defining smallholders (clara aida khalil, fao)
TRANSCRIPT
Criteria to define smallholders and their implications
RuLIS Expert Consultation: 08 November 2016
Aida KhalilFAO Statistics Division
Background
Need for an international definition of smallholders:
Target 2.3: “By 2030, double the agricultural productivity and incomes of small-scale food producers, inparticular women, indigenous peoples, family farmers, pastoralists and fishers, including through secure andequal access to land, other productive resources and inputs, knowledge, financial services, markets andopportunities for value addition and non-farm employment”
FAO proposed as custodian agency of indicators:2.3.1: The volume of production per labour unit by classes of farming/pastoral/forestry enterprise size2.3.2: The average income of small-scale food producers, by sex and indigenous status
Both indicators are classified as Tier III : lack on an internationally agreed methodology
Pre-requisite: International harmonized definition of “small-scale food producers”
Smallholder agriculture is one of RuLIS “qualifiers”: indicators are presented separately for smallholder andnon-smallholder farms
Outline
Characteristics of a workable definition
Overview of existing definitions
Based on single criteria Factors of production (land, labour)
Market orientation
Economic size
Based on multiple criteria
Absolute versus relative approaches
A numerical simulation
Final remarks
Characteristics of a workable definition
The selected definition should be based on criteria that: Can be operationalized in the largest possible number of countries; Can be used to define the entire distribution of farms (as a continuum from small to large farms);Are not dependent from the outcomes to be measured (i.e. income and labour productivity).
Steps for identifying a workable definition: Selection of the variable used as definitional criterion (e.g. land, labour, market orientation, etc.); Use of one or multiple criteria; Assessment of data availability and accessibility; Choice between an absolute or a relative approach; Selection of a threshold to separate smallholders from other farms.
Definitions based on land endowment
Limited access to land is the most widely used criterion: about 70% of the literaturereviewed defines smallholders in terms of the physical size of the farm, primarily in terms ofhectares of operated land or number of tropical livestock units (TLU).
Focus is on operated land, instead of cultivated land or owned land.
Most popular: small farms are those with less than 2 hectares of land.
Definitions based on land endowment (2)
• Data on land size is generally available from:
Agricultural censuses Agricultural surveys Integrated household surveys (e.g. the LSMS and other multipurpose surveys)
• Land measurement methods are well established in the statistical practice. Techniques based on the use ofGPS devices are gaining popularity.
But:
• The same land size can correspond to highly heterogeneous socio-economic outcomes.
• Land size needs to be combined with other information: e.g. land quality and land use parameters.
• Small can mean different things in different countries.
Definitions based on input use: labour
A smallholder is likely to use little labour input but measuring labour input in agriculture is notstraightforward
• High prevalence of seasonal and part-time workers, whose contribution is difficult to capture insurveys
• Need to compute labour units – full time equivalents associated with each worker of the farm(e.g. the Annual Working Units computed by the EU).
Further considerations:
• Better data on labour input are also needed for monitoring SDG2.
• The implementation of a set of integrated agricultural surveys (AGRIS) can improve data availability.
• The same labour input can correspond to different socio-economic and agro-ecological conditions.
Definitions based on the share of contributing family workers
Small holdings rely largely on family labour – overlap with the “familyfarm” concept
Widely used :IFAD (2009): less than 2 hectares, relying on household members formost of the labour”.Hazel et. al. (2007): “those depending on household members formost of the labour, or those with a subsistence orientation, where theprimary aim of the farm is to produce the bulk of the household’sconsumption of staple foods”
Lipton (2005): “units that derive most of labour and enterprise fromthe farm family”.
Issues:• Family farms and small farms may overlap, but do not coincide: some family farms can be large holdings.
Large contribution of family members is not necessarily typical of small holdings.
• The labour contribution of family members is difficult to capture accurately (LFS vs HH surveys ).
Definitions based on market orientation
Example: OECD (2015): smallholders are farmers that “struggle to be competitive and hence to produce anincome to support themselves and their families”. Furthermore “they often live in poverty and produce at leastpart of their produces for self-consumption”.
Possible criterion : share of agricultural production devoted to own-consumption.
Provides information on economic conditions: high level of own-consumption usually implies lowrevenues.
Provides indirect information on competitiveness.
Data on own-consumption is often available (HBS, LSMS, other integrated surveys)
but
Only information on own-consumption does not capture vulnerability.
Definitions based on economic size
Some national definitions use concepts related to the economic size of the holding.
Examples:
USA: farm size is defined on the basis of the “gross cash farm income”, which is the annualtotal sales of the holding. A small farm is defined as one that grows and sells less than$250,000 per year.
EU: the economic size is progressively replacing the land size as a criterion for definingsmallholders and it is measured through the Standard Output (SO).
• SO: Average monetary value of the agricultural output (crop or livestock) at the farm-gate price. It is aunit value expressed in euros per ha or per head of livestock
Definitions based on multiple criteria
Examples:
CFS HLPE (2013): a small farm is “..an agricultural holding run by a family using mostly (oronly) their own labour … [that] relies on its agricultural activities for at least part of the foodconsumed …[and] with limited reliance on temporary hired labour … ”.Narayanan and Gulati (2002): a smallholder “is a farmer (producing crop or livestock)practicing a mix of commercial and subsistence production…, where family provides themajority of labour and the farm provides the principal source of income”.
But:• Different criteria may be in conflict with each other, and interpretation could be less
straightforward.
• Never operationalized in the statistical practice
Absolute versus relative approaches
Thresholds to separate large from small holdings can be either absolute or relative:
An absolute approach assigns, for a given criterion variable, the same threshold for allcountries;
A relative approach assigns, a threshold that corresponds to the same point of thedistribution of the criterion variable in each country.
Example on land endowment:
Absolute approach: the 2-hectare threshold;
Relative approach: Parameters describing the distribution are used to set the threshold;e.g. a weighted median or weighted percentile approach – the threshold is the farm sizethat accounts for a given share (50% if median) of the total acreage.
Absolute versus relative approaches (2)
Absolute approaches enhance comparability across countries.
Relative approaches recognize countries’ specificity, while reducing comparability
The two approached could co-exist, with each country identifying a standardized relativethreshold together with the international absolute one
A numerical exercise to show the difference between the two approaches: data fromsurveys processed in the RuLIS
Country % of smallholders - 2 ha approach
% of smallholders -land wm approach
land weighted median
Albania 2005 87.7 72.1 1.09
Armenia 2010 37.7 38.6 2.21
Burkina Faso 2014 40.4 83.0 6.00
Ethiopia 2014 80.1 78.0 1.90
Georgia 2010 91.8 79.4 1.10
Ghana 2013 45.5 70.7 4.04
Guatemala 2011 87.3 85.7 1.75
Iraq 2012 54.1 82.4 8.50
Kenya 2005 88.7 86.9 1.62
Malawi 2013 87.4 73.0 1.09
Mali 2014 33.2 90.2 21.00
Nepal 2011 44.2 62.8 3.76
Nicaragua 2014 22.8 43.0 42.25
Niger 2011 26.4 79.6 8.00
Nigeria 2013 86.5 86.9 2.05
Pakistan 2014 11.1 88.1 24.71
Peru 2014 73.2 97.0 18.00
Tanzania 2013 35.1 88.0 9.79
Timor-Leste 2007 81.2 76.0 1.20
Uganda 2010 55.2 82.7 4.40
Vietnam 2010 91.1 83.1 1.00
Final remarks
Need to balance accuracy and data requirements: The most widely used approach is the 2-hectares threshold of operated land: easy to measure
The economic size helps to take into account different economic outcomes of the same land size: dataquality and availability however is problematic.
Multiple criteria could be promising if combined hierarchically: A unique standard criterion – e.g. land size - can be used in countries where other data sources are not
available. In countries where data on the economic size are available, a more complex definition can be adopted
and farmers with small land size can be divided in 2 categories:1) small land size with small economic size;2) small land size with large economic size.
Projects like AGRIS, the LSMS-ISA and RuLIS could increase data availability and facilitate theadoption of progressively more refined criteria.
Choice between absolute and relative approaches: needs to be taken, and specificthresholds need to be identified
Thank you for your attention
RuLIS Expert Consultation: 08 November 2016
Definitions based on economic size (2)
The Standard OutputWhat is it: The SO is the average monetary value of the agricultural output (crop or livestock) at the farm-gateprice. It is a unit value expressed in euros per ha or per head of livestock. The SO coefficients for each productsin each region are computed as an average over a reference period of five successive calendar or agriculturalyears. The sum of all the SO per hectare of crop and per head of livestock in a farm is a measure of its overalleconomic size, expressed in euro.
Which data is needed: Yield; Physical quantities produced; Farm-gate prices; Cultivated area; Number ofanimals present and slaughtered; other technical information.
Main limitations: Larger data requirements. Data accuracy and reliability.More calculations are needed compared to other criteria.
Definitions based on economic size (3)
Example of SO calculation for wheat in Lithuania for 2005Data
Calculation of the SO
The computation of the economic size for year N will be based on a coefficient obtaibed as avarage of the SOsfor years N, N-1,.., N-4
Definitions based on economic size (4)
Example of SO calculation for livestock (other pigs) in Lithuania for 2005Weighting per category of product
Calculation of the SO
Observation: fattening period = 175-180 days
Absolute versus relative approaches (3)
The weighted median approach: example of calculation
Operated land (ha) Cumulated land (ha)
Farm 1 2 2 Farm 2 5 7 Farm 3 7 14 Farm 4 10 24 Farm 5 20 44 Total operated land 44 Median ha 22