The current debate over "opening up" various cities and states, as opposed to "keeping them closed", in response to the local state of the COVID-19 epidemic, makes absolutely no sense to me.  First of all, "opening up" and "keeping closed" are not binary choices--there are infinitely many gradations between a complete lockdown and the pre-epidemic state.  In fact, no place (in America, at least) has been placed in a "complete lockdown" state since the epidemic began--every place, for example, allows "essential services", defined variously, to continue to operate.  Secondly, the continuum isn't even single-dimensional--rather, there are all kinds of specific rules that can be imposed that allow or prohibit various activities in various locations under various rules.  The possible combinations are endless.  So how do we decide among them?

The reason so many people are so baffled by these choices, in my opinion, is that we lack clear information on the effects of various restrictive countermeasures.  Suppose we require everyone in a particular location or pursuing a particular activity to wear some kind of cloth mask.  How much would that cut down on the spread of COVID-19?  What about requiring 6 feet of space between people?  12 feet?  6 feet plus a mask?  Does it matter whether it's indoors or outdoors?  As far as I know, we simply don't yet have good, quantitative answers to these questions, and until we do, we can't really determine what policy to impose on activities that involve interactions among people from different households.  (That's why most locales have been resorting to the one big hammer they know works:  fairly complete lockdown.)

Our first priority, therefore, should be to gather the data necessary to answer these questions as quickly and accurately as possible.  Once we have the answers, the issue of "essential" vs. "non-essential" activities will likely disappear.  Instead, we will be able to ask the question, "what set of restrictions makes a particular activity--whether essential or non-essential--under a particular set of circumstances acceptably low-risk for infection?  For example, shopping--whether for essential groceries or non-essential fashion apparel--is presumably reasonably safe under the right set of conditions, and once we have determined those conditions, we should apply them equally to all retail environments.  The same can be said of other activities, whether recreational, commercial, social or political.  (It's of course possible that for some previously popular activities, a reasonably safe set of conditions simply doesn't exist.  But one hopes that that set of activities is fairly small.)

Perhaps more importantly, approaching the issue in these terms is likely to mitigate a lot of the raucous political and social conflict surrounding it.  Right now, discussions about how and when to "open up" are dominated by loud, belligerent and largely ignorant voices, because more reasonable ones have little in the way of concrete proposals or supporting evidence to back them up.  Armed with a set of specific, transparent and scientifically supported policies, though--rather than a pair of vague, broad options--the reasonable voices might actually have a fighting chance.