sites. Travel costs were assumed to be a function 

 of both monetary expenditures and the cost of 

 travel time. 2 Ignoring time costs will cause biased 

 estimates of demand and economic value (Cesar- 

 io and Knetsch 1970). Cost of travel time was cal- 

 culated by multiplying estimated travel time en 

 route to the site by an hourly wage rate (Knetsch 

 et al. 1976). Sample size was included as an inde- 

 pendent variable in the participation equation 

 because others have found that visitation in- 

 creases at a nonlinear rate with increases in popu- 

 lation (Cesario and Knetsch 1976; Grubb and 

 Goodwin 1968). Travel costs to substitute sites, 

 TC, k , were represented in an index of travel costs 

 reflecting the availability of substitute angling 

 opportunities. 3 The attractiveness of available 

 recreation sites can also be an important determi- 

 nant of visitation patterns. The decision to visit a 

 particular site depends, in part, on the attractive- 

 ness of that site compared with other available 

 sites. Site attractiveness measures used by others 

 have included angling success rates (Stevens 

 1966), size of the recreation area (Ravenscraft 

 and Dwyer 1978), congestion at the site(McCon- 

 nell 1977), and accessibility (Cesario and Knetsch 

 1976). Data limitations reduced the possible 

 choices for attractiveness variables in this study 

 to fishing success rates and shoreline distance. 



Site Demand and Economic Value 



The second step of the travel cost method de- 

 rives the demand for and economic value of the 

 recreation site from the participation equation. 

 The usual procedure is to derive a demand curve 

 for a specific site by estimating demand from 

 each origin and aggregating over all origins for 



"'Travel costs were converted to price per angler day by 

 taking into account travel distance and whether lodging ex- 

 penditures were reported by anglers. Analysis of the survey 

 data indicated that anglers who resided at a (one-way) dis- 

 tance between 125 and 175 mi from the site generally incurred 

 lodging expenditures, indicating an overnight stay at the site. 

 Accordingly, price per angler day was assumed to equal one- 

 half the estimated travel costs for anglers residing more than 

 150 mi from a site. For anglers closer to the site, price per 

 angler day was assumed to equal estimated travel costs. Mone- 

 tary costs were assumed to be IOC/mile. Travel time costs 

 were calculated by multiplying estimated travel time at 50 

 mph by a value equal to 35% of the wage rate in the angler's 

 county of origin. Hotel costs were not included in the cost esti- 

 mates since they could not be determined on a per angler day 

 basis. 



3 Use of an index reflects the overall availability of substi- 

 tutes. Dividing the index by four would give the average price 

 of a substitute site in this fishery. A generalized approach to 

 the treatment of substitute sites is preferable to a specific sub- 

 stitute site in a regional travel cost model (Cesario and Knetsch 

 1976; Dwyer et al. 1977: Ravenscraft and Dwyer 1978). 



each increment of a hypothetical fee until aggre- 

 gate demand for the resource is reduced to zero 

 (Grubb and Goodwin 1968; Cesario and Knetsch 

 1976; Knetsch et al. 1976). This study estimated 

 NEV for each origin using a separate site-spe- 

 cific demand curve. Then the site's total NEV 

 was found by numerical aggregation across all 

 origins. This procedure estimates NEV more 

 accurately than the usual procedure because 

 there is less aggregation in deriving the site de- 

 mand curve (McConnell and Norton 1976; Menz 

 and Wilton 1982). 4 Demand was estimated from 

 the participation equation for each site with the 

 following: 



Dij = C tJ + pi (Tdj + p) + e 



(2) 



where D,j = the observed days of participa- 

 tion when the fee is zero 5 

 Tdj = travel costs from county of origin 

 i to site j 

 C>j = the composite of all other vari- 

 ables 

 p — the hypothetical fee charged for 



use of the site 

 e = an error term. 



The site's NEV to anglers in each origin was ob- 

 tained by integrating the demand equation be- 

 tween the limits of current travel costs and the 

 cost at which Dij would become zero. 



Results 



Some anglers may fish exclusively for small- 

 mouth bass, others for largemouth bass, and 

 some may be unconcerned about the specific type 

 of bass caught. Therefore, three separate analy- 

 ses were conducted: one each for the smallmouth 

 and largemouth bass fisheries and one for the 

 "combined" bass fishery. The value of the com- 

 bined fishery was determined in a separate anal- 

 ysis because addition of the smallmouth and 

 largemouth bass results would double-count 

 anglers who fish for both species. The same fish- 

 ing sites were used for each analysis. 



Characteristics of anglers and sites are pre- 

 sented in Tables 1 and 2. Smallmouth and large- 



"This method will be more accurate than if an aggregate de- 

 mand curve were used, but it will not provide as accurate an 

 estimate of economic value as aggregation of individual eco- 

 nomic values (Brown et al. 1965; Smith 1975a). 



5 The value of Dy.was set equal to zero whenever a negative 

 quantity resulted from the calculation. 



170 



