Policy options for improving the value of land use in smallholder Fijian agriculture

Administration
The project was formally approved on 1 August 2005. The initial administrative tasks involved formalising the research team and establishing lines of accountability and financial control between Deakin and project staff within the Ministry of Agriculture in Suva (MA) and the Secretariat of the Pacific Community.
We have established:
A good working team with members from Deakin, MA the Fiji Islands Bureau of Statistics and the Secretariat of the Pacific Community.
Appointed a project officer within the MA.
Financial systems for ensuring the smooth flow of funds from Deakin to our counterparts in Fiji and Noumea and acquittal mechanisms to ensure accountability.
A project planning seminar was held in Suva to bring together interested parties from the groups formally associated with the project plus staff from the University of the South Pacific, Ausaid and The Ministry of Planning and importantly the Fiji Islands Bureau of Statistics.

Research
Sound progress has been made by the project team. The initial tasks related to Objective 1 and involved:
refining the research plan
developing a detailed primary data collection plan
undertaking a comprehensive overview of the existing agricultural systems in Fiji
reviewing the policy environment in Fiji
reviewing data reliability and
familiarisation of the Fiji team with different policy evaluation approaches.

The planning of the primary data collection has involved the development of a set of questionnaires for use in the surveys. In doing so we trialled alternative ways of eliciting estimates of production and consumption from smallholders. The end result has been the development of 2 comprehensive surveys that address issues not adequately tested in the past. The production survey focuses on measuring producer responsiveness to output price changes as well as measuring input and output information. The consumer survey will provide estimates of household responsiveness to food price changes. These surveys are to be considered by the Deakin Ethics Committee as the next stage in this process.

The study of the exiting data sources involved extensive consultation with staff from The Bureau of Statistics in Suva and the MA. As a result of these consultations a system has been established for leveraging our survey work on that already undertaken by the MA and the Bureau of Stats. This review procedure highlighted issues associated with the existing data collection processes in Fiji. For example, the estimates of the value of Agricultural GDP have been critically analysed. As part of this analysis apparent inconsistencies with the current procedure were identified and a set of alternative estimates were derived. These estimates were published in a conference paper that is attached. A version of this paper will be published later this year in the collection of papers from the Conference on Governance and Accountability in the Pacific.

As part of the review of the agricultural systems and existing data sources a paper was prepared on the structure and vulnerability of smallholder agricultural systems to shocks. This paper was presented at the 2006 Annual Conference of the Australian Agricultural and Resource Economics Society. (See attached.) A related paper has been submitted to the Journal of Pacific Studies. This paper highlighted the dual nature of the smallholder agricultural system in Fiji. The Indo-Fijian and Fijian smallholders operate very different agricultural systems and have very different interactions with the commercial sector. While the members of both groups are heavily involved with the commercial sector, the nature of this involvement means that the impact of market shocks such as the sugar price regime change will have very different impacts on these 2 groups. This is a finding of importance in the formulation of adjustment policy.

Trip reports covering the first 12 months are attached.

Data Availability
A major success has been obtaining the agreement of the Fiji Islands Bureau of Statistics (FIBoS) to the use of data from their Household Income and Expenditure Survey, 2002 - 2003 (HIES) in the project. The HIES data will provide the sampling frame for both the urban and rural surveys so the project will have the benefit of a recent and full enumeration of urban and rural households to work from. In addition FIBoS have - very generously - agreed to the "importation" of their unit record data into the unit record sets for the project surveys. This means that our consumer can now include a food expenditure diary (more accurate than recall) and once the survey is complete we should be able to construct a cross section/time series data set that should provide a back-up to the principal ie survey means of deriving the consumer demand elasticities required for the project.

During the period covered by this report the field work for the project has been completed. This involved 2 major surveys and some informal pilot testing of the survey approach and instruments. The first survey targeted consumers in urban centres. This survey involved face to face interviews and the completion of food expenditure and consumption diaries. The urban survey sought information on responses to food price changes as well as information food consumption and expenditure patterns The survey covered some 600 households in all major urban centres in Fiji. The survey was based on a stratified random sample. The stratification was carried out on location, income and ethnic background.

The second survey covered food producers in rural areas. Approximately 900 rural households were interviewed to determine what they were producing and what production was being undertaken. As for the urban survey, a key part of the rural survey was the collection of stated preference data on the likely household production responses to food price changes. The survey was stratified on the basis of, location, farm characteristics and ethnicity.

Reports from the field staff suggest that the surveys were undertaken without major problems and the survey instruments were appropriate. Some difficulty was encountered in finding some of the identified sample households as there has been significant mobility in the population in Fiji in recent years.

The data from both surveys have been recorded in a data base and the records checked and cleaned. Data analysis to derive estimates of food demand elasticities and food supply elasticities has commenced.

Two conference papers were presented at the 2007 annual conference of the Australian Agricultural and Resource Economics Society in Queenstown. The first paper presents preliminary estimates of food demand elasticities from the consumer or urban survey. This paper uses data drawn from responses from the Suva region only. The results are encouraging in that they show the elasticities derived from our stated preference approach appear to be consistent with theory and local market knowledge. In the second paper we analysed the efficiency of food markets in Fiji. This paper raised issues concerning the lags in food price adjustments between local markets in Fiji. While we found prices do move together, the extent of the adjustment lags - sometimes 2 months - suggests market price information services could be improved.

The market price efficiency paper has been submitted to the Australian Journal of Agricultural and Resource Economics. An earlier paper concerning the agricultural policy responses to shocks to household incomes in rural areas has been accepted by the Journal of Pacific Studies. In the latter paper we report a case study of the income position of rural households in the Northern Region of Fiji. The potential sensitivity of these households to likely market changes is reviewed and the likely efficacy of possible government policy responses is analysed. Particular attention is given to the history of government intervention in agriculture in the region to encourage pineapple production and the reasons for its limited success.