1-2 days a week is a sweet spot, according to human resources guidelines and some actual research.

long-run effects of telecommuting are all described by bell-shaped curves: Telecommuting first increases skilled and unskilled workers’ productivity and GDP up to some threshold. Beyond that level, a higher share of home-workers reduces the strength of the knowledge and information spillovers which, therefore, do not produce desirable effects. Too much WFH may thus be detrimental to long-run innovation and growth due to limitations of information and communication technologies as well as foregone agglomeration economies in the form of face-to-face contact and knowledge spillovers.² Figure 1 illustrates this point via some back-of-the-envelope computations using consensus parameter values. The WFH share that maximizes GDP varies between 20% and 40% in our simulations – one or two working days per 5-day week. This is broadly in line with recommendations made in human resource management (Gajendran and Harrison 2007).

Vox

It wasn’t exactly clear to me whether the model mentioned here distinguished between the share of people work from home and the share of an individual’s work days that would be at home. That may not matter to a mathematical model, but it obviously matters to an individual.

1-2 days sounds about right to me. It’s enough to get the personal collaboration and interaction, which is important both for innovation and psychological reasons. Just having a change of scenery a couple days a week is important for psychological reasons too. That 1-2 days at home does cut down on all that wasted time and pollution caused by typical car commutes. This wouldn’t have to be the case if more people lived in communities where they could have active commutes (walking or biking), because the commute then provides some fresh air, exercise, a change of scenery, and sometimes a little social interaction. Sometimes its nice to stop at a coffee shop on the way in, and sit on a park bench for a few minutes or enjoy an…er…adult beverage on the way home (with no possibility of drunk driving, although angry car commuters can be a danger to mildly inebriated pedestrian. I’ve also noticed that car commuters seem particularly angry on Friday afternoons, while walking commuters seem particularly happy. Why is that? Because the walking commuter’s weekend has started and the car commuter’s psychological weekend doesn’t start until the car is in the garage, and in between that moment and the moment they are in are many forces outside their control.) It helps to have “third places” to unwind a bit between work and home. This is a major reason I live where I do, and one thing I have really missed during the pandemic. (Another thing I have missed is my children to and from school, parks, playgrounds, museums, etc.) Over the past year, the headaches of city living have outweighed the benefits I had taken for granted before that.

A couple more thoughts on working from home:

• Obviously, some kinds of jobs can do it more than others. The kind that can tend to be higher paying. I think we have all learned something over the past year about “essential workers”, which actually means essential jobs done by expendable workers. Here’s a crazy idea – people who volunteer to do dangerous jobs like deep sea diving and drilling for oil in war zones get hazard pay to make it worth their while. It should be possible to have a government program that supplements the pay of ordinary people doing ordinary jobs under emergency conditions.
• Co-working seems to me to hold some promise as a compromise between working in a corporate office and working from home. You get a professional atmosphere, a bit of breathing room between work and home (which let’s be honest, your family members may appreciate as much as you do), and you can significantly cut down on your commute – ideally, your co-working site should be accessible on foot, by 100% safe protected bike lane, or in a pinch by public transportation. Over time, this could allow your employer to downsize the office if that is what they want to do, without transferring 100% of the burden of operating an inefficient and far from ideal professional office to each individual worker in their family home. Some employers may have concerns about confidentiality, but outside high-security industries this should be manageable through things like sound-proof booths in the co-working sites.
• Finally, my observation among professional workers is that some people and some specific jobs are better suited to it than others. I have noticed that the same people who struggled with communication in the office (for a variety of reasons – language barrier, personality type, or just being young and not having figured it out yet) are the ones who have been left behind in the co-working world. If those people are otherwise valuable, employers need to figure out how to bring them along through mentoring, training, carrots and/or sticks of some sort or they won’t realize the career potential they otherwise could have.