Warning: This post may be one of my more “wonkish”/boring efforts to anyone not interested in national income statistics.

I’ve been studying macroeconomics on and off for the past four years. These studies have mostly consisted of trying to make headway through some of the main textbooks (e.g. Mankiw’s intermediate book), along with reading some of the classic papers and, perhaps most usefully, carefully following all of the macro debates on the major econ blogs. This approach has been successful enough, but the one big gap has been anyone actually holding my feet to the fire to make me walk through the math underlying the basic models (IS-LM, Solow Growth, etc.). So, this term I’m sitting in on a Master’s level macro class. We’ll see how it goes.

As is traditional*, the first macro class covered national income accounting. The professor walked through the three standard methods for calculating GDP: sum of the value of final production (plus changes in inventories), sum of value added by industry/sector (plus changes in inventories), and the sum of factor payments plus indirect taxes. The first two involve adding up the value of the stuff that is sold, the last involves tracking the payments to workers and owners (of land and capital). Listening to the third method explained though, something bothered me a bit: government had mysteriously appeared. The first two approaches were explained without any reference to the government sector, but the third added in the proviso about indirect taxes. The problem, as it was framed, is that the observed incomes of individuals and businesses don’t quite sum up to the value of the goods produced because of “indirect” taxes. Think here of a sales tax: a business sells something for $1, but only receives $.94 (in Michigan) because 6% goes to the State. So we have to account for that somehow.

What was notable to me was that at that point, we hadn’t mentioned government as a part of the (simple, model) economy at all. But, of course, many of the goods and services used to produce and distribute goods and services are made by the government. Roads are a pretty obvious example, but there are many others. So, why not just think of Government as another, kind of weird factor of production? In that economic ontology, indirect taxes are no longer an exception added to the sum of factor payments – they are one of the factor payments. And your production function is no longer dependent on just land, labor and capital, but also on government (Y = F(K,L,G)). Ok, now all sorts of wacky things would follow from this set of definitions. But it would be a nice opportunity to think through exactly how and why government outputs are different from other kinds of products, and why payments to government (taxes) are different from other kinds of payment. And it would open up the long closed debate about whether or not some government output should count as intermediate production – like the portion of the value of roads** used up by businesses distributing goods – rather than final production.

All this to say, the basic idea that government is fundamentally not productive works its way into many parts of our economic statistics and economic theories, even as we explicitly acknowledge its importance for production in other ways.

* Judging by the typical placement of a chapter about national income data in a macro textbook – often chapter 2, the first substantive chapter after a short introduction.
** Roads might be a bad example because they are durable, and thus should maybe count as a capital good/investment. On the other hand, I live in Michigan, and here the roads do not appear to be that durable at all…