The paper makes an attempt to glean certain salient aspects of the character of the household savings and investment process in developing country contexts. In part, also, this is concerned with identifying formulations of the determinants of household savings that are better suited to the behavioural-structural polity to be found in developing economies. In its empirical part, the paper goes on to test some models of household saving, including the Friedmanesque hypothesis relating to savings out of transitory income components. It is argued that, as would be expected, a Keynesian type relation describes the saving model of the household reasonably well, in both rural and urban areas of Bangladesh that incidentally, the use of the data from 1973/74 and 1976/77 Household Expenditure Surveys Yield quite aberrant results, and, finally, that savings out of transitory income components in rural areas of Bangladesh are consistently and significantly higher than in urban areas. 

The paper investigates the nature of price variation within the crop-year or season as opposed to price variation from year to year. The seasonal price index provides a measure of the month to month variation in jute prices. It was found that, in a typical year, the actual price peak is 36% higher than the seasonal price low. There appears to be a distinct price seasonal within the crop-year, with the low occurring in August and the high in January. The period of substantive price increase, however, takes place between August and November. To exploit the upward price gradient, three early season storage plans were examined. Though the August-November storage plan to stabilize intra-crop-year prices was found to yield the highest returns, the August-January storage plan was recommended due to its lower uncertainty.An attempt has also been made to identify theoretically and empirically, the determinants of the observed seasonal price pattern. Normally, in a crop that is harvested at one time (and consumed throughout the season) the seasonal price spread is caused by a cost of storage element and an adjustment factor due to errors in demand assessment. However, in case of jute, it was found that due to lack of adequate institutions, error in supply estimation also influences the seasonal price spread.

The paper presents a land allocation model for use in water resources planning in general, and in analysing water sector investment for agricultural development in particular. The model would serve as a tool for framework planning through screening and sequencing of alternative investment modes, and in identifying viable investment options in different regions. The paper illustrates the use of this model for analysing alternative strategies for long term development of water resources in Bangladesh. In particular, the implications of alternative investment strategies underlying different projections are analysed in terms of their contributions in achieving foodgrain demand-supply balance over the next twenty years.