How Does Robert Gordon Know What Innovations The Future Will Bring?

Distinguished economist Robert Gordon’s recent work on innovation and growth melds together two different themes in what is, I think, a confusing and problematic way. First he argues, correctly, that as neat as information technology is that the digital revolution isn’t nearly as transformative as the Second Industrial Revolution associated with electrification was. But a second point that he emphasized in a recent Wall Street Journal op-ed is the idea that future innovation will continue to disappoint.

Some of this is just a failure of imagination, as in his thoughts on self-driving cars, but even worse is straight-line projection:

Pharmaceutical research appears to be entering a phase of diminishing returns. Developing new drugs is increasingly expensive, and the potential pool of beneficiaries is ever smaller, mainly people with esoteric types of cancer. Few of the medical optimists acknowledge a stark historical fact: The rate of improvement in U.S. life expectancy was three times higher in the first half of the 20th century than in the second.

I have no real opinion on this, but I don’t understand why Gordon does either. It’s an interesting fact that recent new pharmaceutical developments have mostly involved incremental advances in the treatment of esoteric cancers. But that hardly means there will be no future dramatic advances. Maybe we’ll develop really potent general purpose anti-cancer medication. Or maybe we’ll develop genetic engineering techniques that make future generations of humans much more robust against illness. The future trajectory of science is inherently impossible to forecast, since a reliable method of projecting what will be invented in the future would entail inventing the thing.