7812 - Laserfiche WebLink

Home Browse Search

(1985) have noted, one cannot a priori de- termine the direction of bias in consumer surplus estimates from omitting a variable for substitutes. In addition, if other factors influencing fishing demand on the North Fork of the Feather River were changing over the period studied, they should be reflected by a specific independent vari- able. We are not aware of any significant changes in factors affecting fishing de- mand other than those included in equa- tion (3); therefore, no additional variables have been included. The equation for CREEL is: CREEL, = Bo • FLOW,"' (TRIPS;, / POPJB2 + v,,, (4) where FLOW, is the average discharge downstream of Rock Creek Dam, in cfs, from May to August for the years t = 1981-1985. A positive correlation is expected be- tween the level of river flow and the level of creel. In some respects, equation (4) is a simple production function which quan- tifies the productivity of water in produc- ing harvestable fish. While it may be de- sirable to focus on weekly or monthly flow rather than seasonal average over the five years, we feel the fishery population dur- ing a given season is more influenced by these seasonal flows rather than weekly flows as long as critical flow and temper- ature thresholds are not exceeded. Since CREEL, is a measure of total creel in time t, it is expected that CREEL, is an increasing function of TRIPS;,/POP;,. Hence, there is the possibility of simul- taneity between equations (3) and (4). Es- timated jointly, equations (3) and (4) form a simple, yet powerful, bioeconomic sys- tem. The nonlinear format in equation (4) provided a better fit of the data than a sim- ple linear model, which performed quite poorly. STATISTICAL RESULTS The regression results are presented in Table 1. The results were obtained through the TSP's Version 5.1's nonlinear least squares regression program. At each iter- ation, this quasi-Newton algorithm com- putes the approximate derivatives with re- spect to each of the coefficients. The dependent variable is then regressed on these derivatives. The disturbance term is assumed to be distributed normally. The Two Stage Least Squares (TSLS) es- timation procedure is used to estimate equations (3) and (4) as a system. The TSLS regression results for equation (3) are pre- sented in Table 1. Since the regression es- timates for equation (4) are mainly of in- terest for estimating CREEL, for the TCM demand equation (3), a TSLS regression is not performed for equation (4). However, for informational purposes, Table 1 also presents the nonlinear least squares results for regression (4). Both regressions are TABLE 1 Pooled time-series cross-section regressions for river section 3 of the North Fork of the Feather River. (1) Nonlinear Two Stage Least Squares regression for TRIPS/POP" INTERCEPT TRVCOST INCOME CREEL Adj. rz Log likelihood 0.001 -2.772 0.223 1.110 0.76 2,257 (0.12)6 (-19.58) (0.38) (2.52) (2) Nonlinear Least Squares results for CREEL INTERCEPT FLOW TRIP/POP Adj. r2 Log likelihood 2,282.291 0.067 0.030 0.232 -21,600 (12.35) (6.34) (7.06) ° The number of observations is 285, or 5 years x 57 counties. The t-statistics are in parentheses. J. Loomis and J. Cooper 27 ??