A review of methods to trace material flows into final products in dynamic material flow analysis: From industry shipments in physical units to monetary input–output tables, Part 1

Abstract

Dynamic material flow analysis (dMFA) is widely used to model stock‐flow dynamics. To appropriately represent material lifetimes, recycling potentials, and service provision, dMFA requires data about the allocation of economy‐wide material consumption to different end‐use products or sectors, that is, the different product stocks, in which material consumption accumulates. Previous estimates of this allocation only cover few years, countries, and product groups. Recently, several new methods for estimating end‐use product allocation in dMFA were proposed, which so far lack systematic comparison. We review and systematize five methods for tracing material consumption into end‐use products in inflow‐driven dMFA and discuss their strengths and limitations. Widely used data on industry shipments in physical units have low spatio‐temporal coverage, which limits their applicability across countries and years. Monetary input–output tables (MIOTs) are widely available and their economy‐wide coverage makes them a valuable source to approximate material end‐uses. We find four distinct MIOT‐based methods: consumption‐based, waste input–output MFA (WIO‐MFA), Ghosh absorbing Markov chain, and partial Ghosh. We show that when applied to a given MIOT, the methods’ underlying input–output models yield the same results, with the exception of the partial Ghosh method, which involves simplifications. For practical applications, the MIOT system boundary must be aligned to those of dMFA, which involves the removal of service flows, sector (dis)aggregation, and re‐defining specific intermediate outputs as final demand. Theoretically, WIO‐MFA, applied to a modified MIOT, produces the most accurate results as it excludes massless and waste transactions. In part 2 of this work, we compare methods empirically and suggest improvements for aligning MIOT‐dMFA system boundaries.

A review of methods to trace material flows into final products in dynamic material flow analysis: Comparative application of six methods to the United States and EXIOBASE3 regions, Part 2

“…From these, end-uses can be inferred and assumed according to sector/market label (e.g., "construction"). Industry shipments data often cover only parts of total, economy-wide material consumption (see Streeck et al, 2022). For that reason, many of the studies identified in Streeck et al (2022) do not use industry shipments as direct input to dMFA models.…”

“…Industry shipments data often cover only parts of total, economy-wide material consumption (see Streeck et al, 2022). For that reason, many of the studies identified in Streeck et al (2022) do not use industry shipments as direct input to dMFA models. Instead, studies calculate end-use shares D by dividing individual by total shipments and apply these shares to total material consumption, as we do here (Equation 1).…”

“…For all methods presented here, and as is common in input-output analysis (Schaffartzik et al, 2014;Weisz & Duchin, 2006), proportionality between monetary and physical flows is assumed as default. For a more detailed description of each method, we refer to Supporting Information S1 and Streeck et al (2022).…”

“…are commonly used to split economy-wide material consumption to different end-use products or sectors (see Figure 1 in Streeck et al, 2022).…”

“…Therefore, in part I of this work, we reviewed the literature that used industry shipments and MIOTs to generate end-use shares, and theoretically assessed five different methods for their similarities, strengths, and weaknesses (Streeck et al, 2022). We concluded that the waste input-output approach to MFA (WIO-MFA; Nakamura et al (2007)) represents the theoretically most accurate choice to derive end-use shares from MIOTs, as it closely sticks to the mass-balance principle of dMFA.…”