Chapter 36 — Further Reading

⚠️ This chapter's literature has an unusually high ratio of advocacy to measurement, in both directions. ⚠️ Sections 1 to 4 below are measurement. Section 8 is where the advocacy lives, and it is worth reading — with the label attached.

1. ⚠️ Environmental impact — the primary evidence

Poore and Nemecek, "Reducing food's environmental impacts through producers and consumers," Science (2018).

⚠️ The single most important source for §36.3, and it is worth reading properly rather than reading about. ⚠️ It synthesizes life-cycle data across a very large number of farms in dozens of countries and multiple impact categories.

⚠️ What to look for:

⚠️ The ordering across food categories, which holds across every impact measure. ⚠️ The within-category variation, which is enormous and is the more surprising finding. ⚠️ And the comparison that matters: highest-impact plant producers versus lowest-impact animal producers.

⚠️ Note the methodological appendices. ⚠️ This is where the functional-unit choices are made, and §36.3b is entirely about how much those choices matter.

Life-cycle assessment methodology critiques — ⚠️ on system boundaries, allocation between co-products, and functional units. ⚠️ Read at least one, because "per kilogram versus per calorie versus per gram of protein" is not a technicality; it is where a great deal of the public argument is actually being conducted.

Food-miles and transport-share analyses — ⚠️ the evidence behind the 🟡 verdict on local sourcing, including the air-freight exception.

2. The convergence

The EAT-Lancet Commission report on healthy diets from sustainable food systems.

⚠️ The most-cited attempt to ask what a diet compatible with both human health and planetary limits would look like. ⚠️ Read the dietary pattern it arrives at and compare it with the recommendations in Parts II and III of this book — that comparison is §36.9.

⚠️ And read the criticisms, which are substantive and come from several directions:

⚠️ On affordability — analyses finding the recommended pattern is beyond the reach of a large share of the world's population, which is Chapter 32's argument arriving at the report. ⚠️ On the modelling assumptions. ⚠️ On the appropriateness of a single global pattern given enormous variation in local food systems, agriculture and culture. ⚠️ And from livestock-sector researchers on the treatment of grazing systems on land unsuitable for crops.

⚠️ The convergence in §36.9 does not depend on EAT-Lancet specifically — independent national analyses reach broadly similar patterns — but the report is where the argument is most visible.

National dietary-and-sustainability analyses — ⚠️ several countries have run their own, and reading one for your own country is more useful than the global version.

3. ⚠️ Waste

UN Environment Programme Food Waste Index and FAO food loss and waste reporting.

⚠️ The source for the upstream/downstream split in §36.8, which is the most practically important fact in the section and is routinely flattened in reporting.

WRAP household waste research — ⚠️ as in Chapters 31 and 32, and it remains the best household-level source available.

4. Policy evidence

Evaluations of sugar-sweetened beverage taxes — ⚠️ multiple countries, and the key finding for this chapter is the share of the effect attributable to REFORMULATION rather than to consumer price response. ⚠️ Case Study 1 turns on it.

Salt reduction programme evaluations — ⚠️ particularly the countries that moved from voluntary to mandatory targets, because the comparison is informative.

Trans fat elimination — ⚠️ the clearest public-health win in this chapter, and worth reading as a model of labelling-then-banning.

Front-of-pack labelling evaluations — ⚠️ especially the Chilean and other Latin American warning-label programmes, which have been studied carefully and which changed both purchasing and reformulation. ⚠️ Chapter 30 §30.9's evidence.

Marketing restriction evaluations — ⚠️ and the displacement literature, which is why §36.6 gives 🟡 rather than 🟢.

⚠️ WHO and OECD policy reviews — ⚠️ useful for the intervention ladder in §36.7, and generally more measured than either advocacy or industry material.

5. ⚠️ How the system got this way

Histories of agricultural policy — ⚠️ the postwar productivity settlement, price supports, and how the commodity mix was determined. ⚠️ Read at least enough to see that the current structure was designed to solve a real and serious problem, and largely did.

Work on retail concentration and buyer power — ⚠️ including listing fees, slotting allowances and category management. ⚠️ This is the evidence for §36.5b, it is comparatively dry, and it explains more about what you can buy than almost anything else on this page.

Commodity trading and market concentration analyses — ⚠️ on the small number of firms handling a large share of traded staples.

Food security and supply-shock analyses — ⚠️ the 2007–08 and 2010–11 price spikes, and the more recent energy-and-conflict-driven disruptions, are the best-studied cases for §36.10.

6. Marketing and children

Systematic reviews of food marketing exposure and its effects on children — ⚠️ the ✅ verdict in §36.6 rests here, and the consistency across decades and countries is what makes it a ✅.

Digital and influencer marketing research — ⚠️ a newer and faster-moving literature, and the reason §36.6 says regulators were not built for the channel that now matters most.

7. True cost accounting

⚠️ Attempts to price food's externalities and add them to retail prices — ⚠️ including work by international agencies and by academic groups.

⚠️ Read these for the DIRECTION and the RANK ORDER, which are robust, and not for the totals, which vary enormously with method and which §36.13 explicitly declines to defend.

⚠️ Look specifically at which categories carry the largest externalities and how well those are reflected in price. That relationship is the chapter's closing point.

8. ⚠️ Read with the label attached

⚠️ Named as categories, because the material is useful and the framing is not neutral.

⚠️ Advocacy books and documentaries about the food industry. ⚠️ Frequently accurate on the mechanisms and frequently overstated on the conclusions, and they tend to reach for conspiracy where §36.2's answer — an optimization target nobody chose — is both more boring and better supported.

⚠️ Industry-funded sustainability and nutrition material. ⚠️ Chapter 3's framework applies: not dismissal, but attention to the denominator chosen, the comparison selected, and the claim made beyond the data. ⚠️ §36.3b is the tool.

⚠️ Anything asserting that individual choices are irrelevant, or that they are sufficient. ⚠️ §36.11 answers both, and both positions are more comfortable than the evidence.

⚠️ And carbon-footprint calculators for individual diets. ⚠️ The uncertainty is substantially larger than the interfaces admit, and §36.13's first row is why this chapter's Project Checkpoint asks you not to use one.

9. Where this goes in the book

⚠️ One recommendation, if you read nothing else

Read Poore and Nemecek's figures — just the figures — and then read the EAT-Lancet dietary pattern.

⚠️ Half an hour, both widely available, and between them they contain the whole of §36.9.

⚠️ And then notice that the pattern you have just read is approximately the one Chapter 35's personalization algorithms arrived at from the opposite direction.