Boundary issues

Last week, Grist published a piece guaranteed to hook the attention of anyone under 40 -- its headline proclaimed that your iPhone (or equivalent device) uses more electricity than your fridge. Which is, of course, a load of bunk. Anyone writing (or editing; yes, Scott Rosenberg, I'm looking at you) for Grist should know better.

Once I got past the ridiculous headline, what caught my eye was that the article was based on the conclusions of a report funded by the National Mining Association and the American Coalition for Clean Coal Electricity. The argument presented in the report is that all modern information devices rely entirely on communications networks and servers and other electronic equipment which delivers its value largely by its ubiquity.  Which means that it has to be running constantly, which creates what electrical utilities call "base load" (as opposed to the additional "peak load" created when, for instance, we all get home from work and turn our TVs on).  At the present moment, the cheapest way to supply base load electricity in most parts of the country is by burning coal (externalities aside), so the purpose of the report is to create a cognitive association between usage of information/communications devices (which is increasing) and coal-fired electrical generation (which is obsolescing).  I guess if you're selling a clunky old wagon, your best bet is to hitch it to a rising star.  Any rising star.  Any form of hitch.  However insubstantial.

By the second paragraph of the executive summary, the report's illogic becomes apparent.  Attempting to create as much concern as possible, the writers note that "as the world continues to electrify, migrating towards one refrigerator per household, it also evolves towards several smartphones and equivalent (sic) per person. (Emphasis in the original.)  Since a household consists of one or more people, anything that ramps up as a multiple of people must increase more rapidly than anything driven by the number of households, right?

One flaw, of course, is that the electrical load with which the report claims to concern itself is the load created not by each individual smartphone, but by all the infrastructure needed to keep the smartphone (and other information-communication-technology) network running behind the scenes.  That infrastructure doesn't scale with the number of devices utilizing it, it scales primarily with the volume of information passing across the network.  The fact that I have a smartphone and a work computer and a personal computer doesn't mean that I'm simultaneously using all three.  The demand for network services isn't strictly additive -- some of what I do on my home computer is stuff I otherwise would have done at work (and, of course, vice versa).  Substitution occurs, slowing the rate of increasing demand.

But the bigger flaw in the overall logic of the headline is that for smartphones, all the electrical demand of the supporting infrastructure is figured in, while for refrigerators -- you guessed it -- none of the supporting infrastructure's demand is even mentioned.  Sure, a smartphone is of limited use without its supporting communications network, but how useful is the average home refrigerator without its supporting network of supermarkets, refrigerated delivery trucks, cold-storage warehouses, frozen food manufacturers, meat lockers and the like?  The value derived from a smartphone correlates with the amount of information that passes through it, but the value derived from a refrigerator seems pretty closely tied to the volume of food that passes through it.  Most of that food, in this country and increasingly around the world, is an industrial product.  And that industry (like the telecommunications industry) uses a whole lot of electricity and other forms of energy. 

(Ironically, one of the points the report makes in its very first paragraph is that the information-communications-technology industry now uses more energy than global aviation.  The irony lies in the fact that agricultural production uses more energy globally than all forms of transportation combined.)

The lesson here, for Grist writers and for all of us, is that boundaries matter.  When somebody says that A is greater than B, are A and B really being defined in the same terms?  Under the same conditions?  Within the same constraints?  When that somebody is an interested party, or (today, increasingly) funded by an interested party, skepticism is warranted and often rewarded.  Figures don't lie, but . . .