National accounting and national accounts concepts have very strong links to business accounting and business accounts concepts, and economic theory. The current trend in national accounting work is to carry out the task based on a framework which functions as an integrated general purpose statistical information system. The principal constraints to this approach are data availability which is a function of the standard and serviceability of the country's national statistical system, and general conventional accounting principles. It is submitted that one of the important properties that a system of national accounts should possess is inherent flexibility in order to allow such a system to respond sensitively to developments in economic theory and specialized needs/demands. Even though the development of certain national accounting principles can be traced back to the seventeenth century, national accounting generally is a twentieth century invention, with the development of the SNA as a very major breakthrough. One of the SNA's most valuable contributions is its provision of a consistent and comprehensive framework in terms of national accounting concepts, definitions and principles. The application of the conventional national accounting principles to Tonga and other developing countries has been confronted with many conceptual and statistical problems. The existence of imperfect national price systems in developing countries has led many critics to question the validity of national income estimates for such countries. It has been submitted that such imperfection of national price systems has been essentially a function of constraints on commodity and factor movements. It is argued however that many such restrictions cannot be applied with significant force to the case of Tonga and that there exists great validity and usefulness in making national income estimation. Non-market portion of total production in the economy of Tonga is significant. Its estimation for national accounting purposes is faced with the conceptual problems of setting appropriate boundary of production and imputation. The recommendations of the SNA with regard to the former are endorsed for national accounting practice in Tonga with special attention to transactions such as the unilateral provisions of free and/or partially paid services by Tongans to government and church authorities. Due to the predominantly agricultural character of the Tongan economy and the existence of a partial subsistence sector, certain conceptual problems arising from the need to make clear distinctions between production and consumption, and saving and investment in accordance with the Keynesian framework are bound to prevail with regard to national accounting in the Kingdom. There are also the problems associated with any attempt to distinguish the various factor shares of national income. Recommendations and comments concerning possible means of minimizing these problems or perhaps putting such, problems into some balanced perspective, are presented. Statistical problems faced with national accounting work in Tonga are generally associated with the inadequacy and unreliability of basic national accounting statistics, and the shortage of appropriate material and human resources. While most of the statistical problems can be substantially alleviated by more liberal direction of resources to the process of data collection and national income estimation, some of the statistical problems like the conceptual problems would persist in Tonga as long as the general standard of the Kingdom's national statistical system is low and the general state of under-development prevails. A close scrutiny of the general performance and structure of the Tongan economy, current and anticipated future needs/demands for various national accounting statistical series, and current and anticipated future status of data availability, suggests that a system of national accounts for Tonga which could bring out the important economic characteristics and inter-relationships of important sectors in the economy would be considered as very desirable. With the increasing complexity of the Kingdom's economy and associated increasing diversity of needs/demands for appropriate statistical series, it is felt that a system of national accounts which offers greater disaggregation of important aggregates and which views the production sector at some component levels such as "Agricultural", "Non-Agricultural, Non-Financial" and "Financial" production sub-sectors even though it is a diversion from the general recommendations of the SNA, would merit a high level of serviceability. A system of national accounts which is essentially based on a system of sector accounts with the incorporation of the above stated partitioning of the production sector is therefore proposed for Tonga. The proposed system also consists of sets of supplementary tables and aggregative accounts which aim at providing essential supplementary information via the adoption of different classification systems or appropriate aggregations of information exhibited in the sector accounts. Issues concerning suggested future expansion of the proposed system to include appropriate input-output matrices and flow of funds tables and the need for a proper development and application of an appropriate social accounting matrix (SAM) for Tonga are also discussed. The proposed system of national accounts for Tonga is relatively comprehensive and possesses inherent flexibility which is a reflection of its primary inheritance from the SNA. Even though it might take some time before the proposed design can be implemented in full, its relative comprehensiveness would automatically serve a lot of useful purposes e.g. its explicit facilitation of attention to gaps in the existing data series. The implementation of the proposed system is regarded as a feasible long term project for the Tonga national statistical agency. The extent to which it can be implemented in the very near future given Government's acceptance, would depend heavily on the priority level the Tonga Government attaches or would attach to the task of developing and improving the Kingdom's national accounting system.

Past development efforts in the Kingdom of Tonga are reviewed in relation to the goals of reducing poverty, unemployment and inequality. It is shown that progress has been strictly limited and evidence is provided to support the view that inadequate attention has been paid to agricultural improvement and rural development. The components, practices and institutional framework of the Tongan agricultural system are reviewed and possible future avenues of agricultural development are examined. Attention is drawn to those aspects where improvements are needed if agricultural development is to be promoted. These include an expanded programme of research, an upgrading of the agricultura1 extension effort, an overhaul of the 1and tenure system and a substantial improvement in agricultural product marketing; or, in short, an integrated programme of agricultural improvement and rural development. From a planning viewpoint, there is evidence of inadequate elaboration of development objectives, superficial identification of resources and insufficient knowledge concerning resource interactions, all of which have created confusion and uncertainty, particularly as regards the role of agriculture in the Kingdom's development. As a result, there is a lack of commitment to development efforts generally, and a tendency for activities on the fringes to be pursued at the expense of more fundamental issues. It is concluded that this situation results largely from the lack of an adequate information base, and that before an integrated plan for agricultural improvement can be developed, the network of interrelationships operating within the agricultural system needs to be understood. An information system is developed with the basic aim of providing information which will facilitate an understanding of how the agricultural system is integrated into the everyday lives of the people of Tonga. The concept adopted recognises the reciprocal relationships involved between agriculture and the rest of the economy. Using a systems research framework, the Tongan household agricultural system is defined in terms of its component parts, their interactive behaviour and their interrelationships. The key variables required for analysis are isolated and the production relationships in which rural households participate are specified. The information system provides for the observation of household decision making with particular reference to the allocation of labour, land use and the disposal of farm produce, and the disposition of cash earnings. The information framework is based on the premise that land is the most critical resource in Tongan agriculture. Information requirements are specified which will enable an assessment to be made of the effect of a number of factors on 1and productivity. The data collected will also enable valuable insights to be gained into issues such as land utilisation and methods of production; composition and disposition of agricultural output; household food consumption; marketing of agricultural produce; land tenure; sources of household income; net incomes from agriculture; participation in aid projects; and utilisation of loan finance.