This
** preview**
has intentionally

**sections.**

*blurred***to view the full version.**

*Sign up*
**Unformatted text preview: **Valuing the environment u on visitors and nonwvisitors by means of a,
._ ly postal, survey of a sample of the popula-
'-considered to be relevant as potential visitors.
respondents could be asked about their travel
it but this is rare and typically travel costs are
'_ u -- to respondents by the analyst on the basis
distance travelled, which itself is usually estim-
by the analyst by assigning respondents, on the
of information supplied by them such as a
, . . or zip code, to a number of zones and meas-
in" g distance from the centre of each zone using a
| p. The regression is then. usually. of the number
visits per unit population from zone r' on travel
from zone t'. For some sites and some surveys it
possible to have 1' index individuals, when the
. ndent variable is the number of visits in a period
-. time by individual i and the explanatory variable
it'suavel costs per visit for individual i. Where the
ﬁtlata are such that either approach could be followed,
there is some dispute as to whether it is better to use
the individual data or to average costs over indi-
E'yiduals in given zones and regress total zonal visits
per unit population on average zonal travel costs.
Most TCM applications employ the zonal average
approach, often simply because of data limitations.
The MCS ﬁgure that is produced by the TCM
'I-as described here is the total MCS for the sample
;_of visitors included in the survey. Unless the survey
has been such that the sample is the population, Box 12.1 An Illustrative zonal average TOM example The basic data for a national park with no
admission charge are: _______._____———————
Visits Population Distance
{thousands} [miles] ‘15 000 2 000 10
48 000 8 000 15
11 25!} 2 500 20
45 000 ‘15 000 25
34 [100 22 560 30 where distance is measured from the centre of
the zone, and we are assuming, in the interest of
keeping the story simple, that we know the total
number-of visits in the year from each zone. We
will also assume that we know the travel cost per- there remains the question of how to go from this
ﬁgure to the MCS ﬁgure to be used in ECBA. This
can be quite complicated and the answer depends
on the nature and timing of the survey in relation to
the characteristics of the site concerned. One fairly
standard procedure is to divide the MCS ﬁgure
produced as described here by the total number of
visits covered by the survey to get a ﬁgure for MCS
per visit, which is then multiplied by the (usually
estimated) number of visits per year to get a ﬁgure
for MCS per year for use in the ECBA. In Willis and
Garrod (1991a) MCS per visit is estimated using
both individual data on visits per year and zonal data
on visits per unit population as dependent variable,
with explanatory variables deﬁned appropriately in
each case. Across six forest sites in the UK MCS
per visit ranged from £1.43 to £2.60 (average £2.03)
using the zonal averaging approach. and from £0.06
to £0.96 (average £0.48) using individual data. The
two methods did not even rank the sites in the same
order. Clearly, given annual total visitor numbers
of the order of one million, one could draw quite
different conclusions as to the use value of one of
these sites according to which of these approaches
was used. Boa l2.l works through an illustrative, zonal aver—
age, TCM application where the numbers have been
constructed so as to make the calculations simple
and bring out the basic ideas as clearly as possible. mile to be £1. The ﬁrst step is to estimate the
parameters 'of the trip generating function
vi=ot+ﬂtTt+PJ +2, where ii, is visits per thousand population from
the ith zone, T, is travel cost from the it]: zone,
P is the admission price which. is zero. and e, is
the error term. We get ordinary least squares
estimates for a and b using; . Zapata-r1
B: Eta—rt 414 Project appraisal __g'dtscuss fou ' . .. _ . Edie discuss ”13 _ - _ ' _ - :1 do occurs
2B
' 25
30 “10.0
20. we get the. estimated trip- generating equation 'as [[153 250 — 23 000] .x 5 55:50.5] + [78 0003.5]
91‘" Jilly—5 D3[T+P} plus The seeond step is to use this estimate to [[78 000 - 35 750] x_-5._><_0.51+ [36' 750 X 5']
derive the rolationinip between visits and the
price of admission. which is often referred to in _ - -
the literature as the surrogate demand function. [(36 750 +- _-18 goo] X5 x- 0.5] + [1.8 000 x _5.l.-_ ' 'Wa mil eonaider =P varying in. steps.- of £5. For
P: £5 predicted visits from each Zone V, and _ _ _ _ _
total predicted visits are. calculated. using the [[18 000 — 30001 x 5 x115] +' [3000 x 5] -
estimated trip generating- function as follows: ' ' " " ' ' ' - plus plus plus
[3000 x5 x 0.5]
which is £1 081 875.. q=hsss=1ds "730‘ 191'
' " ”s‘. 2. Laos 12000
4.5 - a m. at, 36 000
'3 _ 7500
1.5 22 500
0 o
' 73 0.00 Proceeding in- the same way for P- =_ £10 and so _ . - . __ -' forest sil
on, we get the following simulated prioehrisits ' - ' of the TI
data forithe'l'smsHtB'-damsnd bastion:- . . _ not prov
_ _ . .' ' , . . _ . _ I reportin, ..V" _ 3 ' ' ' ' ' ' _ II ' t;- -- -- form. : 12.4.
.10 - - .' ' -. ‘ . .- " - - - - 15 ' . " ., .. - .' -' - . -. -. .. '_ . .- '. In many - 20. ' . .' ' _ ; ' . . . _ . _ -. _ _. ; : - -- . . . . .. . -- genera“ .—'———-——-—-—-—'25 ' ' I ' . ; ' - 3 '_ . f :1 ' ' ' ' . .-'.'.:” method . Figure 1'2. _8 shows the surrogate. demand
function. The third step is- to get ﬁom- this the
. estimate of eonsuiners' -surplus for the year. Given
thatiii feetP; g.tot_a_1e0nsumets somlusisthe ' _- '-_ .-' - _ __ _ _J-
demand function which' is Figure 12;8._.f.\n Hlus'trative "surroga'te‘déh'lahti function J - L'J ...

View
Full Document