22
The surplus associated with access at the current conditions are reported in table
7. These figures refer to the average visitor at each site, i.e., they average in-sample
predictions.16 Briefly, they imply that the value of accessing the site in its current
conditions is over 21,000 AMD at Garni, 19,000 AMD at each of Haghardzin and Khor
Virap, and 13,850 AMD at Tatev.
Table 7. Surplus at current conditions and price by site based on the estimates in table 6.
All figures in thousand AMD.
Site |
Mean |
Standard |
Minimum |
Maximum |
All sites_________ |
18.44 |
5.32 |
8.21 |
44.00 |
Garni__________ |
21.68 |
4.78 |
9.59 |
29.13 |
Haghardzin |
19.07 |
6.00 |
8.69 |
44.00 |
Khor Virap |
19.24 |
4.49 |
8.21 |
29.38 |
Tatev__________ |
13.85 |
2.87 |
8.58 |
24.19 |
B. Combining Actual and Hypothetical Data
We combine actual and contingent behavior trips to estimate the value of
programs that restore the sites and improve their quality. Each respondent contributes
three observations to our sample, and we pool the observations from the four sites to
estimate a Poisson model with the correction for on-site sampling where
λij = exp(xijβ1 + pijβ2 + qjβ3) .17 Results based on the full panel of data are displayed in
table 8.
16 As detailed in Englin and Shonkwiler (1995), the value of access for a visitor, or his consumer surplus at
the current conditions, is thus the number of visits predicted by the model for this visitor (λi+1, where
λi=exp(xiβ1+piβ2)), divided by the negative of the coefficient on price (-β2). This formula applies to our
sample of visitors, who are likely to visit more frequently than the population of visitors at large.
17 We constructed the dependent variable for the hypothetical visits (j=2, 3) as follows. For j>1, we
assigned the number of trips respondents said that they would take. If they said that they would visit the
same number of times as during the previous year, then yj=y1. Once again, correcting for the on-site
intercept nature of the sample implies that we estimate a Poisson equation where the dependent variable is
the number of visits minus 1.