Home | Login | Contact | About | Site Map | Print

Home  « What we do  « Methods  « Data

• Who we are
• What we do
• Activities
• Methods
• Typical Farms
• Data
• Cost Analysis
• Benchmarking
• International Collaboration
• Beef & Sheep
• Cash Crop
• Dairy

For the farm-level analysis in agri benchmark we use two different kind of data:

• Farm level data
• Projection data

 

Farm level data

Origin

All farm level data used in agri benchmark is based on own surveying and interviewing of farmers and their advisors. The data collection is done via so-called ‘panels’ (one to six farmers, an advisor and an agri benchmark scientific partner). It is based on the standard operating procedure (see Typical Farms) and on the use of a multi-language standard questionnaire worldwide (see Questionnaire). Depending on the information required, a so-called ‘pre-panel’ or a ‘full panel’ is established for each typical farm.

 

Participants		Data used for
Pre-panel
1 agri benchmark scientist		Production cost analysis
1 advisor		Description of production systems
0-2 farmers		Policy analysis
Full panel
1 agri benchmark scientist		See above
1 advisor		Farm strategies and adjustments
4-6 farmers

 

 

Additionally, individual farm data can be used as a basis for typical farm data, if the particularities of the individual farm are ’typified’, i.e., eliminated or replaced by more typical data.

 

Individual farm data and pre-panel information are sufficient for a status-quo analysis of returns, cost and profitability (benchmarking), whereas full panels are required for discussing, defining and specifying farm strategies.

 

In all cases, data obtained are returned to the farmers and advisors for a cross-check of the data and results obtained (for example, a profit and loss account).

 

 

The panel process – consensus data gathering with crosschecking of results

 

Characteristics

The main characteristics of our farm-level data are:

• Data are collected on whole farm level to reflect an entire picture of the typical farm, its enterprises and production systems. Apart from benchmarking individual enterprises, it enables the analyst to perform policy and farm strategy analysis with its implications for the whole farm level.
• The data sets are consistent due to the whole farm approach. This means that all data on whole farm level and enterprise level fit to each other and together provide a very realistic picture of the farm.
• Almost all return and cost figures can be split up into a quantity and a price component. This allows detailed analysis of the reasons for differences in costs and returns  as well as productivity analysis. Examples are:

Total beef returns
Number of animals * kg carcass weight per animal
(calculated as kg live weight * dressing percentage)
*price per kg carcass weight

 

Total wheat returns
Total hectares of wheat planted * yield per ha in tons * price per ton

 

Labour productivity in beef production
Total carcass weight produced / Total labour hours (hired and family) in the beef enterprise

 

Projection data

Information on the past and the future development of prices and yields is required to simulate typical farms over a ten year period (see also Model section) or to project prices, yields and productivity.

 

Therefore it is necessary to combine the farm-level data of the typical farms with agricultural sector and trade models that are able to develop longer term projections of macroeconomic and market indicators such as input and output prices.

 

Projections are taken from the following sources:

• Projections that are regularly published by certain institutions
  (e.g., USDA, FAPRI, OECD, FAO, European Commission)
• Cooperation with consortiums that apply highly aggregated models (such as GTAP)
• Own trend projections and ad hoc expert consensus assessments if the sources mentioned above turn out to be inadequate to provide the necessary information – for example, if the projection data is not available in sufficient regional or product-related detail

Typical Farms

 

Questionnaire