Abdu Muwonge

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Dr. Sally Wallace - Chair

Second Advisor

Dr. James R. Alm

Third Advisor

Dr. Jorge L. Martinez-Vazquez

Fourth Advisor

Dr. Ragan Petrie

Fifth Advisor

Dr. Yongsheng Xu

Sixth Advisor

Dr. Stephen D. Younger


Decentralization is a key governance reform which many developing countries have embarked on. Local governments are expected to use their informational advantage to improve the delivery of public goods. This result implied by Tiebout’s (1956) model requires fully informed citizens who “vote with their feet.” The model’s application to developing countries has been limited, since local decisions may not be responsive to local demands. Practitioners are shifting to innovations that minimize institutional constraints so that decentralized programs can lead to improved outcomes. Examples of such innovative ways include decentralized agricultural extension programs, which embrace farmers’ empowerment, local government, and private sector participation. Few impact evaluation studies on agricultural extension have combined qualitative and quantitative methods. This dissertation contributes to the literature by applying these methods and survey data to study the impact of a decentralized extension program in Uganda, known as the National Agricultural Advisory Services (NAADS) on the value of farm production per acre. The program is non-randomly assigned to local governments and farmers self-select in or out within participating sub-counties. Using a sample of 305 participating and non-participating farmers and local government assessment indicators as instruments, we cannot reject the null that the NAADS program has had an impact. The 2SLS results show no program impact; however, the OLS results show that the program had a positive impact on the value of farm production per acre of about 20 percent. Qualitative results show that NAADS farmers: participate in local decision making processes through farmers’ institutions; have increased knowledge on farming; and practice enterprise diversification. The quantitative finding must be treated with caution; for example, the study did not account for spillover effects. The NAADS program faces challenges inherent in Uganda’s decentralized structure; particularly the low financial and human capacity, and the weak monitoring at the local level. The policy implications include: the need to strengthen farmers’ institutions; development of a marketing strategy; clear policy guidelines for local government support to NAADS; improved coordination of NAADS activities among line ministries; need for additional resources for NAADS activities; and improved capacity of service providers.


Included in

Economics Commons