an average of 18 months and expansions for 35 months. And the period from 1945 to 2001 saw 10 cycles, recessions lasted an average 10 months, and expansions an average of 57 months. So recessions got shorter and expansion periods longer over time.
Since the third quarter of 2009, the U.S. economy has expanded every quarter but one (in 2014) and is poised to continue the trend through the final quarter of 2016.