When making long-term investments, it's often a bad idea to optimize for efficiency (if we want to measure efficiency as return on investment), because it's difficult to predict ahead of time which projects will work out the best in 10 or 30 years. Sometimes the best bets are the ones no one thinks will work out, and therefore no one wants to make.
I think PG has spilled a lot of ink about this in the context of VC investment, but the same idea doubly applies to government funding because their time horizons are even longer term. If all you do is optimize for next quarter, next year, or even the next 5 years, you're going to miss the things that at first seem innocuous or useless, but then in 30 years blow up into something huge.
I think PG has spilled a lot of ink about this in the context of VC investment, but the same idea doubly applies to government funding because their time horizons are even longer term. If all you do is optimize for next quarter, next year, or even the next 5 years, you're going to miss the things that at first seem innocuous or useless, but then in 30 years blow up into something huge.