Work from home is ending. But the pandemic is still here.

JPMorgan Chase & Co. announced last week that at least some employees — largely those in sales and trading — would be expected to be back in the office September 21. Although many organizations have encouraged employees to come back, JPMorgan's move was notable because it was framed not as a choice, but an expectation. Exceptions will be made, but this is an opt-out, not an opt-in. Other employers are wondering how hard to push people to return, and employees are wondering if their at-home time is ending. 

There is some feeling of abruptness to this. If it wasn't safe to go back in August, why is it safe now? In some locations (for example, Texas and Georgia) cases have dropped in the last month, so it's possible to see why returning to work might have become a better idea. But JPMorgan's announcement was about its offices in New York. Cases there have been stable and low since the middle of June. This makes the decision seem arbitrary; that employers have just decided "it's time" rather than evaluating the risks to employees. 

One hallmark of good decision-making is consistency. Changing your decision when nothing has actually changed is not consistent. In my view, though, something has changed. But it's not the facts about the virus. Instead, it's how we're thinking about the alternatives. The choice of whether or not to open a workplace — or to open a school, or to go to the hairdresser — depends crucially on defining what will happen in the "or not" scenario. 

For months we've been, either explicitly or implicitly, operating under the assumption that — pretty soon — this will "end." That is: If we just wait until a vaccine is approved, or rapid antigen tests are available, we'll be able to reopen everything normally. Decisions not to reopen workplaces over the summer were, I think, rooted in the expectation that if we just waited a bit longer, things could be more normal. Employers wouldn't have to go through all the work of setting up a Covid-safe workplace.   

Because re-opening workplaces before a vaccine is available is going to require a lot of additional work — more cleaning, more distancing, weird partial pod-oriented schedules, messing around with the elevator settings. Masks! Routine testing! People will object. It will be annoying, to say nothing of the fact that some employees are likely to get Covid-19, whether or not they get it at work. Investing in this sort of office homecoming really wouldn't make sense if everything is going to return to normal in a few weeks.

When we ask what's changed, I think the answer is that we have started to realize — and I know we do not want to hear this — that it's not a few weeks, or even a few months. Employers have realized they need to plan for the long term. It's dawning on all of us that there is not an "end" in sight. Vaccine trials are moving along, and it does seem possible we will have some vaccine by the end of the year. But it will take time to vaccinate everyone, and the vaccine will not be perfect. Antigen tests are here, but they aren't a magic bullet and they are not yet widely available. I think Dr. Fauci may have said it best when he said that we'll be dealing with this is some way for the next 18 months … at least.

With that realization, employers now face the choice of opening "Covid-safe" now or waiting perhaps 18 months to open "normally." This is a very different choice. Although a few companies (tech firms, largely) have decided that work from home is a long-term solution, others are realizing that simply won't work for them. JPMorgan says its internal data show workers are not as productive from home. The result is a decision to make the investments to bring people back as safely as possible in this new normal. These investments in safety make a lot more sense than they would if we were going to return to normal in a few weeks.

We must acknowledge that the outcome is uncertain. It is possible that cases will increase if we open workplaces. Employers will need to be diligent about monitoring, open to backtracking, and willing to learn from what works and what doesn't. This learning, too, has more value as we acknowledge a longer road ahead.

This logic extends to other types of reopenings. Schools are one obvious example, but so are things like gyms, restaurants and other businesses. Keeping schools closed now under the assumption that you'll be able to open normally in October may make sense; once we realize that we're looking at an entire academic year at least, it may make sense to invest in ways to conduct in-person instruction more safely.

I can hear the howls of protest already. People will object that this is privileging the economy over health, and it's morally reprehensible to consider asking people to return to work (or school, or anything else) until we are sure it is 100% safe. I see this point, but staying closed also has serious risks in terms of employment, income and, yes, health — both mental and physical. There are no perfect options here. Making good choices among the many imperfect options requires realism about how long this pandemic is likely to last.

Emily Oster, is a professor of economics at Brown University. She is the author of "Cribsheet" and "Expecting Better."

Disclaimer: This article first appeared on bloomberg.com, and is published by special syndication arrangement