I think this Fortune article (paywalled, but I was able to read it the first time I clicked) drawing conclusions about remote work based on productivity statistics is off base. Labor productivity, as I understand it, is dollars changing hands in the economy divided by hours people say they worked. There are a number of measurement problems here. First, in the short term it is just going to fluctuate with dollars changing hands, which fluctuates for all sorts of reasons, so it makes more sense to look at longer-term averages. Second, dollars changing hands is not a perfect measure of value – we could be paying the same number of dollars for crappier goods and services as our expectations are gradually lowered over time. I really suspect this is what is happening.

It does make sense to me that self-reported hours worked at home would be less productive. Even if most people are honest most of the time, some people are going to be less honest some of the time than they would be in an office. People are going to be more distracted. But in all these cases, they are going to report the same number of hours worked and get paid the same number of dollars they would have in the office. So there will be no effect on calculated productivity, while we get used to gradually shifting baseline of crappier goods and services over time.

I think another effect is that training and onboarding are getting harder in some sorts of jobs. Some jobs have a playbook telling a worker exactly what to do, but many jobs do not. In my field of engineering, there is not much of a playbook because we are often trying to apply existing knowledge to solve novel problems under changing external conditions. I learned this job in the 1990s and 2000s by spending a lot of extra time in the office at the end of the day shooting the breeze with colleagues, mentors, and clients. Somewhat frequently, someone would suggest moving these sessions to a local drinking establishment and they would go well into the evening. This was not necessarily healthy for work-life balance or for my liver and waistline, but it’s an important part of how I learned my job and industry and why I am good at it today. This time didn’t go on my time sheet, and yet it boosts my subjectively measured productivity today.

I don’t want to complain about today’s crop of young people, who are just as intelligent as my generation (perhaps more since they’ve been exposed to less lead and air pollution) and seem to have better health habits overall. But the combination of working from home, less informal interaction with mentors, and job hopping means it is much harder for them to learn to do jobs really well. In decade, they will be the ones doing most of the work and trying to train the generation under them, and again we will just get a gradually shifting baseline of lower expectations and worse outcomes, even if we may not be measuring that effectively in dollars.