This paper, like many, is the product of a journey that began many years ago with a conversation. I owe it to Susanna Klassen[1], a then undergraduate student at the McGill School of Environment, for planting the seed back in Nov 2013.
I was meeting with her to discuss potential research opportunities in my group. At one pointSusanna mentioned her interest in studying smallholder agriculture. She had read that smallholders produce 70% of the world’s food and wondered how that was known and what the implications were. I was astounded by that statistic. My work involves mapping the world’s agricultural land use, and knowing that nearly 50% of the world’s land is in wheat, maize, rice, and soy, I couldn’t imagine how this could be possible. While there are many small-scale cereal producers, especially with rice, major cereals today are grown on large scales. On the other hand, as Susanna and I discussed then, maize and soy are mainly animal feedcrops, and many rice producers in Asia are smallholders, so maybe it’s true that food crops are mainly produced by smallholders.[2] I looked into the original source of the statistic and realized that it was a back-of-the-envelope estimate that could be improved upon with more data. Given that my research group prides itself on pulling together global data on the world’s agriculture, this looked like an interesting question to examine. It is also an important question to examine -- more than 80% of the world’s farmers are smallholders and they constitute a large portion of the world’s poor and vulnerable populations. Further, some argue that smaller farms are also more sustainable.
Fast forward to summer 2015, at the University of British Columbia in Vancouver. I was having lunch with my UBC colleagues, Dr. Hannah Wittman, a rural sociologist studying sustainable food systems, and Dr. Jeanine Rhemtulla, a landscape ecologist studying the ecology of working landscapes. I happened to mention my interest in the smallholder agriculture question. We all got excited and decided we would apply for a grant from the Social Science and Humanities Research Council (SSHRC) of Canada. We succeeded in receiving a 5-year grant starting Apr 2016 to study “The role of small farms in global food security and sustainability”.
Enter Dr. Vincent (Vinny) Ricciardi. Vinny started a PhD program at UBC under my supervision in Sept 2015. He came with several years of experience as a development consultant in South East Asia and West Africa working on issues of interest to smallholders. When we got the SSHRC grant in 2016, Vinny was a natural fit for the project, expressing interest in leading some questions in the proposal as part of his PhD research.
This paper started as a systematic review Vinny wrote as part of his PhD comprehensive exam in summer 2016. Vinny expanded on that work further, to conduct a quantitative meta-analysis of farm size and its relationships to various environmental and social outcomes of interest. The bulk of the empirical work was done by Vinny and Dr. Zia Mehrabi, a Research Associate in my group at UBC, while Hannah Wittman and UBC PhD student, Dana James, contributed to interpretation of results and writing. I supervised the project from research design to writing.
Our paper quantitatively summarizes the literature on what is known about the relationships between farm size and different indicators of production, environmental, and socio-economic performance. Our review of the literature indicated that there were enough literature samples to examine farm-size relationships with yields (measured either in terms of weight or crop value), non-crop biodiversity, resource-use efficiency, and farmer profitability. Our initial search resulted in 1474 studies, of which 118 met our inclusion criteria and yielded 318 observations for analysis. For crop diversity and greenhouse gas emissions (GHG), there were too few literature samples to conduct a meta-analysis. We therefore simply reported on a previously conducted analysis of crop diversity relationships with farm size based on census and household survey data. We also gathered life-cycle analysis (LCA) estimates of GHG emissions reported in another study of the environmental impacts of food systems.
We conducted 3 types of analysis. First, for yields, biodiversity, resource-use efficiency and profitability, we examined “vote counts”, i.e., how many studies reported that smaller farms improved these outcomes versus not, and used hierarchical cumulative link models to synthesize these in probabilistic terms, i.e., what’s the probability that a study will find that farm size will have a particular outcome? Second, for a fewer set of studies for yields, resource-use efficiency, and profitability, where the studies reported regression coefficients for farm-size relationships, we conducted hierarchical meta-regressions to calculate pooled effect sizes summarizing the magnitude of change in outcome per 1 ha change in farm size. Pooled effects sizes were also calculated for GHG emissions using robust linear mixed-effects models.
So, what did we find? 79% of studies reviewed reported that smaller farms have higher yields, with yields increasing 5% per 1ha decrease in farm size. This is not an altogether surprising result – the inverse farm size-productivity relationship has fascinated agricultural economists for nearly a century, first reported by Chayanov in 1926. We also showed in a previous study that smaller farms have higher crop species richness; in that study we further found that smaller farms and larger farms grow a different portfolio of crops, and therefore, crop diversity at landscape scale may best benefit from having a mix of farm sizes. Our study also found that non-crop biodiversity increases with decreasing farm size, with 77% of studies finding that smaller farms have greater biodiversity at both farm and landscape scales. The primary studies we reviewed suggest that this is because of more ecological management practices on smaller farms (lower pesticide use, organic management, etc.), edge effects (smaller farms having larger margins that support biodiversity), or more diverse land cover in small-farming dominated landscapes. Our study found no statistical evidence for farm-size relationships with resource-use efficiency, greenhouse gas emissions, or profitability. We should note that we do not assess causality directly, i.e., that farm size is driving all the outcomes we reviewed (thanks to a reviewer for helping us clarify this in our paper). However, with both yields and biodiversity, we do believe there is strong theoretical support for a causal relationship as evidenced in the primary literature we reviewed.
Our summary of the evidence suggests that smaller farms have benefits for both productivity and biodiversity. This alone would suggest the need for increased policy support for smaller farms. But furthermore, there are nearly 500 million small farming households around the world, the majority of whom are found in low and middle-income countries, and who constitute some of the world’s poorest populations. Thus, smallholders are also the target of development support, with SDG 2.3 aiming to doubling the productivity and incomes of small-scale producers by 2030. Clearly, while smaller farms have benefits, smallholders are finding it challenging to make a living from agriculture. Supporting them would satisfy both humanitarian and environmental sustainability goals.
[1] Susanna Klassen is currently a PhD student in the Resources, Environment and Sustainability program at the University of British Columbia. Until July 2014, I was a professor at McGill University, in Montreal. I moved to the University of British Columbia in July 2014.
[2] We have since estimated that farms less than 2 ha produce 30-34% of the world’s food (Ricciardi et al., 2018; https://doi.org/10.1016/j.gfs.2018.05.002).
