Value-added, externalities and eggs

Today, what I initially thought was a mildly controversial statement about upstream vs. downstream profitability and value-added led to me finding myself with a little bit of egg on my face and also completely baffled about the way we use the term value-added.

Let me start off by saying that, as a economist, I tend to think of all resources as having an opportunity cost. In rare cases, that opportunity cost may be zero or close to it, but there are very few inputs to production which exist in infinite supply. Some things are certainly priced well below their actual opportunity cost, but we’ll get to that later.

When I think of the term value-added, I’ve always defined it in a real sense – the value of the product, net the value of the resources used to produce that product.  It turns out that matches Statistics Canada’s definition which states that value added is, “total output (or sales) less intermediate inputs.” In such a calculation, I’d take as given that labour inputs have a value (perhaps reflected imperfectly by wages paid), capital has a value, and resources such as oil, gas, pollution releases, or water used have values as well (even if these are not usually priced efficiently).   I’ve been scornful of those who confound this definition of value-added with the production of higher value products too many times to count.

Why do I have egg on my face? Well, after I posted the initial tweet, Erin Weir and Ian Gillespie both jumped on me for confounding value-added and profit.  I went off to the Economist to arm myself with a useful definition for the argument and here’s what I found: “the value of the firm’s OUTPUT minus the value of ALL its inputs.” I thought I was on solid ground, but I missed the next line. “It is therefore a measure of the PROFIT earned by a particular firm PLUS the wages it has paid.” This baffled me, since I tend to think of labour as an input to production and treating all wages as added value assumes implicitly that the labour input itself has no value or opportunity cost.

I thought this could not be right, so I jumped over to Industry Canada and sought out a definition of manufacturing value-added. It turns out that this definition applies here too.  “Manufacturing value-added consists of the value of manufacturing revenues plus net change in the inventory of goods in process and finished goods, less the costs of materials and supplies and of the energy, water and vehicle fuel used.” So, energy, water, fuel, materials and supplies all count as valuable inputs to production, and are deducted from revenues to yield value-added, but wages paid to labour (as though somehow that is not the purchase of a valuable input to production) are left out of the net calculation.

The final nail in the coffin came from Mr. Gillespie, with a link to StatsCan’s Guide to Economic Accounts, which makes it very clear that value added is defined gross of labour income. It turns out that Erin Weir actually put it well when he described the wages as being the part of value-added which is distributed to labour.

As Erin Weir suggested, I shouldn’t try to redefine value-added, so I won’t go that far, but I will suggest that treating labour as different from other purchasable inputs ignores its opportunity cost (at best, it ignores the value of leisure or the foregone potential productivity in non-labour market roles). In other words, from the national accounts’ definition of value-added, the opportunity cost of labour is an externality – it’s a cost which is borne by someone (the worker providing labour to a given activity) but which is not reflected in the measure of value-added. Wages are not gifts, but rather payment for services provided and I would humbly suggest that they should be treated as such.

Now, I have always been on the wrong side of national accounts because of my focus on environmental and energy economics.  In this respect, there are clear externality problems in that the measures of value-added we use do not include non-priced inputs to production such as emissions (waste disposal service) or potentially under-priced inputs (resource royalties which do not reflect the true opportunity cost of the resource being provided to the extracting firm). It had not occurred to me that these calculations also implicitly treated the opportunity cost of labour as zero – it just seems un-natural.  That raises some serious questions, so I guess it’s off to do some reading. Perhaps there is something I am missing in space between micro and macro.

Thanks to Erin and Ian for the challenge. I learned something today – normally I’d say that was value-added for your time spent arguing with me, but I don’t think I can say that with any confidence now.