The idea of working with small-scale producers, also known as small-scale farmers, is ubiquitous in the development literature, so much so that the term made it into SDG target 2.3: “By 2030, double the agricultural productivity and incomes of small-scale food producers…” However, an exact definition of what is meant by a small-scale producer remains elusive. Imprecise definitions risk failing to target the vulnerable farmers who actually need development interventions or who could catalyze agricultural transformation.  Intuitively, you might picture a farming family with a small plot of land, and a few livestock, living primarily off of what they grow with their own labor, selling any excess on the side, likely living under the poverty line.  Yet, this image includes several indicators which are not always correlated and the world often presents greater heterogeneity than we might anticipate a priori.  This first of two blogs posts explains why this seemingly intuitive household is difficult to pin down, both conceptually, and by the literature.  In the second post we look at some tools EPAR has developed to better define small-scale producer households.  (Note, we prefer the term small-scale producer (SSP), to smallholder because it is more inclusive of different agricultural livelihoods and does not imply land tenure as many farmers rent or lease land).

Why Defining Small-scale Producers Is Difficult

To illustrate this challenge, I apply several potential definitions to farms I worked with in The Gambia. First, take a simple definition of small-scale producer based solely on land size.  If you have less than two hectares of farm land, you are a smallholder.  However, my friend Mustapha Bah’s family has only a small, 1ha plot of vegetables, and a herd of 50 cattle.  This family would be a classified as a small-scale produce under this definition, though they clearly do not fit the initial description.  Perhaps we should add a livestock limit to our definition, such as the requirement that you must own less than 4 TLU (tropical livestock units).  Yet, what of my friend Kaddy Ceesay’s family which has no livestock and a small 1.5ha plot of land, but who grows only high value cash crops, and who sells all of the produce for a substantial profit?  Again, there is some way in which her household does not fit into our preconceived notion of what a small-scale producer is, because they do not eat what they grow. 

To resolve this dilemma, we might add another indicator, that a small-scale producer must consume a certain portion of their harvest, say two thirds of their crop value.  And yet, there remain households which we may include in such a definition who defy our original formulation of small-scale producers.  Take Baba Senghore’s family that has a small, 1ha rural plot of land where they grow millet, most of which they eat, three chickens and a horse (0.53 TLU), but receives 80% of their income from Baba’s work as a tailor in the village.  To address this concern, we may want to limit a family’s income diversification, to be a smallholder, you must receive at least a certain portion of your income, say 50%, from farming.


Before continuing, some quick terminology.  An indicator is one of the criteria which determines whether a farm is considered a small-scale producer or not.  In the preceding examples we looked at land size, livestock holdings, proportion of crop value sold, and share of non-farm income.  A threshold is the cutoff for a particular farm falling into a category or not.  In the previous example we had 2ha, 3 TLU, a third of crop value sold, and half of income from non-farm sources.  A small-scale producer definition is some combination of indicators and thresholds, along with appropriate Boolean operators (whether you are considered a small-scale producer if you meet one of the indicators, or must meet all of them).  Therefore, a definition might be: a small-scale producer farm household is one with less than 4 hectares which sells less than half of its harvest.  The indicators used are farm size and proportion crop value sold, the thresholds are 4 hectares and 50%, and the operator is “AND” (to be a small-scale producer you must satisfy both criteria, if you fail to satisfy either you are considered a largeholder).

We also note that thresholds can be measured in multiple ways.  The thresholds listed above are all absolute thresholds which remain the same from year to year and country to country, even as farms change.  There is another way of defining a threshold which uses a relative measure.  This might be something like, farm size less than the median farm size, or less than the 40th percentile of livestock.  Relative measures may account for cross country differences, but make it more difficult to demonstrate across the board progress.

We also make a distinction between a definition which categorizes households into either smallholders or largeholders, and a more nuanced typology.  As we use the term, a typology is similar to a definition but it divides a population in multiple groups each with different characteristics.  Baba's and Kaddy's households might both be smallholders based on an asset focused definition (say less than 2ha and less than 2 TLU), but we might consider Baba's household a diversifying household (as it is bringing in non-farm income), whereas Kaddy's household might be a commercializing household (as it is selling most of what it grows). 

Literature Review

In order to determine the definitions in the existing literature, we systematically review journal articles from the last year from the top ten agricultural economics journals which mentioned small-scale producer farms, small-scale farms, or a range of other synonymous terms.  Of the 49 articles found, only eight offered any definition of small-scale producer.  Of those eight, six used an absolute measure of land size, one used proportion of family labor, one used a relative measure of land size and one used dry sheep equivalent (an indicator which tries to take into account the quality of the land defined as the amount of land required by a two-year-old 45-50 kg Merino Sheep (non-lactating and non-pregnant ewe) to maintain its weight). 

Pie charts showing proportion of articles showing smallholder farmers

Of those six studies which did use an absolute value for land size, the actual thresholds were quite varied, the lowest being 2 hectares, the highest being 28.3 hectares.  No two studies used the same threshold.  The combination of the majority of articles failing to define the term, and those that do define it, disagreeing by an order of magnitude on its definition suggests there is no consensus definition on what we mean when we say small-scale producer. 

There are also a range of different typologies to be found in grey literature.  AGRA offers a typology of farms that are between zero and four hectares on the basis of income diversification (share of non-farm income) and market orientation (share of production value sold).  Using this typology, they separate farms into five categories: subsistence farms, precommercial farms, specializing commercial farms, transitioning farms, and diversified commercial farms.  Mellor also offers a similar typology focused just on market orientation. 

Figure 2. AGRA Small-scale Producer Farm Typology

AGRA Small-scale Producer Farm Typology showing importance of farm versus non-farm income

In Part 2 of this series we cover the visualizations that EPAR has created to investigate this problem.  These visualizations provide information about the consequences of choosing a particular definition of small-scale producers.  They track different sections of the typology over time to determine how farms move between these categories.  They illustrate the year over year variability of the share of smallholders in a country if particular criteria are set at particular thresholds.

By Terry Fletcher

Summarizing original EPAR research by Didier Alia, C. Leigh Anderson, Travis Reynolds, Pierre Biscaye, and Terry Fletcher