Research the Past, Predict the Future

Research is backward looking.

You take a bunch of studies over past time-slices and come up with theories that explain the data of the studies.

Even if the theories accurately explain the past (the majority of the time they don’t), they don’t explain the present or the future.

The future can’t be researched.  It has to be predicted.  How do you predict the future?

Past data can be a useful item in the toolkit.  But the theories that explain it can’t automatically be applied to the future.  The data is the effect, and what changes in the future are causes.  Will yesterday’s causes hold tomorrow?

To predict that takes more than familiarity with past data.  Understanding of the unchanging things, like the basics of human behavior and natural limits (rational self-interest, scarcity, etc.), are paramount.  Then it takes some observational power of the less tangible trends and moods of the day, extrapolation not just of data but also trends and beliefs that cause it.

It also takes creativity.  The future isn’t only predicted, it’s created.  The most powerful predictors help create it through their predictions.

I enjoy research findings as much as the next guy.  But I try to remain disciplined and remind myself all research is about the past.  If conclusions are true, a separate set of arguments are needed to predict what it means for the future.