In my previous piece I focused on a simple trap that we all fall into when we count what happens (jobs) but forget to count what doesn't happen (other jobs). We forget to net it out.
In this piece, also mercifully short, I hope, I would like to make the case that "job creation" is not a worthy goal of economic policy anyway. Sure, we want jobs for everybody who wants one, but jobs should come as a by-product of producing something or providing a service that is needed. Focusing on the job per se, rather than on what is produced with the job, leads to waste. There is always more work to be done than there are people able and willing to do it. It is the work that should drive the process; not just a job for a job's sake.
Some examples will make this obvious. The government could create enough jobs to achieve full employment by having the previously unemployed dig holes on Monday, Wednesday and Friday and fill them up on Tuesday, Thursday and Saturday. Or, in my long-time favorite example, if you want more jobs, replace the heavy equipment on a construction site with shovels. If that doesn't do it, replace the shovels with spoons. Silly? Sure, but not much sillier that some of the rhetoric we take all too seriously.
When we say we have to "create jobs," we are setting ourselves up for a misallocation of scarce resources and a sub optimization of our economic efficiency. When we say that, we imply that there are just so many jobs to go around and we need to find and fill them. Pretty soon that kind of thinking leads us into the French fallacy of spreading a given (fixed) amount of work among more workers by shortening the legal work week. Only so much work to be done; let's spread it out.
Don't forget that the demand for workers will increase with the work being done. Don't forget Say's Law, "Supply creates its own demand." We sometimes forget it in times like this because we get caught up in Keynes' Law, "Demand creates its own supply." Come to think of it, neither law is a rationale for wasting good workers by putting them in sup-optimal jobs. The lesson is simple: Don't count jobs; make jobs count."