Since the early 1960s, global crop production has increased by over 250%. The social, economic, and environmental ramifications of this growth in output—including the loss of biodiversity and carbon—are hard to overstate. To understand what lies behind this remarkable development, in 2020, Luke Yates and Barry Brook of the University of Tasmania and I, ran an analysis that breaks down the overall increase in crop production into five components, or drivers.
We did the breakdown, or decomposition, in two stages. In the first stage, we found that only 11% of the overall increase in crop production (including crops used for food, feed, fiber, and fuel) came from an expansion of cropland area. The bulk of the growth, 89%, came from growth in what we call aggregate yield—total crop output per unit area per year (Figure 1). Increasing yields is not without its downsides, but overall this is good news since it means that less natural habitat had to be converted to farmland. Other research (including papers by Thomas Hertel et al and Nelson Villoria) suggests that without yield improvements, several hundred million hectares of land would have been taken over by cropland.
Aggregate yield alone, however, is a blunt measure of progress in farming. Growth in aggregate yield does not necessarily imply that farm management or agricultural technology has improved. This is because there are many different ways that aggregate yield could increase over any given year, including through farmers switching from low-yielding to high-yielding crops and through crop production shifting to countries with on average higher yields.
To address this shortcoming, in the second stage of our analysis, we broke down aggregate yield into four contributing drivers, which we call pure yield, cropping intensity, country share, and crop composition (Figure 2).
Pure yield is defined as the output of a given crop per unit of harvested cropland area in a given location. It’s as specific and narrow a definition of yield as you can get. It could, for example, be the number of bushels of soybeans produced per hectare in a given harvest in Brazil. Pure yield pinpoints the ability of farmers to leverage technology to produce as many crops as possible on as little land as possible, and it contributed 63% of the overall increase in aggregate yields, with much of this progress coming from high- and upper-middle-income countries.
The remaining 37% of aggregate yield is divided among the other three factors. The first one we looked at is cropping intensity, or the average frequency per year with which each hectare of cropland is harvested. For example, planting and harvesting two crops of rice per year would imply a cropping intensity of 2, whereas leaving a field fallow every other year would imply a cropping intensity of 0.5. Increased cropping frequency contributed 23% of the increase in aggregate yield, a finding that aligns with other research pointing to an important role of this factor in expanding crop output.
The last two factors are much more seldom considered as drivers of global aggregate yields. But they should not be ignored because in contrast to pure yield and cropping frequency, they don't necessarily have anything to do with farming practices or technology, but rather broader phenomena like diets and trade.
First, we have crop composition: the proportion of cropland dedicated to different crops in each country. Shifts in crop composition drove up aggregate yields by 19%. This means that the world overall shifted towards relatively higher-yielding crops, especially sugar cane, palm oil, corn, and soybeans. A larger share of high-yield crops means greater output per hectare, even if the pure yields of all individual crops remain the same. And the overall increase hides substantial variation across the world: many high-income countries actually saw their crop composition shift towards lower-yielding crops.
The final factor is what we call country share, which represents the geographic distribution of cropland. Between 1961 and 1985, this component had a net-zero contribution to global crop production, but after 1985, this factor reduced aggregate yields to such a degree that it offset 4% of the positive contributions from other factors. What this means in practice is that, in the last few decades, crop production has generally shifted to countries with lower yields. Shifting the location of crop production comes with an environmental cost since the areas in which production is growing—especially low- and middle-income countries in the tropics—often have much higher levels of biodiversity and more carbon stored in soils and forests.
Beyond explaining the factors underlying increased crop production in the past, our study also provides hints about the future. For example, it’s possible that the global crop mix will shift towards lower-yielding crops, as we show it already has in high-income countries. Similarly, crop production might continue shifting toward countries with lower yields. Both of these trends could be a drag on global aggregate yields, meaning it would take more cropland expansion to meet future crop demand.
There’s also evidence from other studies that some parts of the world are seeing pure yields stagnate, possibly as a result of crops getting close to their upper yield limits: what you could get under ideal environmental and management conditions. At that point, there is no currently practiced combination of agronomic techniques or inputs that could raise yields. There are probably other factors contributing to the yield stagnation, and yield limits could be raised through crop breeding or biotechnology. The stagnation is nevertheless a cause for concern especially because recent evidence points to a slowdown in overall productivity growth in global agriculture.
All in all, keeping up the historical trend of getting most of the increase in crop production from aggregate yields rather than cropland expansion is going to be challenging. The stakes are high: further encroachment of cropland onto natural habitats risks worsening carbon emissions and biodiversity losses. Heading off such a future will require a broad suite of interventions ranging from technological innovation to extension services and infrastructure.