Académique Documents
Professionnel Documents
Culture Documents
=
b
i
$+
gdph
t
i
+
d + e
t i
f+airl
ib
t
i
+
g
+
h
0 i t i i t ii i i i
i
=1 i=1 i =1 i =1
i
=1 i =1 i
tr
+gd
h
t
+gd
pf
t
1
+
relative
t
1
airlib+
+roo
t
1
+
proo
t
+
1
t
1 2 3 4 5 t 16 7 t
/20
Since we have annual observations, we chose n1 Ior the maximum order oI lags in the ARDL
model in both cases and carry out the estimation over the period oI study. In Iact the same lag
length was chosen when using the Iinal prediction error due to SBC.
For the model, the hypothesis that is being tested is the null oI non-existence oI the long run
relationship` deIined by
Ho%
&
'
(
'
)
'
*
'
+
'
,
'
-
'.
"e Meridien #otel$ Mauritius$ %&'%(
August %)*) 9
International Research Symposium in Service Management ISSN 1694-0938
And the alternative hypothesis is
H
&
%
&
.!
(
.!
)
.!
*
.!
+
.!
,
.!
-
.
The recommended statistic is the F statistics Ior the joint signiIicance oI
&
!
(!
)!
*!
+
,
and
-
. Computation oI this F statistic requires running the Iollowing regression
tr = +b tr +$ gdh
+d
gdpf
+
e
relat
ive
1
+ f airlib + g roo
+h
proo
1
+
t 0
t
1
t
1 t1 t
t
1
t
1 t t
/&0
and a variable addition test is subsequently made by including the Iollowing.
1
tr
t
1
,
2
gdph
t
1
,
3
gdpf
t
1
,
4
relative
t
1
,
5
airlib
t
1
,
6
roo
t
1
,
7
proo
It should be noted that the distribution oI the F statistic is non-standard, irrespective whether
regressors are I(0) or I(1). Pesaran and al (1997) have tabulated the appropriate critical values
Ior diIIerent number oI regressors and whether the regressors contain an intercept or a time
trend.
The F-Statistics F(tr/gdph!gdpf! airlib!roo!relative! proo) turned out to be 6.21 and
exceeds the upper bound oI the critical value band. We thus reject the null hypothesis oI no
long run relationship between the variables irrespective oI their order. The test results thus
suggest that there is a long run relationship between the variable. (This is also conIirmed by
the Maximum Eigein values and Trace Values oI the Johansen test Ior cointegration)
1stimation results
Given that the speciIication is cointegrated, the unrestricted error correction representation oI
the ARDL model is given by equation 3. The next stage oI the procedure would be to estimate
the coeIIicients oI the long run relations and the associated Error Correction Model (ECM)
using the ARDL approach. The order oI the distributed lag on the dependent variable was
selected by the Schwartz Bayesian Criterion (SBC)
2
and turned out to be one.
"e Meridien #otel$ Mauritius$ %&'
%( August %)*)10
International Research Symposium in Service Management ISSN 1694-0938
The SBC criteria selects the ARDL (1,1,1,1,0,0,0) Ior the model. The long run estimated
coeIIicients are shown in the table 1.
Ta+le *3 1stimated "ong run 4oefficients +ased on ARD" approach3 Dependent
varia+le is tr
Regressor 4oefficient t'ratio
/S54 *$*$*$*$)$)$)0
gdph ..*,/ &.0*
gdpf ).-,/// (.1(
relative 2..0+/ 2&.00
airlib ..&0// (.(*
roo ..)0/// (.1*
proo ..&,/ &.-0
$o"#ta"t +.)*/// *.0&
It is observed that tourism promotional eIIort may have indeed contributed positively to the
tourist arrival in the country in the long run thus conIirming its importance as a potential
ingredient in tourism destination. In Iact a 1 increase in the promo is associated with a 1.6
increase in the number oI tourist arrival in the island. The level oI development oI the island
and tourism inIrastructure is also seen to be important Iactors. The income in the country oI
origin, a measure oI income elasticity seems to indicate that Mauritius is a luxury destination.
As such tourist are reported to be price sensitive as revealed by the negative and signiIicant
coeIIicient oI the variable relative. Air Access Liberalisation is also seen to be an important
element in the development oI tourism in the country and this is in line with the theoretical
underpinnings and also with the studies oI Seetanah (2007).
We proceed to the estimation oI the autoregressive distributed lag estimates and the error
correction model associated with our long run estimates using SBC. The results are reported
in table 2 below.
Ta+le %3 Autoregressive Distri+uted lag estimates3 Dependent varia+le is tr
2
Pesaran et al (1997) Iound that SBC is preIerable to AIC, as it is a parsimonious model that selects
the smallest possible lag length, while AIC selects the maximum relevant lag length
"e Meridien #otel$ Mauritius$ %&'%( August %)*)
11
International Research Symposium in Service
Management
ISSN 1694-
0938
Regressor 4oefficient t'ratio
S54 /*$*$*$*$)$)$)0
tr(lagged ) 0.345*** -.&&
gdph ..))*/ &.00
gdph(lagged) ..&,* ..+*
gdpf (.()// (.&)
gdpf (lagged) ..,1// (.(&
relative 2..+)*/// 2).()
relative(lagged) 2..(() 2..1,
airlib ...1+/ &.-1
roo ..)(+// (.)+
proo ..&*)/ &.-*
3o"#ta"t ).--/// ).()
R sqr ..1&(
Dw (.(
The selected ARDL model passes the standard diagnostic tests (Serial Correlation, Functional
Form, Normality and Heteroscedasticity) and highlights once again the positive link between
tourism marketing and promotion eIIort and tourism arrival in a dynamic Iramework and
where there is the presence oI lagged eIIects. The positive coeIIicient oI the lagged dependent
suggests that repeat tourism is a common phenomenon Ior the Mauritian case. Moreover, the
positive constant term indicates that visitors have a good opinion oI the island as a destination
resort.
Ta+le 23 1rror 4orrection Representation for the Selected ARD" model3 Dependent
varia+le is tr
Regressor 4oefficient t'ratio
S54 /*$*$*$*$)$)$)0
gdph ..))*/ &.00
gdpf (.()// (.&)
relative 2..+)*/// 2).()
airlib ...1+// &.-1
"e Meridien #otel$ Mauritius$ %&'%( August %)*)
12
International Research Symposium in Service
Management
ISSN 1694-
0938
roo ..)(+// (.)+
proo ..&*)/ &.-*
Ecm(-1) ..)1/// *.&(
R sqr ..-&
DW (.(
Table 3 suggests that the impact oI tourism promotion on the tourism development has been
positive and signiIicant in the short run as well with a coeIIicient oI 0.143. This which is
slightly lower than the estimated long run coeIIicient oI 0.16 suggesting that it these
marketing and promotional eIIorts may take some time to have their Iull eIIect. The
coeIIicients Ior the other explanatory variables are well behaved and have the expected sign
and signiIicance. In particular, the long run coeIIicient Ior the Ioreign income is slightly more
than reported in the literature and conIirms that Mauritius is seen as a more luxury destination
in the market. Relative prices and exchange are Iound to negatively impact on the
attractiveness oI Mauritius as a destination. The negative and signiIicant adjusted CPI
indicates the tourists are relatively price sensitive. It is however estimated to be less than in
the literature which reported that the price elasticity oIten Iall with the range oI unitary
(Crouch, 1999) thus indicating that tourists might be less price sensitive. The positive and
signiIicant coeIIicient oI domestic income, used as an indicator Ior development, suggests
that a higher level oI development is consistent with more tourist arrival.
Lastly, the coeIIicient oI the error correction model (ECM) oI the selected ARDL is negative
and highly signiIicant at 1 level. This conIirms the existence oI a stable long-run
relationship and points to a long-run co-integration relationship between variables. The ECM
represents the speed oI adjustment to restore equilibrium in the dynamic model Iollowing a
disturbance. The coeIIicient oI the ECM is around -0.4 in implies that a deviation Irom the
long-run equilibrium Iollowing a short-run shock is corrected by about 40 per cent aIter each
year.
Summary 6 Policy implications
The link between tourism promotion eIIort and tourism arrival has been analysed using an
Autoregressive Distributed Lag Iramework Ior the small island state oI Mauritius Iollowing
rigorous time series properties tests. Results Irom the analysis show that tourism promotion
eIIorts may have contributed positively, though relatively not to a sizeable level to the number
oI tourist arrival in both short and long run. The income and price elasticities Irom the study
Iurther revealed that the Mauritius is considered to be a luxury and that tourists are
"e Meridien #otel$ Mauritius$ %&'
%( August %)*)13
International Research Symposium in Service Management ISSN 1694-0938
also price sensitive. Tourism inIrastructure and the country development level are also part oI
the overall ingredient Ior Mauritius`s successIul tourism development. Moreover the
phenomenon oI repeat tourist is well present in the country.
As Iar as policy recommendations are concerned, it is believed that the government should
have a consistent policy with respect to tourism promotion allocations and grants especially in
times oI budget constraint. The case oI private Iinancing and joint public/private Iinancing
arrangements should thus be less ambiguous so long there is addition to the Iunding oI such
eIIorts and government should ensure that the private sector have suIIicient incentive to
contribute. There should be an awareness between all stakeholders that such promotion is Ior
a common cause and mutually beneIicial Ior all. A spillover beneIit oI public private
partnerships in tourism is that private sector has more sense oI entrepreneurship and is more
responsive than the government these industry members and can thus provide valuable
expertise Ior a market driven promotional and marketing activities.
References
Buhalis, D (2000) Marketing the competitive destination oI the Iuture, Tourism Management 21 (116)
Clarke C P (1981) An analysis oI the determinants oI demand Ior tourism in Barbados`, Unicersity
MicroIlims International, Ann Arbor, Michigan; London UK.
Durbarry, R, (2002), The economic contribution oI tourism in Mauritius, Annals oI Tourism Research,
Vol 29, No 3, pp 862865.
Durbarry, R (2004), Tourism and economic growth: the case oI Mauritius, Tourism Economics, 10(4),
pp 389-401
Eilat Y and L Einav, (2004) Determinants oI international tourism: a three- dimensional panel data
analysis, Applied Economics 36, 13151327
Ferrario, F (1979), The evaluation oI tourist resources, 4o5r"al of Travel Re#ear$h 17(3):18-22
Gunn, C.A. (1988) Tourism Planning (second edition), New York: Taylor & Francis.
"e Meridien #otel$ Mauritius$ %&'
%( August %)*)14
International Research Symposium in Service Management ISSN 1694-0938
Gearing, C.E (1974), Establishing a Measure oI Touristic Attractiveness, Journal oI Travel Research,
12, 1-8.
Heneghan P (1976) Resource allocation in tourism marketing, Tourism International
Press, London, UK 1976
Kulendran, N. (1996) Modeling quarterly tourist .ows to Australia using cointegration analysis,
Tourism Economics, 2, 20322.
Lim C (1997), Review oI international tourism demand models, Annals oI Tourism research, Vol. 24,
No 4, pp 835-849
Martin, C. A., & Witt, S. F. (1988). Substitute prices in models oI tourism demand. Annals oI Tourism
Research, 15, 255-268.
McElroy J, L (2003) Small Island Tourist Economies Across the LiIecycle, Paper prepared Ior the
International ConIerence, Beyond MIRAB: The Political Economy oI Small Islands in the
21st Century, School oI Economics and Finance, Victoria University, Wellington, New
Zealand, 23-25 February 2004
Naudee W A, Saayman A (2004) The determinants oI tourist arrivals in AIrica : A Panel Data
Regression Analysis, Paper presented at the International ConIerence, Centre Ior the Study oI
AIrican Economies, University oI OxIord
Nordstrom J (2002) Dynamic and Stochastic Structures in Tourism Demand Modeling, Umea
Economic Studies No 596, Umea University, Sweden
Papadopoulos, S I and S F Witt (1985) A Marketing analysis oI Ioreign tourism in Greece, in : S Shaw,
L Sparks and E Kaynak, eds, Marketing in the 1990s and beyond: Proceedings oI the Second
World Marketing Congress (University oI Stirling, Stirling) 682-693
Pesaran, H. M. and Y. Shin. (1995) Autoregressive distributed lag modeling approach to
cointegration analysis, DAE Working Paper Series! No. 9514, Department oI Economics,
University oI Cambridge.
Pesaran, M. H., Y. Shin and R. Smith. (1996), Testing Ior the Existence oI a Long -run Relationship,
DAE Working Papers 9622, Department oI Applied Economics, university oI Cambridge.
Pesaran, M.H and Y. Shin (1999), An Autoregressive Distributed Lag Modelling Approach to
Cointegration Analysis, Chapter 11 in S.Storm(ed.), Econometrics and Economic Theory in
the 20
th
Century: The Ragnar Firsch Centennial Symposium, Cambrigde University Press,
Cambridge.
Phillips, P.C.B and Perron, P (1988), Testing Ior a Unit Root in Time Series Regression, Biometrica,
75, pp 335-46
Ritchie, B. J. R. and Zins M. (1978) Culture as a determinant oI the attractiveness oI a
tourist region, Annals oI Tourism Research, 5(2): 252-267.
Sinclair, T. (1998), Tourism and economic development: a survey, Journal oI Development Studies, Vol
34, No 5, pp 151.
Smith, S. L. J. (1994). The tourism product. Annals oI Tourism Research, 21(3), 582-595.
Tang and Rochananond (1990), Attractiveness as a tourist destination; a comparative study oI Thailand
and selected countries, Socio-Econ Planning, Vol 24, No 3
Tourism Task Force (TTF) (2003), Down the track; better ways to deliver tourism`s land transport
inIrastructure. Available on www.ttI.org.au
"e Meridien #otel$ Mauritius$ %&'
%( August %)*)15
International Research Symposium in Service Management ISSN 1694-0938
Uysal M and Crompton J L (1984) Determinants oI demand Ior international tourist Ilows to Turkey,
Tourism Management, 5(4), 288-297
Witt S and Witt C (1995) Forecasting tourism demand: A review oI empirical research, International
Journal oI Forecasting, Vol. 11 pp 447-475
Appendi7 A
Ta+le A*3 Some key figures a+out the Mauritian Tourism Sector
"e Meridien #otel$ Mauritius$ %&'%( August %)*)
16
International Research Symposium in Service
Management
ISSN 1694-
0938
*89) *88) *889 *888 %))) %))* %))% %))2 %)):
Population
/;)))0 1060 1080 1159.7 1174.4 1186.1 1189 1190.3 1195.4 1197.5
No! #otels 43 75 90 92 95 95 95 97 110
#otel 2101
Rooms 4603 6809 7267 8255 8657 9024 9647 10233
Tourist
arrival
/;)))0 115 291.5 558.1 578 656.5 660.3 681.6 702 810
Tourism
Receipt
/million0 7500 9207 11890 14668 14234 18166 18238 19397 21865
Tourism
Receipts
/< of
=DP0 6 10 13 13 14 15 16 17 20
6o5r$e% 36O 7(..08
"e Meridien #otel$ Mauritius$ %&'%(
August %)*) 17