A production function analysis of commercial dairy farms in the Highlands of Eritrea using ridge regression

This study presents a production function analysis of fresh milk production in the Highlands of Eritrea, where most dairy farmers in Eritrea are located. To ensure representative production functions, this region was divided into three relatively homogenous study areas, namely Central Zone, Mendefera and Dekemhare. Most data for the study were collected in a survey of 120 respondents using a structured questionnaire. To obviate the problem of multicollinearity among explainatory variables, ridge regression was used to estimate milk production functions for each study area. Production elasticities of variable inputs, marginal products (MPx), values of marginal products (VMPx), marginal rates of input substitution (MRS) and least-cost combinations of purchased concentrates and forage were estimated for the three regions. The VMPs of all inputs for Central Zone dairy farmer respondents were estimated to be greater than their input prices, implying that the resources were under-utilized from a profit-maximising perspective (i.e. where VMPx = Px). However, respondents in Mendefera and Dekemhare used concentrates in excess of optimum levels (i.e. VMPx<Px). Analysis of the least-cost combination of purchased concentrates and forage suggests that dairy farmer respondents were also not allocating these resources on a minimum-cost basis. However, the profit maximizing and least-cost criteria assume perfect knowledge, a risk-free environment and competitive markets. Improved information, farmer training and better infrastructure (roads and telecommunications) to promote competitive markets could help to enhance resource allocation decisions by dairy producers.

Issue Date:
Publication Type:
Journal Article
PURL Identifier:
Published in:
Agrekon, Volume 45, Issue 2
Page range:
Total Pages:

 Record created 2017-04-01, last modified 2017-11-13

Download fulltext

Rate this document:

Rate this document:
(Not yet reviewed)