The aim of this paper is to select dependent variables and to specify the functional form of hedonic rent equation for evaluating the externality of paddy fields.
The first problem has been takled by the bootstrap P test, which is a kind of non-nested test to select appropriate models among non-nested alternatives. To treat the second problem, we used single index model, which is a semiparametric modeling strategy. It has an advantage to avoid curse of dimensionality. The single index model is estimated by a projection pursUit regression.
The practical procedures are as follows: First, we have used or set different varieties of rents. Second, we have estimated three varieties of hedonic rent equation by the projection pursuit regression. Third, the bootstrap P test was used to select appropriate models. Finally, we have evaluated extemality of paddy fields by the estimated model. The result shows that single-index model is superior than the previous log-log modeL A bootstrap 90% confidence interval of the money metric externality was [3000 bi1lon/7000 billion] (yen) on standardized rent and [2600 billion/5900 bi11ion] (yen) on Jutaku Kinyu Kohko, which isless than the previous results given by Tanaka (1998).

View full abstract

In recent years, watershed desolation is going on in many river basins of many countries. The causes of it are different in each country. In some countries under developing, heavy shift cultivation, irrational reclamabonof farmland, and unscrupulous cuts of firewood etc. are causing this kind of desolation. Whi1e in other countries, abandonment of forest management caused by the depopulation of rural communities is the major cause of desolation.
These kinds of desolation are producing problems of new kinds. Unexpectedly heavy floods are suffering people after the rainfalls of ordinary magnitude in some watersheds. The water sources are suddenly exhausting in some other watersheds.
To fight against these difficulties. periodical monitoring of the environmental conservation capability of each watershed seems to be necessary. For example, the capability of flood prevention should be eva1uated once in every 5 years or so to watch out and forecast the deterioradon of the capability.
And some countermeasures should be prepared if necessary to prevent fiood disasters in future.
The authors are going to propose the ways necessary to evaluate these kinds of capabi1ity of watershed like flood prevention, water source fostering, soil erosion prevention etc. Several reports are now being prepared for this purpose.
In this report. the authors proposed one practical method to evaluate the f1ood prevention capability of watershed. ‘Forest-Plan Map’and‘Forest Data-book’were utilized as the materials for evaluation. And in this evaluadon, the roughness coefficient N for kinematic wave model was estimated to represent the capability of flood prevention. An example of evaluation in a watershed of Tsubo-river basin of Gifu-Prefecture was also shown in this report.

View full abstract