Measuring the productivity of dairy buffalo farms helps in determining which are performing less efficiently than the others in the same geographical location. This is important to develop benchmarking strategies to improve the performance of inefficient dairy buffalo farms.

Along this premise, a research study by Eric Palacpac (Chief of DA-PCC’s Knowledge Management Division or KMD) and Erwin Valiente (former KMD researcher and current Instructor at the Central Luzon State University) employed a nonparametric linear programing method called Data Envelopment Analysis (DEA) in measuring the efficiencies of 75 dairy buffalo farms in Nueva Ecija.

Given that these farms run at different scales of operation, the researchers employed a variable return to scale-input oriented DEA (VRS-IODEA) model. The VRS helps to have an autonomous data analysis of the variables as the increase in input would not reflect a proportional increase in output. Also, in developing countries like the Philippines, the input-oriented DEA model helps to save agricultural resources and services, as it intends to distinguish efficient guidelines and practices in using entire assets such as farm equipment, machinery, labor, etc.

With this VRS-IODEA model, only farms, referred to as decisionmaking units or DMUs, of similar characteristics (i.e., homogeneous) in terms of inputs and output were examined. The inputs are the available resources and services provided to the farms, such as biologics, feeds, forage, and labor. On the other hand, the output was the milk produced by the buffaloes.

Measures of Efficiency

In applying the DEA, three measurements of efficiency were considered in estimating the performances of dairy buffalo farms, namely technical efficiency (TE), allocative efficiency (AE), and economic efficiency (EE). The TE refers to the ability of the DMU to produce the largest possible quantity of output from a given level of inputs (output-oriented) or produce a given level of output with the smallest possible level of inputs (input-oriented). The AE measures the ability of a technically efficient DMU to use inputs and produce outputs in optimal proportions given their respective prices. Economic efficiency (EE) denotes the entire computation of the performances and is calculated as EE = TE x AE.

The analysis were applied to three classifications of DMUs: smallholders or those with 3-5 buffaloes (n=58), family module or those with 6-10 buffaloes (n=12), and semi-commercial or those with 11-20 buffaloes (n=5). Likewise, only farmers with at least three buffalo cows and have at least five years of engaging in dairy buffalo production were considered.

Efficiency Scores

In utilizing the VRS-IODEA model, the level of efficiencies of 75 dairy buffalo farms was measured. Overall, the mean TE and AE were 0.80 and 0.81, respectively. These results signify that if the farms were to be technically efficient in their operations, it could still increase its output by 20%. Also, if the farms were to allocate the costs of its inputs more efficiently, it could still increase its output by 19%. Finally, with a mean EE of 0.65, the farms can still produce 35% more milk if it were to be both technically and allocatively efficient.

Inputs Adjustments

Inefficiencies were more commonly observed among smallholder farms. Smallholder DMUs with low efficiency (scores of 0.3782-0.5854) produced an average of 3,990 li of milk in one production cycle. They need to reduce their inputs (e.g., biologics, feeds, forages, labor) by 53.31% without affecting the amount of milk produced. Those that are moderately efficient (scores of 0.5855-0.7927) have to reduce their inputs by 40% to attain efficiency. Meanwhile, family module farms and semi-commercial farms were already considered highly efficient (scores of 0.7928-0.9999) and only need to reduce their inputs by 14% and 24%, respectively, to become fully efficient (i.e., score of 1.0000).

Benchmarking

Using DEA also made possible the determination of best practice frontiers (fully efficient DMU peers) that inefficient DMUs can emulate. The DEA software generated lambda values (raw weights) assigned to the peer DMUs in a table format. Those DMUs listed horizontally at the top of the table are the fully efficient ones while those listed vertically in the first column of the table are the inefficient DMUs. Those DMUs that shared lambda values of more than 0.00 are considered peers. Inefficient DMUs can then benchmark with the efficient DMUs if they share the highest lambda values. In other words, the inefficient DMUs can evaluate their practices (i.e. how they allocate resources or inputs, what technologies do they apply, etc.) by comparing with those of the fully efficient farms, which could serve as their standards. Such benchmarking is seen to improve the production efficiency of the inefficient DMUs.

Ways Forward

The study demonstrated that DEA can serve as a valuable analytical tool in determining efficiency scores of various categories of dairy buffalo farms, in recommending adjustments in inputs of inefficient farms without affecting output, and in benchmarking of inefficient farms with the best practice peer frontiers (fully efficient farms). The researchers recommended providing assistance to the smallholders in particular to increase their animal holding thereby allowing the utilization of more appropriate technological inputs for increased productivity. Likewise, they recommended future studies that can analyze the various socio-economic factors of the DMUs vis-à-vis their levels of efficiency.