Consumption and Systems Inventories: A Short Comparison
Two popular alternatives to "traditional" (geographic) GHG inventories are called "consumption-based inventories" and "systems-based inventories". These two alternatives are sometimes assumed to be the same, although they are actually very different. This page introduces, compares, and contrasts the two approaches.
The term "systems-based inventory" refers to an emissions inventory where "systems" represent and comprise all the parts of the economy working to fulfill a particular need. For example, the provision of food system includes all emissions from the electric power, transportation, industrial, and agricultural sectors associated with growing, processing, transporting, and disposing of food. The systems view is helpful for framing opportunities to reduce GHG emissions through prevention-oriented mitigation strategies that act across an entire system.
The concept of a "systems-based inventory" first gained widespread use in EPA's report Opportunities to Reduce Greenhouse Gas Emissions through Materials and Land Management Practices. This report estimated that 42% of the nation's GHG emissions (in 2006) were associated with the "systems" of "providing materials" (food and non-food). These emissions are associated with producing, transporting and disposing of materials, but not using them. The inventory presented in EPA's "systems" view is (for the purposes of materials management) identical to the conventional domestic GHG inventory - both inventories include exactly the same emissions. The difference is that the conventional GHG inventory assigns the emissions to "sectors" (such as "electric power industry" and "transportation"), while the systems inventory allocates the emissions into "systems".
Working independently of EPA, one of the lead authors of the EPA report, Dr. Josh Stolaroff, married EPA's systems-based inventory with an analysis of GHG emissions embodied in US trade, prepared by Dr. Scott Matthews and Dr. Chris Weber of Carnegie Mellon University. Published by the Product Policy Institute, this report portrays US domestic emissions, less emissions associated with the domestic production of exported goods, plus emissions associated with the foreign production of imported goods, but still assigned to the same "systems" as EPA's report. In this frame, materials contribute 50% of the GHG emissions – equal to all of the emissions from all other systems (heating/cooling/lighting buildings; transporting people; use of appliances; and infrastructure).
A consumption-based inventory covers the total global GHG emissions occurring from economic consumption within a set region (e.g., a country). Consumption based inventories attempt to estimate all emissions - both inside and outside the community - that arise as a consequence of consumption activities within that community. Consumption-based inventories are explained in greater detail on the Greenhouse Gas Inventories page of this toolkit.
Consumption-based inventories may assign emissions to categories using several different logic rules. For example, emissions might be assigned to categories of commodities as consumed. Alternatively, emissions might be assigned to categories of commodities as produced. To illustrate the difference between these two approaches, consider the natural gas that is burned on-site at a factory that sews apparel. This hypothetical apparel company produces a variety of clothes - some purchased directly by households, and others purchased by businesses (such as uniforms worn by hotel housekeeping staff, or clean room suits worn by employees at semiconductor factories). If a consumption-based inventory assigns emissions to categories of products as consumed, then the emissions from burning of natural gas at the garment factory will be assigned to several different sectors: clothing, hotels, and computer purchases, for example. These represent the different categories of consumption ("final demand") that ultimately drive production in the garment factory. Alternatively, emissions could be assigned to categories of commodities as produced. In this example, all emissions at clothing factories satisfying (directly or indirectly) "final demand" in the community would be assigned to the "clothing" sector, as this is the producing sector where the emissions occur.
As of 2016, relatively few consumption-based inventories have been performed, and no standard has been set for assigning emissions to different categories. However, all inventories produced so far are using what appears to be the more intuitive approach: assigning emissions based on the category of final demand (commodities as consumed). From the perspective of materials management, it is important to understand how this relates to emissions over the life cycle of materials. In this approach, not all of the emissions associated with the use of materials (in the course of satisfying local consumption) will be assigned to categories that appear to relate to materials. For example, the emissions associated with production of office paper used by professional service firms that serve local residents will be assigned to service-related categories such as "legal services" or "real estate" (and not "paper manufacturing"). Similarly, emissions associated with the manufacture of materials used in residential construction and remodeling will be assigned to a category of "building materials" only if consumers (households) are purchasing building materials directly (for example, at a hardware store). To the extent that many building materials are purchased by contractors that in turn are providing a service (remodeling) to households, emissions associated with these materials will be classified as "construction and remodeling services", not "building materials" - the household spends money on a remodeling contractor (a service provider), who in turn spends a portion of that money purchasing materials (lumber, drywall, etc.).
"Systems-based" and "consumption-based" inventories differ in two - or possibly three - important regards: their geographic scale, the emissions they include, and how the emissions are categorized.
It is important to understand that the "systems-based" inventory has only been performed once: by EPA, for the nation as a whole. In theory, a systems-based inventory could be conducted at a smaller geographic scale (a state or a local government), although as of this writing, there are no such examples to point to. In contrast, consumption-based inventories have been performed (or are underway) at multiple geographic scales: nations, states, counties and cities.
As used by EPA, a "systems-based" inventory counts all emissions that physically originate within the borders of the United States. These are exactly the same emissions counted in the "official" domestic GHG inventory. In contrast, a "consumption-based" inventory counts emissions everywhere in the world, but only if they are associated with satisfying final demand ("consumption") by households, governments, and investment expenditures associated with the community. A consumption-based inventory excludes in-community emissions if these emissions are not associated with satisfying in-community final demand (for example, a local cement factory that produces cement used in other communities). At the national scale, Dr. Stolaroff's report for the Product Policy Institute represents an important cross-over between these two methods, reporting results in a manner consistent with the "systems" view but counting out-of-country emissions associated with the import of goods to the US, while excluding in-nation emissions associated with production of exported goods, in a spirit similar to that of consumption-based accounting.
Categorization of emissions
The final difference is how emissions are categorized. In theory, a systems-based view could categorize emissions using a variety of categorization schemes. The systems-based inventory most familiar to the materials management community (EPA's 2009 report) uses categories such as "provision of goods (non-food)" and "provision of food". Again, we currently have no other examples to point to, but can imagine alternatives. For example, "non-food" and "food" might be combined. Emissions associated with making cars could be moved from the "provision of goods" category and into the "transportation of people" category.
As described above, consumption-based inventories may categorize emissions according to several different standards. One approach ("products as consumed") sheds light on materials/productswhen purchased as materials/products by consumers; this doesn't readily show the full scope of materials as materials, as materials used to satisfy other forms of demand (such as services) will have their emissions assigned to those services. A second approach ("products as produced") is more akin to the EPA systems view, grouping together all emissions associated with production of goods, regardless of whether the goods are purchased directly by consumers (e.g., pajamas) or rather are used to satisfy final demand for services (e.g., hotel housekeeping uniforms) or even electricity or fuels (e.g., service station uniforms).