National FSA TrainingModule 15: Economic analysis of on-farm trials

Module 15Economic analysis of on-farm trials

15.1Introduction

Evaluation of on-farm trials is done to determine the impact of the technology/intervention. It may be ex-ante, on-going as well as ex-post. Traditionally experimental results are assessed using statistical techniques only. Over the years it has been demonstrated that the criteria used by farmers to evaluate and adopt technologies may be totally different to that of researchers. Very often socio-economic considerations play a major role in accepting or rejecting a technology. Assessment of trials is, therefore, one of the important stages of an on-farm research programme. The main elements of such an assessment are farmer assessment, agronomic evaluation, statistical analysis and economic analysis.

In this module various socio-economic criteria and simple techniques used to assess economic performance of a technology are discussed. The economic analysis helps researchers:

  • to look at the results from the farmers’ view point;
  • to decide which treatments merit further investigation, and
  • to decide on the recommendations to make to farmers.

On-farm data upon which the recommendations are based must be relevant to the farmers’ own agro-ecological conditions, and the evaluation of those data must be consistent with the farmers’ goals and socio-economic circumstances.

15.2The need for socio-economic evaluation

Traditionally, agronomic trials are evaluated by biological scientists using statistical techniques and the dominant evaluation criteria used is yield per unit area. The primary objective here is the maximum exploitation of the biological potential and the test used often is statistical significance. Normally the experiments/treatments are tested at 0.01, 0.05, 0.10 percent level. If they do not pass this test they are not considered any further. This approach has many limitations in deriving recommendations in a systems context:

(a)Evaluation criteria that a farmer may use are varied, location specific and depend greatly on the degree of market orientation. The appropriateness of any technology should always be evaluated in relation to their priority objectives and resource use pattern. The relevant criteria for the target group should be identified and used in the design and interpretation of the experiment. A number of physical, biological and socio-economic variables influence the farmers' choice of a technology. Very often the socio-economic factors such as lack of marketing, shortage of labour, clash with food priority etc., determine these choices.

(b)Agronomic data only establishes the technical relationships that can be used to determine the technical optimum. It should be complemented with market information in order to establish the economic optimum. The 'economic optimum' for any input is always lower than the 'technical optimum'. If the treatment means are not significantly different but an economic analysis shows that one treatment is a better recommendation than the others then we need to have a more careful analysis. Examples of situations where farmers choose technologies even without significant differences in yield may be:

  • varietal means may not be different but one could observe differences in palatability (preference), prices, storability, early maturing etc;
  • technology may not increase the yield/profit, but may reduce the labour requirement at the critical period;
  • A new technology may provide possibility of introducing a second crop i.e. increasing the total production of the systems;
  • The intervention may reduce malnutrition in the target group.

It should be remembered that these variable attributes are also valid even in situations where there is significant difference in yield between treatments. It is important to remember that a farmer is managing a system and he/she is interested in improving the total production of the system without creating serious contradictions to his/her priority objectives and resource use patterns.

c)Farm level decisions are made and actions are taken in an attempt to reach goals in a world of uncertainty and scarcity of resources. These two elements are therefore crucial in assessing the impact of any technology. Thus it is necessary for the biological scientist to conduct economic analysis in a similar manner, as they are responsible for the statistical analysis of their trials. The usefulness of the result of many bio-physical research experiments can be greatly enhanced if relevant economic analysis can be applied to the results. In this regard, therefore it makes sense for biological scientists and agricultural economists to jointly evaluate experiments to establish both biological and economic viability.

The majority of economic analysis with reference to on-farm trial work is in three categories:

1.Average Returns Analysis. This analysis consists of a listing of the average costs of producing a particular product and the average value of the product for each technology being compared. The cost-and-returns analysis requires information on both variable and fixed inputs. A more limited average variable cost-and-returns analysis is often used to compare different technologies that used the same fixed inputs. This information can be used to compare the average returns above variable costs (RAVC) (i.e., sometimes called gross margin analysis) for different technologies and the returns to other production factors such as total labour or labour for a single operation e.g. weeding. The process of valuing inputs and commodities is of particular importance in making realistic cost-and-return analyses.

The average returns (gross margin) analysis allows a comparison between various technologies being tested based on the inputs the farmer must provide. The procedure is as follows:

  • Calculate an average yield or an average amount of product for each separate technology.
  • Calculate average inputs -- often with particular emphasis on labour inputs -- for each technology separately.
  • Calculate the gross return (i.e., gross total value product), which is the adjusted yield times the appropriate field price for the product, or products, if there are more than one.
  • Calculate the variable costs associated with each technology. The variable costs for a crop trial usually include labour, seed, draught hire (i.e., if hired draught is used), and a charge for equipment depreciation.
  • The average return above variable cost (RAVC) -- also called gross margin -- is then calculated by subtracting the variable costs from the gross total value product.
  • These average net returns can then be compared between technologies. According to Norman, et. al. (1994) some scientists believe that the return for a new technology must be at least 30 percent higher than that for the traditional technology before farmers will be willing to consider adopting the new technology.

This analysis is generally on a per hectare basis, and the return calculated is a return to management, assuming that land is fixed and that labour has been valued at the price of its best alternative use (i.e., opportunity cost). In order to maximise profits, it is necessary to maximise returns to the most limiting resource. For example, the most limiting resource may labour for weeding. When the most limiting resource is known, it is possible to calculate an average net return to that resource for the different technologies being compared and choose the most favourable technology.

Measures of return are calculated as follows:

(a)Return above variable cost (RAV) i.e. Gross Margin (GM)

= Gross Return - Total Variable Costs

(GR)(TVC

(b)Returns to Factor A.

The general equation for the rate of return to Factor A is:

Rate of Return to A

= GR - all costs other than Costs of A

Costs of A

Therefore:

(i) Return to TVC = GR

TVC

(ii) Return to labour and power Cost = GR - Material Cost

Labour and power cost

(iii) Return to material Cost= GR - Labour and power cost

Material Cost

2.Budget Analysis. There are several types of budget analysis. The enterprise budget is statement of costs-and-returns (i.e., both variable and fixed) for a particular enterprise or technology. This type of budget can be used as a building block in making whole farm budgets. The partial budget, on the other hand, is a direct comparison of the elements within enterprise budgets that change between technologies. This type of budget requires fewer data than the enterprise budget and offers the advantage of direct comparison. Finally, whole farm budgets can be used to look at allocation of resources between enterprises and at the impact of a new technology on the allocation of resources to other enterprises on a farm. The partial budget technique is used most frequently in FSR.

3.Risk Analysis. When a farmer undertakes a crop or livestock enterprise she/he always faces the risk of failure and loss of their time, cash or other inputs invested in the enterprise. When farmers consider a new technology, they are concerned about the risk involved in the new technology compared to the risk of their present technology. Measuring risk is difficult and is of somewhat limited value because different farmers look at risk differently. Risk analysis needs to be kept as simple as possible. Some indications of risk can be obtained from doing sensitivity analysis with partial budgets.

15.3The partial budget

The first step in doing an economic analysis of on-farm experiments is to calculate the costs that vary for each treatment. In developing a partial budget all outputs and inputs are measured in currency units, as a common denominator. This is necessary because otherwise it would be impossible, for example, to add hours of labour to litres of herbicide and compare these with kilograms of grain. The use of currency units does not, however, necessarily imply that farmers spend money on inputs or receive money for the outputs. Neither does it imply that farmers are concerned only with money. It is simply a device to represent the process that farmers go through when comparing the value of the things gained and the value of the things given up.

In a partial budget not all costs (e.g. family labour) necessarily represent the exchange of cash. An important concept used in the calculation of such cost items is that of opportunity cost. This cost is defined as the value of any resource in its best alternative use. In partial budgeting one has to be concerned with the differences in costs and benefits between experimental treatments.

Preparation of a good partial budget requires:

  1. A detailed understanding of farmers objectives, practices, resource use pattern, constraints etc.
  2. Some understanding of the fundamental concepts and principles of economic theory e.g. profitability, risk, opportunity cost, scarcity of resources, marginal concepts etc.
  3. Good judgement and common sense. Educated guesses are always better than ignoring a cost or a benefit. It should be remembered that:
  1. all source of benefits and costs to farmers are included in the analysis.
  2. the realism of the costs and yield assumptions are as important as the type of analysis chosen.
  3. a partial budget does not give the net effect of a technology/production process. It only gives the change in NET BENEFITS. It does not show the profitability of the enterprise or the farm. In order to get this one has to do an enterprise/whole farm budget.

In order to construct a Partial Budget one should calculate:

  1. benefits of different treatments
  2. costs that vary across treatments

A. Estimation of benefits

1.Identify the sources of benefits

(a)Direct Benefits - main products

- by products

(b)Indirect benefits - often difficult to identify and quantify if the systems interaction is not known.

2.Quantify the benefits derived from the technology or treatment

(a)Estimate the yield

(b)Estimate the yield adjustment coefficient

(c)Calculate the adjusted yield

(d)Determine the field price of the product (and by-products).

3.Calculate the Gross Field Benefit (GFB)

GFB = Adjusted Yield X Field Price

B. Estimation of Costs

Here the total costs that vary across treatments are estimated:

1.Identify and list the input items that vary across treatments.

2.Quantify the level of input in each treatment.

3.Estimate the unit value of the input. Once again we use the field price for input.

4.Estimate the Field Costs of the inputs

Field Cost = Field Price X Quantity of the Input

5.Calculate the total costs that vary for each treatment.

C. Calculation of NET BENEFIT (NB)

NET BENEFIT = Gross Field Benefit = Total Cost(s) that vary across treatments.

Definitions

  • The adjusted yield for a treatment is the average yield adjusted downward by a certain percentage to reflect the difference between the experimental yield and the yield farmers could expect from the same treatment.
  • The field price (of an input) - is the value which must be given up to bring an extra unit of input into the field.
  • The field price of an output is the value of one unit of output to the farmer, less harvest costs that are proportional to yield.
  • Costs that vary are the costs (per hectare) of purchased inputs, labour, and machinery that vary between experimental treatments.
  • Field cost is the field price multiplied by the quantity of the input needed for a given area.
  • The total costs that vary is the sum of all the costs that vary for a particular treatment.

Yield estimation

Farmers often do not receive the same yield as researchers even when they apply the same treatment. Yields obtained by researchers are generally high due to various reasons:

1.The management standard of the experimentation is very high.

  • Small plots - easier to manage and the quality of work is good.
  • Adlib resource input - often researchers have less resource restrictions than the farmers.
  • Timely operation - e.g. planting, plant spacing, fertiliser application, weed control, pest and disease control and harvesting.
  • High level of non-experimental variables. These are strictly controlled.
  • Control of experimental errors.

On a management scale the (experimental station) researchers are likely to be at the top end but in a farming population both yield and management level are more likely to be normally distributed. The recommendations generated should address the major proportion of the farmers and not the good managers. Depending on the location of the trials and who manages the trial one could observe 3 major differences.

a)Differences in natural factors - soil fertility, pest and disease incidence etc.

b)Differences in management factors - this in fact is a random variable in

farm level situation.

c)Differences in resource base - small farmers have much less resources than

the researchers.

Therefore there is a need to adjust the yields obtained by the farmers.

2.Use of Different Techniques

Researchers often use different harvesting and drying techniques which minimises

the field loses.

3.Storage losses

Some of the local varieties store better than the improve varieties. If farmers store

improved varieties for late sale or house hold consumption then they incur heavy storage losses. Therefore the effective production for the farmer is less than the actual harvest.

4.Abandoning of treatments

Very often researchers abandon some sites or some treatments due to drought, floods, insect attack etc. This may inflate the yield of the experiment, but the farmers will have to face these situations. Therefore it is important to include them in analysing the experiments.

Example of a partial budget:

Table 15.1 shows a partial budget for a weed control trial. There are two columns, representing two treatments (hand weeding and herbicide). The first line of the budget presents the average yield from all locations in the recommendation domain for each of the two treatments. The second line is the adjusted yield and the third one shows the gross field benefits. The last two lines of the table represent the total costs that vary and the net benefits.

Table 15.1Example of a partial budget

Hand weeding / Herbicide
Average yield (kg/ha) / 2,000 / 2,400
Adjusted yield (kg/ha) / 1,800 / 2,160
Gross field benefits (Tsh/ha) / 153,000 / 183,600
Cost of herbicide (Tsh/ha) / 0 / 5,600
Cost of labour to apply herbicide (Tsh/ha) / 0 / 2,000
Cost of labour for hand weeding (Tsh/ha) / 25,000 / 0
Total costs that vary (Tsh/ha) / 25,000 / 7,600
Net benefits (Tsh/ha) / 128,000 / 176,000

15.4Marginal analysis

15.4.1Introduction

In the previous section a partial budget was developed to calculate the total costs that vary and the net benefits for each treatment of an experiment. In this section a method for comparing the costs that vary with the net benefits is described. This comparison is important to farmers because they are interested in seeing the increase in costs required to obtain a given increase in net benefits. Calculation of net benefits for each treatment is only an intermediate step in the economic analysis of on-farm data. The treatment with the highest net benefit does not always make the best recommendation. The net benefit calculations also do not explicitly treat some crucial aspects of small farmers’ conditions, namely scarcity of resources and risk. In order to include these two critical aspects which are crucial in decision-making one has to perform a ‘marginal analysis’. This marginal analysis can be made more efficient by an intermediate step known as Dominance Analysis. This intermediate step enables one to discard the clearly unprofitable treatments.

15.4.2Dominance analysis

Dominance analysis in fact divides the treatment set into 2 categories, namely “dominated” and “un-dominated” treatments. A dominated treatment has net benefits that are less than or equal to those of a treatment with lower costs that vary. Dominated treatments need not be considered further in the analysis. One could perform dominance analysis using one of the two approaches.

  1. Net Benefit Curve
  2. Tabular Method

A. Net Benefit Curve

  • A net benefit curve could be drawn for any factor
  • A net benefit curve shows the relationship between the level of input and the corresponding net benefit from the alternatives.
  • To construct a net benefit curve each treatment is plotted on a graph, the vertical axis representing the net benefits and the horizontal axis representing the cash outlay/labour input.
  • Beginning with the point that corresponds to the least expensive treatment. Aa line is drawn to a point that represents the next most expensive treatment - but only an upward slopping line is allowed. The un-dominated treatments will be on the net benefit curve but dominated treatments will be below the net benefit curve. In other words, the treatments or choices which are not dominated when connected together form the ‘NET BENEFIT CURVE’. The alternatives or treatments which are falling below the Net Benefit Curve are known as dominated alternatives; because for each of these there is another alternative with a higher net benefit and lower variable cost. Under normal circumstances a farmer will never choose one of these dominated alternatives.
  • The dominated alternatives are eliminated from further analysis and un-dominated alternatives are used to compute Marginal Rates of Returns (MRR).

C. Tabular method