8. CLASSIFICATION OF INDONESIAN DAIRY CO-OPERATIVES - Paul Riethmuller and Dominic Smith Department of Economics, The University of Queensland, Brisbane, Australia

Paul Riethmuller and Dominic Smith
Department of Economics, The University of Queensland, Brisbane, Australia

This chapter analyses the current system of co-operative classification and highlights its advantages and disadvantages. Two alternative productivity based classification systems are proposed. While these alternative classification systems are not necessarily better than the existing system, they are presented to contribute to the discussion as to how best the government should classify Indonesian co-operatives.

CURRENT CLASSIFICATION SYSTEM

The current system of classification of Indonesian dairy co-operatives used by the Ministry of Co-operatives, the government ministry responsible for overseeing the operations of Indonesia’s co-operatives, is based on the daily milk production of the co-operative, averaged over a three year period. Table 8.1 shows the classification levels currently in use.

Table 8.1. Existing classification criteria

Milk Production	Classification
(litres/day)
more than 20000	Strong
10000-20000	Potential
5000-10000	Weak
less than 5000	Very weak
0	Non-active

Source: GKSI (1995)

Figure 8.1. Status of Indonesian dairy co-operatives, 1994

As this classification system is based purely on milk production per day, it is heavily biased towards large scale co-operatives with many farmers and cattle. Although Figure 8.1 presents what seems to be quite a dismal picture of the state of Indonesian dairy co-operatives, it may well merely be a reflection of the size distribution of co-operatives. The vast majority of Indonesian dairy co-operatives are small scale operations, and although they might be technically efficient operations, they could be classified as weak or very weak using the criteria of the Ministry of Co-operatives. Of course the co-operatives classified as non-active (almost 50 per cent of the total) would be non-active under any classification system. The following two sections present alternative classification systems, based on productivity measures, rather than on gross output levels. Non-active co-operatives are not included in these two alternative classifications.

PRODUCTION PER FARM AS A CLASSIFICATION CRITERIA

Unlike the previous classification system, these criteria are based on the average milk production per farm within each co-operative. This classification system can assist policy makers identify which co-operatives have a high productivity per farm and help them identify the reasons for different levels of performance. The classification criteria are shown in Table 8.2.

Table 8.2. Farm productivity based criteria

Criterion	Classification
(litres/farm/day)
under 0-5	A
5-10	B
10-20	C
over 20	D

Figure 8.2 shows the number of co-operatives classified according to the criteria of Table 8.2. It is immediately obvious that the distribution of co-operatives when classified according to production per farm is far more even than the distribution gained by aggregate production from the co-operative. Figure 8.2 does not show, however, whether the productivity and output based criteria give similar rankings to the co-operatives.

Figure 8.2. Co-operatives classified according to production per farm, 1994

Figure 8.3 shows the proportion of co-operatives in each farm productivity criteria ranked by their original (output based) rankings.

Figure 8.3. Comparison of farm productivity and output based rankings

The following points may be made from the information presented in Figure 8.3.

• Highly productive co-operatives (Category D) - those where the average farm produces more than 20 litres per day - do not necessarily have a high level of output.
• Low output co-operatives - these are the co-operatives classified as very weak on the basis of total production - could have members able to produce a very high output per farm.
• Co-operatives where the average farmer is able to produce 10 or more litres of milk per day (categories C and D) are less likely to be classed as very weak or weak on the basis of daily output as compared to those co-operatives where the average farmer produces less than 10 litres per day.

A problem with using this approach to rank co-operatives is that it is biased toward co-operatives that consist mainly of large farms. Thus a high per farm output may be achieved simply by having a large number of animals per farm.

PRODUCTION PER ANIMAL AS A CLASSIFICATION CRITERIA

To overcome the bias just referred to, co-operatives can be ranked according to the output per animal per day. This gives a measure of the productivity of each animal, but does not take into account farm or co-operative size. Nor does it take into account the use of other inputs besides the cows, nor the quality of milk delivered. The distributions of co-operatives according to this criteria are shown in Table 8.3 and Figure 8.4.

Table 8.3. Criteria for ranking dairy co-operatives based on production per cow per day

Output	Ranking
(litres/cow/day)
0-1	A
1-2	B
2-3	C
3-4	D
4-5	E
>5	F

Figure 8.4. Co-operatives classified according to productivity per animal, 1994

Once again, the distribution of co-operatives according to productivity per animal is more even than the distribution according to milk output level. Figure 8.5 shows the proportion of co-operatives in each animal productivity criteria based on their original (output based) rankings.

Figure 8.5. Comparison of animal productivity and output based rankings

It is apparent from Figure 8.5 that:

• although not as evenly distributed as the farm productivity rankings, gross output is not highly related to productivity per animal; and
• over half the co-operatives ranked as having a per animal output of greater than five litres per day were ranked as very weak in the output based criteria.

IS THERE A RELATIONSHIP BETWEEN THE RANKINGS?

To quantify the degree of agreement in the rankings provided by the three criteria so far described, Spearman rank correlation coefficients were calculated for the rankings of the co-operatives. The Spearman rank correlation coefficient (r_s) was calculated using the SPSS computer package. It enables testing of the hypothesis that the correlation between the rankings is significantly different from zero. A value of -1 for r_s means that there is perfect negative correlation between the rankings. A value of +1 implies perfect positive correlation while a value of 0 implies that there is no correlation. The Spearman’s rank correlation co-efficients calculated using the three ranking systems are presented in Table 8.4.

Table 8.4. Calculated Spearman Rank correlation coefficients, 1994 data

Rankings	Correlation	Significantly different from zero?
Per cow production and total production	0.25	Yes, at the 10 per cent level.
Per farm production and total production	0.18	No
Per farm production and per cow production	0.91	Yes, at the 5 per cent level.

The following points may be made from the results in Table 8.4.

• The weak positive relationship between the output based ranking and the per cow production ranking may be a result of the larger scale co-operatives being able to provide a higher level of extension advice and veterinary care than the smaller co-operatives. This would increase production per cow.
• There is no statistically significant relationship between the production ranking and per farm ranking. This indicates that the co-operatives that are classified as weak using the current criteria may have farm members able to produce relatively large volumes of milk.
• There is a very strong positive relationship between the per cow and per farm productivity ranking criteria. The strong relationship between per farm and per cow productivity ratings could be explained by the fact that most Indonesian dairy farms have at most three to five milking cattle. Furthermore, there is very little variation in this farm size (Chai, Riethmuller, Smith and Hutabarat, 1999). Thus it would be expected that farms with a relatively high level of production per cow would also have a relatively high level of production per farm.

CONCLUDING COMMENTS

Each of the approaches to ranking co-operatives detailed above has advantages and disadvantages. The technique being used in the late 1990s focuses on co-operative size as a measure of strength by ranking co-operatives on average daily total output. This is useful for measuring the relative sizes of co-operatives, but does not take productivity into account and is biased toward large scale co-operatives. The second technique suggested focuses on productivity per farm. Whilst being a useful tool for measuring the ability of the farm unit to produce milk, it does not take productivity of animals or size of the co-operative into account. Moreover, this measure is biased toward those co-operatives that consist of large farm units. The final criteria is based on productivity per animal. Whilst avoiding the biases implicit in the first two approaches, it does not take into account farm size (and associated economies of scale) or the size of the co-operative. All of these are important factors in the success or failure of a dairy farming enterprise. A general criticism of all of the measures is that they ignore financial considerations, in that no account is taken of the financial performance of the dairy operation, either at the farm level, or at the co-operative level.

A practical solution to the problems posed by the individual techniques outlined above may be to use all three approaches in tandem, to give an overall ranking based on co-operative size and infrastructure, farm productivity and economies of scale and productivity per animal.

The value of this approach is that it would enable the Indonesian government to target co-operatives that are found to be weak in particular areas. For example, the total output of a co-operative is an important consideration when decisions are being made about the provision of capital. It would probably make most sense to provide cooling equipment or milk pasteurizing equipment to the largest co-operatives. Hence in allocating funds for this purpose, a ranking of co-operatives based on the co-operatives output would be a valuable adjunct in decision making. If however, a program to improve the on-farm technical efficiency was being set up, a ranking of co-operatives on the basis of output per animal would enable the poorest performing co-operatives - and those likely to be in greatest need of assistance - to be identified.

A criticism that is sometimes made of policy makers is that they try to use a simple policy measure to achieve a range of objectives. International trade is one area where this is likely to happen. Protection of a sector or industry is given to preserve foreign exchange, to maintain employment in the protected sector, for national security reasons and so on. The general principle that should be followed is to use different instruments to achieve particular objectives. The same philosophy should be used in the Indonesian dairy industry to rate co-operatives. No simple rating measure is likely to be useful for all policy issues.

REFERENCES

Biro Pusat Statistik (1993) Expenditure for Consumption of Indonesia 1993, BPS, Jakarta.

Chai, J., Reithmuller, P., Smith, D. and Hutabarat, D., (1999) “Employment, income and efficiency in Indonesian farms: an exploratory analysis,” Scandinavian Journal of Development Alternatives and Area Studies, Volume 18, No. 2 & 3, pp. 203-228.

Erwidodo and Fadhil Hasan (1993) Indonesian Dairy Industry: Problems and Alternative Policy Options, Center for Policy and Implementation Studies, mimeo.

Gabungan Koperasi Susu Indonesia (1993) GKSI: Union of Dairy Co-operatives of Indonesia, GKSI, Jakarta.

Irawan, B. and Rusastra, I.W. (1990) “Economic efficiency and protection rates of the milk production in Central Java”, in F. Kasryno and P. Simatupang (ed), Comparative Advantage and Protection Structures of the Livestock and Feedstuff Subsectors in Indonesia, Center for Agro Economic Research, Bogor, pp: 133-47.

Smith, D. and Riethmuller, P. (1995) The Indonesian Dairy Industry, Dairy Policy in Indonesia Discussion Paper Number 1. Department of Economics, The University of Queensland.