One of the more substantive moments of the debate came when Erin Burnett asked Sanders and Yang about how to respond to automation. So it's probably worth pointing out that (a) Burnett's premise was basically all wrong and (b) Sanders had a much better answer 1/
Burnett's premise was that automation is about to eliminate large numbers of jobs. This could be true — predictions are hard, especially about the future — but at least so far nothing like that is happening 2/
Of course there are always jobs being eliminated by technology somewhere. Remember when there were hundreds of thousands of longshoremen? But the idea that machines are replacing people en masse presumes that overall labor productivity is soaring, when the truth is the reverse 3/
Here's labor productivity over the past 30 years. It's slowing down! Output per hour rose 36% from 1996 to 2007, only 15% from 2007 to 2018. If automation is replacing people, why aren't we producing more with fewer people? 4/ 
What we do have is a persistent problem of weak demand; yes, we have full employment now, but only with extremely low interest rates, which means little ability to respond to the next downturn. This makes a strong case for a big government investment program 5/
As advocated by Sanders, but also left-wing radicals like, um, Larry Summers. Meanwhile, Yang's suggestion that we give everyone $1,000 a month doesn't address the real problem — and it's not enough to live on 6/
I know that both the DC elite and Silicon Valley types are attracted to the idea that technological unemployment is our biggest problem. But it isn't; and in this case Sanders is the realist, Burnett and Yang engaged in fantasy economics 7/
PS: Just realized that Biden said "the fourth industrial revolution is costing jobs." Oh, dear. He's bought into the automation thing too. 8/
