Modelling ski choice
1017
Table 6. The e ect of accommodation
From\To
Cairngorm
- 0.473
Nevis Range
Glencoe
Glenshee
The Lecht
Cairngorm
Nevis Range
Glencoe
Glenshee
The Lecht
0.204
0.644
0.04
0.065
0.007
0.204
0.018
0.184
0.065
0.007
0.204
0.214
0.039
0.393
0.033
0.204
0.214
0.039
0.308
0.089
7
7
7
7
7
7
7
7
7
7
0.214
0.04
0.065
0.007
7
7
7
7
7
7
Note: Own elasticity shown in bold.
VI. CONCLUSION
become increasingly important. This work shows that
choice models are possible and useful for a product as
unique as Skiing.
When undertaking new projects most analysts tend to use
the stated preferences of consumers. As is well known
stated preferences always overstate the e ect of a change,
for example many individuals would seriously believe (and
state) that they would go skiing if prices were lower and
slopes closer to their homes. In reality, however, existing
social and recreational patterns would act as a strong but
unrecognized deterrent and there would be substantially
less change than predicted. The alternative therefore is to
examine the revealed preferences and try to model their
choices in the past. As this paper shows this is possible
but is by no means an easy task.
The results presented here almost certainly underesti-
mate change, largely because of sample bias discussed in
Section III. In the models we assume that the parameter
linking utility and journey time is constant for all individ-
uals with individual variations from the mean being part of
the stochastic term. Since those who place least cost on
journey time will be sampled at the sites furthest away,
the stochastic term will be correlated with the distance
which will, in turn, bias the coe cient downwards. A simi-
lar e ect is apparent with price elasticity where the least
price conscious will be sampled at the location with the
highest prices.
R EFER ENCES
Baltas, G., Doyle, P. and Dyson, P. (1997) A model of consumer
choice for national versus private label brands, Journal of
±
Operational Research Society, 48, 988 95.
Ben-Akiva, M. and Lerman, S. R. (1985) Discrete Choice
Analysis, MIT Press, Cambridge, MA.
Bucklin, R. E., Gupta, S. and Han, S. (1995) A brands eye view of
response segmentation in consumer choice behaviour,
±
Journal of Marketing Research, 32, 66 74.
Cogan, J. F. (1980) Married women’s labor supply: a comparison
of alternative estimation techniques, in: Female L abor
Supply: Theory and Estimation, (Ed.) J. Smith, Princeton
University Press, Princeton, NJ.
Eymann, A. and Ronning, G. (1997) Microeconometric models of
tourist destination choice, Regional Science and Urban
±
Economics, 27, 735 61.
Greene, W. (1998) L imdep V 7.0/Nlogit V 2.0, Econometric
Software, Plainview NY.
Haab, T. C. and Hicks, R. L. (1997) Accounting for choice set
endogeneity in random choice models of recreation demand,
Journal of Environmental Economics and Management, 34,
±
127 47.
Hausmann, J. and McFadden, D. L. (1984) A speci®cation test
±
for the multinomial logit model, Econometrica, 52, 1219 40.
Kamakura, W. A. and Kim, B. D. (1996) Modelling preference
How useful, then, are these models and results? Choice
modelling is undoubtedly complex and often makes suspect
assumptions. It tries to extract cardinal values from ordinal
data, and place values on the subjective assessments of
individuals. Econometrically it is probably subject to sig-
ni®cant downward bias. This paper shows, however that,
despite these caveats, valuable information can be
obtained. First, in general terms, the models act as an im-
portant counterbalance to stated preference. Secondly in
this case they do reveal quite vividly the importance of
cost to the day-tripper and the signi®cant competition
that does exist between centres (suggesting the sense in
pursuing strategies that tie individuals to a centre).
Finally they do provide site owners and local development
agencies with guides to the most e ective development stra-
tegies.
and structural heterogeneity in consumer choice, Marketing
±
Science, 15, 152 72.
Kling, C. L. and Thomson, C. J. (1996) The implications of model
speci®cation for welfare estimation in nested logit models,
±
American Journal of Agricultural Economics, 3, 103 17.
Luce, R. (1959) Individual Choice Behaviour: A Theoretical
Analysis, John Wiley and Sons, New York.
Mackay (1986) Expenditure of Skiers at Cairngorm and Glencoe,
Final Report for the Highlands and Islands Development
Board, Mackay Consultants, Inverness.
Mackay (1995) Scottish Tourism Commentary, Mackay
Consultants, Inverness.
McFadden, D. L. (1984) Econometric analysis of qualitative
response models, in: Handbook of Econometrics (Eds)
Z. Griliches and M. Intrigilator, Elsevier, Amsterdam, pp.
±
1395 1457.
Milne, N. (1997) The appropriate speci®cation of the multiplicand
and the incorporation of displacement e ects within eco-
nomic impact assessments; a study of the net additionality
of individual Scottish ski centres, MPhil Thesis, Glasgow
Caledonian University, November.
As computing power expands and data-bases increase in
number, we are convinced that this type of modelling will