Who are small-scale food producers in Italy? Comparisons among different approaches¹

1.Sustainable agriculture and the concept of small-scale food producer33

Sustainable agriculture is not a well-defined goal. Social, economic, and environmental sustainability are closely intertwined and necessary components for a truly sustainable agriculture.

Agricultural sustainability rests on the principle that we must meet the needs of the present without compromising the ability of future generations to meet their own needs. That is not an easy task for farmers. For example, farmers faced with poverty are often forced to mine natural resources like soil fertility to make ends meet, even though environmental degradation may hurt their livelihoods in the long run.

Several statistical indicators deal with sustainability: a key question is to assess how many of them can be regularly calculated by each country each year. For example, Gismondi [1] and Jouzi et al. [2] underlined that conversion to organic farming is one of the key indicators of sustainable agriculture, even if it is not the only one and it is not easy to ensure international comparability of organic farming figures.

In this framework, the need to focus on small-scale farming is due to many reasons. Small-scale farming systems are often more environmentally sustainable than large-scale. They are often more productive than other types of agricultural businesses, including industrial farms as well. Small-scale farms are especially critical for the food security and nutrition of vulnerable groups since these farms serve predominantly domestic and local markets. Small-scale farming also contributes to culture and community. Last but not least, increased revenues and profits among small-scale farmers and their businesses are often invested back into local economies, where they create jobs and equitable growth.

By the way, the Sustainable Development Goals (SDGs) were adopted by all United Nations Member States in 2015 as a call to action to end poverty, protect the planet, and ensure that all people enjoy peace and prosperity by 2030. The 2030 Agenda for Sustainable Development is organized into goals, targets, indicators, and sub-indicators. The 6 goals strictly concerned with the sustainable agriculture concept are 2. Zero Hunger, 5. Gender equality, 6. Clean water and sanitation, 12. Responsible consumption and production, 14. Life below water, 15. Life on land. There are 21 SDGs indicators for which FAO plays the role of custodian. FAO is a contributing agency for other 4 SDGs indicators as well. A highlight of the main indicators with the greatest gaps in country reporting is given by FAO [3].

In particular, goal 2 has these purposes: end hunger, achieve food security and improved nutrition, and promote sustainable agriculture. Indicator 2.3 has this purpose: by 2030, double the agricultural productivity and revenues of small-scale food producers, in particular women, indigenous peoples, family farmers, pastoralists, and fishers, including through secure and equal access to land, other productive resources and inputs, knowledge, financial services, markets and opportunities for value addition and non-farm employment. Finally, the sub-indicator 2.3.2 is defined as the average income of small-scale food producers (SSFPs), by sex and indigenous status.

A critical point is the definition of the term “small-scale food producer”. The FAO has proposed the definition described in Section 2. Using the data sources described in Section 3, we analyze the main findings in Section 4. The FAO has proposed the SSFP definition described in Section 2 so that each country can apply the same criterion to identify SSFPs, using baseline data that should be available in each country, even in years when there is no census. However, in each country, the definition of “small”, “medium” and “large” farms may depend on the level of agricultural development. Furthermore, other definitions of “small” farmers may result from the availability of additional indicators relating to the structural characteristics of farms and/or farm managers.

Definitions of “small-scale food producers” that are found in the scientific literature and policy documents are mostly based on four criteria: size of operated land, amount of labor input employed for agricultural production (especially of family members), market orientation, and economic size of the holding. Examples of alternative classification criteria for agricultural holdings are given in Sotte [4], Arzeni and Sotte [5], Italian Ministry of Agriculture [6], Davis et al. [7]. For this reason, in Section 5 we propose other methods to identify SSFPs and compare the results with those in Section 4. Section 6 contains some perspective conclusions.

2.The FAO definition of SSFP

The reference population includes food producers. They are farmers, herders, foresters, fishermen and aquaculture holders engaged in the production of food products. According to FAO [8, 9], small-scale food producers are producers who meet the following criteria.

3.Data sources for identifying SSFP

In Italy, the most updated identification of SSFPs is based on data from the last census of agriculture and the yearly survey on the economic results of agricultural holdings (FADN). Some details are described as follows.

4.Identification of SSFPs

Based on the census data, the utilized agricultural area (UAA), and the number of animals by species are available for each farm. The available data on livestock was used to calculate the number of adult livestock units (LU), which are representative of the livestock units reported by the FAO. The census did not collect data on farm revenues, but it allowed the SO to be calculated. For this reason, the method used in Italy is not exactly the same as that used by the FAO, as the economic size was estimated using SO instead of revenues. To ensure the robustness of the results, the following farms were not taken into account in the calculations: 1) with less than 0.2 hectares and less than 0.2 livestock units and/or 2) with a standard output = 0. Starting from the original figure of 1,133,023 active farms counted in the census, 1,102,729 farms were taken into account in the following elaborations, which represent the reference population.

Table 1 shows some preliminary results depending on the physical size of the farms. In the table, the virtual SSFPs amount is given when we consider the criteria based on UAA, LU and SO separately (numbers on the diagonal) or coupled (numbers not on the diagonal).

For example, looking only at the UAA criterion, the number of farms that fall into the bottom 40% of the cumulative distribution is 1,013,836, i.e. 91.9% of the total. In addition, the number of farms falling in the bottom 40% of the cumulative distribution of LU is 1,095,871 (99.4%) and the number of units falling in the bottom 40% of the cumulative distribution of SO is 1,065,860 (96.7%). These results confirm the high degree of concentration of land, livestock, and standard output in the left tail of the frequency distribution.

If we consider as “potential” SSFPs those farms that fall in the bottom 40% of the cumulative distribution for all three criteria, the number of units is slightly lower: 987,553 (89.6% of the total). Excluding the FAO’s other economic criterion, almost nine out of ten farms in Italy would be small holdings.

According to the FAO, producers with revenues of more than 34,387 purchasing power parity dollars per year should be excluded from the subgroup of small farmers. To take this restriction into account, we had to estimate the revenues. The basic idea is to estimate revenues based on standard production. For this purpose, we used the 2020 FADN data at the unit level.

The FADN data available until 2020 concern 10,764 agricultural holdings for which both standard output (independent x-variable) and revenues (dependent y-variable) are available. The basic idea is to obtain estimates of revenues through the linear model y=α+βx+ε. The main limitation of this approach is that the FADN survey does not include very small farms, e.g. those with a SO < 8,000 euros. To obtain reliable estimates, 19% of units were excluded as they are outliers; the final database used for the calculation therefore included 8,731 farms. The farms were broken down by their main agronomic orientation (AO), which indicates the main type of farming practiced by each farm. There are 8 AOs: 1) Arable farming; 2) Horticulture; 3) Permanent crops; 4) Grazing livestock; 5) Grain eaters; 6) Mixed crops; 7) Mixed livestock farms; 8) Mixed crops with livestock. For each AO, we used a specific regression model, as shown in Table 2. Even though the model performs better for some AOs (1, 5, 7, 8), the results are always satisfactory as all models are statistically significant.

Using the estimated α and β coefficients, it was possible to estimate the revenue for all census farms. Therefore, following FAO recommendations, within the resulting group of producers identified using the physical and economic criteria, producers with revenues greater than 23,883 euros were excluded from the smallholder subgroup (23,883 = 34,387 × 0.68, where 0.68 is the purchasing power parity coefficient in dollars applied to revenues expressed in euros).

The final results can be found in Table 3. In 2020, the number of small farmers amounted to 753,884, or 68.4% of the total. Results are available for specific profiles.

Even in 2020, there is a dramatic gap between the estimated average revenue of non-SSFPs and SSFPs: 69,115 euros against 10,038, the ratio between them is 6.9.

5.Alternative criteria for identifying SSFPs

According to the FAO methodology, SSFP is determined by the cumulative distribution of three indicators: 1) agricultural area; 2) livestock; 3) revenues (estimated by standard output), plus the fourth criterion given by the threshold applied to revenues. Is it possible to obtain similar results using alternative (and perhaps simpler) methods? Possible answers are suggested in the next subsections. The basic idea is to examine the possibility of determining the SSFP more simply, possibly using an annually repeatable methodology (see also Gismondi [15]).

6.Perspective conclusions

Why are SSFPs important? Though their larger, industrial counterparts may dwarf the land they work and the total quantity they produce, their impact on the world is anything but minimal: According to [16], farms smaller than 5 acres produce roughly 35% of the world’s food, and SSFP provide up to 80% of the food supply in Asia and sub-Saharan Africa. When smallholder farmers produce a higher quality and quantity of food, they can earn more revenues, better feed their families, and provide more food for the local marketplace, reducing prices and improving diets.

Of course, the importance of SSFPs may differ from country to country, and SSFPs tends to reduce especially in Italy, as the last 2020 agriculture census confirmed. Nevertheless, worldwide it is important to monitor how many there are and which are their main features.

In this paper, we have presented the main results obtained for Italy by using the main concepts and definitions recommended by the FAO for the determination of SSFPs. In addition, we have proposed three other calculation methods that represent an alternative to the FAO method. They could be equally simpler and could be applied in years other than the survey years. Further work is needed to better identify SSFPs and to assess whether it is possible to derive information on farmers’ revenues without using estimates based on standard output.

Agricultural censuses represent useful benchmarks to update the list of SSFPs in a country. However, the great challenge for next future consists in the strategy to be used by national statistical institutes for obtaining reliable data on land, livestock, and revenues of farms without carrying out traditional censuses anymore. Censuses are very costly and time expensive, while administrative sources may often supply the backbone of a permanent statistical system on agriculture including all basic statistical data at the single farm level. In this way, the Farm Register may include additional indicators (such as those related to revenues) and may be updated every year instead of with a ten-year time lag (the number of years between two traditional censuses).

The paper showed that the FAO methodology aimed at identifying SSFPs may be substituted by alternative and maybe simpler methodologies. Of course, these results are valid in the Italian context, and reasonable concerns are raised about the validity of this exercise beyond the Italian context. However, we must consider that nowadays yearly data on surfaces, livestock and SO are available in many countries. Even the concept of the poverty threshold (used for methodologies 3 and 4) is quite used in many countries, not only in Europe.

The paper is intended as a starting research point, from which other applications may try to check the validity of some conclusions for other countries as well. To replicate calculations for other countries, access to microdata is a mandatory key issue