Increasing agricultural productivity and protecting the environment: the same battle?
Last November, we reported here the results of a study by Kansas State University researcher Nelson Villoria, according to which, if global agricultural productivity had not increased between 1991 and 2010, it would have been necessary to increase agricultural land by 173 million hectares, the equivalent of about 10 % of tropical forests.[1].
The same researcher recently published a new article in which he explores the link between agricultural production efficiency and environmental protection.[2]. His estimates are based on a sample of 70 countries covering 75,000,000 of the world's agricultural area. According to his calculations, if agricultural productivity had not increased, in these countries between 2001 and 2010, agricultural area would have had to increase by 125 million hectares to meet the increase in food demand, whereas it actually decreased by 2 million hectares over the period studied. This would have resulted in additional greenhouse gas (GHG) emissions of, depending on the assumptions used, between 17 and 84 billion tonnes of CO2 equivalent, a level well above the additional 1 to 15 billion tonnes of CO2 equivalent actually released into the atmosphere.[3]Furthermore, half of the increase in agricultural land would have occurred in four of the world's most biodiverse regions.
Nelson Villoria warns that due to the slowdown in agricultural productivity growth in more than half of the countries considered, and taking into account the increase in demand, the area used for agriculture could increase by 67 million hectares over the period 2018-2023, which would reverse the decline observed between 2001 and 2010. Around 60% of the increase would take place in the most diverse biomes on the planet.
The search for continued growth in agricultural productivity therefore appears essential to mitigate climate change and biodiversity loss. However, it is important to clarify some of the results of this study so as not to misinterpret them and to understand their implications.
First, the productivity in question here is total factor productivity (TFP), defined as the ratio between the volume of agricultural production and that of all the production factors (land, equipment, inputs, labor) used to obtain it. TFP growth is equal to the difference between the growth in agricultural production and the growth in the total quantity of production factors used: it corresponds to the gain in production that cannot be explained by the increase in the volume of production factors, but is due to a greater efficiency of these factors or their combination (use of improved varieties, more efficient equipment, better organization of work, etc.). The increase in TFP, which is responsible for three-quarters of the increase in global agricultural production since the 1990s, should not be confused with that of yields, generally linked to a greater use of capital or labor.
Unfortunately, the increase in total factor productivity is difficult to measure. In addition to the frequent lack of reliable data, there are several methods for quantifying it, which are not unanimous and produce different results. It is therefore interesting to note that the improvement in TFP is closely correlated with that of yields – or more precisely with the increase in land productivity, defined as the value of agricultural production (plant and animal) per hectare, which also takes into account cropping intensity (number of harvests per year). According to our estimates, based on data published by the United States Department of Agriculture (USDA), over the period 1961–2016, the correlation between the growth in TFP and that of the value of agricultural production per hectare, at the global level, is very strong (with a correlation coefficient r = 0.98), which suggests a linear relationship between these two indicators (Figure 1).
This linear correlation is very robust in all regions analyzed by the USDA (r between 0.90 and 0.99), with the exception of the former USSR where, for reasons that remain to be clarified, it is low (r = 0.17). This means that, in all regions except the former USSR, when agricultural productivity per hectare increases, it is extremely likely that TFP will also increase (keeping in mind, however, that a correlation coefficient between two variables, however high, does not necessarily indicate a causal link between them). From a practical point of view, this observation gives full meaning to the search for an increase in yields. But it should also be noted that the rate of increase in TFP differs significantly from that of the increase in production per hectare. Overall, since the beginning of the 1960s, total factor productivity has increased about half as fast as land productivity (Figure 2).
Second, the link between the increase in TFP and the evolution of agricultural land is complex. Two effects combine. On the one hand, in a given country, the fall in prices induced by the increase in agricultural production due to the growth in TFP in this country can be more than offset by the rise in demand, if the latter is sufficiently elastic (in other words, if it increases significantly when prices fall). This is particularly the case for countries that export a significant share of their production and thus benefit from the increase in foreign demand: they then tend to increase their agricultural area. On the other hand, this "rebound effect" is more or less mitigated by the impact of the fall in prices caused by the improvement in TFP in other countries; on the contrary, this favors a reduction in the agricultural area in the country in question. Overall, taking these two effects into account, over the period 2001-2010, the area used for agriculture decreased in the United States and Europe, but increased in Latin America, Southeast Asia and sub-Saharan Africa.
Finally, the choice of increasing total factor productivity, rather than increasing yields, as the central indicator of the efficiency of the agricultural system, sheds new light on the debate between agriculture and the environment. Traditionally, this debate pits supporters of the intensification of agricultural production, that is, an increase in land productivity, against those in favor of agricultural extensification. For the former, increasing yields makes it possible to reduce cultivated areas or limit their expansion (land sparing), which promotes the conservation of ecosystems richest in biodiversity as well as the reduction of GHG emissions due to the clearing of grasslands and forests. For the others, a slowdown in growth or even a reduction in yields across the entire cultivable area (land sharing), thanks to a reduced use of fertilizers and plant protection products, maintains greater biodiversity in cultivated agro-systems and leads to a reduction in GHG emissions linked to the manufacture and use of synthetic inputs.
If we base our discussion on the improvement of TFP, the debate is no longer about the choice between land sparing and land sharing, but about the different avenues that can be explored to best reconcile agricultural production and the environment – even if, as we saw above, the increase in total factor productivity and that of land productivity are strongly correlated. In a recent article[4], Canadian and American researchers propose a distinction between two approaches, one "technology-based" and the other "ecosystem-based." The first approach is based on four pillars: genetic selection, equipment and data, and precision agriculture. The second, which we could call agroecological, prioritizes pollination management, biological pest control, the integration of crops and livestock, as well as crop rotations and soil conservation. Each of these trajectories influences TFP in a specific way; each has different implications for the environmental sustainability and resilience of agricultural production systems, and each induces distinct costs for society. One can, of course, ask whether they are completely mutually exclusive or whether, given the diversity of local situations, a hybrid path, borrowing from both models, is possible.
The authors of the article are careful to point out that while growth in total factor productivity is beneficial because it reflects the fact that resources are being exploited with a concern for efficiency, it is not, in itself, a sufficient condition for judging the social or environmental acceptability of agricultural practices. The differentiated impacts of these practices depending on the economic size of farms, ethical considerations linked to production methods, etc. also come into play. In other words, whether we take growth in yields or growth in TFP as an indicator, public policy choices, particularly in terms of agricultural research and development, remain decisive.[5].
[1] Jean-Christophe Debar, “Increasing yields, declining deforestation?”, March 19, 2019, https://fondation-farm.org/rendements-en-hausse-deforestation-en-baisse/
[2] Nelson Villoria, 2019. “Consequences of agricultural total factor productivity growth for the sustainability of global farming: accounting for direct and indirect land use effects”, Environmental Research Letters, Volume 4, Number 12.
[3] The low estimate of the GHG emissions range corresponds to the assumption that cropland expansion occurred at the expense of grasslands. The high estimate corresponds to the assumption that this expansion resulted from deforestation.
[4] Olivier T. Coomes, Bradford L. Barham, Graham K. MacDonald, Navin Ramankutty and Jean-Paul Chavas, 2019. “Leveraging total factor productivity growth for sustainable and resilient farming”, Nat. Sustain. 2 22.
[5] Christian Huygue, “Agricultural production and environmental preservation: is it possible?”, Science et pseudo-sciences no. 331, January-March 2020.
