7.1 SCOPE AND PURPOSE OF EVALUATION
7.2 LIMITED EVALUATION
7.3 COMPARATIVE ANALYSIS
7.4 DATA SOURCES
7.5 DIFFICULTIES IN EVALUATION AND COMPARATIVE ANALYSIS
7.6 REFERENCES
'We must study the present in the light of the past for the purposes of the future.'John Maynard Keynes (1883-1946)
Evaluation of whole-farm systems consists essentially of measuring how adequate and how effective an existing system has been in achieving its planning objectives over some past operating phase (season or operating year). In the simplest case assumed here, the objective will be generation of maximum net income in money terms - either directly for market-oriented farms or indirectly by imputation of values for subsistence-oriented farms. Thus the evaluation will refer primarily to financial criteria and to factors directly underlying those criteria; it reduces largely to a systematic consolidation of the costs and returns which arise in various parts of the system - i.e., in the separate enterprises or activities and in the whole-farm service matrix. This emphasis on costs, returns, income and profitability, however, is not meant to decry the importance of those other farm system properties - such as sustainability, environmental compatibility etc. - outlined in Section 6.2. Their achievement, too, must be evaluated. Some criteria for such evaluations are suggested in Table 6.7.
The demarcation lines which set 'whole-farm evaluation' apart from other areas of farm management analysis are somewhat elastic and shift according to the intended purpose of the evaluation. The sequence and scope of evaluative analysis are shown in Figure 7.1. The basic component of evaluation is a farm operating statement (or a whole-farm budget in financial terms). As indicated, this draws on and consolidates data from each activity as well as data on unallocatable overhead or common costs relating to the farm service matrix or farm system as a whole. An important requirement is that, whatever data are selected for executing the evaluation, they must be consistent with the system's actual planning objectives (Section 6.1). The next step is to derive measurements of performance from the base whole-farm budget table. This defines the scope of what is here termed limited evaluation. The purpose of a limited evaluation is simply to report on the condition or record the performance of a farm system. In itself it is not intended as a basis for further analysis or action. It is intended only to report the facts. Nor does it explicitly consider the farm's household system.
In contrast, an expanded evaluation is intended as a basis for further analysis and action, and it does consider the household system as well as the farm system. As shown in Figure 7.1, expanded evaluation most commonly takes the form of a comparative analysis of (A) the individual resources used by the system and (B) the individual activities which comprise the system. Such a comparative analysis requires the same kind of farm data as used in a limited evaluation. However, for comparative analysis the data must be in disaggregated (i.e., activity-specific) form and be supplemented by data relating to the farm household. Comparative analysis is intended to diagnose problems and provide a basis for remedial action. This defines the scope of an expanded evaluation. Its purpose is not only to record the achievements of a system but also to look for evidence of under-achievement and to identify structural weaknesses within the system. It leads to the kind of follow-up action indicated in the bottom section of Figure 7.1. If weaknesses within either the A stream (resource allocation) or B stream (activity mix) of the diagram are minor, they might be corrected on the basis of a little partial budgeting. If they are major, they might require, in the A stream, the application of response analysis (Chapter 8) and, in the B stream, restructuring of the system using allocation budgeting or simplified programming or linear programming (Chapter 9). System weaknesses might also be due to factors not directly related to the activities or the resources used in the activities, e.g., they might arise in the service matrix with too many/not enough work oxen, too much/not enough investment in soil conservation, drainage, farm storage, etc.
FIGURE 7.1 Sequence of Analysis in Farm-system Evaluation
To summarize Figure 7.1:
· Limited evaluation is concerned with reporting the facts relating to the past performance of a whole-farm system (the operating statement).· Expanded evaluation using comparative analysis is concerned with diagnosing system weaknesses which might be implied by those facts.
· Subsequent analysis (Chapters 8 and 9) is concerned with prescribing remedial action to remove weaknesses, possibly to the extent of restructuring the whole system.
7.2.1 Using data aggregated on a whole-farm basis
7.2.2 Using data on an activity-specific basis
7.2.3 Productivity of individual resources on a whole-farm basis
7.2.4 Productivity of individual resources on an activity basis
7.2.5 Total productivity measures
As noted, a limited evaluation of a farm system consists of preparing an operating statement (or whole-farm budget) covering an appropriate period (usually a year) such as exemplified by Table 7.1 and deriving appropriate measurements of whole-farm performance from this operating statement as shown in Table 7.2.
Table 7.1 records annual costs/returns performance on a whole-farm or pooled basis for a 2.8 ha farm having paddy, maize, bamboo and cattle activities. (The capital investment inventory of this farm was presented earlier in Table 5.5.) The data of Table 7.1 are shown as they would typically be recorded on the farm or received as memory estimates from members of the household. The production/disposal data are activity-specific but the inputs/costs are 'all mixed up' with no indication of which inputs have been used by which activities. For the moment, since the intention is to obtain a picture of the farm as a whole, this does not matter. (But it will cause problems later; Section 7.2.2.) Also, on small farms it will usually be necessary to distinguish between different classes of outputs: final products that are sold or consumed by the household, and intermediate products to be used in the next farm operating phase as resources. For subsistence-oriented farms it will be necessary to impute prices/values to products which are not sold for cash.
In section A of Table 7.1, the values of all final products are consolidated to a total gross return of Rs 44 000. In section B, all direct input costs of all activities are also consolidated to give a total direct cost of Rs 12 500. This excludes the value of farm-generated resources (family labour with an imputed value of Rs 4 500). Section C records farm fixed costs (discussed in Sections 5.3 and 5.4). Depreciation is shown as a separate item in line D.
TABLE 7.1 - End-of-year Operating Statement for a Mixed Farm of 2.8 Ha
|
A. All Outputs/Returns (Pooled) |
(Rs) |
Notes |
|||
|
Paddy
|
grain
|
sold |
25 000 |
|
|
|
food |
6 000 |
|
|||
|
(seed) |
(1 500) |
Used on farm |
|||
|
straw |
(used) |
(900) |
Used on farm. |
||
|
Maize
|
grain
|
sold |
7 000 |
|
|
|
food |
800 |
|
|||
|
stover |
(used) |
(700) |
Used on farm. |
||
|
Bamboo
|
poles
|
sold |
1 200 |
|
|
|
(used) |
(200) |
Used on farm. |
|||
|
Cattle
|
calves |
sold |
2 500 |
|
|
|
milk
|
sold |
600 |
|
||
|
food |
900 |
|
|||
|
Total Gross Return: |
44 000 |
Omits (items) used as resources on farm. |
|||
|
B. All Purchased Activity Direct Inputs (Pooled) |
|
||||
|
HYV seed |
2 000 |
|
|||
|
Fertilizers |
3 000 |
|
|||
|
Agricides |
1 400 |
|
|||
|
Veterinary |
1 200 |
|
|||
|
Irrigation fuel |
1 900 |
|
|||
|
Transport |
600 |
|
|||
|
Hired labour |
2 400 |
|
|||
|
Family labour (450 days) |
(4 500) |
Based on opportunity cost of Rs 10/day. |
|||
|
Total Direct Costs: |
12 500 |
Omits family labour and farm-generated inputs. |
|||
|
C. All Farm Fixed Costs (except Depreciation) |
Depreciation recorded in D below |
||||
|
General Charges |
|
||||
|
|
Land tax |
500 |
|
||
|
|
Livestock tax |
90 |
|
||
|
|
Road-repairs levy |
250 |
|
||
|
|
Water-use fee |
70 |
|
||
|
Capital Equipment Repairs, Operation |
|
||||
|
|
All capital-item repairs |
1 600 |
From Column (5) of Table 5.5. |
||
|
|
All capital operating costs |
1 000 |
From Column (6) of Table 5.5. |
||
|
Total Fixed Costs: |
3 510 |
Omits depreciation. |
|||
|
D. All Capital Depreciation |
4 480 |
From Column (4) of Table 5.5. |
|||
Based on the data of Table 7.1, six whole-system evaluation measures are presented in Table 7.2. The origin of and conditions attached to these factors are noted in the table. There is usually a degree of subjectivity involved in selecting the 'best' evaluation factors, i.e., in determining the 'best' way of measuring performance. But which of the measures E, F,... J are most appropriate will, to some degree, be indicated by analytical circumstances.
TABLE 7.2 - Derived Measures for Annual Whole-farm Evaluation
|
Measure |
Calculationa |
Value |
Notes |
|
|
(Rs) |
||||
|
E. |
Farm Gross Margin |
A - B |
31 500 |
|
|
F. |
Farm Net Actual Returns |
E - C |
27 990 |
Depreciation not yet charged. |
|
G. |
Farm Net Sustainable Returns |
F - D |
23 510 |
Depreciation charged; system now sustainableb. |
|
H. |
Family Farm Available Income |
H = F |
27 990 |
But only if depreciation is not covered. |
|
I. |
Family Farm Sustainable Income |
I = G |
23 510 |
Long-term Sustainable farm income. |
|
J. |
Total Available Family Income |
(H or I) + S |
23 510 |
S is non-farm income, here assumed to be zero. |
a A, B, C and D are as specified in Table 7.1.
b Sustainable refers to the non-depletion of capital assets excluding land.
Measure E, total farm gross margin (of Rs 31 500), is the easiest to derive: it consists of the sum of all activity gross margins (Chapter 4), or if activity costs/returns have been pooled, the total value of farm output of final products less total farm direct costs (A - B in Table 7.1). This is a good measure of performance if the evaluation is needed for making comparisons between similar farms and if their capital structures (levels of fixed costs, item C in Table 7.1) are similar or relatively unimportant.
An alternative is to use factor F (farm net returns, i.e., TGM - FC). But this level of 'income' (of Rs 27 990) is not yet stable over the long term because it makes no provision for replacing capital equipment as this wears out. If this is an important consideration, measure F should be decreased by a depreciation charge to obtain measure G (farm net Sustainable returns) in which case the 'income' (of Rs 23 510) would be Sustainable over the long run in terms of its fixed-capital requirement. Note that this is clearly a narrow use of the word Sustainable. It says nothing about the possible loss of sustainability due to degradation of the land base, falling terms of trade etc.
As discussed in Chapter 5, a depreciation 'charge' is not one involving actual cash; it is only a bookkeeping convention which, at the farmer's discretion, may or may not be covered by putting money aside either in a depreciation fund (not a common choice) or (more commonly) as an investment in the farm to assist future replacement needs. If no money is actually put aside to cover depreciation, the amount of cash income actually available as family income would be H (the same as F). This would imply the (temporary or continuing) run-down of farm capital. Finally, concern might be with total family income from all sources rather than only income from the farm. If so, factor J would be relevant in that it measures the performance of both the farm and the household components of the system. (But, here again, income of the farm component might or might not be defined to be sustainable over the long run depending on whether it is measured as H or I.)
Inter-farm comparisons
The evaluation analysis so far has presented records of performance for a single farm in terms of measures E, F,... J. However, at the risk of contradicting what was said previously regarding the flow of analysis in Figure 7.1, it might be noted that comparisons with other farms can also be introduced at this present stage. Such comparisons would involve comparing measures E, F,... J on the subject farm with those of similar farms. In fact this would be desirable because although E, F,... J are adequate as a record and over time would indicate the degree of variability in performance of the subject farm, in themselves they do not provide any basis at all for judging if the farm's income levels are good or bad or better or worse than those of other farms. If introduced at this point, whole-farm comparison would take the form shown in Table 7.3. Thus the subject farm (Farm 1), primarily a paddy farm, would be compared in terms of the measures E, F,... J with other farms of the village or area having similar size, soils, water supply etc. but not necessarily the same mix of activities. Indeed, for farm management analysis in other fields (e.g., Field D, policy guidance), it might be necessary to compare the subject farm with, e.g., livestock or cotton farms.
TABLE 7.3 - Example Format for Whole-farm Comparative Analysis (Annual Basis)
|
Evaluation measure |
Farm 1 |
Farm 2 |
Farm 3 |
Farm 4 |
|
|
(Rs) |
(Rs) |
(Rs) |
(Rs) |
||
|
E. |
Farm Gross Margin |
31 500 |
33 600 |
30 100 |
28400 |
|
F. |
Farm Net Actual Returns |
27 990 |
28 100 |
27 200 |
22900 |
|
G. |
Farm Net Sustainable Returns |
23 510 |
26 700 |
25 050 |
21 700 |
|
H. |
Family Farm Available Income |
27 990 |
28 100 |
27 200 |
22 900 |
|
I. |
Family Farm Sustainable Income |
23 510 |
26 700 |
25 050 |
21 700 |
|
J. |
Total Available Family Income (I + S) |
28 510 |
28 300 |
29 600 |
22 090 |
All farms differ to some degree in their size, available capital, labour force etc. If more detailed inter-farm comparisons are to be made, it will be necessary to convert the measures E, F,... J to a comparable unit basis such as per ha, per Rs 100 of capital, per labour day etc. Thus, if factor I (sustainable family income) is of interest, further measures would be derived from this such as:
I (Sustainable Family Income) for the subject farm:
|
|
(Rs) |
|
Per ha of land |
23 510/2.8 = 8 396 |
|
Per Rs 100 of capital |
23 510/1 066 = 22 |
|
Per family labour day |
23 510/450 = 52 |
|
Per family member |
23 510/6 = 3 918 |
|
Per adult family member |
23 510/3 = 7 837 |
For some purposes, e.g., when it is intended to use the evaluation as a basis for further comparative analysis, it is necessary to present the operating statement in both aggregated whole-farm terms and in terms specific to each separate activity. Here, as previously, the analyst might possibly obtain the necessary data from farm records. Usually, however, no useful records will be kept and the needed information must be obtained via intensive discussions with members of the farm household. In either case, farmers tend not to distinguish clearly - or as clearly as the analyst would like - either between inputs going to the various activities or between outputs flowing from each of them. The task of disaggregating the data into activity-specific subsets is often a major but necessary one if later comparative analysis is to be possible.
Disaggregation will result in a set of operating statements, one for each activity and another for the farm as a whole, as shown in Table 7.4 (which again refers to the same farm as discussed in the previous section). Because of their nature, it will usually be possible to 'son out' the direct inputs/costs belonging to each activity. Some fixed costs also will be obviously activity-specific; where they are not, because they are of an overhead nature, they are assigned or charged to the farm as a whole as shown in the second-last column of Table 7.4. Data for the whole farm, aggregated across the activity and farm-as-a-whole columns, are given in the last column of the table.
The several components of fixed costs were defined in Chapter 5 and would be obtained as in Table 5.5. To calculate these and also depreciation, it is usually necessary to construct a separate worksheet in which all capital investment is allocated item-by-item to the respective activities on the basis of relative use of each item by each activity. Such a worksheet is illustrated by Table 7.5 in which, e.g., based on the farmer's advice or other knowledge, the tractor is allocated in the proportions of 3:2:0.5:4.5 to paddy, maize, cattle and the farm as a whole, respectively; but all of the thresher to paddy; and since farm fences serve each of the activities, these are allocated to the farm as a whole.
In Table 7.4 the evaluation measures E, F, G have exactly the same definitions when applied at activity level as when applied to the whole farm in Table 7.2. However, measures H, I, J apply only to the whole farm. The evaluation then continues, on a per unit of resource basis, in line items K and L of Table 7.4. Measure K repeats measure E but now on a per ha, per Rs 100 capital and per family labour-day basis, while L repeats G but now on a similar per unit of land/capital/labour basis.
TABLE 7.4 - Farm-system Evaluation using Activity-specific Data (Annual Basis)
TABLE 7.5 - Worksheet for Allocation of Farm Capital Investment to Specific Activities or Farm as a Whole as required for Analysis of Table 7.4 (Annual Basis)
|
Fixed Capital Inventorya |
Specific Activities |
Farm as a Whole |
||||
|
Item |
Value |
Paddy |
Maize |
Bamboo |
Cattle |
|
|
(Rs) |
(Rs) |
(Rs) |
(Rs) |
(Rs) |
(Rs) |
|
|
Land |
50 000 |
- |
- |
- |
- |
50 000 |
|
House |
10 000 |
- |
- |
- |
- |
10 000 |
|
Sheds |
5 000 |
2 000 |
2 000 |
- |
500 |
500 |
|
Tractor |
10 000 |
3 000 |
2 000 |
- |
500 |
4 500 |
|
Thresher |
2 000 |
2 000 |
- |
- |
- |
- |
|
Cultivators |
3 000 |
2 000 |
1 000 |
- |
- |
- |
|
Ox gear |
600 |
300 |
200 |
- |
- |
100 |
|
Barn |
5 000 |
1 000 |
1 000 |
- |
1 000 |
2 000 |
|
Fences |
6 000 |
- |
- |
- |
- |
6 000 |
|
Dam |
8 000 |
3 000 |
- |
- |
- |
5 000 |
|
Pump (house) |
4 000 |
- |
- |
- |
- |
4 000 |
|
Oxen |
3 000 |
- |
- |
- |
3 000 |
- |
|
Totalb |
106 600 |
13 300 |
6 200 |
0 |
5 000 |
82 100 |
a As per Table 5.5.
b As transferred to Table 7.4.
To summarize, the right-hand column of Table 7.4 contains essentially the same information regarding the whole farm as did Tables 7.1 and 7.2. In Table 7.4, however, costs/returns are 'split up' or assigned as far as possible to the respective activities. When this cannot be done, they are assigned to the farm as a whole. Obviously this conveys much more information than did Tables 7.1 and 7.2; but whether this is sufficient to warrant the much greater clerical effort depends on analytical circumstances. It is unavoidable if the intention is to proceed to the next stage, comparative analysis.
So far no information has been obtained regarding the productivity of the various resources - or more specifically, the value of output which might be attributed respectively to the resources of land, capital and labour.
In the whole-farm situation (Section 7.2.1), 2.8 ha of land were used in the production of Rs 23 510 of annual sustainable family income (measure I of Table 7.2). But this does not mean that the resource land produced Rs 23 510/2.8 = Rs 8 396 of income per ha or that the productivity value of land is Rs 8 396 per ha. Similarly, in line L(iii) of Table 7.4 the net sustainable income G per day of family labour used in each activity is respectively Rs 23 030/150 days = Rs 153 in paddy, Rs 29 in maize, Rs 60 in bamboo and Rs 11 when used for cattle. Again these are only input-output coefficients: they do not indicate, e.g., that the productivity value of family labour in maize is Rs 29 per day.
The reason that these coefficients are not measures of resource productivity is obvious: if, e.g., in the case of the whole farm, all of the farm's net sustainable returns of Rs 23 510 were attributed to the 2.8 ha of land (at the rate of Rs 8 396 per ha), there would be nothing left with which to reward or 'pay off the other factors of labour and capital - or, stated another way, this would imply that labour and capital have played no role in farm production, or that the value of output due to them is zero.
Residual productivity values of resources
Resource values based on productivity can be estimated (with some qualification) by the residual-value method. This is most conveniently applied to sets or categories of resources rather than to individual resource items which comprise these sets. On Asian farms the most important sets are the classical triad of land, labour and capital. The procedure is to select the category of interest, say land; then 'pay off the other resource categories of labour and capital at their market rate (or other appropriate rate, i.e., at their opportunity cost); then deduct these resource 'payments' for labour and capital from total farm production value, and regard the residual as the productivity-based value of land, the resource of interest. The procedure is then repeated to obtain the analogous values of the other resources - taking each in turn, paying off the others and assigning any residual value to the subject resource. Note that the 'payments' referred to in this procedure are not actual payments. They are simply ascribed values used in the productivity calculation.
The procedure is illustrated in the worksheet of Table 7.6 in which the productivity-based values of land, capital and labour are obtained for the whole-farm system. The upper part of the table records the necessary data and their source. The selected measure of whole-farm performance on which the analysis is based is item G of Table 7.2, i.e., farm net sustainable returns. In section M of Table 7.6 the returns to land, or that part of total income G which is attributable to land, is G (i.e., Rs 23 510) less payments to capital and family labour for their respective services (i.e., Rs 5 660 and Rs 4 500) giving a residual of Rs 13 350 attributable to land, or Rs 4 768 per ha. In this calculation the ascribed payment to capital of Rs 5 660 is found by applying a market rate of 10 per cent to the total value of capital (excluding land) of Rs 56 600 used in the system (Table 5.5). Similarly, the payment to labour of Rs 4 500 is found by applying an opportunity cost of Rs 10 per day to the 450 days of family labour used (Table 7.1). Note that this payment to labour does not include hired labour which has already been accounted for by its inclusion as a direct cost (Table 7.1) in the calculation of G (Table 7.2).
The logic and steps in obtaining the next valuation measure, item N or returns which might be attributed to use of the capital resource, are identical. Again starting with item G then 'paying' Rs 5 000 and Rs 4 500 for use of the land and labour resources respectively, the residual of Rs 14 010 can be viewed as a measure of the productivity value of capital. On a per-unit-of-resource basis this is Rs 25 per Rs 100 of capital (except land) used by the system. And again using the same procedure, the return that can be attributed to family labour in evaluation item 0 is Rs 29 per labour day.
It might be noted that these allocations or residual 'payments' to each of the sets of resources are not to be regarded as holding true in any aggregative or financial accounting sense. If the total residual returns to each of the resources are summed, i.e., 13 350 + 14 010 + 12 850 = Rs 40 210, the result is obviously much greater than the amount of net sustainable returns (Rs 23 510, item G) actually realized. Interpreted in this aggregated way, the results have no meaning because of the double counting inherent in such aggregation. The correct interpretation of the results of the evaluations of items M, N, O is:
If (a) all farm operating and fixed costs are deducted from the total value of farm production, and then (b) the residual from operation (a) is allocated to the payment at normal prices of all production resource sets (land, labour, capital) except one, then the final residual from operation (b) can be viewed as an amount that can be attributed to that one resource or factor for the services it provides to the whole-farm system. This assumes that other resources do not generate returns which are greater than (or less than) their normal price. Insofar as this assumption is incorrect, the productivity value of the resource will be over- (or under-) estimated.
TABLE 7.6 - Worksheet for deriving per Unit Productivity Values of Farm Resources (Land, Labour, Capital) on a Whole-farm (Annual) Basis
|
Basic Data |
|||
|
(a) |
farm land |
2.8 ha |
|
|
(aa) |
value of land |
50 000 RS |
(from Table 7.5) |
|
(aaa) |
'payment' to land |
5000 Rs |
(10% charge on (aa)) |
|
(b) |
farm capital |
56 600 Rs |
(except land, from Table 7.5) |
|
(bbb) |
'payment' to capital |
5 660 Rs |
(10% charge on (b)) |
|
(c) |
family labour used |
450 days |
(from Table 7.1) |
|
(cc) |
value of labour |
4 500 Rs |
(Rs 10/day) |
|
(ccc) |
'payment' to labour |
4 500 Rs |
|
|
M. Net Farm Sustainable Returns (Income) to Land: |
|||
|
(i) |
Total: G less 'payments' to capital and labour: |
Rs (23 510 - 5 660 - 4 500) = Rs 13 350 |
|
|
(ii) |
Per unit (ha) of land: |
Rs (13 350/2.8) = Rs 4 768 |
|
|
N. Net Farm Sustainable Returns (Income) to Capital: |
|||
|
(i) |
Total: G less 'payments' to land and labour: |
Rs (23 510 - 5 000 - 4 500) = Rs 14010 |
|
|
(ii) |
Per unit (Rs 100) of capital: |
Rs 14 010/566 = Rs 25 |
|
|
O. Net Farm Sustainable Returns (Income) to Family Labour: |
|||
|
(i) |
Total: G less 'payments' to land and capital: |
Rs (23 510 - 5 000 - 5 660) = Rs 12 850 |
|
|
(ii) |
Per unit (day) of family labour: |
Rs 12 850/450 = Rs 29 |
|
The information generated by the residual-value method has two main uses. First, comparisons are often required between farms regarding the relative productivity of one particular resource, e.g., the productivity of labour on Java paddy farms vs Sumatra paddy farms, or between farms in the same village. Second, the information may be used as a guide for determining a reasonable economic price at which to buy or sell or exchange resources. In the example of Table 7.6 the productivity value of family labour is Rs 29 per day. Given this information, family members would have a better basis for deciding whether to stay home and work on the farm or to look for off-farm work. Likewise, an economic price to pay for purchasing or renting land can also be approximated. On the farm of Table 7.6 the productivity value of land is Rs 4 768 per ha. This offers a guide to the annual rent the farmer could pay to acquire a little more land.
Low or negative resource productivities
When the residual method is applied to determine the productivity of the family labour resource it is not unusual to find that, by the time all other resources have been 'paid off at their going rates, there is no production value left to assign to labour. Residual labour value might even be negative (meaning that total productivity of this particular farm system is so low that even the other resources of capital and land cannot earn a normal rate of return). This simply reflects the real world. Sometimes the system weaknesses which this indicates can be removed by restructuring the system, using the methods discussed in Chapter 9. But, in the Asian context, more often than not the problems lie at higher levels - in systems of Order Level 13, 14, 15 - and their solution is beyond the capacity of individual farmers.
The management resource
So far, no attention has been paid to system management as a separate factor or resource. Indeed, on small Asian farms of Types 1 and 2 in particular but also often of Types 3 and 4, management is so closely inter-linked with family labour that it might not be practicable to try to distinguish between them. This contrasts with the situation on farms of Types 5 and 6 and on many Western farms where the farmer expends relatively less physical effort in the fields but relatively more mental effort in planning the use and assignment of his or her greater stock of capital. (Indeed, on Type 6 farms (estates) there will usually be specialist hired management.) In such situations it is often desirable to evaluate the contribution of management as an important separate resource.
This can be done by extending the residual-value method to encompass the four resource categories of land, capital, labour and management rather than just the first three. The value of management is then estimated as the residual of total farm output value after land, capital and labour have been 'paid off, i.e., a value of Rs 8350 for the farm of Table 7.4. Management as a resource, however, exists in a qualitative dimension. It can really only be measured quantitatively in terms of what it achieves after the event, i.e., ex post, as discussed in Section 7.2.5.
The resource productivities obtained in the previous section refer to resource use on the farm as a whole. In some situations it may also be necessary to derive such productivity data for each resource used in each activity. As previously, these estimates are obtained using the residual method. Again, based on the farm of Table 7.4, an example is shown in Table 7.7 where the productivity-based values of land, capital and labour when used in producing maize are derived as Rs 1 280, Rs 2 and Rs 5, respectively, per ha of land, per Rs 100 of capital and per day of family labour. Obviously the value of capital employed in maize is very low. Just how low it is in comparison with capital employed in the other activities could only be determined if similar analyses were conducted for paddy, bamboo and cattle.
TABLE 7.7 - Worksheet for deriving per Unit Productivity Values of Farm Resources (Land, Labour, Capital) on an Activity Basis for the Maize Activity of Tables 7.4 and 7.5 (Annual Basis)
|
Data on Resources used for Maize | |||
|
(a) |
land used |
|
1.0 ha |
|
(aa) |
value of land |
(Rs 50 000/2.8) |
17 857 Rs |
|
(aaa) |
'payment' to land |
(Rs 17 857 x 0.10) |
1 786 Rsa |
|
(b) |
capital used (except land) |
|
6 200 Rs |
|
(bbb) |
'payment' to capital |
(Rs 6 200 x 0.10) |
620 Rs |
|
(c) |
family labour used |
|
100 days |
|
(cc) |
value of labour |
(Rs 10 x 100) |
1 000 Rs |
|
(ccc) |
'payment' to labour |
|
1 000 Rs |
|
Net Returns to Land used for Maize | |||
|
(i) |
Activity total: G for maize less payments to capital and labour: |
Rs (2 900 - 620 - 1 000) = Rs 1 280 | |
|
(ii) |
Per ha of land: |
Rs 1 280/1 = Rs 1 280 | |
|
Net Returns to Capital used for Maize | |||
|
(i) |
Activity total: G for maize less payments to land and labour: |
Rs (2 900 - 1 786 - 1 000) = Rs 114 | |
|
(ii) |
Per Rs 100 of capital: |
Rs 114/62 = Rs 2 | |
|
Net Returns to Labour used for Maize | |||
|
(i) |
Activity total: G for maize less payments to land and capital: |
Rs (2 900 - 1 786 - 620) = Rs 494 | |
|
(ii) |
Per day of family labour: |
Rs 494/100 = Rs 5 | |
a Note that this is an overestimate because there is some double cropping of maize and paddy.
Individual productivity values for land, labour, capital and management derived by the residual-value method either on a whole-farm basis (Section 7.2.3) or on an activity basis (Section 7.2.4) can provide a useful guide in farm diagnosis. Such individual productivity values, however, must be used with caution. They suffer from two deficiencies. First, the residual-value method, by its nature, attributes all the 'excess benefits' to the resource being appraised; other resources are only 'paid off at their opportunity or market cost. Second, productivity values on an individual resource basis conflict with a systems view of the farm. From a systems perspective, the essence of farm production is that management is applied to a conglomerate set of factors (i.e., land, labour and capital resources as a totality) which interrelate and interact with one another and with management to produce outputs. From such a perspective, evaluation of the farm system implies that the resources used should be assessed as a totality or unified whole, not as a set which can be broken down into separate categories of land, labour, capital and management. Such a systems view leads to the concept of total factor (or resource) productivity whereby productivity is assessed on a whole-farm or (with some qualification) activity basis (Dillon and Hardaker 1993, pp. 79-91).
Total factor productivity
The essence of productivity is output per unit of input over some specified time period. Total factor productivity is thus total output divided by total input. As in the case of individual resource productivities, the problem of outputs and of inputs each being of diverse physical forms is met by aggregating each to their respective total on the basis of the common unit or numéraire of money value based on market price for outputs and market or opportunity cost for inputs. Using again the example farm of Tables 7.1, 7.4 and 7.5, this is done on a whole-farm basis in the worksheet of Table 7.8. Apart from the aggregation of land, labour and capital into a single money value, the analysis differs from that of Table 7.4 in that the costs of all inputs, whether or not they actually involve cash payments, are included. The appraisal is in economic rather than merely financial terms. Thus family labour is now included as an input cost at its market value of Rs 10 per day and an assumed opportunity cost of 10 per cent interest is now charged on all capital including land. Likewise, depreciation is included because it also is a cost of production. Total costs are thus Rs 35 650 as compared with the cost of B + C + D = Rs (12 500 + 3 510 + 4 480) = Rs 20 490 used in deriving net sustainable returns (item G) in Table 7.4. Gross total factor productivity is thus:
Total gross returns/Total costs = 44 000/35 650 = 1.23,
i.e., Rs 1.23 of output is obtained per Rs of the conglomerate input of land, labour and capital resources. Analogously, net total factor productivity is:
Total net returns/Total costs = 8 350/35 650 = 0.23,
i.e., Rs 0.23 net is obtained per Rs of input or, in percentage terms, the net return per Rs of input is 23 per cent.
Return to capital and to equity
Table 7.8 also presents two other whole-farm evaluation measures of an economic productivity nature. These are return on capital (including land) and return on equity used in the farm. (Equity is total farm capital minus borrowed capital used in the farm - it is the farmer's share of the total farm capital.) Capital constitutes a relevant basis for evaluation because capital (including land) constitutes both the physical resource base and the investment base on which farm production occurs; as well, the capital value of the farm is also its market value if it is to be bought or sold. Thus return on capital indicates how efficiently the farm asset base is used while return on equity can be compared with the rate of return that may be available from alternative investments.
Return on capital is calculated as:
(Total net returns/Total capital)100 = (8 350/106 600)100 = 7.83%.
Now suppose, as specified in Table 7.8, that the farmer has a debt of Rs 10 000 on the farm at an interest cost of 10 per cent per annum, i.e., he or she only has an equity of Rs 96 600 (or 91 per cent) in the total farm capital of Rs 106 600.
TABLE 7.8 - Worksheet for deriving Total Factor Productivity, Return on Capital and Return on Equity on a Whole-farm (Annual) Basisa
|
Measure on a Whole-farm Basis |
Whole-farm |
Value |
||
|
(Rs) |
(Rs) |
|||
|
Total Gross Returns |
|
44 000 |
||
|
Costs |
|
|
||
|
|
Direct costs |
|
|
|
|
|
|
Labourb |
6 900 |
|
|
|
|
Other |
10 100 |
|
|
|
Fixed costs |
|
|
|
|
|
|
General charges |
910 |
|
|
|
|
Other |
2 600 |
|
|
|
Depreciation |
4 480 |
|
|
|
|
Interest on capitalc |
|
|
|
|
|
|
Land |
5 000 |
|
|
|
|
Other |
5 660 |
|
|
Total Costs |
|
35 650 |
||
|
Total Net Returns |
|
8 350 |
||
|
Gross Total Factor Productivity |
|
1.23 |
||
|
Net Total Factor Productivity |
|
0.23 |
||
|
Total Capital |
|
106 600 |
||
|
Return on Capital |
|
7.83% |
||
|
Equity Capitald |
|
96 600 |
||
|
Borrowed Capitald |
|
10 000 |
||
|
Cost of Borrowed Capitald |
|
1 000 |
||
|
Return on Equity |
|
7.61% |
||
a Data, as required, from Tables 7.1, 7.4, and 7.5.b Including family labour at Rs 10 per day.
c Assuming an opportunity cost of 10 per cent per annum.
d In distinction to Table 7.4, it is assumed here that the farm has a debt of Rs 10 000 at an interest cost of 10 per cent per annum.
Return on equity is calculated as:
[(Total net returns - Cost of borrowed capital)/Equity capital]100 = [(8 350 - 1 000)/96 600] 100 = 7.61%.
Reference was made in Section 7.2.3 to the possibility of estimating a return to the farm's management resource based on the residual-value method. A preferable evaluation of management, however, is provided by the measures of total factor productivity, return on capital and return on equity. Other things being equal, the larger these indices, the better the performance of the farm's management. Obviously, these indices are also useful in inter-farm comparisons. It should be recognized, however, that these total productivity measures are not fully under the control of the farm's management. They can be heavily influenced by, in particular, the vagaries of Nature (affecting product yields) and the market (affecting product prices), as well as by other sources of risk (as outlined in Chapter 11).
Activity-specific total productivity measures
Analogously to the case for the whole farm, total factor productivity, return on capital and return on equity can be estimated on an activity-specific basis. This is exemplified by the worksheet of Table 7.9 for the paddy activity of Table 7.4. Such estimates are again useful for inter-farm comparisons of like activities but they are rather qualified. First, they take no account of those farm resources which, while crucial to the operation of the farm, cannot be allocated to any specific activity. For the example farm these are listed in the 'farm-as-a-whole' column of Table 7.4. Note, however, that in contrast to Table 7.4, the analysis of Table 7.9 allows for an interest cost on the capital value (Rs 44 642) of the 2.5 ha of land used for paddy. Second, estimation on an activity basis may be distorted by double counting. Thus, while the example farm has a total area of 2.8 ha, through double cropping of at least 0.7 ha with paddy (2.5 ha) and maize (1 ha) it appears to use 3.5 ha of land. This double cropping is not allowed for in the worksheet calculations of Tables 7.8 and 7.9.
7.3.1 Level I analysis - the whole farm
7.3.2 Level II analysis - the household
7.3.3 Level III analysis - the fixed-capital structure
7.3.4 Level IV analysis - the individual activities
7.3.5 Level V analysis - the underlying processes
7.3.6 Summary
As noted above, comparative analysis of a subject farm with other relevant farms is intended to identify weaknesses which might exist within a farm-household system and to diagnose their causes, i.e., identify causal factors. These might arise at any level within the farm system (Figure 1.2). A full and formal comparative analysis would follow the sequence of steps I to V listed in Table 7.10. This breaks the problem of diagnosis down into successively smaller but more numerous 'bits' for successively more detailed investigation, in much the same way as a mechanic trouble-shoots a recalcitrant engine. These bits are the subsystems of the farm-household system. In practice, some of these levels of analysis can usually be by-passed since the analyst will often have an idea of where the problem lies and will proceed directly to that level, e.g., directly to an examination of the kind of technologies used on the farm at level V, rather than systematically working through analysis at levels I, II, III and IV.
Analysis at this level is concerned with developing measures of performance of the whole-farm system. Some of these measures, those relating to financial performance, were developed as factors E, F,... J in Section 7.2.1. But to the extent that the farm will also have other operating objectives, some of the other performance criteria of Section 6.2 will apply. The next step in level I analysis is to compare the subject farm with other relevant farms, or with some standard farm (as discussed in Section 7.4.2), in terms of these selected performance criteria, as in Table 7.11 (which uses a different subject farm to that of Section 7.2). Sources of data for such whole-farm comparisons are discussed in Section 7.4.
TABLE 7.9 - Worksheet for deriving Total Factor Productivity, Return on Capital and Return on Equity on an Activity-specific (Annual) Basisa
|
Measure on an Activity Basis |
Paddy Activity |
|||
|
(Rs) |
(Rs) |
|||
|
Gross Return |
|
31 000 |
||
|
Costs |
|
|
||
|
|
Directb |
8 000 |
|
|
|
|
Fixed |
970 |
|
|
|
|
Depreciation |
500 |
|
|
|
|
Interest on capitalc |
|
|
|
|
|
|
Land |
4 464 |
|
|
|
|
Other |
1 330 |
|
|
Total Costs |
|
15 264 |
||
|
Net Return |
|
15 736 |
||
|
Total Factor Productivity |
|
|
||
|
|
Gross |
|
2.03 |
|
|
|
Net |
|
1.03 |
|
|
Capital |
|
57 943 |
||
|
Return on Capital |
|
27.2% |
||
|
Equity Capital |
|
52 508 |
||
|
Borrowed Capitald |
|
5 435 |
||
|
Cost of Borrowed Capitale |
|
543 |
||
|
Return on Equity |
|
28.9% |
||
a Data, as required, from Tables 7.1, 7.4 and 7.5.b Including family labour at Rs 10 per day.
c Assuming an opportunity cost of 10 per cent per annum.
d Pro rata share of loan of Rs 10 000 based on paddy activity's share of total farm capital, i.e., Rs (10 000) (57 943/106 600) = Rs 5 435.
e At an interest charge of 10 per cent per annum.
From Table 7.11, relative to the comparative farm, the subject farm is performing badly in terms of productivity and profitability, but well in terms of the stability, diversity and time dispersion of income, and of sustainability. In following discussion, it will be assumed that profitability, i.e., net farm income, is the operating objective of interest, thus the subject farm's low comparative profitability levels revealed in Table 7.11 warrant closer examination.
TABLE 7.10 - Broad Sequence of Steps in Comparative Analysis
|
Stage or Level in Comparative Analysis |
Subsystema |
Action |
|
I |
Whole-farm (Older Level 10) |
Develop measures of whole-farm performance, as in Section 7.2. |
|
II |
Household (Order Level 11) |
Look for causal factors (of I) in the household. |
|
III |
Service Matrix (Order Level 9) |
Look for causal factors (of I) in the amounts/kinds of fixed farm capital used. |
|
IV |
Enterprises and Activities (Order Levels 3 to 7) |
Look for causal factors (of I) arising in the mix of activities. |
|
V |
Processes (Order Levels 1 and 2) |
Look for causal factors (of IV) arising: |
a Relative to the farm-household system of Order Level 12.
b Order Levels as specified in Section 1.3 and Figure 1.2.
Possible causal factors to look for within the household can only be suggested: they will vary with farm type and cultural environment. They fall into two general groups: factors which have a direct effect on farm performance and those that might have only an indirect effect (broadly, sociological non-economic factors). Some of these can be quantified and this will allow comparison with other standard farms. Others have only an implicit effect on farm performance and do not permit quantitative comparisons. A few are listed in Table 7.12. These and similar household factors would be arranged for inter-farm comparisons as in Table 7.11. Only two of the factors of Table 7.12 - (7) health and (8) debt - might require elaboration.
The importance of family health as a factor underlying farm performance is often overlooked, especially by analysts removed from the village who take the absence of malaria, cholera, intestinal diseases, trachoma, etc. for granted. But in many areas - the limestone hills of Java, southern Sri Lanka, South India, the terai of Nepal, the African lakes - comparative analysis aimed at uncovering reasons why farms and indeed whole villages are 'inefficient' need often proceed no further than the factor of poor family health.
Household debt as a causal factor requires more careful interpretation. If incurred for food, debt may be taken as an indicator of poverty and of the inability of the subject farm to provide family subsistence. But where it occurs in comparatively rich areas such as the Kandy hills of Sri Lanka, it is as likely as not incurred for frivolous purposes. Equally, household and social debt might often be incurred to acquire status-giving possessions or to achieve a higher social standing. Whatever its origin, the pressures for debt repayment often offer good insights into why a quick-growing but soil-erosive and less profitable crop has been grown in preference to a more profitable and less erosive crop; and (as often occurs with clove crops in Sri Lanka and paddy crops in some Central Java villages) why a crop has been sold in its standing state at a fraction of the price that the household could have received if they had waited a little longer and harvested the crop themselves. In short, the occurrence of debt in the household component of a system, or more precisely the pressures for its liquidation, can sometimes lead to apparently bizarre decisions within the farm component.
TABLE 7.11 - Level I Analysis: Comparative Analysis with Some Whole-farm Performance Criteriaa
|
Performance Criteria |
Unit |
Subject Farm |
Comparative Farm | |
|
(1) Productivity (e.g., of staple food grain): |
|
|
| |
|
|
total farm |
t |
7 |
10 |
|
|
per ha |
t |
2.8 |
5 |
|
|
per family labour day |
kg |
25 |
40 |
|
(2) Profitability: |
|
|
| |
|
|
total farm gross margin |
Rs |
25 000 |
45 000 |
|
|
per ha |
Rs |
10 000 |
22 500 |
|
|
per Rs 1 000 of fixed capital |
Rs |
500 |
750 |
|
|
per family labour day |
Rs |
89 |
180 |
|
(3) Stability of annual income |
|
|
| |
|
|
Coefficient of variation (CV) |
% |
12 |
36 |
|
(4) Income diversity: |
|
|
| |
|
|
number of activities |
|
6 |
4 |
|
|
number of separate products |
|
14 |
8 |
|
(5) Flexibilityb: |
|
fair |
fair | |
|
(6) Time dispersion of income: |
|
|
| |
|
|
relative time-dispersion (RTD) |
|
0.9 |
0.4 |
|
(7) System sustainabilityb: |
|
high |
low | |
|
(8) Environmental compatibilityb: |
|
fair |
fair | |
a Criteria are as per Section 6.2.b Assessed subjectively in terms of poor/fair/good for criteria (5) and (8) or low/average/high from an overall (i.e., private and social, economic and resource) perspective for criterion (7).
The factors of Table 7.12 might afford diagnosis of a problem, but it is difficult at this level for the farm management analyst, as analyst, to prescribe solutions. Only a few of the factors, e.g., attitudes to and management of debt, might be amenable to direct adjustment. Most of the factors will either be given (household and labour-force structure and age) or they will exist as a consequence of higher-order (social) systems in such fields as literacy, health, education. However, if the analyst happens to be an extension agent or rural development officer (as he or she is likely to be) then he or she can take action: e.g., ensure that health services are delivered to the village, plan a local literacy campaign, develop a relevant farm-and-home extension package.
TABLE 7.12 - Level II Analysis: Indicative Aspects of the Household to be considered in Diagnosis and Comparative Analysis
|
(1) |
Number of household members |
an indication of total needs for food and/or cash, thus of demands made on the system, thus of the kinds of farm activities which will be selected |
|
|
(2) |
Income per household member |
a direct quantitative measure of well-being |
|
|
(3) |
Labour force |
a direct quantitative measure of available labour resources |
|
|
(4) |
Labour force per ha |
a better measure than (3) and an indication of the potential for further farm adjustment, intensification |
|
|
(5) |
Age structure of workforce |
a qualitative measure of the labour resource |
|
|
(6) |
Literacy of all farm adults |
another qualitative measure of labour resource and management potential, as well as an indicator of the potential of the household to absorb new farming ideas |
|
|
(7) |
Health status |
another qualitative measure of the labour resource, but more importantly an indicator of general household morale |
|
|
(8)
|
Household debt
|
for |
|
|
|
(a) essential food, subsistence |
||
|
|
(b) household items |
||
|
|
(c) social purposes |
||
|
|
(d) farm operation or development |
||
|
(9)
|
Indicators of progress, development
|
(a) school attendance |
|
|
(b) participation in non-agricultural activities |
|||
|
(c) contact with extension service |
|||
|
(d) possession of safe water supply, sewing machine, bicycle, radio, etc. |
|||
Assuming that the specific causal factors cannot be diagnosed or removed at this second level, the analysis moves on to level III.
Analysis at level III is concerned with the farm service matrix - fixed-capital investment in land, buildings, irrigation channels, institutional arrangements etc. that serves the whole farm. Again a large number of possible causal factors of poor farm performance might be found within this capital stock; the art is to concentrate on those capital items that are intuitively most relevant. In practice, it will be necessary to consider the most appropriate measure of each item: for some the best measure is their size/quantity/capacity; for others their capital value; others might be measured as 'poor/fair/good'.
An example of comparative analysis applied to the farm service matrix is presented in Table 7.13. It relates to the same farms as in Table 7.11. Might any of the factors listed in Table 7.13 account for the subject farm's poor performance with a total gross margin of only Rs 25 000 vs Rs 45 000 on the comparative standard farm? Looking first at item (1) of Table 7.13, both farms have a roughly similar capital structure. However, looking at item (2), the subject farm has 50 per cent more good land than the comparative farm; but in spite of this it produced much less income, so clearly its poor performance is not due to insufficient or poor quality land. In item (3), irrigation investments are approximately similar, but the irrigation system on the subject farm covers only 40 per cent of the farm area while that on the comparative farm covers 90 per cent. Inadequate irrigation could be the first causal factor in low income.
TABLE 7.13 - Level III Analysis: Aspects of Fixed Farm Capital to be considered in Diagnosis and Comparative Analysis
|
Capital Item |
Unit |
Subject Farm |
Comparative Farm | |
|
(1) Fixed capital value (except land): |
|
|
| |
|
|
total |
Rs |
50 000 |
60 000 |
|
|
per ha |
Rs |
20 000 |
30 000 |
|
|
per labour unit (worker) |
Rs |
(3) 16 700 |
(4) 15 000 |
|
(2) Land: |
|
|
| |
|
|
total |
ha |
2.5 |
2 |
|
|
good paddy land |
ha |
1.5 |
1 |
|
|
poor crop land |
ha |
1 |
1 |
|
(3) Irrigation system: |
|
|
| |
|
|
total value |
Rs |
15 000 |
13 000 |
|
|
value per ha (of farm) |
Rs |
6 000 |
6 500 |
|
|
coverage |
ha |
1 |
1.8 |
|
|
condition (poor/fair/good) |
|
good |
fair |
|
|
water supply |
mths |
7 |
6 |
|
(4) All field equipment: |
|
|
| |
|
|
total value |
Rs |
25 000 |
20 000 |
|
|
value per ha |
Rs |
10 000 |
10 000 |
|
(5) Work oxen: |
|
|
| |
|
|
number |
no. |
3 |
2 |
|
|
number per ha |
no. |
1.2 |
1 |
|
(6) Grain-drying facilities: |
Rs |
0 |
4 000 | |
|
(7) Grain storage: |
Rs |
1 000 |
5 000 | |
Continuing, levels of investment in field equipment and oxen are apparently not causal factors. But the next item could be - there is no grain drying facility on the subject farm. How then does the farmer dry his or her paddy and corn? This possible system weakness is partly confirmed by the next item - only Rs 1 000 are invested in crop storage compared with Rs 5 000 on the comparative farm. These factors provide sufficient reason for proceeding to analysis at level IV relating to the economics and the technical structures of individual activities.
As per Table 7.10, level IV analysis concerns the individual enterprises/activities of Order Levels 3 to 7. (In doing so it also implicitly considers the farm resource pool of Order Level 8.) At this stage, as exemplified in Table 7.14, two types of comparison are required: comparisons (a) between activities on the subject farm, and (b) between pairs of similar activities on the subject and comparative farm. Both types of comparison have two objectives: first, to evaluate the relative performance of each activity in the system with a view to possibly changing the activity mix, i.e., replacing inefficient activities with efficient ones; and second, to identify specific agro-technical-economic factors within each activity which warrant closer investigation at the next stage, level V. (In actual practice, analysis at levels IV and V is often combined.)
The first step is construction of activity unit budgets (although, as noted previously, these are better termed operating statements because they refer to past events rather than to future projections). Obviously the more detailed these budgets/statements are, the more effective they will be in identifying causal factors. Detailed budget analysis may be undertaken, as in Table 7.14, or only summary unit budgets may be developed as shown at the top of the table. These summaries are sufficient to meet the first objective of intra-farm comparison noted above. Considering the budget results for GM per ha, the presence of the relatively unprofitable roselle activity offers one reason why total income is so low on the subject farm. The indicated action would be to eliminate roselle from the plan and increase the size of the more profitable paddy or maize activities. (Even though paddy on the subject farm is poor relative to the standard farm, it is still twice as profitable as roselle.) However, from Section 7.3.3. irrigation shortage would probably prevent expansion of these high water-using paddy and maize crops. The cropping system followed on the more profitable comparative farm offers another possibility: replace roselle with oilseed and thus increase farm TGM by Rs (3 500 - 2 000) or Rs 1 500 per ha. The analyst would, of course, also consider other low water-using crops as possible substitutes. If confident that the problem of low farm income could be solved by substituting better activities, he or she would at this point begin restructuring the system (Chapter 9). If not, the analyst would proceed to the second objective of level IV analysis, i.e., inter-farm comparison, and derive further comparative factors from the activity budgets which might permit identification of specific weaknesses within the three existing activities.
Some of the possible derived factors are listed in Table 7.14. From the budgets there are apparently no problems with the maize activity, and the possibility of removing roselle entirely has been noted; attention would therefore be concentrated on paddy, specifically in relation to the lower price received (Rs 1 500 vs Rs 1 700), the lower net recovered yield (four vs five tonnes) and the higher direct costs, all of which result in a much lower GM for paddy than on the comparative farm. Also, in the prior level III analysis there were indications that poor post-harvest and irrigation practices might be present, thus the first factors to consider at level IV are those which might verify or remove these suspicions.
From Table 7.14, inadequate irrigation is obviously an important factor in low paddy yield. But assuming that water supply is limited, there is little the farmer can do about it - except reduce his or her paddy area and substitute more dryland crops.
The next item, grain-drying method, explains why there was no capital investment in this item in Table 7.13: the farmer spreads the paddy out to dry on the village road (where inevitably there would be a 15 to 20 per cent loss to the village chickens and ducks). This partly explains the low net sales of only four tonnes per ha. The grain also becomes contaminated to some degree - which probably explains the relatively low sale price received in comparison with the standard farm.
TABLE 7.14 - Level IV Analysis: Example of Comparative Analysis of Individual Activities
|
|
Subject Farm |
Comparative Farm |
|||||
|
Paddy |
Maize |
Roselle |
Paddy |
Maize |
Oilseed |
||
|
Unit Budget |
|||||||
|
|
Net sales (t/ha) |
4 |
3 |
2 |
5 |
3.3 |
2 |
|
|
Sale price (Rs/t) |
1 500 |
1 500 |
2 000 |
1 700 |
1 500 |
2 500 |
|
|
Yield value (Rs/ha) |
6 000 |
4 500 |
4 000 |
8 500 |
4 950 |
5 000 |
|
|
Direct costs (Rs/ha) |
2 000 |
1 500 |
2 000 |
1 500 |
2 000 |
1 500 |
|
|
GM (Rs/ha) |
4 000 |
3 000 |
2 000 |
7 000 |
2 950 |
3 500 |
|
Factors Derived from Budget |
|||||||
|
|
Irrigation water (units per ha) |
6 |
naa |
na |
10 |
na |
na |
|
|
Grain-drying method |
road |
road |
fence |
slab |
slab |
slab |
|
|
Grain storage |
shed |
shed |
shed |
silo |
silo |
silo |
|
|
Family labour (days/ha) |
50 |
30 |
40 |
50 |
35 |
30 |
|
|
Hired labour (days/ha) |
40 |
0 |
10 |
10 |
0 |
0 |
|
|
Humus applied (t/ha) |
6 |
0 |
0 |
6 |
0 |
0 |
|
|
N fertilizer (kg/ha) |
150 |
50 |
40 |
150 |
40 |
30 |
|
|
Agricides (Rs/ha) |
50 |
0 |
50 |
0 |
0 |
0 |
|
|
Oxen (days/ha) |
5 |
3 |
4 |
10 |
4 |
4 |
a na: not applicable to these rainfed crops.
Likewise, the next item, storage, explains the low capital investment reported for this item in Table 7.13. Paddy is stored in an open mud-floor thatch-roofed shed whereas on the comparative farm it is stored in a rodent-proof silo. These two practices, roadway drying and insecure storage, could easily result in a loss of 30 to 45 per cent of harvested grain and thus go a long way to explaining the relatively low farm TGM of Rs 25 000 vs Rs 45 000 on the standard farm (Table 7.11).
The next item, crop labour requirements, also looks a promising lead. Noting that direct paddy costs of the two farms are Rs 2 000 vs Rs 1 500, this extra cost on the subject farm may be explained largely in terms of hired labour (30 days x Rs 10). But why was it necessary to employ so much more labour than on the comparative farm?
The last item listed, ox days used per ha, could also be significant. The subject farm used only half as much oxpower inputs for paddy as the comparative farm. Why the large difference? What actual operations were performed with these oxen?
To find the answers to these questions, the analysis now moves to the next level and looks at specific agro-technical processes which underlie these paddy activities.
'Planting rice is never fun;
Bent from mom till set of sun;
Cannot stand and cannot sit;
Cannot rest for a little bit.'From an old Filipino song.
Analysis to this level has broken the search for causal factors down into successively smaller 'bits' or subsystems of the farm-household system. It now becomes a search within the lowest subsystems, the agro-technical processes of Order Levels 1 and 2. Causal factors here might be found in either of the components of a process, the intensity levels at which they are used or the ways - technologies - by which they are applied (Section 5.1).
From the level IV analysis, the problems within the paddy activity may be summarized as:
(i) (implied) low field yield.
(ii) (implied) high wastage in drying and storage.
(iii) low net (recovered/sold) production.
(iv) low sale price (probably due to contamination).
(v) high direct costs (mainly due to high labour inputs).
Accordingly, the causal factors considered in level V analysis will concentrate on these five areas. The analyst would list each step or operation in the activity from land preparation to sale of the crop and, within each operation, list processes which might be relevant in relation to the problematic performance factors (i) to (v) above, as shown in Table 7.15 on a per ha basis.
Factor (1) of Table 7.15 now explains why ox-day inputs on the subject farm were so low in Table 7.14: the farmer performs only the minimum tillage operations in comparison with the more intensive practice followed on the standard farm. This factor has two effects: a direct effect on tilth and thus on yield, and an indirect effect through weed reduction both on yield and on subsequent hand-weeding requirements (and thus on labour and direct costs). Factors (1), (4) and (6) are closely related. These factors imply a process. Since the technology, i.e., use of oxen, cannot be changed, attention would concentrate on the intensity component: ox-day inputs or number of times the field is initially cultivated vs yield and subsequent labour costs for weeding. The kind of response and cost relationships which are probably present are indicated in Figure 7.2. For optimization of this response, the separate relationships shown - number of cultivations vs yield, cost of weeding labour and cost of oxpower - could be consolidated into a single 'response curve' showing net return in relation to number of cultivations. (Ox days would bear no cost unless they were hired.)
The point in introducing simple response analysis into the inter-farm comparison is that, while recommendations for action on the subject farm might be based on what the comparative farm is doing, it is possible that both farms are suboptimal in regard to factors (1) to (6) of Table 7.15. Both may benefit from response analysis of at least the main factors, although probably the subject farm more so than the standard farm.
TABLE 7.15 - Level V Analysis: Example of Comparative Analysis of Processes
|
Process Factors relating to Paddy |
Subject Farm |
Comparative Farm |
|
|
(1) Land preparation: |
|||
|
|
ox days/ha |
5 |
10 |
|
|
plow, times |
2 |
2 |
|
|
harrow, times |
1 |
3 |
|
(2) Seed used: |
|||
|
|
type |
HYV |
HYV |
|
|
quantity, units/ha |
5 |
5 |
|
|
days in nursery |
30 |
23 |
|
(3) Transplanting: |
|||
|
|
spacing pattern |
random |
square |
|
|
distance apart, cm |
20-30 |
25 |
|
|
method |
'eye' |
pegboard |
|
(4) Weeding: |
|||
|
|
hand weed, times |
3 |
1 |
|
|
hired labour, days/ha |
28 |
0 |
|
(5) Inter-row cultivation: |
|||
|
|
method |
none |
sorok |
|
|
number |
0 |
2 |
|
(6) Grain: |
|||
|
|
harvested, t/ha |
4.5 |
5 |
|
|
lost, t/ha |
0.5 |
0 |
|
|
sold, t/ha |
4.0 |
5 |
Moving to factor (2) in Table 7.15, both farms are similar in regard to variety and amount of seed used, but dissimilar in the age at which seedlings are transplanted to the field. This latter factor again suggests the possibility of process optimization through response analysis of rice yield vs seedling age or days in nursery, as suggested in Figure 7.3. Assuming the response of yield to seedling age is as depicted in Figure 7.3, and given that the cost per day of holding seedlings in the nursery is virtually zero, the optimal age of seedlings for transplanting is about 25 days, i.e., where rice yield is a maximum. This is five days less than the seedling age of 30 days used on the subject farm which clearly leads to a suboptimal yield.
The transplanting factors listed in Table 7.15 also indicate a significant difference between the farms in terms of the transplant method used. This technology component is probably more important than the planting-intensity component (which could if necessary be measured in terms of seedling spacing or population). The technologies can be directly compared by partial budgeting (Section 4.5) of the labour and cost of random planting vs the labour and cost of purposeful planting based on guides provided by a pegboard or knotted rope. This transplanting factor is also closely associated with others, particularly (4) and (5). Although random spacing on the subject farm might possibly be cheaper than peg planting, it leads, in factor (5), to an inability to control weeds quickly and with minimum labour by use of a 'sorok' (a light weeding and cultivating wheel which is pushed along by hand between the paddy rows). In turn this inability leads in factor (4) to the necessity to hand-weed and thus hire 28 days of labour, increasing direct costs on the subject farm by Rs 280. As noted above, the need for more weeding on the subject farm, by whatever means, is probably due to inadequate initial land preparation.
FIGURE 7.2 - Yield, Hand-weeding Cost and Oxpower Cost in relation to Number of Cultivations
Factor (6) provides more detail regarding post-harvest operations. Losses here are serious on the subject farm: 0.5 tonne is lost in the roadway drying and storage operations. There are several ways by which these operations could be improved through response analysis; e.g., in drying, quantity of grain lost vs the number of children employed to keep the ducks away; and in storage, alternative levels of investment to give alternative levels of security. But obviously, with factors and processes of this type, the alternative technologies are more important than their intensity levels. This requires only comparative partial budgeting: in drying, the cost of a slab drying-floor and low grain losses vs free drying on the road with high losses; in storage, the cost of one or two alternative types of new storage and no losses vs continued free use of the open shed with high grain losses.
FIGURE 7.3 - Response of Rice Yield to Seedling Age at Transplanting
This concludes the illustrative comparative analysis of the farm-household system of the subject farm considered in this section. The analysis has diagnosed problems at five system Order Levels as per the schema of Table 7.10 and offered prescriptions for their solution:
(i) From level I (whole-farm system), there is a need to increase the farm's profitability.(ii) From level II (the household system), the family should be encouraged to participate more actively in extension activities, particularly those relating to the main crop (paddy). But equally this recommendation might also be directed at the extension service: it should strengthen its efforts in providing an accessible and relevant extension package for under-performing farmers. (The content of this package will be clear from this summary.)
(iii) From level III (farm capital structure), the farm should increase investment in grain drying and storage.
(iv) From level IV (farm activities), weak activities such as roselle should be replaced by more profitable activities and it might also pay to concentrate the limited available water on a smaller area of paddy.
(v) From level V (processes), the indicated action in restructuring the paddy activity is to instigate better land preparation, earlier transplanting of seedlings, purposeful rather than random planting and weed control by sorok rather than by hand.
Expansion to Order Level 13 systems
The above discussion of comparative analysis has been concerned with adjustments and improvements on individual farms. In Asian (and Pacific and African) villages, farm systems are markedly homogeneous with respect to the types of crops grown, types and levels of resources used, technologies employed, and technical and economic problems encountered. If production is efficient (or inefficient) on one farm, it is likely to be efficient (or inefficient) on most other farms in the village. Further, and in spite of the previous discussion at levels IV and V regarding restructuring the activity mix and adjusting the underlying processes on the subject farm, it is not easy to overturn village conventions and traditions. If farm development on individual farms is to occur, it must often be based on village consensus. Thus it will often be more productive to apply comparative analysis to the 'farm-village' system, working with representative or mean or model farm data, rather than with specific individual farms. With a few obvious modifications, the same step-by-step procedures outlined above can be applied to removing many of the weaknesses which might occur in Order Level 13 village systems, which would of course achieve the same goals as analysis aimed at individual farms.
7.4.1 Data for the subject farm
7.4.2 Data for the comparative or standard farm
Data required for limited evaluation (Section 7.2) of the farm component of a farm-house could system and for comparative analysis of the whole system will vary in degree of necessary detail. Since limited evaluation concerns the farm as a whole, but not the household component (Figure 7.1), the data for this may be of an aggregative nature. The degree of required detail increases if the examination then proceeds in prescriptive mode to a comparative analysis of successively lower Order Level systems.
There are four main sources of data relating to a subject farm:
· written records maintained by the household.· intensive interviews with members of the household.
· records maintained for the farm component by a club, cooperative or other farm association.
· records maintained for the farm and/or its household components by an extension program or other agency specially designated for this purpose.
Very few if any farms of Types 1, 2 or 3 keep any sort of useful farm or household records. Type 4 farms might maintain records relating to their specialist activities, but seldom relating to the household component. Large commercial farms and estates of Types 5 and 6 usually keep good records but often not of the degree of detail required for lower Order Level comparative analysis. Thus even if farm/estate records are available they usually must be supplemented by personal interviews. These will usually yield satisfactory results when applied on large farms and estates (where incidentally the household component will be of less importance or entirely absent) but on small family farms they have three limitations.
First, since the data will be based on memory recall, they will contain inaccuracies. These increase in rough proportion to the period of time which has elapsed since the occurrence of events and the degree of complexity or number of activities in the system. Second, successful interviews require a considerable degree of technical and economic understanding, interpersonal skills and sympathy on the part of the interviewer - a combination of attributes often difficult to find within a single individual. Third and most serious, it takes time and patience to gain the confidence of the household regarding its farm, and even more time, observation and patience to gain an insight into those household factors which dominate the farm. One can forget about Oxford PhDs: the best interviewers are usually women with a village background who speak the local language.
Records relating to some aspects of the farm component might be maintained by a farmer club, cooperative or association. But these will often relate only to the most important farm activity (a club keeping records of inputs supplied to and produce received from its member maize growers; similar records kept by an export company on behalf of sugar or banana growers). These will in any case relate only to the farm and not to the household component.
Best of all are data obtained from a continuing small-scale but intensive survey (usually by a government extension or farm management service) of both the farm and household components over time. The Pakistan and Bangladesh panel surveys of the 1960s still stand as models. They attached one enumerator to each cluster of 20 to 25 farm households, with one cluster in each of the main farming regions. Each household was visited at least once every 10 days when farming and relevant household events were recorded (for often illiterate families) more or less as they occurred. Equally important, these panel surveys covered household processing of farm outputs and their marketing or other final disposal as well as non-farm economic activities and household consumption and expenditure. They were successful largely because the enumerators were given time - about a year - to first win the trust of initially suspicious villagers. As ever, however, the cost of such data collection must be assessed against its benefits. These benefits can arise not only through comparative analysis but also through use of the data in policy analysis and formulation.
As outlined by Dillon and Hardaker (1993, Ch. 2), Collinson (1983, Part II) and Upton (1973, Chs 10 and 11; 1987, Ch. 11), standards against which the performance of a subject farm can be measured might be obtained from such sources as:
· an actual 'high-performing' farm.· an actual farm subjectively or objectively chosen as being representative of some relevant group of farms.
· the mean or median performance data of a group of similar actual farms.
· the performance of a model or demonstration farm which has been established (by government) for this purpose.
· a synthetic 'farm' which has been fabricated by the analyst from estimates by research workers regarding how an actual farm should function if the best mix of activities were to be operated and - within practical limits - the best production technologies employed.
· census-type surveys.
Data for comparisons from a 'high-performing' farm require little discussion, except perhaps to note the problem of selecting a farm which is good but not too good. The same problem exists in selecting a representative farm.
The use of model actual farms has declined. They were fairly widely used in the 1950s and 1960s, primarily for testing research results, as an extension vehicle and as a captive data source. They are generally not effective as an extension vehicle because farmers see little relationship between what governments can achieve with unlimited resources on a model farm and what resource-poor farmers can hope for at village level. The same weakness exists in their data-generating role: data generated within a 9-to-5 civil-service structure is often not applicable to real farms. In any case, model-farm data cannot be generated for the household.
Use of synthetic data offers an alternative. These can be selected to refer to both the farm and household - which latter will require information from other than agricultural departments. Concerning agricultural performance standards, the problem of defining appropriate realistic standards and recommendations again arises. It is more than a technical one. Farm development strategy is not infrequently set and standards adopted, perhaps unconsciously, which serve the special interests of the local bureaucracy or research industry. They might also reflect the ambitions of purveyors of (elsewhere unwanted) agricides and of agronomic miracles, the passions of mechanization experts or the innocence of sectoral economists. In most of South Asia the farm performance standards which these things imply are only dreams; they are not aspirations to which the people may reasonably aspire.
Census-type surveys as a means of obtaining farm data are unfortunately common. In some small countries they can run to the attempted coverage of up to 2 500 farms. Apart from excessive cost, they often lack clarity in purpose and precision in design. They might be based on once-only interviews (which fail for reasons discussed above), or purport to be continuing surveys in that the effort is planned to be repeated over several years. They are the most expensive but least effective of all methods of obtaining detailed data for farm and household systems analysis. So voluminous is their output that seldom is more than 10 per cent of it ever used - but, given their typical lack of reliability, this must usually be counted a good thing!
Sources of data for the evaluation and adjustment of agro-technical processes by response analysis are discussed in Chapter 8.
Three difficulties may arise in the evaluation and comparative analysis of a farm system. First, relevant and reliable data need to be obtained. This difficulty was discussed in Section 7.4 above. It may be compounded, however, by the second difficulty. This is the degree to which a farm is subsistence rather than market oriented. To the extent that a farm system is subsistence oriented (as may be expected particularly of Type 1 farms but also for farms of Types 2 and 3), evaluation may have to be conducted mainly in terms of physical inputs and outputs, supplemented where possible by financial analysis based on the use of opportunity or market costs and prices (Dillon and Hardaker 1993, pp. 80-82; Makeham and Malcolm 1986, p.65). Such financial analysis will obviously be less reliable the greater the subsistence orientation of the farm system and the greater its isolation from the market. In the extreme, for a purely subsistence system, the focus of analysis may need to be in terms of family labour effort and staple food production in relation to household nutrition.
The third difficulty relates to a fallacy in comparative analysis of resource productivities. It is a difficulty of which the analyst must be aware and aim to avoid.
The potential fallacy in comparative analysis
Comparative analysis has proved itself in the field as a very useful guide to the improvement of farm systems (Upton 1973, pp. 234-243). It must be used with caution, however, as a guide to the mix of resources that might best be used. This is because, as shown by Candler and Sargent (1962), deductions about improved resource use made on the basis of comparative analysis may be quite erroneous. The reason for this is illustrated in Figure 7.4.
Suppose that for two variable inputs X1 and X2 used in producing a product Y, the locus of all combinations of X1 and X2 which will produce the fixed level of output Y* is shown by the ellipse of Figure 7.4. This locus is known as the isoquant for Y = Y*. In terms of the symbolism introduced in Section 5.1.1, the isoquant is represented algebraically by the function X1 = h(X2/Y*) and is derived from the production function Y = f(X1, X2) by fixing Y at the level Y*.
From Figure 7.4 it can be seen that use of X1 or X2 at levels beyond OC or OF, respectively, is physically inefficient; Y* units of output can always be produced with a lesser quantity of X1 or X2 lying on the isoquant segment AGB. Just where on this segment will be the economically most efficient (i.e., profit-maximizing) combination of X1 and X2 to produce Y* will depend on their relative prices. It will be at the point where the (clearly negative) slope of the segment AB (given by the ratio D X1/D X2 where D Xi denotes a small change in Xi) is equal to the negative of the inputs' inverse price ratio p2/p1, i.e., where
D X1/D X2 = -p2/p1
so that
-p1(D X1) = p2(D X2)
which implies that X1 and X2 are in economic balance in the sense that, at this point, the positive saving in cost of -p1(D X1) from a small decrease (i.e., D X1 <0) in the use of X1 is exactly offset by the increase in cost of p2(D X2) from the small increase (i.e., D X2 >0) in the use of X2 needed to maintain output at the fixed level Y'. This profit-maximizing combination of X1 and X2 could be, say, at the point G on AGB.
Now suppose that in comparative analysis the subject farm is at point A and the comparative farm at point B of Figure 7.4 in terms of their use of X1 and X2. At A, X1 has relatively low productivity of Y*/OC and X2 has relatively high productivity of Y*/OE while at B these productivities are reversed. Based on comparative analysis of the productivity of X1, an analyst might suggest that the subject farm should reduce its use of X1 from OC to OD and increase its use of X2 from OE to OF so as to have a resource allocation pattern like the comparative farm. This, however, would be a mistake. For profit maximization, both the subject farm and the comparative farm should operate at point G, not at A or B respectively. Thus the productivity of individual resources considered one at a time may lead to fallacious conclusions from comparative analysis.
FIGURE 7.4 - Isoquant showing Combinations of Two Variable Inputs X1 and X2 to Produce a Fixed Level of Output Y*
Candler, W. and D. Sargent (1962). 'Farm Standards and the Theory of Production Economies', Journal of Agricultural Economics 15(2): 283-290.
Collinson, M. (1983). Farm Management in Peasant Agriculture, Westview Press, Boulder.
Dillon, J.L. and J.B. Hardaker (1993). Farm Management Research for Small Farmer Development, FAO Farm Systems Management Series No. 6, Food and Agriculture Organization of the United Nations, Rome.
Makeham, J.P. and L.R. Malcolm (1986). The Economics of Tropical Farm Management, Cambridge University Press.
Upton, M. (1973). Farm Management in Africa, Oxford University Press, London.
Upton, M. (1987). African Farm Management, Cambridge University Press.
