Gottlob Frege (1848-1925) belongs to what I think of as a late nineteenth/early twentieth century golden age of ferment in mathematics, philosophy, economics and physics, before all those fields were fully defined and then subdivided. In college I remember being delighted that he actually wrote a text called “What is Number?”, and over the years I’ve been interested to learn a little about the different counting systems (such as “one, two, and many”) that exist in different cultures and are reflected in their diverse languages around the world. I’m thinking of this cluster of concerns again in relation to the Trump tax cuts.

It has already been observed that while most middle-class consumers got a tax cut, the way it was delivered made them feel poorer, not richer. They got a little less tax taken out of their paychecks every week, and as a result, were due a smaller tax refund, or none at all, come April 15th. I think this may damage consumer confidence (and hence, spending and economic activity) because like me, many people are in the habit of getting a big refund every year that can be used for large purchases, or just to re-establish a cushion in the bank account that makes you feel safe and good. Having a few dollars extra in each paycheck doesn’t have the same effect, and the abstract idea that over the course of the year you’re paying less doesn’t really register at all. I know behavioral economists have studied this in depth, and could have told the tax cut framers this upfront; but I doubt any were consulted.

In other words, most people don’t at all mind “lending the government money interest free,” which is what we effectively do when we overpay in anticipation of a refund. It’s true that when we do this, we use the IRS like a zero-interest savings account. Which one can frame as though we are getting taken advantage of; but the fact is that we are receiving a valuable service (what economists call “utility”)  from the IRS in this scenario. Namely: a structure to ensure that we save up a chunk of money over the course of the year, a thing that it is very hard to do if you have to make the decision to do so week by week.

But getting back to old Gottlob’s question: it occurs to me that in practice, lots of us don’t have finely graduated perceptions of different numerical quantities. In household finance, which is where most of us deal most intensively with numbers, there are regions of the numberline that function as sort of fuzzy numbers, but mean something well-understood for people of a similar class and location. E.g., for me I would list “major appliance or car repair expense” (e.g., $400-$600), “family meal out expense” (e.g., $30-$40); “rentlike expense” (in the $2500-3200 range), and so on. Within those intervals, there’s not a ton of differentiation in how things feel, which marketers well understand when they try to push us toward the higher end.

As for me, currently I’m in search of a committed source of “health-insurance-like expenditure,” which at current market rates for a family of four is–deep breath!–around $36K a year. (Alternatively, I want to find an organization that will add my family to its plan on the same premium-sharing basis as its regular employees, in exchange for part-time work from me.) My husband and I had hoped to start businesses this year with the help of Obamacare, but it is failing to encourage entrepreneurship in communities like ours.  You can’t qualify for any meaningful subsidy if you can afford a Newton rent.