variables), a reasonable facsimile of a price-quantity
demand curve if differences in on-site costs reflect
differences solely in unit prices of ancillary inputs,
but not if they reflect differences in demands for
ancillary inputs at given prices. In the latter case, the
number of days taken could plausibly increase with
an increase in daily on-site expenses, despite the
apparent predominance of empirical evidence to the
contrary.
A diagrammatic interpretation of the distinction
is as follows:
On-Site
Costs
In Figure 1 the curves labeled Dj and D2 depict
hypothetical demand curves. They are demand curves
by virtue of their showing the relationship, other
things being equal, between daily on-site costs and
total usage (quantity demanded) of a given facility.
An initial equilibrium, point a, is defined, with
Ds3 visitor days being consumed at daily on-site costs
of Es . Next an increase in daily ancillary on-site
costs, from ESj to Es^, is posited. The type of cost
increase valid for treatment as a price proxy is
completely independent of any shift in on-site-cost
demand for the given site. For that type of price
change the predicted decline in quantity demanded
would be from Dc to Dc as read from demand
curve Di. If, however, some part of the same change
in on-site costs were due to an improvement in site
quality, for example, the quality improvement would
also induce an upward shift of demand from Dj to,
say, D2. Thus, instead of a movement to point b on
Di, the equilibrium would move to a point such as c
on D2 and on the curve labeled F(7).F(?) in this
example would clearly underestimate the on-site-cost
elasticity of demand.
By the same reasoning it can be shown that if a
change in site quality induces a decline in daily
on-site expenditures, the resulting F(?) would
overestimate the true on-site elasticity of demand.
The question, of course, is: Are ancillary on-site costs
more or less in site-proxy than a site-quality demand
shifter? The answer to that question is crucial to
explaining or predicting recreationists’ reactions to
changes in daily on-site costs.
There are no doubt many types of study areas in
which it would suffice merely to mention the absence
of compelling reasons for suspecting that on-site costs
reflect demand shifts instead of price differences. At
the same time, some empirical evidence on the
presence or absence of a relationship between on-site
costs and other possible demand shifters would
enhance the usefulness of on-site costs in their role as
a price proxy.
Travel Costs
Travel costs constitute a tempting price proxy,
both because of their prominence in the typical
recreationist’s budget and because data on them are
so easily obtainable. It is not altogether certain,
however, that travel costs are, in all cases, a better
index of site price than of quantities purchased of
ancillary travel inputs.
Moreover, only if the sole purpose of a trip is to
recreate on a given site can costs of travel be
considered a valid proxy price for recreational
opportunities of that site. The appropriateness of the
proxy price varies inversely with the strength of other
reasons for the trip. It is not necessary to require the
visitor to know precisely where he is going the
moment he leaves his home. It is enough that he gets
no utility from his trip apart from the on-site
pleasures of that particular site. To assume so much
should be done carefully.
There are suggestions as to how total travel costs
might be adjusted to remove the influence of other
benefits. One is to exclude from consideration the
recreationist whose visit to the site is not the sole
reward for his travels. A more typical approximation
is to exclude from the sample of recreationists those
whose visit is not the major reason for the trip. That
might be rational, as approximations go, for visitors
to a facility with such unique and Unduplicatable
facilities as those of a Grand Canyon or a
Yellowstone, where for reasons of remoteness, as well
as uniqueness, the typical visitor may well be
enjoying the high point of his trip.5
Applying the same rule of sample selection to
any campsite may, however, exclude the typical
5The subject matter of Clawson and many others does belong to this resource-based type of facility.
167