We focus on the supply chain of corn and soybeans, two of the most important crops that are used to produce animal meat such as beef, pork, and poultry. We study the long-term permanent effects of crop demand reduced by the displacement of animal meat, therefore, we disregard storage that may smooth price variations in the short run but not on a 10-year horizon. In line with the evidence in the study by Merener (2015) we assume that the markets for corn or soybeans are largely globally integrated. Table 1 displays statistics for corn and soybeans projected in OECD/FAO (2023) for 2032. This is the horizon for the risk assessments of our study. Most of the global soybean output is crushed and transformed into soybean meal, at an approximate conversion ratio of 77% of bean weight and soy oil. Roughly 57% of corn and 70% of soybeans (in the form of soybean meal) are projected to be used as animal feed, or as input for feed. These utilization ratios are similar to those reported by the OECD-FAO for 2020–2022. Crucially, figures in Table 1 are projected by the OECD-FAO under the assumption of no significant AltMeat growth. Therefore, we will use these estimates as the baseline case over which we will impose crop demand reductions assumed to be induced by AltMeat growth. The starting point of our modeling work is a decomposition of the world in a collection of i = 1, ..., N regions in Table 1 that are either leading crop exporters or major crop importers.

As in the simplest model considered in the study by Baldi and Merener (2023), we assume a global crop demand reduction defined in terms of the animal feed no longer needed after an animal meat consumption reduction of a certain magnitude has been induced by AltMeat. We do not model the link between AltMeat growth and animal meat displacement. Local demand reductions occur on top of a baseline demand and are driven by consumer choices. As such, they might occur in exporting and importing regions. In this study we focus on meat consumption in the United States and the European Union (EU), which, by their living standards and disposable income, are likely to lead to the adoption of environmentally friendly innovations (Pfeiffer and Mulder, 2013; Elgin et al., 2023), and the world as a whole. Animal meat market statistics projected by OECD-FAO for 2032 under the assumption of no significant AltMeat growth are reported in Table 1.

The crop demand reduction is a function of animal meat displacement and feed utilization parameters in Table 2. The tonnage of corn used to produce beef consumed in region i in the baseline scenario is M_beef,ii_beeff_beefc_beef. Expressions for other animal meats or crops are analogous. In the case of soybean, we adjust by a factor of 1/0.77 because we work with soybeans in our computations, but the actual feed mix usually contains soybean meal. The decrease in crop demand in region i is determined by exogenous animal meat replacement shares {d_beef,i, d_pork,i, d_poultry,i}. While AltMeat growth could decrease corn and soybean consumption by lowering animal feed demand, it could also increase demand for crops that are used as a direct input to plant-based meat. This is currently the case for soybeans. We include this effect in our analysis through parameters r_c,PBM and r_s,PBM although its size is likely small. Corn and soybeans would not be required as inputs in cultured meat, therefore, the impact of a sustainability transition driven purely by this type of novel meat would be slightly larger than the estimates reported in this study. In net, the exogenous crop demand shift associated to meat displacement in region i is

An exogenous demand reduction caused by animal meat displacement would be expected to induce a price decrease and, therefore, lead to extra demand for other crop uses. For instance, increased biofuel production or higher human consumption. These effects are captured by the elasticity of demand, γ < 0. Let P⁰ and P be global crop prices under the baseline OECD projection without AltMeat and after the adoption of AltMeat, respectively. Baseline and post-AltMeat demand are D⁰ and D(P). Under a linear supply and demand model, appropriate for small changes in crop markets, global demand is

Table 3, summarized from Baldi and Merener (2023), displays approximate global estimates for animal meat production and feed crop utilization to implement Equations 1 in 2. Industrial methods are more prevalent in poultry and pork than in beef, where grass, forage or mixed systems are widely used. Poultry uses the smallest amount of feed per kg of meat, and beef is the least efficient in this regard. Estimates of the proportion of corn and soybean in the feed of ruminants, pork and poultry, show significant variation across animal species. Information on the composition of plant-based meat products is very scarce. Among some of the best-known products, Beyond Meat's burger uses proteins from peas and beans. Impossible Foods includes soy and potato protein, among other inputs. Approximately 50% of soybean mass is protein and the protein weight in a typical plant-based meat burger is about 20% of its total weight. Therefore, soybean mass per kilogram of plant-based meat ranges between zero and 0.40 kg. This is an order of magnitude smaller than the mass conversion ratios from animal feed to animal meat in Table 3.

It is important to recognize that exporters and importers in Table 1 are likely to behave differently in response to a demand reduction. In particular, transportation costs from distant sources, and strategic and political considerations in the EU and China, two main corn and soybean importers, suggest that a decrease in the global demand for crop and soybeans may be absorbed first by a reduction in imports rather than a reduction in local production (Peters et al., 2009, Bellmann and Hepburn, 2017). To explore the potential implications of different regional adjustment policies on the impact of AltMeat, we consider three alternative crop supply structures. In all of them we use a global estimate for the elasticity of crop supply, β, from the literature. Crop supply structures differ on the regional attribution of global supply adjustments. Let Si0 be the supply in region i in the absence of AltMeat and S_i(P), S(P) be the regional and global crop outputs, respectively, under AltMeat adoption.

Supply structure 1. In this case, we assume that all producers respond to reduced demand in the same manner, regardless of their exporting/importing status. Under a linear model, we have

Supply structure 2. Here we assume that the adjustment of global supply falls entirely on large exporters, while importers and balanced regions maintain their output at their pre-AltMeat levels. For global output to be elastic to prices as β, exporters must be more responsive than in the supply structure 1. Linearizing we have

Supply structure 3. In our third supply specification, we focus on the fact that there is significant heterogeneity in profit margins across crop-producing regions. Supply from crop-exporting regions responds to changes in profitability induced by changes in global demand. The price per ton of crop perceived by local farmers in region i is P_i and may differ from the global market crop price P due to the existence of subsidies (as in the United States in certain scenarios adverse for farmers) or taxes (as currently in Argentina). We assume that this effect is captured by a factor I_i so that P_i = I_iP. Production cost per ton in each region is the sum of a fixed and variable component. We aim to capture the notion that certain inputs, such as labor, fuel, and machinery, machinery are generally uncorrelated with crop prices, while fertilizer, seeds, and land rental are strongly and positively correlated with local crop prices (Schnitkey et al., 2021a). Hence the cost per ton in region i is

for positive η_i and ψ_i that represent the fixed and variable cost components. Farmer economic profit per ton is I_iP − C_i(P). Under a local supply function that is linear in local profits, while holding the global elasticity of supply equal to β, exporter supply can be approximated as

In this specification we continue to assume that output in importing regions is unaffected by a relatively small reduction in global demand so that Simporter(P)=Simporter0.

Baseline levels of crop output for producers are in Table 1 and are sufficient to implement supply structures 1 and 2 in Section 2. Supply structure 3 depends on local profitability, determined by the price received by farmers and their cost of production. Farmers' income depends on taxes, subsidies, and transportation costs to main destinations. In Table 4 we report regional revenue and costs in USD per ton for 2015–2019. It is important to note that this was a period of moderate commodity prices by historical standards, after the end of the supercycle in 2014 and prior to the run-up in prices after the COVID-19 pandemic. Average CME prices during this period were 145 and 340 USD/ton for corn and soybeans, respectively.

It is standard practice to differentiate between direct costs, operating costs, and overhead. Up to 90% of direct costs are fertilizers, seeds, and pesticides. A primary input for fertilizers is ammonia, which had a 74 % correlation with corn between 2008 and 2021 (Schnitkey et al., 2021b). Overall direct costs in Illinois had a 68% correlation with corn prices between 2000 and 2021 (Schnitkey et al., 2021a). Operating costs include labor, machinery, repairs, and fuel. These items are much less sensitive to crop prices than direct costs. Overhead costs include land, building depreciation, and insurance. Land costs can be expressed as a ratio of expected future farmer income, itself a function of current crop prices, and long-term interest rates. This standard theoretical construction diverged from market prices in recent years in the US when interest rates approached zero. Following widespread inflation and interest rate increases in the aftermath of the COVID-19 pandemic, land costs are once again in closer agreement with the theoretical prediction (Schnitkey et al., 2021a). Colussi et al. (2023) documented that croplands in Brazil increased 128 % on average from 2019 to 2022, as corn and soybean prices received by local farmers almost doubled in the same period. In sum, we will assume in our computations that direct costs and overhead are highly correlated with crop prices while operating costs are not. In some cases discussed in Table 4, net farmer profits are negative. However, as discussed in the studies by Saavoss et al. (2021) and Vaiknoras and Hubbs (2023), this has been often the case historically due to the inclusion of the opportunity cost of land and unpaid farmer labor in the computation of cost measures. Since farmers are often not paying rent for their land nor wages for their own labor, negative net profits do not imply that farms have negative accounting profits. We translate regional supply information provided in Table 4 to the parameters used in the supply function in Equation 6 through the following algebraic identities

Calibrated values for these parameters are in Table 5.

For any supply structure specified above, the market impact of animal meat displacement is given by the change in crop price, ΔP = P − P⁰, that leads to the satisfaction of the market clearing condition

The collection of local supply functions S_i(P) incorporates information on regional crop production as expressed in Equation 7. Exogenous demand reductions Z_i combine in Equations 1, 2 the magnitude of the animal meat displacement induced by AltMeat. In addition, Equation 7 depends on the elasticities of supply, β, and demand, γ. For these, we rely on supply and demand elasticities for corn, soybeans, and calories, as discussed in the study by Roberts and Schlenker (2009, 2013). Using global data from 1960 to 2007, these studies combined yield variations, weather shocks and variations in inventories to obtain a wide number of elasticity estimates. Roberts and Schlenker (2013) aggregated corn and soybeans by their caloric content given their partial substitutability and complementarity as sources of animal feed. We take the sharpest estimates from the study by Roberts and Schlenker (2013) for calories, which are more strongly identified than those for individual crops, and apply them to corn and soybeans individually. We rely on the notion that very unequal price impacts on corn vs. soybeans would be limited by substitution in the short run and by planting decisions in the long run. The average values for the estimates provided in the studies by Roberts and Schlenker (2009, 2013) are β = 0.11 for the elasticity of supply and γ = −0.05 for the elasticity of demand. In our numerical experiments, we solve Equation 7 for crop price and regional output changes.

We follow the approach discussed in the study by Baldi and Merener (2023) and propose a plausible animal meat displacement scenario based on quantitative and qualitative information. Our results in this study, should be understood as risk assessments conditional on animal meat displacement scenarios rather than precise forecasts for which there is no sufficient information. Acceptance for plant-based meat grew strongly until 2021, and more so among those who are educated, younger, more liberal, from urban areas, and concerned about animal treatment (Slade, 2018; Godfray et al., 2018; Hartmann and Siegrist, 2019; Bryant et al., 2019; Van Loo et al., 2020; Szejda et al., 2021; Neuhofer and Lusk, 2022; Zhao et al., 2022; Taylor et al., 2023). This suggested a gradual, relative decrease in animal meat consumption led by aging and generational change. Forecasts published prior to the pandemic for 2030 plant-based meat market share were in the 2–10% range. This was consistent with a 2020 share in the 1–2% range and subsequent annual growth rate between 6 and 18%. In recent years, plant-based meat made inroads in becoming relatively more affordable, but it is still a premium product. However, plant-based meat sales have stagnated since 2021, as budget-conscious consumers have been reluctant to embrace it during the inflation peak that followed the COVID-19 pandemic¹. Interest among investors and funding for plant-based startups rose dramatically between 2017 and 2021 and cooled off since 2022 in line with the general retrenchment of investment on risky disruptive technologies induced by higher interest rates².

Cultured meat is produced from animal cells through sophisticated lab-based processes (Hocquette, 2016; Treich, 2021). While energy-intensive to produce (Lynch and Pierrehumbert, 2019), cultured meat is projected to have some environmental benefits (Mancini and Antonioli, 2022). Cultured meat is being evaluated in Europe (Lanzoni et al., 2024) and has already been approved for sale in the United States, Israel, and Singapore.

The alternative dairy market is a relevant medium term-growth benchmark for the AltMeat market. Animal milk consumption in the United States has been replaced in part by plant-based milk (Wolf et al., 2020). The American plant-based milk category accounts for about 14% of the dairy milk market. In our numerical experiments, we will consider a potential scenario in which AltMeat will induce the displacement of 3% of the animal meat consumption by 2032 relative to the current OECD-FAO projections, which are, in turn, 14% larger than the animal meat consumption in 2019–2021. We will also consider displacement scenarios for individual types of meat, and for specific regions. Based on cost and consumer preferences, it could be expected for AltMeat to be adopted first in highly developed regions as the United States and Europe, where the population may have more disposable income to allocate toward environmentally friendly products. We stress that this research study is not a forecast of AltMeat growth but an assessment of its potential impact on crop markets.